University of Southern California Dissertations and Theses

The intelligent control strategy of kinetic façades for daylight and energy performance: evaluating the daylight effect of adaptive systems based on parametric workflow

(USC Thesis Other)

The intelligent control strategy of kinetic façades for daylight and energy performance: evaluating the daylight effect of adaptive systems based on parametric workflow

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE INTELLIGENT CONTROL STRATEGY OF KINETIC FAÇADES FOR DAYLIGHT AND ENERGY PERFORMANCE Evaluating the Daylight Effect of Adaptive Systems Based on Parametric Workflow by Chenxi Yang A Thesis Presented to the FACULTY OF THE USC SCHOOL OF ARCHITECTURE UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF BUILDING SCIENCE MAY 2020 Copyright 2020 Chenxi Yang ii ACKNOWLEDGEMENTS I would like to express my sincere gratitude to my thesis committee chair, Professor Kyle Konis, for his continuous supports and suggestions of this research project. His extensive knowledge and his thinking logic have been greatly valuable since I started the thesis at the early stage. His professional guidance and unique ideas always instruct me to develop the thesis project and make progress in the research until the thesis completion. Kyle, thanks again for your encouragement and help to me. I would like to express my warm appreciation to my thesis committee Professor Douglas Noble and Professor Marc Schiler. I am very grateful to Doug for his valuable instruction and guidance. His mind and advice enlighten me to critically think about an argument, and his spirit and attitude for work inspire me to try my best to do the research. Also, I am very thankful to obtain help from Marc. His profound insight into building science helps me resolve many confusions in the specialized field. I am proud of having the opportunity to work with an erudite scholar like him. I would also like to give my special thanks to Professor Karen Kensek. Her broad knowledge and profound ideas enlighten me to start the thesis project from the beginning. She carefully understands my need for the research and offers valuable advice, introducing new ideas and technology to me. I would like to express my special gratitude to Professor Michael Fox and Professor Joon-Ho Choi. Michael’s books on interactive architecture introduce me to the thesis topic and provide me with many interesting projects and inspiring ideas that are helpful for me to develop the thesis. Meanwhile, he gave me many valuable recommendations on the topic. Joon-Ho also instructed me on many helpful ideas and view about environmental performance and control algorithms. Warm gratitude goes to my family, including my parents, my wife, and my kid for their consistent understanding and support to me. My words fail to express my appreciation to them since I cannot complete the thesis without their spiritual support. Finally, I am very grateful to all the faculty members and classmates in the MBS family for their time and expertise. Studying and living with you in the group for two years are really precious experience and memorable occasion to me in my future life. iii TABLE OF CONTENTS ACKNOWLEDGEMENTS .......................................................................................................................................... ii LIST OF TABLES ........................................................................................................................................................vi LIST OF FIGURES ................................................................................................................................................... viii ABSTRACT ............................................................................................................................................................... xii CHAPTER 1 INTRODUCTION ................................................................................................................................ 1 1.1 Responsive Kinetic Facades for Environment and Energy ........................................................................... 1 1.2 Intelligent Kinetic Facades ........................................................................................................................... 6 1.2.1 Intelligent Kinetic Building System ..................................................................................................... 6 1.2.2 Kinetic Facades and Daylighting.......................................................................................................... 7 1.2.3 Typology of Kinetics in Architecture ................................................................................................... 8 1.2.3.1 Kinetic Patterns Prototype of Architectural Facades ................................................................ 8 1.2.3.2 Kinetic Mechanism: Patterns and Motions ............................................................................. 11 1.3 The Composition of the Kinetic Mechanism .............................................................................................. 14 1.4 Dynamic Daylighting .................................................................................................................................. 16 1.4.1 Daylight Function ............................................................................................................................... 17 1.4.2 Daylight Fundamentals ...................................................................................................................... 18 1.4.3 Daylight Performance Indicators ........................................................................................................ 19 1.4.4 Daylight Simulation and Evaluation .................................................................................................. 21 1.5 Computational Design Tool ........................................................................................................................ 22 1.6 Hypothesis and Objectives.......................................................................................................................... 23 1.6.1 Hypothesis .......................................................................................................................................... 23 1.6.2 Objectives ........................................................................................................................................... 24 1.7 Summary ..................................................................................................................................................... 24 CHAPTER 2 BACKGROUND AND LITERATURE REVIEW ............................................................................. 26 2.1 Kinetic Concept .......................................................................................................................................... 26 2.2 Daylight Environment Factors on Kinetics ................................................................................................. 27 2.3 Daylighting Experimentation Methods on Kinetic Facades ....................................................................... 30 2.3.1 Experimental Method Category ......................................................................................................... 30 2.3.2 Performance Simulation ..................................................................................................................... 33 2.3.3 Control Algorithm .............................................................................................................................. 34 2.3.3.1 Rule-based Control ................................................................................................................. 34 2.3.3.2 Model Predictive Control ....................................................................................................... 35 2.4 Daylighting Analysis Approach for Kinetic Facades .................................................................................. 36 2.4.1 Introduction ........................................................................................................................................ 36 iv 2.4.2 Evaluation of interior daylighting performance ................................................................................. 37 2.4.3 Simulation of energy performance ..................................................................................................... 38 2.5 Case Study: Parametric-based designs for kinetic facades to optimize daylight performance ................... 39 2.5.1 The Method of Research Study .......................................................................................................... 39 2.5.2 The Result of Research Study ............................................................................................................ 42 2.6 Case Study: Performance-based Parametric Design Exploration ............................................................... 45 2.6.1 Introduction ........................................................................................................................................ 45 2.6.2 The workflow of the Design Case ...................................................................................................... 45 2.7 Parametric Computational Design Tools .................................................................................................... 47 2.7.1 Rhino and Grasshopper ...................................................................................................................... 48 2.7.2 Ladybug & Honeybee ........................................................................................................................ 48 2.7.3 DIV A .................................................................................................................................................. 48 2.8 Daylight Performance Metrics of Indoor Space ......................................................................................... 49 2.9 Summary ..................................................................................................................................................... 50 CHAPTER 3 METHODOLOGY ............................................................................................................................. 51 3.1 Workflow overview .................................................................................................................................... 51 3.1.1 Methodology background................................................................................................................... 51 3.1.2 Methodology development ................................................................................................................. 51 3.2 Overall workflow ........................................................................................................................................ 53 3.3 Working Model ........................................................................................................................................... 53 3.3.1 Model Definition ................................................................................................................................ 54 3.3.2 Daylighting Variables ......................................................................................................................... 55 3.3.3 Digital Models Instruction.................................................................................................................. 57 3.4 Parametric Simulation Test ......................................................................................................................... 62 3.4.1 Kinetic Shading States ....................................................................................................................... 63 3.4.2 Instruction of Façade Component ...................................................................................................... 63 3.4.3 Solar Heat Gain and Daylighting Metrics .......................................................................................... 67 3.4.3.1 Solar Heat Gain ...................................................................................................................... 67 3.4.3.2 Spatial Daylight Autonomy (sDA) ......................................................................................... 68 3.4.3.3 Annual Sunlight Exposure (ASE) ........................................................................................... 68 3.4.3.4 Electric Lighting Energy ........................................................................................................ 69 3.4.3.5 Glare ....................................................................................................................................... 69 3.4.3.6 View ....................................................................................................................................... 71 3.4.4 Parametric workflow description ....................................................................................................... 71 3.4.5 Experimentation Process .................................................................................................................... 79 3.4.6 Iterative Test Procedure ...................................................................................................................... 82 3.4.7 Software Environment Support .......................................................................................................... 83 3.5 Data Analysis Approach ............................................................................................................................. 84 3.6 Daylighting simulation algorithmic definition of five metrics.................................................................... 86 v 3.7 Summary ..................................................................................................................................................... 87 CHAPTER 4 CASE STUDIES RESULTS ............................................................................................................... 89 4.1 Overview .................................................................................................................................................... 89 4.2 Introduction of the simulation process ........................................................................................................ 89 4.2.1 Background introduction of parametric simulation ............................................................................ 90 4.2.2 Specification of daylighting and solar heat simulation....................................................................... 91 4.3 Results and Data of the case studies ........................................................................................................... 93 4.3.1 Results and Data of shoebox building model ..................................................................................... 93 4.3.2 Results and data of cuboid building model ........................................................................................ 94 4.4 Control strategy decision ............................................................................................................................ 96 4.5 Description of scoring process .................................................................................................................... 99 4.6 Summary ................................................................................................................................................... 101 CHAPTER 5 ANALYSIS AND EV ALUATION ................................................................................................... 103 5.1 Resulting analysis of the shoebox model with kinetic louver ................................................................... 104 5.1.1 Resulting analysis of indoor daylighting illuminance ...................................................................... 104 5.1.2 Resulting analysis of daylight glare probability ............................................................................... 110 5.1.3 Resulting analysis of solar heat gain ................................................................................................ 113 5.1.4 Daylighting effects comparison of the facades on the shoebox model ............................................. 116 5.2 Resulting analysis of the cuboid model with kinetic shading panels ........................................................ 117 5.2.1 Resulting analysis of daylighting illuminance.................................................................................. 117 5.2.2 Resulting analysis of daylight glare probability (DGP) ................................................................... 123 5.2.3 Resulting analysis of solar heat gain ................................................................................................ 126 5.2.4 Daylight effects comparison of the roofs of the cuboid model ......................................................... 129 5.3 Evaluation of the parametric workflow .................................................................................................... 130 5.4 Summary ................................................................................................................................................... 132 CHAPTER 6 DISCUSSION AND FUTURE WORK ............................................................................................ 134 6.1 Methodology discussion ........................................................................................................................... 136 6.2 Limitations ................................................................................................................................................ 139 6.3 Future work............................................................................................................................................... 140 6.4 Conclusion ................................................................................................................................................ 142 BIBLIOGRAPHY ..................................................................................................................................................... 144 APPENDICES ........................................................................................................................................................... 150 APPENDIX A: Simulation Resulting Data of Shading States at Each Hour. ............................................................ 150 APPENDIX B: Control Decision and Evaluation Process. ....................................................................................... 168 APPENDIX C: Façades Patterns Comparison Based on Daylighting and Thermal Metrics. .................................... 184 vi LIST OF TABLES Table 2. 1 Research topics on the kinetic façade using both quantitative and qualitative methods. ............................ 32 Table 2. 2 The outline of the daylight performance metrics. ....................................................................................... 50 Table 3. 1 The shoebox dimensions. ................................................................................................................................................ 59 Table 3. 2 The cuboid box dimensions. .......................................................................................................................................... 60 Table 3. 3 Material parameters of the building surface. ............................................................................................................. 61 Table 3. 4 Information on the site and the climate in Los Angeles (Temperature: monthly averages in 2018) ............ 62 Table 3. 5 Glare comfort criteria based on DGP. Source: Tabadkani et al., 2018. ............................................................... 71 Table 3. 6 Illuminance level of cuboid model based on time and tilt angle........................................................................... 80 Table 3. 7 The parametric simulation of daylighting metrics on the shoebox with different window sizes .................. 87 Table 3. 8 The resulting data of daylight metrics on different window sizes ........................................................................ 87 Table 4. 1 Information of the site and the climate in Los Angeles (Temperature: monthly averages in 2018) ............ 90 Table 4. 20 Recommended light level for working activities. ..................................................................................... 97 Table 4. 21 Control strategy criteria on June 21 st , Sept 21 st , Dec 21 st . ......................................................................... 99 Table 4. 37 Daylighting performance scores of three cases of the shoebox model on June 21 st , Sept. 21 st , Dec.21 st . .......................................................................................................................................................................... 101 Table 4. 38 Daylighting performance scores of three cases of the cuboid model on June 21 st , Sept. 21 st , Dec.21 st . . 101 Table 4. 2 Simulation results of hourly indoor average illuminance on June 21 st . ..................................................... 150 Table 4. 3 Simulation results of hourly Daylight Glare Probability on June 21 st . ...................................................... 151 Table 4. 4 Simulation results of hourly solar heat gain on June 21 st . ......................................................................... 152 Table 4. 5 Simulation results of hourly indoor average illuminance on September 21 st . ........................................... 153 Table 4. 6 Simulation results of hourly Daylight Glare Probability on September 21 st . ............................................ 154 Table 4. 7 Simulation results of hourly solar heat gain on September 21 st . ............................................................... 155 Table 4. 8 Simulation results of hourly indoor average illuminance on December 21 st ............................................. 156 Table 4. 9 Simulation results of hourly Daylight Glare Probability on December 21 st . ............................................. 157 Table 4. 10 Simulation results of hourly solar heat gain on December 21 st . .............................................................. 158 Table 4. 11 Simulation results of hourly indoor average illuminance on June 21 st . ................................................... 159 Table 4. 12 Simulation results of hourly Daylight Glare Probability on June 21 st . .................................................... 160 Table 4. 13 Simulation results of hourly solar heat gain on June 21 st . ....................................................................... 161 Table 4. 14 Simulation results of hourly indoor average illuminance on September 21 st . ......................................... 162 Table 4. 15 Simulation results of hourly Daylight Glare Probability on September 21 st . .......................................... 163 Table 4. 16 Simulation results of hourly solar heat gain on September 21 st . ............................................................. 164 Table 4. 17 Simulation results of hourly indoor average illuminance on December 21 st ........................................... 165 Table 4. 18 Simulation results of hourly Daylight Glare Probability on December 21 st . ........................................... 166 Table 4. 19 Simulation results of hourly solar heat gain on December 21 st . .............................................................. 167 Table 4. 22 Control strategy determination of kinetic louver of the shoebox model on June 21 st ............................. 169 Table 4. 23 Control strategy determination of kinetic louver of the shoebox model on September 21 st . .................. 170 Table 4. 24 Control strategy determination of kinetic louver of the shoebox model on December 21 st . ................... 171 vii Table 4. 25 Evaluation among three cases of the shoebox model on June 21 st . ......................................................... 172 Table 4. 26 Evaluation among three cases of the shoebox model on September 21 st . ............................................... 173 Table 4. 27 Evaluation among three cases of the shoebox model on December 21 st . ................................................ 174 Table 4. 28 Control strategy determination of kinetic shading panel of the cuboid model on June 21 st . ................... 175 Table 4. 29 Control strategy determination of kinetic shading panel of the cuboid model on Sept 21 st . ................... 176 Table 4. 30 Control strategy determination of kinetic shading panel of the cuboid model on Dec 21 st . .................... 177 Table 4. 31 Control strategy determination of tilt-only shading panel of the cuboid model on June 21 st . ................. 178 Table 4. 32 Control strategy determination of tilt-only shading panel of the cuboid model on September 21 st . ....... 179 Table 4. 33 Control strategy determination of tilt-only shading panel of the cuboid model on December 21 st . ........ 180 Table 4. 34 Evaluation among three cases of the cuboid model on June 21 st . ........................................................... 181 Table 4. 35 Evaluation among three cases of the cuboid model on September 21 st ................................................... 182 Table 4. 36 Evaluation among three cases of the cuboid model on December 21 st . .................................................. 183 Table 5. 1 Grid illuminance of the shoebox model with kinetic louver on three typical days. .................................. 110 Table 5. 2 DGP and interior renderings based on the occupant’s view and sun position scenarios. .......................... 112 Table 5. 12 Grid illuminance of the cuboid model with kinetic shading panels on three typical days. ..................... 122 Table 5. 13 DGP and interior renderings of the cuboid model based on the occupant’s view and sun position scenarios. ........................................................................................................................................................... 125 Table 5. 23 Experience comparison of four tools (workflows) for environment simulation. .................................... 132 Table 5. 3 Data results of hourly indoor average illuminance on June 21 st . .............................................................. 184 Table 5. 4 Data results of hourly indoor average illuminance on September 21 st . ..................................................... 185 Table 5. 5 Data results of hourly indoor average illuminance on December 21 st . ..................................................... 186 Table 5. 6 Data results of hourly daylight glare probability on June 21 st . .................................................................. 187 Table 5. 7 Data results of hourly daylight glare probability on September 21 st . ........................................................ 188 Table 5. 8 Data results of hourly daylight glare probability on December 21 st .......................................................... 189 Table 5. 9 Data results of hourly solar heat gain on June 21 st . ................................................................................... 190 Table 5. 10 Data results of hourly solar heat gain on September 21 st . ....................................................................... 191 Table 5. 11 Data results of hourly solar heat gain on December 21 st . ........................................................................ 192 Table 5. 14 Data results of hourly indoor average illuminance on June 21 st . ............................................................ 193 Table 5. 15 Data results of hourly indoor average illuminance on September 21 st . ................................................... 194 Table 5. 16 Data results of hourly indoor average illuminance on December 21 st . ................................................... 195 Table 5. 17 Data results of hourly daylight glare probability on June 21 st . ................................................................ 196 Table 5. 18 Data results of hourly daylight glare probability on September 21 st . ...................................................... 197 Table 5. 19 Data results of hourly daylight glare probability on December 21 st ........................................................ 198 Table 5. 20 Data results of hourly solar heat gain on June 21 st . ................................................................................. 199 Table 5. 21 Data results of hourly solar heat gain on September 21 st . ....................................................................... 200 Table 5. 22 Data results of hourly solar heat gain on December 21 st ......................................................................... 201 viii LIST OF FIGURES Figure 1. 1 The mean percentage of time spent by the NHAPS respondents in six different locations on the diary day.. ............................................................................................................................................................. 2 Figure 1. 2 Energy consumption distribution according to sectors. Source: Phase Change Composite Materials for Energy Efficient Building Envelopes. ............................................................................................................. 3 Figure 1. 3 Shading modules folding or unfolding in different positions of the kinetic envelope.. ............................... 5 Figure 1. 4 Council House 2 (CH2), kinetic timber shutters of western façade. ........................................................... 5 Figure 1. 5 The process of intelligent building – data flowing steps feature an intelligent façade.. .............................. 7 Figure 1. 6 Dynamic shutter module of Institut du Monde Arabe. Six shutter plates rotate to open and close the aperture in response to the daylight environment (Nouvel, 1987). ....................................................................... 9 Figure 1. 7 kinetic facades panels of Al Bahar Tower in Abu Dhabi designed by AEDAS. .......................................... 9 Figure 1. 8 Summary of the five experimental projects with kinetic motions. The diagrams show the kinetic configurations and mechanisms.. ........................................................................................................................ 10 Figure 1. 9 Mechanical movements of rigid architectural elements.. .......................................................................... 12 Figure 1. 10 The diagram demonstrates the kinetic patterns of motion and their features which are potentially used to design kinetic facades in response to environmental performance. ........................................................ 14 Figure 1. 11 The composition of the kinetic Mechanism. ............................................................................................ 15 Figure 1. 12 Veitch’s proposal that lighting quality is determined by the balance of the factors included in the diagram. .............................................................................................................................................................. 18 Figure 1. 13 Indoor daylighting characteristics that affect occupants’ productivity and comfort. ............................... 19 Figure 1. 14 Left: the demonstration of kinetic façade motion in Revit. Right: The interface of hierarchical nodes control in Dynamo Script. ................................................................................................................................... 23 Figure 2. 1 An algorithmic process of kinetic façade design with the multidisciplinary study.. .................................. 26 Figure 2. 2 Digital modeling and daylighting evaluation of kinetic façade in parametric workflow. .......................... 27 Figure 2. 3 Study of the relation between building form and occupant position to design interactive façade. ............ 29 Figure 2. 4 The kinetic façade interacting with sun position and occupant position to change the components configuration in hierarchical mode depending on the attractor and parametric facade logic. ............................. 30 Figure 2. 5 Rule-based control workflow. ................................................................................................................... 34 Figure 2. 6 Model based predictive control architecture. ............................................................................................ 36 Figure 2. 7 The procedure of calculating unshaded area. ............................................................................................ 39 Figure 2. 8 The transformations of the hexagonal grid in three basic geometric motions. .......................................... 39 Figure 2. 9 The transformations of the hexagonal grid: Sliding (1 & 2), Rotation (3), Scaling (4 & 5) in Rhino interface. .............................................................................................................................................................. 40 Figure 2. 10 Office Space in south orientation with dimensions of 6.0 m width, 7.5 m depth, and partially glazed height of 3.0 m on the south. ............................................................................................................................... 41 Figure 2. 11 Pattern modeling in Grasshopper shows the kinetic hexagonal patterns on the south surface. ............... 41 Figure 2. 12 Daylighting simulation of the base case in terms of illuminance levels analyzed in Diva for Rhino. ..... 42 ix Figure 2. 13 Daylighting simulation of the rotation motion best cases in terms of illuminance levels analyzed in Diva for Rhino. ............................................................................................................................................... 44 Figure 2. 14 Daylighting simulation of the translation motion best cases in terms of illuminance levels analyzed in Diva for Rhino. ............................................................................................................................................... 44 Figure 2. 15 The typical floor plan and section drawings of the office building in terms of spaces simulation based on daylight factor and solar irradiation. .................................................................................................... 45 Figure 2. 16 Performance-based parametric design workflow in Grasshopper interface. ........................................... 46 Figure 2. 17 Simulation workflow showing GA Solver cycle in Galapagos. .............................................................. 47 Figure 3. 1 Overall workflow of the research methodology. ....................................................................................... 53 Figure 3. 2 Workflow segment of working model. ...................................................................................................... 54 Figure 3. 3 The shoebox model test with three window sizes. .................................................................................... 57 Figure 3. 4 The shoebox and cuboid box with kinetic shades. .................................................................................... 57 Figure 3. 5 (a) Isometric rendering of the shoebox; (b) Isometric view of the shoebox with dimensions. .................. 59 Figure 3. 6 (a) Isometric rendering of the cuboid box; (b) Isometric view of the cuboid box with dimensions .......... 60 Figure 3. 7 Workflow segment of model testing. ......................................................................................................... 62 Figure 3. 8 Tilt angle of controlling the shading louver state on the shoebox model. ................................................. 64 Figure 3. 9 Workflow of setting each rotation angle to run a parametric simulation. .................................................. 64 Figure 3. 10 Tilt and rotation angle of controlling the shading panel state on the cuboid model. ............................... 66 Figure 3. 11 Workflow of reading weather data to acquire sun position (altitude and azimuth).................................. 66 Figure 3. 12 Workflow of setting each tile angle to run a parametric simulation ........................................................ 67 Figure 3. 13 Graphic representation of the parametric workflow and its components. ............................................... 72 Figure 3. 14 The components for translating Rhino geometry into HB zone as B-rep. ............................................... 73 Figure 3. 15 The shoebox model demonstration as a closed B-rep HB zone with glazing. ......................................... 73 Figure 3. 16 The components for introducing sunlight and weather conditions. ......................................................... 74 Figure 3. 17 The components for assigning HB surfaces with specific material properties. ....................................... 74 Figure 3. 18 The components for creating kinetic shading device and assigning material property. ........................... 75 Figure 3. 19 The components for daylighting illuminance simulation. ....................................................................... 76 Figure 3. 20 The components for glare index simulation. ........................................................................................... 76 Figure 3. 21 The components for solar heat gain simulation. ...................................................................................... 77 Figure 3. 22 The components for iterative parametric simulation. .............................................................................. 77 Figure 3. 23 The example of visualizing daylighting illuminance simulation. ............................................................ 78 Figure 3. 24 The example of visualizing daylight glare probability simulation. ......................................................... 78 Figure 3. 25 The example of visualizing solar heat gain simulation............................................................................ 79 Figure 3. 26 The database - look-up table diagram. ..................................................................................................... 80 Figure 3. 27 Illuminance level grid of cuboid model at a different time and panel tile angles. ................................... 81 Figure 3. 28 Illuminance curves of cuboid model at different time. ............................................................................ 82 Figure 3. 29 Model-based predictive control algorithm diagram. ............................................................................... 82 Figure 3. 30 Workflow segment of analysis approach. ................................................................................................ 84 Figure 3. 31 Parametric simulation workflow for sDA, ASE, glare, solar heat gain, and electric lighting ................. 86 x Figure 4. 1 Daylight illuminance test points on the work plane of the two models. .................................................... 92 Figure 4. 2 The orientation of the occupant’s view in the two models. ....................................................................... 92 Figure 4. 3 Solar radiation test points on the interior surface of the two models......................................................... 92 Figure 4. 4 The parametric workflow of daylighting and thermal simulation. ............................................................ 93 Figure 4. 23 Daylighting performance scores of three cases of the shoebox model on June 21 st , Sept. 21 st , Dec.21 st . ............................................................................................................................................................ 101 Figure 4. 24 Daylighting performance scores of three cases of the cuboid model on June 21 st , Sept. 21 st , Dec. 21 st . ........................................................................................................................................................... 101 Figure 4. 5 Hourly indoor average illuminances on June 21 st based on 11 tilt angles. .............................................. 150 Figure 4. 6 Hourly Daylight Glare Probability on June 21 st based on 11 tilt angles. ................................................. 151 Figure 4. 7 Hourly solar heat gain on June 21 st based on 11 tilt angles. .................................................................... 152 Figure 4. 8 Hourly indoor average illuminances on September 21 st based on 11 tilt angles...................................... 153 Figure 4. 9 Hourly Daylight Glare Probability on September 21 st based on 11 tilt angles. ....................................... 154 Figure 4. 10 Hourly solar heat gain on September 21 st based on 11 tilt angles. ........................................................ 155 Figure 4. 11 Hourly indoor average illuminances on December 21 st based on 11 tilt angles. ................................... 156 Figure 4. 12 Hourly Daylight Glare Probability on December 21 st based on 11 tilt angles. ...................................... 157 Figure 4. 13 Hourly solar heat gain on December 21 st based on 11 tilt angles. ......................................................... 158 Figure 4. 14 Hourly indoor average illuminances on June 21st based on 9 tilt angles. ............................................. 159 Figure 4. 15 Hourly Daylight Glare Probability on June 21st based on 9 tilt angles. ................................................ 160 Figure 4. 16 Hourly solar heat gain on June 21st based on 9 tilt angles. ................................................................... 161 Figure 4. 17 Hourly indoor average illuminances on September 21st based on 9 tilt angles. ................................... 162 Figure 4. 18 Hourly Daylight Glare Probability on September 21st based on 9 tilt angles. ...................................... 163 Figure 4. 19 Hourly solar heat gain on September 21st based on 9 tilt angles. ......................................................... 164 Figure 4. 20 Hourly indoor average illuminances on December 21st based on 9 tilt angles. .................................... 165 Figure 4. 21 Hourly Daylight Glare Probability on December 21st based on 9 tilt angles. ....................................... 166 Figure 4. 22 Hourly solar heat gain on December 21st based on 9 tilt angles. .......................................................... 167 Figure 5. 1 Hourly outdoor horizontal illuminances on June 21 st . ............................................................................. 105 Figure 5. 2 Hourly sun altitude on June 21 st . ............................................................................................................. 105 Figure 5. 3 Hourly outdoor horizontal illuminances on September 21 st . ................................................................... 107 Figure 5. 4 Hourly sun altitude on September 21 st . ................................................................................................... 107 Figure 5. 5 Hourly outdoor horizontal illuminances on December 21 st . .................................................................... 108 Figure 5. 6 Hourly sun altitude on December 21 st . .................................................................................................... 109 Figure 5. 7 Hourly outdoor horizontal radiation on June 21 st . ................................................................................... 114 Figure 5. 8 Hourly outdoor horizontal radiation on September 21 st . ......................................................................... 115 Figure 5. 9 Hourly outdoor horizontal radiation on December 21 st . .......................................................................... 116 Figure 5. 19 Hourly outdoor horizontal illuminances on June 21 st . ........................................................................... 118 Figure 5. 20 Hourly sun altitude on June 21 st . ........................................................................................................... 119 Figure 5. 21 Hourly outdoor horizontal illuminances on September 21 st . ................................................................. 120 Figure 5. 22 Hourly sun altitude on September 21 st . ................................................................................................. 120 xi Figure 5. 23 Hourly outdoor horizontal illuminances on December 21 st . .................................................................. 121 Figure 5. 24 Hourly sun altitude on December 21 st . .................................................................................................. 122 Figure 5. 25 Hourly outdoor horizontal radiation on June 21 st . ................................................................................. 127 Figure 5. 26 Hourly outdoor horizontal radiation on September 21 st . ....................................................................... 128 Figure 5. 27 Hourly outdoor horizontal radiation on December 21 st . ........................................................................ 129 Figure 5. 10 Hourly indoor average illuminances of three facades on June 21 st . ...................................................... 184 Figure 5. 11 Hourly indoor average illuminances of three facades on September 21 st . ............................................. 185 Figure 5. 12 Hourly indoor average illuminances of three facades on December 21 st . ............................................. 186 Figure 5. 13 Hourly daylight glare probability of three facades on June 21 st . ........................................................... 187 Figure 5. 14 Hourly daylight glare probability of three facades on September 21 st . ................................................. 188 Figure 5. 15 Hourly daylight glare probability of three facades on December 21 st . .................................................. 189 Figure 5. 16 Hourly solar heat gain of three facades on June 21 st . ............................................................................ 190 Figure 5. 17 Hourly solar heat gain of three facades on September 21 st . ................................................................... 191 Figure 5. 18 Hourly solar heat gain of three facades on December 21 st . ................................................................... 192 Figure 5. 28 Hourly indoor average illuminances of three facades on June 21 st . ...................................................... 193 Figure 5. 29 Hourly indoor average illuminances of three facades on September 21 st . ............................................. 194 Figure 5. 30 Hourly indoor average illuminances of three facades on December 21 st . ............................................. 195 Figure 5. 31 Hourly daylight glare probability of three facades on June 21 st . ........................................................... 196 Figure 5. 32 Hourly daylight glare probability of three facades on September 21 st . ................................................. 197 Figure 5. 33 Hourly daylight glare probability of three facades on December 21 st . .................................................. 198 Figure 5. 34 Hourly solar heat gain of three facades on June 21 st . ............................................................................ 199 Figure 5. 35 Hourly solar heat gain of three facades on September 21 st . ................................................................... 200 Figure 5. 36 Hourly solar heat gain of three facades on December 21 st . ................................................................... 201 Figure 6. 1 The workflow of the methodology for kinetic facades design and evaluation ........................................ 136 Figure 6. 2 Examples of the complexity of the geometric configurations generated from parametric definition. ..... 137 Figure 6. 3 The diagram of model-based predictive control shows the operation cycle. ........................................... 138 Figure 6. 4 The score results of daylighting environmental performance.................................................................. 139 Figure 6. 5 The parametric workflow of introducing annual daylight and thermal metrics....................................... 141 xii ABSTRACT As one special type of building envelope, kinetic facades are increasingly being applied to contemporary architecture and building practice for the advantages of either aesthetics of architectural motions or better environmental performance. As is known to professionals in the building industry, the building envelope is an essential structure that plays a significant role in building daylight environments and energy savings since it works as an artificial barrier to isolate the indoor condition from the outdoor environment. The kinetic envelope can actively affect the built interior environmental quality, the building energy consumption, and the occupant's visual and thermal comfort. Thus, it is a significant mission to discover the principles and rules of the dynamic façade and its contemporary development and practice on improving environmental performance. The research focuses on the interior daylight environment analysis with kinetic façades in terms of daylight metrics by using parametric workflow. The process of parametric simulation can detect and investigate the existing advantages and deficiencies with respect to the daylight index. By processing simulation data, the workflow uses model-based predictive control as a control algorithm to operate the movable façade for better daylight effects, compared with the static counterpart of a kinetic façade, the research proposes the potential entry point or approach to improve the interior daylight condition. Beyond that, the research applies a systematic approach and integrative view to establish daylighting evaluation and improvement for testing and verifying the interior space daylighting performance of kinetic facades. Keywords: kinetic facades; intelligent systems; dynamic daylighting; daylight metrics; parametric design; adaptive shading. 1 CHAPTER 1 1. INTRODUCTION This chapter introduces the concept of kinetic façades, separately from the perspectives of the application of kinetic facades as an alternative strategy for the building enclosure, and the mechanism as defined by its taxonomy, configuration, characteristics, and functionality. Simultaneously, the chapter investigates its effect on building environment and energy performance and the computational parametric tools that architectural designers can usually apply in developing and analyzing kinetic facades from a conceptual sketch to a parametric model. The building envelope is acting as a barrier layer that isolates the indoor built environment from the outdoor surroundings. Kinetic façades are a type of transformable building skin that can vary its form to adapt to the exterior surroundings. The design of the kinetic mechanism involves the integration of aesthetic appearance, environmental performance, occupants experience, and technological method. In the design stage, architects and engineers integrate numerous design variables and various design parameters; all these factors complicate the kinetic facade to be a comprehensive and versatile mechanism. Specifically, the building envelope consists of a set of components exposed to the external environment such as roof, walls, windows, doors, and foundation; All these primary elements are connected to fabricate a sealed and efficient container. Kinetic facades are one particular type of building envelope that is capable of transforming its pattern or move its unit using a control algorithm in order to target either architectural aesthetics or building interior environment control and energy efficiency. As far as it concerns the correlation between kinetic facades and daylighting control, it is difficult for architectural designers to obtain an effective and efficient evaluation approaches to analyzing daylighting, thermal, and energy control performance of kinetic facades during the early design stage. The reason is that the geometrical configuration of kinetic facades is too complex to be designed by the regular method and simulated by the tools for conventional facades with static properties, and the environmental performance refers to interdisciplinary knowledge and expertise. 1.1 Responsive Kinetic Facades for Environment and Energy The building envelope is an essential structure that plays a significant role in building environment and energy performance. Natural light penetrates the building surface into the indoor space to determine interior daylight illumination and solar heat gain and thereby affects the energy consumption of electrical lighting and air conditioning. On the other hand, building 2 energy use accounts for a large portion of total energy consumption because buildings must meet an essential need for people’s indoor activities. Figure 1. 1 The mean percentage of time spent by the NHAPS respondents in six different locations on the diary day. Source: Klepeis, Neil E. et al. (2001). The National Human Activity Pattern Survey (NHAPS): A resource for assessing exposure to environmental pollutants. According to a statistic in The National Human Activity Pattern Survey (NHAPS): A Resource for Assessing Exposure to Environmental Pollutants published by Lawrence Berkeley National Laboratory (Klepeis et al., 2001), Americans spend nearly 87% of the total time indoors. The routine of modern buildings often leads to much energy usage occurring in the built interior environment (Figure 1.1). Another previous survey by the Energy Information Administration (EIA, 2013) revealed that more energy was consumed in the building industry than the industrial and transportation sectors in the United States. As reported by the U.S. Department of Energy, in 2011, the building energy use in the U.S. accounted for roughly 40% of entirely main energy use and relevant greenhouse gas emissions (Figure 1.2). This amount of energy expenditure fluctuates over different developed countries and areas; However, a consistent global situation of huge energy consumption in a building is constantly emerging in terms of exacerbating resource crisis and environmental due to excessive energy resources exploitation and massive emissions of carbon and pollutants. 3 Figure 1. 2 Energy consumption distribution according to sectors. Source: Phase Change Composite Materials for Energy Efficient Building Envelopes. https://www.seas.ucla.edu/~pilon/PCMIntro.html Historically, as a shelter, the building has prevented the external extreme climate or weather from threatening human life, while the building facade works as a shell between inner and outer spaces because of its capacity to maintain interior comfort and improving environmental control. In recent years, however, the unprecedented development of building equipment such as heating, ventilation, and air-conditioning (HV AC) system engages in facilitating the indoor environment effectively instead of regulating the external envelope systemically; Likewise, the vast use of artificial lighting rather than mainly counting on natural light promotes lighting ambiance in buildings. As a result, buildings assume a conspicuous energy expenditure in sustaining indoor comfort, which produces one-third of the entire greenhouse gas emission. Therefore, the conventional building facades do not function efficaciously enough to treat the issue, and components of the envelope such as roofs, walls, and windows function separately. As a special interface between outdoors and indoors, kinetic facades in the past few years have performed to be increasingly utilitarian as alternatives to static building envelopes since they work as a coordinator to meet the complex and diverse requests with respect to occupant comfort, energy efficiency, and cost efficiency (Sharaidin, 2014). Kinetic facades represent an innovative way of architectural components that influence the indoor environment and energy use (Risen, 2017). Specifically, the dynamic skin system controls the entry of natural light and solar heat using several open-close states, depending on more operation of passive effects and less artificial active system, including HV AC and electrical lighting, and therefore it closely correlates with its occupant’s thermal and visual comfort. By contrast, the conventional building facades are manipulated in relative static means by which the occupant can only manually adjust windows and blinds to alter air ventilation and heating exchange. However, the fact is that the physical state of the façade surface constantly fluctuates in accordance with the exterior environmental factors so that the static envelope cannot automatically vary its state to accommodate the outdoor surroundings simultaneously, but it has to wait for the occupant’s 4 response. Instead, kinetic façades with active mechanism are adaptive to achieve the optimum position based on environmental analysis, implementing a series of motions to fit in the real-time surroundings. The terms “responsive” and “adaptive” in this thesis are used to represent the interaction between the fabric of kinetic facades and the external environment. Typically, kinetic facades are defined as the capability of building envelopes to respond and adapt to the variations of the exterior environment (Sharaidin, 2014). They are mainly assembled to build or maintain a satisfactory indoor environment such as daylight quality and thermal condition. The outdoor environmental factors, including sunlight direction, solar radiation, ambient temperature and humidity, and wind velocity, etc., varies daily, seasonally, and yearly, while the kinetic facade changes its state of motions. A kinetic facade is a skin that makes the building responsive to the environmental factors to keep the occupants comfortable as well as increase energy savings from electrical lighting and air conditioning. Kinetic facades have appeared in many architectural practices all over the world, either to promote sustainability or to be appealing with technology innovation (Grozdanic, 2016). The buildings covered with kinetic façades are capable of performing more energy-efficiently and supplying more visual and thermal comfort than the ones with conventional facades. Because of the specialty of kinetic facades, they can be actuated to alter their forms and states to adapt to real-time daylight for optimal interior conditions such as maximum utilization of daylight and minimum glare. For example, Al Bahar Towers in Abu Dhabi, UAE, assembled with a dynamic “mashrabiya” shading system computerized and actuated to respond to weather changes, has reduced solar heat gain, improved indoor lighting, increased occupant comfort, and decreased energy consumption by 50% for office spaces alone, and up to 20% for the building overall with the reduction of CO2 emission by 1750 tonnes annually (Figure 1.3). Another example is Council House 2 (CH2), and its kinetic timber shutter is claimed to raise the occupant productivity by 10.9%, and satisfy 80% of the occupants about indoor circumstance, reduce power usage by 85% and lower emissions at 13%, and it supplies an outdoor view to 80% of occupants (Figure 1.4). However, there is a rare sophisticated evaluation approach and system on kinetic facades because of its complexity and diversity. Usually, the kinetic facades could end up with higher construction and maintenance costs over the conventional because it requires higher precision of design, manufacture, and assembly to realize pre-programmed motions. Especially, the daylight evaluation approach is crucial to realize an operable kinetic system in the design phase. In other words, the more expensive kinetic facade system is buildable to the owners only if its advantages of daylight control outweigh the negative factors such as increased energy consumption and occupant discomfort. 5 Figure 1. 3 Shading modules folding or unfolding in different positions of the kinetic envelope. Source: https://www.reddit.com/r/architecture/comments/5mflra/sunshades_on_the_al _bahar_towers_abu_dhabi/. Figure 1. 4 Council House 2 (CH2), kinetic timber shutters of western façade. Source: Alotaibi, F., 2015. The fabrication of kinetic facades is not a novel idea that decreases the energy request of lighting and air conditioning; the earliest version of responsive facades could trace back to the 1920s. In recent years, the increasing pursuits of this domain have appeared in many research projects and publication works, while an increasing number of buildings have been designed and fabricated with kinetic facades for environmental concerns. Previous studies emphasized the façade function on daylighting control of interior space. To deal with constraints of designing kinetic façades respectively, a highly integrated design and the research involves architects, computational designers, and enclosure consultants to cooperate by synthesizing various simulation tools. Still, it is vital to discover kinetic envelope patterns in this design process. Usually, dynamic enclosure patterns are composed by fabricating singular unit alteration. The research of the connection between dynamic façade and building environment performance is to explore specific exterior patterns or forms to maximally meet the occupant’s requirement for the 6 interior environment. Particularly, it could be rationally implemented that the occupants can be promised the appropriate solar heat gain, the reasonable penetration of natural light as well as the optimized visual comfort, and so forth. Concerning adequate daylighting rather than excessive glare, architectural designers anticipate abstracting the basic patterns from the discovery of decomposition and composition of fundamental geometries to place static patterns or even modulate a dynamic formation with a relatively high cost, so far. The mechanism of kinetic façades demonstrates to be superior integrity and more intelligent, especially for the buildings with complex and curved interface structures compared with the constraint of traditional blinds and shutters. However, it is difficult for Architectural designers to obtain effective approaches for analyzing and testing interior daylight and energy performance of buildings with kinetic facades during the early design stage, so they cannot gain valid feedback about the daylight performance related to kinetic facades, which should be an adverse situation for them to develop the design of the kinetic façade and even a defect for the whole building. 1.2 Intelligent Kinetic Facades The research on kinetic envelopes has been continuously advancing since technical applications such as the device and the material are being pushed forward all the time within the information and post-industrial age. Along with new technology and the emergence of new materials, the taxonomy of kinetic configurations can be modified or even transformed. However, the current methodology of categorization is an integral and organic system that is based on the existing types of patterns and motions and material properties to analyze the intrinsic connection between the configurations and the variables. 1.2.1 Intelligent Kinetic Building System The word “intelligent” was initially used to depict buildings with the word “smart” at the start of the 1980s (Wigginton & Harris, 2002). Since then, building facades that are characterized as intelligent performance are called “intelligent building skins” where the envelope represents an important aspect of the intelligent building system. Intelligent kinetic facades are described as one form of dynamic active mechanism that enables its interface to interact with external surroundings based on environmental factors changes such as weather changes daily and seasonally, as well as human need in the context, for either enhancing the environmental comfortability or reducing the energy consumption indoors. Designers can flexibly define intelligence in terms of various variables and parameters depending on the intentions and methods. Also, all definitions of intelligence are identified with the impact of behavior and judgment of living organisms. 7 In the AEC (architecture, engineering, and construction) industry, buildings are expected to apply intelligence on facades, materials, sensors, or even systems for strong performance. The radical reason is that human being is an active living organism who has intelligence as a fundamental characteristic to adapt to their existing surroundings and reconstruct it for easy subsistence. Thus, humans can be understood as an original living model for the application of building intelligence. Intelligence is mentioned by people when some objects or substances can function with automatic property and control systems (Fox & Kemp, 2009). The facade intelligence is based on an archetype that connects to human need and energy efficiency indoors by a responsive system that is responding to the environmental factors, and an intelligent façade is actively mediating the indoor environment and the user inside. Based on this point, the study will investigate the intrinsic relation between the building envelope and luminous interior conditions in terms of human demand and behavior. Building operation is commonly driven by the occupant’s inclinations and activities although these differ in diverse climate conditions and space functions. An intelligent building system should be capable of performing optimal status by conducting the following procedures: - Building harmony between the occupant’s activity and indoor environment - Actively respond to exterior surrounding changes and human needs - Expedite cost-effective adaption to users’ behavior shifts An intelligent building system is conscious of operating the interior environment by detecting information, analyzing data, and responding in a proper measure (Figure 1.5). The interactive course encompasses different sections of building operation; still, the interior environment and energy performance is regarded as the eventual expectancy as far as the building intelligent system. Embedding the data about the changes in user behavior and the variations of exterior circumstance into a cognitive and algorithmic building system could dramatically strengthen the space environmental performance. Figure 1. 5 The process of intelligent building – data flowing steps feature an intelligent façade. Source: El Sheikh, 2011. 1.2.2 Kinetic Facades and Daylighting A kinetic envelope acts as a medium through which the built interior spaces interact with the exterior environment. Buildings are prone to variable environmental conditions, where they 8 could be geared up with dynamic facades to intelligently adapting to surroundings changes and the occupant’s needs. Intelligent kinetic facades are one type “smart building skin” that is technically coping with dynamic and mutable daily activities for improving indoor conditions to satisfy occupants; the dynamic optimization of interior spaces is subject to human behavior and interaction, and it should be competent to instantly adopt measures against a fast-changing mode of human interactive and interior conditions The kinetic façade, as a portion of the integrative intelligent building system, can be identified as a performance-based mechanism related to interaction or responsiveness. It can form into a portion of the main system like operable windows or could be a standalone secondary layer such as external louvers. Michael A. Fox calls the latter type as “dynamic” kinetic typology in which the kinetic module is a component of an integrated system but can operate separately (Fox & Kemp, 2009). The application of a secondary kinetic skin can present the potential of improved environment and energy performance other than innovative design and geometric motions. 1.2.3 Typology of Kinetics in Architecture Kinetic architecture can be embodied in numerous and diverse forms and types in building facades according to dynamic features. The module or component of the kinetic façade exists in distinct configurations for the intention of being either individuality or high efficiency. Practically, understanding the current taxonomy of kinetic facades benefits the research on kinetic motions and its environmental impact assessment. 1.2.3.1 Kinetic Patterns Prototype of Architectural Facades Recently, the architectural tendency of material application plays a significant role in dividing the architectural research and practice. In the domain of engineering, it develops through integrating computational simulation and practical experience and advancement of the construction method. The design of the kinetic façade appears to be a wholly different workflow for architects and designers because they often engage in seeking an optimal resolution of enclosure that adapts to the environmental variations. Specifically, designers concentrate on the final phase that defines the physical details of components and materials, whereas designers are demanded to consider the data input and control system other than the modules, assembly approaches, and techniques when it comes to kinetic facades. The design methodology is always changing depending on the competency and proficiency, resulting in different mechanisms to a specific dynamic envelope. It is significant to research kinetic patterns and configuration for fabricating modules and components that can execute a set of functional motions. Kwitter (1992) and Moloney (2011) 9 developed the idea as the basic conception of the kinetic pattern, which depicts the movements of variation as well as the motions of singularities and points. That varying movement can be treated as a prototype of the kinetic process, and the intricacy of kinetic motions is possibly referring to more than two parameters to control the interactive motions. In the process of interaction, the kinetic patterns are determined by variables, namely the forms of geometric variation that are generally categorized into folding, sliding, rotating, retracting, expanding and transforming as the motion variables change. Thus, the recognition of the variable geometry and the kinetic motion facilitates the research on adaptive envelopes that ameliorate the indoor environment in terms of outdoor conditions (Figure 1.6 & 1.7). Figure 1. 6 Dynamic shutter module of Institut du Monde Arabe. Six shutter plates rotate to open and close the aperture in response to the daylight environment (Nouvel, 1987). Source: Sharaidin, 2014 Figure 1. 7 kinetic facades panels of Al Bahar Tower in Abu Dhabi designed by AEDAS. Source: Sharaidin, 2014 10 Many motions have been recognized and employed in the kinetic design according to the available studies of buildings that were assembled with kinetic envelopes (Moloney, 2011). This study elaborates on five forms of kinetic motion that have been widely identified in kinetic facades so as to interpret the mechanism explicitly in the kinetic design process (Figure 1.8). Figure 1. 8 Summary of the five experimental projects with kinetic motions. The diagrams show the kinetic configurations and mechanisms. Source: Sharaidin, K. (2014). The movement of sliding and rotating represents one type of variation modes across the kinetic envelopes for either aesthetic appearance or physical function like environmental conditioning (Moloney, 2011; Razaz, 2010; Schumacher et al., 2010a; Zuk & Clark, 1970). These kinds of motion originated from the engineering industry, usually detected in the use of a pulley system. However, this system seldom showed up in the domain of architecture, particularly in a building, which can be incorporated into kinetic facades such as the building facades of the Nordic Embassies that demonstrate the rotation on a large number of horizontal panels (Schumacher, Schaeffer, & V ogt, 2010b). Additionally, the model of the Wave project presented the motion of sliding and rotating with a simple conception. 11 The motion of sliding and retracting is thought to be related to the deployable structures that generate the adaptive structure such as scissor structures and umbrella-like structures. The motion includes an integrative and holistic translation such as Chuck Hoberman’s scissor structure, which utilizes retracting motion to alter the shape responsively (Hoberman, Davis, Drozdowski, & Wight, 2013). Another motion of contraction and expansion was developed in the model of Scissornet from the previous application by incorporating flexible structures. That system uses an expandable and contractable elastic material, and it can be flexibly deployed on the tier of components like adding, subtracting, as well as changing position and assembly order. Apart from the use of rigid material, there is still one different motion that involves inflating or deflating like a balloon, including soft and flexible elements. The prototype demonstrates to inflate or deflate a balloon to achieve the movement of opening or closing. This application can be found in the kinetic envelopes fabricated in the Media ICT building in Barcelona (Geli, 2010). The expanding and retracting motion of the project Triangular is one type of motion involving the principle of the previous four types. The prototype was conceiving the integration of triangular components that can be tested to respond to analogous conditions such as thermal heat and daylight. The exploration also includes experimentation of stretching the material to produce the balloon-like shape. 1.2.3.2 Kinetic Mechanism: Patterns and Motions In recent years, kinetic architectural design has been applied to the building enclosure. This approach utilized kinetic mechanisms to respond or interact with the environment. To some extent over media or sunlight screens, kinetic mechanisms can be categorized into three main mechanisms (Moloney, 2011). The first approach is a mechanical and structural system in mobile design developed by Hoberman Associates, representing international design and construction consultancy in terms of kinetics. The second approach is considered as a taxonomy of kinetic control mechanism initiated by the MIT Kinetic Design Group, which provides three methods to control kinetic structure like deployable, dynamic, and embedded. The embedded category is most effective because of directly controlling factors such as thermal condition, light, and ventilation. The third approach is proposed by Oosterhuis and Xia (2007), using Hyperbody’siA/Protospace software based on flocking algorithms. Kinetic architecture research in a building’s motion ability and require the responsive ability to environmental changes. The current kinetic system under intelligent management operates its command procedure based on the composite of three major parts: mechanical engineering, embedded computation and responsive architecture (Mahmoud & Elghazi, 2016). Kinetic motions can be simplified into 12 three fundamental forms: rotation, translation, and scaling (Figure 1.9). Furthermore, they can be randomly combined to form a synthesized transformation (Schumacher et al., 2010). Figure 1. 9 Mechanical movements of rigid architectural elements. Source: Schumacher, M. et al. (2010). The section introduces five synthetic forms of kinetic patterns of motion based on different models after a holistic comparison of potential kinetic geometries and mechanisms (Figure 1.10). At first, expanding is one type of kinetic mechanism based on the pattern that configures its material attributes to change and adapt to a specific demand; These materials that perform special attributes could include, for example, Shape Memory Alloy (SMA) and smart materials (e.g. biometal). An example of the expanding type is the active surface which shows to be flat shape as well as single curved and double curved. This sort of material system that reacts to an external stimulus by changing their properties (physical or chemical) is categorized as non- mechanical kinetics. In contrast, mechanical kinetic relies on an exterior source of energy to actuate the mechanical elements. The study focuses on mechanical kinetics because this kinetic behavior is a more active and efficient way that can actuate the façade motion instantly. Besides, transforming is one of the kinetic motions in which differential forms and configuration coexist in the domain as well as being influenced by material ratio. Nonetheless, it depends more on the intention of kinetic design. The example motions illustrate that incorporating dynamic material 13 can validate the potential geometry transformation in specific positions and on tiny movement. As for these two motions, the target of research emphasizes the inherent relation between material attributes and movement mode because they belong to variations that occurred on the micro-level. On the other side, three types of kinetic configurations can be developed to apply to macro dimension motions. Retracting is one of the kinetic patterns which combines the surface and its structure to vary simultaneously. The behavior is based on the synchronic movement due to the interlocked module. Another type of kinetic behavior is folding that is broadly employed in the kinetic façade of the building industry. The reason is that the motion is a linear and straight mode easily acquired in the construction of a dynamic system. The knowledge of folding as a basic form of movement benefits the designers to research and create more complicated mechanisms that can respond to environmental changes in high efficiency. At last, sliding is the adaptive movement and mechanism that is principally relying on a rail track to translate, changing the panels vertically or horizontally. These three types of motions are built on different geometries that are incorporated into the capacity of kinetic facades in response to daylight conditions. The parameters are defined in a flexible way to adapt to diverse variables. In consequence, the variation of variables will reproduce model alteration and indicate different combinations of specific motions. 14 Figure 1. 10 The diagram demonstrates the kinetic patterns of motion and their features which are potentially used to design kinetic facades in response to environmental performance. Source: Sharaidin, K. (2014). 1.3 The Composition of the Kinetic Mechanism The kinetic mechanism stemmed from the engineering industry; it can be treated as an industrial product that involves interdisciplinary applications on architecture such as mechanical engineering, computer technology as well as manufacture industry. In terms of the architecture and building industry, the kinetic mechanisms or patterns are the seemingly advanced application of technology. However, it is not difficult to understand the intrinsic operation principle because it is essentially a gadget like a smart fan in daily life. A kinetic mechanism is composed of movable components and fabric including structure, connections, actuators, materials, and control systems (Figure 1.11). Despite the complexity of the integration of diverse elements, it is not necessary to construct a kinetic system with all the components in a building (Sheriden, 2000). 15 Figure 1. 11 The composition of the kinetic Mechanism. Source: Elkhayat, Y . O. (2014). The movable structure is the main body of the kinetic mechanism; it is physically assembled with several similar or different material parts to constitute the specific functional modules. It also has evolved from the static structure and continuously enhances its flexibility, dynamics, and activeness, for example, convertible structure and cantilevers, as well as the movable roof structure. Also, the movable connection is significant to interlock each part of the structure with the mobile property; for instance, the bearings and hinges can support the different modules to activate responses and implement pertinent motions. Unlike the above two parts, the actuator is a sort of device that converts some commands from the control hub into adaptive movements; that is, it can move the holistic kinetic system through coordinating different parts to conduct a linkage mechanism. It is actually the dynamic unit that consumes energy to sustain the operation, and the power type can be either electrical and pneumatic or hydraulic resources. It works as the terminal to move the physical object or surface according to the pre-set rules by the control system. For example, hydraulic pistons and pneumatic muscles are the actuators based on pressurized adjustment. The material plays a significant role in functioning the kinetic mechanism for a particular motion; Furthermore, the material use is closely interrelated with the structure performance due to the material attributes. The material with high flexibility and lightweight are able to strengthen the applicability and mobility of the modules. The control system, working as a commanding authority, is a series of equipment and devices that commonly analyze collected data to release command for the structure to complete the responsive movement, which will achieve some effect either for aesthetics or for environmental modulation. The control system is generally divided into two parts: inputs and controllers. The inputs provide the relevant surroundings information as data to the controller, and five types of mode are manual input, sensors or detectors, prior internal information, manual programming, and the Internet. On the other hand, a controller is a special form of computer that receives data Kinetic Mechanism Movable Structure Movable Connection Actuator Material Control System 16 from the inputs part and executes the command to move or actuate the fabric. Three modes of controller system are internal control, external control, and complex system. A kinetic mechanism must be actuated for being adaptive and responsive. For instance, a movement input should be delivered to the mechanism to complete the behavior command. This process is commonly operated by control of interior forces, exterior forces or transformations. Actuators receive a command from the control system to execute particular motions when a control system converts imported signals from user interaction or environment variation into behavior commands. Sensors connect the surrounding space to the adaptive system; they detect the external condition and collect environmental data to compare with the relevant expected state (set-point), then send the information to the processor of the control system; Also, sensors monitor all the actions and reactions of the mechanism. Basically, the actuators are a variety of modules that can shift the power into a corresponding action to perform a reaction on the mechanism, altering the properties of the kinetic system such as patterns, dimensions, and motions because they respond to the processor’s instructions related with the stimuli detected by the sensors. The processor in the control system is a controller module or computer system in which the imported data is processed and analyzed by computational calculation to generate a rational report as a command for the actuator to fulfill. The control system consisting of sensors, processors, and actuators are defined as the extrinsic control or active mechanism. Based on detailed external conditions such as the temperature, humidity, wind, solar radiation and so forth, the system is empowered by electric energy to actuate systematic components and modules to move materials or objects by following pre- programmed logic. 1.4 Dynamic Daylighting Daylighting is an environmental system that regularly functions in the field of energy exchange and circadian rhythm. As is defined by V olla and Seinre, “daylight is a balanced mixture between diffused daylight and direct solar” (V olla & Seinre, 2012). Despite the necessity of daylight being controlled like other environmental factors, the fact that the daylighting is always changing its character day by day or even during the day makes it difficult (Khoshroonejad, 2010). The variation of daylight condition significantly influences the building interior space in terms of lighting and thermal energy, and also affects the occupant’s visual and thermal comfort. The state that solar position periodically alters along its path produces variable daylight on the earth. The issue of effectively regulating the daylight is significantly challenging since the mutable daylight benefits nature lighting of interior space but probably cause glare inward, while solar heat comes in to increase the cooling load in summer. 17 Daylighting is an architectural design strategy that exploits natural sunlight to illuminate building indoor spaces, reducing electric lighting energy use, maintaining a productive interior working environment, and benefit indoor occupants for their physiological and psychological health. The daylight means the visible spectrum of light energy irradiated from the sun, which is a source of full-spectrum light with ideal color rendering. Daylight is a powerful resource that architects can use to enhance the indoor environmental quality, energizing the interior lighting ambiance, thus reinforcing the connection between humans and indoor space. Successful daylighting refers to the thoughtful integration of sunlight exploitation that thoroughly resolves the trade-off of sunlight penetration, daylight variation, solar heat gain, glare, and artificial lighting. The satisfactory daylighting design benefits the occupied spaces and occupants with healthy indoor circumstances, improved occupant productivity, and reduced electric power consumption by optimizing the quantity and quality of interior lighting. A well-designed interior space should be empowered with a capacity of limiting and harvesting daylight to rationally utilize solar energy, and it should have a system that is able to interpret the variable situation of natural lighting. 1.4.1 Daylight Function People often prefer natural lighting because of its benefits in spite of the excessive application of artificial lighting. Daylight is usually seen as a vital characteristic in architecture design since it not only connects the occupant with the outdoor view but sustains the opportune indoor environment for body health and working productivity just as it provides energy and environment to the organisms. Besides, daylight is a dynamic environment factor that changes the indoor condition continuously not like artificial light. Daylight strategy is featured with sufficient and suitable utilization of natural sunlight, being actively effective for human’s health, working environment, and working productivity. Most people regard the comfortable daylit condition as a pleasant and productive space for working and living. Natural light is indispensable to body health as Koster concisely interpreted in describing daylighting influence on people’s physiology and psychology in his book: Light synchronizes the human biological clock with day, night and seasonal rhythms. A lack of natural daylight can lead to disorders of the automatic nervous system, loss of energy, fatigue, a tendency towards self- isolation and metabolic disorders. Conversely, intensive light therapy has been shown to support the healing process (Koster, 2004). 18 Figure 1. 12 Veitch’s proposal that lighting quality is determined by the balance of the factors included in the diagram. Source: Veitch, J. A., 2001 Lighting quality indicates the extent to which a luminous environment sustains the occupant’s needs including visual performance, post-visual performance, sociality, mood, health and safety, and aesthetic judgment (Veitch & Newsham, 1996). Good lighting quality builds appropriate lighting situations for sight and task performance, to encourage satisfying communication, to grow comfortable mood, to benefit healthy ambiance, and to create an aesthetic appreciation of the space (Veitch & Newsham, 1998). Therefore, the quality of indoor daylighting conditions significantly determines the comfort level of interior spaces. On the other hand, indoor daylighting strongly influences the energy consumption of indoor spaces. The more daylight strategies an interior space can utilize to improve the occupied condition, the less energy use that space can achieve. In fact, a large amount of electric power can be saved by incorporating constructive daylighting strategies, such as introducing adequate natural light to the interior at daytime instead of lightening artificial lamps and blocking solar heat penetration with shading blinds during summer. Well-designed daylight strategies are capable of reducing the huge power usage of electric light and air conditioning due to thermal loads (Figure 1.12). The ideal controls of solar light ensure sufficient daylight illuminance level; meanwhile, they prevent excessive solar heat gain as well as unfavorable glare. All these controls are often completed by passive approaches or can be better achieved via an active system. 1.4.2 Daylight Fundamentals Daylight, a full-spectrum solar energy source, is generally comprised of direct sunlight, diffuse light over the sky, and reflected light from surfaces. The daylight source can be divided into two 19 major aspects: primary source and secondary source. The primary source means a direct and fundamental origin of light suck as sun and sky, whereas secondary source indicates the object that bounces off the light but not radiates it like surfaces or reflectors. As for the built environment, reflective surfaces are various surrounding elements including the ground, adjacent facades, and interior surfaces. Thus, daylight at a position of the space often consists of direct sunlight, diffused light from the sky, exterior reflected light, and interior reflected light. Figure 1. 13 Indoor daylighting characteristics that affect occupants’ productivity and comfort. Source: El Sheikh, M. M., 2011. The indoor daylight environment is an integrative system that embodies various characteristics including brightness, luminous distribution, color, glare and veiling reflections, visual contact with the outdoor environment and individual control (Figure 1.13). From the occupant’s perspective, visual and thermal comfort is closely related to the amount of daylight penetration. The problem of unsatisfactory indoor daylight can accompany with either inadequate illumination or over-illumination, as well as direct sunlight projecting on the working plane or human eyes. These adverse phenomena are actually caused by the lack of control over the daylight characteristics or variables. 1.4.3 Daylight Performance Indicators In the past decades, several studies on daylighting quality and performance have been conducted to develop differential evaluation criteria or standards for indoor daylight quality. Also, they reached a consensus on the fundamental rating factors, such as illuminance level, luminous 20 distribution, and glare because of the discussion of those in any of these kinds of literature. However, every researcher explained individual additional factors in their studies according to the specific experimentation condition and goal. In 1994, the committee of quality of the visual environment of the Illuminating Engineering Society of North America (IESNA) proposed ten factors affecting lighting quality that can be adopted to rate daylighting quality: - Brightness of room surfaces - Task contrast - Task illuminance - Source luminance (glare) - Color spectrum and rendering - Daylight view - Spatial and visual clarity - Visual interest - Psychological orientation - Occupant control and system flexibility In the IESNA daylighting evaluation standard, some factors are difficult to use for evaluating the interior because of subjective and complicated individual variables included in those factors, like visual interest, psychological orientation, and occupant control. The other factors can be calculated or simulated through the formulas and algorithms. Also, research conducted by Marie-Claude Dubois in 2001 identified five main factors that determine the luminous environment in office spaces for daylight performance assessment (Dubois, 2001): - Daylight factor - Absolute work plane illuminance - Illuminance uniformity on the work plane - Absolute luminance values on the vertical plane - Luminance ratios between the paper task, the walls, and the video display terminal (VDT) screen. The thesis refers to relevant daylighting factors or variables to conduct the interior daylighting simulation and evaluation for kinetic facades based on the two practices for daylight performance indicators. 21 1.4.4 Daylight Simulation and Evaluation Daylight fundamentals encompass daylight metrics and threshold of daylight design criteria, and the approaches to simulate and assess the daylighting environment. As regards daylight metrics, one of the common daylight metrics to evaluate the adequacy of indoor illumination is illuminance on the horizontal plane although the quality of indoor daylight cannot be thoroughly explained with horizontal illuminance (Mardaljevic et al., 2009). Illuminance typically means the illuminating level of light arriving on the working plane for visual tasks. For example, a numerical value of 300 to 500 lux is suggested for the office space or the clerical environment; most of the artificial lighting is designed to meet the requirement of illumination level according to that type of standard (Rea, 2000). Furthermore, other metrics also used for daylight evaluation involve circadian stimulus levels, overheating from solar heat, and glare produced by strong sunlight, as well as daylight factor, daylight autonomy and useful daylight illuminance (UDI). Each daylight metric should usually have several levels of numerical range to differentiate its performance. Better interior daylight quality is represented as the occupant’s visual and thermal comfort in the working condition and the decrease of energy use of electric lighting and air conditioning, therefore symbolizing energy-saving and sustainability; daylight control is a crucial role that kinetic facades serve as solar shading devices to play (Elghazi et al., 2015). During recent decades, daylighting simulation has reached a high maturity. Especially, many advanced design tools such as Diva and Radiance can run complex daylighting simulations on complicated geometries presently (Vartiainen et al., 2000). Likewise, as a parametric energy analysis tool, Honeybee is a plugin in Grasshopper interface to manipulate daylight simulation through interfacing the analytical engines DAYSIM and Radiance. For instance, the approach to simulate and evaluate the horizontal illuminance is mostly defining a grid of simulative points on the horizontal plane, either on the ground level or on the working plane height, then running daylighting simulation based on each point and other input parameters to export feedback. Especially, the horizontal plane can be composed of certain points, areas or zones, and even a continuous grid (Mardaljevic et al., 2009). The daylight calculation method in the research refers to the DOE-2 Daylighting Calculation (Winkelmann & Selkowitz, 1985). The algorithm of the calculation method indicates that the overall illuminance on a reference point includes two parts: direct daylight (daylight from the sun or sky directly) and indirect daylight (reflected daylight mainly from interior surfaces such as walls, ceilings, and floors). The direct portion of interior daylight illuminance is determined by three parameters: window luminance, solid angle subtended by the window element with regard to the reference point, and the angle between vertical and the ray from the window element 22 center to the reference point. As regards these parameters, window luminance can be calculated according to sun positions and sky conditions, whereas a solid angle can be thought about with geometric analytics because of it involving extensive iterative computation for the angles and distances from reference points to window elements. 1.5 Computational Design Tool Along with the revolutionary progress in computer technology such as operational performance and graphical processing, Computer-Aided Design (CAD) tool kits have been increasingly launched into the design and construction market, and these programs are presenting more and more powerful troubleshooting capability. Thus, it is widely acknowledged that AEC (architecture, engineering, and construction) industry has benefited from the development and innovation of CAD tool kits which can conspicuously raise the design efficiency and undoubtedly offer multiple possibilities for a market prospect (Shea et al., 2005). The growth target of developing these tools lies in increasing the accuracy, streamlining the procedure from design to construction as well as improving the abilities to develop the complicated design. The design methodology has evolved into the system of three-dimensional modeling and parametric interaction from traditional two-dimensional drawing so far after a long period of design practice. Parametric design is an interactive process that discovers the design possibilities from the perspective of parameters (Gane et al., 2007). The parametric design engages in the instances discovery process where the design can be altered to optimize the results according to the variations of the parameters (Holzer et al., 2007). In consequence, it is counted as an efficient approach to explore the existing form of design and adapt to the subjective requirement and objective environment. Some parametric design tools are functioning in the design process of different subjects. Revit, a powerful tool of building information modeling, helps the designer develop the design parametrically form the design phase to the construction phase to the commissioning phase. Grasshopper is a visual programming plugin based on its parent software program Rhino which demonstrates the widespread application in architectural design (Viola et al., 2013); Likewise, Dynamo is also an embedded visual programming tool in Revit for precise design and construction (Figure 1.14). 23 Figure 1. 14 Left: the demonstration of kinetic façade motion in Revit. Right: The interface of hierarchical nodes control in Dynamo Script. Source: Shen, Y .T., Lu, P. W. (2016). Both tools are the programming interface that is performing extended capability to configure digital models since these can get access to the Application Programming Interface (API) to execute more complicated and iterative functions. If there were no visual programming tools plugged in the parent software, the architectural designers would be required to match applying various functions and writing code to model complicated conception or structure. Furthermore, Grasshopper has incorporated more programs like Ladybug and Honeybee that employ the same or analogous coding language from a third party recently; as a result, it enabled the program to execute more extensive functions. For instance, Honeybee, as a plugin of Grasshopper, contributes to linking third-party engines like EnergyPlus and Radiance into Grasshopper to operate daylight and energy simulation. Honeybee is functioning in transforming an architectural model into the energy model. 1.6 Hypothesis and Objectives 1.6.1 Hypothesis It is important to generate a control strategy for the kinetic façade using model-based predictive control and define an evaluation system for the daylighting conditions and energy performance of interior spaces with kinetic facades that will provide early advice in the design process. In the early decision-making stage of kinetic facades design, architectural designers do not have enough capability to analyze and test the daylight and energy performance of the kinetic façade that they intend to design. Typically, they cannot yet obtain effective feedback from the daylight simulation experiment, or they do not have an effective method to generate a dynamic control strategy for kinetic façade design. Consequently, the deviation of design orientation can lead to 24 an irretrievable model selection that limits the daylighting development of the building with a kinetic façade from the post-design phase to commissioning. For example, a large percent of occupants still complained about the case of daylighting discomfort in Al Bahar Towers even though the kinetic shading system had been developed with sufficient technical methods. The methodology of exploring and analyzing indoor spaces of kinetic facades with respect to daylighting environment, energy performance, and occupant comfort is proposed to operate in a parametric workflow of modeling and simulation; in addition, by comparing different façade patterns on the identical building model, the study assumes model-based predictive control can generate an efficacious strategy to operate the kinetic façade for improved indoor daylight performance, and the parametric workflow based on daylight metrics can export valid data and outcomes to accomplish an evaluation system of interior daylighting performance. 1.6.2 Research Objectives This research explores an experimental methodology to identify the kinetic patterns of motion and their daylight environment performance. Based on the hypothesis, the research aims to emulate and analyze the indoor daylight condition and test the synchronic daylighting control strategy based on the evaluation criteria of daylight metrics. The specific objectives are as follows: - Use the parametric software and workflows to model and simulate the dynamic behavior of kinetic facades in response to dynamic environmental variables (climate, occupant, structure, and HV AC system, etc.). - Establish an evaluation system for the daylighting condition and energy performance of the interior space with kinetic facades. - Experiment to compare the indoor cases with kinetic facade A, B, and C. 1.7 Summary As a medium that connects the building indoor space to the outdoor surroundings, kinetic facades act in a complex logic of the dynamic variable system to actuate their components and modules to perform an essential role of adaptive and interactive behavior, for example, either presenting aesthetic geometries and creative designs or promising comfortable and appropriate indoor daylight condition. However, it is imperative to study the underlying principle and mechanism of kinetic patterns or motions for developing their application scope. The study on the kinetic patterns or motions is the cornerstone to improve kinetic envelopes, while the operation of a simulative test and analysis is an effective way to practically fabricate kinetic devices on the building envelope. The development of kinetic facades has benefited from available digital modeling tools and parametric design programs that form a powerful and 25 integral experimental platform to design and simulate the kinetics assembly before the final construction. The research objectives are investigating parametric workflows to emulate and analyze the daylighting environment of indoor spaces based on computational control algorithms, also evaluating the daylighting performance of the occupied interior in a scientific system in terms of kinetic facades. The analytical workflow including digital modeling, daylighting simulation, and data analysis is systematically organized to integrate with dynamic control algorithms such as model predictive control (MPC) for recording the optimum kinetic states. The following chapter presents the referable background, method, and workflow used in the kinetic facade or shading study from other literature and researches as regard to the indoor daylighting condition and energy use. The other chapters propose the methodology and workflow that the study developed, as well as the data collection and analysis based on the parametric workflow. Finally, the two chapters describe the research conclusion about data analysis and the future work that the study allows for addition because of experimentation flaws and limitations. 26 CHAPTER 2 2. BACKGROUND AND LITERATURE REVIEW This chapter introduces related background knowledge about kinetic concepts, daylighting environment, as well as daylighting experiment and simulation method; then it reviews some previous research on the methodology of daylighting simulation by using kinetic facades to test indoor daylighting environment, as well as exploring the logic between the variation of kinetic facades and interior daylight performance. For example, the daylighting experimentation of interactive facades and the simulative and analytical practice on kinetic façades with parametric design programs. Furthermore, the chapter analyzes the feasibility, validity, and query of the former daylighting research that provides referable experimental approaches for this study. 2.1 Kinetic Concept A kinetic façade is a versatile and complicated field where interdisciplinary expertise merges in a way of parametric design thinking to accomplish an adaptive or responsive mechanism. Multidisciplinary researches of the pertinent subjects positively function in developing the kinetic facade in terms of dynamic geometries and motions (Figure 2.1). Figure 2. 1 An algorithmic process of kinetic façade design with the multidisciplinary study. Source: Hosseini, S. M. et al. (2019). Employing intelligent kinetic façades involves several phases including architectural conceptual design, mechanism, evaluation, materialization, and maintenance process; these phases of the procedure are depending on the design framework of kinetic architecture proposed by some 27 professionals and scholars. The architectural design concept resolves creative ideas and morphological methods to propose the unique and novel object for moving or transforming from the static state to a dynamic mechanism. As for the mechanism, it means the interactive and technical system that utilizes patterns and materials to build a kinetic module or component. Using morphological methods to combine architectural concept with mechanism results in generating kinetic facades with the responsive or adaptive capability to satisfy the occupants. Furthermore, the evaluation process functions to define the kinetic conception that impacts the forms and constitution of the façade. During the phase, several performance indicators are involved to assess the kinetic functionality and effect (Hosseini et al., 2019). Thus, the rational selection of daylight metrics for assessing the occupant’s comfort and building energy use can significantly influence the façade configurations. Consequently, an integral system of the architectural concept, mechanism, and evaluation is eligible to form a credible circulation for investigating patterns and motions of kinetic facades (Figure 2.2). Figure 2. 2 Digital modeling and daylighting evaluation of kinetic façade in parametric workflow. Source: Hosseini, S. M. et al. (2019). 2.2 Daylight Environment Factors on Kinetics Natural light always varies continuously along with time change. The dynamic flux determines that traditional static facades are not capable of altering their natures to respond to the external environment and keep the internal spaces in optimal daylight condition; in contrast, kinetic facades are an interactive mechanism that has the ability to screen daylight based on spectrum and strength, controlling daylight in real-time condition, and eliminating uncomfortable solar rays. The introduction of adaptive or responsive facades enables occupants to participate in the 28 active daylight control from the façade regulatory mode to interactive function. As a renewable energy resource, the available natural light can apply the direct solar rays and diffusive rays from the sky and circumstances to offset the artificial lighting (Baker, 2002). Basically, the daylight transmits the lighting energy that positively affects the occupants of interior spaces. Meanwhile, it also sends adequate solar heat gain and excessive daylight that cause the occupant’s discomfort. However, it is always challenging to handle both sides of the daylight effect at the same time, namely achieving visual comfort and making use of daylight because “nature is always in motion, never at a standstill” (Plummer, 1995). In terms of studying the relationship between daylighting and human needs, it is worth thinking about certain factors like light quality and quantity, light uniformity, and glare risk (Hosseini et al., 2019). Architectural designers aspire to explore effective methods to develop architectural forms in the early design stage in order to improve building energy performance and meet the occupant's visual and thermal comfort. At first, building form, as a microclimate modifier, is a significant factor that determines the character of buildings and its relationship with surrounding circumstances. An optimal building form influences the amount of available daylight into the interior environment (ASHRAE press, 2006). Besides, a special design with architectural elements such as atriums, courtyards, lightwells, and galleries is the alternative normal approach to introduce daylight into indoor spaces (Baker, 2002). The research on the relationship between building forms and optimal daylighting often involves several factors including geometry, layout, orientation, compactness, opening characteristics, proportions and material and components such as verandas, shading devices, water pools, and vegetation. Otherwise, the researchers also thought that the occupant's daily migration needs to be fully studied because this interactive factor modifies the microclimate to affect occupant’s thermal and visual comfort (Hosseini et al., 2019). Generally, it is acknowledged that geometry and orientation are the two most significant factors which adjust solar radiation and wind flow. For instance, Caruso and Kämpf utilized design parameters such as orientation, compactness factor and shading devices to simulate the environment with 10,000 up to 20,000 times based on numerous variables from 18 to 26 for obtaining the optimal state (Caruso & Kämpf, 2015). 29 Figure 2. 3 Study of the relation between building form and occupant position to design interactive façade. Source: Hosseini, S. M. et al. (2019). Also, Hosseini, Mohammadi, and Guerra-Santin engaged in using kinetic facades to improve visual comfort based on dynamic daylight and occupant positions because of dynamic solar movement along its track and the user's need for real-time control (Figure 2.3). They thought that self-shading form, systematic step for adjusting daylight, and configuration alteration are the potential method to be incorporated into kinetic facades for improving interior daylight comfort. Since the occupant’s comfort is involved in the time change, regulating the façade configuration across a period of time is a potential way to enhance thermal and visual comfort (Hosseini et al., 2019) (Figure 2.4). As one type of configuration on kinetic facades, three-dimensional pattern modules in the façade system are capable of altering its configuration to regulate the microclimate condition. For example, Al Bahar Towers is fabricated with one outer layer of the adaptive facade to control natural light, solar heat gain and glare for the daylight performance and the occupant comfort. The kinetic modules are operated with sensors and actuators to react to the exterior ambiance by transforming its structure in different patterns (Zaera-polo et al., 2014). Most kinetic facades are the outer layers that interact with environmental stimuli, usually being configured into several sorts of motions including translating, rotating, scaling and folding. The forms of kinetic facades function similarly to the other involved components, influencing the quality and quantity of natural light. Moreover, the dynamic envelope affects the interior spaces more powerfully due to the active operation and optimized control mode based on the pre- programmed system. Therefore, with respect to the kinetic facade buildings, the interior 30 daylighting is recognized to be defined in a more complicated building system and influenced by more intrinsic factors. Figure 2. 4 The kinetic façade interacting with sun position and occupant position to change the components configuration in hierarchical mode depending on the attractor and parametric facade logic. Source: Hosseini, S. M. et al. (2019). 2.3 Daylighting Experimentation Methods on Kinetic Facades 2.3.1 Experimental Method Category The research that investigates the general and evident features of kinetic facades through interdisciplinary study across every domain of kinetic facades are rarely presented. Those researches can be categorized into qualitative and quantitative research based on the study method (Hosseini et al., 2019) (Table 2.1). Qualitative research on kinetic facades studied and contrasted the existing examples and strategies to concisely summarize the characteristics and performance of responsive façades. For example, Ramzy and Fayed used a qualitative method to 31 study some types of kinetic facades about their advantages, defects and potential resolutions (Ramzy & Fayed, 2011). Moreover, Megahed studied some kinetic theories and compared different conceptions so as to present a notional system of kinetic classification and design method (Megahed, 2017), while Barozzi et al. analyzed the adaptive facades and summed up these most advanced facades (Barozzi et al., 2016). On the other hand, quantitative research analyzed and assessed the performance of kinetic facades based on multifunctional intrinsic features. Some researchers applied digital model simulation and parametric evaluation for rating a kinetic façade’s performance (Figure 2.2). For instance, Pesenti et al. investigated available origami patterns by applying the parametric digital model to lower down energy use of actuators (Pesenti et al., 2015); Mahmoud & Elghazi studied daylight performance of hexagonal patterns based on the motions of translation and rotation (Mahmoud & Elghazi, 2016). Likewise, Grobman et al. thought over the mechanism of kinetic shading devices using a quantitative and parametric simulation approach (Grobman et al., 2017). 32 Table 2. 1 Research topics on the kinetic façade using both quantitative and qualitative methods. Source: Hosseini, S. M. et al. (2019). Multidisciplinary research on kinetic facades simultaneously involves both quantitative and qualitative methods. Qualitative methods can be applied to propose a hypothesis at first, while quantitative approaches can be used to analyze and test that hypothesis in the following steps. Specifically, a qualitative study with a parametric method offers principles and constraints to dynamic patterns and motions; meanwhile, quantitative research like parametric modeling and simulation discovers a high-performance model with required function. The integration of both approaches in the uniform system probably resolves multidisciplinary principles and versatile mechanism for kinetic facades functionality. Therefore, the multidisciplinary investigation can be thought of as an integrative efficacious research system used on kinetic facades over the study process from hypothesis proposal to experiment conclusion. 33 2.3.2 Performance Simulation Kinetic façades are designed to make the building adaptive to its external environment to meet sustainable targets. The complexity of the property of kinetic facades causes the study and analysis of them to be more difficult than conventional static facades. In the building industry, architectural designers and engineers are usually required to evaluate the ordinary building scheme with the built environment performance and energy efficiency because of sustainability command, so they have to do these same analyses to the architectural conceptual design with a kinetic façade in order to ensure that the completed building will operate in a user-friendly and high-efficient way. The adaptive building envelope is known as a composition of complex systems that simultaneously impacts multiple physical environments such as thermal, luminous, air quality, etc. The performance of kinetic buildings is determined by local climate, interaction with occupants and other building systems like the HV AC system, unlike the conventional buildings. It means that the traditional evaluation method on building envelopes based on metrics such as U-value and g-value is not applicable to the adaptive dynamic façade due to its inherent time- varying behavior. Instead, a more precise and reliable evaluation method can be used to indicate its performance with more integrative indices such as total energy use, energy use intensity (EUI), and indoor daylight metrics (Loonen et al., 2017). Fortunately, building performance simulation (BPS) can potentially calculate these data for the building stakeholders (Clarke & Hensen, 2015). The integrative approach of parametric modeling and simulation for performance analysis of adaptive envelopes appears to function some helpful possibilities in the process of building design and operation, including making a decision of adaptive facades, predicting energy efficiency, evaluating future-oriented systems and materials, recognizing alternatives, exploring high-potential control strategies, sizing HV AC system, and testing the robustness of adaptive systems (Loonen et al., 2017). Because of the functions mentioned above, the workflow of modeling and simulation generates perception and intellect to deliver into the interaction between design and performance in terms of adaptive building facades, and thus can largely popularize it in the building construction industry and help develop innovative techniques. However, as is known to us, the modeling and simulation of adaptive facades demonstrate far more complicated than the performance evaluation of the regular static facades since the available simulation tools were initially made for the conventional purpose. The situation requests that the potential BPS users for analyzing adaptive envelopes should develop their simulation method consequently under some challenges. Due to the sporadic information currently, the simulation users have limited access to the instruction about aspects, including software selection, adaptive model availability, exemplary project, and important attention. 34 2.3.3 Control Algorithm 2.3.3.1 Rule-based Control Rule-based control strategies are operated to follow preprogrammed commands such as “if statement” according to the differential between set-point values and real-time values (Figure 2.5). Two rule-based control strategies would be defined to run in the research. The first rule- based controller (RBC1) is set to calibrate control variables in terms of the dynamic shading state for maximally introducing the daylight to arrive on the light sensor and avoiding over-lighting within occupancy time. For example, if the light sensor detects the illuminance value above the threshold of 500 lux, the dynamic shading device will be adjusted to the more light-blocking state which decreases the illuminance value below 500 lux. This position is rated again in the following period (e.g. 1 h). On the other side, the RBC1 controller is programmed to hold the dynamic shading on the most light-penetrating state in the unoccupied hours. By contrast, the second rule-based controller (RBC2) performs the setting on the least lighting state during unoccupied hours when the interior space still obtains the solar heat except that the same control as the RBC1 controller works during occupied hours. Consequently, the RBC2 should perform more energy efficiency during the cooling season. The strategy for perception and adjustment of lighting circumstance is theoretically applicable to thermal conditions likewise (Dussault, Sourbron & Gosselin, 2016). For the control strategy also known as heuristic control, the occupancy schedule is significant to determine the change of daylighting and energy use due to the time-based algorithm. It could provide very different results when the strategy is applied to various types and functions of built spaces. Compared with other advanced controls, the rule-based control excludes the optimization process based on the judgment of previous performance. Therefore, it is a relatively easy and preliminary strategy to conduct dynamic change. Figure 2. 5 Rule-based control workflow. 35 2.3.3.2 Model Predictive Control Model predictive control (MPC) was initially used in the chemical process industry, and the control method currently spreads over diverse industrial fields. The approach means a category of computer control algorithms that apply a specific system model to forecast its subsequent reaction through a finite period of time, also known as the prediction horizon. The MPC algorithm at each step of the control process improves the order of control value to acquire an optimum response through the prediction horizon Hp according to the predictive data of the model. Namely, it means that the system is offered with optimal control based on the predictive control model. The control input of that chosen order at the current time, also known as the control horizon, is transmitted to the system; then the system runs repetitive steps like that for the next control command. The optimization in the standard form is implemented in real-time within the controller (Dussault, Sourbron & Gosselin, 2016). The reason why MPC is popular in the field of the control algorithm is that MPC algorithms are capable of handling constraints that are frequently encountered in the control process without other appropriate approaches. Also, the MPC algorithm can address difficult states, and evaluate constraints on states and inputs included in the algorithm over the optimization process. It is generally acknowledged two types of MPC control algorithms: a linear model-based predictive control approach and nonlinear model-based predictive control (NMPC) approach. The former appeared in the early 1980s and were practically built up; while the latter arose after ten years or so and formed separate method to control in practice, but less popular. MPC is comprised of three components including the observer, the predictor, and the optimizer. The workflow of the MPC model is illustrated, assuming with all the completely gauged control inputs (Uc) and disturbances inputs (Ud= [Tout, Qsol,direct, Qsol,diff]) (Figure 2.6). Sensors of the building measured different data, like the temperature of the specific zone to feed the observer; then the observer estimates the unmeasured temperature conditions of the building according to the gauge and all the previously measured and estimated temperature states. Meanwhile, the predictor forecasts disturbance inputs for the next time section. Then the MPC computes the model control inputs for the next operative time section which minimizes the cost function including energy consumption and discomfort costs. The MPC controller was cooperatively made, respectively developed by MATLAB, optimized in YALMIP toolbox and programming solver Gurobi (Dussault, Sourbron & Gosselin, 2016). 36 Figure 2. 6 Model based predictive control architecture. Source: Dussault, J. M. et al. (2016). In terms of the daylighting evaluation of kinetic facades, the MPC model can be made from a variety of experimental variables and data. In other words, the variable like shades positions and angles, as well as the time point could be the model input, while the simulation data of daylight metrics could be the model output. Meanwhile, the control input is the dynamic shading state that meets a certain required daylighting threshold, and the building output is the actual results of the interior daylighting condition. As a result, the system builds an interconnected network with a large number of inputs and outputs. Within the network, the observer is a device that collects the feedback of the indoor daylighting performance on the basis of all aspects of daylighting data; the predictor is a device based on the database comprised of shading states and daylighting results, and it usually offers several predictive commands about the shading states; the optimizer is a controller that is capable of filtrating to produce more opportune control inputs like the shading state. 2.4 Daylighting Analysis Approach for Kinetic Facades In order to develop a more effective workflow, it is necessary to interpret the importance of the current daylighting simulation and analysis of kinetic façade design. 2.4.1 Introduction The current design process of kinetic facades requires the consideration of interior environment performance on the basis of the occupant’s comfort to define the dynamic pattern of façade modules other than the daylighting condition and energy performance by simulating the number of solar rays passing through façade components. Sometimes blocking direct sunlight with movable shading modules at some surfaces may degrade daylight conditions of the interior space since it can raise the amount of electric lighting use to meet the occupant’s demand or fit the required illuminance on the work plane, thereby causing the reduction of energy efficiency. 37 Additionally, the daylighting simulation is a more complicated process, and it produces more workloads due to analyzing each zone independently as well as defining different shading movement states. 2.4.2 Evaluation of interior daylighting performance Daylight is supposed to penetrate into the occupied area to offer the user a medium between indoor space and outdoor surroundings; Also, at least 75% of the commonly occupied area is required to access sufficient daylight. Four approaches can be mainly adopted to assess daylighting conditions: computational simulation, prescriptive calculation, field measurement, and the integration of three previous options (LEED, 2009). Typically, some computational tools of daylighting simulation are available with advanced functions, such as Radiance, DIV A, DAYSIM, and Ecotect; and they are able to process complex geometric spaces and surfaces, outputting exact data analyses (Elghazi, 2015). The process of manipulating daylighting simulation is tough and cumbersome because of divisional simulation on each zone or surface and integrative definition. Especially in the early design phase, an explicit and efficient workflow is more preferable when running plenty of parametric simulations. Besides, the prescriptive calculation is a simplified means on the basis of certain assumptions and building geometry- related variables like room dimensions and window size, etc. It is mainly used to predict and appraise the percentage of indoor area accessing daylight (LEED, 2009), however, not being an ideal tool for complex and detailed daylighting simulation. In addition, field measurement is defined as the method of indicating that a minimum of 75% normally occupied area should attain the illumination level of 250 lux through prescribed or recorded periods of indoor lighting measurement (LEED, 2009). The problem is that it can only be conducted based on real building space. Finally, three-approach integration can be more flexible and versatile for diverse situations. Some previous research demonstrated the methodology and algorithm of calculating daylighting illuminance values, solar heat gain, and lighting energy use. For example, Dong-Seok Lee et al. at Inha University put forward a new idea to evaluate changeable shading devices and infer the optimal operation, including calculating daylight factor, solar heat gain, as well as thermal load under the impact of the exterior shading devices on kinetic facades (Dong-Seok et al., 2016). Daylight factor (DF) is defined as the proportion of the internal illuminance value on a horizontal work plane from daylight to the unshaded, external horizontal illumination value under a CIE overcast sky. The indoor illuminance can be calculated by multiplying the daylight factor with outdoor illuminance by using the value of DF; the external illuminance is derived from weather data or other resources. In the research, the daylight factor is calculated for each occupied hour on the basis of movable shading states, interior space geometry, and material properties of glazing and shading devices. The researchers presented a manual calculation method to analyze 38 indoor daylighting environment instead of computational calculation, and the optimization is more straightforward when combining with other daylighting metrics later. 2.4.3 Simulation of energy performance Parametric energy simulation of kinetic façades is helpful to determine an advantageous configuration for high energy efficiency in the early design phase, rather than running energy simulation in building energy simulation program (BESP) because of the complexity. Most of the available BESPs are available for static elements analysis. The commonly used programs such as DesignBuilder, IES-VE, and eQuest, etc. merely allow for running annual energy simulation on a static building model, rather than one with dynamic or movable elements; also, BESP like eQuest cannot support complicated facade geometry for energy calculation. These limitations of BESPs will cause fault or invalidity in energy performance evaluation of kinetic facades. However, compared with building energy simulation, solar radiation analysis on the facade is easier to complete in Grasshopper with Ladybug and Honeybee. Some scholars have completed research to prove the feasibility and efficacy of interior energy analysis through solar radiation simulation. In order to compensate for the weakness of the approach with regard to the occupant thermal comfort, the research can involve more analyses about thermal comfort such as solar heat gain through the fenestration and visual comforts such as indoor illuminance and glare. The research conducted by Dong-Seok Lee et al. also proposed a new calculation method that contains solar heat gain through fenestration, electric lighting energy use, and heating and cooling load (Dong-Seok et al., 2016). They manually calculated solar heat gain according to the ASHRAE definition. The given variables were sun position, exterior solar radiation, window area, and glazing property, etc. Besides, they discussed three factors to precisely describe the daylight effect of dynamic shading as follows: 1) unshaded fraction, it shows the ratio of solar incident area to the total window area; 2) exterior solar attenuation coefficient, it indicates the proportion of solar radiation penetrating through shading devices; 3) exposure coefficient, it depicts the proportion of the exposed window area in the total window area. The research calculated the shaded area, depending on the sun vector and the coordinates of windows and shadings (Figure 2.7). 39 Figure 2. 7 The procedure of calculating unshaded area. Source: Lee, D.S. et al. (2016). 2.5 Case Study: Parametric-based designs for kinetic facades to optimize daylight performance 2.5.1 The Method of Research Study This research is based on morphology to study the conceivable kinetic synthesis of façade pattern in terms of rotation and translation motion. The integration is arranged to follow comparable movement trajectory among separately movable units along with the change of time and space like cluster, scatter, propagate, and decompose. This study experiment was reduced to merely tests two basic kinetic design movements - rotation and translation (Figure 2.8 & 2.9). In the case, rotation alters the orientation of the objects through revolving each hexagon around the axes; by contrast, a translation moves each hexagon parallel to the coordinate axes of the basic plane. The movement can be controlled by setting the degree variables of changing orientation or position according to 3D axes. Figure 2. 8 The transformations of the hexagonal grid in three basic geometric motions. Source: Mahmoud, A. H. A., Elghazi, Y . (2016). 40 Figure 2. 9 The transformations of the hexagonal grid: Sliding (1 & 2), Rotation (3), Scaling (4 & 5) in Rhino interface. Source: Mahmoud, A. H. A., Elghazi, Y . (2016). The research work was comprised of two sequential steps. The first step highlights the analysis of daylighting performance on a window size of 3 m width by 1.2 m height, which is counted as a base case at the window to wall ratio (WWR) of 20%. The second step applies parametric software to presents daylighting simulation of hexagonal pattern façade to fulfill optimal effect. The base model is a side-lit office box in the area of 45 m sq. And sizes of 6.0 m width, 7.5 m depth and 3.0 m height, which is located on the first floor and south facing (Figure 2.10). The experiment merely kept changing the façade structure rather than model size. The experiment preliminarily ran simulation under clear and cloudy skies in different formations of the fabric movements. The façade of the model appears to be fabricated in a series of hexagonal panels of a constant radius of 0.5 m which can move or rotate in the spatial area (Moloney, 2011). The process of daylighting simulation is conducted by categorizing it into three cases: the base case, the kinetic rotation movement, and the kinetic translation case, and it conforms to LEED V4 daylighting standard with three illuminance levels for the floor: “daylit” (300 lux to 3000 lux), “partially daylit” (below 300 lux) and “overlit” (above 3000 lux) (Heidarinejad et al., 2014; Moschetti et al., 2015). The simulation and DIV A parameters were set to compute the daylight 41 illumination of the room and the proportion of calculation points between 300 and 3000 lux. The simulation ran uniformly in the setting of Cairo, Egypt, clear weather and separately set within each system at representative four days, March 21, June 21, September 21 and December 21, at three moments, 9 am, 12 pm, and 3 pm. The study demonstrates a dynamic process of the algorithmic and parametric method on the platform of Rhino, Grasshopper, and DIV A, respectively representing the modeling tool, parametric setting, and daylight appraisal approach (Figure 2.11). Figure 2. 10 Office Space in south orientation with dimensions of 6.0 m width, 7.5 m depth, and partially glazed height of 3.0 m on the south. Source: Mahmoud, A. H. A., Elghazi, Y . (2016). Figure 2. 11 Pattern modeling in Grasshopper shows the kinetic hexagonal patterns on the south surface. Source: Mahmoud, A. H. A., Elghazi, Y . (2016). 42 2.5.2 The Result of Research Study Simulation of daylighting performance results in the analysis of the base case and two dynamic fabric. The exported data stood for 48 points, each is on 0.9 by 0.9 by 0.9 cubic grid, indicating the numerical value of surface luminance and illuminance, then they were imported into a spreadsheet to compare the three cases. In the base case of traditional windows, almost 50 percent of the floor area fell into “partially daylit”, and the other was in “daylit” floor during summer, while almost 50 percent fell into “daylit” area with comparatively high “overlit” area, and the other was in “overlit” and “partially daylit” floor during winter. This indicates that merely half of the floor area obtained suitable daylight annually (Figure 2.12). Figure 2. 12 Daylighting simulation of the base case in terms of illuminance levels analyzed in Diva for Rhino. Source: Mahmoud, A. H. A., Elghazi, Y . (2016). In the stage of kinetic skin system, ten cases in different time propagated multiplied into 120 results to be analyzed. All the kinetic motions presented to be satisfactory in a better daylighting performance than the base case. The result indicated that the “daylit” area became larger when the “partially daylit” area comparatively lessened in summer; the “daylit” are became larger when the “over-lit” area relatively diminished in winter. The rotation motion showed the satisfactory “daylit” area, 90% of the floor area at all timings when the motion ranged from 45 to 150 deg. Specifically, 90 deg opening of the motion showed optimal “daylit” area in June while 45 and 60 deg augments the “daylit” area to nearly the whole space in March and December (Figure 2.13). The translation motion displayed the “daylit” area to be acceptable daylighting at fully and partially opening of 60% in winter. The motion increased the “daylit” area to 90% at noon (12:00 pm). However, that area was mostly lower than 75% in the early morning (9:00 am) and afternoon (3:00 pm) (Figure 2.14). Both kinetic skins worked as sun shading to contribute to 43 optimizing daylighting gain of the office space. The “overlit” area is comparatively large mainly at four hours of the day during winter. The base case showed the “overlit” at 13% of the space in winter, while the rotation motion displayed to be better at 45, 60 and 150 deg where the “overlit” area was kept below 4%. By contrast, the translation motion showed the “overlit” area below 21% at noon during winter when fully opening and decreased to 8% when partially opening. Therefore, the rotational type rather than the translational type is climatically advantageous. The research explored the influence of kinetic skins towards the daylighting performance of south-facing office spaces. The results generally display daylighting data in different façade patterns and motion types, concluding that the kinetic motion can optimize the daylighting environment. The research demonstrated a helpful methodology on daylighting simulation and analysis of kinetic facades. However, it had some limitations in terms of the daylighting environment. It only selected four specific days and three specific hours to run the daylighting simulation to compare the data, which can be merely used as a speculation of the daylighting trend although it concluded some detailed analysis of the daylighting influence of kinetic motions. A whole-hours daylighting simulation should be more accurate and persuasive as far as the interior daylighting condition although it could spend much more time running daylight simulation. Otherwise, the researchers have used two specific metrics, “point in time” illuminance measurements and “daylit” percentage, to assess the daylighting condition of the office space and compare three different cases of south facades. These metrics were integrated into daylighting simulation at several selected hours for numeric values, representing a profile of daylighting effect comparison among three cases including the base case, rotation motion, and translation motion. The results offered conceptual merits of kinetic facades compared with the static façade. As for the integrity of interior daylighting analysis, the metrics that reflect the annual cumulative index are considered to better indicate the actual daylighting performance from the viewpoint of building operations. 44 Figure 2. 13 Daylighting simulation of the rotation motion best cases in terms of illuminance levels analyzed in Diva for Rhino. Source: Mahmoud, A. H. A., Elghazi, Y . (2016). Figure 2. 14 Daylighting simulation of the translation motion best cases in terms of illuminance levels analyzed in Diva for Rhino. Source: Mahmoud, A. H. A., Elghazi, Y . (2016). 45 2.6 Case Study: Performance-based Parametric Design Exploration 2.6.1 Introduction Performance-based design, also known as performative design, is comprised of two digital design phases as follow: geometry generation and performance simulation (Oxman, 2006). The processes of generative design and performance simulation are complicated with a huge workload because they involve plenty of design parameters and complex inherent linkage. Thus, it is imperative to apply computational tools to design and analyses. The integration of generic algorithms and parametric modeling can help produce and optimize the building geometry in terms of individual performance specifics. Architects request to conceive and assess the building geometry based on the performance test approach from conceptual design to detailed design so as to propose a resolution of high-performance design (Ercan & Elias-Ozkan, 2015). 2.6.2 The workflow of the Design Case The design case involves evaluating an office building in a hot and humid climate that is considered to provide a comfortable level of daylight while preventing excessive solar heat. The building, located in Cyprus, is eight-floors office building (including basement) with a total area of 6,500 m 2 that consists of two rectangular blocks and a north-south axis narrow atrium in between. The building enclosure encompasses a flat concrete roof and hollow concrete-block masonry walls with a layer of thermal insulation and light-colored natural stone cladding on the outside. The west and east façades of office blocks were installed with strip windows and vertical shading fins. The glazed roof of the atrium was shaded with tilted photovoltaic panels (Figure 2.15). Figure 2. 15 The typical floor plan and section drawings of the office building in terms of spaces simulation based on daylight factor and solar irradiation. Source: Ercan, B., Elias-Ozkan, S. T. (2015). The multiple program tools and plug-ins used in the study were Rhino 3D, Grasshopper, Diva, and Galapagos. Rhino 3D was used to do non-uniform rational B-spline (NURBS) modeling and conduct user-based procedures. Grasshopper, as a visual programming language, is a parametric 46 tool plugged in Rhino 3D. DIV A plugged in Rhino was used to analyze daylighting together with Radiance, while Galapagos was applied to improve the shading panels. The application of digital tool kits for building performance simulation was generally composed of four phases. First of all, the organized system connected the diverse programs and plug-in form a comprehensive operation environment based on Grasshopper interface, which is built into a personalized simulation workflow (Figure 2.16). The system relied on Radiance through DIV A simulating the performance of building facade models, as well as Galapagos Evolutionary Solver to produce parametric shading devices by genetic algorithms (Figure 2.17). Then, the simulation model was generated as a simplified version in Rhino 3D to decrease the workload of the simulation process. Furthermore, the research used Grasshopper to model all the shading panels parametrically. Lastly, the study compared the results and properties of the design exploration process, then resolve the optimal parameters and configurations in order to explicate the main design determinations. Figure 2. 16 Performance-based parametric design workflow in Grasshopper interface. Source: Ercan, B., Elias- Ozkan, S. T. (2015). 47 Figure 2. 17 Simulation workflow showing GA Solver cycle in Galapagos. Source: Ercan, B., Elias-Ozkan, S. T. (2015) The research methodology is worth thinking and applying for analyzing the performance of a detailed building model or prototype. The various software tools are powerful and utilitarian to explore the design process of building envelopes; Especially, all these can be integrated into a stable and smooth operation environment that promises the project timeline. Moreover, the workflow of the performance simulation and analysis of building envelope is worth learning in order to develop a reasonable and efficient research procedure. 2.7 Parametric Computational Design Tools The design and test of the kinetic façade is a complex workflow that involves various patterns and motions of basic elements to form functional units or modules. In other words, these processes of modeling and fabricating the kinetic envelope are fundamentally operated with diverse parameters or variables. Consequently, parametric design tools are useful to mimic the dynamic mechanism in a virtual environment. Designers are able to model kinetic envelopes and simulate the indoor daylighting environment covered with them from the pre-design to shop drawings, which can help determine their ideal environmental performance. Precise simulation is a difficult assignment since it requires to perform the experiment with the correct model and tool in a real environment (Godfried, 2011). The problem is that the available tools were made to adapt to static design, for example, the material characteristic cannot be continuously changed. Yet, the kinetic façade should be analyzed in a range of variable 48 parameters for correct system conditions when assessing the real-time performance of the dynamic system (Sharaidin et al., 2012). It requires the development of effective tools to conduct the simulation process of the kinetic façade. 2.7.1 Rhino and Grasshopper Rhino 3D is a 3D computational graphics and computer-aided design (CAD) program based on the Non-Uniform Rational Basis Spline (NURBS) to make mathematical models. It has a complete plug-in system to support many other application programs to function. Grasshopper, a parametric modeling plug-in of Rhino 3D, is a visual programming tool to connect architectural designers to the complicated algorithmic structure because of its convenient use and visual availability of modeling scripting. It is a powerful tool that helps the architect model various complex building concepts based on logical generation. The modeling process within it involves numerous parameters and data to control the variations of volume, pattern, and form. Because of its parametric property, Grasshopper can be used to model kinetic envelopes to analyze its movement track and simulate the variable process in the virtual environment. 2.7.2 Ladybug & Honeybee Ladybug and Honeybee are environmental plug-ins that specialize in thermal and daylight simulation in the built environment as well as energy analysis based on Grasshopper interface. Ladybug can import standard EnergyPlus weather files (.epw) into Grasshopper to generate various 3D graphics such as sun path, radiation analysis, and shadow studies. It provides a great number of analytical charts and graphs as the reference for the design process. On the other hand, Honeybee is building energy and daylight environment analyzer that specially connects Grasshopper to five different engines like EnergyPlus, Radiance, DAYSIM, Therm, and OpenStudio. It converts architectural models into analytical models, then does the simulation work in Grasshopper environment, and export analytical data by graphics at last. Particularly, Radiance involves the analyses of daylight, glare, electric lighting and annual daylight while DAYSIM focuses on annual daylight, and Therm is used for envelope heat flow. Those engines are able to process imported data and deliver the analytical results to the parametric interface. 2.7.3 DIVA DIV A is a plug-in program for Rhino 3D, and it runs thermal, daylight, solar radiation and glare simulation. It can be straightly run in Grasshopper interface using a pre-set definition that allows data exchange between DIV A and Rhino. Rhino is used as the interface indicating the simulation results and the data visualization. DIV A is also connected to some third-party software like Radiance, Evalglare, GenCumulativeSky, and DAYSIM. 49 2.8 Daylight Performance Metrics of Indoor Space In this section, the study refers to some literature to summarize common daylight metrics that the academic study usually mentions and uses. The following table is an outline of various ever used daylight metrics in academic research. Daylight Metrics Definition Property on Kinetic Facades Daylight Factor (DF) the ratio of internal illuminance on a horizontal working plane to the unshaded, external horizontal illumination under a CIE overcast sky Not accounting for clear skies, partially cloudy, or direct sunlight. Not allow for the evaluation of dynamic façade shading. Useful Daylight Illuminance (UDI) useful illumination level for the user that is between 100 and 2000 lux Define the illuminance range in the occupied hours. View (to the outside) LEED rating system: 90 percent of regularly occupied spaces should have a direct line of sight to the outside through a vertical window that is located between 2’6” (76 cm) and 7’6” (228 cm) above the floor Occupants preferred kinetic shading state and active control. Daylight Autonomy (DA) the percentage of annual occupied hours when a minimum illuminance is met by daylight alone Binary threshold not perceived as difference by human visual sense. Continuous Daylight Autonomy (DAcon) partial credit is attributed to time steps when the daylight illuminance lies below the minimum illuminance level Soft threshold of daylighting evaluation. maximum Daylight Autonomy (DAmax) the percentage of the occupied hours when direct sunlight or exceedingly high daylight conditions are present Measure of direct sunlight and potential glary condition. Spatial Daylight Autonomy (sDA) the percentage of an analysis area that meets or surpasses a horizontal daylight illuminance threshold 300 lx for a specified fraction such as 50% of annual operation period Annual daylighting evaluation without upper limit. Annual Sunlight Exposure (ASE) the percentage of analysis points which obtains more than 1000 lx for more than 250 occupied hours annually Annual daylighting evaluation to control direct sunlight or glare. Daylight Glare Probability (DGP) the latest index used for rating glare from daylight, is represented as the proportion of occupants influenced by the glare in the daylight, ranging between 0 to 1 Glare prevention Solar Heat Gain Coefficient (SHGC). the fraction of incident solar radiation that enters into the building interior through the glazing layers, in fact, both directly transmitted and Solar heat management related to heating and cooling loads 50 diffused inward Electric Lighting Energy multiplying the lighting power density (LPD) and the floor space area Reflection of daylight utilization Table 2. 2 The outline of the daylight performance metrics. 2.9 Summary The chapter respectively explains the kinetic concept, daylight environmental factors, daylight experimentation method, and daylight analysis process, which are the professional knowledge and theory of kinetic system and daylighting environment. The review of this literature helps develop logic and rational experimental approaches to study kinetic facades in terms of the simulation and analysis of the interior daylighting condition. Especially, the control algorithm of kinetic shading state is significant to determine the post-occupied operation of kinetic facades, largely impacting the interior daylighting environment and energy consumption. Furthermore, it also refers to two differential but interrelated research cases as follows: Parametric-based designs for kinetic facades to optimize daylight performance, Performance-based Parametric Design Exploration. The first research reviews a parametric-based design methodology to optimize the daylighting of the building with kinetic envelopes. The study was actually based on a space prototype to compare two different kinetic motions of hexagonal façade patterns about their daylight performance by using DIV A based on Rhino 3D and Grasshopper environment. As a result, it provides a useful type of workflow to evaluate the daylight performance of kinetic motions. Likewise, the second research provides an effective workflow to analyze the daylight performance of a specific office building in a hot and humid climate. The method in this study is significant to reference for exploring a valid approach to improve building daylight performance. 51 CHAPTER 3 3. METHODOLOGY This chapter proposes a resolution of parametric workflow based on the control algorithm of the kinetic facade prototype of built indoor space in occupancy; it thoroughly elaborates on the workflow development in terms of daylighting effects of kinetic envelopes. Furthermore, it specifies the daylighting simulation method and the daylighting data analysis, analyzing two cases on the building performance of occupant comfort, daylighting condition, and energy efficiency. Consequently, the parametric workflow will conclude by evaluating the daylighting effects of kinetic patterns of motion with the hourly daylighting values over an entire year. 3.1 Workflow overview This section explains the methodology development and the detailed methodology, as well as its underlying principles. 3.1.1 Methodology background Daylight penetrating through the building envelope is seen as a potential energy reduction strategy since it is a renewable energy and light source at high luminous effectiveness which makes a daylighting strategy more compelling than conventional electric lighting (Konis, 2017). On the other hand, cooling loads account for a huge part of total energy consumption (14%). The data also indicates that one-third of end energy use comes from electrical lighting, and another one-third is composed of removing solar heat gains through windows (Huang & Franconi, 1999). Thus, solar energy control is an effective approach to reduce cooling loads and prevent heating loads from surpassing the cooling capacity of the HV AC system. Furthermore, fenestration and façade strategies function as transmitting adequate natural light into indoor spaces to compensate for or decrease the artificial lighting, as well as regulating solar gain and glare in terms of occupant comfort and energy efficiency. Consequently, the research on kinetic facade mechanisms involves many factors and variables to be coordinated for high daylighting performance. 3.1.2 Methodology development The workflow is based on the hypothesis that a well-designed facade using kinetic mechanism (pattern and motion, etc.) will probably be superior to the existing regular facades or static envelopes as far as indoor daylighting effects because of the proposed analysis of daylight environment variables towards indoor spaces. These variables are significant to define the indoor 52 daylight condition, occupant visual comfort, and lighting energy efficiency. However, these variables that separately impact different benchmarks of daylight evaluation often contradict mutually when the indoor space is required to achieve any one or two optimal states. For example, the direction of a particular solar ray may benefit the energy savings due to reduced indoor artificial lighting, but it could evoke much visual discomfort for occupants. As for retrofitting the interior daylighting environment of a building with a kinetic envelope, the key to determine dynamic daylighting performance is the kinetic system that balances the objectives and weighs the priority hierarchy. After ranking the variables, the design process of the kinetic envelope can be streamlined to develop the conceptual prototype, even the practical facade model, more effectively. The modeling process involves a variety of parameters that transform or transpose the kinetic patterns and motions; additionally, these parameters allow incoming light to be filtered to optimize the daylighting performance of the dynamic facade. The daylighting simulation refers to several computational tools to analyze the daylighting performance of the building indoor space covered with kinetic envelopes. Also, these tools need to set some relevant parameters about specific information like climate, location and period. The daylighting simulation result will provide reference information for the analytical objects from several dimensions as well as representative graphics such as charts and tables. Moreover, the comparison of the daylighting analysis is the process that produces an anticipated outcome between different kinetic mechanisms. The evaluation system is the cornerstone that compares the daylighting variation and effect in a logical way. The methodology involves widely used computational tools including Rhino 3D, Grasshopper, Ladybug and Honeybee in the design and analysis process of the AEC industry. The usual routine is that architectural designers use 3D modeling tools to build the digital model according to 2D drawings for the further design operation and analysis. Afterward, the other Grasshopper-based plug-ins come in to analyze the digital model using available parameters and data to emulate the real environment. The simulation results can be represented in some formats such as charts and tables. 53 3.2 Overall workflow Figure 3. 1 Overall workflow of the research methodology. The overall workflow of the thesis is generally divided into three phases (Figure 3.1). The first phase is developing a working model that involves defining digital models, clarifying shading states of kinetic facades, and analyzing daylighting variables. The second phase is processing a model test that contains control mode, algorithm application, and daylighting simulation. The final phase is collecting and analyzing the data values from the experimental simulation, also establishing a daylighting evaluation system for kinetic facades. The program tools applied in the workflow include Rhino 3D, Grasshopper, Ladybug and Honeybee, as well as Microsoft Excel. The research uses Rhino 3D to make the model geometries and loads them into Grasshopper to parametrize the elements; meanwhile, it controls the kinetic parts in the parametric environment. After that, the components related to daylighting analysis in Ladybug and Honeybee were used to calculate the hourly values and annual results of daylighting metrics like illuminance and solar heat gain according to the relevant algorithm and formula. Based on applying the control algorithm, the study collects the hourly data into Excel to process for the optimal hourly state of kinetic facades. Finally, the integration of data analysis according to the metrics priority offers an objective evaluation of daylighting effects for each kinetic façade. 3.3 Working Model The building prototype and kinetic modules are modeled in Rhino 3D and Grasshopper based on applying the parameter setting to form the object in the simulation and data analysis phases 54 (Figure 3.2). Figure 3. 2 Workflow segment of working model. 3.3.1 Model Definition The parametric working model involves three categories of parameters which are separately related to general conditions, structure configuration, and cladding system. At first, the general parameters decide the common conditions and forms of building envelope based on morphology, position, and dimension. The morphology and relative position of the surface affect its location relative to the exterior circumstance. Furthermore, the structural parameters define the structural configuration that is thought of as a frame of construction, controlling the connectivity between individual parts and affecting variations to the structural work sections. Also, the cladding parameters determine the interactivity between the façade system and the surrounding environment, influencing the form and amount of diverse energy exchange. Therefore, the process of parametric modeling on the building with a kinetic facade includes many variables such as location, grid type, and material type but not stuff like convection on the surface. However, this study only focuses on the following daylighting variables that are associated with the interior daylighting environment. 55 3.3.2 Daylighting Variables The daylighting environment of built interior space is a complex and variable situation since the process involves six different variables that determine the detailed environmental nature together. The six variables that influence the illuminance level of the daylight environment are as follows: building area, orientation, windows size, glazing type, shading device, and internal condition and exterior surroundings (Li et al., 2006). The building plan and dimension closely affect the interior daylight condition because of the rectilinear propagation and reflectance of natural light. The built area includes three factors such as plan form, dimension, and width to depth ratio which affect the daylight penetration. The plan form means a geometric shape of space that is circumscribed by walls. It could be regular geometries such as circles and polygons or irregular forms. The irregular geometric plan directs daylight transmission more complicated than the regular shape, so it is more difficult to control daylight in the irregular space. Dimension and ratio of width and depth define the spatial scale for natural light to penetrate and spread because it outlines the receptacle of daylight motion. The small space is usually relatively dark since the daylight cannot irradiate it enough through the linear emission and reflection of light whereas the large space is sufficiently bright as the natural ray can penetrate the space plan easily and bounce onto the interior surfaces many times. In terms of interior space, the ratio of width and depth is considerably influential to the room brightness. Also, the building or space orientation defines the opening position acting as a pathway where natural light enters the space. It is explicitly acknowledged that space orientation impacts the interior illuminance level. Specifically, a measurement found that in Hong Kong, spaces towards the south and east obtain more amount and longer time of daylight than the north due to high solar altitude on the south and the lopsided mention of east but not west (Li et al., 2006 & Li et al., 2000). The orientation is primarily seen as an efficacious strategy taken into account to design the interior daylighting condition due to its determinate influence as the building site, the surroundings and sun track (Phillips, 2004). Furthermore, the window size is another variable that impacts internal daylighting. Designing appropriate window size is an intricate course because it not only allows for the satisfactory view for the occupant, but controls the interior daylight condition including illuminance, solar heat gain, and glare. In building facades, the indicator of window area is defined as the window-to- wall ratio (WWR), which is formulated to be the percentage of the whole area of windows to the gross area of exterior surface (including walls and windows). The window area is closely related to energy efficiency in the building. For example, a larger window area can reduce artificial lighting energy use but increase heating or cooling load (EL-Deeb, 2013). A literature review indicates that 25% of WWR is the rational ratio in the south facade for building energy efficiency in a hot climate, and a larger window surface is mostly acceptable in a cold climate for solar rays’ penetration. Also, the values of WWR differ based on the orientation and location (Rathi, 56 2012). Generally, the reasonable range of window proportion is from 15% to 30% of the exterior wall area. The visual connection to the outdoors also affects WWR to a large extent, on which the window size is preferably between 50% and 80% for visual satisfaction (Rathi, 2012). Therefore, it can be inferred that the acceptable WWR scope for ideal occupant visual comfort is between 15 % and 80%, and the optimal range of WWR for energy efficiency is from 15% to 25 %. Aside from window dimensions, the glazing type also impacts the interior daylighting condition as it controls the transmittance of solar rays. Glazing is categorized into numerous types based on thickness, layer number, and coating (Li et al., 2006). A Solar shading system is an effective approach used on the building facades to control sunlight entry and glare. Shading devices function to reflect or diffuse the solar rays into the interior space rather than directly penetrating the indoor environment. Also, shading devices are classified into many types including fins, overhangs, louvers and light shelves. Shading devices are capable of increasing or decreasing daylight penetration depth without altering the window- to-wall ratio (WWR) (Li et al., 2006). The internal condition and external surroundings also impact the built interior daylight. Internal factors can directly alter the quality of the interior luminous environment because the need of illuminance is fully different as the interior scenario varies. First, the required illumination level differs from one to another based on the task complexity. Also, the factors such as the room dimensions, height of work plane, material, especially surface reflectance, and accumulated dirt should be defined to reflect the real daylight condition. Particularly, the height of the work plane is somewhat constant in a certain space, for example, the working plane of a computer desk is usually approximately 2’-7” (90 cm). In addition, the layout of interior furniture and equipment also influences the daylight ambiance as well as the space partition and proportion. However, the study has omitted the interior layout of furniture and equipment for quick simulation because more furniture surfaces could bounce, block and absorb the light. In terms of external surroundings, the built internal spaces are always impacted by the external obstruction which acts from two aspects: the sky condition and surrounding building features (El-Dabaa, 2016). At first, the sky condition specifically describes the extent of the daylight obstructed by the sky, indicating the strength of solar light arriving on the ground. The sky condition is determined by the weather states that can produce relevant meteorological elements such as cloud and rain to control natural light emission. Second, the shape of surrounding buildings and the distance between two buildings also impact the daylight circumstance due to the obstruction of sunlight. Commonly, the daylight arriving at the lower floor is blocked more than the higher floor, while the wider building obstructs more sunlight for the behind one than the narrower one. 57 3.3.4 Digital Models Instruction The research defines an office space as the interior environment, and the process of daylighting simulation and analysis is separated into two sequential sections. The first section emphasizes the analysis of daylighting performance in a shoebox space with three window sizes including 5 ft. (Width) *2 ft. (Height), 10 ft. (W) *4 ft. (H), and 15 ft. (W) *6 ft. (H), which are counted as the base cases with the window to wall ratio (WWR) of 5%, 20%, and 45% (Figure 3.3); afterward, the kinetic horizontal louver is assembled out of the window at 45% WWR, being capable of altering the tilt angle to adjust daylight penetration. The second section applies parametric tools and digital modeling to operate daylighting simulations of the cuboid box with adaptive skylight shading panels, and it tests the interior daylighting effect in the cubic box. Both of these two cases are concentrating on collecting the data of pertinent daylight metrics and validating data (Figure 3.4). Figure 3. 3 The shoebox model test with three window sizes. Figure 3. 4 The shoebox and cuboid box with kinetic shades. Based on the test of the cubic box with dynamic shading devices and the theory of kinetic mechanism, the research proceeds to experiment on the one floor of office building space with dynamic façades. The study hypothesizes that a kinetic facade with model-based predicted control of daylighting conditions is theoretically superior to the others and uses the digital 58 models to simulate the indoor daylighting environment in the same situation as the prototype. The data based on the valid daylight metrics are available in terms of the different facades on the identical interior space. Shoebox building model The first prototype is a simple shoebox model of a single office room, which was created in Rhino 3D and Grasshopper. The building is a side-lit office space with a floor area of 500 sq. ft. and with dimensions of 20’ wide, 25’ deep, and 10’ high with a conventional static single-layer glazed window; it is defined as the ground floor, and the only window is toward the south for longest hours of direct sunlight exposure (Figure 3.5) (Table 3.1). The window size is 15’-6” wide and 7’-9” high, and glass type of the window is uncoated single glazing that refers to the glass material library in ASHRAE Handbook (ASHRAE, 2017) because higher transmissivities would result in more light. The shoebox model is made as a normal office room in the building; thus, occupants are supposed to do normal office work in the workplace. The dimensions of the room interior are assumed to approximate the exterior dimension due to unconsidered wall thickness in the model. The interior surface is comprised of four walls (only the south wall assembled with a window), one floor, and one ceiling. As for the material parameters, walls, floor, and ceiling are defined as opaque materials with separate color reflectance of grey, black, and grey, and individual roughness values of 0.02, 0.1, and 0.02. All the material properties of these surfaces are listed in Table 3.3. A work plane is defined virtually inside of the shoebox building at the height of 2’-7” from the floor horizontally to generate analytical images in terms of parametric daylighting and solar energy simulation. The interior space is regarded as clean enough to ignore the impact of dirt and dust to daylight reflectance, and no furniture and fixtures are laid in the space also for calculating simplification. As far as the exterior environment, the sky condition depends on the weather data of Los Angeles in terms of the actual condition such as clear, overcast or raining that affects the daylight transmission and solar radiation toward the interior space. The study dismissed the surrounding objects that could block the sunlight to make the result more general and straightforward to be associated with the louver motions. Further study of the impact of freestanding landscaping and shading devices would be valuable future work. The prototype is defined to simulate the daylighting conditions and compare the simulative results in a uniform exterior condition that means the identical parameter setting for daylight factors. 59 (a) (b) Building depth (ft) 25 Width (ft) 20 Height (ft) 10 Window Width (ft) 15.5 Height (ft) 7.75 Table 3. 1 The shoebox dimensions. Cuboid building model The second prototype is a cuboid box building that has a matrix of nine open apertures as skylights on the top. The cuboid model has a floor area of 900 sq. ft. with 30’ wide and deep respectively, and 20’ high. It is built on the roof with nine circular skylights integrated with circular shading panels outside of each aperture; the diameter of both the circular aperture and shading panels is 6.5’ (Figure 3.6) (Table 3.2). Due to natural daylighting from the top, the research sets four fronts of the cuboid box to separately face toward every orientation such as north. The glass type of the skylight is also defined as uncoated single glazing, the same material as the first prototype. The circular shading panel is a dynamic part moving in both horizontal and vertical planes based on sun position. In the physical model, it is fixed on the stand of vertical axis through the circle center, and the stand is fixed on the horizontal holder and installed with two servo motors, respectively controlling rotation angle by the lower motor and tilt angle by the Figure 3. 5 (a) Isometric rendering of the shoebox; (b) Isometric view of the shoebox with dimensions. 60 upper one, but in the digital model, the study object omits the stand parts to simplify the process of modeling and daylighting calculation for higher analysis speed. (a) (b) Building Length (ft) 30 Width (ft) 30 Height (ft) 20 Skylight Diameter (ft) 6.5 Distance (ft) 3.5 Table 3. 2 The cuboid box dimensions. The cuboid box is designed as an office atrium or a typical warehouse building; thus, occupants are supposed to do normal office work in the workplace. The dimensions of the room interior can be seen to approximate the exterior size because of negligible wall thickness. The interior surface is comprised of four walls, one floor, and one ceiling with nine adaptive skylights. As for the building material, the parameter setting follows the instruction of Radiance Materials Library. To be specific, the walls, the floor and the roof (ceiling) are set as the opaque materials with separate reflectance (separately defined as white, grey, and white color), as well as separate roughness values of 0.05, 0, and 0.05. All the material properties of these surfaces are listed in Figure 3. 6 (a) Isometric rendering of the cuboid box; (b) Isometric view of the cuboid box with dimensions 61 Table 3.3. A work plane is defined virtually inside of the cuboid building at the height of 2’ 7” from the floor horizontally to generate analytical images in terms of parametric daylighting and solar energy simulation. The interior space is regarded as clean enough to ignore the impact of dirt and dust to daylight reflectance, and no furniture and fixtures are laid in the space also for calculating simplification. As for the exterior environment, the sky condition depends on Los Angeles downtown weather data for a specific condition such as clear, overcast or raining that affects the daylight penetration and solar heat gain of the cuboid interior space. The study dismisses the surrounding objects that could block the sunlight to make the result more general and straightforward to associate with the shading panel motions. Material Name R_G_B Rf/Tr Color Roughness Specularity Wall 0.5_0.5_0.5 Grey 0.02 0.05 Floor 0.29_0.26_0.23 Black 0.1 0 Ceiling 0.5_0.5_0.5 Grey 0.02 0.05 Window/Skylight 0.7_0.7_0.7 Clear - - Shades 0.7_0.7_0.7 Silver 0.1 0.05 Table 3. 3 Material parameters of the building surface. The location of the building model in the case study is defined as in the downtown area of Los Angeles, CA, United States, thus the parametric workflow imports the weather file of Los Angeles (USA_CA_Los.Angeles.Downtown.722874_CTZ2016.epw), which is the average records within consecutive five years (last modified Aug 31, 2016). According to climate data, the climate of Los Angeles (LA) is mild-to-hot and mostly dry over an entire year, cooling- dominated, classified as a Mediterranean climate (Table 3.4). The summer (ranging from March to September) in LA is dry and hot, whereas the winter is dry and mild with a few rainy times. Also, sunlight in LA is adequate for the whole year, so it is meaningful to regulate daylight for the building to improve its daylighting and energy performance. The experiment preliminarily ran daylighting and solar energy simulation in various formations of the dynamic pattern and movement under clear or cloudy skies depending on the weather data. 62 Table 3. 4 Information on the site and the climate in Los Angeles (Temperature: monthly averages in 2018) (source: Luo, K., 2018) 3.4 Parametric Simulation Test The research in this phase uses a complete model of building geometry and kinetic components to run simulations including indoor illuminance analysis and solar heat gain calculation in Grasshopper with Ladybug and Honeybee (Figure 3.7). Figure 3. 7 Workflow segment of model testing. 63 3.4.1 Kinetic Shading States Kinetic facades can be variant geometries because of the complexity of their patterns and motions. As mentioned before, kinetic motions include three basic types: rotation, scaling, and translation, even a more complicated combination. Anyhow, the movement of dynamic components can be set at a series of consecutive positions or called shading states. For example, the states of rotation are usually divided into sequential rotational angles like 15°, 30°, 45°, 60°, 75°, and 90°. These shading states basically simplify a consecutive movement into some specific positions; the method is practicable to run a computational simulation on the hourly scale because the kinetic motion is defined as one certain position or shape at each hour, the same as static facades. 3.4.2 Instruction of Façade Component The kinetic façade in the shoebox model is built as a horizontal shading device, consisting of a set of horizontal louvers installed on the outside of the south-oriented glazing layer. The shade consists of ten pieces of plastic polymer slats in the size of 15’ length and 1’ depth plus a negligible thickness. The dynamic shading slat is able to rotate 90° up (-) and down (+) from the horizontal position (0°) around the rotational axis along the back edge; the workflow used 11 tilt angles to run a simulation, including -75°, -60°, -45°, -30°, -15°, 0°, 15°, 30°, 45°, 60°, 75° (Figure 3.8). A series of dynamic shading states is defined in Grasshopper components based on the specific control algorithm which makes the shading device perform an optimal daylighting control via the comparison of different states (Figure 3.9). The south wall is embedded with a single pane glass with a transmittance of 0.7. The dynamic shading louvers covering the windows are made of plastic with a reflectance of 0.7 (defined as silver color) with roughness at 0.1 and specularity at 0.05, also its properties are listed in Table 3.3. 64 Figure 3. 8 Tilt angle of controlling the shading louver state on the shoebox model. Figure 3. 9 Workflow of setting each rotation angle to run a parametric simulation. The kinetic facade in the cuboid model consists of nine adaptive shading modules, that is, circular shading panels on top of the skylights of the roof for daylight regulation and solar heat control. Dynamic facade components function with constant parameters (e.g. panel size and material) and variable parameters (e.g. panel position and tilt angle). Each dynamic circular 65 panel is designated to be able to complete consecutive motions respectively around the horizontal axis and vertical axis; it is essentially an integrative motion mode of translation and rotation. The adaptive panel installed on the skylight moves to change its relative position and angle of the vertical plane to determine the relationship between the panel plane and solar rays, regulating the daylight flux and solar radiation. For example, in the same clear sky condition, more direct sunlight can go through the skylight when the circular panel keeps on the position parallel to the direct solar rays, leading to higher illuminance and more solar heat, whereas less direct sunlight penetrates into the interior space as the circular panel hold on its plane normal direction consistent to direct sunlight, in other words, the panel plane retains perpendicular to solar rays for blocking them. The adaptive circular panel follows the control strategy based on the sun position, namely sun altitude, and sun azimuth. As known, the sun’s altitude and azimuth vary depending on different seasons and time. Specifically, for the area on the northern hemisphere, the sun gets to its higher altitude point on June 21 st in summer and lower point on December 21 st in winter; sun moves from east in the morning to west in the afternoon during a whole day. On the other hand, the adaptive panel moves based on its rotation angle and tilt angle, independently corresponding to sun azimuth and sun altitude (Figure 3.10). According to features of the components in Ladybug and Honeybee, the dynamic control strategy in the study is that the circular panel rotates in the horizontal plane to correspond to sun azimuth that is connected to the hour during the daytime, thus the hour of the day determines a unique rotation angle, while the circular panel tilts in the vertical plane to match sun altitude that is related to the seasonal change over a whole year (Figure 3.11). Therefore, the shading state of the panel is determined by two factors: hour and tilt angle. The study applies available hours in the weather data to run the simulation, and the time range includes every hour from 7 AM to 5 PM. The shading panel rotates continuously as time goes on to always keep the vertical tilt plane in the consistent direction with sun azimuth. The title angle of shading panels is classified into nine sequential levels from 10° to 90° with an increment of 10° (10°, 20°, 30°, 40°, 50°, 60°, 70°, 80°, 90°) (Figure 3.12); the shading panel will open at the maximum to allow more sunlight going through the skylight when the tile angle increases close to 90°; in contrast, the shading panel will close at the minimum to obstruct more sunlight from penetrating the skylight when the tile angle decreases to near 0°. As regards the module number, the more adaptive shading devices the roof of the cuboid box is installed, the more complex the daylight environment the interior space will display; a large number of adaptive shading panels lead to more difficulties for organizing and processing the data efficaciously. Thus, the prototype cuboid box is designed to make nine skylights with kinetic shading panels. 66 Figure 3. 10 Tilt and rotation angle of controlling the shading panel state on the cuboid model. Figure 3. 11 Workflow of reading weather data to acquire sun position (altitude and azimuth). 67 Figure 3. 12 Workflow of setting each tile angle to run a parametric simulation The skylight aperture is embedded with a single pane glass with a transmittance of 0.7. The dynamic shading panel is made of plastic with a reflectance of 0.7 (defined as silver color) with roughness at 0.1 and specularity at 0.05, also its properties are listed in Table 3.3. It is important to define all these settings to ensure an advantageous kinetic shading system with effective daylighting alteration during both hot and cold seasons. 3.4.3 Solar Heat Gain and Daylighting Metrics 3.4.3.1 Solar Heat Gain Daylight is one primary source of solar heat gain for buildings, while indoor spaces obtain solar heat through the envelope, such as roofs, walls, and windows. The sun constantly radiates rays and heat outwards to transmit its energy, and daylight as a carrier of solar energy radiates the heat and the rays to continuously reach the ground buildings. In other words, from the perspective of solar radiation, it is difficult to separate the heat from the light because these two factors always blend and coexist to influence the surroundings. Therefore, it is inevitable to investigate the solar heat gain of the indoor daylighting environment. Solar heat gain is like a double-edged sword that can supply free heat in the winter but can also result in overheating in the summer. As solar radiation is cast on a glazing surface of a window, the resistance state of the glass surface is named the Solar Heat Gain Coefficient (SHGC). It is defined as the fraction of incident solar radiation that enters the building interior through the glazing layers, in fact, both directly transmitted and diffused inward. It is expressed as a dimensionless decimal between 0 and 1. SHGC consists of two parts based on the type of solar radiation. The SHGC of direct solar radiation is written as SHGC(Q) ，while the SHGC of diffuse solar radiation as <SHGC>D. A low 68 coefficient indicates less heat gain while a high coefficient signifies more heat gain. The nationally ranking method approved by the National Fenestration Rating Council (NFRC) is for the entire window, including the effects of the frames. Additionally, the relation between solar heat gain and SHGC depends on the climate, orientation, shading condition and other factors. Solar heat gain is usually determined by the glazing type, the number of panes and glass coatings of the glazing system. Concerning kinetic facades, also, it is impacted by dynamic mechanisms such as patterns and motions because the building interior is shaded from sunlight by partially or fully closing the facade system, while the solar ray will enter into the interior spaces when opening it. As a result, the study measures indoor solar heat gain by calculating solar radiation on a virtual surface near the fenestration with the unit of kWh. 3.4.3.2 Spatial Daylight Autonomy (sDA) In the approach named as Climate Based Daylight Modeling (CBDM) (Mardaljevic, 2006), which involves inferring the interior conditions across a whole year within the sky models based on the hourly weather data of the project location, researchers have suggested and developed several metrics to thoroughly evaluate the dynamic and variable features of a project and its performance. Daylight Autonomy (DA), initially defined by Reinhart as “The percentage of occupied times of the year when a minimum work plane illuminance threshold of 500 lx can be maintained by daylight alone”, is used to explain the proportion of the annual occupied hours in which occupants can avoid using electrical lighting by applying adequate daylight. Also, Spatial Daylight Autonomy (sDA) is interpreted as the percentage of an analysis area that meets or surpasses a horizontal daylight illuminance threshold 300 lx for a specified fraction such as 50% of the annual operation period. (IES, 2012), and its written format is using the subscript sDA300,50 %, showing as a percentage ranging from 0 to 100%. Besides, the operation period is every day from 8:00 AM to 6:00 PM, amounting to 3650 hours per year. The illuminance thresholds and performance benchmark are mainly based on the field research that is composed of gauged data and professional evaluation (Heschong, 2012). According to sDA values, two performance standards are determined. “Preferred” spaces are the analysis areas that reach or surpass sDA300, 50% with more than 75% of analysis points, while “Nominally Accepted” spaces are the satisfactory analysis areas with more than 50% of analysis points. 3.4.3.3 Annual Sunlight Exposure (ASE) The fact is that sDA cannot indicate an upper limit of daylight illuminance, thereby not being used to rate the underlying hazard of interior overexposure from solar rays. In consequence, the IES daylighting metrics committee proposed an additional metric named Annual Sunlight Exposure (ASE), which represents the probability of visual discomfort in internal environments (IES, 2012). The calculation of ASE is similar to sDA because of the same analysis points and 69 time, but it restricts the percentage of analysis points that obtain more than 1000 lx for more than 250 occupied hours annually, so its written format is expressed as ASE1000,250h. Based on this definition, the assessment of excessive solar rays entry is divided into three performance standards as follow: the daylit spaces are regarded as “unsatisfactory visual comfort” spaces when they attain to more than 10% ASE1000,250h, “nominally acceptable” spaces as they are less than 7% and “clearly acceptable” spaces as they are less than 3%. 3.4.3.4 Electric Lighting Energy Since the first incandescent electric lamp came out, electrical lighting has gone through continuous development by following the tendency of moving towards being smaller, brighter and more efficient. However, people continue to seek to use natural light to illuminate the building interior with a more effective approach. When natural lighting and indoor artificial lighting are used for illumination simultaneously, there is a complementary relation between them in building interior lighting; Lighting energy use changes in terms of the practical condition whether the natural lighting offers proper illuminance to the working plane of interior space. The more effective daylight the interior spaces obtain, the less electrical lighting energy the occupants consume. The daylight factor (DF) is defined as the proportion of exterior diffuse illuminance on an indoor horizontal working plane, while the energy requirement for lighting defines the calculation method to acquire the daylight factor in a daylight area. Indoor space can be partitioned into the daylight area and non-daylight area. Illuminance requests can be supplied through natural light in a daylight area, while the illuminance requirement is satisfied through electrical lighting rather than daylight in a non-daylight area. A daylight area is determined by the zone dimensions, window size and position (Lee et al., 2016). The electrical lighting energy amounts to the multiplication of the lighting power density (LPD) and the floor area. When calculating the electrical lighting energy for a daylit area, a determination coefficient for satisfaction or non-satisfaction of the required illuminance for the indoor working plane was used in order to determine whether the daylight area needs the electrical lighting. Electrical lighting Energy is usually counted by using energy metrics such as kilowatt-hour (kWh), Btu and Joules. The survey on electrical lighting energy use is significant because of the massive quantity. For example, in the United States, lighting energy constitutes one of the largest power uses in commercial buildings (0.78 exajoules (EJ)) (724 Trillion Btu) (EIA, 2012), which mainly occurred in daylighting time. Consequently, it is helpful to analyze artificial lighting energy to measure daylight utilization, but the thesis should not consider calculating it at the current time due to the workload on the direct daylighting effect. 3.4.3.5 Glare Glare is defined as the effect of contrast decrease in the visual scenario because of the existence 70 of shiny light sources, and it has been quantified with several indices. Interior space visual discomfort is tightly related to a glare situation. Effective daylighting largely depended on the control of daylight transmission and the prevention of glare. Studying glare during the design phase is rare because it is complicated to perceive and analyze the dynamic forms of luminance in daylight space as well as plotting out the influence of these forms for occupant comfort. Daylight Glare Index (DGI) reflects the detection of glare that is determined by the brightness of the source, the size of the source in the viewer’s eyes, the viewer’s position and the surrounding luminosity. DGI can be used to calculate the extent of visual discomfort from the glazing. Another metric of glare is Daylight Glare Probability (DGP) by Wienold and Christofferson (2006), the latest index used for rating glare from daylight, is represented as the proportion of occupants influenced by the glare of direct sunlight based on the full glare equation, ranging between 0 to 1. DGP is thought to be part of the primary climate-based daylight metrics for evaluating daylight quality. There are four criteria of glare performance: “imperceptible,” “perceptible,” “disturbing” and “intolerable glare, separately in response to three DGP thresholds of 0.35, 0.40 and 0.45 (Table 3.5). Based on field research, both of the above metrics involve complex formulas to calculate the values of the proxy. Several studies found that the DGP method has a better correlation with a human’s reaction than DGI, although there is still some discrepancy. (Jakubiec & Reinhart, 2012; Suk & Schiler, 2013). Ls: luminance of source (cd/m 2 ) Lb: background luminance (cd/m 2 ) Lw: luminance of the window in function of the relative areas of sky, obstruction and ground (cd/m 2 ) Ω: solid angle of the window (sr) ω: solid angle of the source, modified in function of the line of sight (sr). EV: vertical eye illuminance (lux) Ls: luminance of source (cd/m 2 ) ωs: solid angle of source (sr) P: position index 71 Table 3. 5 Glare comfort criteria based on DGP. Source: Tabadkani et al., 2018. 3.4.3.6 View The window view is a key factor in building interior daylighting performance because it is the occupant’s access to connect their indoor environment to outdoor surroundings. Also, the content of the window view connection to the outdoors helps occupants keep psychological well-being, and the identical discovery is that occupants prefer natural views to artificial ones (Farley & Veitch, 2001). In a study related to a view out and visual requirements, providing a minimum transparency ratio of 20% in the daylit space is suggested in order to supply optimum view out and occupant well-being (Keighley, 1973). The view is influenced by many structure and construction factors that consist of window size or aperture configuration, the distance of occupants from windows, provision of multiple views, view content, visual transparency and openness factor, and visual clarity. Also, the LEED rating system furthers the supply of a view connected to the exterior. The system states that 90 percent of regularly occupied spaces should be met with a straight linear view toward outside though a perpendicular window that is 2’6” (76 cm) and 7’6” (228 cm) higher than the floor. The credit acknowledges that the view access to outside is highly recommended as an advantage of a window, which is technically challenging to satisfying the occupant’s thermal and visual comfort along the space perimeter. Because of its complexity, it is more complicated to analyze the view than just running daylight simulation, so the thesis should not involve simulating it at the current time. Still, it could be a possible factor considered in future work. 3.4.4 Parametric workflow description In this study, a conceptual parametric workflow was created in Grasshopper to offer a basic method to generate the control strategy of kinetic facades based on the principle of model-based predictive control. The entire workflow is comprised of several sections and uses the available components in Ladybug and Honeybee (Figure 3.13). 72 Figure 3. 13 Graphic representation of the parametric workflow and its components. 1. Establishing building type parametric inputs The workflow introduces the building model and façade configuration with parameters which the architectural designer only needs to concern, including dimensions, material properties, structure features, etc. Anything else seen in the workflow, such as the simulation engine (e.g. Radiance) setting in daylighting and thermal simulation, was created to provide full transparency and control to the designer. 2. The process of transferring the Rhino geometry to Honeybee The workflow was specifically created to accommodate all the geometries so the architect would simply connect the B-rep or parametric design components to start the workflow. The proposed building prototype is modeled in Rhino and translated into Grasshopper by assigning the model to a Boundary Representation (B-rep). This step demonstrates the interoperability between modeling and parameterization. Additionally, it is also available to build the geometry in Grasshopper by parametric components. After parametrization, the B-rep geometry is converted into an energy model with thermal zones. In this step, the components include “Split Building Mass to Floors”, “Split Floor to Thermal Zones”, “Masses to Zones”, “Solve Adjacencies”, and “add HB Glazing”. The figures illustrate the forefront input as a Closed B-rep and the resultant output as the same Closed B-rep with Honeybee (HB) zones (Figure 3.14 & 3.15). 73 Figure 3. 14 The components for translating Rhino geometry into HB zone as B-rep. Figure 3. 15 The shoebox model demonstration as a closed B-rep HB zone with glazing. 3. Creating the building performance scenario for daylight and thermal simulation using specific parameters This step introduces experimental or practical data as the parameters for daylighting variables to generate the building performance sketch. This portion introduces the exterior environment including sunlight and weather conditions by using the components “import EPW” and “Sun Path”, also separates the whole HB zone into HB surfaces for assigning material properties to the building by applying the main components “Decompose By Type”, “Radiance Material”, and 74 “Create HB surfaces”. These surfaces decomposed from B-rep were recollected to form a new B- rep with the material property that determines the daylight and thermal exchange. The process provides the model with reliable data to simulate environmental performance in a virtual situation (Figure 3.16 & 3.17). Figure 3. 16 The components for introducing sunlight and weather conditions. Figure 3. 17 The components for assigning HB surfaces with specific material properties. 75 4. Creating shading for thermal and daylight simulation The workflow creates shading devices for the building model automatically, and additional shading devices can be added by the user. This part involves defining the shading element and the shading material to form the kinetic shading mechanism, being comprised of the component “Shading Designer”, “Shading Material”, “Radiance Material”, and “Create HB surfaces”. The definition based on the experimental data from Radiance Materials presents confidence on the imported shading parameters (Figure 3.18). Figure 3. 18 The components for creating kinetic shading device and assigning material property. 5. Simulating the model with Los Angeles climate data In the workflow, this part is a pivotal feature that enables parametric setting with climate data to simulate both the daylight and thermal environment by using the identical model in the same tool for the full interoperability of data. The main components for daylighting illuminance include “Radiance Parameters”, “Grid Based Simulation”, and “Run Daylight Simulation” which meet the condition of daylight metrics and simulate the indoor daylighting illuminance on the grid- based work plane with 48 test points as light sensors above the floor, and calculate the indoor average illuminance for data collection. The main components for the glare index involve “Radiance Parameters”, “Image Based Simulation”, and “Run Daylight Simulation” which calculate the glare probability using the DGP formula based on a digital occupant’s view in the space center. The primary components for solar heat gain contain “Generate Cumulative Sky Matrix”, “Select Sky Matrix”, and “Radiation Analysis” which calculate the solar radiation on the grid-based surface with 18 test points as thermal sensors inside of the fenestration. According to the property of each daylight or thermal metric, the three parts respectively simulate the real daylighting and thermal environment to calculate the values on the basis of relevant theoretical 76 formulas (Figure 3.19 - 3.21). 6. Processing the simulation results by applying the thermal and daylight metrics These components enable the workflow to obtain a daylighting schedule from the model to be used for the daylighting and thermal simulation results, allowing for an evaluation basis for a daylighting control schedule for daylighting performance and energy saving. Figure 3. 19 The components for daylighting illuminance simulation. Figure 3. 20 The components for glare index simulation. 77 Figure 3. 21 The components for solar heat gain simulation. 7. Running iterative parametric simulations This part is an essential feature for speeding up data collection within the early stages of design (Figure 3.22). Figure 3. 22 The components for iterative parametric simulation. 78 8. Setting up results to be visualized through Rhino interface This part is a helpful feature for interoperability between Rhino and Grasshopper for the resulting visualization of environmental simulation (Figure 3.23 - 3.25). Figure 3. 23 The example of visualizing daylighting illuminance simulation. Figure 3. 24 The example of visualizing daylight glare probability simulation. 79 Figure 3. 25 The example of visualizing solar heat gain simulation. 3.4.5 Experimentation Process The study process is targeted at the integration of algorithm-based control and daylighting data statistics. The macroscopic view lies in building a database consisting of 4380 tables (roughly equivalent to diurnal hours) (Figure 3.26). The daylighting data of all hours are technically stored in the database, meaning that daylight metrics information of each hour can be searched in terms of kinetic states if it pointing to detailed given time. As far as each table of hourly data, it contains specific hours at the vertical column and different shading states at the horizontal row, so the intersections are daylight metrics values (Table 3.6). Therefore, the table can export a determinate value of daylight metrics when it proposes a shading state at a certain hour (Figure 3.27). Usually, there are several ranges of daylight metrics value to define disparate levels of daylighting performance, so the expected threshold results in a rational recommendation for the dynamic shading state at a given hour (Figure 3.28). 80 Figure 3. 26 The database - look-up table diagram. Prototype Illuminance Level on Sept. 21 st , Clear Sky (equinox) Annotation: the value uses lux as the unit, and it is an arithmetic average of interior illuminance. Table 3. 6 Illuminance level of cuboid model based on time and tilt angle. 81 Figure 3. 27 Illuminance level grid of cuboid model at a different time and panel tile angles. The daylighting database forms the fundamental of one parametric algorithm – Model Predicted Control (MPC). The control algorithm of kinetic facades on the building is interpreted as follows (Figure 3.29): at first, the system requires all daylighting database comprised of daylighting tables to move the shading devices according to the sun position based on solar azimuth and solar altitude. The algorithmic control starts with inputting basic environmental information including date, hour, and target metrics level, for example, the required illuminance value is around 500 lux for the indoor office space on September 21 st under a clear sky. Then the system controller can search the database to acquire the range of acceptable shading states (e.g. tilt angle of shading panels at 20-30 degree). After that, the controller can select one specific shading states based on other factors such as heating or cooling need and occupant control. The command of the shading state will be sent to the actuator (e.g. DC motors) after the completion of system judgment. The building management system conduct to collect the data of the actual effect to validate the control mode by using physical sensors. At last, the system can readjust the shading device state one hour later to cycle the control process. 82 Figure 3. 28 Illuminance curves of cuboid model at different time. Figure 3. 29 Model-based predictive control algorithm diagram. 3.4.6 Iterative Test Procedure The façades of the prototypes are fabricated respectively from a regular glazed window and two sets of kinetic shading devices with disparate mechanisms that can translate or rotate in the three- dimensional area. The process of daylighting simulation and analysis is conducted by categorizing it into three cases including the baseline case, the static case, and the kinetic case, and it conforms with the criteria of daylighting metrics. The simulation approach and parameters are defined to calculate the daylight illuminance of the indoor space and the proportion of calculation points between 300 and 3000 lux over a range of hours. The study can run some 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 10° 20° 30° 40° 50° 60° 70° 80° 90° Average Lux of Work plane Panel Tilt Angle Illuminance Chart (Sept. 21st, clear sky) 8:00 AM 12:00 PM 5:00 PM 100-2000 Lux 83 batches of daylighting simulations consistently in location setting and weather condition, as well as a period of diurnal hours of consecutive days over an entire year (4380 hours) for all two types of façades since the dynamic facade always changes at each hour or in the hourly interval, and the experiment analysis allows for comparable data between two kinetic facades. 3.4.7 Software Environment Support The experiment primarily aims to assess the property of kinetic façades by comparing the disparate motions reacting to real-time daylighting conditions and analyzing the motions at the optimum states for efficacious daylight regulation. For that purpose, the study demonstrates a dynamic course consisting of the algorithmic control and parametric method based on the platform of Rhino, Grasshopper, and Ladybug & Honeybee, respectively representing the digital modeling interface, parametric setting application, and daylighting evaluating approach. Especially, Honeybee can be interfaced with some environmental simulation engines such as DAYSIM, Radiance, and EnergyPlus by which the tool is enabled to analyze the daylighting condition and solar heat gain. The simulation data are shown in the format of the analytical charts and tables, being consistent with daylighting metrics. These tools are integrated into a toolkit in the uniform operating environment to do daylighting analysis and simulation depending on the integral system with superior compatibility and instant feedback. As mentioned, the research is based on some cubic boxes with different window types and facade structures to simulate the daylighting condition of the interior space. The digital models of cubic boxes are simply built in Rhino 3D, then they are input into Grasshopper interface through the “B-rep” (boundary representation) component. During this period, composing an appropriate component kit of daylighting simulation in Grasshopper needs to wire the components of specific functionality and rationalize the intrinsic program logic. Due to the specialty of individual metric, the experiment allows for creating five branches of exclusive component combinations except that all the parametric component kits share the identical parameters of internal state such as the dimensions of the model and the external context like weather data and time. Because of the daylight metrics study, every component kit is able to generate a corresponding set of data that shows the daylighting metric values as well as the analytical charts and tables that demonstrate daylight distribution on detailed indices. The chart based on the work plane in the cubic box is a visualized approach that explicitly represents the level of daylight metrics. Also, the component kits support to export data as a format of the numerical spreadsheet to process the daylighting data in another analytical tool such as Microsoft Excel and comparing the results explicitly. Basically, the process of producing all the daylighting data over one entire year usually needs simulating many rounds because of the workload of whole year hourly analysis, so 84 the workflow would process numerous variables to produce a huge mass of data. Therefore, the research requires to arrange the data in a systematic approach for efficient data extraction and analysis. In the research, the available data can be converted into a “CSV” file in order to be imported into Microsoft Excel for data analysis and process. 3.5 Data Analysis Approach The phase focuses on analyzing the daylighting data of different facades and comparing daylighting results of building interior space based on the daylight metrics, then verifies the practical daylighting improvement of the kinetic facades. By studying kinetic patterns and motions, the design process of kinetic façades can perform to optimize building interior daylighting environment. Daylight simulation operated by specialized tools can contribute valuable data and results to the facade design process, which presents an essential reference to alter the forms of kinetic mechanism, improving the building envelopes to be responsive and adaptable to different weather, season, and time. As a result, architectural designers can get access to the simulative data of the kinetic façades operation process and predict the daylighting effect at the early design decision-making stage (Figure 3.30). Figure 3. 30 Workflow segment of analysis approach. Data analysis is a process that involves comparison among several groups of data results based on standardized codes and empirical values. According to the previously proposed daylighting metrics, the metrics performance criteria are defined by some professional organizations such as 85 IES to precisely weighing the building interior daylighting environment. Based on the MPC principle, the analysis process can literally compare the metrics resulting value with the desired value to generate a control strategy by following the predefined priority among all selected metrics. Then it means that the data analysis actually results in the kinetic facade with the control algorithm to compare with other facades (either dynamic or static) according to the resulting data of metrics. Also, the evaluation method is elaborated to score the daylighting performance in terms of these specific metrics. However, as for comprehensively evaluating an interior daylight condition enclosed with kinetic envelopes, it is complicated but significant for kinetic facade design to establish an integrative evaluation system for daylighting and thermal environment performance. The research focuses on a parametric method of developing and evaluating kinetic facades initially based on the priority of five daylighting or thermal metrics, as it concerns the evaluation criteria for kinetic facades. During the simulation process, the numerical values of five metrics are calculated for each hour as the base of assessing daylighting performance. The optimum state of kinetic facades at a certain hour is weighed by considering the trade-off among these five factors in the analytical scope. Afterward, systematic analysis can finally determine the manipulation of kinetic facades. The study takes into account glare as the first rank to impact the state of kinetic facades because glare, especially direct solar rays, objectively causes severe discomfort for the occupant thus reduces working productivity. The kinetic façade is required to prevent direct solar rays from showing in the occupant’s sight because of the destructive effect of glare. The glare index (DGP) can be calculated in Grasshopper with Ladybug and Honeybee. The priority between daylight quality (illuminance level) and energy performance is not a constant issue, which means either of these two aspects can be thought about behind glare. The daylight illuminance is an index that reflects indoor daylight quality. sDA and ASE are two indices based on similar principles using the percentage of the proper area to indicate illuminance situation; sDA can be weighed before ASE because sDA reflects illuminance threshold in a longer time range than ASE. Furthermore, these two indices are closely related to artificial lighting use. Specifically, the more usable daylight illuminance there is in the indoor space, the less electric lighting energy the occupant needs. Generally, the ideal indoor illuminance is between 500 lux and 1500 lux, whereas solar heat gain should be maximally utilized during the heating season and effectively avoided in the cooling season. The control criteria of solar heat gain follow the instruction from the ASHRAE standard by comparing the hourly outdoor temperature with the thermal comfort temperature range to determine whether the solar heat is needed or not. Based on the simulation data, the evaluation system for kinetic facades works to grade a group of façades in daytime hours depending on the scoring criteria and the priority of these metrics. The score of each façade is calculated through the comparison of their values. The sum of the hourly 86 score indicates the control performance of kinetic façades in terms of the indoor daylighting and thermal environment. 3.6 Daylighting simulation algorithmic definition of five metrics The research in the initial phase was conducted to analyze the interior daylighting of the shoebox model assembled with regular static windows based on five selected daylight and thermal metrics. The process can be seen as an equivalent approach where the dynamic shading motion of kinetic facades is converted into the static states with different sizes of windows. This step of the experiment aims to verify the simulation results of daylight metrics in order to compare the simulation data among different conditions or states. Each calculation of the daylight metrics refers to a special workflow segment of parametric simulation that encompasses various coding components in Grasshopper plus Ladybug and Honeybee (Figure 3.31). During the initial stage, the research obtained some referable data from the baseline prototype, the shoebox model with a regular static single-glazed window. The simulation result varies to indicate some regularity when different sizes of windows are assigned to the south façade of the shoebox model. In conclusion, the values of sDA, ASE and solar heat gain increase much higher along with the expansion of the window size; DGP numbers indicate that the larger size window generates slightly more visual discomfort for the occupants within interior space; Electric lighting energy on three static windows shows an equivalent value that could be a limited problem of the component definition to investigate (Table 3.7 & 3.8). Figure 3. 31 Parametric simulation workflow for sDA, ASE, glare, solar heat gain, and electric lighting 87 Table 3. 7 The parametric simulation of daylighting metrics on the shoebox with different window sizes Window Size sDA ASE DGP Solar Heat Gain Electrical Lighting Energy (W)x(H) inch kWh/ft.2 kWh/ft.2 Static Glazing 5x2 8.40% 6% 0.29 14.74/147.40 38.31 Static Glazing 10x4 41.60% 22% 0.31 58.94/589.42 38.31 Static Glazing 15x6 70.40% 35% 0.37 132.61/1326.11 38.31 Table 3. 8 The resulting data of daylight metrics on different window sizes 3.7 Summary This chapter proposes the research methodology with its complete workflow of the simulation and analysis. The research focuses on studying regularity between kinetic facades and interior daylighting and thermal environment for investigating a practical approach to promote the daylight condition through developing an efficacious kinetic mechanism. This chapter describes the entire workflow that defines kinetic models and simulates the interior environment and analyzes daylighting and solar heat data for building interior daylighting performance of kinetic facades. The workflow introduces five essential daylighting and thermal metrics to rate the daylit spaces by using parametric design tools. These metrics are used as the evaluation criteria that 88 differentiate daylighting performance results and assess the superiority of diverse kinetic facade cases. Also, the control algorithm, model-based predictive control, is applied to the kinetic façade for deciding the kinetic patterns of motion or shading states based on simulation data. Two kinetic facade prototypes, including the shoebox model and the cuboid model, are defined to test the validation of the computational simulation programs and connectivity between the daylighting criteria and the analytical model. The baseline case of the shoebox model, a rectangular room with regular static windows, helps to establish the analytical model and verify the feasibility of the experimental procedure. After the test, the method of the baseline model can be applied to two groups of building models with different kinetic patterns of mechanism for collecting the simulation data. Meantime, the parametric workflow (including parametric modeling and simulation) has to be upgraded or updated to adapt to the new dynamic daylighting environment. The analysis approach aims to initiate a scientific and effective evaluation system of façade performance in terms of the interior daylight analysis and judgement. The evaluation system is tested and applied by comparing three forms of facades on the identical building model. Once the research institutes an evaluation system with grading criteria, the workflow of the methodology can constructively instruct architectural designers about the developing process of dynamic facades, particularly for the project during the early design stage of kinetic facades. 89 CHAPTER 4 4. CASE STUDY RESULTS The chapter reports the results of two cases of kinetic facades by using the parametric workflow presented in chapter 3 for the daylighting test and validation. The report specifies the experiment process which encompasses the brief model description, the daylight parameter setting, as well as the daylighting and thermal simulation. The study applied modeling tools in Rhino 3D and parametric components in Grasshopper integrated with Ladybug and honeybee; it also processed many rounds of daylighting simulation on the model for the data collection of hourly values and annual data in terms of daylighting metrics. 4.1 Overview The chapter demonstrates two building prototypes to test the effectiveness of the proposed parametric workflow for kinetic facades simulation and evaluation. Two prototype models are presented as follows: 1) a shoebox model of an office room with one single-layer glazed window and horizontal shades outside on the south orientation, and 2) a cuboid model of typical atrium space with nine skylight apertures installed with circular panels on the top. This chapter concisely introduces the two models as the case studies, then presents simulation results in data and data processing results. The chapter is written in the corresponding order with the last chapter for readability. 4.2 Introduction of the simulation process The section illustrates the basic background settings and workflow definition of the parametric simulation in terms of the daylighting and thermal environment in the case study. The building model in the case study is simple, so it not only reduces the workload on the daylighting and solar energy calculation and related data processing but also facilitates the generation of kinetic control strategy and the comparison of different façade patterns. However, the simplified model is not created as a virtual counterpart of a real building (e.g. neglected thickness of building components) and follows an ideal and theoretical setting in terms of daylighting variables, so it cannot fully reflect the real environment due to the more complicated parameter for the daylight variables. 90 4.2.1 Background introduction of parametric simulation The case study is merely used to demonstrate the proposed workflow where the prototype is not a real building, and also simulating the whole year daylighting and solar heat conditions of the interior is a complicated calculation and time-consuming process. Therefore, the study selected three representative days to conduct an hourly simulation and analysis. By checking the weather condition of Los Angeles based on climate data, three dates including June 21 st , September 21 st , and December 21 st were chosen to separately represent the sun altitude and the seasons of summer and winter because the sun altitude depends on the earth’s position on the revolution track, or known as the season of the year. Also, the hottest day in summer shows near the period between August and September, whereas the coldest day generally appears approximately from December to early January (Table 4.1). The time of hourly simulation for built interior space ranges from 7:00 AM to 5:00 PM respectively on these three dates. Table 4. 1 Information of the site and the climate in Los Angeles (Temperature: monthly averages in 2018) (source: Luo, K., 2018) In this case, the research has separately investigated the dynamic shading louver on the shoebox model and the responsive skylight shading panels on the cuboid model on the basis of three metrics that includes indoor average illuminance, daylight glare probability (DGP), and solar heat gain. Because of its complexity and iteration, the parametric workflow features three branches of the component definition of simulation to correspond to these daylight and thermal metric calculations based on the identical model definition and environment setting. 91 4.2.2 Specification of daylighting and solar heat simulation Based on the climate pattern in Los Angeles, the parametric workflow for analyzing the kinetic facades conducts daylighting and solar radiation simulation respectively on three dates, June 21 st , September 21 st , and December 21 st , separately standing for the hot season and mild (or cool) season over the year. Because of the component property and analytical principle in the interface of Ladybug and Honeybee, daylight illuminance is calculated by doing a grid-based daylighting simulation throughout the work plane at 2’-7” above the floor (Figure 4.1), and the glare index (DGP) is measured by running an image-based daylighting simulation through a virtual occupant’s south-oriented view at 4’-6” above the floor (Figure 4.2), also the value of solar heat gain is calculated by analyzing solar radiation on a grid-based surface near the fenestration (2’ to 3’ away from it) based on hourly solar heat cumulation (Figure 4.3). Respectively simulating the shoebox model and the cuboid model, the parametric workflows for indoor illuminance are defined as a grid-based simulation with a 6x8 (shoebox) and a 7x7 (cuboid) grid of test points on the work plane, and those test points are used as daylight sensors to measure the illuminance level of each grid. The Honeybee component “RADParameters” (setting Radiance parameters for ambient light simulation) involves defining the simulation rule that apparently decides the calculation amount of light transmission thus impacts simulation hours to a large extent. Also, the part of glare calculation is defined as an image-based simulation where a digital person facing south is seated in the central point of the indoor space as the face is used as a light sensor. This part involves the setting of the component “RADParameters” as well, and the script on the component “Glare Analysis” (using evalglare for glare calculations) can eventually output both DGP and DGI values, but this research only refers to the DGP value. Besides, the partial workflow for solar radiation is using the components from Ladybug rather than Honeybee applied in the previous two parts, and the “Radiation Analysis” (calculating the radiation falling on) is able to simulate the solar energy exchange by being integrated with other environmental parameters. The overall parametric workflows of two kinetic models are illustrated below, and the component definitions of the two cases have a few differences due to the different façade patterns (Figure 4.4). The result data of diverse daylight or thermal metrics for the interior daylight effects evaluation of kinetic facade are produced from standalone computational logic depending on the arithmetic algorithm of these metrics. The calculation results of hourly indoor average illuminance, daylight glare probability (DGP), and solar heat gain for the three days are summarized in the spreadsheet of Microsoft Excel. Especially, the seasonal difference in daylight hours results in varied simulation hours on three days since the summer season has a longer daytime and the winter has a shorter one. However, the daylighting simulation was conducted with the same hour range (7 AM to 5 PM) for consistent comparison. 92 Figure 4. 1 Daylight illuminance test points on the work plane of the two models. Figure 4. 2 The orientation of the occupant’s view in the two models. Figure 4. 3 Solar radiation test points on the interior surface of the two models. 93 Figure 4. 4 The parametric workflow of daylighting and thermal simulation. 4.3 Results and Data of the case studies This section separately describes two mentioned cases with different kinetic facades in terms of the simulation results and the related data. As for the three selected metrics, a set of tables and charts are listed to illustrate the daylighting and thermal condition of diverse shading states on the shoebox model (Table 4.2 - 4.10) (Figure 4.5 - 4.13) and the cuboid model (Table 4.11 - 4.19) (Figure 4.14 - 4.22). 4.3.1 Results and Data of shoebox building model On June 21 st , the indoor illuminance through the daytime hours (7 AM to 5 PM) appears to reach the maximum at 2 PM. Besides, another lower peak shows at 10 AM, and the value goes down not much until 12 PM, then goes up. The louver rotation angle for easiest daylight penetration is -45° (down) tilt where illuminance curve almost takes up the highest position in the chart whereas the louver at 75° (up) tilt angle obtains the least illuminance. The trend of glare index (DGP) arrives at a lower peak at 10 AM then falls down with a different extent, and reaches the highest point at 2 PM, but the difference is that -15° (down) tilt curve is on the top of all the other curves. The chart of solar heat gain displays a similar curve trend as the indoor illuminance. The maximum point of solar radiation shows at 2 PM but another lower peak arises at around 11 AM. Typically, the louver at -45° (down) tilt provides the interior space with the most solar radiation. On September 21 st , the indoor illuminance through the daytime hours (7 AM to 5 PM) demonstrates that the illuminance level at -60° (down) tilt is plotted out of the range of UDI from 94 before 9 AM to past 3 PM, being much larger than that of the other tilt angles. Most of the tilt angles reach the maximal point at 12 PM except that the state of -45° (down) tilt gets to its peak at 10 AM, and it is the highest curve in the normal scope of illuminance level. Meanwhile, the louver at 75° (up) tilt angle stays at the lowest level of illuminance. The curves of the glare index (DGP) appears to be regular, reaching the top point at 12 PM, but some portion of the tilt angle at -15° (down) or -30° (down) tilt angle receives the largest DGP value at different hours, unlike the illuminance value. The chart of solar heat gain looks irregular because fluctuations occur on three rotation angles including -30° (down), -45° (down), and -60° (down). Particularly, the louver at -45° (down) rotation obtains the most solar heat among these tilts at 1 PM while the louver at -60° (down) gets to its maximum at 2 PM. On December 21 st , the indoor illuminance through the daytime hours (7 AM to 5 PM) shows that the largest illuminance is given by -30° (down) tilt angle almost during the whole day, and the highest portion lasts between 11 AM and 12 AM. Some irregular fluctuations exist in the curves of -60° (down)up, -45° (down)up, -15° (down)up, and 0° tilts, making the illuminance trend complex. In contrast, the curves of glare index (DGP) manifest a regular variation that the maximum glare arises at 12 PM, and -15° (down) tilt generates the most glare issue over the daytime, while the louver at -30° (down) and 0° tilts result in disturbing condition between 10 AM and 3 PM. In the chart of solar heat gain, the curves present regularity to some extent though several curves fluctuate much. The louver at -45° (down) tilt gets the maximal solar heat at 1 PM, and a similar trend happens on -30° (down) and -60° (down) tilts. The opposite trend of the curve shows on -15° (down), 0°, and 15° (up) tilt angles, and they reach a relative nadir at 1 PM. 4.3.2 Results and data of cuboid building model On June 21 st , the peak of indoor illuminance through the daytime (7 AM to 5 PM) shows up around 2 PM, and before that point, another lower vertex appears at 10 AM. The indoor illuminance value is generally increasing as the tilt angle of the shading panel gradually gets larger from 10° to 90° with more openness of the skylight. The glare index (DGP) arrives at a lower peak around 10 AM and drops down a little around 11 AM, then gets to the maximum point at 2 PM. The index of solar heat gain shows a similar tendency but the maximum value around 2 PM is much higher than the before peak at 11 AM. The sunlight illumination and radiation stay at a relatively low level in the early morning (7 AM and 8 AM) and late afternoon (5 PM). On September 21 st , the peak of indoor illuminance through the daytime (7 AM to 5 PM) shows up at noon (11 AM and 12 PM), and the curves of value variation appear regular hump pattern. The indoor illuminance value is generally increasing as the tilt angle of the shading panel gradually gets larger from 10° to 90° with more openness of the skylight. The glare index (DGP) 95 arrives at the highest peak at 12 PM and displays slight slop on the curves. The chart of solar heat gain shows a similar tendency as the illuminance but the maximum value around 12 PM and 1 PM. The sunlight illumination and radiation stay at a relatively low level in the early morning (7 and 8 AM) and late afternoon (5 PM). On December 21 st , the apex of indoor illuminance through the daytime (7 AM to 5 PM) displays at noon (12 PM), and the value gently drops until 3 PM then keeps at the flat level till 4 PM. The indoor illuminance value is generally increasing as the tilt angle of the shading panel gradually gets larger from 10° to 90° with more openness of the skylight. The glare index (DGP) arrives at the highest peak at 12 PM and displays slight slop from 9 AM to 4 PM on the curves. The image of solar heat gain shows a similar tendency as the illuminance but the maximum value delays a little around 1 PM. The daylight intensity and solar radiation are very weak in the early morning (7 and 8 AM) and late afternoon (5 PM). The charts of indoor average illuminance indicate that the bigger tilt angle of the adaptive panels leads to higher indoor average illuminance and vice versa. Useful Daylight Illuminance (UDI) ranges from 100 lux to 2000 lux, so the dynamic skylight shades should avoid placing the tilt angle at 70°, 80° and 90° between around 12 PM and 3 PM on June 21 st , and at 80° and 90°between around 10 AM and 2 PM on September 21 st , and the illuminances at 80° and 90° around 10 AM on June 21 st are slightly higher than the high boundary of UDI, while all the other tilt angles over the whole day are useful. Most of the illuminances at 10° tilt angle on June 21 st and September 21 st and December 21 st are below 100 lux, not helpful to indoor daylighting. Also, the illuminance at 7 AM on December 21 st is not available. As for normal work in the office space, the ideal illuminance value is 500 lux. Therefore, advisable tilt angles can be selected from the tabular data to decide the optimal effect of indoor illuminance. In contrast, the indoor illuminance cannot be regulated to meet required light condition when the roof skylights are assembled with conventional static shading panels such as panel tilt angle at 60°. The illuminance level fluctuates mostly higher than required light condition for the normal office work (500 lux) as time goes on, though many of the values belong to the range of UDI. Specifically, most of illuminances are higher than 1000 lux and even some value exceed high boundary of UDI (2000 lux) in hot season, while most of illuminances are between 500 and 1000 lux except for some below 500 lux in mild season. The charts of daylight glare probability show that the bigger tilt angle of the adaptive panels results in a higher probability of glare, and vice versa. According to the performance criteria of DGP, all of the values fall into the range of “imperceptible” glare, thus the uncomfortable sunlight from the skylight is not a conspicuous influence on indoor occupants on the three days. The DGP value reaches the maximum point at around 12 PM on September 21 st and December 21 st , respectively smooth increase in the morning and decrease in the afternoon, but on June 21 st 96 the DGP value rises up to a peak at 10 AM then falls down slightly and climbs up again to the apex at 2 PM. In comparison, most of the DGP values with the tilt angle at 60° are in the range between 0.09 and 0.26, which is below the imperceptible threshold 0.35; the lighting condition is acceptable. Thus, the selection of a tilt angle should not be affected by the glare issue. The charts of solar heat gain signal that the larger tilt angle of the adaptive panels causes the larger solar radiation penetration, namely more solar heat gain in the indoor space, vice versa. The indoor thermal comfort for occupants is affected by the amount of solar heat gain depending on the season during a year or the hour within a day. The total solar radiation reaches the apex at about 12 PM on September 21 st and at 1 PM on December 21 st ; it separately goes up before noon and goes down in the afternoon. The curve on June 21 st appears different than the other, indicating that the solar heat gain rises up to a peak at 11 AM then falls down slightly, and climbs up again to the maximum at around 2 PM. Generally, the climate feature in Los Angeles is dry and hot with cooling dominant. The control strategy of adaptive skylight panels infers that the lower tilt angle reduces solar radiation coming in the indoor environment during the hot season (June 21 st and September 21 st ), while the higher tilt angle increases solar heat gain during the mild season (December 21 st ) to adjust the indoor temperature to thermal comfort range. The solar heat gain falls in the range between 0.21 and 18.87 kWh on September 21 and between 0.14 and 22.77 kWh on June 21 st when the indoor space needs cooling. Thus, the excessive solar heat will cause an uncomfortable thermal environment. The solar heat gain falls in the range between 0.04 and 3.31 kWh on December 21 when heating improves the indoor thermal environment. 4.4 Control strategy decision The standard of determining a control strategy is based on the control algorithm and the simulation results, counting on the climate pattern of different seasons or periods over a year, and also the weather conditions during a single day. The control strategy mainly aims to make an indoor space to meet the occupant’s demand of visual comfort by adequately supplying indoor daylight illuminance and avoiding the glare problem caused by direct sunlight penetration in the view; meanwhile, it anticipates to prevent undue solar radiation from coming into the interior space so that it can promise the indoor thermal comfort and reduce the energy use of mechanical systems in the building. Thus, the control strategy of kinetic facades is determined based on making trade-offs among these three metrics, including indoor illuminance, daylight glare probability (DGP), and solar heat gain. The desired illuminance values indoors are usually different among various space functions and diverse occupant activities. For example, illumination at 250 lux can meet the need for easy office work and classwork, whereas the illuminance for normal office work is recommended as 500 lux. The required value climbs up as the accuracy and complexity of the work increases. 97 Supermarkets or mechanical shops are advised to achieve 750 lux of lighting condition; normal drawing workshops, detailed mechanical workshops, and operation theaters are supposed to get 1000 lux; For detailed drawing work and very detailed mechanical work, the lighting condition is recommended to be equal to or high than 1500 lux. The illuminance requirement of some types of working activities is outlined below (Table 4.20). The case study aims at a normal office space, it thus considers the illuminance range between 500 lux and 1500 lux as the opportune threshold for the workplace to control the kinetic façade. Therefore, the control algorithm for indoor daylight illuminance is defined to adjust the shading device state to offer 500 to 1500 lux of illuminance value. Table 4. 20 Recommended light level for working activities. A glare issue usually arises in the built interior space when the direct solar rays enter the window in the occupant’s view. Because of that, the simplest and most direct way to reduce sun glare is to use the kinetic façade shades to obstruct the direct sunlight by placing them in a certain position or state. As mentioned, the glare issue is defined to be a higher priority to other daylighting factors in the case study. That is, as glare appears, the scenario of the kinetic façade has to be altered eventually, though the daylighting illuminance is satisfactory to the working activity of occupants under that façade shading state. According to glare performance criteria, the DGP Activity Illuminance (lx, lumen/m2) Public areas with dark surroundings 20 - 50 Simple orientation for short visits 50 - 100 Areas with traffic and corridors 100 Working areas where visual tasks are only occasionally performed 100 - 150 Warehouses, homes, theaters, archives, loading bays 150 Coffee break room, technical facilities, ball-mill areas, pulp plants, waiting rooms 200 Easy office work 250 Classrooms 300 Normal office work, PC work, study library, groceries, show rooms, laboratories, check-out areas, kitchens, auditoriums 500 Supermarkets, mechanical workshops, office landscapes 750 Normal drawing work, detailed mechanical workshops, operation theaters 1000 Detailed drawing work, very detailed mechanical works, electronic workshops, testing and adjustments 1500 - 2000 Performance of visual tasks of low contrast and very small size for prolonged periods of time 2000 - 5000 Performance of very prolonged and exacting visual tasks 5000 - 10000 98 value of imperceptible glare is not higher than 0.35. The ideal shading state to control kinetic facades should be making the DGP value fall in the range of imperceptible case. Furthermore, the worst case of glare control is keeping the value under 0.45 which is the threshold of disturbing glare. Solar radiation produces positive or negative effects on the indoor thermal environment according to a specific season or period of one year, but it can be controlled rationally to maintain a thermally comfortable indoor space on the basis of outdoor ambient temperature and indoor set-point temperature. Theoretically, according to ASHRAE standards, a building with south-oriented windows can be determined about the necessity of solar radiation to ensure indoor thermal comfort by looking at the outdoor ambient temperature and the indoor thermal comfort temperature (summer season: 75.1 – 80.1 ℉, winter season: 68.5 – 75.7 ℉). Thus, four cases of indoor temperature range and corresponsive adjustment appear as follows: (1) the outdoor ambient temperature below the low threshold of indoor comfort temperature (75.1 ℉) in the summer; (2) the outdoor temperature in the indoor comfort range or above the high threshold of it (80.1 ℉) in the summer; (3) the outdoor temperature above the high threshold of indoor comfort temperature (75.7 ℉) in the winter; (4) the outdoor temperature in the indoor comfort range or below the low threshold of it (68.5 ℉) in the winter. The control criteria to decide the appropriate operation scenario for dynamic facade shading devices are summarized as: the shading state in the summer meets that, DGP values of the interior space for glare issue could be requested to be equal or lower than 0.35 for guaranteeing no apparent direct sunlight (lower than 0.40 in the worst case), then the fenestration offers illuminance values of daylighting between 500 lux and 1500 lux, finally, the kinetic facades control the solar radiation going through to transmit minimal solar heat; the shading state in the winter meets that, DGP values of the interior space for glare issue should be requested to be equal or lower than 0.35 for guaranteeing no apparent direct sunlight, then the fenestration offers illuminance values of daylighting between 500 lux and 1500 lux, lastly the dynamic envelopes control the solar radiation going through to transmit maximal solar heat (Table 4.21). 99 Date Control Criteria June 21 st (Cooling Season) The interior DGP values no more than 0.35 Indoor daylight illuminance between 500 lux and 1500 lux Minimize solar heat gain Sept 21 st (Cooling Season) The interior DGP values no more than 0.35 Indoor daylight illuminance between 500 lux and 1500 lux Minimize solar heat gain Dec 21 st (Heating Season) The interior DGP values no more than 0.35 Indoor daylight illuminance between 500 lux and 1500 lux Maximize solar heat gain Table 4. 21 Control strategy criteria on June 21 st , Sept 21 st , Dec 21 st . After data collection for three representative days, a group of specific tilt angles at daytime hours are calculated to form a control strategy by complying with the priority from glare through illuminance to solar heat gain. The process of parametric simulation figures out a series of tilt angles along with hourly intervals during daytime including June 21 st , September 21 st , and December 21 st . Based on the mentioned rule of control criteria, the study filtered all the tilt angles of every hour in accordance with corresponding daylight performance to acquire an optimal value for the shading state at each hour. Therefore, the course that picks up tilt angles will lead to a whole group of hourly tilt angles during the daytime; the tilt angle results of three dates are respectively listed in tabular format (Table 4.22 - 4.24, 4.28 - 4.30 & 4.31 - 4.33). Meanwhile, the values of three daylight metrics for the static window or shading and simple kinetic shading are generated through the daylight simulation and organized in these tables as well. Furthermore, the evaluation system for façades or shading devices is designed with a logical method to score the indoor lighting and thermal performance based on three daylight metrics. Typically, the façade fabric that meets the required threshold of the daylight indices will be scored one point in the evaluation system. Especially for the item of solar heat gain, the score is given to some certain façade structure through the comparison of the values depending on the seasonal and hourly thermal need for heating or cooling. Generally, in the course of control decision, the determination of the priority among several selected metrics are flexible, depending on the occupant’s need and design consideration to better meet specific demand of the architectural design. This study only used the proposed priority to show how the designer can conduct selecting shading states for the control algorithm. 4.5 Description of scoring process The scoring process concerns assessing several discrepant facade objects of the shoebox model and the cuboid model. Specifically, the facade comparison involves kinetic louver, fixed louver 100 (0° tilt), and static glazing on the shoebox model versus kinetic shading panels, tilt-only shading panels and static shading panels (60° tilt) on the cuboid model. The evaluation of the facades is concluded based on the determination of scoring criterion that is working to reflect the superiority and merit among different façade systems. The scoring process sums up the points separately corresponding to three daylight metrics to record total scores. The study measures the glare issue by using the criterion of three thresholds of DGP value including “≤ 0.35” (imperceptible), “≤ 0.40” (perceptible), and “≤ 0.45” (disturbing). The DGP value that falls in the range of imperceptible glare at a certain hour will be scored 1 point, whereas the case gives 0.5 points toward the perceptible situation and 0 points toward disturbing conditions. Additionally, the initial instruction of acceptable indoor average illuminance for office space is the range between 500 lux and 1500 lux. However, for the purpose of precisely controlling illuminance level based on the occupant activity, the study assumes normal office work to the indoor space, referring to 500 lux as the benchmark to score the facades about indoor illuminance. Specifically, the facade case with an indoor illuminance value out of the required range will not score; besides, the facade case being closer to the threshold 500 lux will score 1 point if all the illuminance values of different facades fit into the acceptable indoor illuminance range, and the facade secondly close to the threshold will score 0.5 points when the façade cases are more than two. Lastly, the scoring for solar heat gain depends on the seasonal and hourly thermal need for heating or cooling. The interior scenario with the kinetic or static facade that gains less total solar radiation during the summer (June 21 st and September 21 st ), or that obtains more solar radiation during the winter (December 21 st ) will be recorded 1 point with a gradient of 0.5 kWh, for example, 0 – 0.500, 0.501 – 1.000, 1.001 – 1.500, and so forth, that is, the cases in the same rank will be given the same point, and the second place will be given 0.5 point when the comparison happens with more than two façade cases. On the basis of the scoring criteria for three aspects, the evaluation system can add up the hourly points throughout the whole day to calculate the total scores for each façade fabric. The scoring results of two case studies on three different days are listed below (Table 4.25 - 4.27 & 4.34 - 4.36). Consequently, the kinetic facades generally score higher than the other facades (Table 4.37 & 4.38) (Figure 4.23 & 4.24). The scoring criterion in the study is considered as a reference to demonstrate the operational process. In fact, the scoring criterion of the evaluation system is not constant but parametric. The scoring criteria can be determined by the designer by considering the scale and range of data as well as the code related to interior lighting and thermal condition, in order to make a unified scoring system. 101 Daylighting Performance Score Façade Type June 21st Sept. 21st Dec. 21st Kinetic Louver 27.5 27.5 19 Fixed Louver 20.5 26 12.5 Static Glazing 18.5 7.5 13 Table 4. 37 Daylighting performance scores of three cases of the shoebox model on June 21 st , Sept. 21 st , Dec.21 st . Figure 4. 23 Daylighting performance scores of three cases of the shoebox model on June 21 st , Sept. 21 st , Dec.21 st . Daylighting Performance Score Façade Type June 21st Sept. 21st Dec.21st Kinetic Shading Panel 29 28.5 19.5 Tilt-only Shading Panel 23.5 25 21 Static Shading Panel 12 12 16.5 Table 4. 38 Daylighting performance scores of three cases of the cuboid model on June 21 st , Sept. 21 st , Dec.21 st . Figure 4. 24 Daylighting performance scores of three cases of the cuboid model on June 21 st , Sept. 21 st , Dec. 21 st . 0 5 10 15 20 25 30 June 21st Sept. 21st Dec. 21st Score Daylighting Performance Evaluation Kinetic Louvers Static Louvers Static Glazing 0 5 10 15 20 25 30 June 21st Sept. 21st Dec. 21st Score Daylighting Performance Evaluation Kinetic Shading Panel Tilt-only Shading Panel Static Shading Panel 102 4.6 Summary The chapter demonstrates the process and operability of the workflow proposal using two case studies to evaluate the daylight performance of kinetic facades. The case study shows the applicable parametric workflow in Grasshopper plus Ladybug and Honeybee. The workflow generated several simulation datasets with respect to three days of different seasons by categorizing the variables including shading state and hour. Based on model-based predictive control, the processes of data collection and data analysis conclude the theoretical control strategy to optimize the daylight performance according to the priority of glare, indoor illuminance, and solar heat gain. The case studies respectively defined two prototypes, the shoebox building model with kinetic horizontal shading louver and the cuboid building model with adaptive skylight shading panels. The parameters in Grasshopper operate the shading state or position of the kinetic facades. Particularly, the tilt angle is adjusted in the shoebox model whereas the tilt angle and rotation angle are controlled in the cuboid model for daylighting performance simulation. On the basis of climate pattern in Los Angeles, the research involved three days, June 21 st , September 21 st , and December 21 st , to represent cool and hot seasons for analyzing the indoor daylighting condition based on three daylight metrics including glare (DGP), indoor illuminance, and solar heat gain. Hourly daylighting conditions (7 AM to 5 PM) were simulated depending on a series of shading states or positions of two dynamic facades. According to the priority of three daylight metrics, the optimum state at every hour was determined as a control strategy of model-based predictive control (MPC) to regulate the façade motion for improved performance compared with the baseline model. The evaluation criteria were proposed to score daylight performance among kinetic facades models and conventional static models, demonstrating the kinetic façade’s contribution to the indoor daylight environment optimization. 103 CHAPTER 5 5. ANALYSIS AND EV ALUATION The research methodology proposed a parametric workflow to design an intelligent control system of kinetic facades that can be adaptive or responsive to the external surroundings. By using a parametric definition, the workflow is activated to simulate the interior daylighting environment of kinetic facades with simulation engines like Radiance in terms of the performance of direct glare, indoor illuminance, and solar radiation energy, and deduces the appropriate control strategy of kinetic motions according to the computational algorithm of model-based predictive control. From the global perspective, this workflow functions normally, but the objectivity of the metrics priority for control decision and scoring criteria for evaluation is still questionable. However, the workflow offers a systematic method as a reference to develop kinetic facades. Two case studies were conducted to examine the applicability of the workflow for informing the control of kinetic facades, with the objective of improving on the performance of static exterior façade shading systems and dynamic shading systems that use simple dynamic control algorithms (e.g. solar tracking). Furthermore, although this research was carried out in simplified models, it is feasible to apply this workflow to a more complicated configuration because of the common kinetic features and daylighting variables. The analytical results of the case studies depended on internal and external factors such as climate, location, building structure, surrounding condition and façade configuration, etc. The approach was prototypical to resolve the design and control issue of kinetic facades. In addition to the time limitation, the research thus merely chose two representative prototypes to carry out the daylight analysis and discussion to demonstrate the process of a parametric method during the early design phase. The determination of dynamic control strategy is based on the predefined conditions in the case study, for example, building geometry, façade configuration, location, climate reference, surface material, etc. The setting of the various parameters can be modified to fit other cases or situations. The research method is feasible and applicable to other building geometries, glazing patterns, and façade configurations. During the early design stage, the workflow can be used to generate the kinetic control strategies and judge the potential profit and deficiency of a kinetic façade in terms of indoor daylighting and energy environment. The chapter discusses the results of the evaluation process and scoring system and evaluates the parametric workflow for simulating daylight effects of kinetic facades from several significant aspects. 104 5.1 Resulting analysis of the shoebox model with kinetic louver Based on the shoebox model, the section separately introduces the process of data analysis on daylighting illuminance, daylight glare probability (DGP), and solar heat gain according to the outdoor environment including sun position, exterior daylight, and solar energy, and it presents a complete result analysis on three typical days including June 21 st , September 21 st , and December 21 st . The variation of daylight effects is closely relevant to the weather condition, the sun position at a certain time, the façade geometric property, and the facade shading state. 5.1.1 Resulting analysis of indoor daylighting illuminance On June 21 st , the indoor illuminance through the simulation time (7 AM to 5 PM) rises to a lower crest at 10 AM and lowers down slightly until 12 PM, and after that, it arrives at the peak at 2 PM then goes down till sunset. The order of the illuminance curves of all tilt angles is somewhat irregular as the kinetic louver rotates from down to up. The illuminance curve at a tilt angle of - 45° (down) almost produces the highest position in the chart, whereas the case of 75° (up) tile angle obtains the least illuminances. During the early morning (7 AM and 8 AM), indoor daylight illuminance appears to stay at a low level. One cause for the pattern of the indoor illuminance on this day is the daylight pattern of the weather condition. The hourly level of global horizontal illuminance on June 21 st reflects its fluctuation trend that is similar to the indoor situation described above. The weather condition is that a haze could usually happen on the morning of June (from approximately May to October due to either onshore migration of cloud cover). Also, the pattern of the horizontal louver is good at blocking solar rays from the top like the sun at noon during summer. The result shows that tilt angles between -15° (down) and -65° (down) are beneficial to sunlight penetration on this day. The sun at 10 AM is at a relatively low altitude to transmit more direct illuminance from the east side into the room through the south window because of the availability of the direct sunlight, and the louver components cannot simply prevent the direct sunlight without daylighting loss. The sun altitude at 11 AM and 12 PM increases to a higher level in the sky and thus it is hard for sunlight to pass through the window due to the direction of direct sunlight; besides, the global horizontal illuminance lowers down near 10000 lux. As a result, the louver modules can easily block the direct sunlight entry. After that moment, the sun goes down gradually in the afternoon, and the exterior horizontal illuminance becomes larger to reach the maximum at 2 PM, so the sunlight goes through the window more easily at a high direct illuminance. Later until 5 PM, both of the sun altitudes and the global horizontal illuminance lower down, resulting in the reduction of indoor illuminance level. Especially at 3 PM, the global horizontal illuminance (94369 lux) is slightly lower than that at 2 PM (99612 lux), but the gap of the influenced indoor illuminance is relatively large because the sun goes 105 down a little to the west so that solar rays penetrate the window to mainly hit the east side of the indoor space (Figure 5.1 & 5.2). Figure 5. 1 Hourly outdoor horizontal illuminances on June 21 st . Figure 5. 2 Hourly sun altitude on June 21 st . 0 20000 40000 60000 80000 100000 120000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Illuminance (lux) Outdoor Horizontal Illuminance_June 21st Global HI Diffuse HI Direct NI 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Degree Sun Altitude_June 21st 106 On September 21 st , the indoor illuminance through the simulation time (7 AM to 5 PM) goes up to the apex at 12 PM except that 45° (up) tilt gets its maximal point at 10 AM, and it is the highest curve in the normal scope of illuminance level. The illuminance level at -60° (down) tilt appears out of the UDI range after 8 AM till near 4 PM, much larger than those of the other tilt angles. Meanwhile, the illuminance at 75° (up) tilt angle stays at the lowest level of all the illuminance curves. As the louver tilt changes from down to up, the overall level of indoor illuminance is lined up to reduce except the curve at -75° (down) tilt. In the early morning (7 AM) and the late afternoon (5 PM), indoor daylight illuminance is also available but appears relatively low. The trend of the illuminance curves is consistent with the weather condition on September 21 st , namely the variation of the global horizontal illuminance on this day. Although the horizontal louver mainly functions to obstruct the sunlight from the top, the sun is at a medium altitude to emit its rays, so it is beneficial for the direct sunlight to come in the interior space. Especially, the tilt angles at between -15° (down) and -60° (down) fit the sunlight travel orientation into the room. The sun position at 10 AM is comparatively low, and the sunlight direction matches the louver at -45° (down) tilt well to allow more sunlight through the window from the east side, then the illuminance value falls at 11 PM because of the higher sun. The sun rises to the highest point in the sky at 12 PM, also the exterior horizontal illuminance increases to the peak, so the indoor illuminance reaches the maximum at this point despite the increased height of the sun. Generally, the indoor illuminance stays at a similar level from 11 AM to 1 PM due to the small differences of the sun altitude and the global horizontal illuminance at those hours. After 1 PM, the sun goes down gradually with a decrease of the exterior horizontal illuminance till sunset. Thus, the indoor illuminance level falls linearly (Figure 5.3 & 5.4). 107 Figure 5. 3 Hourly outdoor horizontal illuminances on September 21 st . Figure 5. 4 Hourly sun altitude on September 21 st . On December 21 st , the indoor illuminance through the simulation time (7 AM to 5 PM) shows discrepant trends on each tilt angle. In the early morning (7 AM) and the late afternoon (5 PM), indoor daylight illuminance is also nearly unavailable for the indoor space. The largest illuminances are offered by -30° (down) tilt angle almost over the whole daytime, and the top portion of those lasts from 11 AM to 12 AM, showing a rough parabolic curve. A similar trend 0 20000 40000 60000 80000 100000 120000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Illuminance (Lux) Outdoor Horizontal Illuminance_Sep. 21st Global HI Diffuse HI Direct NI 0.00 10.00 20.00 30.00 40.00 50.00 60.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Degree Sun Altitude_Sep. 21st 108 happens on -75° (down) tilt. The irregular fluctuation runs in the illuminance curves at -60° (down), -45° (down), -15° (down), and 0° tilts with different peak time and complicates the illuminance trend. The other tilts perform similarly to get to the maximum at 3 PM. The trend of global horizontal illuminance appears as a parabolic curve, displaying an increase in the morning and a decrease in the afternoon along with the maximal point at 12 PM. The louver states with the tilt angles between -70° (down) and 0° favor daylight penetration indoors although some of them fluctuate unsteadily, including -45° (down), -15° (down) and 0°. The reason for this unsteadiness is that the light propagation is controlled by the angle between the direct solar rays and the position of tilted louver slats due to rectilinear propagation of light. Typically, the tilt angle at -45° (down) apparently reduces the direct daylight illuminance at 1 PM, while the -15° (down) tilt prominently obstructs the direct sunlight penetration at 12 PM and 2 PM, as well as the 0° tilt at 11 AM, 12 AM, and 2 PM but being helpful at 3 PM due to the lower sun altitude. The tilt at -30° (down) gets to the peak of the indoor illuminance with two approximate values at 11 AM and 12 AM, which is a little ahead of the appearance of the highest global horizontal illuminance at 12 PM since the sun position plays a significant role in daylight penetration, the same as the exterior illuminance. The sun goes up slightly higher at 12 PM than at 11 AM. Likewise, the same principle works on -60° (down) tilt at 2 PM, showing that indoor illuminance is the largest at that hour with slightly lower exterior illuminance and sun altitude (Figure 5.5 & 5.6). Figure 5. 5 Hourly outdoor horizontal illuminances on December 21 st . 0 10000 20000 30000 40000 50000 60000 70000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Illuminance (lux) Outdoor Horizontal Illuminance_Dec. 21st 109 Figure 5. 6 Hourly sun altitude on December 21 st . The indoor illuminance comparison of June 21 st , September 21 st , and December 21 st basically indicates that the highest illuminance level happens on December 21 st , whereas the lowest level appears on June 21 st during the period of adequate daylight supply outdoors. By contrast, the comparison of the outdoor illuminance level is an inverse result. The main reasons for this phenomenon are that a lower sun altitude facilitates the direct sunlight spread with less shade formed on the facade, and it is hard for the louver to block the direct sunlight with a tilt of low absolute value, so a larger amount of the sunlight gets through the window to hit the work plane inside, and vice versa (Table 5.1). In summary, the pattern of the indoor illuminance variation is correlated with either the exterior environment like sun altitude and weather conditions or the geometric pattern of the façade. 0 5 10 15 20 25 30 35 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Degree Sun Altitude_Dec. 21st 110 Table 5. 1 Grid illuminance of the shoebox model with kinetic louver on three typical days. 5.1.2 Resulting analysis of daylight glare probability On June 21 st , the glare index (DGP), viewed facing south, through the simulation time (7 AM to 5 PM) increases to a lower peak at 10 AM then falls in different extents at 12 PM, and reaches the maximal point at 2 PM then goes down until sunset. The glare curve at -15° (down) tilt is on the top of all the glare index curves. The remaining curves of either up or down tilts are lined up alternately from top to bottom in the chart. The overall situation of glare is acceptable by the occupant due to the DGP maximum at near 0.36. The main reason is that sun altitudes are high from 9 AM to 3 PM, and sun position is either in the east in the early morning or the west in the late afternoon, so direct sunlight cannot obtain a right angle and an opportune timing to penetrate through the window to hit the occupant's view in the interior space. The glare probability reduces a little at noon with the shading states of partially open louver because of the sun’s high altitude. Additionally, the kinetic louver components work to shade the interior space against the sun glare. The result proves that tilt angles at between 15° (up) and -30° (down) raise the chance of glare generation on this day. On September 21 st , the glare index (DGP) through the simulation time (7 AM to 5 PM) appears to be a series of irregular parabolas, reaching the maximal point at 12 PM. A range of -15° (down) and -30° (down) tilt angles receive the largest DGP values at different hours. Likewise, all the illuminance curves are distributed alternately from 0° tilt on. As the sun path changes to a medium level, the glare issue appears worse than that in June, so the direct sunlight can pass through the window to appear in the occupant’s view sometime. The sun glare is perceptible from 9 AM to 3 PM, especially, the glare becomes disturbing between 10 AM and 1 PM with some specific tilt angles of the horizontal louver. However, the kinetic louver components are able to adjust to an appropriate state to stop or reduce the glare issue. The data demonstrates that 111 the kinetic louver should avoid tilt angles at -15° (down) and -30° (down) from 10 AM to 3 PM and at 0° from 10 AM to 1 PM. On December 21 st , the glare index (DGP) through the simulation time (7 AM to 5 PM) also manifests a group of irregular parabolic variations, and the maximal probability of glare arises at 12 PM. During the early morning (7 AM) and the late afternoon (5 PM), the glare issue fully disappears from the indoor space. The distribution of illuminance curves demonstrates an alternate pattern. Typically, -15° (down) tilt louver generates the most glare issue during the daytime, especially after 9 AM until 3 PM, and the -30° (down) and 0° tilt result in disturbing glare conditions from before10 AM to around 3 PM. The sun path is located at the lowest level of the whole year, and it is the easiest way that sun glare goes into the room to interfere with the occupant. The sun glare could generate interference from 9 AM to 3 PM. Typically, the glare achieves a disturbing level between 10 AM and 2 PM at some specific tilt angles. Thus, the glare control strategy should focus on avoiding -30° (down), -15° (down), and 0° between 10 AM and 3 PM based on the glare data, and the other tilts should be considered according to the other two metrics. The glare comparison among June 21 st , September 21 st , and December 21 st basically indicates that the higher glare probability is on December 21 st , whereas the lower chance appears on June 21 st during the period of adequate sunlight outdoors (from 9 AM to 4 PM). The main cause for this phenomenon could be that the lower sun altitude favors direct sunlight penetration with less shadow formed, and it is not easy for the louver to stop the direct sunlight with a tilt of low absolute value, so more direct sunlight gets through the window to reach the occupant’s sight indoors (Table 5.2). In summary, the change of the glare issue relates to either the outdoor environment like sun altitude and weather conditions or the geometric pattern of the façade. 112 Table 5. 2 DGP and interior renderings based on the occupant’s view and sun position scenarios. 113 5.1.3 Resulting analysis of solar heat gain On June 21 st , solar heat gain goes up to a lower peak at 11 AM then goes down a little at 12 PM, and it arrives at the maximum at 2 PM before falling until 5 PM. The louver at -45° (down) tilt angle provides the interior space with most solar radiation. The overall solar heat curves are alternately lined up from top to bottom in the chart. The chart of solar heat gain displays a similar curve trend to the indoor illuminance. During the early morning (7 AM and 8 AM), solar radiation is available but keeps relatively low. One reason for the pattern of the solar heat gain on this day is the solar radiation effect of the weather condition. The hourly level of global horizontal radiation on June 21 st reflects its fluctuation trend that is similar to the indoor case described above. The other reason is that solar heat penetration usually decreases as the sun hits the higher altitude for the south facades. The sun at 10 AM is at a relatively low altitude to radiate more direct solar energy from the east side into the indoor space through the south window, and the louver modules cannot merely stop the direct solar radiation because of daylighting loss. Meantime, the global horizontal radiation reaches a lower peak. However, the factual peak of indoor solar heat gain in the morning arises at 11 AM since it could be the time delay of solar heat conduction. The sun at 11 AM and 12 PM rises to a higher altitude in the sky and thus it is hard for direct solar heat to pass through the window due to the orientation of direct sunlight; furthermore, the global horizontal radiation falls down a little. Therefore, the louver modules can easily block the solar energy entry. Afterward, the sun goes down gradually in the afternoon, but the exterior horizontal radiation goes up to the higher peak at 2 PM, so solar heat penetrates through the window more easily at a high direct solar radiation. Later until 5 PM, both of the sun altitudes and the global horizontal radiation decrease to lead to the fall of solar heat gain. The result shows that tilt angles between 0°and -60° (down) are detrimental to heat insulation on this day. Also, the other tilt states from 30° (up) to 75° (up) primarily help prevent solar radiation from heating the indoor space (Figure 5.2 & 5.7). 114 Figure 5. 7 Hourly outdoor horizontal radiation on June 21 st . On September 21 st , solar heat gain through the simulation hours (7 AM to 5 PM) shows to change irregularly because fluctuations occur on three tilt angles including -30° (down), -45° (down), and -60° (down). Particularly, the louver state at -45° (down) tilt reaches the summit of solar heat at 1 PM among these tilt angles, while -60° (down) gets to its maximum at 2 PM. The values of solar heat gain at 60° (up) and 75° (up) tilts are totally the same to stay at the lowest level. Generally, the louver states with a tilt angle at -60° (down) and -45° (down), as well as - 30° (down) favor solar heat gain indoors, though their waves are unstable. Meanwhile, the global horizontal radiation appears as a parabolic curve during the simulation hours. Therefore, the analysis concludes that the fluctuations at these three tilts are caused by the correlation change between the direction of solar rays and the passage of tilted louver slats. Typically, the tilt at -45° (down) conspicuously reduces the direct solar radiation from 10 AM to 11 AM and at 3 PM, while the tilt at -60° (down) prominently obstructs the direct solar radiation at 12 PM, as well as the tilt at -30° (down) at 10 AM, 1 PM, and 4 PM. Additionally, the other tilt states mainly work to shade the space against solar radiation (Figure 5.4 & 5.8). 0 100 200 300 400 500 600 700 800 900 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Wh/m 2 Outdoor Horizontal Radiation_June 21st Global HR Diffuse HR Direct NR 115 Figure 5. 8 Hourly outdoor horizontal radiation on September 21 st . On December 21 st , solar heat gain through the simulation hours (7 AM to 5 PM) presents regularity to some extent, though several illuminance curves fluctuate a great deal. The louver state at -45° (down) tilt gets to the maximal point of solar heat gain at 1 PM, and a similar trend happens on -30° (down) tilts as well. By contrast, the opposite trend of the solar heat curves displays on -15° (down), 0°, and 15° (up) tilt, and the curves reach a relative nadir at 1 PM. The values of solar heat gain at 60° (up) and 75° (up) tilts are the same. Generally, the louver tilt angles between -45° (down) and 0° promote the solar radiation to be transmitted through the window even though some tilt angles result in an unsteady variation of the solar heat gain. In contrast, the tendency of global horizontal radiation is a regular parabolic curve over the simulation period. Thus, the louver tilt angle largely impacts the solar radiation through the window in the case of a low sun altitude integrated with the corresponding azimuth. Specifically, the louver tilt angles at -45° (down), -30° (down) and -60° (down) helps direct solar rays radiate the indoor space at 1 PM while the louver tilts at -15° (down), 0°, and 15° (up) decreases the effect of direct solar radiation, but they work to facilitate that effect at 10 AM, 11 AM, and 3 PM. Besides, the other tilt states majorly function to shade the space from solar heat gain (Figure 5.6 & 5.9). 0 100 200 300 400 500 600 700 800 900 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Wh/m 2 Outdoor Horizontal Radiation_Sep. 21st Global HR Diffuse HR Direct NR 116 Figure 5. 9 Hourly outdoor horizontal radiation on December 21 st . To compare solar heat gain among June 21 st , September 21 st , and December 21 st , it basically shows that a higher solar heat gain is generated on December 21 st , whereas a lower level appears on June 21 st during the period of adequate sunlight outdoors. However, the level of global horizontal radiation is the opposite. The main reason for this result could be that a lower sun altitude helps solar radiation be transmitted with less shade formed on the facade, and it is hard for the louver to prevent the direct solar radiation with a tilt of low absolute value, so more amount of solar rays penetrates the indoor space and vice versa. Consequently, the variation pattern of the solar heat gain is determined by either the outdoor environment like sun altitude and weather conditions or the geometric pattern of the façade. 5.1.4 Daylighting effects comparison of the facades on the shoebox model By comparison, a kinetic louver based on the intelligent control strategy performs better on the indoor daylight effects than the other façade patterns through the three typical days (June 21 st , Sep. 21 st , and Dec. 21 st ). (Table 5.3 - 5.11) (Figure 5.10 - 5.18) The kinetic louver shows superior performance on indoor illuminance because it is able to regulate the direct and diffuse sunlight penetration in real-time to make the indoor space achieve or get close to a required illuminance level for a specific working activity (e.g. 500 lux for normal office work). The static louver plays a role in the second position, functioning better in 0 100 200 300 400 500 600 700 800 900 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Wh/m 2 Outdoor Horizontal Radiation_Dec. 21st Global HR Diffuse HR Direct NR 117 the summer and worse in the winter due to the sun altitude change. The static glazing without any shading device definitely is ranked in the last position. As for the glare index, the kinetic louver is able to operate a real-time control to avoid or reduce the direct solar rays that reach the occupant’s view. It generally performs slightly better with relatively low value but near glare probabilities with the others on June 21 st and September 21 st , except that the static glazing works badly sometimes. However, it appears vastly superior to the other facades on December 21 st since its glare index remains in a tolerable range at daytime, whereas those of the other facades go beyond the threshold of the disturbing level. Concerning the solar heat gain, the static louver performs slightly better on preventing solar heat than the kinetic louver on June 21 st except for 2 PM and on September 21 st , apart from the worst performance of the static glazing. By contrast, the static glazing results in the best performance on December 21 st by introducing the most solar radiation to heat the indoor space. Also, the kinetic louver lets less solar radiation come into the room than the static louver. The reason for the bad performance of kinetic louver on solar heat gain is that the proposed method for control decides the glare prevention as the first priority and obstruct the solar radiation at the same time. The study also evaluates the daylighting effects of three different facades by introducing a scoring method to establish an evaluation system on the basis of integrated consideration. The result shows that the kinetic louver obtains the highest score on three separate days, whereas the static glazing is graded the lowest point in summer, and the tilt-only scores the lowest point in winter. According to the conclusion from the three dates, it can be inferred that the kinetic louver is estimated to be the best regulator or strategy for the indoor daylighting performance over the year when compared to the specific static louver strategy. 5.2 Resulting analysis of the cuboid model with kinetic shading panels Based on the cuboid model, this section separately introduces the process of data analysis on daylighting illuminance, daylight glare probability (DGP), and solar heat gain according to the outdoor environment including sun position and exterior daylight and solar energy, and it presents a complete result analysis on three typical days including June 21 st , September 21 st , and December 21 st . The change of daylight effects is closely connected to the weather condition, the sun position at a time, the façade geometric property, and the facade shading state. 5.2.1 Resulting analysis of daylighting illuminance On June 21 st , the indoor illuminance through the simulation hours (7 AM to 5 PM) shows a peak around 2 PM, and before that point, another lower vertex appears at 10 AM. The indoor illuminance value is generally increasing as the tilt angle of the shading panel gradually gets larger from 10° to 90° with the growing openness of the skylight. In the early morning (7 AM), 118 indoor daylight illuminance appears to stay low. The pattern of the indoor illuminance on this day is consistent with the hourly level of the global horizontal illuminance based on the weather condition. The weather condition of exterior horizontal illuminance shows that there could be a hazy or cloudy situation on the morning of June. Also, the pattern of circular shading panels works well in obstructing solar rays from the top throughout the daytime. Hence, the two curve trends of exterior and interior illuminance are completely matching. The result shows that the tilt angle between 70° and 90° facilitates more sunlight penetration in the daytime, but the values much exceed the high boundary of UDI between 12 PM later and near 3 PM. The sun at 10 AM is at a relatively low altitude, but still, many direct solar rays from east top illuminate the indoor space through the skylights because of a lower peak of the exterior illuminance. The sun goes up higher in the sky at 11 AM and 12 PM to make it easy for the direct sunlight to penetrate the skylights; nevertheless, the global horizontal illuminance goes down to some extent. As a result, the indoor illuminance level lowers down. After that, the sun goes down gradually in the afternoon, but the exterior horizontal illuminance grows larger and larger to reach its apex at 2 PM, so the direct sunlight passes through at a high direct illuminance. Later until 5 PM, both of the sun altitudes and the global horizontal illuminance are lower, leading to the decrease of indoor illuminance (Figure 5.19 & 5.20). Figure 5. 1 9 Hourly outdoor horizontal illuminances on June 21 st . 0 20000 40000 60000 80000 100000 120000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Illuminance (lux) Outdoor Horizontal Illuminance_June 21st Global HI Diffuse HI Direct NI 119 Figure 5. 20 Hourly sun altitude on June 21 st . On September 21 st , indoor illuminance through the simulation hours (7 AM to 5 PM) shows the vertex at noon (11 AM and 12 PM), and the curves of value variation appear regular hump pattern. The indoor illuminance value generally rises up as the tilt angle of the shading panel gradually becomes larger from 10° to 90° with increasing openness of the skylight. During the early morning (7 AM), indoor daylight illuminance stays at a low level. The pattern of indoor illuminance is relevant to the hourly trend of the global horizontal illuminance on the basis of weather conditions. The shading panel tracks the sun at a medium altitude of the year to block the direct sunlight in the daytime. When the sun is at a relatively high level to emit solar rays, it is helpful for the direct sunlight to penetrate the interior space from the skylight. Typically, the tilt angles between 80° and 90° meet the direct sunlight direction to come into the room from 10 AM to 2 PM, showing much higher illuminance than the other tilts. The sun rises up to the highest point in the sky at 12 PM, also the exterior horizontal illuminance increases to the peak, so the indoor illuminance curves achieve the maximum at this point except that the 90° tilt angle reaches the peak at 11 AM. Generally, the indoor illuminance values except for 80° and 90° tilts stay at an approximate level from 11 AM to 1 PM because of the subtle distinction of the sun altitude and the global horizontal illuminance. Furthermore, those two tilts can largely enhance indoor illuminance during the period due to the highest sun position in the daytime. After 1 PM, the sun descends gradually along with the decline of the exterior horizontal illuminance till sunset (Figure 5.21 & 5.22). 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Degree Sun Altitude_June 21st 120 Figure 5. 21 Hourly outdoor horizontal illuminances on September 21 st . Figure 5. 2 2 Hourly sun altitude on September 21 st . On December 21 st , the indoor illuminance through the simulation hours (7 AM to 5 PM) goes up in the morning and reaches the apex at noon, and the value drops down more gently at the lower tilt until 4 PM, then falls off. The indoor illuminance value basically increases as the tilt angle of the shading panel gradually gets larger with more openness of the skylight. 0 20000 40000 60000 80000 100000 120000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Illuminance (Lux) Outdoor Horizontal Illuminance_Sep. 21st Global HI Diffuse HI Direct NI 0.00 10.00 20.00 30.00 40.00 50.00 60.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Degree Sun Altitude_Sep. 21st 121 The shading panel tracks the sun at the lowest altitude of the year to regulate the direct sunlight in the daytime. As the sun goes up to higher angles, the direct sunlight enters the interior space more easily from the skylights. The curves are distributed homogeneously following the tilt order from 10° to 90°, which is correlated with the low altitude of the sun. The sun rises up to the highest position in the sky at 12 PM, as well as the exterior horizontal illuminance, so the indoor illuminance attains the maximum at this point. The indoor illuminance varies steadily as a whole, and its changes between 9 AM and 10 AM, 3 PM and 4 PM show a small difference despite a large discrepancy of the global horizontal illuminance since the sun altitude is low at this point and the diffuse horizontal illuminance with little difference predominates in daylight spread (Figure 5.23 & 5.24). Figure 5. 2 3 Hourly outdoor horizontal illuminances on December 21 st . 0 10000 20000 30000 40000 50000 60000 70000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Illuminance (lux) Outdoor Horizontal Illuminance_Dec. 21st 122 Figure 5. 24 Hourly sun altitude on December 21 st . The comparison of the indoor illuminance on June 21 st , September 21 st , and December 21 st basically shows that the highest illuminance level arises on June 21 st , whereas the lowest level occurs on December 21 st during the period of adequate daylight supply outdoors. Meanwhile, the comparison of outdoor illuminance levels appears to be a consistent result. The result is because a higher sun altitude helps the direct sunlight penetration from the skylight rather than blocks it to form shadow. The kinetic shading panel functions well on both allowing sunlight and obstructing it depending on the panel tilt. In brief, the exterior environment and the shading panel geometry can determine the pattern of indoor illuminance together (Table 5.12). Table 5. 12 Grid illuminance of the cuboid model with kinetic shading panels on three typical days. 0 5 10 15 20 25 30 35 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Degree Sun Altitude_Dec. 21st 123 5.2.2 Resulting analysis of daylight glare probability (DGP) On June 21 st , the south-facing glare index (DGP) over the daytime simulation (7 AM to 5 PM) demonstrates that it arrives at a lower peak around 10 AM and then drops down a little at around 11 AM to fluctuate until the value gets to the maximum point at 2 PM. After the summit, the values gradually fall off till sunset. The DGP value basically increases as the tilt angle becomes larger, so the curves are lined up from 90° at the top to 10° at the bottom. The overall curves of DGP value are acceptable, with the values below 0.35 that are imperceptible glare. The reason for the result is because sun altitudes are high between 9 AM and 3 PM, and the sun is positioned in the east in the early morning and the west in the late afternoon so that the direct sunlight cannot find a proper angle to reach the occupant’s view easily. Additionally, the skylight is a type of advantageous fenestration to avoid the glare disturbance for the occupant due to the entry position of direct sunlight. The glare probability declines at noon because of the decreasing global horizontal illuminance based on the weather pattern when the sun could be shaded by the haze or the cloud, while it reaches the peak value at 2 PM to be consistent with the exterior horizontal illuminance. The kinetic shading panel is capable of preventing glare since lower tilts can reduce the glaring possibility, but it is unnecessary to operate it for glare issues. On September 21 st , the south-facing glare index (DGP) over the daytime simulation (7 AM to 5 PM) shows a group of approximate parabolic curves, arrives at the peak at 12 PM and displays slight slope on the curves. The DGP curves are line up from the bottom to top as the tilt angle grows up. The whole DGP curves fall in the range of imperceptible glare with the threshold of 0.35. The result is caused by sun altitude and skylight. The sun is positioned at a medium altitude so that it is not easy for the sunlight to reach the height of the work plane and appear in the occupant’s view from the skylight where the direct sunlight passes through. The glare index mainly displays a similar tendency as the global horizontal illuminance. The kinetic shading panel could not be considered to use for controlling the glare in this case. On December 21 st , the south-facing glare index (DGP) over the daytime simulation (7 AM to 5 PM) demonstrates a group of approximate parabolic curves, also it arrives at the maximum point at 12 PM and displays relatively small change with a gentle slope from 9 AM to 4 PM on the curves versus abrupt slope before 9 AM and after 4 PM. The DGP curves line up from the bottom to top as the tilt angle grows up. Like the other cases, all the DGP values are acceptable for the occupant with imperceptible glare since the sun altitude becomes at the lowest path of the whole year, so it is difficult for the direct sunlight to penetrate the skylight and reach the height of the occupant’s eyes near the work plane. Therefore, the glare issue could be ignored in the case of regulating the tilt angle of the shading panel. 124 The comparison of the glare issue on June 21 st , September 21 st , and December 21 st basically proves that the glare shows a higher probability on June 21 st and lower chance on December 21 st during the period of abundant sunlight outdoors. Anyway, the indoor space can meet the occupant’s visual comfort since the overall glare indices belong to the imperceptible range. Consequently, the skylight is a significantly effective approach to avoid glare interference for the indoor space (Table 5.13). 125 Table 5. 13 DGP and interior renderings of the cuboid model based on the occupant’s view and sun position scenarios. 126 5.2.3 Resulting analysis of solar heat gain On June 21 st , the index of solar heat gain over the daytime simulation (7 AM to 5 PM) shows to rise up to a lower peak point at 11 AM and drops down slightly at 12 PM, then it goes up to the maximum point at around 2 PM or 3 PM. The maximum is far larger than the previous peak at 11 AM. After the peak value, the value of solar heat gain falls off till sunset. The solar radiation stays at a low level in the early morning (7 AM). The solar heat gain is closely related to the openness of the skylight because the value of solar heat gain increases as the tilt angle becomes larger. The trend of solar heat gain is influenced by the global horizontal radiation based on the weather pattern. Two trends about solar heat change perform similar regularity. The sun at 10 AM is at a lower altitude than it is at 11 AM, but the exterior horizontal radiation reaches a lower peak. However, a lower peak of indoor solar heat gain appears at 11 AM. The reason for the result could be a higher sun altitude is more beneficial to the solar radiation entering the indoor space from the skylight, as well as the delay effect of solar heat conduction. The sun goes up after 10 AM until 12 PM, but the global horizontal radiation goes down, resulting in a decreasing solar heat gain indoors. After the noon, the exterior horizontal radiation rises up continuously to attain the maximum at 2 PM along with the decline of sun altitude. As a result, most tilt angles of the shading panel reach the peak value of solar heat gain at 2 PM apart from the 90° tilt with the peak at 3 PM because the direct solar radiation is stronger at 3 PM integrated with the role of the shading panel tilt. Therefore, the exterior solar radiation plays a major role in influencing indoor solar heat gain. Later, solar heat gain gradually lowers down when the exterior solar radiation is reducing (Figure 5.20 & 5.25). 127 Figure 5. 2 5 Hourly outdoor horizontal radiation on June 21 st . On September 21 st , the curves of solar heat gain over the daytime simulation (7 AM to 5 PM) appear to rise up in the morning and get to the maximum value at 12 PM apart from the 80° tilt reaching the maximum at 1 PM. After the noon, the solar heat gain drops down at different rates until sunset. The value of solar heat gain shows to stay low in the early morning (7 AM). The solar heat gain is closely connected to the openness of the skylight since the value of solar heat gain increases as the tilt angle enlarges. The indoor solar heat gain is largely affected by the situation of the global horizontal radiation depending on the weather condition, and both of their curves are plotted as a regular parabolic pattern. The sun path rises to a medium altitude on this day. However, the sun is positioned at a high altitude at 12 PM during the daytime, so it is beneficial for the direct solar radiation to transmit into the indoor space from the skylight. Also, the global horizontal radiation meets its maximal value over the day, which offers the most solar heat for the existing building. Therefore, most of the maximum indoor solar heat gain happens at 12 PM in spite of the small distinction of exterior horizontal radiation between 12 PM and 1 PM. For the exception of 80° tilt, the result explains that this tilt helps the entry of direct solar radiation effectively because of the larger direct radiation at 1 PM (Figure 5.22 & 5.26). 0 100 200 300 400 500 600 700 800 900 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Wh/m 2 Outdoor Horizontal Radiation_June 21st Global HR Diffuse HR Direct NR 128 Figure 5. 26 Hourly outdoor horizontal radiation on September 21 st . On December 21 st , the solar heat gain over the daytime simulation (7 AM to 5 PM) appears to increase in the morning and attain the peak value at 1 PM except that the shading state at 50° tilt reaches the peak at 12 PM. Later, the solar heat gain falls off with different extents till sunset. The solar radiation is not available in the early morning (7 AM) and the late afternoon (5 PM), and it is also very weak at 8 AM. The value of solar heat gain is affected by the openness of the skylight because the solar heat gain rises when the tilt angle of the shading panel becomes large. Besides, the global horizontal radiation outdoors influences the solar heat gain indoors to a large extent. The sun is positioned at the lowest path of the whole year. Still, the global horizontal radiation mounts up to the maximum when the sun goes up to the highest position in the sky at 12 PM. Therefore, it is beneficial for the direct solar radiation to be transmitted into the indoor space through the skylight. Most of the tilts reach the maximal point at 1 PM since it is a little discrepancy of the sun altitude and global horizontal radiation between 12 PM and 1 PM, and also the conduction of solar heat delays in dozens of minutes. After 1 PM, the solar heat gain at the tilt angles of the shading panel between 60° and 90° goes down faster, but that at the other tilts drops down slowly because the direct solar radiation decreases at a high rate, while the diffuse radiation descends at a low rate. Specifically, the shading panel at a low tilt angle restricts more direct solar radiation and less diffuse radiation, and vice versa (Figure 5.24 & 5.27). 0 100 200 300 400 500 600 700 800 900 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Wh/m 2 Outdoor Horizontal Radiation_Sep. 21st Global HR Diffuse HR Direct NR 129 Figure 5. 27 Hourly outdoor horizontal radiation on December 21 st . To compare the solar heat gain on June 21 st , September 21 st , and December 21 st , it basically demonstrates that more solar heat gain is generated on June 21 st and a lower level displays on December 21 st during the period of adequate exterior sunlight. In addition, the level of outdoor solar radiation is showing the same tendency. The result is caused by two factors: one is the sun altitude, the other is exterior solar radiation. According to its top orientation of the skylight, a higher sun and stronger outdoor solar radiation in summer help more solar radiation come in the space, leading to more solar heat gain, and vice versa. In contrast, the interior space needs less or no solar heat gain in summer and more of it in winter. Therefore, the kinetic shading panel should work on controlling solar radiation based on seasonal demand. 5.2.4 Daylight effects comparison of the roofs of the cuboid model In contrast, the kinetic shading panels depending on an intelligent control strategy show an environmental performance on the indoor daylight effects superior to the other facades over the three days (June 21 st , September 21 st , and December 21 st ). (Table 5.14 - 5.22) (Figure 5.28 - 5.36) The kinetic shading panels present a better performance of indoor illuminance since it can track the sun azimuth and operate the panel tilt to adjust the real-time amount of direct solar rays and diffused daylight for meeting a required illuminance level of a specific working activity (e.g. 500 0 100 200 300 400 500 600 700 800 900 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM Wh/m 2 Outdoor Horizontal Radiation_Dec. 21st Global HR Diffuse HR Direct NR 130 lux for normal office work in this case). Generally, the kinetic panel adjusts the indoor illuminance level to be nearer the threshold of 500 lux than the tilt-only panel, but the grades of two shading panels do not demonstrate a large discrepancy. The static panel works worse on indoor illuminance level especially in summer, the sun is positioned at a high altitude with a higher exterior illuminance thus more daylight penetrates into space; but its performance can be accepted in winter. As far as the glare index, because of the advantage of the skylight, all three types of shading panels can basically provide the indoor occupant with visual comfort condition in the imperceptible level except that an intolerable glare under the static panel occurs at 2 PM on June 21 st , but it is a serious glare problem with the DGP value of 0.73. Additionally, the kinetic panel allows less solar radiation to be transmitted into space than the other panels in summer (June 21 st and September 21 st ), profiting from the panel rotation of sun tracking. On the contrary, the static panel introduces much solar heat into space, so it leads to thermal discomfort and more energy use on air conditioning. However, the kinetic panel still restricts the solar heat gain at a relatively low level in winter (December 21 st ) due to sun tracking, whereas tilt-only panel and static panel help space gain more solar heat, and benefit for reducing the energy consumption on space heating. The study also introduced a scoring method to establish an evaluation system depending on the integrated consideration for assessing the daylighting effects of three different façade patterns. The result shows that the kinetic shading panel scores the highest point in summer, and the tilt- only panel gets the highest point in winter, whereas the static shading panel is graded the lowest point in three days. Based on the conclusion from the three dates, it can be inferred that the kinetic shading panel is estimated to be the best regulator or strategy for the indoor daylighting performance over the year. 5.3 Evaluation of the parametric workflow The parametric workflow used in this study is based on the interface of Rhino 3D and Grasshopper and the Ladybug and Honeybee plugins. The approach involves a parametric design process and parametric simulation depending on the daylighting and energy simulation engine such as Radiance, Daysim, and EnergyPlus. The workflow structure and component effectiveness are evaluated. By using the proposed workflow, the kinetic model can be set up through modeling in Rhino or scripting in Grasshopper. The parametric design and analysis can be completed in Grasshopper, while the parametric simulation for daylight and energy analysis can be realized in Ladybug and Honeybee. Specifically, the process of model definition can involve all the kinetic states in a model. The simulation process allows the automatic mode over a period because a batch simulation can be set up using the component in Ladybug. These advantages help save the user 131 much time. Daylighting simulation can be completed for illuminance and luminance with either grid-based mode or image-based mode in Grasshopper, and Honeybee works faster than DIV A especially for point-in-time simulation. The performance of solar energy can be calculated by simulating the solar radiation amount that hits the building façade or the surface near the fenestration. Anyway, doing parametric design and simulation on kinetic facades requires the user to be capable of knowing the visual scripting in Grasshopper. Additionally, the interface also allows advanced users to build customized components for personalized demand. By using these available tools to conduct an environmental simulation on the building with kinetic facades, the study generalizes the merits and demerits of the tool kit and its workflows in the table (Table 5.23). The evaluation factors for the tool kit are listed as follows: 1) Modeling • Parametric Capability. • Modeling Difficulty. 2) Energy Performance Assessment • Solar Heat Gain Calculation. • Solar Radiation Calculation. • Dynamic Shadow Calculation & Analysis. 3) Daylighting Performance Assessment • Indoor Illuminance Calculation. • Daylight Glare Probability Calculation • Other daylighting Metrics Availability. 4) General Properties • Parametric Analysis Capability. • Popularity & Accessibility. • Work Stability. 5) Data Processing • Optimization Program Availability. (e.g. Octopus and Galapagos in Grasshopper, and Optimo in Dynamo). • Generative Design Program Availability. (e.g. Project Fractal). 132 Tools Rhino, Grasshopper, Ladybug and Honeybee Modeling Parametric Capability Yes Modeling Difficulty Neutral Daylighting Performance Assessment Indoor Illuminance Calculation Yes Daylight Glare Probability Calculation Yes Other daylighting Metrics Availability Yes Energy Performance Assessment Solar Heat Gain Calculation Yes Solar Radiation Calculation Yes Dynamic Shadow Calculation & Analysis No General Properties Parametric Analysis Capability Yes Popularity & Accessibility Neutral Work Stability Good Data Processing Optimization Program Availability Yes Generative Design Program Availability Yes Table 5. 23 Experience comparison of four tools (workflows) for environment simulation. 5.4 Summary This chapter evaluates the parametric workflow of environmental simulation on kinetic facades by analyzing and explaining the functionality of the components in Grasshopper, Ladybug, and Honeybee. The study discusses two different kinetic façade models about their daylight and solar energy performance. Furthermore, for each case, the study analyzes the probable reason for the simulation results of three representative days over the year (separately representing summer and winter) and infers the potential connection between the exterior environment and the interior daylighting condition. Specifically, not only the exterior environment such as the sun position and the weather condition affect the indoor daylighting and thermal indices, but also the façade 133 geometry and pattern influence those conditions. Also, it compares the indoor daylighting and thermal results in three days. The study found the indoor daylighting and thermal environment results from the combined effects of both the environmental factor and the geometric feature of the kinetic facade pattern. As far as the daylighting variables in the study, sun position and weather conditions play a significant role in determining the indoor environment. Specifically, for the kinetic louver on the shoebox model, due to the sunlight entry from the side fenestration, the sun at a low altitude in winter benefits the daylighting penetration and solar heat radiation but simultaneously brings about the glare interference, whereas a higher sun in summer only can transmit relatively little daylight and solar energy into the indoor space with lower probability of glare, and the sunlight direction can combine with the specific louver tilt to affect the daylighting and thermal condition (Figure 4.5 -4.13); for the kinetic shading panel on the cuboid box, because of the sunlight entry from the top skylight, the sun at a high altitude in summer results in helping more daylight and solar radiation reach the building interior, while a lower sun in winter brings about less daylight and solar heat inside the building, and the skylight is a strong passive strategy to avoid the glare (Figure 4.15 – 4.23). Also, the study has conducted the daylighting and solar heat simulation on three individual cases of each building model on three representative days, including the shoebox model with kinetic louver, Fixed louver, and static glazing and the cuboid model with kinetic shading panels, tilt-only shading panels, and static shading panels. Through doing data organization and analysis, the study concludes that the kinetic louver and the kinetic shading panel demonstrate to control the indoor daylighting and thermal condition superior to the other counterparts from the integrative score result. That is, the kinetic louver scored 27.5, 27.5, and 19 on June 21 st , September 21 st , and December 21 st , and the kinetic shading panel was graded 29, 28.5, and 19.5 (the last point lower than tilt-only panels) (Figure 4.24 & 4.25, Table 4.36 &4.37). Therefore, the result shows that the model-based predictive control can be treated as a control algorithm to be used for generating the kinetic control strategy based on the interior environmental simulation process. The parametric workflow in Grasshopper is a useful tool to design and analyze kinetic facades in terms of their performance for the interior daylight environment in the early design stage since it is powerful in parametric modeling and simulation along with many sophisticated components in Ladybug and Honeybee. As a result, the architectural designer can complete kinetic facade design and analysis using the tool kit. 134 CHAPTER 6 6. DISCUSSION AND FUTURE WORK Kinetic facades are increasingly proposed to apply to the building envelope for the reason of either architectural aesthetics or indoor environmental control. As a result, it is important to develop means to evaluate kinetic systems and their various possible controls options from the perspective of performance in regard to indoor environmental control. Besides, it is often seen that a group of architectural designers come up with several different façade patterns in the same building. When this situation occurs, the designers still can compare different kinetic façade patterns based on different control strategies by using the proposed evaluation system to score each kinetic façade. Therefore, the superiority of particular kinetic facades is actually discerned by comparing several different facades within the uniform scoring system that the architectural designer can determine based on the data scale and range. For architectural or facade designers, the research provides a method for them to develop a well- performing kinetic façade compared to static facades and simple kinetic facades depending on the model-based predictive control algorithm since the workflow supports a more opportune shading scenario at each specific hour through daylighting calculation and evaluation. In the early design process, they can follow the procedure of generating a control strategy shown in the study to produce a group of control strategies on the identical kinetic façade plus the static façade from the simulation data. Then the best façade (highest score) of the building model based on a certain control (static facade seen as static control) can be concluded through comparing all the facades in the same condition (time and weather). During the early design stage of kinetic façade structure design, Rhino and Grasshopper, as a tool of modeling and managing data for architectural designers, are being used extensively in the AEC industry to accomplish a qualified design work. This research aims to create a workflow that develops a daylight control of kinetic facades by simulating the interior daylight condition and analyzing the interior environmental performance of façade patterns, as the specific objectives are the followings:  Use the parametric software and workflows to model and simulate the interior daylight and thermal conditions of the building model to develop a dynamic control for kinetic facades in response to dynamic environmental variables (climate, occupant, and structure, etc.).  Establish an evaluation system for scoring the daylighting condition and energy performance of the interior space with kinetic facades.  Experiment to compare the indoor cases with kinetic facade A, B, and C. 135 This research completed several of the objectives at least in part. Specifically, it demonstrated the process of using a parametric workflow to develop control strategies for two kinetic façade cases (kinetic louver on the shoebox model and kinetic shading panels on the cuboid model) (Figure 3.5 & 3.6) by processing the simulation data. Also, it establishes an evaluation system with a detailed scoring criterion to grade the performance of each façade. Lastly, it analyzed the cause of metrics results on the kinetic shading states and compares three different façade patterns on the same building model. However, this research did not complete the comparison of diverse kinetic facades on an identical building model. For example, the kinetic shading panels and tilt- only shading panels are merely two forms of the same façade pattern. Additionally, it neither completed conducting the simulation of all daytime hours of the whole year nor annual daylight metrics to simulate the kinetic facade to verify its performance in one year. In summary, the research investigates a systematic methodology with daylight simulation experiments to discern kinetic patterns of motion based on the indoor daylight and thermal performance. Three lessons were learned in the study by completing the three main objectives:  The investigation of the process that uses the parametric workflow to model and simulates dynamic shading states of kinetic façade buildings and generates the control strategy in response to exterior environment change;  The approach of data analysis that respectively compares daylight and thermal metrics of the identical kinetic model under the different exterior environmental condition and different facade patterns in terms of the interior daylight and energy performance;  The method to establish an evaluation system that scores the kinetic or static facades based on daylight metrics of the built interior environment by completing two case studies on comparing different façade patterns on an identical building model. Based on these tools, the methodology involves a parametric workflow to create a database with daylighting and solar heat data that could be used to implement the control algorithm (model- based predictive control). The process includes defining a material model, testing an environmental model, and data processing. The overall workflow proposed two cases of kinetic facades with two specific control strategies separately based on the result analysis of environmental simulation and the performance evaluation between the kinetic systems and their counterparts. The process and results can be informative to architectural designers to develop an effective kinetic façade during the early stage of architectural design. Specifically, the process demonstrates the method that applies a workflow in Rhino and Grasshopper to simulating the daylighting and thermal model with moveable facades for collecting hourly data of the proposed metrics. For the designer, the metric values compose a database by which they can weigh the values to decide an operational state (tilt) at each hour with the combination of different metric values. These operational states are collected to form a control model of the kinetic façade which 136 can be compared with a static façade or other control modes of the identical façade based on the evaluation criteria. However, the study can still improve on workflow development, data processing methods, and other aspects. For example, the current workflow concerns only three hourly metrics (average illuminance, DGP, and solar heat gain) as evaluation factors, so a further workflow can be developed to involve annual metrics such as sDA and ASE. Also, the range of simulation hours and the scoring criterion are negotiable to be refined. 6.1 Methodology discussion The objective of the overall workflow is to develop a systematic and logical method for the architectural designer. The method applies the model-based predictive control (MPC) algorithm as a basis for designing a kinetic façade based on interior daylighting and thermal simulation and evaluates the proposed kinetic façade by comparison between it and its other counterparts by introducing an evaluation system with scoring criterion. For example, the kinetic louver on the shoebox model was compared with the static louver at 0° tilt and the static glazing, while the kinetic shading panels on the cuboid model were contrasted with the tilt-only panels and the fixed panels at 60° tilt. Another basis of the method involves using daylight and thermal metrics as the common criteria to quantify the indoor environmental performance in terms of kinetic facades. The methodology for developing the control strategy for kinetic façades by simulating the daylighting and thermal energy and evaluating the result of metric values in the digital interface of Rhino, Grasshopper, and Microsoft Excel was developed and validated (Figure 6.1). Figure 6. 1 The workflow of the methodology for kinetic facades design and evaluation. 137 A literature review was initially fulfilled to primarily inspect the kinetic concept, the current research with its method, and control algorithms, as well as simulation and evaluation methods with available tools for the kinetic facades, in addition to the application of parametric design programs including Rhino and Grasshopper. Based on the literature review, four aspects of issues were investigated in the research as follows:  The design method for kinetic facades without an effective parametric workflow possibly causes difficulties in developing and improving kinetic systems during the early design stage because of the complexity of kinetic facades (Figure 6.2).  Architects have to do too many simulations to evaluate too many options and this is hard or impossible to do “by hand”. The simulation process for kinetic facades without an effective parametric workflow probably operates inefficiently and inaccurately due to the huge workloads of indoor environmental simulation especially for daylight hours of the whole year.  Current research rarely presented an integration of the parametric method and kinetic control algorithm to propose the feasible control strategy for a kinetic façade with improved indoor daylighting and thermal performance, as well as energy savings.  Few current studies proposed an evaluation method integrated with scoring criteria to quantify the performance of a designed kinetic façade by comparing it with its counterpart facades based on several valuable daylight and thermal metrics. Figure 6. 2 Examples of the complexity of the geometric configurations generated from parametric definition. Source: Velasco, R. & Robles, D., 2011. Based on the parametric tools, the methodology involves a parametric workflow that achieves to model, simulate and analyze the kinetic models with the following functionality: it parameterizes the building body and the shading device to generate the design case, thus the digital model can 138 be evaluated and modified easily and instantly, saving much time on developing the kinetic model. Also, it parameterizes the whole simulation process that includes exterior environment (such as weather conditions and sun positions), interior factors (like surface materials), and the simulation definition (including lighting calculation, analysis recipe, and simulation engine adjustment), so the simulation definition in the complex and diverse process is quantified to automatically execute the procedure within the cyclic operation. Furthermore, it explores an effective approach that simulates the environmental performance of each static state to define kinetic motions at regular intervals based on the control algorithm (Figure 6.3). Lastly, it designs a set of assessment criteria for scoring the facades by comparison for verifying the merits or demerits of a proposed kinetic façade. Figure 6. 3 The diagram of model-based predictive control shows the operation cycle. Therefore, the study basically discovered that a well-performing kinetic façade essentially relies on a control strategy to manage the indoor condition. The control strategy in the study was developed by calculating daylighting conditions (daylight illuminance and daylight glare probability) and solar radiation and balancing all three factors to produce a satisfactory indoor condition, and it is an intrinsic feature of physical kinetic systems. As for the workflow procedure, the models of the façade and building body were generated to demonstrate in Rhino and Grasshopper, and the digital building model was converted into thermal zones by Honeybee, preparing for the simulation process. The necessary parameters of daylight and thermal simulation were added with Ladybug and Honeybee components, including the exterior environment like weather conditions and interior condition like material settings, as well as the simulation algorithm. Besides, the function of iterative operation emulates practical kinetic motions or shading states to produce the environmental simulation data; in other words, the process mimics a sequential movement of dynamic facades. Consequently, the hourly data of 139 daylight metrics and solar heat gain in the daytime was exported from Grasshopper into Microsoft Excel to analyze for the control strategy of kinetic operation and process for the conclusion of building performance based on each façade (Figure 6.4). The results of these case studies demonstrate that the parametric workflow using Rhino, Grasshopper, Ladybug, and Honeybee is feasible and helpful to develop the control algorithm for kinetic facades and evaluate their performance in terms of the indoor daylight and energy conditions. Meanwhile, the data result of parametric simulation can be processed to analyze and visualize in the interface of either Rhino or Microsoft Excel for a more visible and accurate determination. Figure 6. 4 The score results of daylighting environmental performance. 6.2 Limitations The research refers to creating a parametric workflow to design and evaluate kinetic facades by simulating the indoor daylight and thermal environment and analyzing the daylighting and energy performance. However, the research has some limitations that can be improved with future work.  The days of the daylight and thermal simulation are only three representative dates in summer and winter. The complete results should be concluded by simulating annual daytime hours to generate the kinetic control strategy and compare different facade patterns for scoring.  The evaluation process did not involve the yearly daylight metrics such as sDA, ASE, and annual solar heat gain, etc. A reasonable conclusion of indoor daylight and thermal performance should include complete annual indices that can reflect a case holistically. June 21st Sept. 21st Dec. 21st Kinetic Shading Panel 29 28.5 19.5 Tilt-only Shading Panel 23.5 25 21.5 Static Shading Panel 12 12 17 0 5 10 15 20 25 30 Score Daylighting Performance Evaluation 140  Both the shoebox building and cuboid spaces are a small area of space that represents a room or a hall, so the simulation result cannot exactly reflect the interior daylighting and thermal performance in a real building scale.  The precision of current environmental simulation in this study is one-hour intervals. Thus, whether the control strategy can be improved to a higher level to present a better daylighting and thermal performance if the study shortens the sampling interval.  The process of the facade control did not involve the occupant’s feedback that can reflect the practical human need to a large extent. The kinetic control that the occupants can participate in or overrule will make a distinct comparison of the interior condition between the computational control and human control.  The evaluation process is based on the theoretical criteria of the daylighting and thermal condition of a normal office workplace without the occupant’s evaluation. Whether the dynamic control can meet the human’s requirement for lighting conditions, visual and thermal comfort is still inconclusive. This process proposed metrics priority and scoring criteria from the perspective of the subjective study. These affect the evaluation results to some extent. Especially, the choice of variables and scales on daylight metrics is not adequately objective so that it can influence the precision of the score and comparison result.  Due to digital simulation, the data results in the study are determined with the input parameters and calculation algorithm in terms of daylighting and thermal condition, so the data precision still needs validating from a real physical model or a test facility. 6.3 Future work For the research, several aspects of the current methodology can be ameliorated and developed in further work. Further development from additional work will propose more effective kinetic control strategies based on the kinetic facade pattern and more efficient workflow for collecting and processing simulation data, as well as a more accurate evaluation of daylighting and thermal performance of kinetic facades. Possible further developments include as follows:  The study will run simulations for a longer time and process additional data on daylighting and thermal values. The study merely simulates the hourly daylighting and solar heat gain of kinetic facades on three representative days (June 21 st , September 21 st , and December 21 st ) in one year to indicate typical season changes of the control strategy and the indoor environment performance based on the climate pattern. For more reliable operation and superior performance, the study will further simulation process to obtain more daylighting and thermal data in two steps. The first step is to select four typical days per month (one day per week) to run the simulation and analyze the data for more exact control and better performance on kinetic facades monthly. The second step is to simulate through the daytime hours of the whole year for yearly data to derive a more accurate control strategy and a more 141 authentic reflection of the indoor environment for kinetic facades. The annual simulation can be more informative and instructive for designers in helping them know more about the kinetic operation scenario throughout one year to optimize the kinetic pattern.  The study will introduce annual daylight and thermal metrics for an integrative evaluation: the current workflow only calculates the hourly metrics including daylight illuminance, daylight glare probability, and hourly solar heat gain for shading state selection and evaluation at each hour. The study will further simulate the annual daylighting and thermal performance of kinetic facades based on the metrics such as spatial daylight autonomy (sDA), annual sunlight exposure (ASE), daily or seasonal solar heat gain, and electric lighting energy use, etc., improving the evaluation to be more complete and holistic for indicating the quality of proposed kinetic facades. This course may involve creating the customized components for the specific workflow concerning dynamic control and simulation algorithms, resulting in a large workload. Anyhow, these calculations can make the conclusion more persuasive. This portion of work is initially planned to complete in the thesis currently, but there should be more workload on the parametric definition, simulation, and data processing for a reliable conclusion, so it is scheduled in the future work due to the limited thesis time (Figure 6.5). Figure 6. 5 The parametric workflow of introducing annual daylight and thermal metrics.  The study will improve the determination criterion of solar heat gain for shading device control. The current study determined whether the interior space needs solar heat by comparing outdoor ambient temperature with indoor space temperature. The result only offered qualitative selection (whether needs solar heat or not) but the quantitative value that indicates how much solar heat gain is needed to prevent overheating. The value discrepancy of heat loss and internal heat gain is helpful to calculate the desired amount of solar heat 142 gain. For indoor thermal comfort, the criterion can be improved by studying the relation between the mean radiation temperature (MRT) and the solar heat gain.  The study could test on a physical model and a real full-scale lab facility. For the confidence of the methodology, the proposed method is objectively required to validate the project within the real environment by architectural professionals and facades specialists. Instead, to validate the digital simulation, a scaled-down physical model can be made to conduct the environmental test using the light and thermal sensors to obtain data for comparing the results to those from the digital model, or the data from testing a real full-scale lab with sensors for environmental conditions can be used to compare with the lighting and thermal data from the digital simulation. It is significant to validate the simulation data in a physical condition for validity and reliability. Especially, the practical project is much more complicated because of flexible façade patterns as well as complex building configuration and space layout.  The study should use generative design tools to process data more efficiently. In the study, the determination of the control strategy of kinetic facades was completed by exporting the simulation data into Microsoft Excel that stores the hourly daylight and thermal values to analyze shading state cases for generating the operation scenario and compares different façade patterns to assess their indoor environmental performance with the graphic charts of the visualized data. The process is done by the designer manually. Instead, a potential tool for selecting an optimum resolution, called Octopus, is available as a plugin in Grasshopper to make the process straighter and more efficient. It can help the designers find the shading state of the kinetic façade that mostly meets the criteria at different hours by defining constants and variables. 6.4 Conclusion The research explored a methodology with an integrative parametric workflow to investigate the relationship between the dynamic patterns of motion of kinetic façade shading systems compared with static and simple dynamic systems in regard to the indoor daylighting and thermal performance. Also, it used environmental analysis software and simulation tools including Grasshopper, Ladybug & Honeybee, and Radiance, etc. to conduct the calculation and analysis process, and it developed a parametric workflow to simulate kinetic operation scenarios (shading states) and design the control strategy for kinetic facades about the pattern of motions. The parametric modeling and simulation produced daylighting and thermal data that is exported to Excel for further data processing. During data analysis, appropriate kinetic shading states were selected to form a complete control strategy of kinetic facades on each day by the proposed criteria on metrics priority and threshold. Consequently, the kinetic façade can follow the 143 appropriate shading states by making an opportune trade-off among avoiding adverse glare, offering required illuminance, as well as utilizing solar heat (winter) or avoiding overheating (summer). Therefore, the kinetic facades based on model-based predictive control has been developed from the analysis of simulation data. Then the research also established an evaluation system with a specific scoring standard for comparing the proposed kinetic façade and two other counterparts. The score calculation demonstrates that on the identical building model, the kinetic façade based on the dynamic control strategy is superior to the other facades in terms of the daylighting and thermal performance as a whole. According to the presented method and the case study, it provides the architectural designer with an applicable idea or logic on the design and evaluation of kinetic facades, though the workflow has some aspects of limitations. Consequently, the further development of the method can be a superior system based on the parametric design for kinetic façade study. 144 BIBLIOGRAPHY ASHRAE press. 4 - architectural design impacts, The ASHRAE GreenGuide, seconded., Butterworth-Heinemann, Burlington, 2006[cited 2019 Mar 5]. pp. 55–72. Available from: http://www.sciencedirect.com/science/article/pii/B9781933742076500074. Baker, N., & Steemers, K. (2014). Daylight design of buildings: a handbook for architects and engineers. Routledge. Barozzi, M., Lienhard, J., Zanelli, A., & Monticelli, C. (2016). The sustainability of adaptive envelopes: developments of kinetic architecture. Procedia Engineering, 155, 275-284. Caruso, G., & Kämpf, J. H. (2015). Building shape optimisation to reduce air-conditioning needs using constrained evolutionary algorithms. Solar Energy, 118, 186-196. Dubois, M. C. (2001). Impact of Solar Shading Devices on Daylight Quality. Department of Construction and. Dussault, J. M., Sourbron, M., & Gosselin, L. (2016). Reduced energy consumption and enhanced comfort with smart windows: Comparison between quasi-optimal, predictive and rule-based control strategies. Energy and Buildings, 127, 680-691. EIA, U. (2012). Commercial buildings energy consumption survey. United Sates Department of Energy, Ed., ed. El Razaz, Z. (2010). Sustainable vision of kinetic architecture. Journal of Building Appraisal, 5(4), 341-356. El-Dbaa, R. (2016). The Use of Kinetic Facades in Enhancing Daylight Performance for Office Buildings. Master’ s degree. Arab Academy for Science, Technology and Maritime Transport. EL-Deeb, K. (2013). Combined Effect Of Window-To-Wall Ratio And Wall Composition On Energy Consumption: Where Is The Turning Point? In Sustainable Architecture for a Renewable Future. In Proceedings – PLEA2013 - 29th Conference, (pp. 10–12). Munich, Germany. Elghazi, Y ., Wagdy, A., & Abdalwahab, S. (2015). Simulation driven design for kinetic system; optimize kaledocycle facade configuration for daylighting adequacy in hot arid climates. In Conference of International Building Performance Simulation Association (pp. 182-189). Ercan, B., & Elias-Ozkan, S. T. (2015). Performance-based parametric design explorations: A method for generating appropriate building components. Design Studies, 38, 33-53. 145 Farley, K. M., & Veitch, J. A. (2001). A room with a view: A review of the effects of windows on work and well- being. National Research Council of Canada (NRCC) technical report IRC-RR-136. Fox, M., & Kemp, M. (2009). Interactive architecture. New York: Princeton Architectural Press. Gane, V ., & Haymaker, J. (2007). Conceptual design of high-rises with parametric methods. Geli, E. R. (2010). Media-ICT. Actar. Godfried, A. (2011). The role of simulation in performance based building. Building Performance Simulation for Design and Operation, 15-36. Grobman, Y . J., & Elimelech, Y . (2016). Microclimate on building envelopes: testing geometry manipulations as an approach for increasing building envelopes' thermal performance. Architectural Science Review, 59(4), 269-278. Grozdanic, L. (2016, November 7). 8 Impossibly Dynamic Facades That Were Actually Built. Retrieved from Architizer: https://architizer.com/blog/8-impossibly-dynamic-facades-that-were-actually-built/ Handbook-Fundamentals, A. S. H. R. A. E. (2017). ASHRAE-American Society of Heating. Ventilating and Air- Conditioning Engineers. Heschong, L. Heschong Mahone Group. (HMG) (2012). Daylight metrics. California Energy Commission. Publication number: CEC-500-2012-053. Holzer, D., Hough, R., & Burry, M. (2007). Parametric design and structural optimisation for early design exploration. International Journal of Architectural Computing, 5(4), 625-643. Hosseini, S. M., Mohammadi, M., & Guerra-Santin, O. (2019). Interactive kinetic façade: Improving visual comfort based on dynamic daylight and occupant's positions by 2D and 3D shape changes. Building and Environment, 165, 106396. Hosseini, S. M., Mohammadi, M., Rosemann, A., Schröder, T., & Lichtenberg, J. (2019). A morphological approach for kinetic façade design process to improve visual and thermal comfort. Building and Environment. Huang, J., & Franconi, E. (1999). Commercial heating and cooling loads component load analysis. Berkeley, CA, Building Technologies Department, Lawrence Berkeley National Laboratory: LBNL-37208. 146 IES Daylight Metrics Committee. (2012). IES Spatial Daylight Autonomy (sDA) and Annual Sunlight Exposure (ASE), Daylight Metrics Committee. Approved Method IES LM-83-12. Illuminating Engineering Society of North America. Jakubiec, J. A., & Reinhart, C. F. (2012). The ‘adaptive zone’–A concept for assessing discomfort glare throughout daylit spaces. Lighting Research & Technology, 44(2), 149-170. Keighley, E. C. (1973). Visual requirements and reduced fenestration in office buildings—A study of window shape. Building science, 8(4), 311-320. Khoshroonejad, S. (2010, August). A Comparison of Daylight Prediction Methods. Eastern Mediterranean University, Gazimağusa, North Cyprus. Klepeis, N. E., Nelson, W. C., Ott, W. R., Robinson, J. P., Tsang, A. M., Switzer, P., ... & Engelmann, W. H. (2001). The National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to environmental pollutants. Journal of Exposure Science and Environmental Epidemiology, 11(3), 231. Konis, K., & Selkowitz, S. (2017). Effective daylighting with high-performance facades: emerging design practices. Springer. Koster, H. (2004). Dynamic Daylighting Architecture: Basics, Systems, Projects. Springer Verlag. Lee, D. S., Koo, S. H., Seong, Y . B., & Jo, J. H. (2016). Evaluating thermal and lighting energy performance of shading devices on kinetic facades. Sustainability, 8(9), 883. Li, D. H. W., & Lam, J. C. (2000). Vertical solar radiation and daylight illuminance data for Hong Kong. International Journal of Lighting Research and Technology, 32(2), 93-98. Li, D. H. W., Wong, S. L., C.L. Tsang, & Cheung, G. H. W. (2006). A Study of the Daylighting Performance and Energy Use in Heavily Obstructed Residential Buildings via Computer Simulation Techniques. Energy and Buildings, 38(11), 1343–1348. Mahmoud, A. H. A., & Elghazi, Y . (2016). Parametric-based designs for kinetic facades to optimize daylight performance: Comparing rotation and translation kinetic motion for hexagonal facade patterns. Solar Energy, 126, 111-127. Mardaljevic, J. (2006). CIBSE national conference 2006: engineering the future. 21–22 March 2006. Oval Cricket Ground, London, UK. 147 Mardaljevic, J., Heschong, L., & Lee, E. (2009). Daylight metrics and energy savings. Lighting Research & Technology, 41(3), 261-283. Megahed, N. A. (2017). Understanding kinetic architecture: typology, classification, and design strategy. Architectural Engineering and Design Management, 13(2), 130-146. Moloney, J. (2011). Designing kinetics for architectural facades: state change. Abingdon, Oxon; New York: Routledge. Oosterhuis, K. (2012). Hyperbody: First Decade of Interactive Architecture. Jap Sam Books. Oosterhuis, K., Xia, X., & Sam, E. J. (Eds.). (2007). Interactive Architecture. episode publishers. Oxman, R. (2006). Theory and design in the first digital age. Design studies, 27(3), 229-265. Pesenti, M., Masera, G., Fiorito, F., & Sauchelli, M. (2015). Kinetic solar skin: a responsive folding technique. Energy Procedia, 70, 661-672. Phillips, D. (2004). Daylighting Natural Light in Architecture. London, England: Architectural Press publications. Plummer, H. (1995). Light in Japanese architecture. a+ u Publishing. Ramzy, N., & Fayed, H. (2011). Kinetic systems in architecture: New approach for environmental control systems and context-sensitive buildings. Sustainable Cities and Society, 1(3), 170-177. Rathi, P. (2012). Optimization of Energy Efficient Windows in the Office Buildings for Different Climate Zones of the United States (Master of Architecture). College of Architecture and Environmental Design of Kent State University, Kent State University. Rea, M. S. (2000). The IESNA lighting handbook: reference & application Risen. Clay (2017, July 11). Award: Dynamic ETFE Facade. Retrieved from. http://www.architectmagazine.com/awards/r-d-awards/award-dynamic-etfe-facade_o Schumacher, M., Schaeffer, O., & V ogt, M. M. (2010). MOVE: Architecture in Motion - Dynamic Components and Elements. Basel: Birkhauser. Sharaidin, K., Burry, J., & Salim, F. (2012). Integration of digital simulation tools with parametric designs to evaluate kinetic façades for daylight performance. In Physical digitality: proceedings of the 30th eCAADe 148 conference (V ol. 2, pp. 701-709). Sharaidin, M. (2014). Kinetic facades: towards design for environmental performance. Shea, K., Aish, R., & Gourtovaia, M. (2005). Towards integrated performance-driven generative design tools. Automation in Construction, 14(2), 253-264. Sheriden, J. (2000). The engineering of Style. Naval Architecture, Current Thinking on Smart Products and Materials. Jonathan Sheridan Ltd. http://www.designpartners.co.uk/Smart.html. 2012 Suk, J., & Schiler, M. (2013). Investigation of Evalglare software, daylight glare probability and high dynamic range imaging for daylight glare analysis. Lighting Research & Technology, 45(4), 450-463. Tabadkani, A., Banihashemi, S., & Hosseini, M.R. (2018). Daylighting and visual comfort of oriental sun responsive skins: A parametric analysis Vartiainen, E., Peippo, K., & Lund, P. (2000). Daylight optimization of multifunctional solar facades. Solar Energy, 68(3), 223-235. Veitch, J. A. (2001). Psychological processes influencing lighting quality. Journal of the Illuminating Engineering Society, 30(1), 124-140. Veitch, J. A., & Newsham, G. R. (1996). Determinants of Lighting Quality II: Research and Recommendations. Veitch, J. A., & Newsham, G. R. (1998). Determinants of lighting quality I: State of the science. Journal of the Illuminating Engineering Society, 27(1), 92-106. Velasco, R., & Robles, D. (2011). Eco-envolventes: A parametric design approach to generate and evaluate façade configurations for hot and humid climates. In Respecting Fragile Places [29th eCAADe Conference Proceedings (pp. 539-548). Viola, A., & Roudsari, M. S. (2013). An innovative workflow for bridging the gap between design and environmental analysis. Proceedings of BS2013, 1297-1304. V olla, H., & Seinre, E. (2012). Cooling Demand and Daylight in the New Tallinn Town Hall Buildings the Influence of Facade Design. Energy Procedia, 30, 1243–1249. Wienold, J., & Christoffersen, J. (2006). Evaluation methods and development of a new glare prediction model for daylight environments with the use of CCD cameras. Energy and buildings, 38(7), 743-757. 149 Wigginton, M., & Harris, J. (2002). Intelligent Skins. Reed Educational and Professional Publishing Ltd. Winkelmann, F. C., & Selkowitz, S. (1985). Daylighting simulation in the DOE-2 building energy analysis program. Energy and Buildings, 8(4), 271-286. Zaera-polo, A., Truby, S., Koolhaas, R. (2014). I. boom, Façade. Elements of Architecture – 14.. International Architecture Exhibition, la Biennale di Venezia. Zuk, W., & Clark, R. H. (1970). Kinetic architecture. New York: Van Nostrand Reinhold. 150 APPENDICES APPENDIX A: Simulation Resulting Data of Shading States at Each Hour. 6/21 Indoor Average Illuminance Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 51.75 104.88 144.35 140.09 105.24 55.53 22.91 12.49 7.71 3.69 1.05 8:00 AM 104.71 213.26 292.86 282.83 210.04 108.45 45.28 25.65 15.54 8.11 2.03 9:00 AM 155.06 311.93 425.35 417.55 305.35 155.70 69.40 37.99 23.81 11.95 3.24 10:00AM 282.13 567.54 770.18 741.86 514.12 273.07 120.55 69.59 42.03 21.63 6.24 11:00AM 204.30 429.48 577.09 559.49 412.48 212.01 91.74 51.94 34.49 17.14 5.23 12:00 PM 212.07 416.87 565.40 537.97 384.32 192.66 68.43 30.27 18.50 9.45 2.85 1:00 PM 396.52 735.99 935.36 867.43 617.37 343.86 152.93 88.64 56.55 27.97 7.31 2:00 PM 517.73 822.55 985.96 926.24 710.90 443.30 260.17 177.56 118.20 57.41 16.22 3:00 PM 254.17 456.70 582.88 614.99 521.44 368.03 227.43 155.69 102.42 47.47 11.80 4:00 PM 169.36 339.09 474.11 518.73 454.67 298.51 171.49 110.34 72.87 32.52 8.38 5:00 PM 154.28 301.85 435.63 463.16 379.49 227.68 118.21 68.58 45.32 22.21 4.95 Table 4. 2 Simulation results of hourly indoor average illuminance on June 21 st . Figure 4. 5 Hourly indoor average illuminances on June 21 st based on 11 tilt angles. 0.00 200.00 400.00 600.00 800.00 1000.00 1200.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Time Indoor Average Illuminance 6/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 151 6/21 Daylight Glare Probability Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 0.010 0.020 0.084 0.200 0.228 0.226 0.154 0.037 0.012 0.010 0.006 8:00 AM 0.027 0.085 0.215 0.241 0.253 0.255 0.247 0.155 0.036 0.023 0.011 9:00 AM 0.054 0.168 0.233 0.256 0.272 0.272 0.260 0.222 0.090 0.051 0.017 10:00AM 0.180 0.223 0.256 0.289 0.313 0.307 0.284 0.247 0.207 0.170 0.045 11:00AM 0.114 0.209 0.244 0.272 0.292 0.289 0.271 0.238 0.161 0.099 0.031 12:00 PM 0.078 0.195 0.237 0.268 0.287 0.283 0.265 0.229 0.101 0.056 0.025 1:00 PM 0.220 0.235 0.273 0.308 0.334 0.325 0.295 0.255 0.225 0.206 0.073 2:00 PM 0.254 0.260 0.311 0.333 0.357 0.351 0.316 0.273 0.240 0.230 0.126 3:00 PM 0.249 0.249 0.275 0.305 0.330 0.331 0.305 0.266 0.234 0.214 0.074 4:00 PM 0.197 0.233 0.258 0.290 0.313 0.315 0.292 0.257 0.223 0.178 0.049 5:00 PM 0.121 0.213 0.245 0.275 0.296 0.296 0.277 0.244 0.183 0.111 0.024 Table 4. 3 Simulation results of hourly Daylight Glare Probability on June 21 st . Figure 4. 6 Hourly Daylight Glare Probability on June 21 st based on 11 tilt angles. 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Time Daylight Glare Probability 6/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 152 6/21 Solar Heat Gain (kWh) Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 0.033 0.081 0.121 0.094 0.089 0.077 0.056 0.005 0.004 0.001 0.001 8:00 AM 0.079 0.194 0.289 0.225 0.211 0.183 0.133 0.013 0.010 0.002 0.002 9:00 AM 0.128 0.320 0.476 0.369 0.343 0.295 0.215 0.021 0.016 0.003 0.003 10:00AM 0.205 0.534 0.791 0.604 0.548 0.469 0.341 0.034 0.027 0.006 0.006 11:00AM 0.228 0.598 0.888 0.676 0.609 0.521 0.379 0.037 0.030 0.007 0.007 12:00 PM 0.196 0.503 0.750 0.573 0.519 0.445 0.324 0.032 0.025 0.006 0.006 1:00 PM 0.285 0.696 1.020 0.752 0.675 0.580 0.424 0.043 0.034 0.009 0.009 2:00 PM 0.363 0.950 1.272 0.935 0.847 0.728 0.530 0.055 0.045 0.013 0.013 3:00 PM 0.301 0.766 0.980 0.767 0.757 0.658 0.474 0.047 0.037 0.009 0.009 4:00 PM 0.240 0.489 0.669 0.572 0.634 0.559 0.396 0.036 0.027 0.004 0.004 5:00 PM 0.214 0.424 0.600 0.521 0.581 0.511 0.358 0.032 0.024 0.003 0.003 Table 4. 4 Simulation results of hourly solar heat gain on June 21 st . Figure 4. 7 Hourly solar heat gain on June 21 st based on 11 tilt angles. 0.000 0.200 0.400 0.600 0.800 1.000 1.200 1.400 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain 6/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 153 9/21 Indoor Average Illuminance Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 95.17 258.60 261.65 296.90 257.24 172.65 89.91 46.64 25.56 12.91 3.82 8:00 AM 231.98 1698.11 889.89 936.31 576.09 391.21 230.48 133.77 75.76 37.79 10.67 9:00 AM 347.34 3825.37 1040.57 1043.74 819.02 562.18 329.56 189.73 109.89 53.09 13.79 10:00AM 434.95 6268.42 1967.12 1285.82 993.07 686.79 408.89 245.90 144.24 69.70 17.24 11:00AM 497.86 6782.82 1543.61 1489.10 1117.63 732.97 427.06 252.62 149.55 74.86 19.31 12:00 PM 523.58 7242.05 1585.66 1505.82 1134.42 754.12 448.60 263.01 154.78 73.87 21.58 1:00 PM 493.27 6945.22 1512.26 1453.46 1100.80 741.71 437.33 262.32 157.81 78.50 20.75 2:00 PM 438.03 5272.34 1279.31 1270.71 986.95 678.76 407.55 243.88 142.13 71.19 18.05 3:00 PM 351.20 3139.00 1028.79 1026.29 817.71 554.60 324.46 192.25 111.98 53.90 15.09 4:00 PM 262.21 978.41 749.41 764.47 609.70 377.45 210.00 116.88 71.95 32.59 10.16 5:00 PM 156.70 296.52 433.37 445.29 359.48 211.48 102.65 59.89 28.96 14.92 5.03 Table 4. 5 Simulation results of hourly indoor average illuminance on September 21 st . Figure 4. 8 Hourly indoor average illuminances on September 21 st based on 11 tilt angles. 0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00 7000.00 8000.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Time Indoor Average Illuminance 9/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 154 9/21 Daylight Glare Probability Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 0.049 0.153 0.226 0.250 0.263 0.263 0.250 0.222 0.097 0.041 0.013 8:00 AM 0.226 0.236 0.273 0.307 0.317 0.308 0.286 0.257 0.233 0.201 0.045 9:00 AM 0.241 0.260 0.309 0.353 0.362 0.345 0.314 0.275 0.244 0.228 0.114 10:00AM 0.252 0.281 0.344 0.432 0.395 0.373 0.336 0.290 0.253 0.233 0.159 11:00AM 0.255 0.289 0.356 0.476 0.414 0.387 0.345 0.295 0.255 0.237 0.176 12:00 PM 0.258 0.295 0.364 0.490 0.419 0.391 0.348 0.298 0.258 0.238 0.173 1:00 PM 0.257 0.290 0.359 0.483 0.412 0.385 0.345 0.296 0.256 0.237 0.167 2:00 PM 0.251 0.278 0.338 0.389 0.389 0.368 0.332 0.289 0.253 0.233 0.151 3:00 PM 0.251 0.278 0.338 0.389 0.389 0.368 0.332 0.289 0.253 0.233 0.151 4:00 PM 0.219 0.232 0.267 0.304 0.319 0.308 0.284 0.255 0.231 0.204 0.053 5:00 PM 0.073 0.183 0.234 0.257 0.275 0.272 0.257 0.230 0.129 0.065 0.022 Table 4. 6 Simulation results of hourly Daylight Glare Probability on September 21 st . Figure 4. 9 Hourly Daylight Glare Probability on September 21 st based on 11 tilt angles. 0.000 0.100 0.200 0.300 0.400 0.500 0.600 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Time Daylight Glare Probability 9/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 155 9/21 Solar Heat Gain (kWh) Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 0.054 0.088 0.122 0.117 0.203 0.188 0.143 0.007 0.005 0.000 0.000 8:00 AM 0.242 0.484 1.122 0.915 0.585 0.443 0.306 0.022 0.016 0.001 0.001 9:00 AM 0.411 1.479 2.571 1.167 0.789 0.603 0.395 0.038 0.028 0.003 0.003 10:00AM 0.411 3.657 2.104 0.829 0.878 0.762 0.534 0.084 0.073 0.040 0.040 11:00AM 0.517 4.946 2.846 1.734 1.100 0.981 0.762 0.239 0.227 0.192 0.192 12:00 PM 0.781 3.686 5.708 1.777 1.090 0.976 0.769 0.204 0.192 0.158 0.158 1:00 PM 0.728 4.574 7.076 1.016 0.940 0.834 0.637 0.078 0.067 0.034 0.034 2:00 PM 0.589 5.502 5.018 1.748 1.076 0.967 0.762 0.230 0.218 0.185 0.185 3:00 PM 0.464 3.861 2.210 1.646 1.031 0.913 0.691 0.205 0.192 0.159 0.159 4:00 PM 0.477 2.428 2.141 0.961 0.871 0.719 0.483 0.048 0.037 0.005 0.005 5:00 PM 0.434 1.030 1.754 0.997 0.850 0.689 0.475 0.039 0.030 0.003 0.003 Table 4. 7 Simulation results of hourly solar heat gain on September 21 st . Figure 4. 1 0 Hourly solar heat gain on September 21 st based on 11 tilt angles. 0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000 8.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain 9/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 156 12/21 Indoor Average Illuminance Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 0 0 0 0 0 0 0 0 0 0 0 8:00 AM 78.77 169.72 326.44 577.32 826.08 548.27 231.37 104.16 48.65 20.70 5.55 9:00 AM 1211.81 1476.18 3119.54 4333.13 3964.04 1938.11 696.04 495.96 124.39 52.31 13.05 10:00AM 1989.54 2373.36 4514.19 7587.01 5841.11 2607.16 599.37 327.55 153.22 67.61 17.80 11:00AM 2722.62 3183.35 6030.59 10828.3 8410.91 1226.06 726.80 394.11 186.89 81.86 21.91 12:00PM 2742.42 3225.41 10655.1 10865.9 6199.78 1247.60 707.74 382.08 189.54 86.20 21.99 1:00 PM 2539.15 3021.75 4188.25 10067.8 9876.95 3776.58 1228.96 365.04 187.77 83.26 20.89 2:00 PM 2117.67 3827.10 6122.05 7653.19 4501.84 1106.07 676.43 375.49 175.56 81.02 20.67 3:00 PM 1227.51 1554.39 2062.17 2422.04 4532.07 4064.63 1420.43 982.60 591.71 288.47 244.39 4:00 PM 283.43 451.01 700.76 899.17 902.44 625.45 278.54 136.49 67.40 27.90 7.98 5:00 PM 0 0 0 0 0 0 0 0 0 0 0 Table 4. 8 Simulation results of hourly indoor average illuminance on December 21 st . Figure 4. 1 1 Hourly indoor average illuminances on December 21 st based on 11 tilt angles. 0 2000 4000 6000 8000 10000 12000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Time Indoor Average Illuminance 12/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 157 12/21 Daylight Glare Probability Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 0 0 0 0 0 0 0 0 0 0 0 8:00 AM 0.042 0.210 0.275 0.296 0.319 0.301 0.274 0.233 0.214 0.079 0.013 9:00 AM 0.228 0.260 0.280 0.340 0.383 0.365 0.318 0.285 0.258 0.233 0.089 10:00AM 0.245 0.281 0.329 0.425 0.513 0.440 0.368 0.313 0.270 0.244 0.169 11:00AM 0.255 0.301 0.372 0.495 0.599 0.478 0.396 0.333 0.281 0.248 0.197 12:00 PM 0.255 0.304 0.381 0.514 0.620 0.492 0.404 0.334 0.280 0.248 0.206 1:00 PM 0.252 0.298 0.367 0.487 0.591 0.480 0.396 0.331 0.280 0.249 0.196 2:00 PM 0.247 0.290 0.335 0.428 0.487 0.443 0.374 0.322 0.279 0.245 0.187 3:00 PM 0.235 0.267 0.292 0.352 0.398 0.377 0.328 0.294 0.264 0.239 0.136 4:00 PM 0.115 0.226 0.264 0.310 0.339 0.321 0.284 0.243 0.232 0.166 0.032 5:00 PM 0 0 0 0 0 0 0 0 0 0 0 Table 4. 9 Simulation results of hourly Daylight Glare Probability on December 21 st . Figure 4. 1 2 Hourly Daylight Glare Probability on December 21 st based on 11 tilt angles. 0 0.1 0.2 0.3 0.4 0.5 0.6 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Time Daylight Glare Probability 12/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 158 12/21 Solar Heat Gain (kWh) Tilt Time -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 AM 0 0 0 0 0 0 0 0 0 0 0 8:00 AM 0.567 0.689 0.712 1.094 1.270 1.157 0.075 0.008 0.005 0.000 0.000 9:00 AM 0.934 1.316 1.425 3.771 4.576 2.383 0.420 0.137 0.127 0.001 0.001 10:00AM 0.554 1.496 2.253 4.690 6.997 4.433 1.839 0.645 0.628 0.002 0.002 11:00AM 0.387 2.396 5.936 5.872 5.734 4.678 2.683 0.545 0.528 0.003 0.003 12:00 PM 0.425 3.223 10.283 8.360 2.646 2.024 1.644 0.056 0.042 0.003 0.003 1:00 PM 0.447 3.146 11.714 9.189 1.210 1.110 0.910 0.055 0.041 0.004 0.004 2:00 PM 0.426 3.247 8.733 7.664 3.998 2.808 1.885 0.174 0.159 0.003 0.003 3:00 PM 0.404 2.371 4.157 5.538 7.397 5.305 2.426 0.800 0.781 0.003 0.003 4:00 PM 0.884 1.631 2.415 4.537 6.474 4.506 1.226 0.680 0.663 0.002 0.002 5:00 PM 1.007 1.435 1.548 2.560 3.158 2.139 0.290 0.029 0.021 0.001 0.001 Table 4. 1 0 Simulation results of hourly solar heat gain on December 21 st . Figure 4. 1 3 Hourly solar heat gain on December 21 st based on 11 tilt angles. 0 2 4 6 8 10 12 14 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain 12/21 -75 -60 -45 -30 -15 0° 15° 30° 45° 60° 75° 159 6/21 Indoor Average Illuminance (Lux) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 15.83 51.17 104.87 172.60 242.38 308.66 365.65 397.83 411.67 8:00 AM 31.32 102.07 209.35 342.30 481.00 622.36 726.80 807.47 844.31 9:00 AM 45.24 151.23 304.42 505.34 714.73 917.14 1079.58 1196.80 1251.69 10:00 AM 73.75 264.38 542.51 922.43 1290.19 1679.94 1996.22 2232.71 2334.44 11:00 AM 61.70 211.85 424.18 704.49 1001.29 1275.79 1518.36 1678.68 1739.56 12:00 PM 61.61 202.23 418.39 708.35 1010.03 1304.51 1554.92 1706.57 1758.78 1:00 PM 87.92 287.44 607.75 1028.31 1656.42 2193.40 2913.76 3494.62 3905.38 2:00 PM 135.63 373.49 680.46 1102.82 1581.86 2114.02 3360.44 5862.55 7683.26 3:00 PM 104.82 268.13 465.78 702.42 957.39 1274.76 1599.64 2012.62 2627.02 4:00 PM 70.21 189.48 343.97 520.63 714.27 946.68 1199.74 1500.58 1887.82 5:00 PM 46.74 140.72 263.40 423.01 590.24 758.02 949.08 1158.09 1383.79 Table 4. 1 1 Simulation results of hourly indoor average illuminance on June 21 st . Figure 4. 1 4 Hourly indoor average illuminances on June 21st based on 9 tilt angles. 0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00 7000.00 8000.00 9000.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Time Indoor Average Illuminance Jun 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 160 6/21 Daylight Glare Probability (DGP) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 0.005 0.010 0.023 0.048 0.089 0.119 0.139 0.151 0.151 8:00 AM 0.008 0.031 0.098 0.157 0.179 0.184 0.188 0.189 0.190 9:00 AM 0.012 0.073 0.159 0.182 0.190 0.196 0.200 0.202 0.204 10:00 AM 0.032 0.164 0.188 0.200 0.211 0.222 0.231 0.242 0.278 11:00 AM 0.018 0.129 0.180 0.190 0.199 0.206 0.212 0.216 0.218 12:00 PM 0.019 0.132 0.179 0.189 0.232 0.230 0.230 0.231 0.232 1:00 PM 0.039 0.173 0.191 0.205 0.219 0.234 0.248 0.258 0.270 2:00 PM 0.067 0.182 0.198 0.214 0.230 0.248 0.267 0.290 0.313 3:00 PM 0.026 0.156 0.185 0.195 0.205 0.217 0.230 0.241 0.255 4:00 PM 0.016 0.107 0.174 0.187 0.194 0.203 0.214 0.222 0.231 5:00 PM 0.012 0.070 0.156 0.180 0.188 0.195 0.202 0.209 0.214 Table 4. 1 2 Simulation results of hourly Daylight Glare Probability on June 21 st . Figure 4. 1 5 Hourly Daylight Glare Probability on June 21st based on 9 tilt angles. 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Time Daylight Glare Probability Jun 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 161 6/21 Solar Heat Gain (kWh) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 0.137 0.276 0.417 0.562 0.665 0.743 0.832 0.862 0.880 8:00 AM 0.319 0.665 0.993 1.321 1.615 1.793 1.972 2.097 2.163 9:00 AM 0.536 1.109 1.625 2.205 2.669 2.950 3.277 3.493 3.615 10:00 AM 0.911 1.806 2.717 3.687 4.435 5.004 5.722 5.963 6.218 11:00 AM 1.044 2.055 3.130 4.147 5.097 5.864 6.407 6.823 7.042 12:00 PM 0.882 1.816 2.748 3.632 4.473 4.886 5.583 5.699 5.988 1:00 PM 1.125 2.226 3.482 4.705 6.209 7.174 8.220 9.217 9.834 2:00 PM 1.225 2.519 3.843 5.932 7.800 10.274 13.921 16.833 20.495 3:00 PM 0.795 1.680 2.768 4.670 5.651 7.406 11.168 16.265 22.773 4:00 PM 0.495 1.400 1.862 2.338 2.926 3.875 4.560 6.565 10.422 5:00 PM 0.487 0.936 1.357 1.809 2.166 2.464 2.868 3.161 4.404 Table 4. 1 3 Simulation results of hourly solar heat gain on June 21 st . Figure 4. 1 6 Hourly solar heat gain on June 21st based on 9 tilt angles. 0.000 5.000 10.000 15.000 20.000 25.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain Jun 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 162 Table 4. 1 4 Simulation results of hourly indoor average illuminance on September 21 st . Figure 4. 1 7 Hourly indoor average illuminances on September 21st based on 9 tilt angles. 0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00 7000.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Time Indoor Average Illuminance Sep 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 9/21 Indoor Average Illuminance (Lux) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 17.43 48.97 89.49 130.74 173.18 212.10 257.18 314.71 386.98 8:00 AM 42.81 115.07 202.03 304.29 406.04 522.61 671.35 857.19 1065.73 9:00 AM 70.17 187.15 325.48 485.19 655.85 851.94 1087.72 1340.60 1728.03 10:00 AM 99.97 251.29 437.78 637.98 866.02 1120.87 1434.08 1813.38 2343.62 11:00 AM 121.65 305.59 537.07 793.79 1095.41 1437.25 1796.66 3932.71 6113.51 12:00 PM 127.26 323.12 561.76 839.49 1143.00 1502.83 1897.63 4126.03 5622.22 1:00 PM 118.49 296.72 513.60 769.05 1060.71 1373.31 1732.96 3035.60 4529.35 2:00 PM 93.84 241.02 418.59 622.18 855.93 1115.86 1404.82 1754.45 2266.31 3:00 PM 67.26 180.84 322.21 487.31 673.24 865.99 1125.97 1399.64 1756.58 4:00 PM 45.93 133.62 261.43 410.57 570.33 741.95 922.48 1121.68 1347.53 5:00 PM 25.84 79.76 164.49 266.49 372.35 485.39 594.38 699.76 798.16 163 Table 4. 1 5 Simulation results of hourly Daylight Glare Probability on September 21 st . Figure 4. 18 Hourly Daylight Glare Probability on September 21st based on 9 tilt angles. 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Time Daylight Glare Probability Sep 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 9/21 Daylight Glare Probability (DGP) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 0.005 0.009 0.019 0.036 0.061 0.089 0.123 0.149 0.165 8:00 AM 0.012 0.067 0.150 0.176 0.185 0.190 0.197 0.205 0.210 9:00 AM 0.027 0.152 0.183 0.192 0.202 0.212 0.224 0.231 0.243 10:00 AM 0.055 0.177 0.192 0.206 0.220 0.235 0.250 0.269 0.280 11:00 AM 0.088 0.185 0.201 0.218 0.236 0.255 0.274 0.290 0.305 12:00 PM 0.104 0.187 0.203 0.222 0.241 0.261 0.282 0.298 0.312 1:00 PM 0.085 0.184 0.200 0.216 0.230 0.247 0.266 0.289 0.303 2:00 PM 0.048 0.176 0.192 0.204 0.214 0.231 0.246 0.259 0.275 3:00 PM 0.026 0.154 0.183 0.193 0.202 0.212 0.223 0.235 0.246 4:00 PM 0.018 0.115 0.175 0.186 0.193 0.200 0.203 0.211 0.222 5:00 PM 0.009 0.043 0.120 0.165 0.179 0.184 0.188 0.191 0.192 164 9/21 Solar Heat Gain (kWh) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 0.335 0.598 0.467 0.400 0.409 0.404 0.403 0.396 0.389 8:00 AM 0.210 0.405 0.560 0.705 0.819 0.901 1.051 1.123 1.205 9:00 AM 0.333 0.750 1.033 1.264 1.504 1.649 1.826 2.097 2.303 10:00 AM 0.465 0.980 1.366 1.839 2.197 2.487 3.198 4.534 5.972 11:00 AM 0.531 1.183 1.662 2.457 2.897 3.547 5.722 8.199 11.814 12:00 PM 0.775 1.444 2.219 3.391 4.726 6.294 9.095 11.605 18.870 1:00 PM 0.754 1.397 2.178 2.745 4.179 5.567 8.353 12.629 18.463 2:00 PM 0.544 1.081 1.610 2.527 3.352 4.629 6.517 10.419 15.210 3:00 PM 0.450 0.891 1.276 1.800 2.239 3.035 4.352 5.883 8.294 4:00 PM 0.449 0.875 1.261 1.653 2.189 2.578 2.982 3.368 3.936 5:00 PM 0.549 1.044 1.399 1.759 2.120 2.329 2.586 2.794 3.048 Table 4. 16 Simulation results of hourly solar heat gain on September 21 st . Figure 4. 19 Hourly solar heat gain on September 21st based on 9 tilt angles. 0.000 2.000 4.000 6.000 8.000 10.000 12.000 14.000 16.000 18.000 20.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain Sep 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 165 12/21 Indoor Average Illuminance (Lux) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 0 0 0 0 0 0 0 0 0 8:00 AM 9.26 27.64 49.44 73.11 95.42 117.45 142.72 178.61 222.56 9:00 AM 28.69 79.70 146.06 219.79 298.47 382.44 486.27 625.78 780.44 10:00 AM 46.51 124.24 221.33 325.28 445.67 575.46 740.77 928.81 1181.05 11:00 AM 58.94 152.00 266.81 391.55 531.00 679.29 891.94 1139.23 1457.69 12:00 PM 62.95 167.20 295.73 441.58 592.83 757.31 1001.00 1242.10 1595.83 1:00 PM 59.40 155.79 278.17 418.99 574.45 730.92 950.43 1192.84 1503.59 2:00 PM 49.66 129.84 241.50 364.10 501.24 645.92 839.12 1057.58 1339.34 3:00 PM 36.47 105.09 197.13 311.98 430.32 573.07 722.53 892.74 1071.79 4:00 PM 23.57 75.97 146.69 237.05 338.07 430.03 529.07 618.90 703.27 5:00 PM 0 0 0 0 0 0 0 0 0 Table 4. 1 7 Simulation results of hourly indoor average illuminance on December 21 st . Figure 4. 2 0 Hourly indoor average illuminances on December 21st based on 9 tilt angles. 0 200 400 600 800 1000 1200 1400 1600 1800 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Time Indoor Average Illuminance Dec 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 166 12/21 Daylight Glare Probability (DGP) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 0 0 0 0 0 0 0 0 0 8:00 AM 0.004 0.006 0.009 0.015 0.023 0.034 0.049 0.069 0.095 9:00 AM 0.006 0.023 0.071 0.137 0.168 0.179 0.185 0.191 0.196 10:00 AM 0.010 0.057 0.148 0.177 0.186 0.193 0.200 0.208 0.217 11:00 AM 0.013 0.092 0.170 0.184 0.193 0.202 0.211 0.223 0.235 12:00 PM 0.015 0.110 0.175 0.187 0.196 0.206 0.217 0.228 0.241 1:00 PM 0.013 0.098 0.172 0.185 0.193 0.202 0.212 0.223 0.234 2:00 PM 0.011 0.070 0.158 0.180 0.188 0.196 0.204 0.214 0.222 3:00 PM 0.009 0.040 0.123 0.169 0.181 0.188 0.194 0.200 0.206 4:00 PM 0.006 0.019 0.060 0.120 0.157 0.174 0.180 0.185 0.187 5:00 PM 0 0 0 0 0 0 0 0 0 Table 4. 18 Simulation results of hourly Daylight Glare Probability on December 21 st . Figure 4. 2 1 Hourly Daylight Glare Probability on December 21st based on 9 tilt angles. 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Time Daylight Glare Probability Dec 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 167 12/21 Solar Heat Gain (kWh) Tilt Time 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM 0 0 0 0 0 0 0 0 0 8:00 AM 0.038 0.072 0.099 0.118 0.132 0.144 0.150 0.158 0.163 9:00 AM 0.141 0.277 0.385 0.459 0.521 0.574 0.624 0.680 0.713 10:00 AM 0.264 0.489 0.681 0.826 0.977 1.058 1.191 1.291 1.358 11:00 AM 0.282 0.565 0.807 1.006 1.166 1.304 1.558 1.708 1.864 12:00 PM 0.365 0.677 0.948 1.283 1.758 1.893 2.073 2.331 2.813 1:00 PM 0.350 0.863 1.161 1.435 1.700 1.963 2.289 2.656 3.311 2:00 PM 0.342 0.661 0.941 1.163 1.395 1.791 2.001 2.256 2.932 3:00 PM 0.318 0.622 0.963 1.159 1.373 1.588 1.827 2.011 2.277 4:00 PM 0.308 0.615 0.889 1.104 1.320 1.475 1.630 1.748 1.893 5:00 PM 0.316 0.619 0.892 1.113 1.353 1.517 1.761 1.926 2.106 Table 4. 19 Simulation results of hourly solar heat gain on December 21 st . Figure 4. 2 2 Hourly solar heat gain on December 21st based on 9 tilt angles. 0 0.5 1 1.5 2 2.5 3 3.5 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain Dec 21st 10° 20° 30° 40° 50° 60° 70° 80° 90° 168 APPENDIX B: Control Decision and Evaluation Process. Legend Imperceptible glare (DGP ≤ 0.35) Recommended illuminance (500 lux – 1500 lux) Selected hourly solar heat gain Annotation Glare: daylight glare probability E: illuminance SHG: solar heat gain 169 6/21 Control Strategy of Kinetic Louver based on Three Metrics Tilt Time Metrics -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 Glare 0.010 0.020 0.084 0.200 0.228 0.226 0.154 0.037 0.012 0.010 0.006 AM E 51.75 104.88 144.35 140.09 105.24 55.53 22.91 12.49 7.71 3.69 1.05 SHG 0.033 0.081 0.121 0.094 0.089 0.077 0.056 0.005 0.004 0.001 0.001 8:00 Glare 0.027 0.085 0.215 0.241 0.253 0.255 0.247 0.155 0.036 0.023 0.011 AM E 104.71 213.26 292.86 282.83 210.04 108.45 45.28 25.65 15.54 8.11 2.03 SHG 0.079 0.194 0.289 0.225 0.211 0.183 0.133 0.013 0.010 0.002 0.002 9:00 Glare 0.054 0.168 0.233 0.256 0.272 0.272 0.260 0.222 0.090 0.051 0.017 AM E 155.06 311.93 425.35 417.55 305.35 155.70 69.40 37.99 23.81 11.95 3.24 SHG 0.128 0.320 0.476 0.369 0.343 0.295 0.215 0.021 0.016 0.003 0.003 10:00 Glare 0.180 0.223 0.256 0.289 0.313 0.307 0.284 0.247 0.207 0.170 0.045 AM E 282.13 567.54 770.18 741.86 514.12 273.07 120.55 69.59 42.03 21.63 6.24 SHG 0.205 0.534 0.791 0.604 0.548 0.469 0.341 0.034 0.027 0.006 0.006 11:00 Glare 0.114 0.209 0.244 0.272 0.292 0.289 0.271 0.238 0.161 0.099 0.031 AM E 204.30 429.48 577.09 559.49 412.48 212.01 91.74 51.94 34.49 17.14 5.23 SHG 0.228 0.598 0.888 0.676 0.609 0.521 0.379 0.037 0.030 0.007 0.007 12:00 Glare 0.078 0.195 0.237 0.268 0.287 0.283 0.265 0.229 0.101 0.056 0.025 PM E 212.07 416.87 565.40 537.97 384.32 192.66 68.43 30.27 18.50 9.45 2.85 SHG 0.196 0.503 0.750 0.573 0.519 0.445 0.324 0.032 0.025 0.006 0.006 1:00 Glare 0.220 0.235 0.273 0.308 0.334 0.325 0.295 0.255 0.225 0.206 0.073 PM E 396.52 735.99 935.36 867.43 617.37 343.86 152.93 88.64 56.55 27.97 7.31 SHG 0.285 0.696 1.020 0.752 0.675 0.580 0.424 0.043 0.034 0.009 0.009 2:00 Glare 0.254 0.260 0.311 0.333 0.357 0.351 0.316 0.273 0.240 0.230 0.126 PM E 517.73 822.55 985.96 926.24 710.90 443.30 260.17 177.56 118.20 57.41 16.22 SHG 0.363 0.950 1.272 0.935 0.847 0.728 0.530 0.055 0.045 0.013 0.013 3:00 Glare 0.249 0.249 0.275 0.305 0.330 0.331 0.305 0.266 0.234 0.214 0.074 PM E 254.17 456.70 582.88 614.99 521.44 368.03 227.43 155.69 102.42 47.47 11.80 SHG 0.301 0.766 0.980 0.767 0.757 0.658 0.474 0.047 0.037 0.009 0.009 4:00 Glare 0.197 0.233 0.258 0.290 0.313 0.315 0.292 0.257 0.223 0.178 0.049 PM E 169.36 339.09 474.11 518.73 454.67 298.51 171.49 110.34 72.87 32.52 8.38 SHG 0.240 0.489 0.669 0.572 0.634 0.559 0.396 0.036 0.027 0.004 0.004 5:00 Glare 0.121 0.213 0.245 0.275 0.296 0.296 0.277 0.244 0.183 0.111 0.024 PM E 154.28 301.85 435.63 463.16 379.49 227.68 118.21 68.58 45.32 22.21 4.95 SHG 0.214 0.424 0.600 0.521 0.581 0.511 0.358 0.032 0.024 0.003 0.003 Table 4. 2 2 Control strategy determination of kinetic louver of the shoebox model on June 21 st 170 9/21 Control Strategy of Kinetic Louver based on Three Metric Tilt Time Metrics -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 Glare 0.049 0.153 0.226 0.250 0.263 0.263 0.250 0.222 0.097 0.041 0.013 AM E 95.17 258.60 261.65 296.90 257.24 172.65 89.91 46.64 25.56 12.91 3.82 SHG 0.054 0.088 0.122 0.117 0.203 0.188 0.143 0.007 0.005 0.000 0.000 8:00 Glare 0.226 0.236 0.273 0.307 0.317 0.308 0.286 0.257 0.233 0.201 0.045 AM E 231.98 1698.11 889.89 936.31 576.09 391.21 230.48 133.77 75.76 37.79 10.67 SHG 0.242 0.484 1.122 0.915 0.585 0.443 0.306 0.022 0.016 0.001 0.001 9:00 Glare 0.241 0.260 0.309 0.353 0.362 0.345 0.314 0.275 0.244 0.228 0.114 AM E 347.34 3825.37 1040.57 1043.74 819.02 562.18 329.56 189.73 109.89 53.09 13.79 SHG 0.411 1.479 2.571 1.167 0.789 0.603 0.395 0.038 0.028 0.003 0.003 10:00 Glare 0.252 0.281 0.344 0.432 0.395 0.373 0.336 0.290 0.253 0.233 0.159 AM E 434.95 6268.42 1967.12 1285.82 993.07 686.79 408.89 245.90 144.24 69.70 17.24 SHG 0.411 3.657 2.104 0.829 0.878 0.762 0.534 0.084 0.073 0.040 0.040 11:00 Glare 0.255 0.289 0.356 0.476 0.414 0.387 0.345 0.295 0.255 0.237 0.176 AM E 497.86 6782.82 1543.61 1489.10 1117.63 732.97 427.06 252.62 149.55 74.86 19.31 SHG 0.517 4.946 2.846 1.734 1.100 0.981 0.762 0.239 0.227 0.192 0.192 12:00 Glare 0.258 0.295 0.364 0.490 0.419 0.391 0.348 0.298 0.258 0.238 0.173 PM E 523.58 7242.05 1585.66 1505.82 1134.42 754.12 448.60 263.01 154.78 73.87 21.58 SHG 0.781 3.686 5.708 1.777 1.090 0.976 0.769 0.204 0.192 0.158 0.158 1:00 Glare 0.257 0.290 0.359 0.483 0.412 0.385 0.345 0.296 0.256 0.237 0.167 PM E 493.27 6945.22 1512.26 1453.46 1100.80 741.71 437.33 262.32 157.81 78.50 20.75 SHG 0.728 4.574 7.076 1.016 0.940 0.834 0.637 0.078 0.067 0.034 0.034 2:00 Glare 0.251 0.278 0.338 0.389 0.389 0.368 0.332 0.289 0.253 0.233 0.151 PM E 438.03 5272.34 1279.31 1270.71 986.95 678.76 407.55 243.88 142.13 71.19 18.05 SHG 0.589 5.502 5.018 1.748 1.076 0.967 0.762 0.230 0.218 0.185 0.185 3:00 Glare 0.251 0.278 0.338 0.389 0.389 0.368 0.332 0.289 0.253 0.233 0.151 PM E 351.20 3139.00 1028.79 1026.29 817.71 554.60 324.46 192.25 111.98 53.90 15.09 SHG 0.464 3.861 2.210 1.646 1.031 0.913 0.691 0.205 0.192 0.159 0.159 4:00 Glare 0.219 0.232 0.267 0.304 0.319 0.308 0.284 0.255 0.231 0.204 0.053 PM E 262.21 978.41 749.41 764.47 609.70 377.45 210.00 116.88 71.95 32.59 10.16 SHG 0.477 2.428 2.141 0.961 0.871 0.719 0.483 0.048 0.037 0.005 0.005 5:00 Glare 0.073 0.183 0.234 0.257 0.275 0.272 0.257 0.230 0.129 0.065 0.022 PM E 156.70 296.52 433.37 445.29 359.48 211.48 102.65 59.89 28.96 14.92 5.03 SHG 0.434 1.030 1.754 0.997 0.850 0.689 0.475 0.039 0.030 0.003 0.003 Table 4. 2 3 Control strategy determination of kinetic louver of the shoebox model on September 21 st . 171 12/21 Control Strategy of Kinetic Louver based on Three Metrics Tilt Time Metrics -75° -60° -45° -30° -15° 0° 15° 30° 45° 60° 75° 7:00 Glare 0 0 0 0 0 0 0 0 0 0 0 AM E 0 0 0 0 0 0 0 0 0 0 0 SHG 0 0 0 0 0 0 0 0 0 0 0 8:00 Glare 0.042 0.210 0.275 0.296 0.319 0.301 0.274 0.233 0.214 0.079 0.013 AM E 78.77 169.72 326.44 577.32 826.08 548.27 231.37 104.16 48.65 20.70 5.55 SHG 0.567 0.689 0.712 1.094 1.270 1.157 0.075 0.008 0.005 0.000 0.000 9:00 Glare 0.228 0.260 0.280 0.340 0.383 0.365 0.318 0.285 0.258 0.233 0.089 AM E 1211.8 1476.2 3119.54 4333.13 3964.0 1938.1 696.04 495.96 124.39 52.31 13.05 SHG 0.934 1.316 1.425 3.771 4.576 2.383 0.420 0.137 0.127 0.001 0.001 10:00 Glare 0.245 0.281 0.329 0.425 0.513 0.440 0.368 0.313 0.270 0.244 0.169 AM E 1989.5 2373.4 4514.19 7587.01 5841.1 2607.2 599.37 327.55 153.22 67.61 17.80 SHG 0.554 1.496 2.253 4.690 6.997 4.433 1.839 0.645 0.628 0.002 0.002 11:00 Glare 0.255 0.301 0.372 0.495 0.599 0.478 0.396 0.333 0.281 0.248 0.197 AM E 2722.6 3183.4 6030.59 10828.3 8410.9 1226.1 726.80 394.11 186.89 81.86 21.91 SHG 0.387 2.396 5.936 5.872 5.734 4.678 2.683 0.545 0.528 0.003 0.003 12:00 Glare 0.255 0.304 0.381 0.514 0.620 0.492 0.404 0.334 0.280 0.248 0.206 PM E 2742.4 3225.4 10655.1 10865.9 6199.8 1247.6 707.74 382.08 189.54 86.20 21.99 SHG 0.425 3.223 10.283 8.360 2.646 2.024 1.644 0.056 0.042 0.003 0.003 1:00 Glare 0.252 0.298 0.367 0.487 0.591 0.480 0.396 0.331 0.280 0.249 0.196 PM E 2539.2 3021.8 4188.25 10067.8 9876.9 3776.6 1228.96 365.04 187.77 83.26 20.89 SHG 0.447 3.146 11.714 9.189 1.210 1.110 0.910 0.055 0.041 0.004 0.004 2:00 Glare 0.247 0.290 0.335 0.428 0.487 0.443 0.374 0.322 0.279 0.245 0.187 PM E 2117.7 3827.1 6122.05 7653.19 4501.8 1106.1 676.43 375.49 175.56 81.02 20.67 SHG 0.426 3.247 8.733 7.664 3.998 2.808 1.885 0.174 0.159 0.003 0.003 3:00 Glare 0.235 0.267 0.292 0.352 0.398 0.377 0.328 0.294 0.264 0.239 0.136 PM E 1227.5 1554.4 2062.17 2422.04 4532.1 4064.6 1420.43 982.60 591.71 288.47 244.39 SHG 0.404 2.371 4.157 5.538 7.397 5.305 2.426 0.800 0.781 0.003 0.003 4:00 Glare 0.115 0.226 0.264 0.310 0.339 0.321 0.284 0.243 0.232 0.166 0.032 PM E 283.43 451.01 700.76 899.17 902.44 625.45 278.54 136.49 67.40 27.90 7.98 SHG 0.884 1.631 2.415 4.537 6.474 4.506 1.226 0.680 0.663 0.002 0.002 5:00 Glare 0 0 0 0 0 0 0 0 0 0 0 PM E 0 0 0 0 0 0 0 0 0 0 0 SHG 1.007 1.435 1.548 2.560 3.158 2.139 0.290 0.029 0.021 0.001 0.001 Table 4. 2 4 Control strategy determination of kinetic louver of the shoebox model on December 21 st . 172 6/21 Kinetic Louver Fixed Louver Static Glazing Time Tilt Metrics Value Score Tilt Metrics Value Score Tilt Metrics Value Score 7:00 AM 0.084 1 0.226 1 0.229 1 -45° 144.35 0 0° 55.53 0 - 268.63 0 0.121 1 0.077 1 0.206 1 8:00 AM 0.215 1 0.255 1 0.256 1 -45° 292.86 0 0° 108.45 0 - 539.18 1 0.289 1 0.183 1 0.492 1 9:00 AM 0.233 1 0.272 1 0.278 1 -45° 425.35 0 0° 155.70 0 - 793.66 1 0.476 0.5 0.295 0.5 0.809 1 10:00 AM 0.313 1 0.307 1 0.328 1 -15° 514.12 1 0° 273.07 0 - 1430.11 0.5 0.548 0.5 0.469 1 1.340 0 11:00 AM 0.272 1 0.289 1 0.302 1 -30° 559.49 1 0° 212.01 0 - 1081.88 0.5 0.676 1 0.521 1 1.501 0 12:00 PM 0.268 1 0.283 1 0.295 1 -30° 537.97 1 0° 192.66 0 - 1038.73 0.5 0.573 0.5 0.445 1 1.268 0 1:00 PM 0.334 1 0.325 1 0.351 0.5 -15° 617.37 1 0° 343.86 0 - 1791.03 0 0.675 1 0.580 1 1.733 0 2:00 PM 0.254 1 0.351 0.5 0.383 0.5 -75° 517.73 1 0° 443.30 0 - 1893.27 0 0.363 1 0.728 0.5 2.242 0 3:00 PM 0.330 1 0.331 1 0.352 1 -15° 521.44 1 0° 368.03 0 - 1155.28 0.5 0.757 1 0.658 1 1.790 0 4:00 PM 0.290 1 0.315 1 0.332 1 -30° 518.73 1 0° 298.51 0 - 954.66 0.5 0.572 1 0.559 1 1.227 0 5:00 PM 0.275 1 0.296 1 0.306 1 -30° 463.16 0 0° 227.68 0 - 843.77 1 0.521 1 0.511 1 1.096 0 27.5 20.5 18.5 Table 4. 2 5 Evaluation among three cases of the shoebox model on June 21 st . 173 9/21 Kinetic Louver Fixed Louver Static Glazing Time Tilt Metrics Value Score Tilt Metrics Value Score Tilt Metrics Value Score 7:00 AM 0.250 1 0.263 1 0.327 1 -30° 296.90 0 0° 172.65 0 - 2277.37 0 0.117 1 0.188 1 0.286 1 8:00 AM 0.317 1 0.308 1 0.380 0.5 -15° 576.09 1 0° 391.21 0 - 4711.79 0 0.585 0.5 0.443 1 1.598 0 9:00 AM 0.345 1 0.345 1 0.420 0 0° 562.18 1 0° 562.18 1 - 7311.69 0 0.603 1 0.603 1 3.492 0 10:00 AM 0.373 0.5 0.373 0.5 0.445 0 0° 686.79 1 0° 686.79 1 - 8047.79 0 0.762 1 0.762 1 5.297 0 11:00 AM 0.387 0.5 0.387 0.5 0.452 0 0° 732.97 1 0° 732.97 1 - 8532.55 0 0.981 1 0.981 1 6.775 0 12:00 PM 0.391 0.5 0.391 0.5 0.441 0 0° 754.12 1 0° 754.12 1 - 8183.02 0 0.976 1 0.976 1 7.747 0 1:00 PM 0.385 0.5 0.385 0.5 0.412 0 0° 741.71 1 0° 741.71 1 - 6314.95 0 0.834 1 0.834 1 8.359 0 2:00 PM 0.368 0.5 0.368 0.5 0.372 0.5 0° 678.76 1 0° 678.76 1 - 4004.57 0 0.967 1 0.967 1 7.724 0 3:00 PM 0.368 0.5 0.368 0.5 0.329 1 0° 554.60 1 0° 554.60 1 - 1687.48 0 0.913 1 0.913 1 5.795 0 4:00 PM 0.319 1 0.308 1 0.279 1 -15° 609.70 1 0° 377.45 0 - 747.80 0.5 0.871 1 0.719 1 4.153 0 5:00 PM 0.257 1 0.272 1 0.266 1 -30° 445.29 0 0° 211.48 0 - 530.91 1 0.997 1 0.689 1 2.630 0 27.5 26 7.5 Table 4. 2 6 Evaluation among three cases of the shoebox model on September 21 st . 174 12/21 Kinetic Louver Fixed Louver Static Glazing Time Tilt Metrics Value Score Tilt Metrics Value Score Tilt Metrics Value Score 7:00 AM 0 0 0 0 0 0 - 0 0 0° 0 0 - 0 0 0 0 0 0 0 0 8:00 AM 0.301 1 0.301 1 0.325 1 0° 548.27 1 0° 548.27 1 - 963.37 0 1.157 1 1.157 1 1.423 1 9:00 AM 0.318 1 0.365 0.5 0.411 0 15° 696.04 1 0° 1938.11 0 - 4961.65 0 0.420 0 2.383 0.5 5.108 1 10:00 AM 0.368 0.5 0.440 0 0.533 0 15° 599.37 1 0° 2607.16 0 - 8142.24 0 1.839 0 4.433 0.5 8.780 1 11:00 AM 0.396 0.5 0.478 0 0.621 0 15° 726.80 1 0° 1226.06 0.5 - 11446.7 0 2.683 0 4.678 0.5 11.400 1 12:00 PM 0.404 0.5 0.492 0 0.649 0 15° 707.74 1 0° 1247.60 0.5 - 11590.41 0 1.644 0.5 2.024 0.5 12.651 1 1:00 PM 0.396 0.5 0.480 0 0.614 0 15° 1228.96 1 0° 3776.58 0 - 10747.2 0 0.910 0 1.110 0.5 12.641 1 2:00 PM 0.374 0.5 0.443 0 0.533 0 15° 676.43 1 0° 1106.07 0.5 - 8253.45 0 1.885 0 2.808 0.5 12.452 1 3:00 PM 0.264 1 0.377 0.5 0.428 0 45° 591.71 1 0° 4064.63 0 - 5230.27 0 0.781 0 5.305 0.5 11.277 1 4:00 PM 0.321 1 0.321 1 0.349 1 0° 625.45 1 0° 625.45 1 - 1288.30 0 4.506 0.5 4.506 0.5 8.110 1 5:00 PM 0 1 0 1 0 1 -15° 0 0 0° 0 0 - 0 0 3.158 0.5 2.139 0 3.746 1 19 12.5 13 Table 4. 2 7 Evaluation among three cases of the shoebox model on December 21 st . 175 6/21 Control Strategy of Kinetic Shading Panel based on Three Metrics Tilt Time Metrics 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM Glare 0.005 0.010 0.023 0.048 0.089 0.119 0.139 0.151 0.151 E 15.83 51.17 104.87 172.60 242.38 308.66 365.65 397.83 411.67 SHG 0.137 0.276 0.417 0.562 0.665 0.743 0.832 0.862 0.880 8:00 AM Glare 0.008 0.031 0.098 0.157 0.179 0.184 0.188 0.189 0.190 E 31.32 102.07 209.35 342.30 481.00 622.36 726.80 807.47 844.31 SHG 0.319 0.665 0.993 1.321 1.615 1.793 1.972 2.097 2.163 9:00 AM Glare 0.012 0.073 0.159 0.182 0.190 0.196 0.200 0.202 0.204 E 45.24 151.23 304.42 505.34 714.73 917.14 1079.58 1196.80 1251.69 SHG 0.536 1.109 1.625 2.205 2.669 2.950 3.277 3.493 3.615 10:00 AM Glare 0.032 0.164 0.188 0.200 0.211 0.222 0.231 0.242 0.278 E 73.75 264.38 542.51 922.43 1290.19 1679.94 1996.22 2232.71 2334.44 SHG 0.911 1.806 2.717 3.687 4.435 5.004 5.722 5.963 6.218 11:00 AM Glare 0.018 0.129 0.180 0.190 0.199 0.206 0.212 0.216 0.218 E 61.70 211.85 424.18 704.49 1001.29 1275.79 1518.36 1678.68 1739.56 SHG 1.044 2.055 3.130 4.147 5.097 5.864 6.407 6.823 7.042 12:00 PM Glare 0.019 0.132 0.179 0.189 0.232 0.230 0.230 0.231 0.232 E 61.61 202.23 418.39 708.35 1010.03 1304.51 1554.92 1706.57 1758.78 SHG 0.882 1.816 2.748 3.632 4.473 4.886 5.583 5.699 5.988 1:00 PM Glare 0.039 0.173 0.191 0.205 0.219 0.234 0.248 0.258 0.270 E 87.92 287.44 607.75 1028.31 1656.42 2193.40 2913.76 3494.62 3905.38 SHG 1.125 2.226 3.482 4.705 6.209 7.174 8.220 9.217 9.834 2:00 PM Glare 0.067 0.182 0.198 0.214 0.230 0.248 0.267 0.290 0.313 E 135.63 373.49 680.46 1102.82 1581.86 2114.02 3360.44 5862.55 7683.26 SHG 1.225 2.519 3.843 5.932 7.800 10.274 13.921 16.833 20.495 3:00 PM Glare 0.026 0.156 0.185 0.195 0.205 0.217 0.230 0.241 0.255 E 104.82 268.13 465.78 702.42 957.39 1274.76 1599.64 2012.62 2627.02 SHG 0.795 1.680 2.768 4.670 5.651 7.406 11.168 16.265 22.773 4:00 PM Glare 0.016 0.107 0.174 0.187 0.194 0.203 0.214 0.222 0.231 E 70.21 189.48 343.97 520.63 714.27 946.68 1199.74 1500.58 1887.82 SHG 0.495 1.400 1.862 2.338 2.926 3.875 4.560 6.565 10.422 5:00 PM Glare 0.012 0.070 0.156 0.180 0.188 0.195 0.202 0.209 0.214 E 46.74 140.72 263.40 423.01 590.24 758.02 949.08 1158.09 1383.79 SHG 0.487 0.936 1.357 1.809 2.166 2.464 2.868 3.161 4.404 Table 4. 28 Control strategy determination of kinetic shading panel of the cuboid model on June 21 st . 176 9/21 Control Strategy of Kinetic Shading Panel based on Three Metrics Tilt Time Metrics 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM Glare 0.005 0.009 0.019 0.036 0.061 0.089 0.123 0.149 0.165 E 17.43 48.97 89.49 130.74 173.18 212.10 257.18 314.71 386.98 SHG 0.335 0.598 0.467 0.400 0.409 0.404 0.403 0.396 0.389 8:00 AM Glare 0.01 0.07 0.15 0.18 0.18 0.19 0.20 0.21 0.21 E 42.81 115.07 202.03 304.29 406.04 522.61 671.35 857.19 1065.73 SHG 0.21 0.41 0.56 0.71 0.82 0.90 1.05 1.12 1.20 9:00 AM Glare 0.03 0.15 0.18 0.19 0.20 0.21 0.22 0.23 0.24 E 70.17 187.15 325.48 485.19 655.85 851.94 1087.72 1340.60 1728.03 SHG 0.33 0.75 1.03 1.26 1.50 1.65 1.83 2.10 2.30 10:00 AM Glare 0.06 0.18 0.19 0.21 0.22 0.24 0.25 0.27 0.28 E 99.97 251.29 437.78 637.98 866.02 1120.87 1434.08 1813.38 2343.62 SHG 0.47 0.98 1.37 1.84 2.20 2.49 3.20 4.53 5.97 11:00 AM Glare 0.09 0.18 0.20 0.22 0.24 0.25 0.27 0.29 0.31 E 121.65 305.59 537.07 793.79 1095.41 1437.25 1796.66 3932.71 6113.51 SHG 0.53 1.18 1.66 2.46 2.90 3.55 5.72 8.20 11.81 12:00 PM Glare 0.10 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.31 E 127.26 323.12 561.76 839.49 1143.00 1502.83 1897.63 4126.03 5622.22 SHG 0.77 1.44 2.22 3.39 4.73 6.29 9.09 11.61 18.87 1:00 PM Glare 0.08 0.18 0.20 0.22 0.23 0.25 0.27 0.29 0.30 E 118.49 296.72 513.60 769.05 1060.71 1373.31 1732.96 3035.60 4529.35 SHG 0.75 1.40 2.18 2.75 4.18 5.57 8.35 12.63 18.46 2:00 PM Glare 0.05 0.18 0.19 0.20 0.21 0.23 0.25 0.26 0.28 E 93.84 241.02 418.59 622.18 855.93 1115.86 1404.82 1754.45 2266.31 SHG 0.54 1.08 1.61 2.53 3.35 4.63 6.52 10.42 15.21 3:00 PM Glare 0.03 0.15 0.18 0.19 0.20 0.21 0.22 0.23 0.25 E 67.26 180.84 322.21 487.31 673.24 865.99 1125.97 1399.64 1756.58 SHG 0.45 0.89 1.28 1.80 2.24 3.04 4.35 5.88 8.29 4:00 PM Glare 0.02 0.11 0.17 0.19 0.19 0.20 0.20 0.21 0.22 E 45.93 133.62 261.43 410.57 570.33 741.95 922.48 1121.68 1347.53 SHG 0.45 0.87 1.26 1.65 2.19 2.58 2.98 3.37 3.94 5:00 PM Glare 0.01 0.04 0.12 0.16 0.18 0.18 0.19 0.19 0.19 E 25.84 79.76 164.49 266.49 372.35 485.39 594.38 699.76 798.16 SHG 0.55 1.04 1.40 1.76 2.12 2.33 2.59 2.79 3.05 Table 4. 29 Control strategy determination of kinetic shading panel of the cuboid model on Sept 21 st . 177 12/21 Control Strategy of Kinetic Shading Panel based on Three Metrics Tilt Time Metrics 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM Glare 0 0 0 0 0 0 0 0 0 E 0 0 0 0 0 0 0 0 0 SHG 0 0 0 0 0 0 0 0 0 8:00 AM Glare 0.004 0.006 0.009 0.015 0.023 0.034 0.049 0.069 0.095 E 9.26 27.64 49.44 73.11 95.42 117.45 142.72 178.61 222.56 SHG 0.038 0.072 0.099 0.118 0.132 0.144 0.150 0.158 0.163 9:00 AM Glare 0.006 0.023 0.071 0.137 0.168 0.179 0.185 0.191 0.196 E 28.69 79.70 146.06 219.79 298.47 382.44 486.27 625.78 780.44 SHG 0.141 0.277 0.385 0.459 0.521 0.574 0.624 0.680 0.713 10:00 AM Glare 0.010 0.057 0.148 0.177 0.186 0.193 0.200 0.208 0.217 E 46.51 124.24 221.33 325.28 445.67 575.46 740.77 928.81 1181.05 SHG 0.264 0.489 0.681 0.826 0.977 1.058 1.191 1.291 1.358 11:00 AM Glare 0.013 0.092 0.170 0.184 0.193 0.202 0.211 0.223 0.235 E 58.94 152.00 266.81 391.55 531.00 679.29 891.94 1139.23 1457.69 SHG 0.282 0.565 0.807 1.006 1.166 1.304 1.558 1.708 1.864 12:00 PM Glare 0.015 0.110 0.175 0.187 0.196 0.206 0.217 0.228 0.241 E 62.95 167.20 295.73 441.58 592.83 757.31 1001.00 1242.10 1595.83 SHG 0.365 0.677 0.948 1.283 1.758 1.893 2.073 2.331 2.813 1:00 PM Glare 0.013 0.098 0.172 0.185 0.193 0.202 0.212 0.223 0.234 E 59.40 155.79 278.17 418.99 574.45 730.92 950.43 1192.84 1503.59 SHG 0.350 0.863 1.161 1.435 1.700 1.963 2.289 2.656 3.311 2:00 PM Glare 0.011 0.070 0.158 0.180 0.188 0.196 0.204 0.214 0.222 E 49.66 129.84 241.50 364.10 501.24 645.92 839.12 1057.58 1339.34 SHG 0.342 0.661 0.941 1.163 1.395 1.791 2.001 2.256 2.932 3:00 PM Glare 0.009 0.040 0.123 0.169 0.181 0.188 0.194 0.200 0.206 E 36.47 105.09 197.13 311.98 430.32 573.07 722.53 892.74 1071.79 SHG 0.318 0.622 0.963 1.159 1.373 1.588 1.827 2.011 2.277 4:00 PM Glare 0.006 0.019 0.060 0.120 0.157 0.174 0.180 0.185 0.187 E 23.57 75.97 146.69 237.05 338.07 430.03 529.07 618.90 703.27 SHG 0.308 0.615 0.889 1.104 1.320 1.475 1.630 1.748 1.893 5:00 PM Glare 0 0 0 0 0 0 0 0 0 E 0 0 0 0 0 0 0 0 0 SHG 0.316 0.619 0.892 1.113 1.353 1.517 1.761 1.926 2.106 Table 4. 3 0 Control strategy determination of kinetic shading panel of the cuboid model on Dec 21 st . 178 6/21 Control Strategy of Tilt-only Shading Panel based on Three Metrics Tilt Time Metrics 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM Glare 0.005 0.011 0.024 0.047 0.076 0.102 0.119 0.129 0.134 E 15.59 52.47 105.24 173.41 245.15 310.27 363.27 400.36 417.64 SHG 0.140 0.294 0.432 0.561 0.689 0.756 0.820 0.863 0.882 8:00 AM Glare 0.008 0.034 0.105 0.157 0.174 0.180 0.184 0.186 0.187 E 32.45 101.29 210.12 347.32 490.02 623.60 733.11 812.81 839.53 SHG 0.339 0.711 1.047 1.363 1.675 1.838 1.994 2.099 2.145 9:00 AM Glare 0.012 0.082 0.165 0.181 0.188 0.192 0.196 0.199 0.200 E 47.53 151.44 314.94 526.15 741.78 944.41 1087.46 1206.23 1259.84 SHG 0.560 1.178 1.737 2.265 2.790 3.067 3.334 3.513 3.594 10:00 AM Glare 0.034 0.170 0.257 0.284 0.282 0.278 0.279 0.280 0.282 E 82.54 278.16 580.85 950.28 1389.01 1764.03 2098.08 2372.29 2456.48 SHG 0.948 2.004 2.974 3.904 4.832 5.377 5.900 6.268 6.464 11:00 AM Glare 0.019 0.139 0.180 0.190 0.198 0.205 0.210 0.213 0.215 E 59.72 212.26 427.72 708.61 1007.84 1286.29 1515.74 1675.48 1745.76 SHG 1.058 2.239 3.324 4.367 5.409 6.018 6.606 7.017 7.239 12:00 PM Glare 0.019 0.133 0.180 0.190 0.252 0.251 0.251 0.252 0.252 E 61.50 201.45 424.14 692.32 1001.64 1300.39 1534.36 1696.87 1798.17 SHG 0.876 1.853 2.747 3.615 4.471 4.944 5.414 5.739 5.923 1:00 PM Glare 0.045 0.177 0.194 0.209 0.224 0.238 0.251 0.260 0.266 E 95.01 387.02 738.44 1226.45 1895.45 2472.97 3273.81 3823.38 4179.36 SHG 1.078 2.329 3.522 4.878 6.227 7.200 8.296 9.164 10.016 2:00 PM Glare 0.132 0.204 0.226 0.250 0.270 0.729 0.743 0.754 0.761 E 203.24 564.46 1038.93 2274.28 2924.21 6609.86 7279.49 8428.92 11143.86 SHG 1.807 3.705 5.685 8.476 11.561 14.405 17.990 20.732 24.637 3:00 PM Glare 0.117 0.203 0.219 0.238 0.257 0.271 0.285 0.294 0.300 E 208.74 554.75 1679.05 2898.26 4142.68 4585.63 5098.80 6985.47 7931.80 SHG 2.618 5.222 8.311 11.776 15.527 20.639 25.810 30.449 35.398 4:00 PM Glare 0.128 0.203 0.218 0.233 0.248 0.261 0.271 0.278 0.275 E 206.24 549.63 1017.38 1449.05 1880.69 2311.48 2637.27 2897.94 3286.22 SHG 2.933 5.539 9.300 12.342 15.528 20.178 24.529 29.149 31.621 5:00 PM Glare 0.095 0.188 0.200 0.211 0.220 0.226 0.228 0.228 0.226 E 149.57 390.04 674.88 948.21 1247.17 1474.94 1604.05 1639.00 1587.42 SHG 2.748 5.162 8.533 10.925 13.933 16.219 18.650 20.781 20.879 Table 4. 3 1 Control strategy determination of tilt-only shading panel of the cuboid model on June 21 st . 179 9/21 Control Strategy of Tilt-only Shading Panel based on Three Metrics Tilt Time Metrics 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM Glare 0.007 0.022 0.056 0.098 0.127 0.146 0.156 0.163 0.166 E 35.55 87.03 147.20 203.52 251.90 295.26 333.68 366.63 387.51 SHG 0.142 0.262 0.403 0.484 0.605 0.650 0.660 0.712 0.779 8:00 AM Glare 0.017 0.117 0.175 0.186 0.192 0.199 0.204 0.212 0.224 E 91.88 230.83 371.75 558.89 705.77 872.23 1027.04 1206.29 1360.61 SHG 0.768 1.227 2.043 2.525 3.280 3.618 3.773 4.027 4.584 9:00 AM Glare 0.054 0.220 0.229 0.234 0.245 0.249 0.261 0.271 0.283 E 104.78 285.17 481.68 715.81 979.72 1765.71 2020.96 2857.33 3170.50 SHG 1.387 2.483 3.794 4.863 6.419 7.502 8.406 9.577 11.230 10:00 AM Glare 0.037 0.172 0.190 0.202 0.216 0.230 0.243 0.256 0.265 E 119.37 300.36 499.45 778.72 1750.89 2086.68 2419.48 3615.46 4862.38 SHG 1.235 2.695 3.799 5.032 7.277 8.791 10.709 12.881 16.315 11:00 AM Glare 0.046 0.177 0.193 0.207 0.223 0.240 0.257 0.272 0.282 E 122.41 314.06 550.26 1588.67 1905.58 2247.65 2645.33 4727.15 6914.29 SHG 0.893 1.960 2.729 3.659 6.071 7.590 9.970 12.312 17.403 12:00 PM Glare 0.047 0.178 0.193 0.208 0.225 0.243 0.262 0.275 0.286 E 126.30 320.56 561.27 830.12 1156.26 1496.21 1925.37 4131.78 5590.41 SHG 0.718 1.438 2.067 2.764 4.200 5.597 8.014 10.408 17.295 1:00 PM Glare 0.046 0.178 0.193 0.208 0.223 0.240 0.256 0.271 0.283 E 125.50 321.62 564.20 828.66 1135.46 1499.94 2720.65 4071.60 5505.50 SHG 0.629 1.277 1.912 2.554 3.337 4.699 6.917 9.423 17.154 2:00 PM Glare 0.041 0.174 0.191 0.204 0.217 0.231 0.243 0.255 0.263 E 120.92 323.02 569.87 1510.34 1800.79 2145.49 2492.19 3580.98 4709.34 SHG 0.795 1.485 2.150 2.873 4.680 6.172 8.679 10.985 17.424 3:00 PM Glare 0.032 0.167 0.188 0.198 0.208 0.217 0.226 0.234 0.240 E 118.28 306.24 549.18 809.48 1081.63 1798.64 2100.13 2361.13 2639.45 SHG 0.986 2.285 3.251 4.508 6.941 8.789 10.979 13.227 18.106 4:00 PM Glare 0.025 0.149 0.183 0.193 0.201 0.209 0.217 0.224 0.234 E 86.75 245.02 422.94 631.90 837.90 1061.15 1254.65 1398.20 1505.66 SHG 1.260 2.804 4.133 5.598 7.558 9.023 10.262 12.021 14.761 5:00 PM Glare 0.011 0.062 0.143 0.172 0.181 0.185 0.188 0.189 0.191 E 46.63 127.91 244.88 370.61 502.02 624.29 728.90 800.35 833.60 SHG 1.229 2.237 3.468 4.375 5.713 6.268 6.856 7.579 8.489 Table 4. 3 2 Control strategy determination of tilt-only shading panel of the cuboid model on September 21 st . 180 12/21 Control Strategy of Tilt-only Shading Panel based on Three Metrics Tilt Time Metrics 10° 20° 30° 40° 50° 60° 70° 80° 90° 7:00 AM Glare 0 0 0 0 0 0 0 0 0 E 0 0 0 0 0 0 0 0 0 SHG 0 0 0 0 0 0 0 0 0 8:00 AM Glare 0.004 0.007 0.011 0.017 0.026 0.036 0.049 0.066 0.087 E 12.46 33.19 58.30 79.96 103.49 123.19 144.86 173.89 205.15 SHG 0.054 0.098 0.164 0.188 0.225 0.250 0.264 0.280 0.296 9:00 AM Glare 0.008 0.032 0.095 0.151 0.173 0.182 0.187 0.191 0.196 E 36.27 96.09 172.58 246.39 326.34 412.31 511.47 620.51 767.50 SHG 0.190 0.460 0.772 0.872 1.103 1.212 1.369 1.526 1.614 10:00 AM Glare 0.011 0.071 0.156 0.179 0.187 0.195 0.203 0.210 0.218 E 51.03 134.34 236.90 346.74 469.24 606.46 761.68 943.34 1179.37 SHG 0.313 0.763 1.135 1.355 1.825 2.047 2.259 2.508 2.896 11:00 AM Glare 0.013 0.098 0.172 0.185 0.194 0.203 0.213 0.224 0.236 E 60.12 157.45 270.36 395.89 532.66 691.33 910.72 1124.28 1427.01 SHG 0.360 0.730 0.978 1.226 1.706 1.905 2.209 2.431 2.979 12:00 PM Glare 0.016 0.115 0.176 0.187 0.196 0.207 0.217 0.228 0.241 E 64.37 166.34 295.11 442.31 586.12 771.38 1010.35 1247.89 1591.67 SHG 0.363 0.680 0.947 1.175 1.512 1.648 2.055 2.307 2.711 1:00 PM Glare 0.015 0.106 0.174 0.186 0.195 0.204 0.214 0.224 0.236 E 61.79 160.25 287.43 426.94 574.77 749.70 980.28 1208.14 1514.55 SHG 0.359 0.699 0.987 1.235 1.500 1.650 2.106 2.386 2.734 2:00 PM Glare 0.012 0.086 0.165 0.183 0.191 0.198 0.206 0.215 0.224 E 55.88 147.55 265.59 394.23 537.77 699.65 875.68 1075.99 1339.31 SHG 0.380 0.727 1.006 1.244 1.697 1.858 2.280 2.498 2.987 3:00 PM Glare 0.010 0.058 0.143 0.175 0.184 0.190 0.196 0.201 0.206 E 44.79 126.86 233.75 348.19 479.47 612.50 759.84 910.25 1077.62 SHG 0.381 0.866 1.190 1.476 2.092 2.357 2.614 2.844 3.465 4:00 PM Glare 0.007 0.026 0.076 0.134 0.165 0.175 0.181 0.184 0.187 E 27.73 84.07 164.09 258.56 363.75 461.05 555.59 640.91 702.49 SHG 0.383 0.906 1.368 1.666 2.163 2.449 2.688 2.952 3.339 5:00 PM Glare 0 0 0 0 0 0 0 0 0 E 0 0 0 0 0 0 0 0 0 SHG 0.223 0.451 0.749 0.885 1.098 1.216 1.337 1.450 1.534 Table 4. 3 3 Control strategy determination of tilt-only shading panel of the cuboid model on December 21 st . 181 6/21 Kinetic Shading Panel Tilt-only Shading Panel Static Shading Panel Time Tilt Metrics Value Score Tilt Metrics Value Score Tilt Metrics Value Score 7:00 AM 0.151 1 0.134 1 0.102 1 90° 411.67 0 90° 417.64 0 60° 310.27 0 0.880 0.5 0.882 0 0.756 1 8:00 AM 0.184 1 0.180 1 0.180 1 60° 622.36 1 60° 623.60 0.5 60° 623.60 0.5 1.793 1 1.838 0.5 1.838 0.5 9:00 AM 0.182 1 0.181 1 0.192 1 40° 505.34 1 40° 526.15 0.5 60° 944.41 0 2.205 1 2.265 0.5 3.067 0 10:00 AM 0.188 1 0.257 1 0.278 1 30° 542.51 1 30° 580.85 0.5 60° 1764.03 0 2.717 1 2.974 0.5 5.377 0 11:00 AM 0.190 1 0.190 1 0.205 1 40° 704.49 1 40° 708.61 0.5 60° 1286.29 0 4.147 1 4.367 0.5 6.018 0 12:00 PM 0.189 1 0.190 1 0.251 1 40° 708.35 0.5 40° 692.32 1 60° 1300.39 0 3.632 0.5 3.615 1 4.944 0 1:00 PM 0.191 1 0.194 1 0.238 1 30° 607.75 1 30° 738.44 0.5 60° 2472.97 0 3.482 1 3.522 0.5 7.200 0 2:00 PM 0.198 1 0.204 1 0.729 1 30° 680.46 0.5 20° 564.46 1 60° 6609.86 0 3.843 0.5 3.705 1 14.405 0 3:00 PM 0.195 1 0.203 1 0.271 1 40° 702.42 0.5 20° 554.75 1 60° 4585.63 0 4.670 1 5.222 0.5 20.639 0 4:00 PM 0.187 1 0.203 1 0.261 1 40° 520.63 1 20° 549.63 0.5 60° 2311.48 0 2.338 1 5.539 0.5 20.178 0 5:00 PM 0.188 1 0.200 1 0.226 1 50° 590.24 1 30° 674.88 0.5 60° 1474.94 0 2.166 1 8.533 0.5 16.219 0 29 23.5 12 Table 4. 3 4 Evaluation among three cases of the cuboid model on June 21 st . 182 9/21 Kinetic Shading Panel Tilt-only Shading Panel Static Shading Panel Time Tilt Metrics Value Score Tilt Metrics Value Score Tilt Metrics Value Score 7:00 AM 0.165 1 0.166 1 0.146 1 90° 386.98 0 90° 387.51 0 60° 295.26 0 0.389 1 0.779 0 0.650 0.5 8:00 AM 0.19 1 0.186 1 0.199 1 60° 522.61 1 40° 558.89 0.5 60° 872.23 0 0.90 1 2.525 0.5 3.618 0 9:00 AM 0.20 1 0.234 1 0.249 1 50° 655.85 1 40° 715.81 0.5 60° 1765.71 0 1.50 1 4.863 0.5 7.502 0 10:00 AM 0.21 1 0.190 1 0.230 1 40° 637.98 0.5 30° 499.45 1 60° 2086.68 0 1.84 1 3.799 0.5 8.791 0 11:00 AM 0.20 1 0.193 1 0.240 1 30° 537.07 1 30° 550.26 0.5 60° 2247.65 0 1.66 1 2.729 0.5 7.590 0 12:00 PM 0.20 1 0.193 1 0.243 1 30° 561.76 1 30° 561.27 1 60° 1496.21 0.5 2.22 0.5 2.067 1 5.597 0 1:00 PM 0.20 1 0.193 1 0.240 1 30° 513.60 1 30° 564.20 0.5 60° 1499.94 0 2.18 0.5 1.912 1 4.699 0 2:00 PM 0.20 1 0.191 1 0.231 1 40° 622.18 0.5 30° 569.87 1 60° 2145.49 0 2.53 0.5 2.150 1 6.172 0 3:00 PM 0.20 1 0.188 1 0.217 1 50° 673.24 0.5 30° 549.18 1 60° 1798.64 0 2.24 1 3.251 0.5 8.789 0 4:00 PM 0.19 1 0.193 1 0.209 1 50° 570.33 1 40° 631.90 0.5 60° 1061.15 0 2.19 1 5.598 0.5 9.023 0 5:00 PM 0.19 1 0.181 1 0.185 1 70° 594.38 0.5 50° 502.02 1 60° 624.29 0 2.59 1 5.713 0.5 6.268 0 28.5 25 12 Table 4. 3 5 Evaluation among three cases of the cuboid model on September 21 st . 183 12/21 Kinetic Shading Panel Tilt-only Shading Panel Static Shading Panel Time Tilt Metrics Value Score Tilt Metrics Value Score Tilt Metrics Value Score 7:00 AM 0 0 0 0 0 0 - 0 0 - 0 0 60° 0 0 0 0 0 0 0 0 8:00 AM 0.095 1 0.087 1 0.036 1 90° 222.56 0 90° 205.15 0 60° 123.19 0 0.163 0 0.296 1 0.250 0.5 9:00 AM 0.191 1 0.187 1 0.182 1 80° 625.78 0.5 70° 511.47 1 60° 412.31 0 0.680 0 1.369 1 1.212 0.5 10:00 AM 0.193 1 0.195 1 0.195 1 60° 575.46 1 60° 606.46 0.5 60° 606.46 0.5 1.058 0.5 2.047 1 2.047 1 11:00 AM 0.193 1 0.194 1 0.203 1 50° 531.00 1 50° 532.66 0.5 60° 691.33 0 1.166 0 1.706 0.5 1.905 1 12:00 PM 0.196 1 0.196 1 0.207 1 50° 592.83 0.5 50° 586.12 1 60° 771.38 0 1.758 1 1.512 0 1.648 0.5 1:00 PM 0.193 1 0.195 1 0.204 1 50° 574.45 1 50° 574.77 0.5 60° 749.70 0 1.700 1 1.500 0 1.650 0.5 2:00 PM 0.188 1 0.191 1 0.198 1 50° 501.24 1 50° 537.77 0.5 60° 699.65 0 1.395 0 1.697 0.5 1.858 1 3:00 PM 0.188 1 0.190 1 0.190 1 60° 573.07 1 60° 612.50 0.5 60° 612.50 0.5 1.588 0.5 2.357 1 2.357 1 4:00 PM 0.174 1 0.181 1 0.175 1 60° 573.20 0.5 70° 555.59 1 60° 461.05 0 1.475 0 2.688 1 2.449 0.5 5:00 PM 0 0 0 0 0 0 70° 0 0 90° 0 0 60° 0 0 2.106 1 1.534 0.5 1.216 0 19.5 21 16.5 Table 4. 3 6 Evaluation among three cases of the cuboid model on December 21 st . 184 APPENDIX C: Façades Patterns Comparison Based on Daylighting and Thermal Metrics. 6/21 Indoor Average Illuminance (lux) Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 144.35 55.53 268.63 8:00 AM 292.86 108.45 539.18 9:00 AM 425.35 155.70 793.66 10:00 AM 514.12 273.07 1430.11 11:00 AM 559.49 212.01 1081.88 12:00 PM 537.97 192.66 1038.73 1:00 PM 617.37 343.86 1791.03 2:00 PM 517.73 443.30 1893.27 3:00 PM 521.44 368.03 1155.28 4:00 PM 518.73 298.51 954.66 5:00 PM 463.16 227.68 843.77 Table 5. 3 Data results of hourly indoor average illuminance on June 21 st . Figure 5. 1 0 Hourly indoor average illuminances of three facades on June 21 st . 0.00 200.00 400.00 600.00 800.00 1000.00 1200.00 1400.00 1600.00 1800.00 2000.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Indoor Average Illuminance_June 21st Kinetic Louvers Static Louvers Static Glazing 185 9/21 Indoor Average Illuminance (lux) Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 296.90 172.65 2277.37 8:00 AM 576.09 391.21 4711.79 9:00 AM 562.18 562.18 7311.69 10:00 AM 686.79 686.79 8047.79 11:00 AM 732.97 732.97 8532.55 12:00 PM 754.12 754.12 8183.02 1:00 PM 741.71 741.71 6314.95 2:00 PM 678.76 678.76 4004.57 3:00 PM 554.60 554.60 1687.48 4:00 PM 609.70 377.45 747.80 5:00 PM 445.29 211.48 530.91 Table 5. 4 Data results of hourly indoor average illuminance on September 21 st . Figure 5. 11 Hourly indoor average illuminances of three facades on September 21 st . 0.00 500.00 1000.00 1500.00 2000.00 2500.00 3000.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Indoor Average Illuminance_Sep. 21st Kinetic Louvers Static Louvers Static Glazing 186 12/21 Indoor Average Illuminance Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 0 0 0 8:00 AM 548.27 548.27 963.37 9:00 AM 696.04 1938.11 4961.65 10:00 AM 599.37 2607.16 8142.24 11:00 AM 726.80 1226.06 11446.74 12:00 PM 707.74 1247.60 11590.41 1:00 PM 1228.96 3776.58 10747.24 2:00 PM 676.43 1106.07 8253.45 3:00 PM 591.71 4064.63 5230.27 4:00 PM 625.45 625.45 1288.30 5:00 PM 0 0 0 Table 5. 5 Data results of hourly indoor average illuminance on December 21 st . Figure 5. 12 Hourly indoor average illuminances of three facades on December 21 st . 0 500 1000 1500 2000 2500 3000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Indoor Average Illuminance_Dec. 21st Kinetic Louvers Static Louvers Static Glazing 187 6/21 Daylight Glare Probability Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 0.084 0.226 0.229 8:00 AM 0.215 0.255 0.256 9:00 AM 0.233 0.272 0.278 10:00 AM 0.313 0.307 0.328 11:00 AM 0.272 0.289 0.302 12:00 PM 0.268 0.283 0.295 1:00 PM 0.334 0.325 0.351 2:00 PM 0.254 0.351 0.383 3:00 PM 0.330 0.331 0.352 4:00 PM 0.290 0.315 0.332 5:00 PM 0.275 0.296 0.306 Table 5. 6 Data results of hourly daylight glare probability on June 21 st . Figure 5. 13 Hourly daylight glare probability of three facades on June 21 st . 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Daylight Glare Probability_June 21st Kinetic Louvers Static Louvers Static Glazing 188 9/21 Daylight Glare Probability Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 0.250 0.263 0.327 8:00 AM 0.317 0.308 0.380 9:00 AM 0.345 0.345 0.420 10:00 AM 0.373 0.373 0.445 11:00 AM 0.387 0.387 0.452 12:00 PM 0.391 0.391 0.441 1:00 PM 0.385 0.385 0.412 2:00 PM 0.368 0.368 0.372 3:00 PM 0.368 0.368 0.329 4:00 PM 0.319 0.308 0.279 5:00 PM 0.257 0.272 0.266 Table 5. 7 Data results of hourly daylight glare probability on September 21 st . Figure 5. 14 Hourly daylight glare probability of three facades on September 21 st . 0.000 0.100 0.200 0.300 0.400 0.500 0.600 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Daylight Glare Probability_Sep. 21st Kinetic Louvers Static Louvers Static Glazing 189 12/21 Daylight Glare Probability Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 0.000 0.000 0.000 8:00 AM 0.301 0.301 0.325 9:00 AM 0.318 0.365 0.411 10:00 AM 0.368 0.440 0.533 11:00 AM 0.396 0.478 0.621 12:00 PM 0.404 0.492 0.649 1:00 PM 0.396 0.480 0.614 2:00 PM 0.374 0.443 0.533 3:00 PM 0.264 0.377 0.428 4:00 PM 0.321 0.321 0.349 5:00 PM 0.000 0.000 0.000 Table 5. 8 Data results of hourly daylight glare probability on December 21 st . Figure 5. 15 Hourly daylight glare probability of three facades on December 21 st . 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Daylight Glare Probability_Dec. 21st Kinetic Louvers Static Louvers Static Glazing 190 6/21 Solar Heat Gain (kWh) Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 0.121 0.077 0.206 8:00 AM 0.289 0.183 0.492 9:00 AM 0.476 0.295 0.809 10:00 AM 0.548 0.469 1.340 11:00 AM 0.676 0.521 1.501 12:00 PM 0.573 0.445 1.268 1:00 PM 0.675 0.580 1.733 2:00 PM 0.363 0.728 2.242 3:00 PM 0.757 0.658 1.790 4:00 PM 0.572 0.559 1.227 5:00 PM 0.521 0.511 1.096 Table 5. 9 Data results of hourly solar heat gain on June 21 st . Figure 5. 16 Hourly solar heat gain of three facades on June 21 st . 0.000 0.500 1.000 1.500 2.000 2.500 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain_June 21st Kinetic Louvers Static Louvers Static Glazing 191 9/21 Solar Heat Gain (kWh) Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 0.117 0.188 0.286 8:00 AM 0.585 0.443 1.598 9:00 AM 0.603 0.603 3.492 10:00 AM 0.762 0.762 5.297 11:00 AM 0.981 0.981 6.775 12:00 PM 0.976 0.976 7.747 1:00 PM 0.834 0.834 8.359 2:00 PM 0.967 0.967 7.724 3:00 PM 0.913 0.913 5.795 4:00 PM 0.871 0.719 4.153 5:00 PM 0.997 0.689 2.630 Table 5. 1 0 Data results of hourly solar heat gain on September 21 st . Figure 5. 17 Hourly solar heat gain of three facades on September 21 st . Figure 5.30 0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000 8.000 9.000 10.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain_Sep. 21st Kinetic Louvers Static Louvers Static Glazing 192 12/21 Solar Heat Gain (kWh) Type Time Kinetic Louver Fixed Louver Static Glazing 7:00 AM 0.000 0.000 0 8:00 AM 1.157 1.157 1.423 9:00 AM 0.420 2.383 5.108 10:00 AM 1.839 4.433 8.780 11:00 AM 2.683 4.678 11.400 12:00 PM 1.644 2.024 12.651 1:00 PM 0.910 1.110 12.641 2:00 PM 1.885 2.808 12.452 3:00 PM 0.781 5.305 11.277 4:00 PM 4.506 4.506 8.110 5:00 PM 3.158 2.139 3.746 Table 5. 1 1 Data results of hourly solar heat gain on December 21 st . Figure 5. 18 Hourly solar heat gain of three facades on December 21 st . 0.000 2.000 4.000 6.000 8.000 10.000 12.000 14.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain_Dec. 21st Kinetic Louvers Static Louvers Static Glazing 193 6/21 Indoor Average Illuminance (lux) Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 411.67 417.64 310.27 8:00 AM 622.36 623.60 623.60 9:00 AM 505.34 526.15 944.41 10:00 AM 542.51 580.85 1764.03 11:00 AM 704.49 708.61 1286.29 12:00 PM 708.35 692.32 1300.39 1:00 PM 607.75 738.44 2472.97 2:00 PM 680.46 564.46 6609.86 3:00 PM 702.42 554.75 4585.63 4:00 PM 520.63 549.63 2311.48 5:00 PM 590.24 674.88 1474.94 Table 5. 1 4 Data results of hourly indoor average illuminance on June 21 st . Figure 5. 28 Hourly indoor average illuminances of three facades on June 21 st . 0.00 500.00 1000.00 1500.00 2000.00 2500.00 3000.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Indoor Average Illuminance_June 21st Kinetic Panel Tilt-only Panel Static Panel 194 9/21 Indoor Average Illuminance (lux) Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 386.98 387.51 295.26 8:00 AM 522.61 558.89 872.23 9:00 AM 655.85 715.81 1765.71 10:00 AM 637.98 499.45 2086.68 11:00 AM 537.07 550.26 2247.65 12:00 PM 561.76 561.27 1496.21 1:00 PM 513.60 564.20 1499.94 2:00 PM 622.18 569.87 2145.49 3:00 PM 673.24 549.18 1798.64 4:00 PM 570.33 631.90 1061.15 5:00 PM 594.38 502.02 624.29 Table 5. 1 5 Data results of hourly indoor average illuminance on September 21 st . Figure 5. 29 Hourly indoor average illuminances of three facades on September 21 st . 0.00 500.00 1000.00 1500.00 2000.00 2500.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Indoor Average Illuminance_Sep. 21st Kinetic Panel Tilt-only Panel Static Panel 195 12/21 Indoor Average Illuminance (lux) Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 0.00 0.00 0.00 8:00 AM 222.56 205.15 123.19 9:00 AM 625.78 511.47 412.31 10:00 AM 575.46 606.46 606.46 11:00 AM 531.00 532.66 691.33 12:00 PM 592.83 586.12 771.38 1:00 PM 574.45 574.77 749.70 2:00 PM 501.24 537.77 699.65 3:00 PM 573.07 612.50 612.50 4:00 PM 529.07 555.59 461.05 5:00 PM 0.00 0.00 0.00 Table 5. 1 6 Data results of hourly indoor average illuminance on December 21 st . Figure 5. 30 Hourly indoor average illuminances of three facades on December 21 st . 0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00 800.00 900.00 1000.00 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Illuminance (Lux) Indoor Average Illuminance_Dec. 21st Kinetic Panel Tilt-only Panel Static Panel 196 6/21 Daylight Glare Probability Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 0.151 0.134 0.102 8:00 AM 0.184 0.180 0.180 9:00 AM 0.182 0.181 0.192 10:00 AM 0.188 0.257 0.278 11:00 AM 0.190 0.190 0.205 12:00 PM 0.189 0.190 0.251 1:00 PM 0.191 0.194 0.238 2:00 PM 0.198 0.204 0.729 3:00 PM 0.195 0.203 0.271 4:00 PM 0.187 0.203 0.261 5:00 PM 0.188 0.200 0.226 Table 5. 1 7 Data results of hourly daylight glare probability on June 21 st . Figure 5. 31 Hourly daylight glare probability of three facades on June 21 st . 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Daylight Glare Probability_June 21st Kinetic Panel Tilt-only Panel Static Panel 197 9/21 Daylight Glare Probability Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 0.165 0.166 0.146 8:00 AM 0.190 0.186 0.199 9:00 AM 0.202 0.234 0.249 10:00 AM 0.206 0.190 0.230 11:00 AM 0.201 0.193 0.240 12:00 PM 0.203 0.193 0.243 1:00 PM 0.200 0.193 0.240 2:00 PM 0.204 0.191 0.231 3:00 PM 0.202 0.188 0.217 4:00 PM 0.193 0.193 0.209 5:00 PM 0.188 0.181 0.185 Table 5. 1 8 Data results of hourly daylight glare probability on September 21 st . Figure 5. 32 Hourly daylight glare probability of three facades on September 21 st . 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Daylight Glare Probability_Sep. 21st Kinetic Panel Tilt-only Panel Static Panel 198 12/21 Daylight Glare Probability Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 0.000 0.000 0.000 8:00 AM 0.095 0.087 0.036 9:00 AM 0.191 0.187 0.182 10:00 AM 0.193 0.195 0.195 11:00 AM 0.193 0.194 0.203 12:00 PM 0.196 0.196 0.207 1:00 PM 0.193 0.195 0.204 2:00 PM 0.188 0.191 0.198 3:00 PM 0.188 0.190 0.190 4:00 PM 0.174 0.181 0.175 5:00 PM 0.000 0.000 0.000 Table 5. 19 Data results of hourly daylight glare probability on December 21 st . Figure 5. 33 Hourly daylight glare probability of three facades on December 21 st . -0.050 0.050 0.150 0.250 0.350 0.450 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM DGP Daylight Glare Probability_Dec. 21st Kinetic Panel Tilt-only Panel Static Panel 199 6/21 Solar Heat Gain (kWh) Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 0.880 0.882 0.756 8:00 AM 1.793 1.838 1.838 9:00 AM 2.205 2.265 3.067 10:00 AM 2.717 2.974 5.377 11:00 AM 4.147 4.367 6.018 12:00 PM 3.632 3.615 4.944 1:00 PM 3.482 3.522 7.200 2:00 PM 3.843 3.705 14.405 3:00 PM 4.670 5.222 20.639 4:00 PM 2.338 5.539 20.178 5:00 PM 2.166 8.533 16.219 Table 5. 2 0 Data results of hourly solar heat gain on June 21 st . Figure 5. 34 Hourly solar heat gain of three facades on June 21 st . 0.000 2.000 4.000 6.000 8.000 10.000 12.000 14.000 16.000 18.000 20.000 22.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain_June 21st Kinetic Panel Tilt-only Panel Static Panel 200 9/21 Solar Heat Gain (kWh) Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 0.389 0.779 0.650 8:00 AM 0.901 2.525 3.618 9:00 AM 1.504 4.863 7.502 10:00 AM 1.839 3.799 8.791 11:00 AM 1.662 2.729 7.590 12:00 PM 2.219 2.067 5.597 1:00 PM 2.178 1.912 4.699 2:00 PM 2.527 2.150 6.172 3:00 PM 2.239 3.251 8.789 4:00 PM 2.189 5.598 9.023 5:00 PM 2.586 5.713 6.268 Table 5. 2 1 Data results of hourly solar heat gain on September 21 st . Figure 5. 35 Hourly solar heat gain of three facades on September 21 st . Figure 5.48 0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000 8.000 9.000 10.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain_Sep. 21st Kinetic Panel Tilt-only Panel Static Panel 201 12/21 Solar Heat Gain (kWh) Type Time Kinetic Panel Tilt-only Panel Static Panel 7:00 AM 0.000 0.000 0.000 8:00 AM 0.163 0.296 0.250 9:00 AM 0.680 1.369 1.212 10:00 AM 1.058 2.047 2.047 11:00 AM 1.166 1.706 1.905 12:00 PM 1.758 1.512 1.648 1:00 PM 1.700 1.500 1.650 2:00 PM 1.395 1.697 1.858 3:00 PM 1.588 2.357 2.357 4:00 PM 1.475 2.688 2.449 5:00 PM 2.106 1.534 1.216 Table 5. 2 2 Data results of hourly solar heat gain on December 21 st Figure 5. 36 Hourly solar heat gain of three facades on December 21 st . 0.000 0.500 1.000 1.500 2.000 2.500 3.000 3.500 4.000 7:00 AM 8:00 AM 9:00 AM 10:00 AM 11:00 AM 12:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM Total Radiation (kWh) Time Solar Heat Gain_Dec. 21st Kinetic Panel Tilt-only Panel Static Panel

Abstract (if available)

Abstract As one special type of building envelope, kinetic facades are increasingly being applied to contemporary architecture and building practice for the advantages of either aesthetics of architectural motions or better environmental performance. As is known to professionals in the building industry, the building envelope is an essential structure that plays a significant role in building daylight environments and energy savings since it works as an artificial barrier to isolate the indoor condition from the outdoor environment. The kinetic envelope can actively affect the built interior environmental quality, the building energy consumption, and the occupant's visual and thermal comfort. Thus, it is a significant mission to discover the principles and rules of the dynamic façade and its contemporary development and practice on improving environmental performance. The research focuses on the interior daylight environment analysis with kinetic façades in terms of daylight metrics by using parametric workflow. The process of parametric simulation can detect and investigate the existing advantages and deficiencies with respect to the daylight index. By processing simulation data, the workflow uses model-based predictive control as a control algorithm to operate the movable façade for better daylight effects, compared with the static counterpart of a kinetic façade, the research proposes the potential entry point or approach to improve the interior daylight condition. Beyond that, the research applies a systematic approach and integrative view to establish daylighting evaluation and improvement for testing and verifying the interior space daylighting performance of kinetic facades.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Kinetic facades as environmental control systems: using kinetic facades to increase energy efficiency and building performance in office buildings

PDF

Multi-domain assessment of a kinetic facade: determining the control strategy of a kinetic façade using BIM based on energy performance, daylighting, and occupants’ preferences; Multi-domain asse...

PDF

Performative shading design: parametric based measurement of shading system configuration effectiveness and trends

PDF

Evaluation of daylighting circadian effects: Integrating non-visual effects of lighting in the evaluation of daylighting designs

PDF

Adaptive façade controls: A methodology based on occupant visual comfort preferences and cluster analysis

PDF

Environmentally responsive buildings: multi-objective optimization workflow for daylight and thermal quality

PDF

Dynamic shading and glazing technologies: improve energy, visual, and thermal performance

PDF

Evaluation and development of solar control screens: using daylight simulation to improve the performance of facade solar control screens

PDF

Intelligent building skins: Parametric-based algorithm for kinetic facades design and daylighting performance integration

PDF

Daylight and health: exploring the relationship between established daylighting metrics for green building compliance and new metrics for human health

PDF

Energy use intensity estimation method based on building façade features by using regression models

PDF

Effective light shelf and form finding: development of a light shelf design assistant tool using parametric methods

PDF

The effectiveness of enviro-materially actuated kinetic facades: evaluating the thermal performance of thermo-bimetal shading component geometries

Zip

Intelligent building skins: Parametric-based algorithm for kinetic facades design and daylighting performance integration [program]

PDF

Simulation-based electric lighting control algorithm: integrating daylight simulation with daylight-linked lighting control

PDF

A proposal for building envelope retrofit on the Bonaventure Hotel: a case study examining energy and carbon

PDF

Energy savings by using dynamic environmental controls in the cavity of double skin facades

PDF

Double skin façades performance: effects on daylight and visual comfort in office spaces

PDF

A simplified building energy simulation tool: material and environmental properties effects on HVAC performance

PDF

Daylighting in Riyadh and Los Angeles: comparison of cultural factors in potential market penetration

Asset Metadata

Creator Yang, Chenxi (author)

Core Title The intelligent control strategy of kinetic façades for daylight and energy performance: evaluating the daylight effect of adaptive systems based on parametric workflow

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 04/26/2020

Defense Date 03/23/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag adaptive shading,daylight metrics,dynamic daylighting,intelligent systems,kinetic facades,OAI-PMH Harvest,parametric design

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Konis, Kyle (committee chair), Fox, Michael (committee member), Noble, Douglas (committee member), Schiler, Marc (committee member)

Creator Email chenxiy@usc.edu,yangchenxi1987830@126.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-288090

Unique identifier UC11665928

Identifier etd-YangChenxi-8341.pdf (filename),usctheses-c89-288090 (legacy record id)

Legacy Identifier etd-YangChenxi-8341.pdf

Dmrecord 288090

Document Type Thesis

Rights Yang, Chenxi

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA