University of Southern California Dissertations and Theses

CFD visualization: a case study for using a building information modeling with virtual reality

(USC Thesis Other)

CFD visualization: a case study for using a building information modeling with virtual reality

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i

CFD VISUALIZATION

A Case Study for Using a Building Information Modeling with Virtual Reality

By

Jiayi Yan

Presented to the
FACULTY OF THE
SCHOOL OF ARCHITECTURE
UNIVERSITY OF SOUTHERN CALIFORNIA
In partial fulfillment of the
Requirements of degree
MASTER OF BUILDING SCIENCE

AUG 2017
ii

To family and all the people that help me in my life,
Thank you for all of your support.

iii

ACKNOWLEDGEMENT
First and foremost, I would like to express my deepest gratitude to my committee chair, Professor Karen Kensek, who
has contributed great efforts on this project. Her patience and academic advice helped me with the development of all
the work. Her positive attitude and patience towards work and her dedication and enthusiasm in education will inspire
all the way along my life.
I would also like to express my special appreciation and thanks to my committee members, Professor Kyle Konis, and
Professor Douglas Noble. They provided great support and encouragement. I really appreciate their valuable advice
on improving the project and my writing skills.
Last but not least, I would like to say thank you to the wonderful MBS family. I feel so warm with the support of all
the family members. I learned a lot and had memorable experience with all of our big family, which greatly support
me to complete the project along the way.

iv

COMMITTEE

Karen M. Kensek, LEED AP BD+C, Assoc. AIA
Associate Professor of the Practice of Architecture
USC School of Architecture
kensek@usc.edu
(213)740-2081

Kyle Konis, Ph.D., AIA
Assistant Professor
USC School of Architecture
kkonis@usc.edu

Douglas Noble, FAIA, Ph.D.
Associate Professor
USC School of Architecture
dnoble@usc.edu
(213) 740-2723

v

ABSTRACT
Building energy simulation is an important procedure during the building’s design. Not only can building energy
simulation help identify opportunities for saving energy, but if interpreted correctly and implemented, it can contribute
to the occupants’ comfort. Scientific visualization has been adopted for a long time in engineering field, tracking large
scale simulation data and providing intuition and understandable graphs and models displaying the data. For
computational fluid dynamics (CFD) data, the need for scientific visualization is of more importance, due to the
complicated spatial data structure and large quantities of data points characteristic of CFD data. In the CFD simulation
engines, there are built-in CFD simulation data visualization method such as streamlined 3D air flow visualization of
an analyzed space. The raw CFD data could also be exported and processed in other software programs to add clarity.
Given the consideration of better taking advantages of the CFD results for buildings, the potential of the use of virtual
reality (VR) techniques cannot be overlooked. VR techniques bring about immersion and presence (Gutierrez et al.
2008). One can start with a building information model (BIM), produce CFD simulation results, and visualize those
results in VR. There are existing achievements of CFD data or similar energy simulation data visualization in VR
environment, such as visualizing Ecotect simulation data in VR game engine 3DVia Studio (Bahar 2014) and
visualizing OpenFOAM CFD simulation data in Unity 3D (Hosokawa 2016). Some researchers have created CFD
visualization tool as a way to streamline the workflow from CFD raw data to VR environment (Berger et al. 2015).
The existing methods have given positive answers of the possibility to convert numerical CFD data to virtual
environment objects. However, the previous workflows include problems in data exchanging complication (the
software applied may be from totally different working areas), incomplete data representation, etc. As a result, a BIM
to CFD data to VR visualization workflow is proposed.
Building information modeling (BIM) as a lifecycle tool for buildings as includes as much as possible information for
further applications. The proposed methodology starts from the building information model Autodesk Revit. CFD
simulation is followed using Autodesk CFD Simulation through the add-in tab in Revit. Multiple kinds of CFD
simulation including indoor natural and mechanical ventilation and outdoor wind pattern simulation can be done. The
data type of nodal is exported for later visualization processing. To realize the visualization in VR, Autodesk 3Ds Max
and Autodesk Stingray are applied continuously to manage the data and generate the VR objects. An actual case study
of a kitchen space works is applied with the proposed methodology. In the final result, users are able to interact with
the virtual environment wearing VR headset like HTC Vive, observing visualized CFD streamlines with color.
Evaluation of the methodology is based on CFD data representation accuracy, user’s experience in both immersion,
and presence perspectives and BIM integration level of the whole workflow.

KEY WORDS: CFD simulation, BIM, scientific visualization, virtual reality

HYPOTHESIS
Starting with a building information model, architects and engineers are able to visualize CFD simulation results in a
virtual reality VR environment that can aid simulation results representation and scientific analysis.

vi

CONTENTS

ACKNOWLEDGEMENT ............................................................................................................................................ iii
COMMITTEE ............................................................................................................................................................... iv
ABSTRACT ....................................................................................................................................................................v
1. Introduction .................................................................................................................................................................1
1.1 Energy simulation ................................................................................................................................................1
1.2 Building information modeling (BIM) ................................................................................................................1
1.3 Visualization ........................................................................................................................................................4
1.4 Virtual reality (VR) ..............................................................................................................................................6
1.5 Thesis framework ...............................................................................................................................................13
1.5.1 Abstract ......................................................................................................................................................13
1.5.2 Hypothesis ..................................................................................................................................................14
1.5.3 Research objectives ....................................................................................................................................14
1.5.4 Chapter outlines ..........................................................................................................................................14
1.6 Summary ............................................................................................................................................................14
2. Background of energy simulation result visualization in virtual reality ...................................................................15
2.1 Understanding of energy simulation data ..........................................................................................................15
2.1.1 Introduction of energy simulation ..............................................................................................................15
2.1.2 Different types of energy simulation data for visualization .......................................................................16
2.1.3 CFD data ....................................................................................................................................................18
2.2 Applying BIM for energy simulation including CFD simulation ......................................................................19
2.3 Scientific visualization for CFD simulation .......................................................................................................22
2.3.1 Introduction ................................................................................................................................................22
2.3.3 Scientific visualization for CFD simulation ...............................................................................................23
2.4 Understanding of virtual reality (VR) ................................................................................................................25
2.4.1 Applications of VR in architectural field ...................................................................................................25
2.4.2 VR in scientific visualization .....................................................................................................................26
2.4.3 Previous research on VR in scientific visualization ...................................................................................27
2.5 Summary ............................................................................................................................................................28
3. Methodology .............................................................................................................................................................29
3.1 Methodology overview for CFD ........................................................................................................................29
3.1.1 Overall workflow .......................................................................................................................................31
3.1.2 Software selection ......................................................................................................................................33
3.2 BIM geometry ....................................................................................................................................................34
3.2.1 Detailed geometry preparation ...................................................................................................................34
3.2.2 Geometry simplification .............................................................................................................................36
3.3 CFD simulation ..................................................................................................................................................37
3.3.1 Material and devices ...................................................................................................................................39
3.3.2 Assigning boundary condition ...................................................................................................................41
3.3.3 Meshing ......................................................................................................................................................42
3.3.4 Solver .........................................................................................................................................................43
3.3.5 Result simulation in CFD ...........................................................................................................................43
3.4 Three-dimension visualization ...........................................................................................................................45
3.4.1 Import geometry .........................................................................................................................................46
3.4.2 Wind flow visualization .............................................................................................................................47
3.4.3 Air flow animation .....................................................................................................................................50
3.5 Virtual reality (VR) ............................................................................................................................................53
3.5.1 VR Environment build ...............................................................................................................................54
3.5.2 HMD test with HTC VIVE ........................................................................................................................55
3.6 Conclusion .........................................................................................................................................................55
4. Case study .................................................................................................................................................................56
vii

4.1 Case study overview ..........................................................................................................................................56
4.2 BIM geometry ....................................................................................................................................................56
4.3 CFD simulation in Autodesk CFD .....................................................................................................................57
4.2.1 Set material .................................................................................................................................................58
4.2.2 Set boundary ...............................................................................................................................................58
4.2.3 Mesh sizing ................................................................................................................................................58
4.2.4 Solver .........................................................................................................................................................59
4.2.5 Result viewer ..............................................................................................................................................59
4.2.6 Export Nodal Results data ..........................................................................................................................60
4.3 Visualization in Autodesk 3ds Max ...................................................................................................................61
4.3.1 Import CFD data and Revit model .............................................................................................................61
4.3.2 Create CFD streamlines .............................................................................................................................62
4.3.3 Add CFD vertex color modifier .................................................................................................................63
4.4.4 Create VR ...................................................................................................................................................64
4.3.7 HTC VIVE test ...........................................................................................................................................70
4.4 Conclusion .........................................................................................................................................................71
5. Evaluation and analysis .............................................................................................................................................72
5.1 Evaluation of the workflows ..............................................................................................................................72
5.1.1 Workflow 1 evaluation: EPW to VR ..........................................................................................................73
5.1.2 Workflow 2 evaluation: BIM solar simulation to VR ................................................................................76
5.1.3 Workflow 3 evaluation: BIM to CFD to VR ..............................................................................................77
5.2 Comparison of visualization results ...................................................................................................................82
5.2.1 CFD results .................................................................................................................................................82
5.2.2 3ds Max results ...........................................................................................................................................86
5.2.3 VR results ...................................................................................................................................................89
5.3 Summary ............................................................................................................................................................90
6. Conclusion and future work ......................................................................................................................................91
6.1 Summary of proposed methodology ..................................................................................................................91
6.1.1 BIM geometry stage ...................................................................................................................................91
6.1.2 CFD simulation stage .................................................................................................................................92
6.1.3 3D visualization stage ................................................................................................................................93
6.1.4 Virtual reality stage ....................................................................................................................................93
6.2 Evaluation of the methodology ..........................................................................................................................94
6.3 Future work ........................................................................................................................................................94
6.3.1 BIM geometry ............................................................................................................................................95
6.3.2 CFD simulation ..........................................................................................................................................95
6.3.3 3D visualization ..........................................................................................................................................96
6.3.4 Virtual reality .............................................................................................................................................96
6.3.5 Validation of the visualization and VR environment .................................................................................97
6.4 Other applications in AEC of VR with BIM ......................................................................................................97
6.5 Summary ............................................................................................................................................................99
Appendix A .................................................................................................................................................................101
Bibliography ...............................................................................................................................................................103

1

1. Introduction
This chapter provides an introduction to the concept of building energy simulation, building information modeling
(BIM), data visualization and virtual reality (VR), the scope of the thesis, and thesis output.
1.1 Energy simulation
As the global environment issue is becoming more and more severe, people in many fields are trying to contribute
their efforts to tackle this issue. According to a report from Architecture 2030, U.S. Energy Information
Administration (EIA) points out that the building sector consumes 47.6% of all energy produced in the United States,
and 74.9% of all the electricity produced in the U.S. is used up just to operate the building (Architecture 2030). The
urgency of saving energy consumed by building industry to the maximum extent calls for reactions from architects
and engineers.
Building energy simulation, known also as building energy modeling, helps architects and engineers predict the
behavior of buildings before they are built and lets them test virtual improvements. It is a series of applications of
software that can predict the performance of a single building or a complex of building (Wikipedia contributors.
“Building energy simulation.” Wikipedia, The Free Encyclopedia. Wikipedia, The Free Encyclopedia, 25 Apr. 2017.
Web. 28 Apr.2017.)
The building environment consists of and depends on outdoor climate and indoor diverse types of heat sources and
ventilation. At the same time, the building environment is adjusted according to the surrounding changes to reach the
comfort level. It is often difficult to tell how the combined parameters can affect the building environment. Hence, by
applying building energy simulation with the building’s computer aided design model, the potential problems of the
buildings can be predicted. The fluctuation of indoor temperature, timely energy consumption of a mechanical
ventilated zone, and yearly energy use of the whole building are examples of building energy simulation.
Computational fluid dynamics (CFD) is one method used for complex calculations, which deals with problems related
to fluid flows. For example, in architectural and urban design field, CFD is very helpful for ventilation and air flow
analysis in the built environment (Chung et al. 2010).
While the building’s performance can be improved during the building’s operation stage, it is best to comprehensively
simulate the possible scenarios during the early design phase. Building energy simulation can be classified into six
categories with subdivided simulation functions, especially aimed at net-zero energy building (Attia et al. 2011) (Table
1.1). Different simulation software may include different features related to these functions.
Table 1.1 Building energy simulation category (based on Attia, 2011)

1.2 Building information modeling (BIM)
Building information modeling (BIM) is often misunderstood as a specific brand of software or is confused with CAD.
In fact, it is neither a single software nor CAD. Computer-aided design (CAD) is traditionally means using program
to produce two dimensional drawings of objects e.g. buildings, mechanical components, electrical layouts, etc., or
more recently it might even refer to 3D modeling. BIM is a different paradigm (Kensek 2014). It is an integrated
2

database of building components that combines building information throughout the building’s lifecycle including 3D
graphics, parametric modeling, and user-supplied data that creates the virtual design and construction model. BIM
provides the opportunity for each discipline to talk to each other by offering exchangeable data format. No matter
whether the users are architects, structural engineers, mechanical engineers or construction workers and operators, the
same database is provided to share the building’s updates in time, which avoids unnecessary waste of money and time
(Kensek 2014). BIM is not a single software; however, the selection of a combination of software applied in a building
project is important. Proper selection enjoys more qualifications of data format, level of detail, operability, export
availability, which comprise the concept of interoperability.
BIM is being applied starting from the early conceptual design stage, until the demolition of the building (Fig. 1.1).
The necessity of BIM can be explained in architecture, engineering, construction and operation (AECO) industry.

Figure 1.1 BIM in a building’s lifecycle (Calvert, 2016)
AECO in the building industry refers to architecture, engineering, construction, and operations. BIM is used in all
these areas (Eastman et al. 2015).
Architecture: In BIM software like Autodesk Revit, architects are able to use Revit to do conceptual design, drawing
whatever shapes they need to show clients the idea and making changes at will. What is more, 2D drawings are
available anytime without much extra effort, and the 3D designs are ready for early stage energy simulation.
Engineering: BIM allows structural or MEP (mechanical, electrical, plumping) engineers to use the same architectural
model for engineering design. In this case, engineers do not need to worry about errors caused by inaccurate
dimensions and lack of clarity. For example, a model created by the architect can be updated with additional data and
used by the energy consultant for simulation.
Construction: during the construction stage, the schedule of construction can be made ahead of the real work, avoiding
possible rework cost resulted from inappropriate arrangement. Managers can follow any updates using the BIM
database. Clash detection and construction sequencing are also possible, sometimes starting with the models produced
by the architecture.
Operation: since the building comes with an entire database from the beginning, the operation and maintenance
become more efficient and effective. Every single component can be pursued if any problem happens. Even for
demolition or renovation, the database provides convenience for safe and efficient work.
In order for the models to be passed to the different users and software programs, the data must be able to be translated
3

into a common format. This might be the native format of the BIM software (such as RVT for Revit), or also a format
that only includes the 3F model such as DWG, which unfortunately does not encode all the data in a building
information model (Kensek 2014). To realize the lifecycle support from BIM, currently the BuildingSmart
International is developing Industry Foundation Class (IFC) standard for BIM software interoperability, and
previously Green Building XML (gbXML) was developed for data exchange between energy programs. Both IFC and
gbXML create a common language to transfer BIM information between different BIM and building analyses
applications while maintaining the meaning of different pieces of information in the transfer. This reduces the need to
remodel the same building in each different application. It also adds transparency to the process (Haymaker, et al.,
2007). For example, Autodesk Revit supports import and export file formats (Fig. 1.2). Data (but not the 3D model)
can also be exported using CSV, a common format that can be imported and exported from Microsoft Excel.

Figure 1.2 file format supported by Autodesk Revit Architecture
There are three main characteristics of BIM are data-rich model, object-based tools, and high interoperability in the
whole life cycle. (Eastman 2015):
• Data-rich model: one of the most attractive qualities of BIM is that it comes with data. People can use the
basic elements, parameters to create more complex data like 2D draws, component schedules. By combining
different drawings, architectural, structural, MEP models are created. There is a building information
hierarchy within the BIM model.

• Object-based tool: BIM tool is an object-based innovation that is pretty much like the idea in object-oriented
computer science programming. Architects and engineers can create different types of objects with many
parameters, which make up the richness and thoroughness of a BIM (Kensek 2014) (Table 1.2).
4

Table 1.2 Example of parametric objects (Kensek, 2014)

• High interoperability: interoperability indicates the ability for the data to be exchanged among different
applications within BIM tools (Eastman 2015). As an example of one of the most famous BIM tools,
Autodesk Revit provides more than ten types of data format including IFC, the common BIM “language” to
make it possible talking with other BIM tools within the entire building lifecycle (Fig. 1.2). National BIM
Standards project (NBIMS) has been and will be contributing to the interoperability for BIM in a long-term
run (Eastman 2015).
1.3 Visualization
Visualization of complicated figures and words is both an art and a science (Aparicio et al. 2015). The aim of
visualization is to create a bridge between the raw data gathered from all kinds of fields and the people who need them,
which improve the efficiency to understand the raw material for people with whatever backgrounds.
From ancient ages, human being’s ancestors knew how to use graphics to convey information (Tufte 2003). As the
society developed, numbers and languages were invented for better expression. However, nowadays people are
surrounded by too much information living in this ‘big data age, leaving people difficulties to distinguish the most
important and conclusive data for future use. Visualization helps us out.
Edward Tufte who is noted as a pioneer in data visualization field, defines graphical displays and principles for
effective graphical display in the following passage: “Excellence in satistical graphics consists of complex ideas
communicated with clarity, precision and efficiency.” According to Tufte, graphical displays should do the following
(Tufte 1991):
• represent the data;
• trigger the viewer’s thoughts about the substance instead of methodology, graphic design or others;
• clearly express what the data want to express;
• condense the data;
• link large data sets continuously;
• inspire users to compare different parts of data;
• represent the data with at different levels of information, from broad to narrow view;
5

• clarify a reasonable purpose;
• use both statistical and verbal description to make the data set clear.

According to Tufte’s listed requirements of data visualization, there are certain types of graphics commonly used by
almost everyone for decades. The quantity of information the graphic can represent is regarded as “visual dimension.”
For example, the most common bar chart has y axis indicating length/count of a certain unit, category and color, which
are the four visual dimensions of a bar chart. Color is an optional parameter that can be removed in the discretion of
users. There are many ways of visualizing graphics that are commonly used (Table 1.3).

Table 1.3 Examples of data visualization (Images by Li, Kelin)

Based on the design of two-dimension visualized graphics, there are three-dimension graphics extended with more
visual dimensions. For instance, when recording weather, a bar chart with two visual dimensions can store
temperatures of different cities of one day; three visual-dimension chart, on the other hand, is added a visual dimension
date, which can record temperatures of different cites of several dates (Fig. 1.3). To be more creative, it is possible for
people to use virtual reality to experience how the data influence the analyzed scene of the data.

6

Figure 1.3 2D and 3D bar chart comparison, another visual dimension is added (Images by author)

1.4 Virtual reality (VR)
The idea of virtual reality is to create an imaginary environment using advanced computing technology, that would be
indistinguishable from the real world (Gutierrez et al. 2008). When talking about virtual reality, people may first think
of using computers to create the 3D virtual environment that users can navigate and interact (Gutierrez et al. 2008).
Most VR relies on the fact that humans have two eyes set apart that provide the brain two slightly different images.
Stereopsis is known as the experience human being’s brain combining images from left and right eyes, in order to
create the three-dimensional stereoscopic image (Howard 1995). Binocular vision is used to describe the way people
perceive the world with a slight difference between left and right eyes at the same time. And the difference between
what the left and right eye generate is called binocular disparity. As a result, stereopsis occurs whether people are
seeing different viewpoints of the physical world from each of the two eyes or two flat pictures with appropriate
differences between them (Oculus Documentation 2017). For example, to make the stereopsis, two separated images
of left and right eyes of an object are shot (Fig 1.4). By overlapping the two images and filtering them into anaglyph
images, the stereopsis can be observed with blue and red anaglyph glasses (Fig. 1.5).

Figure 1.4 Stereoscopic shooting of left eye (left) and right eye (right), the difference is slight, which is like the human’s left and right eyes
(Images by author)
7

Figure 1.5 Grey anaglyph combing left and right eyes images. With blue and red anaglyph glass, the 3D stereoscopic image can be seen i.e.
stereopsis. (Images by author)
Besides the simple example of stereoscopic image, more advanced but low cost 3D head-mounted display (HMD)
with calibrated lenses are available to see images or videos via phone (Fig. 1.6)

Figure 1.6 Users are able to see images and videos from the phone screen through the lenses which correct the images. (IQ by Intel 2014)
As VR technology develops, stationary views using stereoscopic glasses cannot satisfy the need users require. Higher
immersive design and interactions with the environments are expected, as to better create the reality. Thus, some high-
end VR headset with computer-aided virtual reality design emerged to enrich the applications of VR.
Two important parameters, immersion and presence, that are useful for developers, users to evaluate the VR:
“Immersion is related to the physical configuration of the user interface of the VR application,” and “presence is a
subjective concept, associated with the psychology of the user” (Gutierrez et al. 2008). There are three categories of
8

immersive VR systems: fully immersive (VR with HMD), semi-immersive (CAVE), and non-immersive (desktop-
based VR) (Gutierrez et al. 2008). As for presence, i.e. the psychological status, is highly depending on the judgement
by users themselves. For example, if the VR environment with sea view and the sound of the wave aiming at calming
people down helps people ease their anxiety greatly, it can be considered a fully presence VR (Wang et al. 2006).
Tracing back the history of VR, the milestones were established from 1960s (Fig. 1.7). The Sensorama, a multisensory
vehicle simulator, was invented by Morton Heilig in 1960s. Users could sit in front of a screen to experience the
prerecorded rides by choosing from motorcycles, bicycles and even helicopter (Sensorama Machine 1962). The Aspen
Movie Map provided users with simulated rides through the city of Aspen, Colorado and allowed users to choose
different directions, which were considered the start of VR interaction, and in 1980s and 1990s, VR head-mounted
display (HMD) and the CAVE room with projected graphics were experimented gradually (Gutierrez et al. 2008).
Since 2000, various topics towards VR applications are blossoming, which are driving the development of VR related
equipment (Wang et al. 2006).

Figure 1.7 Brief history of VR (Gutierrez et al. 2008) [1] http://www.mortonheilig.com/InventorVR.html;[2]
http://www.inventinginteractive.com/2010/03/18/aspen-movie-map/; [3] https://www.nasa.gov/ames/spinoff/new_continent_of_ideas/; [4]
https://en.wikipedia.org/wiki/Cave_automatic_virtual_environment#/media/File:CAVE_Crayoland.jpg; [5]
https://store.google.com/product/google_cardboard; [6] https://www.amazon.com/Oculus-Rift-Virtual-Reality-Headset-Pc/dp/B00VF0IXEY; [7]
https://arstechnica.com/gaming/2016/10/best-vr-headset-2016-psvr-rift-vive/; [8] https://www.wareable.com/vr/best-smartphone-headsets-
mobile-vr-apps-1655.)
• Display
Displays in VR are the googles that displays the virtual environment for the user. In general, there are six categories
of displays, and the degree of immersion are different: desktop displays, head-mounted displays (HMD), arm-mounted
displays, single screen displays, surround screen displays, and volumetric displays (Shneiderman 2010; Stuart 1996).
Among these displays, head-mounted displays are able to provide better stereoscopic views and wearing-convenience
is better. Table 1.4 lists the popular HMD in the market.
Table 1.4 HMD available for the market [1] htcvr.com; [2] oculus.com; [3] sony.com; [4] samsung.com; [5] https://vr.google.com/cardboard/;
[6] https://www.microsoft.com/microsoft-hololens/en-us; [7] razerzone.com; [8] getfove.com; [9] zeiss.com; [10] avegant.com; [11]
freeflyvr.com; [12] https://www.vuzix.com/Products/m100-smart-glasses; [13] http://atheerair.com/; (based on Viatechnik 2016)
9

• Input device
Input device offers users possibility to interact with the virtual environment. For example, in the set of HTC VIVE
devices, there are a couple of joysticks. While users are wearing the HMD, they can use the joysticks to point objects,
change scenes to play with the environment, which bring more experience to users. There are also simple types of VR
googles such as Google Cardboard that are no need of input device. Users are capable of simply walking around and
watching things using these googles.
• Software
VR software has to contain 3 major functions: ability to render, interaction with the environment, and create the
simulation (Jubertie 2007). Render indicts how the software can convert the information users need of the virtual
environment into readable images by the displays or the googles; the ways people can play with environment are the
interaction; and the simulation means the function and evolution of a VR application.
Given the three modules as a framework for thinking about VR software, developers choose the software and platforms
based on their needs. However, there is a problem that major virtual reality 3D modeling software are not the primary
choice for architecture 3D modeling; major virtual reality game engines such as Unity, Unreal Engine 4 (UE4) and
Autodesk Stingray mainly talk to the so-called digital content creation (DCC) systems (Boeykens 2013). DCC system
is a way that a category of software is used to provide the combination functions of modeling, rendering, and
animation. 3D modeling programs like Autodesk 3ds Max or Autodesk Maya are of DCC systems. 3D models from
such software can be read and saved directly inside the game engines. Otherwise, for other 3D modeling software that
10

are not part of DCC systems, the intermediate is needed for importing. For example, from Autodesk Revit to Unity,
the user needs to export FBX file from Revit first and then import the Revit FBX file to Autodesk 3ds Max. Material
and model adjustments are allowed in 3ds Max; the next step is to export FBX file from 3ds Max and import to Unity
3D (Fig. 1.7). Besides the example, commonly used 3D modeling software in architecture field are available for VR
modeling via the intermediate software (Fig. 1.9). Currently, interoperability between software programs is not
seamless, and often the user will have to move the data between the BIM application to the VR application by going
through other software programs.

Figure 1.8 Example importing architecture 3D modeling software Autodesk Revit models into game engine Unity

Figure 1.9 Importing 3D architecture modeling to game engines
For game engines, there are three programs enjoy the most popularity: Unity, Unreal Engine 4, and Autodesk Stingray
(Table 1.5). They vary in operation, supported languages and scripting methods, and other internal simulating and
rendering methods. One advantage may attempt architects and engineers when choosing a game engine to work with
is whether they are visual programming available. In visual programming environment, users are able to create the
script they want using program elements or modules of different functions, instead of writing line-to-line detailed
scripting (Fig. 1.10). This is beneficial for users without coding background like many architects and engineers.
Table 1.5 Brief comparison of three mainstream game engines

Figure 1.10 Visual programming script in Autodesk Stingray (left), UE4 (middle), and coding in Unity (right).
11

In addition to VR, there is augmented reality (AR) which is a variation of VR, creating imaginary objects based on
the real-world objects recognition (Furht 2011). Augmented reality is an overlay of VR on a view of the real world in
real time simultaneously. The workflow to realize AR is different from VR. Because it needs the information of the
real world, the first step is to record, i.e. deliver the real-world imagery to the computing system, and then the real
world information will be processed by the vision based tracking algorithm. In order to correspond the virtual data
with the real data, the tracking process is very crucial. The position and direction of the object that the user is going
to add animated things on should be tracked using mathematical algorithm (Wang et al. 2006). After tracking
successfully, the final step is rendering, which means to apply the digital data on top of what have been record in the
desired place (Kalkofen et al. 2011). Researchers have applied AR to view the interior lighting condition using head-
mounted display (Fig. 1.11).

Figure 1.11 User is able to view the lighting condition of an interior space using head-mounted display (Kensek et al. 2000)
For AR systems, there are four parts: displays, input device, tracking, and computers (Carmigniani et al. 2011).
• Display: There are three types of displays including head-mounted display (goggles), handheld displays
(smart phone), and spatial displays (mirror-like and station-based display that can trace the user’s movement,
e.g. the AR fitting equipment in cloth stores that people can watch themselves while selecting clothes from
the screen to “dress up” users);
• Input device: the input device for AR works the same way as joysticks in VR. There are other kinds of input
device like gloves, wristband, or smart phone itself;
• Tracking: digital cameras, optical sensors, GPS, accelerometers, solid state compasses, wireless sensors, etc.
These tracking devices can provide digital and numerical data of the aimed object, which is the basis to be
added visualized information or graphics on;
• Computer: computers are need no matter in VR or AR, high speed of CPU and large amount of RAM are
required. A smart phone can have enough computing power for some VR applications.
12

The attractive merit of AR is that it can reveal hidden objects, like the weather or environment information existed in
the space but people cannot see; and satisfy people’s imagination based on the real world, such as the interior design
of the space (Carmigniani et al. 2011). In architectural-related scientific visualization field, researchers have created
a human-building interaction model, that users can interact with the visualized CFD simulation data using gestures
while wearing head-mounted display (Malkawi et al. 2005). False color isoplanes represent the visualized CFD result
measured by wireless sensors (Fig. 1.12).

Figure 1.12 Human-interaction model applied visualized CFD data and AR technology (Malkawi et al. 2005)
Besides VR and AR, mixed reality (MR) is combining all the technologies to relate real environment and virtual
environment (Fig. 1.13). At first, unreal objects are included according to the real-world objects; gradually the
immersion increases as the real-world objects reduce; finally, with virtual reality, users are experience the digital
world entirely (Milgram et al. 1994).
13

Figure 1.13 Virtuality continuum by Milgram and Kishino (Milgram et al. 1994)
1.5 Thesis framework
BIM, energy simulation (specifically CFD), scientific visualization, and virtual reality combined into the topic area
of the thesis. Abstract, hypothesis, objectives and chapter outlines are given.
1.5.1 Abstract
Building energy simulation is an important procedure during the building’s design. Not only can building energy
simulation help with saving energy, but if interpreted correctly and implemented, it can contribute to the occupants’
comfort. Scientific visualization has been adopted for a long time in engineering field, tracking large scale simulation
data and providing intuition and understandable graphs and models displaying the data. For computational fluid
dynamics (CFD) data, the need of scientific visualization is of more importance, due to the complicated spatial data
structure and large quantities of data points characteristic of CFD data. In the CFD simulation engines, there are built-
in CFD simulation data visualization method such as streamlined 3D air flow visualization of an analyzed space. The
raw CFD data could also be exported and processed in other software programs to add clarity.
Given the consideration of better taking advantages of the CFD results for buildings, the potential of the use of virtual
reality (VR) techniques cannot be overlooked. VR techniques bring about immersion and presence (Gutierrez et al.
2008). One can start with a building information model (BIM), produce CFD simulation results, and visualize those
results in VR. There are existing achievements of CFD data or similar energy simulation data visualization in VR
environment, such as visualizing Ecotect simulation data in VR game engine 3DVia Studio (Bahar 2014) and
visualizing OpenFOAM CFD simulation data in Unity 3D (Hosokawa 2016). Some researchers have created CFD
visualization tool as a way to streamline the workflow from CFD raw data to VR environment (Berger et al. 2015).
The existing methods have given positive answers of the possibility to convert numerical CFD data to virtual
environment objects. However, the previous workflows include problems in data exchanging complication (the
software applied may be from totally different working areas), incomplete data representation, etc. As a result, a BIM
to CFD data to VR visualization workflow is proposed.
14

Building information modeling (BIM) as a lifecycle tool for buildings as includes as much as possible information for
further applications. The proposed methodology starts from the building information model Autodesk Revit. CFD
simulation is followed using Autodesk CFD Simulation through the add-in tab in Revit. Multiple kinds of CFD
simulation including indoor natural and mechanical ventilation and outdoor wind pattern simulation can be done. The
data type of nodal is exported for later visualization processing. To realize the visualization in VR, Autodesk 3Ds Max
and Autodesk Stingray are applied continuously to manage the data and generate the VR objects. An actual case study
of a kitchen space works is applied with the proposed methodology. In the final result, users are able to interact with
the virtual environment wearing VR headset like HTC Vive, observing visualized CFD streamlines with color.
Evaluation of the methodology is based on CFD data representation accuracy, user’s experience in both immersion,
and presence perspectives and BIM integration level of the whole workflow.
1.5.2 Hypothesis
Starting with a building information model, architects and engineers are able to visualize CFD simulation results in a
virtual reality VR environment that are able to aid scientific analysis as well as representation.
1.5.3 Research objectives
Four research objectives have been set based on the overall goal to create a BIM model that can be used for CFD
simulation and the results viewed in a VR game engine:
• Complete the CFD simulation using the imported BIM model;
• Visualize the CFD simulation result and realize VR accessible environment to display the data;
• Determine an overall methodology that can integrate into the BIM workflow.
1.5.4 Chapter outlines
Chapter 1 Introduction. This chapter introduces the concepts of energy simulation, data visualization, building
information modeling, and virtual reality.
Chapter 2 Background. The background chapter provides a detailed description of energy simulation data, especially
computational fluid dynamic (CFD) data, how can work for scientific visualization, scientific visualization of CFD
data, and the way to realize scientific visualization in VR.
Chapter 3 Methodology. Chapter three describes the workflow of modeling in BIM integrated software, energy
simulation in CFD program, visualizing in three-dimension modeling software and realizing VR in game engine.
Chapter 4 Case Study. Chapter four demonstrates the methodology to test a specific case study of an interior space
with the methodology and shows the results.
Chapter 5 Evaluation. Chapter five presents the findings, discuss the findings according to case study, and evaluates
the overall BIM workflow’s pros and cons and visualization quality.
Chapter 6 Conclusions and Future work. The final chapter summarizes the whole research, states conclusions from
methodology and results analysis, and discusses what should be done to improve the proposed method as future work.
1.6 Summary
It is important to include energy simulation in the design of a building project. BIM has its strength to cooperate and
interact while increasing the whole lifecycle efficiency. At the same time, scientific visualization is common and
helpful during the analysis of the energy simulation results. VR, as a renovated visualization technique has the
potential to be applied for scientific research. It is worth considering the integration of these four areas.

15

2. Background of energy simulation result visualization in virtual reality
Fulfilling the visualization of energy simulation result demands a high-level understanding of the energy simulation
data, BIM tools, data visualization, as well as virtual reality. This chapter is going to introduce background material
and previous research, including the characteristics of energy simulation data, BIM tools to realize the simulation,
the principles of data visualization, and the application of virtual reality in building science field. Background of
2.1 Understanding of energy simulation data
To realize the goal of energy simulation data visualization, it is important to learn about the data first. Because of the
diversity of energy simulation, the type of the data varies depending on users’ needs. Deeper background research of
energy simulation, different types of energy simulation data, and thus CFD data are given.
2.1.1 Introduction of energy simulation
Throughout the building’s lifecycle, it is of great importance to study energy performance and thermal comfort
(Maile et al. 2007). Some projects take more effort to reach their energy performance goals, at the same time
hopefully of not sacrificing aesthetic or other goals. High performance buildings are sought-after among architects,
engineers and stakeholders because that they bring benefits not only to the global environment but also to the
financial investment in the long run (Nall 2011). The building’s performance can be improved during the whole life
cycle of the building project; however, a great deal of efforts should be put in the design phase, which requires
different types of energy simulation.
There are six different stages that a building project must experience (Laine et al., 2007):
1. feasibility study based on the site information;
2. programming to determine the functional requirements and space needs;
3. schematic design for early architectural design without yet thinking of things like R-values and specific
materials;
4. detailed design with materials and specific windows and doors;
5. construction for the actual creating of the building; and
6. operations and maintenance after the building is done where lifecycle building data is important.
During each stage, some energy simulation can be carried out and last throughout the project. For example, in a
zero-energy building project, there are four segments (comfort and climate, passive solar, energy efficiency,
renewable) including more than 20 types of energy simulation during the whole life cycle (Fig. 2.1) (Laine et al.,
2007).
16

Figure 2.1 different types of energy simulation help with decision making during different stages of a project for a zero energy building project
(based on Laine et al., 2007 and Attia et al. 2011).

2.1.2 Different types of energy simulation data for visualization
As there many types of energy simulations, a variety of data types are provided. There are two main kinds of output
from energy simulation software: numerical-based output and graphic-based output. (Bahar, 2014). These two types
of contents cover almost all the simulation results formats, such as the energy simulation results which can be
represented either in graphic-based (daily, monthly and yearly) reports with graphs or raw numerical sheets. The report
from Revit built-in energy simulation tool gives an example of the graphic-based output (Fig. 2.2 left). The numerical-
based data is in detailed raw data often in XML or CSV format. For example, the solar insolation simulation results
from Autodesk Insight 360 includes the coordinates of each calculated point and specific number of that point (Fig.
2.2 right).

Figure 2.2 Time-based building performance evaluation from Autodesk Revit (left); numerical solar insolation data report (.csv format) from
Autodesk Insight 360 (right)
17

However, there may be possible inaccuracy caused during the workflow of data exchange. For example, during the
design stage, a building is modeled in the 3D modeling software Rhinoceros (abbr. Rhino, a commercial 3D computer
graphics and computer-aid design application software). All the 2D drawings deliverables are exported from the 3D
model. Besides, at the same time the 3D model is going to be imported into energy simulation engines for further
analysis, e.g. IES-VE. Although the compatibility has developed to allow data exchange among different software,
there is no guarantee that all the details are recognized by the simulation engines (Hijazi 2015). For example, as the
model is imported from Revit to IES-VE through the IES-VE Revit add-in tool, there is the possibility that some of
the windows are recognized as walls and roofs may be recognized as walls. Although the IES-VE add-in is designed
for Revit, these kinds of errors due to data exchange are inevitable.
In addition, analysists need to assign different settings to the model back and forth to make sure the building reach the
best performance, as a “round-tripping” process during the whole design stage (Hijazi 2015). As a result, it is necessary
to use software that compatible with each other or to make the workflow more streamlined to reduce the loss of detain
and accuracy in the building’s whole life cycle.
The common energy simulation software are available for different applications, with a variety of input and output
data formats (Table 2.1). The input data formats mainly indicate geometry files that are used for energy simulations.
Besides these formats, the EPW files are the weather data providing the basis weather information of the building for
energy simulation including temperature, humidity, velocity etc. collected from weather stations of places around the
world (Fig. 2.3).
Table 2.1 Overview of common energy simulation

18

Figure 2.3 EPW weather data (.csv)
2.1.3 CFD data
Computational fluid dynamics (CFD) is a branch of fluid mechanics that utilizes numerical methods to solve and
analyze problems involving fluid flows (Chung et al. 2010). CFD has been commercially available since the early
1980s in the engineering community for applications such as turbo machinery, aerospace, combustion, and mechanical
engineering (Kaijima et al., 2013). CFD is one method used for complex calculations, which deals with problems
related to fluid flows. Air is a type of fluid that can be studied by CFD.
Dividing the common CFD software from an aspect of availability, there are commercial and non-commercial CFD
simulation software (Table 2.2). For non-commercial software, they not designed and developed for licensing or sale
to end users or that serves a commercial purpose, like OpenFOAM. Some are commercial software like Autodesk
CFD, ANSYS FLUENT. They also differ from each other by focusing on different fields, e.g. architectural CFD or
industrial CFD. The calculation method each program adopts also vary depending on developers.
What is more, the data formats exported are different from program to program, therefore, the ways to post-process
or visualize the numerical data are different as well. For example, OpenFOAM is able to export a data format called
VTK file, which can be read directly by post-processing scientific visualization program like Paraview, taking
advantages of Paraview’s quick and easy tools to create the visualization result (OpenFOAM 2016) (Fig. 2.4);
However, in Autodesk CFD, the results can be exported as nodal results (CSV) file, which can be processed in
Autodesk 3Ds Max for 3D visualization results (Fig. 2.5).

Figure 2.4 Post-process CFD data in OpenFOAM using built-in tool simpleFOAM (left) and add-in tool Paraview (right) (Image by
OpenFOAM )

Figure 2.5 Post-process data in Auotdesk CFD 2016 built-in visualization tool (left and middle) and in Auotdesk 3ds Max (right) (Image by
author)
The variety of output data formats lead to more possibilities to post-process the numerical data into more graphic
information. As other energy simulation software, CFD software have different interoperability as well. The
interoperability can be obvious in two stages in CFD software. On one hand, before the simulation starts, directly
19

importing geometry from other modeling software can be efficient. For example, in Autodesk CFD, the program can
import Revit file (RVT file) directly avoiding model re-build step. On the other hand, after the simulation in data post-
processing step, high interoperability allows the CFD software export multiple data formats that can be used in other
programs. The example of OpenFOAM using the software Paraview for data post-processing is a valuable
demonstration.
Table 2.2 Overview of common CFD simulation software

From an architectural perspective, the application of CFD simulation can be divided into two areas: indoor and outdoor
simulation (Chung et al. 2010). For indoor environment, the main usage is to simulate the natural ventilation or
mechanical ventilation, heat transfer of certain spaces; for the outdoor environment, the wind flow (velocity, pressure,
turbulence and temperature) is calculated. For example, by setting up the inlet and outlet in a kitchen space, the air
flow is studied and visualized based on the numerical data in Autodesk CFD 2016 (Fig. 2.6).

Figure 2.6 A CFD study of the air flow in a kitchen space (Image by author)
2.2 Applying BIM for energy simulation including CFD simulation
Building simulation and analysis is a part of the building’s lifecycle (Autodesk 2016) (Fig. 2.7). The simulation and
analysis results enrich the building information models. There are a large number of simulation tools as BIM tools
concerning energy simulation in buildings. Research on energy simulation programs for net zero energy building
(NZEB) resulted in 10 energy simulation software programs graded and interpreted by their performance including
usability, accuracy, intelligence, process adaptability and interoperability (Attia et al. 2011). As well, quantitative
20

benchmarks were established to compare the capability of these software from aspects as metric, comfort level and
climate, passive strategies, energy efficiency etc. The results of this study included diagrams showing the usability,
intelligence, interoperability, process adaptability, and accuracy of ten energy simulation software (Fig. 2.8). The
interoperability of the researched programs is not very high, allowing easy exchange either of geometry or output data
(Attia et al. 2011). Researchers studied the interoperability between a BIM-based architectural model and performance
analysis programs (EnergyPlus, eQUEST, Ecotect, IES) based on gbXML protocol and concluded that users should
select appropriate analysis programs considering the interoperability of energy analysis programs (Moon et al. 2011).

Figure 2.7 Building simulation and analysis is a part of the building’s lifecycle (Autodesk 2016)

Figure 2.8 Results of the NZEB tools mechanics (Attia et al. 2011).
Researchers have illustrated the workflow that defines the area input of BIM in the energy simulation process (Maile
et al. 2007). The first step Maile et al. did was to define the location of the building that provides a link to weather
21

data. The second step ideally provided information through importing data from a BIM (Fig. 2.9). This information
includes the needed 3D geometry information, material definitions, and space types that are typically defined by the
architect. From the workflow, it is worth noticing that every step has the potential to lose a certain amount of data,
which leads to the result that the fewer changes of simulation software, the more accurate of the simulation.

Figure 2.9. An ideal workflow for energy performance in energy simulation tools (Maile et al, 2007).
Considering previous studies, using energy simulation of high interoperability is crucial to avoid data loss during
exchange. In addition, the emphasis on high-performance building requires BIM-based thermal analysis during the
design (Welle et al. 2011). BIM tools help with better decision-making for a building because of building information
models’ coordinated and consistent information (Bahar 2014). As a result, energy simulation programs with high
interoperability enrich the information of the building information models, and BIM tools make energy simulation
easier and more accurate.
CFD simulation, as part of the building’s analysis, is beneficial to be carried in software with high interoperability.
Applying BIM tool like Autodesk Revit with CFD simulation helps decision-making of the design (Fig. 2.10)
(Bergman 2012). The streamlined workflow included geometry modeling in Autodesk Revit, simulation and
visualization and analysis in Autodesk Simulation CFD (Bergman 2012), avoiding as much data loss as possible
during the “round-tripping” process (Hijazi 2015). If the CFD simulation software has better BIM support, the “design-
to-analysis” workflow is more convenient and precise for users because of the interoperability between the geometry
input software and simulation software. Besides, the post-process is also part of the interoperability. Multiple output
data formats allow more post-process software to read and use. For example, in IES-VE MicroFlo, the data exchange
exists within the software itself, there is no possibility of using the simulation results for visualization in other
visualization software (Table 2.2).
22

Figure 2.10 Workflow of taking advantages of BIM for CFD simulation
2.3 Scientific visualization for CFD simulation
The post-process of simulation data results is important for decision making. Proper visualization makes the data
straightforward and graphically favorable.
2.3.1 Introduction
Visualization has been divided up into three main types: scientific visualization, data visualization, and information
visualization (John et al., 2001).
• Scientific visualization: visualize scientific simulation results to help scientists better understand and
analyze the results (Friendly, 2008).
• Data visualization: present data graphically to lets people see data presented visually, in which way they
can get difficult concepts or identify new patterns (SAS, 2017).
• Information visualization is the study of (interactive) visual representations of abstract data to help people
understand better. The abstract data include both numerical and non-numerical data, such as text and
geographic information (Wikipedia contributors. "Information visualization." Wikipedia, The Free
Encyclopedia. Wikipedia, The Free Encyclopedia, 10 Feb. 2017. Web. 28 Apr. 2017).
There are debates about the difference and application among these three types. One view holds that scientific
visualization indicates the visual representations of scientific data from simulation results to make the data
understandable (Nagel 2016).; instead, information visualization is to visualize abstract information that is not from
simulation, such as the directory or file structures on a computer, the information content of books and the website
structures (Bahar 2014).
Information visualization is the broadest term that could be taken to subsume all the developments
described here. At this level, almost anything, if sufficiently organized, is information of a
sort. ... scientific visualization ... is primarily concerned with the visualization of 3-D+ phenomena
(architectural, meterological, medical, biological, etc.), where the emphasis is on realistic renderings
of volumes, surfaces, illumination sources, and so forth, perhaps with a dynamic (time) component.
Instead, we focus on the slightly narrower domain of data visualization, the science of visual
representation of "data", defined as information which has been abstracted in some schematic form.
(Friendly and Denis 2014)
23

In the architectural domain, the energy simulation results could benefit from better scientific visualization in
addition to realistic renderings of volumes, surfaces, illumination sources, and so forth, i.e. the spaces and their
ambient environment. This is especially important for time-based results.
2.3.3 Scientific visualization for CFD simulation
For numerical data such as EPW weather data and solar insolation data from Revit in CSV data format, they allow
data post-process in many ways in order to visualize the data. For example, the weather data is manageable using
different software with CSV reader tool. Researcher had tried to process weather CSV data downloaded from
Wunderground.com in Unity 3D (game engine) by programming to plot the weather data in 3D space (Hodgson 2013)
(Fig 2.11). Although the test provided a good method to visualize weather data, the information displayed is partially
confusing to users and the meaning to plot the weather data of different dates around 3D space was not sufficient.
From the architectural design perspective, the CSV data can be managed in parametric design tools like Dynamo for
Revit, Grasshopper for Rhino. An experiment was made using solar insolation data from Revit to generate the 3D
visualization result by visual programming in Dynamo, to show the solar insolation value of each simulated points on
a façade (Fig 2.12). The visualization results were not ideal, because on one hand, the graphics with bumpy meshes
did not decently represent the data information; on the other hand, even the graphics were designed better, the
calculation speed to generate the visualization was much longer than expected. For large-scale data sets, this method
is inappropriate. The previous research indicated that the visualization should be meaningful and clear no matter
whether they are presented in 2D or 3D. Proper visualization tools should be applied to guarantee the visualization
process goes smoothly.

Figure 2.11 Weather CSV data downloaded from Wunderground (left); programming in Unity 3D (middle); visualization results in Unity 3D
(right) (Hodgson 2013)

Figure 2.12 Solar insolation results from Revit Insight 360 (left); Dynamo script to process the CSV results (middle); bumpy meshes represent the
value of each calculated point on the façade (right) (Images by author)
CFD simulation often contains high-dimensional data in a three-dimensional volume. The display of phenomena
associated with this data may involve complex 3D structures (Bryson, 1996). CFD results are often contain hundreds,
if not thousands, of points’ coordinates and each point’s several parameters with values, recorded in numerical data
spreadsheet (Fig. 2.13 left). If people read the numbers, they lose the idea of where it is that the point locates. If these
points are mapped into the space, it may be difficult for people to tell the values and confused by these points. In order
to clarify and present the accurate simulation results, certain visualization methods are in need to explain and interpret
the data points properly. Moreover, for architectural wind flow simulations, the big picture means more than the single
point, which means the tendency of the wind and what is good or bad for a certain area of the building count more.
Cloud points of a mechanical ventilation CFD simulation for an office, easily causes confusions of what the points
24

stand for, why a part of the point cloud is intense while other parts are loose, whether each point indicates a specific
value (Fig. 2.13 right).

Figure 2.13 Cloud points spreadsheet (left); cloud points of CFD simulation in Autodesk 3Ds Max(right)
The visualized CFD data can be still images including 2D graphs and 3D spatial images, as well as animation
indication the direction of air flows using built-in visualization tools in the software. 2D graphs are plotting the data
using XY axis (Fig. 2.14 left) and 3D spatial images are screenshots from the visualized 3D CFD data (Fig. 2.14
right).

Figure 2.14 2D graph (Image by OpenFOAM user guide) (left); 3D spatial images (Image by author) (right);
For commercial CFD simulation engines such as Autodesk CFD, ANSYS Fluent, IES-VE MicroFlo, 3D
visualization has several representation types with false color representing number of degree. The common types are
plane, volume, vector, particle tracking lines (Fig. 2.15). Besides, the numerical data exported is a kind of data
representation.
• Plane: plane is a section of the 3D simulated model with simulation data (Fig. 2.15 upper left);
• Iso-surface: iso-surface is a surface that has values on it. Different from the plane, iso-surface can be a
combination of surfaces of an object (Fig. 2.15 upper middle);
• Iso-volume: iso-volume is the volume of the part with a certain range of values (Fig. 2.15 upper right);
• Vector: vector representation uses the plane as a source to create the arrow-like flowing effect indicating
the direction of simulated flow (Fig. 2.15 lower left);
• Particle tracing lines: particle tracing lines uses the plane as a source to trace the path of each points on the
plane, from inlet to outlet in the simulated space (Fig. 2.15 lower middle);
• Numerical data exported from CFD simulation (Fig. 2.15 lower right).

25

Plane iso-volume iso-surface

Particle tracing lines vectors numerical data
Figure 2.15 CFD data visualization in plane, volume, vector, particle tracking lines (Autodesk 2017)
As for the animation of the results, they are based on the 3D still visualization. For example, the particle tracing
lines can be animated recording how the air flows from inlet to outlet (Fig. 2.16).

Figure 2.16 Air flow animation (Image by author)
In conclusion, the large amount of CFD simulation data are representing in different ways, commonly visualization
results are 3D results of plane, iso-surface, iso-volume, vectors and particle tracing lines. The researched CFD
software Autodesk CFD, ANSYS Fluent, IES-VE Microflo and OpenFOAM all have these kinds of visualizations.
2.4 Understanding of virtual reality (VR)
When talking about the applications of VR from a broad point of view, the first many people think of are VR movies
and games. VR has been adopted by the military for training purposing, e.g. in a real combat situation; education,
for instance astronomy majors are able to observe the stars and planets through the VR headset. VR also has a huge
potential to develop in rail construction, car design etc. There are also uses for VR in art, which may contain creative
art work viewed in immersive environment or masterpieces from great artists who has passed away (VRS, 2017).
VR can be used in creative ways in almost any field. The use of VR could be helpful with visualizing big data also.
Specially, VR can be used for scientific visualization in architect applications.
2.4.1 Applications of VR in architectural field
26

Not only the data can be interpreted by VR engines, but also the 3D model of the building project can be seamlessly
imported into the environment. By combining the result with the original model, VR offers an accurate and intuitive
exploration of the simulation result.
In the building industry, there is no doubt that VR can be utilized. Possible applications of VR from the whole life-
cycle perspective of a building project are investigated previously (Fig. 2.17). Designers are take advantage of VR’s
powerful capabilities for visualization of existing designs. The are many other areas where VR could be used.

Figure 2.17 Application of VR from a life-cycle perspective ([1] https://i.ytimg.com/vi/P_UHZfAtZJg/maxresdefault.jpg; [2]
https://www.texomamedicalcenter.net/sites/texomamedicalcenter.net/files/thumbnails/image/surgery-room.png; [3]
http://www.bandt.com.au/information/uploads/2016/04/IMG_7071-1260x840.jpg; [4] http://www.interior-design-
academy.com/images/2016/interior-design-course10.jpg; [5]image by author; [6] https://4.bp.blogspot.com/-
jjdnDSSFcCw/VtIUCcDr9EI/AAAAAAAAA-g/KGl4ehLD_t8/s1600/2015-02-17_16-44-04-695x522.png; [7] https://nextstl.com/wp-
content/uploads/collaboration-image-2.jpg; [8] http://www.1ssh.com/files/photogallery/preview/Workers-waiting-for-tarp.JPG; [9]
http://symposia.asse.org/images/image2.jpg; [10] https://www.vrroom.buzz/sites/default/files/styles/article_top_banner/public/vr-in-
construction.jpg?itok=18Qf_6os&c=84bb9be94eb0ca84a6b51d24be9c42f1; [11] http://rockwallcontracting.com/wp-
content/uploads/2010/08/Sample-Project-Schedule.jpg; [12] http://www.cs.unc.edu/~marc/tutorial/img848.png; [13]
http://www.roadtovr.com/wp-content/uploads/2014/06/positional-tracking-technology-overview-virtual-reality.jpg.
2.4.2 VR in scientific visualization
VR becomes more and more popular in the field of scientific visualization, based upon using computer graphics to
express complex ideas and scientific concepts (VRS, 2017). VR technology assists the unambiguous display of data
structures by providing a rich set of spatial and depth cues (Bahar, 2014).
There are many topic areas that use scientific visualization that could also use VR, like physics, chemistry, biology,
medicine, astronomy as well as engineering. People may doubt that if most of the simulation programs are able to
export 3D visualized graphics, why VR visualization are crucial. There are a couple of reasons:
27

• When the required simulation space is relatively large: larger space means more enormous amount of data
points. If the data points are pretty intensive, there may be important points being neglected. Walking around
the actual site with simulation result helps greatly to examine every specific area. For instance, comparing
the CFD air flow simulation for a small engine and for a building complex, the representation of the wind
flow for a building complex requires better visualization in VR (Fig. 2.18).
• When the type of simulation need daily experience to examine or analyze: when analyzing heating or cooling
load of a space, architects and engineers care more about the value of a specific area. But when things come
to some daily experience simulation e.g. wind flow and lighting level, it is very important to stand from a
human view point to feel whether the place needs a higher or lower level of ventilation/lighting level (Fig.
2.19).

Figure 2.18 CFD model with cloud points that can be better represented in VR (Kaijima et al. 2013)

Figure 2.19 Lighting illuminance level visualization in VR (Li, Hang 2017)
Virtual reality, to some extent, not only creates the reality human beings live in, but also create the vision that can
never be shown in the real world.
2.4.3 Previous research on VR in scientific visualization
While the scientific visualization in VR is relatively brand new topic, there are great works researchers have developed.
Researchers have developed a Human-Building Interaction model that combines four components including wireless
sensor data, CFD analysis Human-Computer Interaction and AR visualization (Malkawi et al. 2005). The Human-
Building Interaction (HBI) model allows for efficient processing and transfer data. The example is very impressive
with the techniques using gestures to interact with the real world environment and recognizes the importance to
visualize CFD data. After the BIM topic becomes a focus, there is research that talks about visualizing thermal building
simulation data using simulation tool of Autodesk Ecotect Analysis and CAD model tool of Autodesk Revit and 3D
visualization tool of 3Dvia to generate the augmented reality effects of the project Gunzo room (Bahar 2014).
Considering the application of BIM tool, this research offers clues of how to include scientific visualization in the
BIM platform using more simplified workflow. Another methodology provides study to visualize weather data
worldwide. Each region is assigned a texture that may represent the characteristic of the climate. The selection of
28

texture for climates is a good example for weather data visualization considering graphical issue. The management of
data is also a method worth learning from (Tang et al. 2014). There is a previous research on temperature distribution
in a computer room (data center), which is a study of optimum solutions through simulations using VR simulation
technology (Taisei Corporation Japan, 2013). Some brings meterological raw data to 3D visualization using symbols
representing parameters like temperature, precipitation, velocity from a worldwide view and realize VR effects in the
CAVE (Helbig et al, 2014) (Fig. 2.20). Recently a research is completed that by simulating the CFD natural ventilation
of a single room in OpenFOAM, the data are exported in Unity 3D, a VR game engine, programmed using the
visualization toolkit (VTK) to get the CFD wind flow splines in VR environment and observed using Oculus Rift
headset (Hosokawa et al. 2016) (Fig. 2.21). However, there is minimal research on visualizing energy simulation data
in VR environment, especially in BIM workflow (Hosokawa et al. 2016).

Figure 2.20 VR Visualization of meterological data in the CAVE (Helbig et al, 2014)

Figure 2.21 CFD visualization in VR using Oculus Rift to observe (Hosokawa et al. 2016)
2.5 Summary
Three main conclusions were reached about the critical nature of BIM interoperability, visualization in energy
simulation, and the potential for a BIM to VR workflow.
The interoperability of BIM is of great importance to make sure a data-rich and data-accurate BIM model are valuable
throughout the life cycle. High interoperability of the software used in BIM allows the data exchange to happen, so
that the building information can be used for energy simulation or other applications like scheduling, maintenance. In
this way, the energy simulation or other applications enrich the model’s information.
Energy simulation results need a high level of visualization to be understood, interpreted, and explained. Especially
for CFD data, which have complicated structures and multiple parameters should be applied with better and more
intuitive visualization method. Virtual reality becomes a good choice for CFD, as an innovation of graphic
representation.
The workflow of BIM in a whole life-cycle is gradually developing and updating new methods; VR is an innovative
way of viewing and sometimes manipulating data. It can to be added onto the workflow by itself as a design tool, but
also as a method of visualizing simulation data (Fig. 2.22).

Figure 2.22 workflow from BIM to simulation to VR
29

3. Methodology
In order to find out a more efficient way to visualize CFD simulation data in a VR environment, a BIM based workflow
has been proposed, before which a couple of workflows are studied as tests. This chapter documents the overall
methodology applied to achieve the goal, and details each step of the research process, including geometry modeling,
CFD simulation, 3D visualization, and virtual reality. These four parts form the entire workflow. First though, a test
case of VR visualization was attempted using solar radiation data (from a weather file) to BIM to VR.
3.1 Methodology overview for CFD
The methodology was developed based on the background research about visualization of CFD data. CFD simulation
data has the properties of 1) the large amount of data points, 2) spatially distributed around the simulated area. Given
these characteristics the method of VR visualization of CFD data in BIM workflow considers overcoming the
complexity of visualizing data sets.
Before the proposed methodology was fully developed, two workflows were tested as experiments to study the pros
and cons of possible tools to realize simulation data visualization in VR (Fig. 3.1). The first workflow applied a case
study using EPW file as easy-to-grab Excel data set that was imported in to VR game engine Unity 3D (Fig. 3.2). A
scale-changing column was built symbolizing the degree of temperature, which reached the basic goal of visualizing
data in graphics (Fig. 3.3). The script was written in two part: 1) converting the CSV data into TXT data so that the
engine can use it; 2) assigning the values in “Dry Bulb Temperature” column one by one very quickly to scale the box
to different size in order to show the change of the temperature every day in a year. A Unity 3D add-in called Dotween
was applied to make the box geometry transform smoother. However, the disadvantages of this workflow were pretty
obvious. Since EPW files record the meteorological data of a single weather station yearly, there were no XYZ
coordinates to plot each data set (daily data set) with clarity and immersion around the space.

Figure 3.1 Workflow 1 & 2 are non-idea workflows to realize simulation data visualization in VR, and workflow 3 is the proposed methodology
to develop and create demonstrable virtual environment.

Figure 3.2 Workflow 1
30

Figure 3.3 Workflow 1:script in Unity 3D (upper) column scaling based on temperature value in Los Angeles (lower)
The second workflow used geometry from the BIM tool Autodesk Revit to complete the solar study simulation and
generate Excel data sets. Later, the data was applied in visual programming Revit add-in Dynamo as parameters that
controlled adaptive components symbolizing solar study values. Finally, all the visualized components were imported
into Unity to create the virtual environment (Fig. 3.4 and 3.5). This workflow brought data with spatial coordinates
that can create VR immersive effects, at the same time using Dynamo brings convenience for architects and engineers
to achieve the visualization goal. Nevertheless, processing the large-scale data sets in Dynamo consumes unexpected
long time, which limits the visualization greatly. And importing components from Dynamo to Revit to Unity 3D
encounters difficulties such as bringing in the color to Unity that have been visualized in Dynamo.
31

Figure 3.4 Workflow 2

Figure 3.5 Workflow 3: visualizing solar study data sets using Dynamo to parametrically control adaptive components that symbolize the
insolation values.
As a result, after performing previous workflows, the initial goal of showing solar radiation and other weather data
information was changed to viewing CFD data. CFD simulation data have the characteristics of spatially scattered
data points and easily exported data formats that cater the basis of VR visualization, especially for wind flow
simulation such as natural ventilation, mechanical ventilation or urban wind flow simulation. In order to observe the
CFD wind flow using VR HMD, modeling, a third workflow was developed.
3.1.1 Overall workflow
The proposed method included four parts: BIM geometry, CFD simulation, 3D visualization, and virtual reality
visualization (Fig. 3.6):
• BIM geometry: this part included two sub-steps. For the first thing, a completed 3D model containing basic
elements (all/one of architectural, engineering, mechanical contents) were created. Especially for CFD
simulation, in the second step, the model was simplified in advance in order to improve the CFD calculation
speed and avoid unnecessary errors (Fig. 3.6).
32

Figure 3.6 Original geometry (upper) and simplified geometry (lower)
• CFD simulation: after modeling, CFD simulation was carried out by importing the simplified 3D models.
Before starting the simulation, the geometry was set up from aspects of material, boundary conditions,
meshing, and solver requirements. These four setup tasks could be complicated or simplified depending on
user’s simulation accuracy requirements. If the simulation is processed successfully, the ingredient data for
visualization is almost ready for exporting. The second step was to manage the data to trace points from inlet
to outlet of a certain plane, which at the same time creates numerical data (Nodal Results) in CSV format
that can be used for VR visualization.
• 3D visualization: in this step, the detailed BIM geometry was imported into Autodesk 3ds Max. Then the
main step was to visualize the wind flow using Nodal results from CFD simulation by the built-in CFD
Vertices Modifier tool in Autodesk 3ds Max to editable splines that can be imported into VR game engine.
This step provides recognizable material for later steps to create VR environment.
• Virtual reality visualization: this section actually brought users possibility to view the VR results using HMD.
The material created from 3D visualization step was imported as assets into Autodesk Stingray. Interactions
and animations were made in Stingray. The test using HTC VIVE HMD was also completed in the game
engine as adjustments were needed.
33

Figure 3.7 Methodology
3.1.2 Software selection
The selection of the software to complete the test of the methodology was important for the overall efficiency and
effects. According to the background research, the software used for the methodology should have high
interoperability, in order to avoid data loss as much as possible. Besides, BIM tools should be included to guarantee
the richness of the building information models and make the “round-tripping” design convenient among the four
proposed stages.
• BIM geometry: Autodesk Revit was applied as the basis of the methodology. Revit has great user interface
and tutorials for architects and engineers to engage while the powerful modeling and information storage
functions allows models of every kind to be built. Student version for Revit is available for research use.
• CFD simulation: as for simulation, although the representations of the results are adequate of all CFD
simulation software, the priority was on interoperability. On one hand, Autodesk CFD simulation provides
launchers in Revit, from which the model built in Revit can be imported to CFD simulation without extra
labor; on the other hand, Autodesk CFD simulation has multiple choices of output data formats that can be
used in visualization programs. The simulation engine of Autodesk CFD is powerful as well satisfying
interior, exterior air flow simulation, heat transfer simulation etc. A student version is available.
• 3D visualization: the visualization of CFD results was of great importance since the software needs to read
the CFD numerical data results at the same time to generate graphically recognizable representation that can
be used for VR environment. Autodesk 3ds Max works as the intermediate not only converting the Revit
geometry into VR accepted geometry, it also has a significant CFD data modifier that can visualize the data.
A student version is available for 3ds Max.
• Virtual reality: the selection of game engine considered the interoperability at first as well, since the
mainstream game engines on the market are all free to use and have powerful functions. Autodesk Stingray
can talk to 3ds Max directly because there is built-in tab of Stingray in 3ds Max sending geometry and
materials back and forth or linking the files between 3ds Max and Stingray. A student version is available.
Overall, all the software used for the proposed methodology is of Autodesk series, which, to some extent, facilitates
the data exchange and functions compatibility (Fig. 3.8).
34

Figure 3.8 Software selection for the overall workflow
3.2 BIM geometry
BIM geometry, as the basis of all the further process, was created. On one hand, a detailed 3D model was needed to
display the project; one the other hand, a simplified model for simulation use is required as well. From the detailed to
simplified models, both of their level of development and level of details were reduced to some extent. For example,
in the proposed methodology, the detailed model contained furniture with delicately designed details and walls with
parameters of structured layers, the model for CFD simulation only consisted of furniture with basic dimensions
without details (much fewer faces), and the parameters were eliminated to lessen the data exchange process and
calculation burden.
3.2.1 Detailed geometry preparation
Autodesk Revit (Revit) works as BIM tools to provide 3D geometry and building information (Fig. 3.9 & Fig. 3.10).
Users are able to build the necessary elements of the project including (Fig. 3.11):
• Architectural: site, landscape, structural layered walls, floors and roofs, doors, windows, furniture etc.;
• Structural: columns, beams, truss, foundations etc.;
• MEP: duct, pipe, conduit, cable tray, electrical wire etc.
In the proposed methodology, the detailed geometry was used for VR environment, but later parts of it were simplified.
There is no need to bring all these three catogories into VR, while selection depends on user’s preference. Architectural
elements are for design display mostly including interior and exterior. Important and large structural and MEP
elements should also be included if they influence the user’s experience very much. If the VR environment is used for
clash detection study, all these three parts should be contained. For most of the cases, architectural elements are in
primary need (Fig. 3.9 upper). Besides using the model for visualization, the model can be used for CFD simulation,
additional elements are deleted based on the detailed model (Fig. 3.9 lower).

Figure 3.9 Exterior model (upper left); interior model (upper right); exterior model for CFD simulation (lower left); interior model for CFD
simulation (lower right) (sample model from Autodesk Revit was used).
35

Figure 3.10 Autodesk Revit user interface: 1) Application Menu; 2) Quick Access Toolbar; 3) InfoCenter; 4) Options Bar; 5) Type Selector; 6)
Properties Palette; 7) Project Browser; 8)Status Bar; 9); View Control Bar; 10) Drawing Area; 11) Ribbon; 12) Tabs on the ribbon; 13) A
contextual tab on the ribbon, providing tools relevant to the selected object or current action; 14) Tools on the current tab of the ribbon; 15)
Panels on the ribbon. (Autodesk Revit 2017)

36

Figure 3.11 Geometry preparation for geometry simplification and 3D visualization (usee to display the model).
In conclusion, the detailed model is the foundation for later model simplification for CFD simulaiton process and VR
environment built.
3.2.2 Geometry simplification
Based on the detail project model geometry, the model was tailored for wind flow simulation considering the
computing speed and accuracy. There are four items should be reduced or eliminated before importing for simulation
(Autodesk CFD simulation 2017):
• Gaps between different parts and holes: make sure that the space for wind flow simulation is divided clearly
for the computer to recognize;
• Fasteners: delete unnecessary fasteners components that increase the computing faces but have no impact on
flows like lighting switches, detailed buckles for MEP use etc.;
• Large assemblies: large assemblies may include large amount of faces to represent their complicated design
which cause computing burden, while do not influence the flow at all. Large items can be replaced by box-
like mass;
• Interferences: interferences in the conditions should be eliminated, for example, some of the surfaces of a
duct that can confuse the boundary between fluid air and solid objects.
In addition, two types of items should be added for the CFD simulation:
• Exterior void: if the wind flow simulation is for urban wind flow design, like a group of buildings in a certain
area, an exterior void around all the building geometries can be created before simulation representing the
air volume for simulation (Fig. 3.12). The size of exterior void depends on specific simulation requirements.
If the simulation is for indoor wind flows, the program would automatically create the void enclosed by the
boundaries that it can recognize.

37

Figure 3.12 Set the exterior air volume of the object (Autodesk 2017)
• Mass representing large assemblies: given large assemblies may influence the wind flow, especially small
indoor space, it is improper to delete them for the sake of improving computing speed. To compromise,
creating the mass of large assemblies is beneficial, such as replacing large cabinets with simple mass of box
shape.
The detailed geometry and simplified geometry are saved as separated files for later use and retrieve (Fig. 3.13).

Figure 3.13 Original geometry (upper) and simplified geometry (lower)
3.3 CFD simulation
While selecting BIM tools for the completed workflow, interoperability was highly considered because the data the
model generates are supposed to be shared with other application within the building’s lifecycle phases (Eastman
2015). By comparing other commercial or non-commercial CFD software, Autodesk CFD Simulation has the best
interoperability with Autodesk Revit, which is one of the most popular BIM tools. From the data exchange perspective,
it accepts 3D models from many CAD modeling tools including Revit, from the user’s operation convenience aspect,
it has direct launchers for Autodesk Inventor, Autodesk Fusion, Autodesk Revit, Pro/Engineer, UGNX, and Solid
Works (Autodesk 2017).
In the CFD simulation section, there was a pre-simulation process to set up the model for simulation and post-
simulation process to generate the wind flow data for further use (Fig. 3.14)
38

Figure 3.14 CFD simulation process
With the simplified model file open in Revit, Autodesk CFD launcher icon was chosen from “Add-Ins” tab (Fig. 3.15).
the model was imported into CFD simulation without complexity. The design study of the building project was
established by saving the design study to the aimed folder.

Figure 3.15 Autodesk CFD 2016 launcher in Revit

The default user’s is similar to Revit with models and operation commands listed in the viewport directly (Fig. 3.16).
39

Figure 3.16 Autodesk CFD Simulation user’s interface: 1) Graphics window; 2) Ribbon; 3 )Context Panels; 4) Context Toolbars; 5) Right-Click
Menus; 6) Quick Edit Dialogs; 7) Application Menu; 8)Quick Access Toolbar; 9) Infor Center; 10) Design Study Bar; 11) Decision Center; 12)
Output Bar; 13) ViewCube; 14) Selection List; 15) Status Bar; 16) Navigation Bar.
The basic and frequently used operation techniques that cooperate with the functions are selection (left click) and
multi-selection (left click without holding any key), hide and unhide (Control key with middle mouse button click).
3.3.1 Material and devices
Materials are physical substances of the components geometry such as wall, floor, air or water (Fig. 3.17); while
devices are models of physical devices including fans, lights, valves etc.
40

Figure 3.17 Examples of fluid material selections (upper) and solid material selections (lower)
After the model is imported automatically from Revit, all the components of the project are left unassigned from
Design Study Bar and lose their properties because the materials and devices that the user previously set in Revit are
not recognized by CFD simulation for CFD analysis. This is definitely an interoperability problem. Materials must be
assigned before running the simulation; otherwise there will be an error.
To assign materials, select each item in the viewport and right click to choose Edit…. In pop-up window, Type menu
defines the property of the material (fluid/solid) and the Name menu offers a list of materials for the user to choose
from. Optionally, the material may vary according to its environment. In Environment menu, there are Fixed (materials
at a specified environment with constant properties) and “Variable” (materials with varying properties as defined in
the materials) selections. To assign device, the same Material window is used. Once the material is assigned, it will
be listed in Design Study Bar as assigned in different categories (Fig. 3.18). This was done for all the objects.

Figure 3.18 Assigned and Unassigned objects were listed in the Design Study Bar
41

For wind flow simulation, there are two types of material are used: fluid and solid. A fluid is a volume of flowing
object like air, while a solid is an opaque object that with materials like gypsum board, wood. Air volume in the project
should be assigned as fluid type with name of air. Other project elements are assign as solid type with name of default
listed material such as gypsum board, wood, glass. After assigning all the materials, the color legend down the left
corner of the graphics window shows all material in the model (Fig. 3.19).

Figure 3.19 All assigned materials type were showed in color legend down the left in the viewport
Besides the functions above, users can customize all the material by clicking “Edit” button in the material window.
All the default and customized materials can be saved in “My Favorite” list.
3.3.2 Assigning boundary condition
A boundary condition in CFD simulation engine indicts the inputs of the simulation model. For example, it may
represent how the fluid (water/air) goes through the model, and it may also represent the energy interchange between
the models and environment.
For an architectural project, multiple kinds of simulation e.g. flow and heat transfer are available. Changing the
boundary conditions differentiates different kind of simulation. For instance, in air flow simulation, key factors like
velocity, pressure, internal fan, or pump and buoyancy from natural convection are driving the flow. Heat transfer
simulation requires boundary conditions of temperature, heat load (such as an electronic chip), radiation, or resistance
to electrical current. While no matter users want to change velocity, pressure, internal fan or pump and buoyancy from
natural convection, or temperature, heat load (such as an electronic chip), radiation, or resistance to electrical current,
they can only differentiate air flow simulation and heat transfer simulation by change the boundary conditions. There
is no automated selection of air flow simulation and heat transfer simulation.
To set up boundary conditions, user should select the model surface or part of the volume and then click the Boundary
Condition button from Setup tab (Setup Tasks Panel). In the Boundary Condition Quick Setup Dialog window (Fig.
3.20 left), important factors are listed in order to indicate the conditions according to the simulation requirements. As
for different kinds of simulation, such as flow and heat transfer, it is the type setting that user need to modify. The
proposed methodology aims for wind flow simulation for VR visualization, then the inlet and outlet surfaces are set
as Velocity and Pressure respectively. The color legend down the left corner of the viewport shows what conditions
have been applied to the model (Fig. 3.20 right). This was done for the case study.
42

Figure 3.20 Boundary Condition Quick Setup Dialog window (left); color legend of assigned boundary conditions (right)
3.3.3 Meshing
CFD simulation requires calculable nodes, which consist of creating meshes from the model. The nodes are from the
small pieces of elements that are broken up from the entire model. For the most convenience, Autodesk CFD
simulation offers Auto Mesh Sizing by clicking the Autosize icon in Setup tab (Fig. 3.21), Automatic Sizing panel.
The split elements are showed with blue dots lines. By using “Frame line” viewing icon, inner elements can be
observed (Fig. 3.22). This was done for the case study (Fig. 3.23).

Figure 3.21 Autosize tool

Figure 3.22 Frame line mode Autosizing result
43

Figure 3.23 The meshed model
3.3.4 Solver
After setting up all the tasks needed for simulation, the user applies the solver to calculate the model. The solve icon
locates in setup tab, simulation panel. A few options are for different customized solving solution (Fig. 3.15 left). In
order to create the proper result of wind flow for visualization, the only modification in this wizard is “iterations to
run”. Iteration indicates how many times the model is calculated to gradually approach to the most accurate results.
But as a tradeoff of accuracy, the more iterations the longer time to calculate. Finally, by hitting solve button, the
engine starts the simulation depicting the convergence plot (Fig. 3.24). The primary criteria for determining
convergence is that the change of each degree of freedom is minimized over a large range of iterations. (Autodesk
2017) The curves shown in the Convergence Monitor are plots of the average value of each degree of freedom
throughout the entire calculation domain (Autodesk 2017). This was done for the case study.

Figure 3.24 Solver Quick Edit Dialog (left); Convergence plot (right).
3.3.5 Result simulation in CFD
After the simulation completes successfully, the simulation result data is stored for users to use. In the proposed
methodology to visualize CFD wind flow result, the wind flow data including coordination and velocity value of the
cloud points must be created in Autodesk CFD simulation. At this point, there is no result that are shown in the
viewport (Fig. 3.16).
It is often easier to visualize the airflow with air traces, also referred to as stream lines. First, create the plane that the
user would like to add points for tracing by using “Planes” tool in results tab, Results Tasks panel. Then in Planes
editing panel, use Traces tool to add points. Users are able to add rectangular grid with automatically generated grid
spacing or customized grid (Fig. 3.25). As a result, streamlines that trace the point on the grid from inlet to outlet are
displayed (Fig. 3.26). This was done for the case study.
44

Figure 3.25 Create the plane with points for tracing

Figure 3.26 Streamlines that trace points on the grid.
Besides for visualization within the CFD software, the reason for creating the streamlines is to export its numerical
Excel data file that can be visualized in 3D visualization program Autodesk 3ds Max. The proposed exported data
format is .csv that records the coordinates and parameters like velocity, pressure of each points on the streamlines and
that is readable in Microsoft Excel (Fig. 3.27 left). By exporting Nodal Results in Autodesk CFD Simulation, the data
file was stored in the project folder.
45

Figure 3.27 NODAL file exported from Autodesk CFD Simulation (left); export NODAL file (right).
3.4 Three-dimension visualization
Adapting Autodesk 3ds Max (3ds Max) as one part of the whole workflow was useful because it is the intermediate
between Revit and game engines, but also because it has a significant function of visualizing CFD simulation data in
terms of velocity, temperature, and pressure in various ways (Autodesk 3ds Max 2017). 3ds Max has been used as a
professional creative 3D modeling tool in architectural domain for a long time because of its easy and powerful
modeling engine as well as the stunning rendering effects. What is more, its high interoperability allows connecting
between Revit, Autodesk Stingray, and Unity. It bridges the gap between all three software programs. Until the
updated version 3ds Max 2015, the function of CFD visualization makes the tool more than a production and animation
tool but a scientific study helper (Fig. 3.28).

46

Figure 3.28 Autodesk 3ds Max user’s interface: 1) Quick Access toolbar; 2) Main toolbar;3) The Ribbon; 4) Scene Explorer; 5) Viewport Layouts;
6) Status Bar controls; 7) Viewport label menus: 8) Quad menu; 9) Time Slider; 10) Viewports; 11) Command Panel; 12) Animation controls; 13)
Viewport Navigation; 14) Slate Material Editor; 15) Rendered Frame Window. (Autodesk 3ds Max 2017).
There are two parts in 3ds Max for visualization: the first is to convert Revit geometry into game engine readable
geometry; the second is to visualize the wind flow Nodal data into splines that can be imported in game engines (Fig.
3.29).

Figure 3.29 3ds Max process
3.4.1 Import geometry
Importing Revit geometry file into 3ds Max created the environment that the CFD data was fitted into. After all the
CFD simulation results are only instructive and informative in the simulated space. Stingray cannot read Revit file
directly; 3ds Max is an intermediate to realize the “Revit to game engine” workflow.
In Revit, the user exports the detailed model as FBX file and imports the FBX to 3ds Max scene (Fig. 3.30). Some of
the material may be lost because the material from Revit is different from material for 3ds Max (Liu et al. 2016). The
way to fix this problem is to re-assign material in 3ds Max using material editor (Fig. 3.31). In order to make sure the
geometry imported into 3ds Max can spatially match the CFD simulation data, making sure the center of the all
elements locate in the original point is important (Fig. 3.32). This was done for the case study.

Figure 3.30 Export Revit file as FBX file (left); Import Revit FBX file into 3ds Max (Middle); Geometry in 3ds Max Scene (right).
47

Figure 3.31 3ds Max material editor; Re-assign material is 3ds Max.

Figure 3.32 Centering the geometry to coordinate (0, 0, 0).
3.4.2 Wind flow visualization
Wind flow visualization is the most crucial step that makes the numerical Nodal data from Autodesk CFD simulation
into graphically represented and colored wind flow splines. The operations were as follows:
• In Command panel, Creation tab, click the dropdown list and choose CFD as the creation category (Fig. 3.33
left);
• Click the CFDImportData node and then left click an empty space in the viewport to locate the coordinate
that is supposed to be the center of the CFD data, and change the coordinate of the center to (0, 0, 0) to make
sure it can correspond to the geometries previously imported from Revit (Fig. 3.33 middle);
• After centering, click the CSV File node as an import window popped up, then choose the Nodal results
exported from Autodesk CFD simulation, the default file name is Scenario 1_nodal.csv (Fig. 3.33 right);

Figure 3.33 Import Nodal results in 3ds Max
• After the file is loaded into the scene, nothing is showing yet; in order to observe the data, right click the
XYZ coordinate in the viewport and choose properties and check Vertex Ticks item in the Display Properties
section; As the scene shows, the cloud points scatter perfectly within the space; check the border points to
make sure the geometry and data points overlap seamlessly (Fig. 3.34 left, middle);
• Right click the cloud points and convert the cloud points to Editable Poly for further edit and open Properties
dialog and check the Display as box item to avoid confusion in the view (Fig. 3.34 right).
48

Figure 3.34 Display the cloud points as box
• The next step is to add splines based on the cloud points. In the Command panel, Creation tab, select the
shape icon and CFD list; click CFDSplinePaths node and two items CFDImport Object and Paths Source
are required to create the splines. The CFD Import Object indicates the cloud points while the “Paths Source”
is the geometry that is closest to the inlet that is regarded as the origin of the CFD splines. In order to create
the origin, three planes are create using “Plane” command (Command panel -> Creation tab -> Primitive
Geometry -> Plane) right above the wind flow inlet of the space (Fig. 3.35). After plugging in the CFD data
and path source of the planes, the short splines tracing from the source planes are shown (Fig. 3.35)

Figure 3.35 CFD spline path command (left); Path source (right).
• To adjust the splines to the proper length, there are three parameters that users can change (Fig. 3.36 left):
“Num Steps” (Total number of segments that the splines will have), “Num Samples” (how many samples to
look at to determine the next velocity and “Step Size”. The default value of the three parameters are 10, 10,
49

1.0 respectively. With the default values, the splines are not obvious at all because the values are not large
enough to cover all the needed data points in the space (Fig. 3.36 upper). When adjusting to proper value, the
splines from the inlet to outlet display as expected (Fig. 3.36 lower). The default name of the splines is
CFDSplinePath001, which can be found in the Scene Explorer.

Figure 3.36 Splines in 3ds Max traveled from inlet to outlet adjusting Num Steps, Num Samples and Step Size
• The Nodal results with data points are traced in 3ds Max as the same way in Autodesk CFD simulation in
splines. Then the color is added to the splines using CFD vertices modifier (Fig. 3.37). The Modifiers in 3ds
MaX are modeling and editing tools that create different effects. For example, in order to edit the splines, the
CFD path should be converted to editable splines by right clicking the CFD path and choosing “Convert to
Editable Spline” (Fig. 3.37 left). Then check the Enable In Render and Enable in Viewport items in the
Command panel so that every time the change made to the splines can be viewed in the viewport under
rendering mode.

After checking the items, the splines show their thickness in the viewport, which can be changed in the
Command panel as well (Fig. 3.37 middle). Then right click the splines again to open properties dialog that,
check “Vertex Channel Display” box in Display Properties section. After all the setups, choose the
CFDColorVerticesMod from the list in Modify tab, Command panel, and plug in the CFD data named
CFDImportData001 (not the splines, easier to choose from Scene Explorer). The vertex color shows on the
50

splines (Fig. 3.38 left). The color represents the value of a certain data point on the spline, as the same in
Autodesk CFD simulation result, the darker the color, the lower the value (Fig. 3.38 right). To make the color
more differentiated, the Red Amount value which enlarges the color range can be increased (Fig. 3.39)

Figure 3.37 Convert the CFD path to Editable Splines (left) (Elizardo 2015); Enable in Renderer and Viewport (middle); Add CFD vertex color
modifier (right).

Figure 3.38 Vertex color on the splines (left); Wind flow streamlines in Autodesk CFD Simulation (right).

Figure 3.39 Add more red amount to the vertex color
• Next step after visualizing is to export the CFD splines and model geometries separately in FBX format in
case of any confusion that can be imported into Autodesk Stingray for VR.
• All these steps were done for the case study.
3.4.3 Air flow animation
After the BIM geometry and visualized CFD splines are ready in 3ds Max, they can be imported into Stingray for VR
creation. For the sake of better animation of the direction of the air flow, a special material called CFD animation
material is scripted in 3ds Max (Fig. 3.40). As to control a material, there are basis parameters like opacity, roughness,
51

metallic, emissive etc. The key point for the material to create the air flowing effect is to control the opacity of the
material. Step 1, step 2 and step 3 are three parts that assemble the special opacity parameter (Fig. 3.40). (see Appendix
A).

Figure 3.40 CFD animation material scripting in 3ds Max to create the animated air flow effect
• Step 1: in this step, a rotation gizmo (rotator) is added to place and tilt the color map like an arrow or a
gradient band based on the coordinates divided on the splines. It is a single vector node, which means the
color map can be tilted only in one direction (Fig. 3.41);

Figure 3.41Step 1 Rotator in the CFD Animation Material
• Step 2: a panner is added to create the flowing effect. A panner is the same as to create the sea waving effect
(Fig. 3.42). the tiling node controls the intensity of the color map, the higher the value, the denser the color
map places onto the splines. Changing the speed U and speed V node is to speed up or down the flowing rate
from U and V directions of the splines. The time node is to inform the user and the program that the speed is
modifiable (Fig. 3.43).
52

Figure 3.42 Sea waving effect using panner in the material’s script (Autodesk Stingray 2017)

Figure 3.43 Step 2 in CFD Animation Material
• Step 3: the color map parameter is added and connected to step 1 and step 2 so that the Rotator, Tiling, Speed
U and Speed V can be control (turned on/off) by the opacity parameter (Fig. 3.44).

Figure 3.44 Step 3 in CFD Animation Material
Combing the three steps and connecting opacity, metallic and roughness to the standard base node, the material script
can be controlled in the Stingray user’s interface after the material script imported to Stingray (Fig. 3.45).
53

Figure 3.45 Final plug-in in CFD Animation Material
All of these steps were done to the case study.
3.5 Virtual reality (VR)
For VR game engines, there are several mainstream popular software to choose from. This proposed methodology
adopted Autodesk Stingray (Stingray) because of its high interoperability with previously mentioned software,
especially with 3ds Max. There is a Stingray add-in tab in 3ds Max that users can send models back and forth directly
through this command. At the same time, the add-in allows real-time tracking with the two software, which means
when users operate in Stingray or 3ds Max the other software will be tracked at the same time (Fig. 3.33). In addition,
Stingray and 3ds Max use visual programming language from material editor perspective that is easier for architects
or engineers that are familiar with 3ds Max. Other than material editor, Stingray uses visual programming language
for other functions as well, which is beginner friendly. The user’s interface is pretty streamlined to use (Fig. 3.34).
In this section, all the geometries and CFD splines are brought into Stingray and are edited ready for VR. Then the
complete VR environment is tested in Stingray using HTC VIVE HMD (Fig. 3.46). All of the steps below were done
in the case study.

54

Figure 3.46 Stingray workflow
3.5.1 VR Environment build
The first thing to do after open the Stingray program is to start a new project by choosing “VR HTC VIVE” template
from the template tab. This template provides a ready-to-test VR room that can be interacted using HTC VIVE. In
addition, the scripts in the template are also useful to be learned and adapted. The VR environment is built following
the steps as below:
• Create and save a new level from “File” tab;
• Create new folders that are ready for imported assets
• Import the geometries and CFD splines separately into different folders to clarify;
• Drag and throw the imported geometries and CFD splines into the scene viewport;
• Centering both of the files to (0, 0, 0) to make sure they overlap;
As shown in the scene, the splines are grey instead of having the vertex color on. This is because the material of the
splines is not set up. Since the material have been created in 3ds Max, users can import the material as an asset as well
and assign it to the splines. To assign material, select the splines first, and drag the material to the Properties panel
under material path. Then the vertex color displays. In order to generate the animation of wind flow effect, select the
material and increase the opacity to a certain value for the best result. (Fig. 3.47)

Figure 3.47 Before and after assign the CFD animation material
In VR environment, all the objects are “solid” so that users cannot walk through them, the same in real world. So the
last step before test using HMD is to set collisions for all solid objects like walls, floors, furniture. By double clicking
the geometries, Unit Editor opens for create interactions and other model editing. In the Unit Editor interface, select
all the objects that are supposed to be solid and right click them to choose Add Physical Actors, and save the changes
in Unit Editor. As a result, when testing in VR, users cannot walk through the objects that have been set for checking
for collisions (Fig. 3.48).

Figure 3.48 Create collisions for objects
55

3.5.2 HMD test with HTC VIVE
Before testing in Stingray using HTC VIVE HMD (VIVE), the headset should be set up in Steam VR program that is
the VIVE start engine. VIVE requires a certain space for play. In this proposed methodology the space allowed users
to walk around and observe the geometry and CFD spline (Fig. 3.49).

Figure 3.49 HTC Vive sensors (left and middle) and HMD (right)
If there is a “Ready” message showing up (Fig. 3.50), VIVE is ready to use. In Stingray, click the Play icon on the
right. The VR viewpoint window will appear (Fig. 3.51). Users can put on VIVE to walk and look around the
environment (Fig. 3.x) At the same time, some problems may be found during the test, e.g. the collision is missing for
a certain object. Adjustments are needed after the tests until the project achieves the requirements.

Figure 350. VR Ready reminder

Figure 351. Screenshots of VR environment that the user is in wearing Vive HMD
3.6 Conclusion
The proposed methodology detailed the entire workflow from the construction of BIM geometry to CFD simulation
to visualization and finally to VR. Although four different software of different fields were applied in this workflow,
users do not need to have a deep command of each software. Software details should be paid attention to since it is
vital to generate the desired visual effects.
To completely test the proposed methodology, a case study was created to study the feasibility of the methodology
56

4. Case study
To test and demonstrate the feasibility and effectiveness of the proposed CFD VR visualization method, a kitchen
space is created as a case study. A kitchen space needs air circulation because of the daily routines of cooking and
dining may affect the environment of the kitchen. In addition, in the kitchen there are often large assemblies like
cabinets, partitions that may influence the air flow. And the size of a kitchen space is proper for a VR case study to
walk around.
4.1 Case study overview
The kitchen space was set locating in Los Angeles of a single-family house. The dimension was 42’ x 16’, 672 square
foot in total. The kitchen space was divided into three separate areas with two partitions in order to create some
variations of the wind flow (Fig. 4.1)

Figure 4.1 Floor plan of the kitchen space
4.2 BIM geometry
In Autodesk Revit 2016, a new project with architectural template was built to start the project. Starting from the
architectural elements, wall, floor, roof, window, door and furniture were placed in the Level 1 floor plan (Fig. 4.2)

Figure 4.2 3D perspective of the kitchen’s floor plan
57

Walls and floor were set as layered structures by creating new types of wall with several layer. Structural elements
were not built since the model is not a fully developed large-scale project. MEP elements are included as inlet ducts
and outlet ducts (Fig. 4.3)

Figure 4.3 inlet (left) and outlet (right) ducts
This is the detailed model for later geometry simplification and VR display, saved as kitchen_detailed.rvt.
After the detailed model was built, the model was tailored according to the principles in Chapter 3.2.2. Detailed
furniture were replaced by box-like mass that have the same dimensions. The wind direction of the kitchen is from
inlet ducts to outlet ducts (from roof to floor). As the model is prepared for wind flow simulation, the glass of window
will not influence the flow so that all the windows and skylights were eliminated from the model (Fig. 4.4). The
simplified model was saved as kitchen_cfd.rvt.

Figure 4.4 Simplified Revit model ready for CFD simulation

4.3 CFD simulation in Autodesk CFD
In the simplified kitchen_cfd.rvt file, click the Launch Active Model icon to start Autodesk CFD 2016 for simulation.
In the Design Study Manager dialog, the file was saved to the proper folder with a proper project name (Fig. 4.5).

Figure 4.5 Design Study Manager dialog to name and save the project
58

4.2.1 Set material
According to the proposed methodology, each component was assigned to a different material. In this project, the
indoor air volume was assigned Fluid type with material name of Air; the floor was assigned as Solid, Timber Plank.
Objects other than these two were all assigned as Solid, Gypsum Board (Fig. 4.6).

Figure 4.6 After set the material the Design Study Bar shows all the assigned material (left); the kitchen after assigning material (indoor air is
hided) (right).
4.2.2 Set boundary
After setting up the material, setting the boundary conditions were the second task. As the kitchen is design to let the
air circulate from the roof inlet ducts to the outlet ducts on the floor, the inlet surfaces were assigned as Velocity type,
with a speed of 85 in/s. The outlet surfaces on the floor are assigned to Pressure type, with 0 psi. Different color of
the bands on the surfaces indicate different type of boundary conditions. In this case, the black band represents for
Velocity type of condition and the orange band is for Pressure type (Fig. 4.7 and 4.8).

Figure 4.7 kitchen wind flow inlet ducts and boundary conditions

Figure 4.8 Kitchen wind flow outlet ducts and boundary conditions
4.2.3 Mesh sizing
In order to generate the nodes for calculation, the model was sized. In this case study, Autosize function was applied
for convenience and accuracy (Fig. 4.9). After sizing, the Visual Style to Outline was changed to view the result (Fig.
4.10).
59

Figure 4.9 Change the visual style to observe the autosizing result

Figure 4.10 Autosize the model
4.2.4 Solver
After all the setup tasks are finished, the model was ready for solving. In the Solve Quick Edit Dialog, set the Iterations
to Run of 20, as a balance of the computing speed and accuracy for a relatively small space. The convergence plot
showed the progress of calculation (Fig. 4.11).

Figure 4.11 Solve Quick Edit Dialog setting (left); Convergence plot during calculation (right)
4.2.5 Result viewer
In order to export the nodal data file for VR visualization, the air flow streamlines were created in CFD first. First, a
plane was added in the middle of the space so that it covers the main points within the space. Then the Trace button
was used to add a rectangular grid onto the plane (Fig. 4.12). In the Trace Quick Edit Dialog, the Add trace
set tool generate the splines that traced each point from inlet to outlet (Fig. 4.13).
60

Figure 4.12 Add points for tracing on the plane

Figure 4.13 The wind flow tracing splines are added
4.2.6 Export Nodal Results data
For later visualization in 3ds Max, Nodal Results were exported via Export -> Nodal Results process in Autodesk
CFD (Fig. 4.14).

Figure 4.14 Export Nodal Results
In the project folder, the file could be found in the “Solver” subfolder. Because the nodal results are in CSV format,
it can be open directly in an Excel spreadsheet. It recorded the nodes on the plane that is previously created, with their
coordinates, velocity, pressure, temperature etc. (Fig. 4.15) The velocity was expressed in X, Y, Z vectors.
61

Figure 4.15 Nodal Results from Autodesk CFD 2016
4.3 Visualization in Autodesk 3ds Max
With the geometry from Revit and Nodal Results from CFD, the 3D visualization can be realized in 3ds Max. In 3ds
Max, new project was started and saved.
4.3.1 Import CFD data and Revit model
In Revit, kitchen_detailed.rvt file was opened and exported the detailed kitchen model as FBX file that can be read by
3ds Max. Then the model shows in the scene. Next, the method in Chapter 3 was followed to import the CFD data as
cloud points (Fig. 4.16). The geometry and cloud points were overlapped seamlessly by centering to the origin (0, 0,
0).

62

Figure 4.16 Import BIM geometry from Revit and CFD Nodal Results from Autodesk CFD
4.3.2 Create CFD streamlines
The cloud points were converted to editable poly and displayed as box, and the kitchen geometry was hidden to better
observe the splines. Using the tool of create CFD splines proposed in Chapter 3, the splines were generated with planes
as the source and CFD data as the tracing points. The Number of Segments was 10000, Number of Example was 30
and Number of Step was 4 (Fig. 4.17). By hiding the building geometry, the splines were created at the correct place
in Autodesk CFD (Fig. 4.18).

63

Figure 4.17 Create the splines

Figure 4.18 Splines displayed with the kitchen geometry
4.3.3 Add CFD vertex color modifier
The splines were checked with properties of Enable In Renderer and Enable In Viewport and Generate Mapping
Coords. The splines’ thicknesses were set to 0.2’ for better visualization (Fig. 4.19 left). In the Object Properties
window, the Vertex Channel Display was checked as well (Fig. 4.19 middle). After the properties, the vertex color
modifier was added on top of the editable splines to generate the vertex color that represent the level of velocity (Fig.
4.19 right). In order to boost the color, the Red Amount was set from default 100 to 500 to enlarge the overall color
range of displaying the data (Fig. 4.20). The final result was shown properly with the kitchen space geometry (Fig.
4.21).

Figure 4.19 Properties to check (left); Object Properties (middle); Add vertex color modifier with Red Amount of 500 (right)

64

Figure 4.20 Red Amount of 100 vs 500

Figure 4.21 Overall result with kitchen geometry
In order to avoid confusion when importing objects to Stingray, the splines and all the other objects were selected
separately and then exported as FBX files respectively. There were two FBX files, one of them is the splines and the
other is all of the kitchen geometry.
4.4.4 Create VR
Opening Autodesk Stingray, VR HTC VIVE was chosen as the template for this case study. After the new project
starts, there was a default vr_learning level with a room. A new level called cfd_level was created (Fig. 4.22).

Figure 4.22 vr_learning level (left) and empty cfd_level (right)
Working in the cfd_level, two folders of kitchen and splines were created in the Asset Brower to store different material
and model (Fig. 4.23). By clicking the Import button on the up left, the kitchen model and splines model were imported
to each folder along with all their material. The kitchen geometry and splines were added to the scene and centered,
the grey version of the visualized objects is displayed (Fig. 4.24)
65

Figure 4.23 Asset Brower

Figure 4.24 Grey objects without material
In order to generate the colorful wind flow animation for the splines, a material script was created in 3ds Max as
mentioned in Chapter 3(Fig. 4.26). With this script imported into Stingray, a Stingray material was created with
parameters (Fig. 4.27). The splines were selected and assigned the material named CFD animation material (Fig.
4.26). The material was assigned a color map of an arrow, which created the animation of the wind flow from inlet
to outlet (Fig. 4.25 & 4.28).

Figure 4.25 CFD animation material
66

Figure 4.26 CFD animation material script in 3ds Max for wind flow animation

Figure 4.27 Material with parameter in Stingray

Figure 4.28 Wind flow animation with arrow (left); the arrow colormap (right)
67

Then add to the color map the black and white gradient image, and the wind flow animation becomes thicker and is
more like the actual wind flowing (Fig. 4.29). Even though both the arrow image and the black and white gradient
image are grayscale, the airflows are still coated with the visualized “CFD color”. The reason is when scripting the
CFD animation material, vertex color was used to represent the CFD velocity level, in which case no matter what
image (colormap) was used only the pattern of the image (colormap) was read. If there is no colormap used, the airflow
will be still instead of the flowing animated effect (Fig. 4.30).

Figure 4.29 Airflow animation with gradient color (left); the gradient color image (right)

Figure 4.30 Airflow without animation in Stingray
The animation effects brought different user experience when in arrow mode and gradient color mode (Figure
4.31).
68

Figure 4.31 CFD air flow animation with arrow mode and gradient color mode in Stingray
In the VR environment, the colored splines represented the velocity of each points on the splines, the visualization
was the same as it was in 3ds Max with the Red Amount 500. Although the coloring in 3ds Max was different from
that in CFD, the values that were visualized were the same. Only the color range was enlarged which meant more
colors were used to represent the same amount of data points with the same velocity values (Fig. 4.32). If the Red
Amount in 3ds Max was decreased to 50, the visualized results were highly corresponded with the results in CFD (Fig.
4.33).
69

Figure 4.32 Visualization in CFD (upper); visualization in 3ds Max (middle) and in Stingray (lower) with the Red Amount of 500
70

Figure 4.33 Visualization in CFD (upper); visualization in 3ds Max (middle) and in Stingray (lower) with the Red Amount of 50
4.3.7 HTC VIVE test
To drive the VR equipment, the requirements for computer were higher than the daily use computers especially with
regard to the graphics card. In this case, a desktop with ASUS STRIX GTX1080 8G Advanced Edition Gaming
Graphics Card and Intel Core i7-7700K 4.2 GHz Quad-Core Processor powered (with speed and memory) the VR
equipment and create the project.
After the result shows up in Stingray, the VIVE equipment was set up including sensors, joysticks and HMD, and then
the Play button was hit to start testing. Users were able to walk around with the joystick to point the position they
would like to go and experience the CFD wind flow with animation in the kitchen space (Fig. 4.34).

Figure 4.34 Test using HTC VIVE

71

4.4 Conclusion
The workflow enabled a user to go from Revit to Autodesk CFD to 3ds Max to Stingray to view the airflow in a VR
environment for a simple kitchen example. Each of the steps usually worked although unexpected errors could
occur. For example, when adding the splines using CFD splines tool, if user did not locate the spline object before
plug in the CFD data the splines might not be able to generate. The overall workflow was functional and effective in
terms of the goal to visualize the CFD simulation results and bring the visualization to VR environment, so that the
final wind flow result can be experienced using VR HMD by the user. Chapter 5 gives a more detailed evaluation of
the workflow.

72

5. Evaluation and analysis
Three workflows were described in Chapter 3 with the end results being a demonstration of how weather data, solar
insolation, and CFD simulations could be visualized in VR. The first two led to the development of the third and the
case study based on it for trying out the methodology to discover its feasibility and effectiveness and to learn about
possible deficiencies and disadvantages. The evaluation and analysis of the workflows is discussed in this chapter in
two sections: the first part is to evaluate the test workflows and the proposed completed workflow and its potentials
and problems; the second part is the evaluation and comparison of the visualization results among the three software
programs that can generate the CFD wind flow splines based on the criteria of accuracy, graphics, and experience (Fig.
5.1). The ability of the visualization for CFD analysis is also going to be discussed.

Figure. 5.1 Evaluation and comparison of the methodology and case study
5.1 Evaluation of the workflows
Three workflows were created (see Chapter 3), two of which were the test workflows that helped to develop the
completed CFD visualization in VR workflow. Eight parameters were chosen to evaluate the workflows, several of
which had been used to evaluate VR workflows previously (Li, 2017):
• Cost: the cost of the workflow includes the software used and the VR equipment;
• Interactivity: the interactivity was classified into four categories: interactive stereo image (observing),
interactive stereo panorama (looking around), motion tracking walkthrough (looking around and walk) and
scene-modifiable motion tracking walkthrough (reacting and receiving feedback from the environment) (Li,
2017);
• Compatibility: the collaboration of different software, such as whether one software has an add-in designed
for the other software in order to make it more convenient for users;
• Interoperability: the data exchange among different software;
• Visualization: the visualized results generated via the workflow in VR;
• Accuracy: whether the visualized results correspond with the original data source;
• User-friendly: whether or not the workflow is friendly for users to setup in each step to generate the
visualization results;
• Process speed: the calculation time that the computer needs to generate the visualization results.
Among the criteria, user-friendly and process speed are indications of the complexity for users to setup the
visualization and the average time consumed by the computer to generate the results.
Based on the discussion of the criteria, an overall evaluation of the workflow is given in form of a radar chart (Fig.
5.2) (based on Attia et al. 2011). Each criterion is evaluated with “poor,” “good,” and “excellent” levels.
73

Figure 5.2 Results of the NZEB tools mechanics (Attia et al. 2011).

5.1.1 Workflow 1 evaluation: EPW to VR
The first test workflow was trying to directly visualizing EPW data in a game engine by writing scripts to aid with
interoperability (Fig. 5.3).

Figure 5.3 Test workflow 1
Cost: the EPW file was downloaded from EnergyPlus website for free, and the conversion of EPW format to CSV
format was realized in EnergyPlus add-in for free as well. In the VR part, Unity 3D was used for free (the commercial
version of Unity 3D is more powerful, but the function used in the workflow can be reached in the free version.). The
VR equipment of HTC VIVE cost $800. The desktop that can drive HTC VIVE cost about $2000 (Fig. 5.4 & Fig.
5.5).
74

Figure 5.4 Key components used in the desktop: ASUS Z170-A Motherboard, Cooler Master Hyper 612 Ver.2 cooler, WD Blue 2TB Desktop
Hard Disk Drive, Corsair LPX 32GB DRAM (Set of 2), Intel 7
th
Gen Intel Core Desktop Processor i7-7700K, Samsung 850 EVO 250GB Internal
SSD, EVGA Supernova G1 650W Power Supply, ASUS GeForce GTX 1080 8GB ROG STRIX Graphics Card (left to right, upper to lower)

Figure 5.5 HTC VIVE VR set: two sensors (upper), HMD and two controllers (lower)
Interactivity: the workflow from EPW file to VR achieved the interactive stereo panorama level, which meant that
users can look around in the VR environment of the visualized box with constantly changing scale indicating the
temperature of different date and the default Unity 3D background. However, since the VR environment was
developed in Unity 3D, which is the tool having great potentials to create better interactivity by additional user
programming, motion tracking walkthrough (looking around and walk) and scene-modifiable motion tracking
walkthrough (reacting and receiving feedback from the environment) could have been achieved.
Compatibility: EnergyPlus as an energy simulation software had no direct add-in h the game engine Unity 3D. The
bridge between the two was a CSV file that can be read by Unity 3D via a script.
75

Interoperability: from the CSV file to Unity, how the data is managed and processed depends on users’ needs. In
workflow 1, the “temperature” column was read through script (temperature records of one month), which meant the
other data in the EPW file were left off intentionally. During the data reading process using the script, all the data
points were read one by one. If programmed properly, the interoperability level could be high when all the data from
the EPW file are interpreted.
Visualization: the test visualization was a box in the space with changing scales representing the temperature data.
The effect was not ideal for several reasons. The speed of scaling change was either too fast or too slow for a large
amount of data points. If the speed was very high to present all the data points within a reasonable period of time, the
change rate was too frequent for people to grab any information from it. If the speed was slowed down, the time was
too long to finish reading all the data points, even for a monthly record. in addition, the information that the users
learned from the visualization was not much. The position of the box in the space did not convey useful and intuition
information as well (Fig. 5.6).

Figure 5.6 Box scaling based on temperature value in Los Angeles
Accuracy: observing the data read by the script, all the data points imported was represented. However, by looking at
the visualized object in the environment, it was hard to tell whether or not the weather data was visualized accurately.
There were user-controlled factors such as the size of the box, the enlarging or shrinking factor multiplied to the
temperature values etc. In this specific case, the visual accuracy was poor even though the data in the VR environment
was correct. But the potential exists that with scripting, a better level of accuracy could be obtained.
User-friendly: the conversion from EPW to CSV in EnergyPlus was easy to operate by just clicking the appropriate
buttons. In Unity 3D, for users as programmers the workflow was not hard to realize; for users like architects and
engineers without aprogramming background, it took time to create and compile even the easy test script.
Process speed: the process speed was acceptable to generated the visualization effect with one box of no more than
1000 values changing constantly (within 2 minutes). A longer time is expected as the data structure and graphic design
become more complexed.
Overall, the ability of “EPW to VR” workflow is given based the discussions. Generally, the “EPW to VR” workflow
scored poorly as actually accomplished (Fig. 5.7 – left). However, the potential of this workflow should not be
overlooked and a second set of results is shown with Interactivity, Interoperability, Visualization, and Accuracy
evaluated based their perceived potential (Fig. 5.7 - right).

Figure 5.7 Evaluation of workflow 1 – left side (as done), right side (potential with additional scripting)
76

Even though the evaluation of the workflow was above average, the workflow was not attempted because of the User-
friendly reason, i.e. the programming requirements were really high in order to create the VR environment with the
high interactivity, interoperability, visualization, and accuracy.
5.1.2 Workflow 2 evaluation: BIM solar simulation to VR
Workflow 2 was tested to visualize the solar insolation simulation results using BIM tools and to realize VR in Unity
3D (Fig. 5.8).

Figure 5.8 Test workflow 2
Cost: Autodesk Insight 360 and Dynamo are the add-ins for Autodesk Revit. The cost of Revit is $2000/year. A free
student version of Revit is available. For the VR part, the free version of Unity 3D is qualified for the workflow. The
VR equipment of HTC VIVE is $800. The desktop that can drive HTC VIVE cost about $2000. Both the desktop and
VR equipment were used the same as workflow 1.
Interactivity: the visualized meshes indicating the solar insolation value on a building façade, as well as the building
geometry were imported into Unity 3D. Users were able to look around wearing HTC VIVE HMD. However, similar
to workflow 1, since the VR environment was developed in Unity 3D, which has a great potential to create better
interactivity by programming, the interactivity score could have been much higher.
Compatibility: Insight 360 and Dynamo as Revit add-ins that are strongly compatible with Revit. The Revit model
can be used for simulation directly through Insight 360. Dynamo, as the visual programming design tool for Revit,
can read the Revit model seamlessly. In workflow 2, Unity 3D was used only to create the VR environment by
importing objects from Revit and Dynamo in order to be observed using HTC VIVE. No extra programming was
needed in Unity 3D.
Interoperability: when using Revit model for solar insolation, the components such as walls and roofs can be
recognized by Insight 360, which indicated the interoperability between Revit and Dynamo was good. In the
visualization stage, the simulation data (CSV) was exported from Insight 360 and imported into Dynamo using CSV
reading tools. In Dynamo, all the data including the calculated points’ coordinates and solar insolation values were
read. However, in the VR stage when importing geometry from Dynamo to Unity 3D via Revit, the visualized color
scheme was lost. Only the geometries were imported.
Visualization: since the visualized geometries that represented the insolation value were created as adaptive
components in Revit, the potential of the visualization was high. Even though in the experiment the visualized colors
were lost, there may be other way to visualize the insolation instead of colors. This was not tried.
Accuracy: the accuracy of the workflow was not high enough since the colors were lost, and users may be really
confusing what are the grayscale meshes. The potential of the accuracy would be better using better scripting
User-friendly: with visual programming in Dynamo, the visualization become easier than directly writing code.
Process speed: even with a relatively good configuration the calculation speed of the computer was very low.
77

The ability of workflow 2 was also measured with radar chart (Fig. 5.9). The reason that workflow 2 was not fully
developed was the process speed was really low that the script created in Dynamo cannot process large-scale data sets.

Figure 5.9 evaluation of workflow 2
5.1.3 Workflow 3 evaluation: BIM to CFD to VR
Workflow 3 has four stages: BIM geometry, CFD simulation, 3D visualization and virtual reality visualization, using
four different software programs (Fig. 5.10).

Figure 5.10 Workflow 3
Cost: Achieving the proposed method requires at least a VR headset with a VR compatible computer. In the case study,
HTC VIVE was cost $800 and a desktop cost around $2,000 (Fig. 5.11 & Fig. 5.12). The software cost is high, with
Revit $2000, Autodesk CFD $3780/year, 3ds Max $1470/year and Stingray $240/year. 3-year free licenses for student
are available for Revit, CFD and 3ds Max.
78

Figure 5.11 Key components used in the desktop: ASUS Z170-A Motherboard, Cooler Master Hyper 612 Ver.2 cooler, WD Blue 2TB Desktop
Hard Disk Drive, Corsair LPX 32GB DRAM (Set of 2), Intel 7
th
Gen Intel Core Desktop Processor i7-7700K, Samsung 850 EVO 250GB Internal
SSD, EVGA Supernova G1 650W Power Supply, ASUS GeForce GTX 1080 8GB ROG STRIX Graphics Card (left to right, upper to lower)

Figure 5.12 HTC VIVE VR set: two sensors (upper), HMD and two controllers (lower); HTC VIVE room setup demonstration
(http://notes.caseorganic.com/2016/11/02/room-setup-for-htc-vive/, retrieved in 2017)
Interactivity: The high quality VR equipment HTC VIVE provides excellent interactivity. Users can observe real-time
rendered image while looking and walking around in the VR environment within the sensor’s tracking area and at the
same time using the joysticks to track different point as an alternative way of “walking” that enlarge the space users
can go through. There were feedbacks from the environment.
Compatibility: the compatibility is good between Revit and CFD because of the Autodesk CFD add-in in Revit. But
a problem existed that the material in Revit cannot be recognized by CFD. All materials need to be re-assigned in
CFD. Besides the CFD add-in in Revit, as 3ds Max is designed to be the modeling tool for Stingray, the connection
between the two is smooth. There is a Stingray tab in 3ds Max that can directly send selected 3ds Max objects to
Stingray. In addition, the Live Camera Tracking tool allows the user to drag and zoom in/out the model and add things
to the model in two windows at the same time (Fig. 5.13). Users are able to observe the changes between the two. The
compatibility and integration of the two software is greatly increased in this way.
79

Figure 5.13 Live Camera Tracking between 3ds Max and Stingray

Interoperability: The data exchange between each stage requires no more than twice conversion, which meant that
there was only one intermediate needed for the program to read the file i.e. from Revit to 3ds Max. The overall
workflow requires high interoperability of each software program. The combination of Revit, CFD, 3ds Max, and
Stingray from Autodesk satisfied that requirement. In return, the data exchange during the workflow almost does not
require extra intermediate software. The CFD results were exported as nodal CSV file, and they were all read in 3ds
Max and represented as cloud points. The geometries were exported as FBX file from Revit. Inside 3ds Max, the FBX
file was imported. However, at the same time, most of the materials were missing due to the different material systems
the data exchange from 3ds Max to Stingray was almost a completed exchange with no data lost in this case study in
terms of materials, geometries. Difficulties exist among data exchange from Revit to 3ds Max to Stingray, due to the
material system’s difference. Most of the materials are missing after importing to 3ds Max, the same kind of material
need to be created again in 3ds Max (Fig. 5.14 and 5.15).

Fig 5.14 Original BIM model in Revit (left); the original model imported to 3ds Max (right)
80

Fig 5.15 Re-assign materials in 3ds Max using material editor
Visualization: the proposed method generates CFD wind flow streamlines in three software programs with different
visual effect (Fig. 5.16) (section 5.2 will discuss the visualization from accuracy, graphics and experience aspects).
As for the final effects in virtual reality environment, the rendering quality of the environment is medium high,
bringing a reasonable VR experience to users. Comparing with previous studies, the breakthrough from the
visualization perspective is create animated color streamlines in the VR environment of the CFD data.

Figure 5.16 Comparison between CFD and VR results
Accuracy: all the visualized data are from the simulation results in CFD. The accuracy of whether the VR splines
represent the actual wind flow scenario depends on the simulation settings in CFD. The more detailed the geometry
and settings, the more accurate the results are. However, in this proposed method and case study, more attention is
paid on the visualization of the data. In terms of the accuracy between CFD splines and VR splines, the position, color,
and direction are generally graphically accurate from human’s observation (Fig. 5.17). But the problem was the lack
of color legend in 3ds Max and Stingray stages affects users’ understanding of the visualization graphics.
81

Figure 5.17 Visualization in CFD (upper); visualization in 3ds Max (middle) and in Stingray (lower) with the Red Amount of 50
User-friendly: in order to set up for the visualization in 3ds Max, there was no need for programming of scripts, which
was easy for users to operate. However, the creation of the CFD animation material needed visual programming skill
to finish, which was still relatively for those without a programming background.
Process speed: because of the CFD visualization tool is in 3ds Max, the post-process of the large-scale CFD data was
acceptable that would not crush the computer down. The recommendation of the computer configuration is the higher
the better, whereas as long as the computers can carry out CFD simulation that would be qualified.
In all, the strengths of the proposed methodology are its user-friendly, high compatibility, high interactivity, and
process speed and the visualization potential (Fig. 5.18).

Figure 5.18 evaluation of workflow 3
82

5.2 Comparison of visualization results
Air flow results in was shown in places: the CFD simulation, 3ds Max, and Stingray. The comparison and evaluation
of these results are worth discussing. The visualization results are discussed in terms of accuracy, graphics and
experience aspects.
• Accuracy: accuracy refers to whether or not the graphics accurately show the data received from the CFD
simulation;
• Graphics: the graphics including the rendering of the visualization and the possible visualization methods
(different symbols);
• Experience: the experience indicates whether the visualization allows users to communicate with the
visualization, such as data analysis or to interact with the visualization.

5.2.1 CFD results
In Autodesk CFD simulation, the results are visualized using the built-in visualization tool. As for a professional
simulation program, CFD simulation allows the scientific visualization to be accurate and informative. In the case
study, the visualized air flows are represented in color indicating velocity, with the color legend on the side to explain
what the colors represent (Fig. 5.19).

Figure 5.19 Air flow visualization with color legend in Autodesk CFD
Besides visualizing the results in streamlines, there are also visualization of planes, iso-surface, and vectors (Fig. 5.20).
Different representations contain different information. For example, the iso-volume representation, the volume in the
graph indicated the portion of the air volume within the certain range of values. In that case, the user can understand
the volume with comfortable air flow in the space using iso-volume visualization (Fig. 5.21).
83

84

Figure 5.20 Representations of the planes (left), iso-surface (middle), and vectors (right)
85

Figure 5.21 Representation of changed iso-volume
The animation of the result can be convenient. As for the animation of air flow directions, it can be saved right after
the particle tracing lines are built (Fig. 5.22).
86

Figure 5.22 Saving the animated particle tracing lines
The disadvantage with the built-in CFD visualization tool was obvious. From the perspective of graphics, the rendering
of the color was limited, the false-color display of the data was scientifically perfect but not ideally graphically decent
in Autodesk CFD. Although the simplified simulated model can be combined with the visualized result, the simplified
model with boxes may be confusing for users that are not good at CFD simulation. This also brought up the necessity
for visualizing the CFD results in other software for the sake of better representation.
Overall, the visualization in CFD was accurate according to the simulation results. With the help of the color legend,
users are able to understand what the visualization represents. in addition, users are able to use the data to generate
different visualizations, such as visualizing the data points with velocity lower than 50 m/s and tracing wherever the
point in the space from inlet to outlet. From the graphics perspective, the visualized data can be represented with
different symbols. But the rendering of the graphics is limited within false colors. As for the experience, users are able
to play with the visualization results through the computer, in order to better analysis the airflow conditions within the
space (Fig. 5.23). As for scientific analysis, the Autodesk CFD visualization tool is an excellent choice.

Figure 5.23 The rate of Autodesk CFD visualization results.
5.2.2 3ds Max results
3ds Max is used to visualize the CFD results and to convert the Revit model file into an VR accepted data format.
As a scientific visualization tool, the built-in CFD modifier is user-friendly with nodes that the user can simply plug
in the materials (CFD nodal data results and spline source). But users should pay highly attention to the details, such
as the properties of the splines and the calibration of the data, model, and the splines source. If the details are ignored,
there would be no visualized results or the results may be not accurate compared with the original visualized result
from CFD simulation. The main deficiency was what the colors exactly represented was ambiguous. Especially when
the color range was extended, more colors were added to represent the same amount of data of with the same values
87

of velocity (Fig 5.24 and 5.25). Even though the colors from blue to red appeared in the visualization in Autodesk
CFD, most of the colors were blue to green which meant that most of the velocities were in the range of 0 – 80
approximately. Adding the Red Amount in 3ds Max indicates using more colors like yellow and red to represent the
same range of 0- 80 (Fig. 5.28). However, there were no color legend coming automatically after the color scale was
changed. As for the number of splines, in CFD simulation, the splines can be removed based on the number of all
splines. In 3ds Max, however, the number (or the density) of the splines are generated according to the source plane.
The more vertices the plane has, the more splines extend from the plane. When there were four vertexes on each source
plane, 12 splines in total were extended from the planes (Fig. 5.26); When there were 25 vertexes on each planes,
there were 75 splines in total that were extended from the planes (Fig. 5.27). Creating visualized CFD splines by this
way brought convenience to users for setup, but the questions remained, such as whether the plane source needed to
be places at exactly the same place of inlet and outlet as was in Autodesk CFD simulation, how to determine how
many vertexes needed for the visualization to increase the accuracy etc. The Red Amount generally indicated the value
of “R” channel in the RGB color coordinate. Increasing the Red Amount referred to increase the value of “R” channel,
thus the color of the splines tended to be “redder” simultaneously. The scale after the Red Amount was added was
assumed instead of automatically updated by the software, which might cause inaccuracy. Future work should lay
more attentions on the scale of the visualization.

Figure 5.24 Comparison between 3ds Max results and Autodesk CFD results before adding Red Amount (Red Amount: 100)

Figure 5.25 Comparison between 3ds Max results and CFD results after adding Red Amount (Red Amount:400)

Figure 5.26 4 vertices on each plane
88

Figure 5.27 25 vertices on each plane

Figure 5.28 More colors are added to represent the same amount of values
Overall, from the accuracy perspective, the lack of color legend and uncertainty of the source planes and the way 3ds
Max CFD simulation tool operated and processed the data resulted in the deficiency in scientific analysis of the CFD
data. But by comparing the 3ds Max results with the CFD visualization results, the tendency and direction were the
same (when the Red Amount was 50). The accuracy would be OK in this case. From the graphics perspective, as 3ds
Max is the professional graphics modeling tool, the rendering of the color was more decent and subtle with lower
saturation. But the representing symbols were not able to changed using proposed methodology. From experience
perspective, users cannot tailor the visualization by velocity in 3ds Max and cannot interact with the environment
either. In 3ds Max, the visualization is more for digital representation (Fig. 5.29).

Figure 5.29 The rate of Autodesk 3ds Max visualization results.
89

5.2.3 VR results
All the objects were brought into Stingray from 3ds Max to create the VR environment. From the desktop screen, the
results between 3ds max and Stingray only have slight graphic-display (rendering) difference. After the CFD splines
was assigned the CFD animation material (Fig. 5.30), the flowing of wind was more obvious with the animation, but
the colors were changed a bit because the animation material was a transparency gradient. In the case study, an arrow
image and a gradient color band were used for the animation, indicating the wind direction from inlets to outlets (Fig.
5.31).

Figure 5.30 CFD animation material scripting in 3ds Max to create the animated air flow effect

Figure 5.31 Using the image of arrow (left) and the image of gradient color band (right) for animation
90

In order to understand the VR results, the “immersion” and “presence” are two categories to consider (Gutierrez et al.
2008).
• Immersion: immersion indicates the physical configuration of the user interface of the VR application, such
as fully immersive (VR with HMD), semi-immersive (CAVE), and non-immersive (desktop-based VR)
(Gutierrez et al. 2008). As the user is surrounded by the digital environment totally using HTC Vive HMD,
it is considered full-immersion virtual reality, especially the user can not only view the environment but can
also interact with the environment by walking around.
• Presence: presence is associated with the psychology of the user that is more subjective. The VR environment
with CFD air flow was created to have the user fully present in the environment (Gutierrez et al. 2008). For
the airflow with arrow image, the indication of the air flow direction was more informative. Graphically the
air flow was subtle, the user may pay more attention to the surroundings. For the airflow with gradient color
band, the user may feel strongly involved in the VR environment with scientific visualization results. The
user may instinctively avoid the airflow as it was “hitting” the user. Besides, users tended to walk to the area
that they would like to see more details, which was the strength of VR that can provide real-world details.
The downsides of the VR environment with CFD simulation result were apparent. First of all, due to the lack of the
color legend, users are confused at first what were the colorful splines represent. This indicates the representation of
CFD splines are not informative enough for users to clarify the function of the visualization and are losing the scientific
meaning of the visualization.
In conclusion, the accuracy of the visualization in VR was good by comparing the images from Autodesk CFD and
Stingray using naked eyes. From the graphics perspective, although the rendering and animation effects were great in
Stingray, the limitation of representing symbols was obvious. Better symbols can be designed in future works. As for
experience, in the VR environment, users were about to walk around the space to examine every details with the CFD
airflow so that possible problems of the airflow in the space may be discovered. For example, if the airflows represent
hazard gas in the room, the dangerous zone can be defined by walking around the space. It was also fun to interact
with the environment (walking around and getting feedback); this feeling should not be overlooked, and it contributed
to the experience of CFD visualization in VR (Fig. 5.32)

Figure 5.32 The rate of Autodesk Stingray visualization results.
5.3 Summary
Overall, the visualizations in different software have different emphases. As for the visualization in CFD, it contributes
more to the scientific application and the analysis of the data and as shown in the existing workflows, has graphic
deficiency. The visualization in 3ds Max is convenient and graphically decent, however it is not informative enough
to represent the meaning of the data without color legend. In Stingray, the VR environment with CFD data brings a
fully-immersive experience with relatively high presence. The visualization in the CFD and 3ds Max stages can be
adopted separately depending on users’ needs such as data analysis or representation, but in order to realize the CFD
visualization in VR environment all these three stages are necessary.

91

6. Conclusion and future work
Starting with a building information model, architects and engineers are able to visualize CFD simulation results in a
virtual reality VR environment. This chapter reviews the proposed methodology and evaluation of workflows and
proposes future work with an emphasis on how VR can be used with BIM in other applications.
6.1 Summary of proposed methodology
The tests of the two workflows demonstrated possible methods that can bring weather or simulation data to VR
environment. However, these two workflows were completely finished because of the necessity for more
programming expertise, especially in Unity 3D. The two tests indicated that in order to create a better VR environment
that is useful for representation and scientific analysis, it would be better to visualize large-scale data sets generated
from a 3D space. High compatibility and interoperability of the applied software in the workflow were necessary.
Based on this work, a methodology was proposed to visualize CFD simulation data in VR environment using BIM as
the starting point and Stingray for VR viewing (Fig. 6.1).

Figure 6.1 final workflow

6.1.1 BIM geometry stage
In this stage, BIM geometry, as the basis of all the further process, was created. On one hand, a detailed 3D model
was needed to display the project; one the other hand, a simplified model for simulation use is required as well. From
the detailed to simplified models, both of their level of development and level of details were reduced to some extent.
For example, in the proposed methodology, the detailed model contained furniture with delicately designed details
and walls with parameters of structured layers, the model for CFD simulation only consisted of furniture with basic
dimensions without details (much fewer faces), and the parameters were eliminated to lessen the data exchange process
and calculation burden (Fig. 6.2).
92

Figure 6.2 In BIM geometry stage
6.1.2 CFD simulation stage
In Autodesk CFD, simplified geometry of a kitchen space from Revit was adopted for CFD simulation. The simulation
needed setup for material, boundary, meshes and solver. After the setups, the simulation results were generated. In
order to export the nodal results of the simulation results, spatial points were traced from the inlet to the outlet to in
forms of streamlines (Fig. 6.3).

Figure 6.3 In CFD simulation stage

93

6.1.3 3D visualization stage
Adapting Autodesk 3ds Max as one part of the whole workflow was useful because it can be used as an intermediate
between Revit and game engines, but also because it had a significant function of visualizing CFD simulation data in
terms of velocity, temperature, and pressure in various ways. There were two parts in 3ds Max for visualization: the
first was to convert Revit geometry into game engine readable geometry including re-assigning materials to the
geometries; the second was to visualize the wind flow nodal data into splines that can be imported in game engines.
In this stage, the geometry of the kitchen space model and visualized airflows were imported to Stingray for VR
environment. Other than the geometries, a special CFD animation material shader was created that would be created
for the splines to animate the “flowing” effect (Fig. 6.4).

Figure 6.4 In 3D visualization stage
6.1.4 Virtual reality stage
The proposed methodology adopted Autodesk Stingray for creating the VR environment because of its high
interoperability with previously mentioned software, especially with 3ds Max. There was a Stingray add-in tab in 3ds
Max that users can send models back and forth directly through this command. Other than material editor, Stingray
uses visual programming language for other functions as well, which is beginner friendly. The user’s interface is pretty
streamlined to use. In this section, all the geometries and CFD splines were brought into Stingray and are edited ready
for VR. Then the complete VR environment was tested in Stingray using HTC VIVE HMD (Fig. 6.5).
94

Figure 6.5 In 3D visualization stage
6.2 Evaluation of the methodology
The test workflows and proposed methodology were evaluated with criteria of cost, interactivity, compatibility,
interoperability, visualization, accuracy, user-friendly and process speed (Chapter 5 has much more detail about this).
The evaluation and comparison of the visualization results among the three software that all generate the CFD wind
flow splines depending on the criteria of accuracy, graphics and experience and the ability of the visualization for
CFD analysis was discussed in the evaluation. In all, the visualizations in different software have different emphasis.
As for the visualization in CFD, it contributes more to the scientific application and the analysis of the data and has
graphic deficiency. The visualization in 3ds Max is convenient and graphically decent, however it is not informative
enough to represent the meaning of the data without color legend. In Stingray, the VR environment with CFD data
brings fully-immersive experience with relatively high presence. The visualization in CFD stage and 3ds Max stage
can be adopted separately depending on users’ needs such as data analysis or representation, but in order to realize the
CFD visualization in VR environment all these three stages are necessary.
6.3 Future work
After the reflection of the research, there are several topics that need future efforts from aspects of BIM, CFD
simulation, 3D visualization, and virtual reality (Fig. 6.6).
95

Figure 6.6 Future work based on the study
6.3.1 BIM geometry
Because BIM was used as the software for providing building information for later simulation and visualization,
compatibility and interoperability issues need to be considered further. Although Autodesk CFD had add-in in Revit
that allowed the Revit geometries import into Autodesk CFD directly, the geometries were still needed to be assigned
materials in order to carry out the simulation. One way to avoid the re-assignment is to create custom scripts that do
this for the user. Revit geometries cannot be directly used in Stingray and 3ds Max as the intermediate was needed;
custom scripts might solve this problem also or the choice of a different set of software programs.
6.3.2 CFD simulation
Autodesk CFD as a simulation software, is able to generate powerful simulation data that could be applied for other
use. In the proposed methodology, the nodal results exported from Autodesk CFD was post-processed in 3ds Max for
visualizations. Other than nodal results in CSV format, other data formats like TECPLOT format that may include
different content of the Autodesk CFD simulation can be post-processed in other ways. How to post-process the data
and increase the interoperability with other software remains a topic to be discussed.
The disadvantage with the built-in CFD visualization tool was obvious. From the perspective of graphics, the rendering
of the color was limited, the false-color display of the data was scientifically perfect but not ideally graphically decent
in Autodesk CFD. Although the simplified simulated model can be combined with the visualized result, the simplified
model with boxes may be confused for users that are not good at CFD simulation. This also brought up the necessity
for visualizing the CFD results in other software for the sake of better representation.
Even though Autodesk CFD had the built-in visualization tool to visualize the simulated results, whether the
visualization is valid and accurate enough is still a question. The workflow as developed had more emphasis on the
visualization features. More work needs to be done on the reliability and accuracy of the final data that is being passed
on for visualization. It is assumed that the software simulations were correct, but was that information being exported
and then imported properly into the visualization software?
96

6.3.3 3D visualization
The powerful tool in 3ds max to visualize CFD data was discussed in previous chapters, further research of using this
tool to visualize data other than CFD data will be useful or would the use of other 3d graphics software be more user-
friendly or fare better in the list of criteria that it was judged on.
Although the CFD data were represented in 3ds Max, the data cannot be played with for data analysis, such as to view
data points within a certain range of values. This could be helpful for scientific analysis.
The potential of visualization in 3ds Max is high; instead of using false color lines, better and more creative graphics
for visualization may be developed in future work. As in the methodology, the CFD animation material shader has
been created for the splines to generate a different graphic representation of the splines to help with scientific analysis
as well as graphic design (Fig. 6.7).

Figure 6.7 CFD animation material scripting in 3ds Max to create the animated air flow effect (upper); CFD splines without animation (lower
left); CFD splines with animation (lower right)
In 3D visualization stage, whether or not the built-in CFD visualization tool in 3ds Max was accurate enough
comparing to the original results remained a question to be further researched.
6.3.4 Virtual reality
As VR equipment needs a fast computer configuration to drive it, and the VR software has not been developed
thoroughly from an architectural perspective, the research on the hardware and software for VR, especially for building
industry use is needed. Further work needs to be done on the quality of graphics. This could start with a comparison
of visualization based on number of colors, pixels, re-draw time in different VR environments (both software and
hardware), and user surveys on what constitutes poor, good, and excellent color/resolution/speed.
In the VR environment, color legends can be added to help users to understand the visualization. Interactive sliders
that may change the Red Amount would be informative. The question remains in whether the visualization in VR
would be useful and helpful for actual scientific analysis or more for representation.
In order to provide more interactions to the VR environment, scheme switching functions between different
visualization (arrow, lines, dots etc.) can be added. What’s more, interactions like moving behaviors, sound, gestures
etc. are all useful elements to be created in VR.
The exploration in Stingray as for building science has not been touched much yet. The topics may exist in the
interaction between users and building environment to better help architectural and engineering design. The CFD
97

simulation results generated the airflow effects should have more applications than just for airflow representation. For
example, the CFD simulation may be used for game design providing actual wind effects. CFD visualization in VR
may be also useful to examine the hazard area within a space when there is dangerous gas. People can better understand
where to avoid the danger.
Then the VR environment may cause users’ VR nausea, especially when there were moving objects like the moving
airflow in the CFD visualization environment. The nausea is associated with the human brain (Barret 2004). Further
research can be laid on scientific visualization and virtual reality sickness.
6.3.5 Validation of the visualization and VR environment
Before the begin of the proposed methodology, a couple of tests workflow were tried but given up for visualization
setup and programming reasons. But the evaluation indicated they had the potentials to present great visualization.
Future works can be developed based on the tests.
The proposed evaluation of the whole methodology and case study is not effective enough. A significant research in
future work may be the quantitative evaluation of the scientific data visualization for diverse aspects, including users’
surveys, and other evaluate methods aiming for VR visualization. Ideally this should be done for different types of
users (students, clients, architects, engineers, owners, etc.) who may be using the VR environment for different reasons
(e.g. education, general understanding, detailed analysis, etc.). The VR results must be compared with other methods
of portraying the same data (e.g. charts of numbers, 2D graphs, 3D models, etc.) to evaluate how the display of data
is useful or not useful. The ease of producing the different visualizations should also be compared. This type of
validation and determining appropriate use is complex and difficult, but necessary.
6.4 Other applications in AEC of VR with BIM
The methodology developed concentrated on the use of VR as a visualization method for CFD data (and originally
weather and solar radiation data). This is just a tiny part of the lifecycle of a project (Fig. 6.8). In an actual building
project, the BIM geometry can be retrieved from conceptual design and detailed design stages. Even if in the proposed
method, the geometry simplification was after detailed geometry, the model in conceptual design stage was
informative enough for CFD simulation. CFD simulation belongs to the analysis stage for sure. During this stage,
different kinds of simulation including CFD were made to increase the building performance and thermal comfort.
The 3D visualization and VR parts were crossing over the analysis and documentation and operation/maintenance
stages. The reasons are, on one hand, the VR visualization is a part of the CFD simulation in different representation
method; on the other hand, the VR result can be documented as part of the deliverables as well as a visualized way to
maintain the building after construction.
98

Figure 6.8 Visualization workflow designed aimed at the BIM tools of the whole lifecycle (based on Calvert 2013)

Given that BIM is not a single software but a digital representation of physical and functional characteristics of a
project and a knowledge source serving the whole lifecycle of the project and storing the data from the beginning of
the project (National BIM standard 2014), thus, the methodology was trying to add information and functions to it
combining the advanced technology i.e. VR to contribute the diversity of BIM for a project while digging through the
possibility of working through multiple disciplines (Fig. 6.9).
99

Figure 6.9 Potential applications using BIM tools with VR in the whole lifecycle (based on Calvert 2013)

6.5 Summary
Revit objects (BIM) and air flow information (Autodesk CFD) were brought through 3ds Max (visualization), and
Stingray (VR environment) as an experiment to provide a better representation method for CFD data. Although there
are barriers between Revit objects and the VR environment, 3ds Max, and other intermediate software programs
helped complete the workflow. Using virtual reality to realize architectural design is feasible. At the same time, the
visualization of CFD simulation in VR offers more details to help architects understand what is happening in the actual
space, thus, to better assist the design.
From the perspective of the client and designer (architects and engineers), VR could better explain the air flow in the
space, which might also lead to refinement of the design. Especially with building environment simulations, clients
are able to give suggestions from their perspective when providing with more details, before the actual construction
is built. Unnecessary investment can be avoided because of the virtual reality. The communication will be increased
greatly as the clients can understand better.
The visualization of CFD results in VR also has an educational function. For students without building environment
knowledge background, the virtual reality environment can stimulate their interest of study learning what the graphics
represent, and the exploration of the environment would aid teaching. For students with building environment
background, the detailed and actual virtual environment with interactions will help them to diagnose the possible
problems better.

Interactive VR potentially has tremendous unrealized potential as another step in the sequence from 2D drafting to 3D
modeling to 4D animation to VR in the building industry. As both the hardware and software continue to develop, one
100

can imagine the game industry fueling these potentials and hopefully architects and engineers taking advantage of the
spin-offs.
The research on CFD visualization in VR environment using BIM tools worked as an experiment to combine several
popular technologies that contribute to the development of building industry in a fast-developing age. BIM tools
provide powerful techniques to help with projects during their entire lifecycle; traditional building analysis such as
CFD can be added newer elements to play a better role in helping the architectural design; and the potential of using
VR for both architectural visualization and scientific analysis should not be overlooked.

101

Appendix A

Figure 2.17 Application of VR from a life-cycle perspective ([1] https://i.ytimg.com/vi/P_UHZfAtZJg/maxresdefault.jpg; [2]
https://www.texomamedicalcenter.net/sites/texomamedicalcenter.net/files/thumbnails/image/surgery-room.png; [3]
http://www.bandt.com.au/information/uploads/2016/04/IMG_7071-1260x840.jpg; [4] http://www.interior-design-
academy.com/images/2016/interior-design-course10.jpg; [5]image by author; [6] https://4.bp.blogspot.com/-
jjdnDSSFcCw/VtIUCcDr9EI/AAAAAAAAA-g/KGl4ehLD_t8/s1600/2015-02-17_16-44-04-695x522.png; [7] https://nextstl.com/wp-
content/uploads/collaboration-image-2.jpg; [8] http://www.1ssh.com/files/photogallery/preview/Workers-waiting-for-tarp.JPG; [9]
http://symposia.asse.org/images/image2.jpg; [10] https://www.vrroom.buzz/sites/default/files/styles/article_top_banner/public/vr-in-
construction.jpg?itok=18Qf_6os&c=84bb9be94eb0ca84a6b51d24be9c42f1; [11] http://rockwallcontracting.com/wp-
content/uploads/2010/08/Sample-Project-Schedule.jpg; [12] http://www.cs.unc.edu/~marc/tutorial/img848.png; [13]
http://www.roadtovr.com/wp-content/uploads/2014/06/positional-tracking-technology-overview-virtual-reality.jpg.
102

CFD animation material scripting in 3ds Max to create the animated air flow effect

103

Bibliography
"Digital Media Design | Weather Data Visualization." Digital Media Design | Weather Data Visualization. Accessed
December 07, 2016. https://digitalmediadesign.io/projects/sector-app/weather-data-visualization.
"Weather Data for Simulation." Weather Data for Simulation | EnergyPlus. Accessed December 07, 2016.
https://energyplus.net/weather/simulation.
Aigner, Wolfgang, Silvia Miksch, Heidrun Schumann, and Christian Tominski. Visualization of Time-oriented
Data. London: Springer, 2011.
B. Jost, M. Ketterl, R. Budde and T. Leimbach, "Graphical Programming Environments for Educational Robots:
Open Roberta - Yet Another One?," 2014 IEEE International Symposium on Multimedia, Taichung, 2014, pp.
381-386.
Becerik-Gerber, Burcin, and Karen Kensek. "Building information modeling in architecture, engineering, and
construction: emerging research directions and trends." Journal of Professional Issues in Engineering Education
and Practice 136, no. 3 (2009): 139-147.
Bordnick, Patrick S., Brian L. Carter, and Amy C. Traylor. "What virtual reality research in addictions can tell us
about the future of obesity assessment and treatment." (2011): 265-271.
Froehlich, Mark A., and Salman Azhar. Investigating Virtual Reality Headset Applications in Construction.
Proceedings of 52nd ASC Annual International Conference.
Grondzik, Walter T., and Alison G. Kwok. Mechanical and Electrical Equipment for Buildings, 12th Edition. John
Wiley & Sons, 2014.
Jadidi, Hossein, Mehdi Ravanshadnia, Mujtaba Hosseinalipour, and Fatemeh Rahmani. "A Step-by-step
Construction Site Photography Procedure to Enhance the Efficiency of As-built Data Visualization: A Case
Study." Visualization in Engineering 3, no. 1 (2015). doi:10.1186/s40327-014-0016-9.
Jiao, Yi, Shaohua Zhang, Yongkui Li, Yinghui Wang, and Baoming Yang. "Towards Cloud Augmented Reality for
Construction Application by BIM and SNS Integration." Automation in Construction 33 (2013): 37-47.
doi:10.1016/j.autcon.2012.09.018.
Johansson, Mikael, Mattias Roupé, and Petra Bosch-Sijtsema. "Real-time Visualization of Building Information
Models (BIM)." Automation in Construction 54 (2015): 69-82. doi:10.1016/j.autcon.2015.03.018.
Langdon, Patrick. Designing Inclusive Systems: Designing Inclusion for Real-world Applications. London:
Springer, 2012.
Natephra, Worawan, Ali Motamedi, Tomohiro Fukuda, and Nobuyoshi Yabuki. "INTEGRATING BUILDING
INFORMATION MODELING AND GAME ENGINE FOR INDOOR LIGHTING VISUALIZATION."
In Proceedings of the 16th International Conference on Construction Applications of Virtual Reality, vol. 11, p.
13. 2016.
Park, Chan-Sik, Do-Yeop Lee, Oh-Seong Kwon, and Xiangyu Wang. "A Framework for Proactive Construction
Defect Management Using BIM, Augmented Reality and Ontology-based Data Collection Template."
Automation in Construction 33 (2013): 61-71. doi:10.1016/j.autcon.2012.09.010.
Sacks, Rafael, Amotz Perlman, and Ronen Barak. "Construction Safety Training Using Immersive Virtual Reality."
Construction Management and Economics 31, no. 9 (2013): 1005-017. doi:10.1080/01446193.2013.828844.
Sampaio, Alcínia Z., Miguel M. Ferreira, Daniel P. Rosário, and Octávio P. Martins. "3D and VR Models in Civil
Engineering Education: Construction, Rehabilitation and Maintenance." Automation in Construction 19, no. 7
(May 14, 2010): 819-28. doi:10.1016/j.autcon.2010.05.006.
Shi, Yangming, Jing Du, Sarel Lavy, and Dong Zhao. "A Multiuser Shared Virtual Environment for Facility
Management." Procedia Engineering 145 (2016): 120-27. doi:10.1016/j.proeng.2016.04.029.
Sun, Shang, Karen Kensek, Douglas Noble, and Marc Schiler. "A method of probabilistic risk assessment for energy
performance and cost using building energy simulation." Energy and Buildings 110 (2016): 1-12.
104

Weidlich, D., L. Cser, T. Polzin, D. Cristiano, and H. Zickner. "Virtual reality approaches for immersive
design." CIRP Annals-Manufacturing Technology 56, no. 1 (2007): 139-142.
105

REFERENCE
"3ds Max Interface Overview." Autodesk Support & Learning. Accessed May 01, 2017.
https://knowledge.autodesk.com/support/3ds-max/learn-
explore/caas/CloudHelp/cloudhelp/2016/ENU/3DSMax/files/GUID-212A9477-E69E-4174-BB6F-
1B7FD97A4281-htm.html.
"Convergence Plot." Autodesk Support & Learning. Accessed April 28, 2017.
https://knowledge.autodesk.com/support/cfd/learn-explore/caas/CloudHelp/cloudhelp/2017/ENU/SimCFD-
Learning/files/GUID-9169AD51-4916-4C43-80D0-C78213DCEC41-htm.html.
"Documentation." Developer Center - Documentation and SDKs | Oculus. Accessed May 01, 2017.
https://developer3.oculus.com/documentation/intro-vr/latest/concepts/bp_app_imaging/.
"IQ by Intel." IQ by Intel. Accessed March 27, 2017. https://blogs.intel.com/intellabs/2014/09/03/idf-2014-virtual-
reality-research.
"OpenFOAM User Guide: CFD Direct, Architects of OpenFOAM." CFD Direct. April 21, 2017. Accessed May 01,
2017. https://cfd.direct/openfoam/user-guide/.
"User Interface." Autodesk Support & Learning. Accessed May 01, 2017.
https://knowledge.autodesk.com/support/revit-products/learn-
explore/caas/CloudHelp/cloudhelp/2014/ENU/Revit/files/GUID-3197A4ED-323F-4D32-91C0-
BA79E794B806-htm.html.
"Virtual Reality." Virtual Reality. Accessed May 01, 2017. https://www.vrs.org.uk/.
A., Gutiérrez Alonso Mario, Frédéric Vexo, and Daniel Thalmann. Stepping into Virtual Reality. London: Springer,
2008.
Anderson, David B., John W. Barrus, John H. Howard, Charles Rich, Chia Shen, and Richard C. Waters. "Building
multiuser interactive multimedia environments at MERL." IEEE MultiMedia 2, no. 4 (1995): 77-82.
Aparicio, Manuela and Costa, Carlos (2015) Data visualization. Communication Design Quarterly Review 3, 1
(January 2015), 7-11. DOI=http://dx.doi.org/10.1145/2721882.2721883
Attia, Shady, and André De Herde. "Early design simulation tools for net zero energy buildings: a comparison of ten
tools." In Conference Proceedings of 12th International Building Performance Simulation Association, 2011.
2011.
Bahar, Yudi Nugraha, Jérémie Landrieu, Christian Pére, and Christophe Nicolle. "CAD Data Workflow toward the
Thermal Simulation and Visualization in Virtual Reality." International Journal on Interactive Design and
Manufacturing (IJIDeM) 8, no. 4 (2013): 283-92. doi:10.1007/s12008-013-0200-5.
Barrett, Judy. Side effects of virtual environments: A review of the literature. No. DSTO-TR-1419. DEFENCE
SCIENCE AND TECHNOLOGY ORGANISATION CANBERRA (AUSTRALIA), 2004.
Berger, Matthias, and Verina Cristie. "CFD Post-processing in Unity3D." Procedia Computer Science 51 (2015):
2913-2922.
BERGMAN, ARYN. "Taking Advantage of BIM for CFD Modeling." Taking Advantage of BIM for CFD
Modeling. 2012. Accessed May 01, 2017. http://au.autodesk.com/au-online/classes-on-demand/class-
catalog/2012/autodesk-simulation-cfd/taking-advantage-of-bim-for-cfd-modeling#chapter=0.
Boeykens, Stefan. Unity for Architectural visualization. Packt Publishing Ltd, 2013.
Bryson, Steve. "Virtual reality in scientific visualization." Communications of the ACM 39, no. 5 (1996): 62-71.
Calvert, Neil. "Why WE care about BIM…." Directions Magazine. Accessed May 01, 2017.
http://www.directionsmag.com/entry/why-we-care-about-bim/368436.
Carmigniani, Julie, and Borko Furht. "Augmented reality: an overview." In Handbook of augmented reality, pp. 3-
46. Springer New York, 2011.
106

Chung, Daniel Hii Jun, and Malone-Lee Lai Choo. "Computational fluid dynamics for urban design: the prospects
for greater integration." International journal of architectural computing 9, no. 1 (2011): 33-53.
Eastman, Chuck. BIM Handbook: A Guide to Building Information Modeling for Owners, Managers, Designers,
Engineers and Contractors,. John Wiley & Sons, 2015.
Elizardo, Jose. "3ds Max 2016 Ext 2 Part 2: Generating and Animating CFD." YouTube. December 18, 2015.
Accessed May 01, 2017. https://www.youtube.com/watch?v=OmNiDBYAO-s.
Elizardo, Jose. "3ds Max 2016 Ext 2 Part1: Generating and Animating CFD." YouTube. December 18, 2015.
Accessed May 01, 2017. https://www.youtube.com/watch?v=TdMlAxQEyG0&t=32s.
Furht, Borko. Handbook of Augmented Reality. New York: Springer, 2011.
Heilig, Morton L. SENSORAMA SIMULATOR. Filed January 10, 1961, and issued April 28, 1962.
Helbig, Carolin, Hans-Stefan Bauer, Karsten Rink, Volker Wulfmeyer, Michael Frank, and Olaf Kolditz. "Concept
and workflow for 3D visualization of atmospheric data in a virtual reality environment for analytical
approaches." Environmental earth sciences 72, no. 10 (2014): 3767-3780.
Hijazi, Mohammed Omar. Bridging the gap: A tool to support bim data transparency for interoperability with
building energy performance software. University of Southern California, 2015.
Hodgson, Eric. "IMS461 - Data Viz in Unity (HD)." Vimeo. May 01, 2017. Accessed May 01, 2017.
https://vimeo.com/59696565.
Hosokawa, Masahiro., Chien, S., S. Choo, M. A. Schnabel, W. Nakapan, M. J. Kim, and S. Roudavski.
"INTEGRATING CFD AND VR FOR INDOOR THERMAL ENVIRONMENT DESIGN FEEDBACK."
Jubertie, Sylvain. "Modèles et outils pour le déploiement d’applications de Réalité Virtuelle sur des architectures
distribuées hétérognèes." PhD diss., Thèse de Doctorat de l’Université d’Orléans, 2007.
Kaijima, Sawako, Roland Bouffanais, Karen Willcox, and Suresh Naidu. "Computational fluid dynamics for
architectural design." Architectural Design 83, no. 2 (2013): 118-123.
Kalkofen, Denis, Christian Sandor, Sean White, and Dieter Schmalstieg. "Visualization techniques for augmented
reality." In Handbook of Augmented Reality, pp. 65-98. Springer New York, 2011.
Kensek, Karen M. Building information modeling. Routledge, 2014.
Kensek, Karen, Douglas Noble, Marc Schiler, and Anish Tripathi*. “Augmented Reality: An Application for
Architecture.” 8th International Conference on Computing in Civil and Building Engineering (ICCCBE-VIII),
Palo Alto, California, August 14-17th, 2000, pp. 294-301.
Laine, Tuomas, Reijo Hänninen, and Antti Karola. "Benefits of BIM in the thermal performance management."
In Proceedings of the Building Simulation. 2007.
Li, Ning. "Comparison between three different CFD software and numerical simulation of an ambulance hall."
(2015).
Liu, Yifan, Geoffrey Tanudjaja, Zhouqian Jiang, and Natasha Beck. "Workflow of exporting revit models to unity -
BIM Wiki." Accessed May 23, 2017.
https://www.bing.com/cr?IG=BD6655D103674D9C9FB59302BD48A77F&CID=33FC87A31DFF6A492F178
D2F1CF96B84&rd=1&h=dPlKJlEASXWIJhtOxkF7iTY0_2PoWoatkexL5Gd_2kM&v=1&r=https%3a%2f%2f
bim.wikispaces.com%2ffile%2fview%2fRevit_3DS_Unity%2bWorkflow.pdf&p=DevEx,5064.1.
Maile, Tobias, Martin Fischer, and Vladimir Bazjanac. "Building energy performance simulation tools-a life-cycle
and interoperable perspective." Center for Integrated Facility Engineering (CIFE) Working Paper 107 (2007): 1-
49.
Malkawi, Ali M., and Ravi S. Srinivasan. "A New Paradigm for Human-Building Interaction: The Use of CFD and
Augmented Reality." Automation in Construction 14, no. 1 (2005): 71-84. doi:10.1016/j.autcon.2004.08.001.
Milgram, Paul, and Fumio Kishino. "A taxonomy of mixed reality visual displays." IEICE TRANSACTIONS on
Information and Systems 77, no. 12 (1994): 1321-1329.
107

Moon, Hyeun Jun, Min Seok Choi, Sa Kyum Kim, and Seung Ho Ryu. "Case studies for the evaluation of
interoperability between a bim based architectural model and building performance analysis programs."
In Proceedings of 12th Conference of International Building Performance Simulation Association, vol. 2011.
2011.
Nagel, Henrik R. "Scientific visualization versus information visualization." In Workshop on state-of-the-art in
scientific and parallel computing, Sweden. 2006.
Nall, Daniel H., P.E. and Drury B. Crawley PhD. 2011. "Energy Simulation in the Building Design
Process." ASHRAE Journal 53 (7): 36-43.
North, Chris, Ben Shneiderman, and Catherine Plaisant. "User controlled overviews of an image library: A case
study of the Visible Human." In Proceedings of the first ACM international conference on Digital libraries, pp.
74-82. ACM, 1996.
Procadsa. "Interactive Visualization with 3ds Max and Stingray." YouTube. January 22, 2016. Accessed May 01,
2017. http://www.youtube.com/watch?v=oH7Ly7KKZDw.
Shneiderman, Ben. Designing the user interface: strategies for effective human-computer interaction. Pearson
Education India, 2010.
Stuart, Rory. The design of virtual environments. New York: McGraw-Hill, 1996.
Taisei corporation. Http://www.taisei.co.jp/english/idc_e/environment/index.html. 2013. Accessed April 5, 2017.
Tang, Ying, Huamin Qu, Yingcai Wu, and Hong Zhou. "Natural textures for weather data visualization."
In Information Visualization, 2006. IV 2006. Tenth International Conference on, pp. 741-750. IEEE, 2006.
Tufte, Edward R. "Envisioning information." Optometry & Vision Science 68, no. 4 (1991): 322-324.
Tufte, Edward R. "Presenting data and information." In Unpublished seminar notes, New York. 2003.
Wang, Xiangyu, and Marc Aurel Schnabel, eds. Mixed reality in architecture, design, and construction. Springer
Science & Business Media, 2008.
Welle, Benjamin, John Haymaker, and Zack Rogers. "ThermalOpt: A methodology for automated BIM-based
multidisciplinary thermal simulation for use in optimization environments." In Building Simulation, vol. 4, no.
4, pp. 293-313. Tsinghua University Press, co-published with Springer-Verlag GmbH, 2011.
Wikipedia contributors, "Building energy simulation," Wikipedia, The Free
Encyclopedia, https://en.wikipedia.org/w/index.php?title=Building_energy_simulation&oldid=777151250 (acce
ssed May 1, 2017).
Wikipedia contributors, "Information visualization," Wikipedia, The Free
Encyclopedia, https://en.wikipedia.org/w/index.php?title=Information_visualization&oldid=764635254 (access
ed April 28, 2017).
Wikipedia contributors, "Information visualization," Wikipedia, The Free
Encyclopedia, https://en.wikipedia.org/w/index.php?title=Information_visualization&oldid=764635254 (access
ed May 1, 2017).
Yudi Nugraha Bahar.2014. “Representation of Thermal Building Simulation in Virtual Reality for Sustainable
Building.” PhD diss., Universite De Bourgogne.

Abstract (if available)

Abstract Building energy simulation is an important procedure during the building’s design. Not only can building energy simulation help identify opportunities for saving energy, but if interpreted correctly and implemented, it can contribute to the occupants’ comfort. Scientific visualization has been adopted for a long time in engineering field, tracking large scale simulation data and providing intuition and understandable graphs and models displaying the data. For computational fluid dynamics (CFD) data, the need for scientific visualization is of more importance, due to the complicated spatial data structure and large quantities of data points characteristic of CFD data. In the CFD simulation engines, there are built-in CFD simulation data visualization method such as streamlined 3D air flow visualization of an analyzed space. The raw CFD data could also be exported and processed in other software programs to add clarity. ❧ Given the consideration of better taking advantages of the CFD results for buildings, the potential of the use of virtual reality (VR) techniques cannot be overlooked. VR techniques bring about immersion and presence (Gutierrez et al. 2008). One can start with a building information model (BIM), produce CFD simulation results, and visualize those results in VR. There are existing achievements of CFD data or similar energy simulation data visualization in VR environment, such as visualizing Ecotect simulation data in VR game engine 3DVia Studio (Bahar 2014) and visualizing OpenFOAM CFD simulation data in Unity 3D (Hosokawa 2016). Some researchers have created CFD visualization tool as a way to streamline the workflow from CFD raw data to VR environment (Berger et al. 2015). The existing methods have given positive answers of the possibility to convert numerical CFD data to virtual environment objects. However, the previous workflows include problems in data exchanging complication (the software applied may be from totally different working areas), incomplete data representation, etc. As a result, a BIM to CFD data to VR visualization workflow is proposed. ❧ Building information modeling (BIM) as a lifecycle tool for buildings as includes as much as possible information for further applications. The proposed methodology starts from the building information model Autodesk Revit. CFD simulation is followed using Autodesk CFD Simulation through the add-in tab in Revit. Multiple kinds of CFD simulation including indoor natural and mechanical ventilation and outdoor wind pattern simulation can be done. The data type of nodal is exported for later visualization processing. To realize the visualization in VR, Autodesk 3Ds Max and Autodesk Stingray are applied continuously to manage the data and generate the VR objects. An actual case study of a kitchen space works is applied with the proposed methodology. In the final result, users are able to interact with the virtual environment wearing VR headset like HTC Vive, observing visualized CFD streamlines with color. ❧ Evaluation of the methodology is based on CFD data representation accuracy, user’s experience in both immersion, and presence perspectives and BIM integration level of the whole workflow.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Building information modeling based design review and facility management: Virtual reality workflows and augmented reality experiment for healthcare project

PDF

Using building information modeling with augmented reality: visualizing and editing MEP systems with a mobile augmented reality application

PDF

Visualizing architectural lighting: creating and reviewing workflows based on virtual reality platforms

PDF

Data visualization in VR/AR: static data analysis in buildings

PDF

Visualizing thermal data in a building information model

PDF

Bridging the gap: a tool to support bim data transparency for interoperability with building energy performance software

PDF

A BIM-based visualization tool for facilities management: fault detection through integrating real-time sensor data into BIM

PDF

MM Electrical Tool: a tool for generating electrical single line diagrams in BIM

PDF

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

PDF

Energy efficient buildings: a method of probabilistic risk assessment using building energy simulation

PDF

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

PDF

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

PDF

Building information modeling: guidelines for project execution plan (PxP) for India

PDF

Simplified acoustic simulation - Rutabaga Acoustics: a Grasshopper plug-in for Rhino

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

Acoustics simulation for stadium design using EASE: analyzing acoustics and providing retrofit options for the Los Angeles Memorial Coliseum

PDF

Augmented reality in room acoustics: a simulation tool for mobile devices with auditory feedback

PDF

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

PDF

BIM+AR in architecture: a building maintenance application for a smart phone

PDF

Quantify human experience: integrating virtual reality, biometric sensors, and machine learning

Asset Metadata

Creator Yan, Jiayi (author)

Core Title CFD visualization: a case study for using a building information modeling with virtual reality

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 06/20/2017

Defense Date 04/28/2017

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

Tag BIM,CFD simulation,OAI-PMH Harvest,scientific visualization,virtual reality

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Kensek, Karen M. (committee chair), Konis, Kyle (committee member), Noble, Douglas (committee member)

Creator Email jiayiyan@usc.edu,yanjiayi0327@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-387195

Unique identifier UC11258286

Identifier etd-YanJiayi-5429.pdf (filename),usctheses-c40-387195 (legacy record id)

Legacy Identifier etd-YanJiayi-5429.pdf

Dmrecord 387195

Document Type Thesis

Rights Yan, Jiayi

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