University of Southern California Dissertations and Theses

Contour crafting construction with sulfur concrete

(USC Thesis Other)

Contour crafting construction with sulfur concrete

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content I

Contour Crafting Construction with
Sulfur Concrete

By

Xiao Yuan

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(INDUSTRIAL AND SYSTEMS ENGINEERING)

August 2016

Copyright 2016 Xiao Yuan

I

ACKNOWLEDGEMENTS

Since my arrival at USC on August 8
th
, 2012, my experience in this campus has been nothing
short of amazing. This dissertation not only represents my research work in Contour Crafting
Technology, but also records the guidance and assistance from my professors, colleagues and
friends during these four years.
First and foremost, I would like to thank my advisor, Professor Behrokh Khoshnevis, Dean’s
Professor of Industrial & Systems Engineering and director of the Center for Rapid Automated
Fabrication Technologies (CRAFT). He has been patient and supportive since the first day I
started to work in the CRAFT lab. Ever since, he has supervised me not only providing technical
support on research problems, but also encouraging me when I faced difficulties. Influenced by
the innovative approach he posed when approaching new problems, I also exercised in
innovative thinking, product development, experiment design, as well as academic
communication.
I would also thank Professor Yong Chen for his priceless suggestion and support on my
dissertation and research. I admire his responsible attitude toward research. He taught me the
course, Mechatronic Systems Engineering, which helps in developing the prototype machines.
Many thanks to Professor Qiming Wang for his suggestion on theoretical analysis of my
dissertation.
II

My colleagues in CRAFT lab have contributed enormously to my personal and professional
development at USC. I really appreciate the time to work in such a motivated and diligent group.
I would like to particularly acknowledge Dr. Jing Zhang, who also worked with me together in
the Rapid Prototyping lab at Tsinghua University during my Master’s career. I extraordinarily
appreciate his passion, intensity, confidence and willingness in hardware development and
experimental design, and remarkable ability in academic presentation. Other previous or present
lab members at USC that I have had a joyful time to work with are Dr. Kai Xu, Behnam Zahiri,
Dr. Yayue Pan, Xuan Song, Payman Torabi, Matthew Petros, Hadis Nouri, Dongping Deng, Pu
Huang, Xiang Gao, Amir Mansouri, Aref Vali, Babak Zareiyan and Dr. Yongqiang Li.
I can’t express by words how thankful I am to my grandparents, my mother-in-law, father-in-
law, my mother, my father for all the unconditional love and endless support they provide. It’s
impossible for me to complete my work without the sacrifices they made. Finally, and most
importantly, I would like express my appreciation to my beloved wife, Shaohua Liu, for her
support, encouragement and understanding.

III

Abstract

Over the past three decades, additive manufacturing (AM), also widely known as 3D printing,
has impacted the traditional manufacturing industry for its specific capability in rapid
prototyping, fabrication of complex geometries, creation of multi-material composites and
customization. In contrast with most traditional manufacturing processes, which subtracts the
unintended material from a blank piece, the AM process fabricates part by successively adding
building material layer-by-layer based on a planned tool path generated from a CAD model. So
far, various additive manufacturing technologies have been developed for applications in areas
such as architecture, space, automotive, fashion, jewelry, energy and even biomedical. Classified
by processes, additive manufacturing contains extrusion, powder bed, light polymerized,
laminated and powder fed [1].
Contour Crafting is an extrusion based additive manufacturing process invented by Dr. Behrokh
Khoshnevis from University of Southern California, which can quickly build constructions such
as houses out of Portland concrete. By depositing wet concrete through a hose against the side
shaping trowel at the nozzle outlet, Contour Crafting can create a smooth surface finish over the
accumulated successive layers. A special designed triple-nozzle assembly can construct hollow
walls with increased strength and less amount of building material. This automatic construction
technology aims to release construction workers from dangerous and tedious work. In the Mars
and Lunar colonization missions led by NASA, contour crafting is considered to be the
promising construction candidate for out space construction, especially in automatically structure
protective hangars for equipment and shelter for astronauts.
IV

In this dissertation, the feasibility of contour crafting in space application is analyzed in
consideration of the accessible local construction raw material, the adversary environmental
conditions, the adjustment of existing Contour Crafting framework for space application and the
optimization on special extruder for construction material. A mobile 6-axis robotic system is
proposed which can fit in cargo area of launching system, deploy easily by itself after landing
and move around to build structures without size limitation. To save the enormous shipping cost
of building material from Earth and maximize the utilization of local resource, the sulfur
concrete, composed of planetary regolith and elemental sulfur, is chosen as the main construction
material for Contour Crafting space application. Unlike the chemical solidification process of
Portland cement, in sulfur concrete, the liquid sulfur in the extruded paste solidifies when the
temperature drops below 130 ℃ and acts as the binder of the aggregates, which is a physical
thermal reaction.
Accordingly, a novel lower pressure extruder is developed for this thermal extrusion process.
Since the temperature of sulfur concrete should be kept higher than 130 ℃ before extrusion, the
original extrusion process for Portland concrete of pumping building material from a terminal all
the way to the nozzle through a hose, is inapplicable due to the multifarious and complex heating
systems. This novel extruder is developed with an embedded sulfur concrete reservoir, gradient
thermal control system and compliant extrusion blades for wide range properties of sulfur
concrete.
To further control the sulfur concrete extrusion flow rate, its workability, includes mobility,
pumpability, formability and finishability, is studied. The mobility and pumpability represent the
ability of material to be extruded, while the formability and finishability represent the ability of
material to be shaped. Technically, the mobility and pumpability of a fluid material conflicts
V

with its formability and finishability. Experiments were carried out to test the influence factors
for both extrusion ability and shaping ability.
Sulfur concrete is a Bingham plastic, which behaves as a rigid body at low stresses and flows as
a viscous fluid at high stress. Without appropriate control, the solidifying sulfur concrete
expands in width and shrinks in height, which causes the failure of construction or terrible
finishing surface. A fraction factorial experiment and a full factorial experiment are carried out
to study the influence factors and the interaction of these factors. Finally, the deformation of
extrudate is controlled within the 5% by weight.

VI

Table of Contents
ACKNOWLEDGEMENTS ............................................................................................................. I
Abstract ......................................................................................................................................... III
Table of Contents .......................................................................................................................... VI
1. Introduction ............................................................................................................................... 1
1.1. Additive Manufacturing ................................................................................................... 1
1.2. Contour Crafting .............................................................................................................. 2
1.3. Previous Research of Contour Crafting ........................................................................... 6
1.3.1. Contour Crafting Process Study Using Ceramic Material ........................................ 6
1.3.2. Full Scale Concrete Wall Construction..................................................................... 7
1.3.3. Vibration Operated Valve for Abrasive Viscous Fluid............................................. 8
1.3.4. Summary of Contour Crafting History ..................................................................... 9
1.4. Application of Contour Crafting in Space Construction .................................................. 9
1.5. Research Challenges and Hypotheses ............................................................................ 11
1.6. Research Outline ............................................................................................................ 15
2. Large Scale Additive Manufacturing Construction Technologies in Space Application ....... 17
2.1. Automation and Robotics in Construction ..................................................................... 18
2.2. Cable Suspended Robotic Contour Crafting System ..................................................... 19
2.3. Powder Based Additive Manufacturing ......................................................................... 20
VII

2.3.1. Selective Separation Sintering ................................................................................ 21
2.3.2. D-shape ................................................................................................................... 23
2.4. Extrusion Based Additive Manufacturing ...................................................................... 25
2.4.1. Concrete Printing .................................................................................................... 26
2.5. 3D Printing in Space ...................................................................................................... 27
2.6. Summary ........................................................................................................................ 28
3. Feasibility Analysis of Martian/Lunar Construction .............................................................. 29
3.1. Space Colonization and Planetary Construction ............................................................ 29
3.1.1. In Situ Resource Utilization .................................................................................... 30
3.2. Construction Material for Space Application................................................................. 31
3.2.1. Introduction to Sulfur .............................................................................................. 31
3.2.2. Introduction to Sulfur Concrete .............................................................................. 33
3.2.3. Introduction of Modified Sulfur Concrete .............................................................. 37
3.2.4. Preparation of Sulfur Concrete ............................................................................... 39
3.3. Research on Regolith Simulant Sulfur Concrete Extrusion Process .............................. 40
3.4. Summary ........................................................................................................................ 46
4. Contour Crafting Equipment Development ............................................................................ 47
4.1. Sulfur Concrete Extruder Design ................................................................................... 47
4.2. Double Roller Extruder .................................................................................................. 48
VIII

4.3. Mini-size Auger Extruder............................................................................................... 49
4.4. Mini-size Auger Extruder with Piezo Vibrators ............................................................ 52
4.5. A Novel Sulfur Concrete Contour Crafting Extruder .................................................... 54
4.6. The Final Contour Crafting Extruder ............................................................................. 56
4.7. Modification for Improving the Extrudate Surface Quality ........................................... 59
4.8. Utilization of KUKA 6-axis Robot ................................................................................ 60
4.9. Sulfur Concrete Deformation Test Bed Design ............................................................. 63
4.10. Summary ..................................................................................................................... 64
5. Research on the Workability of Sulfur Concrete .................................................................... 65
5.1. Challenges in the SCCC Approach ................................................................................ 65
5.2. Extrudate Slump Analysis .............................................................................................. 66
5.3. Finite Element Analysis * .............................................................................................. 70
5.4. Research on Nozzle Design............................................................................................ 74
5.5. Sulfur Holding Capacity................................................................................................. 79
5.6. Research on Sulfur Concrete Extrusion Process ............................................................ 82
5.7. Summary ........................................................................................................................ 85
6. A Design of Experiments for Sulfur Concrete Deformation .................................................. 86
6.1. Fractional Factorial Experiment Design ........................................................................ 87
6.2. Experiment Analysis ...................................................................................................... 92
IX

6.3. Additional Full Factorial Experiment ............................................................................ 94
6.4. Additional Experiment Analysis .................................................................................... 97
6.5. The Extrudate Sample Cross-section Comparison ....................................................... 102
6.6. Summary ...................................................................................................................... 104
7. Conclusion ............................................................................................................................ 105
7.1. Recommendations for future work ............................................................................... 106
7.1.1. Complex Shape Construction ............................................................................... 106
7.1.2. Thermal Transfer Process ..................................................................................... 107
8. Bibliography ......................................................................................................................... 109
9. Appendix: R Code and Result for Experimental Data Analysis ........................................... 122

X

List of Figures
Figure 1.1 The milestones of additive manufacturing .................................................................... 1
Figure 1.2 Contour Crafting assembly and fabricating process ...................................................... 3
Figure 1.3 Contour Crafting construction ....................................................................................... 4
Figure 1.4 Four types of construction processes by Contour Crafting ........................................... 5
Figure 1.5 Ceramic material Contour Crafting technology ............................................................ 6
Figure 1.6 Full scale concrete wall construction ............................................................................ 7
Figure 1.7 Arc “bridging” phenomenon and “Bridging” breaker ................................................... 8
Figure 1.8 6-axis robot used for Contour Crafting ....................................................................... 10
Figure 1.9 Correlated chapters about each challenge and hypothesis .......................................... 14
Figure 1.10 Dissertation outline .................................................................................................... 16
Figure 2.1 Automation and robotic in construction ...................................................................... 19
Figure 2.2 Cable suspended robotic Contour Crafting system .................................................... 20
Figure 2.3 The working principle of powder based processes ...................................................... 21
Figure 2.4 Selective Separation Sintering ..................................................................................... 22
Figure 2.5 Selective Separation Sintering in space application .................................................... 23
Figure 2.6 The D-shape invented by Enrico Dini ......................................................................... 24
Figure 2.7 The demonstration of D-shape moon base .................................................................. 25
Figure 2.8 The working principle of fused deposition modelling ................................................ 26
XI

Figure 2.9 The demonstration of concrete printing ...................................................................... 27
Figure 2.10 On site printing and food printing in space ............................................................... 28
Figure 3.1 Earth, Mars and the Moon ........................................................................................... 29
Figure 3.2 Sulfur concentration in the upper few decimeters of the Martian surface .................. 31
Figure 3.3 Liquid sulfur viscosity variations with temperature .................................................... 33
Figure 3.4 Composition of Portland cement concrete and sulfur concrete ................................... 34
Figure 3.5 Strength development of Portland cement concrete and sulfur concrete .................... 34
Figure 3.6 Comparison of hydraulic concrete and sulfur concrete ............................................... 35
Figure 3.7 Water absorption for various sulfur and aggregate portions ....................................... 36
Figure 3.8 Sulfur components after modification ......................................................................... 38
Figure 3.9 SEM of elemental sulfur and bitumen modified sulfur ............................................... 38
Figure 3.10 Sulfur cement production process ............................................................................. 39
Figure 3.11 The composite of Martian regolith simulant sulfur concrete .................................... 41
Figure 3.12 Martian regolith simulant & Lunar regolith simulant sulfur concrete ...................... 43
Figure 3.13 The microstructure of Martian and Lunar regolith simulant ..................................... 44
Figure 3.14 Martian regolith simulant sulfur concrete printing .................................................... 45
Figure 4.1 Extrusion principle ...................................................................................................... 47
Figure 4.2 Extrusion system of double-roller design .................................................................... 48
Figure 4.3 Experimental setup of the double-roller extrusion system .......................................... 49
XII

Figure 4.4 Diagram of Extrusion System ..................................................................................... 50
Figure 4.5 Diagram of nozzle head and layer binding .................................................................. 50
Figure 4.6 Experimental machine for extrusion testing of sulfur concrete................................... 51
Figure 4.7 Schematic of Mini-size auger extruder........................................................................ 52
Figure 4.8 Pre-heating method on Martian regolith simulant sulfur concrete .............................. 53
Figure 4.9 Auger worn-out during experiment ............................................................................. 54
Figure 4.10 Schematic of multiple propellers extruder ................................................................ 55
Figure 4.11 A novel sulfur concrete Contour Crafting extruder ................................................... 55
Figure 4.12 Final edition of CC extruder and its temperature gradient ........................................ 56
Figure 4.13 Contour Crafting control system ............................................................................... 58
Figure 4.14 The construction process ........................................................................................... 58
Figure 4.15 The leaking and jaggy problems in the extrusion process ......................................... 59
Figure 4.16 The design of trowels on both sides of the nozzle..................................................... 60
Figure 4.17 Framework of Contour Crafting for space application .............................................. 60
Figure 4.18 Tool path of straight wall & half dome taken with long exposure camera ............... 61
Figure 4.19 Tool path coordinates for dome-like structure .......................................................... 62
Figure 4.20 Sulfur concrete deformation test bed model .............................................................. 64
Figure 5.1 Deformation of slump process..................................................................................... 67
Figure 5.2 Deformation process of an element ............................................................................. 69
XIII

Figure 5.3 Mohr-Coulomb failure model...................................................................................... 71
Figure 5.4 Mohr-Coulomb yield surface in meridional and deviatoric planes ............................. 72
Figure 5.5 The geometry of the specimen and the initial mesh used to implement the FEA ....... 73
Figure 5.6 The Mohr-Coulomb based FEA model and the extruded sample 35%wt ................... 73
Figure 5.7 The Mohr-Coulomb based FEA model and the extruded sample 30%wt ................... 74
Figure 5.8 The previous nozzles designed for plastic, gypsum and hydraulic concrete ............... 75
Figure 5.9 The flow path in the nozzles ........................................................................................ 77
Figure 5.10 The flow velocity distribution in the nozzles ............................................................ 78
Figure 5.11 The pressure distribution in the nozzle ...................................................................... 79
Figure 5.12 Sulfur holding capacity experiment........................................................................... 81
Figure 5.13 Extrusion process with different sulfur cement temperature ..................................... 83
Figure 5.14 Micro-structure of extrudate with different sulfur cement temperature .................... 83
Figure 5.15 Extrusion process with different sulfur proportion ................................................... 84
Figure 5.16 Micro-structure of extrudate with different sulfur proportion ................................... 84
Figure 6.1 Monothetic experimental method ................................................................................ 86
Figure 6.2 Horizontal and vertical deformation definition ........................................................... 87
Figure 6.3 Cuboidal representation of a 2
3
design ........................................................................ 88
Figure 6.4 The test process of sulfur concrete deformation.......................................................... 90
Figure 6.5 The deformation extrusion experiment with fractional factorial design ..................... 90
XIV

Figure 6.6 Half-normal distribution of location model in horizontal deformation ....................... 92
Figure 6.7 Half-normal distribution of dispersion model in horizontal deformation ................... 93
Figure 6.8 Half-normal distribution of location model in vertical deformation ........................... 93
Figure 6.9 Half-normal distribution of dispersion model in vertical deformation ....................... 94
Figure 6.10 Half-normal distribution of location model in horizontal deformation ..................... 97
Figure 6.11 Half-normal distribution of dispersion model in horizontal deformation ................. 98
Figure 6.12 Interaction plot of melting temperature against ash proportion ................................ 99
Figure 6.13 Interaction plot of melting time against ash proportion ............................................ 99
Figure 6.14 Half-normal distribution of location model in vertical deformation ....................... 100
Figure 6.15 Half-normal distribution of dispersion model in vertical deformation ................... 101
Figure 6.16 The cross section of sulfur concrete extrudate ........................................................ 103
Figure 7.1 Modified sulfur concrete printing on different platforms .......................................... 106
Figure 7.2 0° printed structures ................................................................................................... 106
Figure 7.3 10° printed structures ................................................................................................. 107
Figure 7.4 40° sloped curved structures ...................................................................................... 107
Figure 7.5 Two topics about thermal transfer ............................................................................. 108

XV

List of Tables
Table 1.1. Comparison between conventional CC and sulfur concrete CC.................................. 11
Table 2.1 Classification of additive manufacturing ...................................................................... 17
Table 3.1 Quantitative analysis and physical properties of aggregates ........................................ 35
Table 3.2 Comparison between sulfur concrete and hydraulic concrete ...................................... 37
Table 3.3 Major element composition of JSC-Mar1A and JSC-1A ............................................. 41
Table 3.4 Materials and processing parameters ............................................................................ 42
Table 4.1 KUKA KR150L robot parameter ................................................................................. 61
Table 5.1 Most efficient factors of SCCC .................................................................................... 66
Table 5.2 Microstructure of sulfur concrete ................................................................................. 80
Table 5.3 Sulfur holding capacity of different mixture ................................................................ 82
Table 5.4 A 2
k
factorial design on extrusion process .................................................................... 82
Table 6.1 Factors and levels of sulfur concrete extrudate experiment ......................................... 88
Table 6.2 Sulfur concrete deformation experiment planning matrix ............................................ 89
Table 6.3 Fractional factorial experiment design deformation data ............................................. 91
Table 6.4 Fractional factorial experiment horizontal deformation dataset ................................... 92
Table 6.5 Factors and levels, sulfur concrete extrudate experiment ............................................. 95
Table 6.6 Full factorial experiment design deformation data ....................................................... 96
Table 6.7 Full factorial experiment horizontal deformation dataset ............................................. 97
XVI

Table 6.8 Factorial effects of horizontal deformation .................................................................. 98
Table 6.9 Full factorial experiment vertical deformation dataset ............................................... 100
Table 6.10 Horizontal and vertical data statistical analysis ........................................................ 102
1

1. Introduction
1.1. Additive Manufacturing
Additive Manufacturing (AM), also called 3D printing, is the process of successively adding
layers of materials on top of previous layers according to a tool path generated by computer
directly from the CAD model of the part. Additive Manufacturing can almost build any
geometrical shape. A variety of materials, such as polymers, metals, and ceramics can be used in
this process [2]. Since the concept of AM was proposed in 1986, different processes have been
invented and developed for various applications during these 30 years. Figure 1.1 shows the
milestones of additive manufacturing technologies.

Figure 1.1 The milestones of additive manufacturing [3]
Additive manufacturing is fundamentally different from subtractive manufacturing, which
removes material from work blank by processes such as lathing, milling and polishing.
Accordingly, additive manufacturing has many advantages as compared with the other
manufacturing processes:
1) Regular manufacturing processes are always limited in the ability to fabricate very complex
shapes, such as hollow mesh, closed cavity and porous metal. Nevertheless, additive
manufacturing is capable of building such parts [2][3].
2

2) During subtractive manufacturing redundant material is removed from the blank part, which
causes significant waste. While, the only consumed material in additive manufacturing is the
fabricated part itself and some necessary mechanical support in certain processes.
3) In mass production using casting all parts are produced exactly the same from the same
casting mold and assembly line. While additive manufacturing can customize products according
to customers’ requirements, which is especially helpful in the dental and hearing aid industry as
well as in personalized fashion and jewelry making.
4) Saving time to market. The traditional prototype process requires tedious machining and
manufacturing work by humans and expensive equipment for a very long period. While, additive
manufacturing can dramatically reduce the R&D process time by fast prototyping in the same
day, which releases the researchers from low efficient work [2][4].
5) Printing with multiple materials. For some specific engineering applications, scientists want to
create a new composite with the characteristics of different materials, which unfortunately can’t
be achieved by a traditional manufacturing process. With the help of powder based additive
manufacturing, scientists can build parts with different material.
6) Additive Manufacturing shortens the steps from product design to creation of final products
has potential effect on the current economy and hence offers a new business model. For example,
customers can customize their new shoes with the help of designers and then directly print the
shoes on a 3D printer. Of course, privacy and copyright might be a concern during this process.
1.2. Contour Crafting
Contour Crafting (CC) is an extrusion based additive manufacturing process for automatic
construction of customized design invented by Dr. Khoshnevis from University of Southern
3

California in 1996. By extruding building material through a nozzle against the side trowel at the
outlet, Contour Crafting can create a smooth surface finish over the accumulated successive
layers. Comparing with other material extrusion additive manufacturing process Contour crafting
can achieve better surface quality, higher fabrication speed and wider range of building
materials. Up to now, thermoplastic and ceramic based Contour Crafting process have been
developed for its potential applications.

(a) Assembly with top and side trowels (b) Ceramic printing by Contour Crafting
Figure 1.2 Contour Crafting assembly and fabricating process
Figure 1.2 (a) shows the Contour Crafting nozzle assembly which consists of a top trowel and a
side trowel. During the extrusion process of each layer, the top and side trowels shape the top
and side to achieve a smooth surface. At the end of printing, the refill material is deposited into
the internal chamber to save the printing time of this volume. Figure 1.2 (b) shows the contour
fabrication process of Contour Crafting using wet ceramic.
Up to now, Contour Crafting can quickly build structures such as houses out of hydraulic (water-
based) concrete using an extrusion based fabrication system that implements a gantry robot
structure. A single house or a series of houses with different designs can be constructed in a
single run due to the benefit of customization of additive manufacturing. No printed support is
4

necessary in this process, support beams for window, door frames and ceiling are picked and
located by a robotic arm such as the one shown in Figure 1.3 [5]. The relative complex features,
such as domes and vaults, can be directly constructed without any external support when
combined with a proper path planning software.

Figure 1.3 Contour Crafting construction
1) This construction process allows architects to create personalized structures for customers.
The complex geometries that are hard to achieve by the existing construction process can be
accomplished with CC.
2) The nozzle can deposit multiple materials according to the construction requirements which
allows the structure to contain different number of components in different regions. Moreover,
this allows to control the chemical reaction in the printing process.
3) Painting can also be completed by the painting nozzle fixed to the robotic arm during the
construction process.
4) The electric and telephone line segments, similar to the industrial bus-bar, can be installed
inside the wall by an automatic robotic arm.
5

More specifically, there are generally four types of construction processes by Contour Crafting in
Figure 1.4. Concrete direct printing can build medium-scale construction and structure
demonstration models as shown in Figure 1.4 (a). While building large scale structures in Figure
1.4 (b), a small scale Contour Crafting nozzle can build a wall mold by cement and later filling
concrete into the mold. A special designed Contour Crafting multi-nozzle can build the wall with
good sound insulation and thermal insulation in Figure 1.4 (c). With the help of automatic arm
installing rebar within the wall, Contour Crafting can construct reinforced anti-earthquake
concrete wall shown in Figure 1.4 (d).

(a) Concrete direct printing (b) Concrete mold printing

(c) Reinforced concrete wall printing (d) Rebar reinforced concrete wall printing
Figure 1.4 Four types of construction processes by Contour Crafting
6

1.3. Previous Research of Contour Crafting
1.3.1. Contour Crafting Process Study Using Ceramic Material
Hongkyu Kwon studied the Contour Crafting process using ceramic material in his PhD
dissertation [6]. In his research, the Contour Crafting mechanical design, such as the shape of the
nozzle, trowel material and mechanical design, was optimized for printing of ceramic
geometries. To achieve complex shapes, the primary shapes were sequentially studied
considering the effect of side trowel and the composition of building material. The feasibility of
Contour Crafting applied in construction was also proven by testing the structures, such as
hollow cavities, reinforcements and impregnations, sandwich structures. To study the flow
characteristic of the extrusion, a computational fluid dynamic simulation was set up to study the
effect of different flow patterns and the influence on the final extrudate. Figure 1.5 (a) shows the
Contour Crafting machine for ceramic material, composed of linear actuator, side trowel control
motor and trowel control mechanism. Figure 1.5 (b) shows the hardened ceramic sample with
smooth external surface built by Contour Crafting.

(a) Contour Crafting machine for ceramic material (b) Sample built with smooth external surface
Figure 1.5 Ceramic material Contour Crafting technology
7

The ceramic material used in this research is the clay which includes Pioneer Talc 2882, Taylor
Ball clay, Barium Carbonate, Soda Ash, Sodium Silicate, and 35% water by mass[6]. The
engineering properties of clay are influenced primarily by its mineral composition and water
content. Placing the wet clay, a Bingham plastic fluid, in a high temperature furnace for hours
makes it become hard. This operation process makes it a very appropriate material for Contour
Crafting. Hongkyu Kwon’s research shows the potential of Contour Crafting in fast construction.
1.3.2. Full Scale Concrete Wall Construction
Dooil Hwang studied the Contour Crafting construction for full scale concrete wall in his
dissertation [7]–[10]. A full scale hydraulic Portland cement Contour Crafting machine was
developed to construct the composite concrete wall with external cement mold and concrete fill.
As it is shown in Figure 1.6 (a), the full scale Contour Crafting machine includes main frame,
piston extrusion system, material carrying tank and extrusion nozzle with CC’s trowels. The
extrusion nozzle can rotate 360°at the end of both sides to close the extrudate with a complete
loop.

(a) Full scale hydraulic Portland cement
Contour Crafting machine
(b) A concrete wall with external cement mold
and concrete fill
Figure 1.6 Full scale concrete wall construction
8

Once a layer is completed, the nozzle moves up a distance equal to the layer thickness. U-shape
form ties are manually inserted between the two solidifying mortar edges to prevent the
displacement caused by the lateral pressure from the fresh concrete filling. After filing any type
of fresh concrete into this cement mold, the fresh concrete gets solidified and works as main the
mechanical structure. Eventually, a 6 x 12-inch concrete cylinder was built for compressive
strength test. The average compressive strength of three samples is 2,946 psi.
1.3.3. Vibration Operated Valve for Abrasive Viscous Fluid
Khashayar Behdinan researched on the extrusion of abrasive viscous fluids using a vibration
operated valve [11], [12]. Concrete flow rate control is a crucial process in Contour Crafting,
especially when the fluid contains abrasive particles that the flow rate control is more
challenging. The resistance in the flow hose is due to the “bridging” phenomenon which is
caused by the interlock of abrasive aggregates. The higher a pressure the incoming material is
pushing against the arc “bridging”, the larger a resistance will be endured caused by the external
aggregate against the hose inner wall until the hosing wall is scraped off or the pipe breaks.
Vibration of the hose breaks the “bridging” phenomenon by agitating these mechanical locking
pattern and makes the abrasive material flow, as shown in Figure 1.7 (a).

(a) Arc “bridging” in a hose (b) “Bridging” breaker in a hose
Figure 1.7 Arc “bridging” phenomenon and “Bridging” breaker
9

Taking advantage of arc “bridging” phenomenon, the vibration of piezoelectric element controls
the flow rate of abrasive viscous fluids by adjusting the fraction of flow material against the hose
inner wall, as shown in Figure 1.7 (b).
1.3.4. Summary of Contour Crafting History
In view of the past research on Contour Crafting, the building material evolves from clay and
cement to Portland concrete with abrasive particles. The import of complex chemical reaction
and abrasive aggregates make the extrusion process more complicated. Therefore, it is necessary
to study the concrete flow phenomenon and its impact on CC process.
1.4. Application of Contour Crafting in Space Construction
Since NASA plans to send astronauts for Mars exploration in 2030, building a Martian base
ahead of this mission to protect astronauts and aircrafts from the extreme environment and
intense cosmic radiation is necessary. Taking advantage of the sulfur material that naturally
exists on the Martian surface, Sulfur Concrete Contour Crafting (SCCC) is regarded as an in-situ
resource utilization (ISRU) approach with the greatest potential for outer space construction. It
should be noted that the gantry structure is less attractive for planetary construction because of
the large size of the gantry which makes fitting in cargo area of launch systems problematic, and
implementation problems due to the requirement of autonomous assembly of the gantry upon
deployment. Furthermore, the CC system with a gantry structure cannot build anything larger
than its footprint.
Accordingly, a mobile robotic system such as the one shown in Figure 1.8 is proposed which will
be: a) more compact when its elements are retracted hence making it more suitable for launching,
10

b) possible to deploy much easier after being landed, and c) be able build structures with
practically unlimited size [12]–[15].

Figure 1.8 6-axis robot used for Contour Crafting
As described in Section 1.2, Khoshnevis invented and developed the Contour Crafting
technology using hydraulic concrete in 1996 and this technology is proven to be the most
efficient large-scale Additive Manufacturing process to this date. Various robotic structures have
been proposed for Contour Crafting. While, the gantry based motion control system which is
suggested for most terrestrial applications of CC, the limitation on size and the need for assembly
and disassembly for transportation, as well as launch rocket cargo space limitations make the
approach problematic for planetary construction. A more versatile approach based on mobile
robotics is hence suggested for Contour Crafting to overcome these obstacles [18]. In 2010,
Khoshnevis proposed the new sulfur concrete Contour Crafting approach for extraterrestrial
settlement infrastructure buildup to NASA. The mobile CC construction machine can work in
complex terrain and build large buildings [18].
Although numerous demonstrations have been made using hydraulic concrete, the NASA project
for the first time focused on construction using sulfur concrete. Compared to hydraulic concrete,
11

sulfur concrete has lots of advantages including acid-resistance, shorter curing time, recycle
ability, mechanical strength, and most importantly, availability on the Moon and Mars. Table 1.1
shows the comparison between conventional CC and sulfur concrete CC.
Table 1.1. Comparison between conventional CC and sulfur concrete CC
Process
Mechanical
performance
Dimensional
accuracy
Printing size Space application
CC Good Excellent Limited No
SCCC Good Good Unlimited Yes
1.5. Research Challenges and Hypotheses
This dissertation aims to study the application of Contour Crafting using in-situ building material
in space colonization. As discussed in Section 1.4, Mars and the Moon are current accessible
planet/satellite by human beings and are the targets of this research. The in-situ regolith is an
appropriate material for space construction. Since the processing of in-situ building material is
different from that of Portland concrete, it is necessary to redesign the extrusion system to fit in
this application. Moreover, there are very few references about using planetary in-situ building
material in the extrusion process. The workability of this material which represents if the
material is easily transported, placed, compacted and finished without any segregation is another
concentration [19]. Overall, the primary research challenges are listed as follows.
Challenge 1) How to develop an in-situ building material Contour Crafting equipment for
Space colonization application?
Challenge 2) How to control the extrusion and deposition parameters to balance the
pumpability and formability of this in-situ building material?
Accordingly, the following two hypotheses response to the previous research challenges.
12

Hypothesis 1) It is hypothesized that the planetary regolith building material can be
processed in a specific Contour Crafting system.
Hypothesis 2) It is hypothesized that the workability of in-situ building material can be
controlled by adjusting effect factors and the extrudate can be fabricated within 5% of
deformation as the outline of the extruder nozzle outlet.
The first challenge of Contour Crafting application in space colonization, as discussed in section
1.4, contains the accessibility of building material on planetary surface, preparation of planetary
regolith based building material, deposition process and durable Contour Crafting equipment
development. Another 2 sub hypotheses are proposed responding the following sub challenges.
Challenge 1a. What kind of material on the planetary surface is suitable for additive
manufacturing construction material?
Hypothesis 1a. There is an in-situ building material on the studied planetary surface which
works as building material of additive manufacturing construction.
Challenge 1b. How to develop a durable Contour Crafting equipment for space
application ？
Hypothesis 1b. A novel Contour Crafting equipment can be developed for the building
material and space application
To meet these challenges, the feasibility of Contour Crafting in space construction, especially on
the Moon and Mars, had been studied in Chapter 3 of this dissertation. Consequently, some
experiments about Martin/Lunar regolith simulant are carried out to prove the Hypothesis 1a.
Moreover, the Contour Crafting extrusion principle is discussed and a series of extruders for the
building material are developed as described in Chapter 4.
13

Khoshnevis indicates that sulfur concrete is a utilizable construction material on the Moon and
Mars due to the abundant inventory of sulfur concentration on the surface[18]. Sulfur concrete is
composed of sulfur and aggregates. Due to the special condition of sulfur concrete, each of the
earlier extrusion methods for Portland concrete was problematic in one way or another.
Workability is the most important property of molten sulfur concrete, which not only affects the
transportation and extrusion but also the strength and surface quality of the final product. Bartos
explained workability with several fundamental characteristics such as viscosity, mobility,
internal friction, pumpability, segregation, bleeding, formability and finishability [30]. In
Contour Crafting with sulfur concrete, mobility and pumpability of molten sulfur concrete affect
the extrudability of the mix. While, the formability and finishability affect the strength and
surface quality of extrudate structures. Unfortunately, the molten sulfur concrete with high
mobility and pumpability always has poor formability and finished quality and this is the
difficult contradiction in this project. Therefore, exploring the appropriate sulfur concrete for
multiple properties and developing effective nozzle assembly and control regimen for
controllably forming sulfur concrete are the two main tasks of this project.
In the Hypothesis 2, it is expected that the workability of ISRU building material can be
controlled by adjusting effect factors and the deformation of extrudate can be controlled within
5%. Identifying the proper values of controllable parameters and creating an effective control
system are two other challenges of this project. The sub challenges and hypotheses are listed to
support the C2 and H2.
Challenge 2a. How to control the extrusion process of Contour Crafting?
Hypothesis 2a. The mobility and pumpability of sulfur concrete can be controlled by its
composites proportion and mechanical optimization.
14

Challenge 2b. How to control the deposition process and extrudate deformation of Contour
Crafting?
Hypothesis 2b. The formability and finishability of sulfur concrete can be controlled by
mechanical optimization.
Challenge 2c. How to predict the dynamic extrudate deformation by a FEA model?
Hypothesis 2c. A FEA model can be established to predict the building material flow and
dynamic extrudate deformation.

Figure 1.9 Correlated chapters about each challenge and hypothesis
To face these sub challenges, the sulfur concrete flow phenomenon in extruders and some
extrusion characteristics of sulfur concrete are studied in Chapter 5. Compared with the previous
study of concrete flow in Contour Crafting system, operation temperature is an additional hard to
control factor in the entire process. To study the interaction of effect factors, a fractional factorial
CC in space
application
Challenge 1
Hypothesis 1
Challenge 1.a
Hypothesis 1.a
Section 3.1,
Section 3.2,
Section 3.3
Challenge 1.b
Hypothesis 1.b
Section 4.1 ~
Section 4.8
Challenge 2
Hypothesis 2
Challenge 2.a
Hypothesis 2.a
Section 5.1,
Section5.4,
Section 5.5
Challenge 2.b
Hypothesis 2.b
Section 5.2,
Section 5.3,
Section 5.6
Challenge 2.c
Hypothesis 2.c
Section 6.1 ~
Section 6.5
15

experiment and a full factorial experiment are carried out to optimize the deposition process in
Chapter 6. The correlated chapters about each challenge and hypothese are shown in Figure 1.9.
1.6. Research Outline
This dissertation concentrates on the application of Contour Crafting in Planetary Construction.
More specifically, the main research objectives of this research project include:
1) Analyze the application feasibility of Contour Crafting in the planetary construction using
Martian/Lunar regolith simulant
2) Develop an appropriate Contour Crafting extruder which is compliant with sulfur concrete and
a space applicable framework
3) Research on the influence factors of sulfur concrete workability and dynamic extrudate
deformation mechanism
4) Optimize the parameters of dynamic extrudate deformation and seek an acceptable process
parameter
Figure 1.10 shows an overview of all chapters in this dissertation.
16

Chapter 1
Introduction

Chapter 2
Literature Review

Chapter 3
Feasibility analysis of
Martian/Lunar construction
Chapter 4
Contour crafting
equipment development
Chapter 5
Research on the workability
of sulfur concrete
Chapter 6
Contour crafting construction
process optimization
Chapter 7
Conclusion and
recommendations for future work
Figure 1.10 Dissertation outline
Problem Clarification
Feasibility Analysis & Equipment Development & Process Optimization Conclusion

 Additive manufacturing & contour crafting
 Space application & ISRU material
 Challenges & hypothesize & contribution

 Related additive manufacturing processes
 Bingham Plastic deformation research
 Portland concrete workability research

 Research on the regolith simulant
 Regolith simulant sulfur concrete
 Sulfur concrete printing process

 Extruder principle analysis
 Mini-scale sulfur concrete extruder
 Full-scale sulfur concrete extruder
 Extrusion study

 Sulfur concrete workability
 Theoretical analysis and FEA
 Sulfur holding capacity

 Optimize sulfur concrete deformation
 Experimental design
 Deformation analysis

 Conclusion
 Recommendations for future work

17

2. Large Scale Additive Manufacturing Construction Technologies
in Space Application
Many additive manufacturing technologies have been developed since 1986. Categorized by
applications, most of additive manufacturing processes are developed for rapid prototyping in the
R & D phase and some costly processes are used to directly fabricate specialized components
with special properties. Table 2.1 shows the classification of additive manufacturing by
processing technology. Extrusion based fused deposition modeling is the most common 3D
printing process in community market. Light polymerization and material jetting can fabricate
high accuracy sample with detail features. Most of metal printing processes belong to powder
bed technology.
Table 2.1 Classification of additive manufacturing [1]
Type Technologies Materials
Extrusion
Fused deposition modeling Thermoplastics, rubbers
Contour Crafting Clay, cement, concrete
Concrete printing concrete
Light
polymerization
Stereolithography Photopolymer
Digital Light Processing Photopolymer
Powder bed
Electron-beam melting Titanium alloys, Stainless Steel, etc.
Selective laser melting Titanium alloys, Stainless Steel, etc.
Selective laser sintering Thermoplastics, metal powders, etc.
Selective separation sintering Ceramic powders, metal alloy powders
Material jetting
PolyJet Photopolymer
D-shape Sand
Laminated
Laminated object
manufacturing
Paper, metal foil, plastic film
Ultrasonic Additive
Manufacturing
Aluminum, titanium
18

As the development of manufacturing technology, advanced material and sensor industry,
additive manufacturing is also enhanced in terms of accuracy, quality, speed and dimension. Up
to now, additive manufacturing not only can provide the regular miniatures, prototypes and
customized jewelry, but also can fabricate full-size automotive and construct single-family
house. Researchers attempts to achieve full-scale construction and the components of a real
buildings, such as wall, column and building facades using additive manufacturing. Up to now,
there are three main constructing additive manufacturing processes that are actively pursued.
These are Contour Crafting, D-shape and Concrete Printing.
2.1. Automation and Robotics in Construction
As early as 1980s, construction robotics and automation were designed to replace or support
human labor in some specific construction related mission [20]. Japanese led this automation
construction revolution, primarily due to the predicted serious labor shortage of the construction
labor community [21]. US focused on the development of remote control and tele-operated
machines to release the construction workers from hazardous work. In Europe, the development
of large communal facilities robots, such as for brick laying and road paving, and automatic
structure construction process attracted more attention.
Figure 2.1 shows the automation equipment developed two decades ago. Figure 2.1 (a) shows a
Japanese developed concrete surface finishing robot. Figure 2.1 (b) shows that a concrete
pumping machine transports fresh concrete from a concrete truck to the architecture mold by
articulated-boom hose. Figure 2.1 (c) shows a remote control robot sprays concrete to a
reinforced framework. Figure 2.1 (d) shows a Germany made automatic concrete slip form road
paving machine. Overall, the development of automation and robotics makes the construction
industry more efficient and safer.
19

(a) Concrete surface finishing robot
(b) Articulated-boom machine for the
pumped delivery of fresh concrete

(c) Tele-operated concrete spraying robot (d) Concrete slip form road paving machine
Figure 2.1 Automation and robotic in construction [20]
2.2. Cable Suspended Robotic Contour Crafting System
Bosscher proposed a cable suspended robotic Contour Crafting system (Figure 2.2 (a)) with
better portability, lower cost, and the possibility to build much larger structures [22]. This
framework is relatively simple and is composed of a cube frame, multiple cables and
construction nozzle. The position and orientation of construction nozzle is controlled by the
crossbar (Figure 2.2 (b)) at each corner and three motors near each cube lower vertices. Four
upper cables, linked to the nozzle through each cube upper vertices, support the weight of nozzle.
While, eight lower cables, warped around the crossbars, controls the motion of nozzle in XY
20

plane. Although this innovation reduces the cost of framework structure, the complex driving
mechanism (12 motors and 4 linear actuators) add significantly to the complexity of the machine.

(a) Robot building a structure with
crossbars raised
(b) Crossbar in lowered and raised configurations
Figure 2.2 Cable suspended robotic Contour Crafting system [22]
2.3. Powder Based Additive Manufacturing
In powder based processes the powder of building material is paved and compacted by a roller or
a blade onto the build platform. Then, a heating source (such as laser or electron beam) scans the
surface of the material and melts/sinters the layer profile to form a single layer of the part. At the
end of each layer, the building platform moves down and another layer of loose powder is paved
on top of it for processing the subsequent layer. Once completed, the redundant powder can be
recycled. An advantage of the powder bed process is that no support structure is needed in the
printing process since the loose powder can support the cantilevered sections of the parts in the
successive layers. Laser and electron beam are two common heat source options shown in Figure
2.3. For both of these processes the build volume is totally sealed or vacuumed depending on the
material characteristics to guaranty the accurate temperature and reaction conditions [2].
21

Upper left. Selective laser melting Right. Electron-beam melting Lower left. EBM Components
Figure 2.3 The working principle of powder based processes [3]
In addition, at MIT a binder jet method was developed which deposits low viscosity binding glue
to join selected areas of powder particles. This technology is capable to process a variety of
materials, such as metals, sand and ceramic and achieve relatively high density solid parts [23].
2.3.1. Selective Separation Sintering
Khoshnevis invented Selective Separation Sintering (SSS) in 2015 and won the first place in the
NASA In-Situ Materials Challenge in 2016.
In the SSS process, two kinds of powders are used, the base power (B-powder) and the separator
powder (S-powder) [24]. The B-powder as the building material, such as steel alloy powder and
ceramic powder, has a relatively lower sintering temperature compared with the S-powder. Most
commonly, tungsten is used as S-powder since it has a higher sintering temperature. As shown in
Figure 2.4 (a), the white S-powder in the cylinder is selectively deposited into the B-powder
(black on the platform) by a vibrator to form a barrier between the surrounding region and the
22

powder that finally becomes the part. At the end of printing, the completed green part is placed
into a sintering furnace. The temperature of sintering furnace is set higher than that of B-powder,
but lower than the S-powder. During the sintering process, the B-powder particles combine with
the adjacent B-powder particles and become a solid part. While, the S-powder remains loose and
can be easily separated from B-powder [24].

(a) Selective separation sintering priciple (b) Stainless steel part built by SSS
Figure 2.4 Selective Separation Sintering [24]
For planetary construction, both NASA and European Space Agency follow a strategy of in-situ
resource utilization, which provides most of materials and energy on site for life support,
structure construction and space exploration. Printing on the current approachable planets, the
Moon and Mars, faces plenty of noticeable difficulties. Firstly, the building material of
construction should be able to resist the large differential temperature of 123℃ ~ –233℃ on the
Moon and 20℃ and -153℃ on Mars. Secondly, the ISRU construction specifically on Mars will
be exposed to the sand storm with the wind speed of up to 30m/s. Thirdly, the vacuum
environment on the Moon and Mars makes the constructing process different than that on the
Earth. Up to now, there are three potential additive manufacturing construct process proposed:
Contour Crafting (Section 1.4) & Selective Separation Sintering funded by NASA and D-Shape
chosen by European Space Agency.
23

(a) Space craft landing pad (b) Martian regolith parts built by SSS
Figure 2.5 Selective Separation Sintering in space application
To protect the landing aircraft from the recoil thrust, Selective Separation Sintering is also
proposed for creating landing pads and roads by means of interlocking tiles created by SSS. In
this arrangement a high sintering temperature S-powder such as alumina or magnesia, which is
abundant on other planets, is deposited into the regolith of planetary surface, as shown in Figure
2.5 (a). Then, the regolith is successively sintered by microwave until the entire landing pad or
road is completed. The landing pad is expected to resist the high temperature of the lander
thrusters’ exhaust plumes. Figure 2.5 (b) shows the sintered lunar regolith simulant parts built by
Selective Separation Sintering.
2.3.2. D-shape
The D-shape process uses a large-scale powder based binder-jetting printer, invented by Enrico
Dini, the founder of Monolite UK Ltd. The D-shape printer frame shown in Figure 2.6 (a) is
composed of a square base and four vertical columns, which defines the maximum volume of the
printer.
The jetting head with 300 nozzles fixed to the frame can deliver an inorganic binding glue
composed of magnesium hexahydrate as a binder to react with the building material and produce
a sandstone-like surface. Since the layer thickness of building material is between 5 to 10mm,
the printer is specifically designed to guarantee that the binder can sink into the bottom of each
24

layer. After about 24 hours until the binder gets solidified the extra sand which performs as
support can be recycled for later use. A finished part is shown in Figure 2.6 (b). Enrico Dini also
proposed the application of D-shape on building habitats on the Moon using Lunar regolith in
2013.

(a) The printing process of D-shape (b) The completed part
Figure 2.6 The D-shape invented by Enrico Dini [7] [8]
To construct the planetary base, a D-Shape chamber (Figure 2.7 (a)) with all construction
equipment is sent to the planetary surface. Then, D shape takes advantage of a high pressure
inflatable “balloon” (Figure 2.7 (b)) to serve as the support of the dome structure. The
construction robots start to transport regolith and to deposit the binder into each layer to bind it
(Figure 2.7 (c)) until the shell is completed (Figure 2.7 (d)). This process can create a foam like
wall (Figure 2.7 (e)) to save binding material and to increase mechanical strength. Lastly, the
original “balloon” will be replaced by a lower pressure “balloon” (Figure 2.7 (f)) in which
astronauts can live and work. At the same time, the void space between the lower pressure
“balloon” and D-Shaped structure can work as an insulated chamber to resist the large
differential temperature [25], [27].
25

(a) Landing on the moon (b) Inflatable structure

(c) Robotic printers (d) Work done

(e) Foam like internal structure (f) Low pressure inflatable dome
Figure 2.7 The demonstration of D-shape moon base [7], [10]
2.4. Extrusion Based Additive Manufacturing
The extrusion process, also known as Fused Deposition Modelling, is the most common 3D
printing process which was firstly developed by Stratasys Company in the early 1990’s [3]. This
affordable technology rapidly advanced until in 2009 due to the expiring of registered patent and
a related open source commerce emerged.
As shows in Figure 2.8, in FDM the filament, mostly ABS or PLA, from material spool is melted
by the heated extrusion head and deposited layer-by-layer onto the building platform to form a
26

solidified model. The nozzle path is planned according to the 3D data imported to the printer.
Mechanical support is printed under the cantilevered or hollowed geometries to support them and
can be easily removed or dissolved by water at the end of the printing process.

Figure 2.8 The working principle of fused deposition modelling [3]
2.4.1. Concrete Printing
The concrete printing is another extrusion based concrete additive manufacturing process
proposed by Sungwoo Lim. Compared with the two additive manufacturing construction
technologies introduced before, this process uses a nozzle in a smaller resolution to achieve a
better control on the detailed feature and a coarse nozzle to fill the internal chamber, as it is
shown in Figure 2.9 (a). At the beginning of each process, the printing head with a hopper on top
is refilled with the building material at the recharging position. Then, it can deposit the cement
mortar or gypsum according to the design until the material in the hopper reaches the low
material level. Figure 2.9 (b) shows the finished part of concrete printing.
27

(a) The printing process of concrete printing (b) The completed part
Figure 2.9 The demonstration of concrete printing [28]
2.5. 3D Printing in Space
As the additive manufacturing starts to impact construction on Earth, researchers also begin to
explore the applications of additive manufacturing in space. Generally, there are three main
distinct areas: manufacturing of special tools, printing sustainable food for astronauts and
construction of planetary base for space exploration.
Considering the cost of shipping tools to the space destination by space shuttle and the waiting
time of weeks for some critical maintenance parts, printing a tool or part locally is deemed to be
the preferred approach. Unfortunately, the zero gravity condition restricts the application of
powder based additive manufacturing processed. A company, named Made In Space, developed
an extrusion based 3D printer shown in Figure 2.10 (a) which can build polymeric parts within a
few hours on demand. Moreover, long term missions, such as Mission to Mars, would benefit
even more from such an on-board machine shop [25].
28

(a) Made in Space 3d printer (b) Crew members having meal
Figure 2.10 On site printing and food printing in space [25]
NASA also funded the Systems and Materials Research Consultancy project phase I at $125,000
to pursue the study of 3D printing food in space. The key mission of this research is to guarantee
that the astronauts can have a healthy and variety diet with abundant diversity. For some long
term trips in outer space, this research offers the potential to meet the nutrition supplies using
only the elementary food components (see Figure 2.10 (b)) [25].
2.6. Summary
As early in 1980s, automation and robotics were used in construction industry to liberate human
labor from hazardous and tedious work. Many semi-automatic construction machines for
infrastructure building, such as paving road and fresh concrete pumping, are developed which
dramatically improved the operation efficiency. The advantages of additive manufacturing will
make it a superior tool in space application. Several large-scale additive manufacturing processes
for construction and their applications in space are discussed.

29

3. Feasibility Analysis of Martian/Lunar Construction
3.1. Space Colonization and Planetary Construction
Moon and Mars have been identified as the most promising planets to offer a variety of
opportunities while being relatively close to Earth. The Moon is the only natural satellite of
Earth, and the shortest travel time to the Moon by Atlas V rocket is 8.5 hours [29]. However, a
trip to Mars takes more than 6 months using the present propulsion system. Figure 3.1 shows the
real size comparison of Earth, Mars and the Moon. The Moon diameter is approximately ¼ of
Earth diameter; and Mars diameter is slightly larger than half diameter of Earth.

Figure 3.1 Earth, Mars and the Moon
Mars has become a considerable planet for the following reasons: 1) The rotation period of Mars
is 1 day and 40 minutes, which is very close to Earth rotation period; 2) The axial tilt of Mars is
very similar to Earth. This means that Mars also has seasons during the 1.88 Earth year
revolution; 3) Most recently, the existence of the ice water on Mars is confirmed by NASA.
Verseux Cyprien proposed a cyanobacteria that can be developed as a self-sustainable pioneer
on Mars [30]. Its application includes producing food, fuel and oxygen, as well as supporting
30

other organisms. However, artificial environment and life support equipment are still necessary
for human survival on Mars [31].
As recently stated, the possibility of existing water at the Lunar poles, attracted new exploration
efforts [32]. Taking the advantages of low gravity and low escape velocity, the Moon can serve
as a transfer station for launching rockets in exploring the solar system. Lunar lava tubes,
discovered in early 2009 [33], are regarded as a natural reliable structure to shelter equipment
and humans from the severe atmospheric condition of the Moon.
In this research, the application of Contour Crafting for construction on the Moon and Mars is
studied.
3.1.1. In Situ Resource Utilization
To reduce the otherwise huge transportation cost of space colonization, in situ resource
utilization is considered, initially for infrastructure and in the future for construction of human
habits. According to NASA, "in-situ resource utilization will enable the affordable establishment
of extraterrestrial exploration and operations by minimizing the materials carried from Earth."
[34] Planetary regolith, also known as “planetary soil”, is a fluffy layer covering the dense rock,
which includes ash, soil, crushed rock and some other minerals. Both Mars and the Moon are
covered with regolith. The regolith on the Moon is mostly formed by the impact of meteoroids
from solar system and galaxy over the last 4.6 billion years [35]. The planet-wide dust storms on
Mars make its regolith very fine. The suspended dust in the atmosphere bears a reddish outlook
to Mars [36]–[42].
In 1985, Larry A. Beyer at the University of Pittsburgh proposed Lunarcrete (Lunar sulfur
concrete) for construction on the moon [29]. This proposal is based on the abundant inventory of
31

mineral troilite (FS) [43] on Lunar surface, which can be refined to acquire elemental sulfur.
Unlike hydraulic Portland cement, sulfur concrete is an energy-saving construction material
which does not require an ultra-high temperature to purify cementitious components [44]. Sulfur
melts at 140 °C and performs as the binder of aggregates.
Similarly, the core of the Mars as well as some of its regional surface is abundantly comprised of
sulfur, as it is shown in Figure 3.2. Therefore, in Contour Crafting space application, sulfur
concrete is researched as the building material [30], [45]–[50].

Figure 3.2 Sulfur concentration in the upper few decimeters of the Martian surface
3.2. Construction Material for Space Application
To better understand the characteristics of sulfur concrete material, its chemical and physical
properties, the preparation process and its current applications on Earth are presented in this
section.
3.2.1. Introduction to Sulfur
As a byproduct of natural gas and oil industry, sulfur has been studied since 17th century and
widely used in construction and modeling fields. Due to its superior mechanical and chemical
32

properties, sulfur concrete is used in cold weather placement, wind and moisture environments,
repairing damages, joint sealing, forming and reinforcement [51]. In all these processes, sulfur
concrete has been conventionally casted manually or semi-automatically. These methods are
inefficient and labor intensive. As the development of Contour Crafting progresses, sulfur
concrete may be finding a huge potential for widespread applications for construction of a
variety of infrastructure elements such as bridges, towers, etc. [18], [52]–[56].
Elemental sulfur is known in #16 as one of the most abundant elements in nature which is also
called brimstone, usually regarded as free resource near volcanoes and hot springs. In industrial
operations, sulfur is the byproduct of the petroleum refining and natural gas industry [57].
Although sulfur has been observed as S2, S3, S4, S5, S6, S7, S8, S12, S ∞ and S π in chemistry,
the orthorhombic and monoclinic sulfur are the only stable solid states under native conditions.
In liquid form polymerization of sulfur takes place at the approximate temperature of 159.4 ℃,
where cyclo-S8 changes to Catena-S8 and the combination of both changes to Catena-S8*2:
–S8 ↔ Catena–S8
Catena–S8 + -S8 → Catena–S8*2
This process is also regarded as irreversible [58].
In solid form, both orthorhombic and monoclinic sulfur are composed of cyclo-S8 structure. The
orthorhombic sulfur is regarded as the most stable and common until the temperature reaches
95.3°C at which orthorhombic sulfur changes to monoclinic sulfur as temperature increases
[58].
33

Bacon and Fanelli discovered that the viscosity of liquid sulfur reduces with the increasing of
temperature at around 160 ℃ [59]. From this point, the viscosity dramatically increases to a
maximum of 10
3
Poise at 190 ℃, as it is shown in Figure 3.3 [57].

Figure 3.3 Liquid sulfur viscosity variations with temperature [57]
3.2.2. Introduction to Sulfur Concrete
Sulfur concrete is a composite construction material composed of sulfur as the binding agent and
aggregate (coarse aggregates such as gravel rocks and stones and fine aggregates such as sand
and ash). The large amount of available sulfur and the relatively low price make sulfur concrete a
potential candidate to replace the use of regular Portland cement.
As it is shown in Figure 3.4, Portland cement concrete is a mixture of coarse aggregates, fine
aggregates, water and cement. While, elemental sulfur concrete includes aggregates and
elemental sulfur [57].
34

Figure 3.4 Composition of Portland cement concrete and sulfur concrete [57]

Figure 3.5 Strength development of Portland cement concrete and sulfur concrete [57]
Compared to Portland cement concrete, elemental sulfur concrete has several superior properties
such as rapid strength development and higher compressive strength compared to most common
concrete. As it is shown in Figure 3.5, to acquire better compressive strength, Portland cement
requires moisture and temperature conditions until it achieves 90% of final strength in 1 month.
While, elemental sulfur concrete only takes a few hours to reach its final strength without
requiring special conditions. The final compressive strength of elemental sulfur concrete is
40~50 MPa which higher that of common Portland cement which ranges from 24~35MPa [60].
35

Besides proportion of elemental sulfur, aggregates also have great impact on the final
compressive strength. Gravel, crushed rocks, natural sand lava and crushed bricks are the most
common aggregates. In the 1960s, Dale and Ludwing in Southern Research Institute carried out
experiments on three aggregates: gravel, granulite and basalt.
The physical properties of aggregates and composition of aggregates are shown in Table 3.1.
After 6 to 12 hours’ solidification, the typical compressive strength of these sulfur concrete
specimens ranges from 33 to 47Mpa. They also pointed out that the major problem with
elemental sulfur as the binder was the poor resistance to thermal cycling [57].
Table 3.1 Quantitative analysis and physical properties of aggregates [57]
Element Gravel Granulit Basalt
SiO2 33.3 70.7 42.7
Al2O3 - 15.5 16.6
Fe2O3 - 3.4 10.9
K2O - 3.8 2.5
Na2O - 4.9 5.0
CaO Rest - -
MgO - - -
TiO2 - - -
Loss at red heat 26.6 - -
Specific gravity 2.64 2.65 2.90
Compressive strength 216 236 305

Figure 3.6 Comparison of hydraulic concrete and sulfur concrete [57]
36

Because of the superior corrosion-resistant of sulfur concrete this material has been used for
fabrication of corrosion-resistant tanks and floor surfaces in highly corrosive areas. Figure 3.6
shows the comparison of hydraulic cement concrete and sulfur concrete in 50 wt% sulfuric
acid [61].
The amount of water absorption by concrete influences the durability of the structure built. The
standard measure is determined by the amount of water absorbed 24 hours after construction
under normal condition. Figure 3.7 shows the changes of water absorption for various sulfur and
aggregate portions. Figure 3.7 shows that with sulfur content higher than 17% by weight, the
water absorption amount is less than 0.05% by weight [51].

Figure 3.7 Water absorption for various sulfur and aggregate portions [57]
Generally, Compared to regular Portland cement, sulfur concrete has the following advantages in
Table 3.2 [57]:

37

Table 3.2 Comparison between sulfur concrete and hydraulic concrete [57]

3.2.3. Introduction of Modified Sulfur Concrete
In the solidification process, liquid sulfur turns into solid sulfur at a temperature of 95.5C with
7% contraction in volume. Then, monoclinic β-phase sulfur transforms into orthorhombic α-
phase in the next 24 hours with a further 6% volume shrinkage. As discussed before, elemental
sulfur concrete also has a large expansion coefficient. This means that it displays large extension
and shrinkage when exposed to thermal cycles, which may cause cracks and defects over a long
period of time. So, modifying sulfur concrete with chemical material and making it plasticized
before use is necessary to overcome these problems [57].
To plasticize and stabilize sulfur, Pryor mixed plasticizer with sulfur at the temperature 140C for
3 hours [62]. The research showed that liquid sulfur reacts with plasticizer in liquid form and
38

forms several types of polysulfide products. Figure 3.8 shows the sulfur components after
elemental sulfur reacting with styrene at 140 ℃ for 3 hours [63].

Figure 3.8 Sulfur components after modification [62]
Bitumen was used to modify molten sulfur by Mohamed and El Gamal [57]. In their report, an
oil bath 2.5 wt% molten bitumen was mixed with 97.5 wt% liquid sulfur at 140℃. After stirring
for 1 hour, samples were cooled down at the rate of 8~10 C/min and exhibited plastic property.
The final sample properties depend on the bitumen content, reacting temperature and time. In the
cooling process, bitumen performs as a yielding layer between the sulfur crystals and alleviates
the influence of the phase transformation from β to α phase, as shows in Figure 3.9 [64].

a) SEM of typical crystalline morphology of
pure elemental sulfur
b) SEM micrographs for sulfur modified with
2.5 wt% bitumen at different magnifications
Figure 3.9 SEM of elemental sulfur and bitumen modified sulfur [64]
39

3.2.4. Preparation of Sulfur Concrete
Sulfur concrete is a composite construction material composed of elemental sulfur and
aggregates, such as sand and crushed rocks. Compared to regular Portland cement which reaches
the maximum strength only after 28 days of curing, sulfur cement reaches its final strength
within several hours. The humidity and temperature of the external environment do not influence
the strengthening process at all. The compressive strength of final sulfur concrete product
reaches up to 50MPa which is comparable with high strength concrete. Its excellent resistance to
acid and salt environments as well as its superior water tightness makes sulfur concrete an ideal
construction material for marine construction and a wide range of other applications [57].

Figure 3.10 Sulfur cement production process [57]
However, due to the variations of elemental sulfur crystal, sulfur cement displays great
expansion and shrinkage rate in thermal cycles. According to the report, the daily temperature
differences on Mars and Moon are about 100 ℃ and 300 ℃, respectively. Therefore, it is
essential to modify the property of elemental sulfur to overcome expansion/retraction issue.
Nimer tried to add chemical additives, such as dicyclopentadiene, styrene or acrylic acid, to
elemental sulfur to plasticize sulfur in its polymeric form [65]. As it is shown in Figure 3.10,
40

sulfur and chemical additives are melted and mixed at the temperature range from130 ℃~150 ℃
for more than 3 hours. Then, modified sulfur is mixed with mineral fillers at the same
temperature until the sulfur is completely plasticized. The availability of comparable additives
on Moon and/or Mars should be studied to evaluate the possibility of making such modifications
to sulfur.
In the sulfur extrusion process, the viscosity and surface tension of the mixture are considered to
be two main factors. Blight approves that elemental sulfur and DCPD react in an exothermic
reaction, which would increase the temperature of reaction and make the sulfur tremendously
viscous [63]. Allan and Neogi also carried out research on the impacts of heating time and
DCPD proportion on viscosity and surface tension of modified sulfur. Both viscosity and surface
tension of sulfur with 5 wt% DCP gradually increase with time, while these parameters in sulfur
with 20 wt% DCP increase to 2~3 times more after 3 hours heating [66].
So far, research is concentrated on the extrusion process. Further effort will be dedicated to
improve the strength, durability and stability of modified sulfur cement using in-situ materials.
3.3. Research on Regolith Simulant Sulfur Concrete Extrusion Process
Martian regolith simulant JSC-Mar1A (aggregate of Martian sulfur cement) and Lunar regolith
simulant JSC-1A (aggregate of Lunar sulfur cement) made by Orbital Technology Corporation
are chosen in this extrusion experiment in Table 3.3. The purity of sulfur powder used in this
experiment is higher than 99.0%.
Sulfur and regolith may be premixed prior to being poured into the extrusion machine. However,
movement and vibration can displace the fine sulfur powder particles, making the mixture less
homogeneous. We have therefore devised a method to assure mix consistency. First, sulfur and
41

regolith were thoroughly mixed, and then heated to the melting point of sulfur. The mixture was
then cooled down to form a solid lump, which was crushed into fine chunks so each chunk
retained the correct proportion of sulfur to regolith. This method creates sulfur-coated regolith
particles that are very stable and resist settling back into their components. In Figure 3.11, the
lower left powder is the element sulfur and the lower right powder is the Martian regolith
simulant. The upper right powder is the mixture of Martian regolith simulant sulfur concrete.
Table 3.3 Major element composition of JSC-Mar1A and JSC-1A
Major Element Composition
Martian regolith simulant
JSC-Mar1A
Lunar regolith simulant JSC-
1A
Silicon Dioxide (SiO2) 34.5-44 46.67
Titanium Dioxide (TiO2) 3-4 1.71
Aluminum Oxide (Al2O3) 18.5-23.5 15.79
Ferric Oxide (Fe2O3) 9-12 12.5
Iron Oxide (FeO) 2.5-3.5 8.17
Magnesium Oxide (MgO) 2.5-3.5 9.39
Calcium Oxide (CaO) 5-6 9.90
Sodium Oxide (Na2O) 2-2.5 2.83
Potassium Oxide (K2O) 0.5-0.6 0.78
Manganese Oxide (MnO) 0.2-0.3 0.19
Diphosphorus Pentoxide (P2O5) 0.7-0.9 0.71

Figure 3.11 The composite of Martian regolith simulant sulfur concrete
42

Regolith simulants and sulfur were mixed together uniformly, and the mixture was extruded with
the parameters shown in Table 3.4.
Table 3.4 Materials and processing parameters

Sulfur
Concentration
(wt%)
Processing
Temperature
(℃)
Auger
Rotation
Speed
(RPM)
Raw Material
Density
(g/cm3)
Sulfur
Concrete
Density
(g/cm3)
Martian Regolith
Simulant (JSC
Mars-1A)
40 150 120 1.05 1.86
Lunar Regolith
Simulant (JSC-1A)
35 135 60 1.84 2.26
Lunar concrete was heated at 135 ℃, and the auger rotation speed was set to 60 rpm. Martian
concrete was heated at 150 ℃, and the auger rotation speed was set to 120 rpm. The relatively
slow auger speed for lunar regolith is due to the high friction of the simulant, and its lower
extrusion temperature is due to its increased time in contact with heating elements. The
temperature for Martian regolith, in contrast, is higher because the material moves faster through
the barrel.
Upon extrusion the nozzle is moved linearly at a constant speed to produce straight extrudate as
shown in Figure 3.12 (a) and (b).
For both materials, the auger experiences a changing frictional force, as noted earlier. Lunar
concrete has a significantly higher friction than Martian concrete, leading to generally better
quality for Martian samples.
Smoothly extruded layers with smooth surfaces and sharp edges were obtained with Martian
sulfur concrete (Figure 3.12 (a)). A rougher layer was obtained for lunar sulfur concrete (Figure
3.12 (b)) due to intermittent nozzle flow caused by the high friction. Short walls with multiple
43

layers were built with both materials (Figure 3.12 (c) and (d)). They are fairly strong, difficult to
break or separate layers by hand.

(a) Single layer of Martian sulfur concrete (b) Single layer of lunar sulfur concrete

(c)Multiple layers wall of Martian sulfur
concrete
(d) Multiple layers wall of lunar sulfur concrete
Figure 3.12 Martian regolith simulant & Lunar regolith simulant sulfur concrete
Due to the interactions among the auger surfaces, abrasive particles, and nozzle wall, the auger
experiences a changing friction. For the Martian simulant JSC-Mars1A, the powder is loosely
bonded together (Figure 3.13 (a)), and can be pulverized into fine grains by finger pressure. In
the case of the lunar simulant JSC-1A (Figure 3.13 (b)), the particles are hard and abrasive. This
makes JSC-1A have poor friction characteristics, for it flows into the gap between the auger
44

blade and the nozzle wall, slowing the rotation of the auger. If the lunar regolith could be ground
into a flour-like powder, the extrusion process might be smoother.

(a) JSC Mars-1A regolith simulant (×92) (b) JSC-1A Lunar regolith stimulant (×92)
Figure 3.13 The microstructure of martian and Lunar regolith simulant
The compression effect during extrusion is obvious in the case of JSC-Mars1A. The density of
JSC-Mars1A is 1.05g/cm
3
alone versus 1.86g/cm
3
for the extruded part, whereas for JSC-1A
these values are 1.84g/cm
3
and 2.26g/cm
3
, respectively. For a solid mixture without any
compression, the mixed density can be given as follows:
The calculated densities are 1.30g/cm
3
for the Martian mixture and 1.89g/cm
3
for the lunar
mixture. The measured densities for the extruded parts are 1.86g/cm
3
and 2.26g/cm
3
. The
compression ratios are therefore 1.42 and 1.19 for Martian concrete and lunar concrete,
respectively.
The surface smoothness of Martian sulfur concrete is better than that of lunar sulfur concrete due
to the better wettability of sulfur to JSC-Mars1A. The flour-like Martian simulant holds sulfur
tightly, and the melted paste does not flow freely as it does in the case of lunar concrete.
45

The material feeding rate must be proportional to the extrusion rate, for if the feeding rate is
higher than the extrusion rate, the nozzle will overflow; if the feeding rate is lower than the
extrusion rate, the product will be discontinuous. The ideal system would incorporate feedback
to properly match these two speeds.
As an initial forming experiment, parts were extruded with Martian regolith because of its
superior plasticity. A 3D model of an incomplete circular wall was built in SolidWorks (Figure
3.14 (a)). Next, based on the dimensions of the nozzle head, the model was divided into layers
with a thickness of 3/8 inch and converted into tool paths, as it is shown in Figure 3.14 (b).
Finally, using pre-heating and anti-bridging vibration, an incomplete circular wall (Figure 3.14
(c)) was built by the CC machine. Subsequently, a Martian regolith dome was constructed in
Figure 3.14 (d).

(a) 3D model of incomplete circle wall (b) Tool path

(c) Incomplete circle wall (d) Martian regolith dome
Figure 3.14 Martian regolith simulant sulfur concrete prinitng
46

3.4. Summary
In this chapter the possibility of applying Contour Crafting in space construction is discussed. To
take advantage of in-situ resources, sulfur concrete is a promising building material in the
process. Compared with Portland concrete, sulfur concrete has many benefits in energy-saving,
accessibility and post-processing. Martian regolith simulant JSC-Mar1A and Lunar regolith
simulant JSC-1A sulfur concretes have been implemented in Contour Crafting construction.
From the workability perspective, these two concrete types are disparate in their physical
composition.

47

4. Contour Crafting Equipment Development
4.1. Sulfur Concrete Extruder Design
Prior to this project the proven extrusion process for Contour Crafting was designed only for
processing several specific Portland cement based concrete types [10]. The system included
heavy mixers, massive concrete pumps, and air compressor. To apply Contour Crafting in the
environment of Mars and the Moon and using sulfur concrete, an entirely different extruder is
necessary considering the large temperature variation, vacuum, low gravity, dust effects and the
necessity for durability [5]. Since construction materials provided by in-situ resource utilization
(ISRU) may be different for different terrains, a new extrusion concept must be developed for
sulfur concrete with different types of constituents, viscosity and plasticity [67].

(a).Design of extrusion system (b) bridging effect
Figure 4.1 Extrusion principle
Generally, the extrusion nozzle is composed of two sub-stages: a compression stage and a
forming stage. Some options for granular material compression are shown above the dotted line
in Figure 4.1 (b). Compression was used to deliver the mixture of sulfur and regolith into the
heated nozzle, where it was melted and extruded. Figure 4.1 (a) shows four kinds of compression
48

methods, from left to right: the auger rotation method, the double-roller method (developed by
our team), the plunger method, and the preheated paste stirring/pressing method. In the auger
method, single or double augers rotate to move the mixture. In the double-roller method, the
knurled surface of the rollers captures and presses down the powder particles. In the plunger
method, a piston connected to a crank periodically pushes against the mixture. In the preheated
method a new flow generation mechanism in a heated hopper provides pressure to push the
molten mixture into the nozzle [18].
4.2. Double Roller Extruder
Method (b) in Figure 4.1 (a) shows the concept which has been developed by our team, for
which the CAD model is shown in Figure 4.2.

(a)The extrusion mechanism illustration (b) (c) Perspectives of extrusion system
Figure 4.2 Extrusion system of double-roller design
First, the mixture of aggregates and sulfur was placed in the chamber. Second, the knurled rollers
rotate toward each other (on top side) to press the mixture between the rollers and drive it into
the nozzle. Vibration was added by a separate motor with an off-center load to prevent the
bridging effect in Figure 4.1 (b).
49

As seen in Figure 4.3 the extruder system was fixed to a frame with a turntable mounted on it.
The nozzle was heated to roughly 130 ℃ (the melting point of sulfur) and the mixture was
extruded in the form of a viscous paste.

(a) Extrusion system (b) Extrusion process
Figure 4.3 Experimental setup of the double-roller extrusion system
In this part of the research we investigated the interaction of several variables in controlling the
density and hence strength of the resulting extrudate. Further experiments are warranted to find
the combination of parameters that uses the least portion of sulfur in the mix while producing the
greatest strength.
4.3. Mini-size Auger Extruder
In the mini-size auger extrusion system (Figure 4.4), the mixture was pushed downward to the
nozzle by a rotating auger while the nozzle body (the barrel) was heated by electric heating
elements. In this design at the bottom of the nozzle the nozzle head controls the flow of the paste
and creates planar surfaces and edges. To increase the inter-layer binding strength, a contour
profile (Figure 4.5) was machined on the nozzle outlet to create interlocking features [68]–[77].
50

Figure 4.4 Diagram of Extrusion System
Figure 4.5 Diagram of nozzle head and layer
binding
Extrusion of granular abrasive material is never straightforward. Lacking water, which serves as
lubricant to ease the concrete flow, the extrusion of the highly viscous and abrasive sulfur and
regolith mix is challenging, especially considering the highly abrasive lunar regolith particles
(Lacking of atmosphere, for billions of years’ lunar dust remains stagnant, unlike the Martian
condition in which frequent storms rub the dust particles together rounding off their sharp
edges.). To study the problem under NIAC Phase I project, a crude experimental device shown
in Figure 4.6 was created. In this device a gear-motor turns an auger that forces the dry mixture
of sand and sulfur out of a vibrating hopper and into the hot barrel and ultimately through the
nozzle-head, which shapes the extrudate. The auger was also equipped with a vibrator to prevent
potential clogging. As the mixture is pushed inside the hot barrel of the extrusion nozzle, the
51

sulfur portion heats up in the upper portion of the barrel and melts in the lower portion. The
extrusion process may take place without clogging only for a few seconds until the “bridging”
phenomenon causes a few particles of sand to create an arch against the flow direction and the
two arch bases push against the inner walls of the barrel causing complete clogging. Under such
a circumstance vibration could overcome the static friction and clear the clogging problem.
Vibration may be applied to the barrel or to the auger. Vibrating the barrel could be problematic
as the barrel is solidly attached to the rest of the structure for stability. Barrel vibration may also
eventually damage the electric heating filaments that heat the barrel to melt the sulfur; therefore,
it is best to vibrate the auger. Furthermore, it is best to minimize the vibration period only to
occasions of clogging because too much vibration causes segregation of sulfur powder at higher
sections of the barrel as well as sulfur melt in the lower sections.

Figure 4.6 Experimental machine for extrusion testing of sulfur concrete
A hard-wired controller that activates the auger vibrator only at clogging times had been devised.
This was done by sensing the surge in the gear motor current which increases as the motor
52

struggles to turn the auger during clogging, which significantly increases the motor load. The
more serious the clogging the higher the motor current becomes. A circuit activates the auger
vibrator with a vigor that is proportional to the gear motor current. At normal gear motor load
levels, the vibration motor remains inactive.
4.4. Mini-size Auger Extruder with Piezo Vibrators
In a later part of this project a new mini-size auger with piezo vibration was developed to
investigate the possibility of reliably extrusion sulfur concrete. As it is shown in Figure 4.7, dry
sulfur and Martian regolith simulant in the funnel were pushed downward through a nozzle by a
rotating auger while sulfur was being melted by electric heaters. The nozzle head formed the
shape of the exiting sulfur concrete and built smooth surfaces and edges. To increase the bonding
strength of the layer, a feature was added to the nozzle to produce an interlocking profile.

Figure 4.7 Schematic of Mini-size auger extruder
To automatically counteract the bridging effect during the extrusion process, a vertical vibration
was induced by a piezo transducer to the auger and a horizontal vibration piezo was added to the
53

nozzle-head, as shown in Figure 4.7. These two piezo actuators partially alleviated friction
during the extrusion process.

(a) Extruded sample with cracks on surface (b) Extruded sample with smooth surface
Figure 4.8 Pre-heating method on Martian regolith simulant sulfur concrete
However, the surface quality of extruded parts during this process was not always acceptable, as
some parts were smooth while others had many cracks. Considering this phenomenon, several
factors such as heat transfer, hardness of the nozzle’s inner wall (which impacts friction with
abrasive particles) and shape of the inner channel of the nozzle head were analyzed. After cutting
and polishing some failed samples (Figure 4.8 (a)) it was revealed that the color of the material
in the core of the layer was significantly different from the color of the peripheral regions, which
indicated that the mixture was not uniformly and completely melted in the nozzle during the
extrusion process. To improve the uniformity of heating, an experiment was carried out by pre-
heating a mixture of sulfur and Martian regolith simulant in water at 98℃ for half an hour. This
process, which resulted in dramatically improved part quality, as it is shown in Figure 4.8 (b). In
comparison with heating in room temperature and setting the heater at melting temperature of
sulfur at 115 ℃, pre-heating method further reduces the extrusion friction and improves the final
surface quality and density of extruded Martian sulfur concrete. Of course, use of water on Moon
54

and Mars is problematic and other approaches such as the use of microwave for preheating may
be considered.
Although the surface quality of the extruded layer has been improved, the rotation torque in this
configuration has been excessive. Also, auger blade wears out quickly when it is used for a few
hours, especially on the upper part of the nozzle. Figure 4.9 shows that after 50 hours of use the
auger blade diameter reduces by 6 mm due to excessive friction. The wear would be much higher
in case of lunar regolith due to significantly higher friction. Silicon dioxide, which has a
durability measuring between 4500~9500 MPa, makes half of the content of lunar regolith
stimulants.

Figure 4.9 Auger worn-out during experiment
4.5. A Novel Sulfur Concrete Contour Crafting Extruder
To ensure durability, stability and reliability of sulfur concrete extruder, a new universal concrete
extruder was conceived and examined. As shown in Figure 4.10, the new mixer/extruder
combination mechanism was devised in which the entire mixing and extrusion chambers were
heated to an accurate temperature. In contrast with the previous auger extruder the stage-wise
mechanisms on the upper sections move the material downward while a special extruder at the
end of the nozzle provides the main extrusion force. In addition, the mixture is completely
55

melted before entering the nozzle and as such its friction with the walls of the nozzle is far less
than the mixture at the ambient temperature or even pre-heated mixture. Besides these two
alterations, an aluminum extender end which also acts as heat sink is added to the outlet of the
nozzle to rapidly lower the temperature of the existing material. Fuzzy logic control, which has
been consistent and accurate in this set up, is employed to control the DC motor that mixes the
material and moves it downward through the nozzle. The maximum length of extrusion length in
this setup is limited by the volume of the nozzle barrel.

Figure 4.10 Schematic of multiple propellers extruder

(a) Linear test bench (b) Three layers’ extrusion sample
Figure 4.11 A novel sulfur concrete Contour Crafting extruder
56

To test the properties of the sulfur concrete extrusion process, a simple linear motion platform
was developed prior to using the 6-axis robot. As shown in Figure 4.11 (a), the linear motion
platform contains a linear rail system, a belt drive system and a control unit. Figure 4.11 (b)
shows a three-layer extrusion sample made by this platform.
4.6. The Final Contour Crafting Extruder
Several extrusion experiments have been carried out with the extruder and the influences of
temperature, sulfur proportion, extrusion speed and linear speed in different stages are
investigated for each experiment. However, the volume of the extrusion barrel has been a serious
limitation. To study the extrusion in a continuous process, the final edition extruder with a larger
reservoir and feeding system has been developed (Figure 4.12). For a large-scale planetary
construction, this reservoir is supposed to be filled intermittently by a supporting robotic system.

(a) Improved novel extruder with a reservoir (b) Temperature gradient
Figure 4.12 Final edition of CC extruder and its temperature gradient
57

Figure 4.13 shows the final Contour Crafting machine control system. A Programmable Logic
Controller (PLC), connected with an Industrial Personal Computer (IPC), controls all three sub-
systems, extruder, KUKA robot controller and process monitoring module. There are three
subcomponents under the extruder system. The heating system controls the operational
temperature gradient of the entire extruder in Figure 4.12 (b). A 400W flexible silicone-rubber
heat sheet is warped around the building material reservoir (Figure 4.12 (b) gray zone),
maintaining the building material at the melting temperature. Another 100W flexible silicone-
rubber heat sheet is warped around the extrusion cylinder (Figure 4.12 (b) black zone),
controlling the temperature of extrusion. Since the viscosity of sulfur dramatically increases at
160 ℃ (Figure 3.3), this temperature is set around 155 ℃ to minimize the flow friction from
inner wall of the cylinder. Also a cartridge heater is installed inside the nozzle outlet (Figure 4.12
(b) brown zone). Since higher viscosity sulfur concrete has better shapeability and finishability,
the temperature of this heater is set around 125℃ to lower the molten sulfur concrete temperature
at the end of the extrusion nozzle. The mixing system keeps running to make the building
material temperature uniform. There are two stages in the extrusion process. The first stage is
located in the cone between the reservoir and cylinder, which pushes the material downward into
the cylinder. The second stage is the main extruder near the bottom of the cylinder. A 6-axis
industrial KUKA robotic introduced in Section 4.8. The monitoring system contains a scanner
and a camera. The camera aims to record the extrusion process. The scanner dynamically scans
the shape of extrudate and sends the feedback to the PLC.
58

Figure 4.13 Contour Crafting control system
A batch process for feeding the nozzle assembly was developed on the basis of the above
approach. Figure 4.14 shows the new fabrication process concept where raw aggregate and dry
sulfur are transported into a hopper installed on top of the nozzle. The hopper heats and mixes
the material at 150 ℃. As the sulfur concrete mixture in the reservoir gets consumed, the robot
moves the extrusion system along with its reservoir under the silo outlet at a refilling station.

Figure 4.14 The construction process
•Melting
sulfur
•Heating
aggregate
Raw
material
preparation
KUKA
robot
Refilling
•Extruder
•KUKA
robot
Extrusion
Axis 3
IPC PLC
Extruder KRC
Heating
system
Mixing
system
Extrusion
system
Axis 1
Monitor
Camera
Scanner
Axis 2
Axis 4
Axis 5
Axis 6
Mixer
(melting
temp.)
Extruder
(155 ℃)
Outlet
(125 ℃)
Stage 1
Stage 2
59

4.7. Modification for Improving the Extrudate Surface Quality
For the extrudate built by the novel extrusion nozzle, the fluid sulfur of the upper layer always
sips down through the sand and runs down on the sides of the lower layer, which is hard to
remove after it get solidified. Meanwhile, the jaggy phenomenon also happens a lot if the nozzle
movement is too fast relative to the extrusion rate and if the sulfur concrete is not prepared to be
well to be uniformly mixed. The leaking problem and the jaggy phenomenon are shown in
Figure 4.15 (a) and (b).

(a) The sulfur sipping problem (b) The jaggy problem
Figure 4.15 The leaking and jaggy problems in the extrusion process
To prevent the discontinuous extrudate in the printing process, the nozzle extrusion flow speed is
set about such that the flow speed is 3% more than that of the nozzle movement speed. In this
case, however, a higher pressure in the nozzle is created which causes material leakage in the gap
between successive layers. To further improve the printing quality, two trowels are installed on
both sides of the nozzle outlet to shape the extrudate surface as shown in Figure 4.16. These
trowels are connected to the nozzle by elastic metallic sheets. So that, the previous cured sulfur
concrete extrudate won’t conflict with the moving nozzle during the construction process. To
prevent the redundant sulfur concrete causing expansion at the outlet of nozzle, a depressurizing
slot is designed at one side of the nozzle. The excess material can escape through this slot to
achieve a good extrudate surface quality.
60

Figure 4.16 The design of trowels on both sides of the nozzle
4.8. Utilization of KUKA 6-axis Robot
For the construction of full-scale structures, a large industrial robot is being used in the Contour
Crafting Laboratory. The robot (KUKA brand) consists of six rotational actuators which provide
six degrees of freedom. One of the most important advantages of using such robot over a gantry
robot is that for a given size of structure the gantry has to be designed in such a way that it can
contain the structure within its work envelope. However, a joint robot such as the KUKA robot
used in this experiment can be mounted on top of a rover to eliminate the size restriction.

(a) KR150 KUKA robot (b) CC robot on JPL rover
Figure 4.17 Framework of Contour Crafting for space application
Figure 4.17 (a) shows the KUKA robot and Figure 4.17 (b) shows the Contour Crafting robot
mounted on JPL rover. The rotational range of each actuator (A1 to A6) is limited which in turn
61

limits the total workspace of KUKA robot. Each layer in the construction must then be carefully
arranged so that it falls within KUKA’s workspace. The robot can handle up to a maximum force
of 1.5 KN exerted to its TCP in Table 4.1 (Tool Center Point).
Table 4.1 KUKA KR150L robot parameter
Voltage 480V
Capacity 150KG
Repeatability ±0.12mm
Horizontal Reach 2410mm
Vertical Reach 2700mm
Axis Robot Motion Range Robot Motion Speed
1-Axis ±185° 110 °/s
2-Axis +93 to -40 110 °/s
3-Axis +155° - 119° 110 °/s
4-Axis ±350° 170 °/s
5-Axis ±125° 170 °/s
6-Axis ±350° 238 °/s

Figure 4.18 Tool path of straight wall & half dome taken with long exposure camera
For construction of vertical walls linear motion is generated by the robot. The speed of the linear
motion is adjusted so that it is synchronized with the feeding rate that the reservoir can provide.
When the robot shifts to subsequent layer, the moving direction of end nozzle has to be reversed,
62

which is executed by using the KUKA’s ending flange rotational degree of freedom. Figure 4.18
shows the KUKA’s TCP path for vertical wall and dome constructions.
Building 3D dome structures requires following more complicated pathways than straight or
circular lines. For a stationary robot (not mounted on a rover) the position of starting layer in the
global coordinate system must be selected in such a way that the whole dome structure falls
completely within the robot workspace. Different shapes and geometries can be selected for the
dome-like structures. One of the candidates for the 3D dome structure is depicted in Figure 4.19.
For this structure inclined layers are built consecutively such that each layer has a slight angle
with respect to the horizontal plane.

Figure 4.19 Tool path coordinates for dome-like structure
Other patterns to achieve dome-like structures are also being considered. For instance, patterns
which inherit geometric advantages of domes and cones can be used to build optimum structures
by a horizontal, rather than inclined, layering method. We plan to investigate relevant merits of
different geometric structures in order to determine the optimum shape such that each layer has
sufficient support and the whole structure shows robust stability and strength.
63

4.9. Sulfur Concrete Deformation Test Bed Design
The performance of sulfur concrete is sensitive to the characteristics of its ingredients and the
variables of the thermal process used for preparing the material. Different proportions of
ingredients would cause dramatically different performances in the extrusion process. To test the
deformation phenomenon in the extrusion process, an experiment was designed to study the
factors that influence the performance of sulfur concrete. Of course the research result is
applicable for improving the performance of sulfur concrete for terrestrial applications as well.
To precisely simulate the extrusion process, an extrusion test bed was developed (Figure 4.20).
This test bed is composed of a square nozzle, a linear actuator and a linear railway. The square
nozzle is heated up by two adhesive flexible heaters and the temperature is adjusted by a PID
temperature controller. After the experimental sulfur concrete is filled into the nozzle through a
funnel, a linear actuator pushes a piston downward against the material at a constant speed.
Meanwhile, the entire nozzle is moved on a straight rail using a stepper motor and a timing belt.
A camera is also set in the front of this test bed to monitor the deformation phenomenon near the
nozzle outlet.
Using this test bed an optimized sulfur concrete parameter can be acquired with much less
material consumption. The experiment data and analysis are presented in Chapter 6.
64

Figure 4.20 Sulfur concrete deformation test bed model
4.10. Summary
This chapter described the development of Contour Crafting extruders for sulfur concrete
application. Four extrusion methods are studied and more than eight extruders have been
implemented to fit the workability and properties of sulfur concrete. A 6-axis robotic system is
employed to achieve complex construction tool paths. Later, a simplified extrusion test bed is
developed to better study the sulfur concrete workability with less building material
consumption.

Stepper motor
Platform
Frame
Linear actuator
Nozzle outlet
Nozzle
Linear railway
65

5. Research on the Workability of Sulfur Concrete
5.1. Challenges in the SCCC Approach
Due to the special condition of sulfur concrete, each of the earlier extrusion methods described
above was problematic one way or another. The double roller extruder, auger extruder, and
vibration valve extruder, as well as some combinations of these were all tested in the early stage.
As described in section 3, none of them worked with sulfur concrete without some drawbacks.
A new novel extrusion concept called Compliant Pressure Flow (CPF) was conceived by Dr.
Khoshnevis which can provide continuous and smooth flow for sulfur concrete without exerting
excessive pressure on any part of the assembly. Accordingly, component wear is not a problem
in the new design. Our major large-scale extrusion experiments have been successfully carried
out using the new extrusion system.
Workability is the most important property of molten sulfur concrete, which not only affects the
transportation and extrusion but also the strength and surface quality of the final product. Bartos
explained workability with several fundamental characteristics such as viscosity, mobility,
internal friction, pumpability, segregation, bleeding, formability and finishability [53].
In Contour Crafting with sulfur concrete, mobility and pumpability of molten sulfur concrete
affect the extrudability of the mix. While, the formability and finishability affect the strength and
surface quality of extruded structures. Unfortunately, the molten sulfur concrete with high
mobility and pumpability always has poor formability and finished quality and this is the
difficult contradiction in this project. So, exploring the appropriate sulfur concrete for multiple
properties and developing effective nozzle assembly and control regimen for controllably
forming sulfur concrete are the two main tasks of this project [18], [78]–[85].
66

In the sulfur concrete Contour Crafting process many factors may influence the workability of
molten sulfur concrete. The following Table 5.1 lists the influence factors that have been found
and the current parameters.
Table 5.1 Most efficient factors of SCCC
Name Parameters Impact
Proportion of modified sulfur 35% High
Proportion of ash 10% High
Proportion of aggregate 55% High
Melting temperature 145 C High
Melting time 2 hours High
Temperature at reservoir 130 C Medium
Temperature at the nozzle 160 C High
Extrusion speed 100 RPM High
Linear speed 10mm/s High
Mixing time 30 seconds Medium
Ambient temperature 25 C Low
Identifying the proper values of controllable parameters in the above list and creating an
effective control system are the main challenges of this project.
It should be noted that because of the deformation caused by less viscous sulfur concrete, the
width of the extruded layer is generally larger than the outlet of nozzle. Accordingly, these
inaccuracies in layer width may cause non-smooth structure surfaces. If the sulfur concrete is too
viscous, the extruder can’t pump it fluently. On the other hand, the low viscosity sulfur concrete
could cause deformation and possibly collapse for tall structures. Predicting and controlling
deformation at an acceptable tolerance is another challenge of this research project [86]–[100].
5.2. Extrudate Slump Analysis
Molten sulfur concrete is similar to fresh hydraulic concrete and can be approximately analyzed
using Bingham model [84]. In this model, molten sulfur concrete that shear stress is less than
67

yield stress are able to stand without flowing under its own mass. The one-dimension law of
Bingham model is:
τ = μ
𝑝 γ ̇ + τ
0
(1)
where τ Yield stress
μ Plastic viscosity
𝛾 ̇ Shear stress

(a) Initial shape (b) Final shape (c) Initial stress distribution (d) Final stress distribution
Figure 5.1 Deformation of slump process
The molten sulfur concrete can keep the shape without flowing and support the material above
until the shear stress is larger than the yield stress. To study the deformation of extrusion, a two-
dimension model is built [78]. Figure 5.1 (a) shows the cross section shape of extrusion. At the
height of z, the pressure p(z) can be calculated as,
𝑝 (𝑧 ) =
𝐹 𝑆 =
𝜌𝑔𝑧𝐷 𝐷 = 𝜌𝑔𝑧 (2)
Where H Extrusion height
z Axial position
68

ρ Material density
g Gravity
D Width of extrusion
Assume the maximum shear stress at the position of z is half of the pressure. Then, the maximum
shear stress τ(z) can be calculated as,
τ(z) =
1
2
𝜌𝑔𝑧 (3)
The shear stress distribution of extrusion before deformation can be illustrate as Figure 5.1 (c).
The material on top of the boundary z0 where the shear stress value is less than the yield value
keeps the same shape and remains as a rigid body. While, the material below this point yields as
shown in Figure 5.1 (b). And, Figure 5.1 (d) shows the shear stress distribution after the
slumping process is complete [79].
Since the top material remains undeformed in the slumping process, only the material underneath
is discussed in the following. In the slump process, the material under the boundary flows until
the shear stress reduce to the yield value.
Assumption:
 The material above the boundary keeps the same shape and remains as a rigid body.
 The boundary of yield and un-yield keeps a flat surface and sink down as the material
underneath flows.
 The yield material flow outward horizontally along the x axis.
 The volume of the yield material stays unchanged.
69

After the deformation, the initial height of unyeild material h0 remains the same, but the initial
height of yield material decreases to the final height h1. This change is illustrated in Figure 5.1
(d) [82].
The final height of yield material h1 can be calculated by intergrading dz1 as
ℎ
1
= ∫ 𝑑𝑧 1
𝐻 ℎ
0
(4)

(a) An element before deformation (b) The element after deformation
Figure 5.2 Deformation process of an element
Since the volume of the yield material stay unchanged, the later thinness of dz1 in Figure 5.2 can
be calculated as,
𝑑𝑧 1
=
𝑥 (𝑧 )
𝑥 (𝑧 1
)
𝑑𝑧 (5)
When the slump is completed, the shear stress value of yield material become identical S yield [80].
Since the weight of material above the x(z) remains the same, the relationship of original element
thickness x(z) and the deformed element thickness x(z1) can be calculated as,
𝜏 (𝑧 )𝑥 (𝑧 ) = 𝜏 (𝑦𝑖𝑒𝑙𝑑 )𝑥 (𝑧 1
) (6)
Substitute equation 5 and equation 6 into equation 4:
70

ℎ
1
= ∫
𝑆 yield
𝜏 (𝑧 )
𝐻 ℎ
0
𝑑𝑧 (7)
Then, substitute equation 3 into equation 7:
ℎ
1
= 2 ∫
𝑆 yield
𝜌𝑔𝑧 𝐻 ℎ
0
𝑑𝑧 (8)

Last, calculate this integration and get:
ℎ
1
=
2
𝜌𝑔
𝑙𝑛 (
𝐻 ℎ
0
) (9)
The slump value can be calculated by:
𝑆 = 𝐻 − ℎ
0
− ℎ
1
= H −
2
𝜌𝑔
[𝑆 yield
+ 𝑙𝑛 (
𝐻 ℎ
0
)] (10)
5.3. Finite Element Analysis *
A finite element analysis for the deformation behavior of sulfur concrete immediately after
extrusion was conducted by using the finite element code. The details of analysis are given
below.
Mohr Coulomb Plasticity model, provided within the material library of ABAQUS, was selected
for the description of the rheological behavior of uncured sulfur concrete right after
extrusion. According to Mohr-Coulomb model, failure in the element begins when the shear
stress in a material reaches a threshold that depends linearly on the normal stress in the same
plane. By plotting Mohr's circle for states of stress at failure in the plane of the maximum and
minimum principal stresses, the failure line is defined as the best straight line that touches these
Mohr's circles as shown in Figure 5.3 [101].
1. This section is entirely the contribution of Behnam Zahiri, who was a participating graduation
research assistant in this project.
71

Figure 5.3 Mohr-Coulomb failure model [101]
According to Figure 5.3, the Mohr-Coulomb model is defined by [101]
𝜏 = 𝑐 − 𝜎 tan 𝜙 (11)
Where 𝜎 would have a positive sign in tension and a negative sign in compression. In Equation
(11) 𝜙 is the friction angle and 𝑐 is the cohesion of the material. With some manipulation in
Equation (11), the Mohr-Coulomb model can be written as
𝑠 + 𝜎 𝑚 si n 𝜙 − 𝑐 cos𝜙 = 0 (12)
This model can also be written in the general form as:
𝐹 = 𝑅 𝑚𝑐
𝑞 − 𝑝 tan 𝜙 − 𝑐 = 0 (13)
Where
𝑅 𝑚𝑐
(Θ, 𝜙 ) =
1
√3 cos 𝜙 sin(Θ +
𝜋 3
) +
1
3
cos(Θ +
𝜋 3
) tan 𝜙 (14)

The essential components of the Mohr-Coulomb model are described below:
72

 Friction angle: In Equation (11-14), 𝜙 is the friction angle which is the slope of the
Mohr-Coulomb yield surface in the 𝑝 _𝑅 𝑚𝑐
𝑞 stress plane (see Figure 5.3), which is
commonly referred to as the friction angle of the material and can depend on temperature
and predefined field variables.
 Dilation angle measured in the 𝑝 _𝑅 𝑚𝑐
𝑞 plane at high confining pressure and can depend
on temperature and predefined field variables.
 Cohesion: In Equation (11-14) 𝑐 is the cohesion of the material.
The geometry of the yield surface in the deviatoric plane is controlled by the friction angle 𝜙 as
shown in Figure 5.4. The range for friction angle variation is from 0 ≤ 𝜙 < 90. In the case of
𝜙 = 0, the Mohr-Coulomb model reduces to the pressure-independent Tresca model with a
perfectly hexagonal deviatoric section. In the case of 𝜙 = 90 the Mohr-Coulomb model reduces
to the “tension cut-off” Rankine model with a triangular deviatoric section and 𝑅 𝑚𝑐
= ∞ [102].

Figure 5.4 Mohr-Coulomb yield surface in meridional and deviatoric planes [101]
The geometry of the specimen is shown in Figure 5.4. Only one half of the specimen could also
be modeled for the analysis to get the results since the deformation analysis of the concrete has
73

axisymmetric nature. The specimen was meshed using four-node axisymmetric elements. The
surface beneath the specimen was modeled as a rigid body with friction interaction with the
model.

Figure 5.5 The geometry of the specimen and the initial mesh used to implement the FEA
The relative geometrical deformation from the FEA model was obtained and compared to the
experimental results (Figure 5.5). By changing the viscosity in each experiment and using an
iterative recursive curve fitting method, we are able to obtain an FEA model for uncured sulfur
concrete which can predict the deformation of the extruded mixture as a function of viscosity.

Figure 5.6 The Mohr-Coulomb based FEA model and the extruded sample 35%wt
74

Figure 5.7 The Mohr-Coulomb based FEA model and the extruded sample 30%wt
Figure 5.6 and Figure 5.7 show the FEA models versus the experimental samples. The essential
parameters in the Mohr-Coulomb model is different for the two samples with different sulfur
contents. The appropriate friction coefficients between the samples and the base were found by
comparing the experimental results with the FEA results so that the friction coefficients were in
the best agreement between the predicted and experimental deformation patterns.
5.4. Research on Nozzle Design
Since the design of sulfur concrete extrusion nozzle has been changed for the new material, it is
necessary to study the flow of sulfur concrete during the extrusion process. The previous nozzles
which are used for gypsum, plastics and hydraulic concrete are shown in Figure 5.8.
In Figure 5.8 (a), the “I” type nozzle is mostly used to extrude materials, like molten plastic and
gypsum, which are smooth and viscous. In Figure 5.8 (b), the “J” nozzle is developed for single-
track concrete which contains both fine cement and coarse aggregates. In Figure 5.8 (c), the “Y”
nozzle is optimized based on the “I” nozzle and is used to build parallel-track concrete. For all
these three designs, the nozzles provide high pressure on the building material by different types
of pumps. Therefore, these mechanical structures would use significant amount of energy and
could wear out fast in presence of hard aggregates. In contrast, the novel design of extruder for
75

sulfur concrete can provide continuous lower pressure to the building material at the end of the
nozzle which would last a much longer time during construction.

(a) “I” nozzle for gypsum and plastic (b) “J” nozzle for hydraulic concrete

(c) “Y” nozzle for hydraulic concrete (d) “L” nozzle for sulfur concrete
Figure 5.8 The previous nozzles designed for plastic, gypsum and hydraulic concrete
To compare the extrusion processes, the situation of material flow in the nozzle is simulated. The
analysis type is internal and the gravity is considered [78]–[80], [82], [84], [101], [103], [104].
By setting the flow velocity at the inlet of each nozzle the 0.05kg/s and the outlet of nozzle air
pressure. The flow path in the nozzle, the variation of flow velocity and the variation of pressure
in the nozzle are compared.
Figure 5.9 shows the simulated flow path in the nozzle. Figure 5.9 (a) and (b) show the material
flow paths in the “I” and “J” nozzle. It’s obvious that the difference of nozzle design doesn’t
76

influence the material flow paths much and material flow paths in these nozzles are very similar
to each other. In Figure 5.9 (c), the “Y” nozzle divides the stream of material into two channels.
However, the material flow paths in both division channels are quite similar to that in the “I”
nozzle. In Figure 5.9 (d), the material flows from a cylinder channel to a cuboid channel and the
flow paths change dramatically. Since the area of inlet section is larger than that of outlet section,
the flow speed increases at the end of nozzle outlet.
Figure 5.10 shows the cut plot of flow velocity distributions in the nozzle. Figure 5.10 (a) and (b)
show that the material flow velocity distributions in the “I” nozzle and “J” nozzle are very
similar to each other. The blue region in Figure 5.10 (a) shows that the material is almost
stagnant in this zone. So, the curving design at the end of nozzle in Figure 5.10 (b) can increase
the flow ability of material. The three cut plots in Figure 5.10 (c) are used to illustrate the
velocity distribution. After the stream is divided by the branch, it is pushed all the way to the
different outlets. The velocity distribution is similar to that in Figure 5.10 (a). In Figure 5.10 (d),
the area of inlet is larger than that of the outlet, the flow velocity at the end of the nozzle is much
higher than these of the other three nozzles. Moreover, the flow velocity at the bottom of the
nozzle outlet is higher than that at the top of the nozzle outlet, which is also different from those
of the previous three nozzles.
Figure 5.11 shows the pressure distributions in the nozzles during the extrusion process.
Generally, the pressure gradually decreases along the flow direction of material. However, the
decreasing speed in Figure 5.11 (d) is slower than that of others due to the change of direction in
the nozzle.

77

(a) the flow path of “I” nozzle (b) the flow path of “J” nozzle

(c) the flow path of “Y” nozzle (d) the flow path of “L” nozzle
Figure 5.9 The flow path in the nozzles
78

(a) the velocity cut plot of “I” nozzle (b) the velocity cut plot of “J” nozzle

(c) the velocity cut plot of “Y” nozzle (d) the velocity cut plot of “L” nozzle
Figure 5.10 The flow velocity distribution in the nozzles
79

(a) the pressure surface plot of “I” nozzle (b) the pressure surface plot of “J” nozzle

(c) the pressure surface plot of “Y” nozzle (d) the pressure surface plot of “L” nozzle
Figure 5.11 The pressure distribution in the nozzle
5.5. Sulfur Holding Capacity
The sulfur holding capacity is described as the amount of liquid sulfur that sand and ash can
retain in the sulfur concrete extrusion process. The sulfur holding capacity directly influences the
80

workability of molten sulfur concrete and the strength of cured structure. Similar studies has
been conducted about the water holding capacity of sand [97], [99], [105]–[114].
Table 5.2 Microstructure of sulfur concrete [106]

Dry grains – cohesion negligible

Partially saturated – at small volume fractions, liquid
bridges are formed between grains near points of
contact. Liquid bridges induce cohesion between
grains.

At higher volume fractions, liquid bridges merge to
given trimers, tetrahedral and pentamers.

At still higher volume fractions, large contiguous wet
cluster form

Slurry – the pore space is fully saturated with liquid.
Cohesion becomes negligible again.
In Table 5.2, the yellow spheres represent the aggregates and the blue region represents the
liquid sulfur. When the mixture is partially saturated, liquid bridges are formed between
aggregates near the points of contact. With the increase of liquid sulfur, liquid bridges merge to
given trimer, tetrahedral and pentamers [106] until the mixture becomes saturated. When the
pore space is filled with liquid sulfur, cohesion of grains becomes negligible again and liquid
sulfur leaking starts to appear.
81

Figure 5.12 Sulfur holding capacity experiment
In previous experiments, sand could not contain and arrest enough liquid sulfur due to the
insufficient surface tension of liquid sulfur. When the well mixed sulfur concrete is filled into the
extruder, the liquid sulfur always flows down to the bottom and the top portion of the mix
becomes quite viscous. As a way of arresting the liquid sulfur, ash is used in this process to make
the sulfur distribution more uniform in the entire volume of the mix. To test the liquid sulfur
holding capability of ash and sand, a comparison experiment is carried out.
In this experiment (Figure 5.12), different proportion of aggregates were mixed with excessive
molten sulfur. These mixtures were then loaded into aluminum cylinders with fine nets in the
bottom and put into a furnace (145℃) for more than 3 hours until the saturated sulfur exudes. At
82

last, the sulfur holding capacity of different materials was calculated by scaling the weight loss.
The result in Table 5.3 shows that finer aggregate can track more liquid sulfur.
Table 5.3 Sulfur holding capacity of different mixture
Group component Sulfur holding capacity
#1 Sand (250~500um) 34.6%
#2 Sand (0~250um) 39.8 %
#3 Ash (0~250um) 41.3%
5.6. Research on Sulfur Concrete Extrusion Process
The performance of the extrusion process varies considerably with different proportions of sulfur
in sulfur concrete. When sulfur proportion in the mix is high the sulfur concrete has low
viscosity, which makes it easy to extrude. However, low viscosity also causes the problem of
having high slump and leakage. On the other hand, low proportion of sulfur lowers the slump,
but a low viscosity mixture is hard to extrude as it always sticks to extrusion barrel and internal
mechanisms.
To explore the significant factors that may influence the extrusion process a 2k factorial
experiment was designed based on temperature at different stages, proportion of sulfur and
extrusion/moving speed as shown in Table 5.4.
Table 5.4 A 2k factorial design on extrusion process
Factors Level - Level +
Sulfur Proportion 30 wt% 35 wt%
Sulfur Concrete Temperature 140 C

150 C
Extruder Temperature 150 C 160 C
Extrusion Rate 3 kg/min 5 kg/min
When the sulfur concrete does not stick to the inner wall of the barrel and with the force of
gravity flows smoothly to the lower sections of the nozzle, the experiment shows that the
extruder temperature and extrusion rate have little effect on the extrusion process.
83

(a) Sample built in 140C (b) Sample built in 150C
Figure 5.13 Extrusion process with different sulfur cement temperature
In the experiment it is obvious that the mixture at 150C has less viscosity and better flow rate.
Some extrusion cross-sections are observed under optical microscope, as shown in Figure 5.13.
The porous structure is more equally distributed in the sample built at 150 C as shown in Figure
5.14 (b), while most of porous structure is shown in Figure 5.14 (a) as sample built in 140C. It is
deduced that sulfur in mixture at 150℃ took longer for the curing process and sank into the
lower region of parts by the force of gravity. The uneven sulfur content between upper and lower
parts may considerably reduce the strength of multi-layer structures.

(a) Sample built in 140C (b) Sample built in 150C
Figure 5.14 Micro-structure of extrudate with different sulfur cement temperature
84

(a) Sample with 30wt% sulfur (b) Sample with 35wt% sulfur
Figure 5.15 Extrusion process with different sulfur proportion

(a) Sample with 30wt% sulfur (b) Sample with 35wt% sulfur
Figure 5.16 Micro-structure of extrudate with different sulfur proportion
Sulfur plays the role of binder in solidified sulfur concrete and the role of lubricant during the
extrusion process. By increasing sulfur proportion, the viscosity of sulfur cement decreases
considerably, therefore, the liquidity increases. Compared to the 30 wt% sulfur concrete samples,
the 35 wt% sulfur concrete is saturated with sulfur and hence the porosity level is not significant.
However, the plasticity of 35 wt% sulfur cement in the molten state is less than that of 30 wt%
sulfur concrete. This may explain the reason behind the round shapes of both edges and top
surface of the sample in Figure 5.15 (b). In such cases it is especially necessary to add side
trowels behind the nozzle to keep the extruded surface smooth.
85

From Figure 5.16 it is clear that the microscopic characteristics of the two samples are different.
The size of microscopic holes in the two samples differs greatly. Porosity directly impacts the
strength of the structures made with these materials. By choosing the appropriate weight ratio for
sulfur in the mixture, the porosity can be decreased.
5.7. Summary
In this chapter, the workability dilemma of sulfur concrete created by mobility/pumpability and
formability/finishability is discussed. A balanced configuration of these two abilities is the key
aspect of working extrusion process. The theoretical and FEA deformation models are created to
study and predict the formability/finishability of molten sulfur concrete. The influence of nozzle
design is studied by simulating the flow of slurry in these four types of nozzle. In the sulfur
holding capacity experiment, fine aggregate, such as dust and ash, can keep more sulfur and
perform as lubricant in the extrusion. While, the coarse aggregate improves the strength of final
part. In the sulfur concrete extrusion process, the influence of temperature and sulfur proportion
on the strength and porosity of final part is studied. However, further experiments still need to be
carried out to study this influence.

86

6. A Design of Experiments for Sulfur Concrete Deformation
As described in Section 5.1, the challenge in the SCCC approach is that more than 10 factors
influence the final extrudate in many ways. To obtain an acceptable set of processing parameters,
the Monothetic method is applied in this experiment. As it is shown in Figure 6.1 (a), a set of
reference parameters is chosen based on the bibliography of sulfur concrete. The approach is to
change-one-factor at a time and compare the response with the previous results. If a better result
is obtained, renew the reference parameter and repeat the previous steps again until the final
result meets the target.
Figure 6.1 Monothetic experimental method
1. Choose parameters
3. Change one out of previous parameters
5. Compare
with
6. Repeat 3~5 until
meets the target
0. Start
2. Measure the defordddmation
4. Measure the deformation
7. END
(a) Monothetic method
(b) Samples built by Monothetic method
87

Although some acceptable extruded parts were obtained by using this method, none of the
experiments which led to these parts has been repeatable due to many external factors. Moreover,
this method is inefficient (Figure 6.1 (b)) and it ignores the interactions between different
parameters. To study the influence of factors as well as the interactions of these factors, a
factorial experiment is expected.
In this chapter, both horizontal and vertical deformation is studied. When the molten sulfur
concrete gets out of the nozzle outlet, gravitational force deforms the extruded part. As it is
shown in Figure 6.2, the dotted line rectangle represents the dimension of nozzle outlet, which is
2.546 × 0.99 inch. The horizontal deformation and vertical deformation are defined as:
horizontal deformation = horizontal width – nozzle outlet width
vertical deformation = vertical height – nozzle outlet height

Figure 6.2 Horizontal and vertical deformation definition
6.1. Fractional Factorial Experiment Design
Factorial experiment design is commonly used for the experiment with two or more factors. This
experiment design method not only concentrates on the effects of these factors, but also analyzes
the interaction effects between these factors. In a factorial experiment design at two levels, 2 ×2
×…×2=2
k
(k is the number of effects) observations are required to achieve the balance and
Horizontal width
Vertical height
Vertical
deformation
88

orthogonality. In a 2
k
design, each factor lever appears in the same number of run [115]. As it is
shown in Figure 6.3, each vertex of the cube represents a run of a 2
3
design.

Figure 6.3 Cuboidal representation of a 2
3
design
The 2
k
full factorial experimental design is seldom used for economic reasons, especially when
the number of factors is larger than 7, since it requires to run the experiment 2
k
times.
Accordingly, a fractional factorial design using only 2
k-p
of experiments, which only contain
subdivisions of the full factorial experimental design, is usually chosen to solve this large-scale
experimental design problem. Considering a fractional factorial experimental design to improve
the extrudate surface quality and deformation, the 5 high impact factors are listed in the
following Table 6.1.
Table 6.1 Factors and levels of sulfur concrete extrudate experiment
Factors
Level
+ -
A. Proportion of Ash 20% 10%
B. Melting Temperature (C) 155 140
C. Proportion of Sulfur (g) 35% 30%
D. Melting Time (hours) 2 1
The objective of this experiment is to minimize the deformation in both horizontal and vertical
direction. In other words, it is desirable that both horizontal deformation and vertical
deformation be as close to 0 as possible. The result of Monothetic method archives the horizontal
89

deformation of 0.4 in and the vertical deformation of 0.1in. Thus, this experiment is expected to
figure out the factors that can be set to minimize the horizontal/vertical deformation non-
uniformity, as well as keep the deformation as close to the nominal value as possible [115].
These factors are chosen based on the previous Monothetic experiment results:
Table 6.2 Sulfur concrete deformation experiment planning matrix
No.
Proportion of
Ash (g)
A
Sulfur
Concrete
Temp. (C)
B
Proportion of
Sulfur (g)
C
Melting
Time
(hours)
D=ABC
Proportion of
Bitumen (g)
Sand
Weight
(g)
1 400 140 1200 1 30 2400
2 800 140 1200 2 30 2000
3 400 155 1200 2 30 2400
4 800 155 1200 1 30 2000
5 400 140 1400 2 35 2200
6 800 140 1400 1 35 1800
7 400 155 1400 1 35 2200
8 800 155 1400 2 35 1800
A: The proportion of ash by weight. B: Melting temperature, includes the temperature of the
nozzle and the molten sulfur concrete. C: Proportion of sulfur by weight. D: Melting time,
referring to the time of sulfur concrete staying in the melting furnace after the bitumen modifier
is added. Proportion of aggregate, equal to the total weight minus the weight of ash and sulfur.
This experiment is considered as a 2
4-1
design, where 4 is the number of factors in research and
2
-1
is the fraction of this experiment. Referring to the eight-run 2
Ⅳ
𝑘 −𝑝 fractional factorial design,
the fraction and resolution are chosen as 2
Ⅳ
4−1
and the design generator is D=ABC, which means
that the main effect of D is also used to estimate the interaction effect between A, B and C [115].
D = ABC or I = ABCD
The experiment matrix is listed Table 6.2. These experiments are taken out randomly to
eliminate the influence of the environment.
90

In the process, the mixture of “Sand” and “Ash” is heated up to the “Melting Temperature” and
then mixed with the “Sulfur” properly. Then, the mixture is placed in the furnace until it is
completely melted. Later, the modification “Bitumen” is added into the sulfur concrete and it is
placed in the furnace again for the “melting time”. Finally, the modified sulfur concrete can be
filled into the nozzle to pursue the experiment. Two replicates are processed and the average is
chosen. The modified sulfur concrete preparation process is shown in Figure 6.4.

Figure 6.4 The test process of sulfur concrete deformation
The extrusion test is shown in Figure 6.5 (a) and the extrudate samples are shown in Figure 6.5
(b). To analyze the extrusion process, the width and height of samples are the two obvious
quantifiable factors that can reflect the extrudate deformation. Random measuring points are
chosen and the primary data are shown in Table 6.3.

(a) Deformation extrusion experiment (b) Fractional factorial samples
Figure 6.5 The deformation extrusion experiment with fractional factorial design
•Mix sand and
ash
•Heat up to
160C
•Mix with dry
sulfur
Sulfur
concrete
preparing
•Heat until
sulfur melted
•Mix with liquid
bitumen
•Keep in
furnace for
hours
Sulfur
concrete
modifying
•Mix up to 1
minute
•Load into
nozzle
Extrusion
experiment
91

Table 6.3 Fractional factorial experiment design deformation data
Group
#
Horizontal Vertical
Width
1.1
Width
1.2
Width
1.3
Width
2.1
Width
2.2
Width
2.3
Height
1.1
Height
1.2
Height
1.3
Height
2.1
Height
2.2
Height
2.3
1 2.561 2.544 2.549 2.543 2.537 2.549 1.006 1.006 0.999 1.014 1.004 0.998
2 2.549 2.549 2.546 2.568 2.557 2.566 1.006 1.01 1.014 1.007 1.002 1.007
3 2.536 2.55 2.546 2.538 2.541 2.545 1.032 1.000 1.017 1.017 1.021 1.006
4 2.519 2.534 2.531 2.533 2.537 2.576 1.000 1.012 1.001 1.015 1.015 1.013
5 2.658 2.666 2.654 2.62 2.659 2.707 1.008 1.019 1.014 1.009 1.02 1.022
6 3.149 3.205 3.193 3.129 3.162 3.141 1.099 1.111 1.134 1.064 1.111 1.129
7 2.638 2.67 2.716 2.671 2.723 2.737 1.017 1.031 1.036 1.007 1.032 1.026
8 3.049 3.071 3.035 3.06 3.043 3.066 1.04 1.042 1.06 1.044 1.058 1.06

92

6.2. Experiment Analysis
In this fractional factorial experiment at two levels, the purpose is to figure out the main effects
and the combination of factors levels for A, B, C, D to achieve the minimum deformation in both
horizontal and vertical directions. The horizontal deformation experiment dataset is listed in the
following (Table 6.4) by mean and log variance.
Table 6.4 Fractional factorial experiment horizontal deformation dataset
Ash
proportion
Melting
temperature
Sulfur
proportion
Melting
time Y bar Log (Si2)
-1 -1 -1 -1 0.001166666 3.47E-05
1 -1 -1 1 0.009833333 1.23E-04
-1 1 -1 1 0.003333334 3.56E-06
1 1 -1 -1 0.007666666 2.14E-04
-1 -1 1 1 0.114666667 3.56E-06
1 -1 1 -1 0.617166666 7.35E-04
-1 1 1 -1 0.1465 6.36E-04
1 1 1 1 0.508 1.09E-05
A regression analysis is conducted on this data using R software, to obtain half-normal plots for
location and dispersion. Half-normal plots are used for testing the significant effects [115].

Figure 6.6 Half-normal distribution of location model in horizontal deformation

93

Figure 6.7 Half-normal distribution of dispersion model in horizontal deformation
As the half-normal plot for the location model plotted in Figure 6.6 shows, the sulfur proportion
is the most significant factor in the process. In Figure 6.5(b), the upper eight samples are built
with 30% sulfur by weight, the lower eight ones are built with 35% sulfur by weight. It is evident
that the middle part of each upper one is discontinuous, which is caused by the workability
dilemma of sulfur concrete discussed in Section 5.1.
As the half-normal plot for the dispersion model plotted in Figure 6.7 shows, the melting time,
which also represents the interaction of three other factors, is a significant factor for variation.

Figure 6.8 Half-normal distribution of location model in vertical deformation

94

Figure 6.9 Half-normal distribution of dispersion model in vertical deformation
Similarly, a regression analysis is conducted on the vertical deformation data using R software to
obtain a half-normal plots for location and dispersion. Neither of factors identified significant in
both half-normal distribution plots.
In this fractional factorial experiment design, the discontinuous samples built at the sulfur
proportion “-” level are unacceptable. To eliminate the influence of this factor, an additional full
fractional experiment is carried out only with ash proportion, melting temperature and melting
time factors.
6.3. Additional Full Factorial Experiment
Based on the previous result an additional full factorial experiment is designed to study the
influence of these factors. Similar to the previous fractional factorial experiment design, the new
experiment matrix is shown in Table 6.5. To assure the integrity of samples, all the following
experiments contain 35% sulfur by weight.

95

Table 6.5 Factors and levels, sulfur concrete extrudate experiment
Factors
Level
+ -
A. Proportion of Ash 20% 10%
B. Melting Temperature (C) 155 140
C. Melting Time (hours) 2 1
To assure the same experimental environment, an additional experiment is carried out by the
same process described in Section 6.1. Each run of this experiment has been carried out
randomly to reduce the influence of unwanted effects that are not planned in this matrix. For
example, the room temperature is a mid-impact unwanted factor. The temperature at noon time is
higher than in the evening. So, molten sulfur concrete gets solidified faster and deforms less in
the evening. Additionally, the melting temperature in this experiment is a hard to control factor,
since the temperature takes a long time to stabilize [115]. Therefore, the “+” level of melting
temperature is within the limit [150, 160], and the “-” level of melting temperature is within the
limit [135,145]. The full factorial experiment design deformation data is obtained by measuring
random points of the extrudate width and height using a caliper, as it is shown in Table 6.6.
96

Table 6.6 Full factorial experiment design deformation data
Horizontal Vertical
Group
Width
1.1
Width
1.2
Width
1.3
Width
2.1
Width
2.2
Width
2.3
Height
1.1
Height
1.2
Height
1.3
Height
2.1
Height
2.2
Height
2.3
1 2.658 2.666 2.654 2.62 2.659 2.707 1.008 1.019 1.014 1.009 1.02 1.022
2 3.149 3.205 3.193 3.129 3.162 3.141 1.099 1.111 1.134 1.064 1.111 1.129
3 2.638 2.67 2.716 2.671 2.723 2.737 1.017 1.031 1.036 1.007 1.032 1.026
4 3.049 3.071 3.035 3.06 3.043 3.066 1.04 1.042 1.06 1.044 1.058 1.06
5 2.635 2.636 2.594 2.64 2.655 2.651 1.013 1.006 1.016 1.004 1.017 1.005
6 3.089 3.092 3.152 3.181 3.153 3.15 0.979 0.967 1.003 0.978 0.962 0.984
7 2.672 2.655 2.664 2.646 2.677 2.684 1.019 1.011 1.004 1.005 1.006 1.004
8 2.905 2.905 2.858 2.883 2.903 2.897 1.003 1.029 1.029 1.008 1.025 1.025

97

6.4. Additional Experiment Analysis
A) Horizontal deformation analysis
Similarly, the horizontal deformation experiment dataset is listed in the following Table 6.7 by
mean and log variance.
Table 6.7 Full factorial experiment horizontal deformation dataset

Ash
proportion
Melting
time
Melting
temperature
Y bar Log (Si2)
1 1 1 1 0.508 1.09E-05
2 1 1 -1 0.34583 1.25E-05
3 1 -1 1 0.590165 1.27E-03
4 1 -1 -1 0.617165 7.35E-04
5 -1 1 1 0.120335 1.42E-05
6 -1 1 -1 0.1465 6.36E-04
7 -1 -1 1 0.114665 3.56E-06
8 -1 -1 -1 0.08917 3.65E-04
A regression analysis is conducted on the data using R software to obtain half-normal plots for
location and dispersion and all factorial effects, as shown by Figure 6.10, Figure 6.11 and Table
6.8.

Figure 6.10 Half-normal distribution of location model in horizontal deformation
98

Figure 6.11 Half-normal distribution of dispersion model in horizontal deformation
The half-normal plot of location model in horizontal deformation shows that the ash proportion
is the significant factors in the process. The half-normal plot of dispersion model in horizontal
deformation do not identify any significant factor for variance.
Table 6.8 Factorial effects of horizontal deformation
Effect y ̅ lns
2

A 0.3976 0.5940
B -0.0726 -1.723
C 0.0336 -2.006
A:B -0.1041 -2.693
A:C 0.0340 2.208
B:C
A:B:C
0.0344
0.0602
0.0351
-0.3779
All the factorial effects include the main factor, two-factor and three-factor interactions are
shown in Table 6.8 for the y ̅ and lns
2
data of horizontal deformation.
Thus, the regression equation for location is
𝑦 ′
̂
= 𝑙𝑛 𝑦 ̂ = 0.31648 + 0.19881 × 𝑥 𝐴
Where y’=lny

99

The interaction plot visualizes the influence of one factor depending on the level of another
factor. Figure 6.12 is an interaction plot of melting temperature against ash proportion. The
average of horizontal deformation is used as vertical axis. The solid and dotted lines show two
levels of melting temperature (B factor), and the horizontal axis represents the influence of ash
proportion (A factor). When melting temperature is at 140 ℃, the change of ash proportion
causes more variance in horizontal deformation than that of 155℃. In Figure 6.13, when melting
time is 1 hour, the change of ash proportion causes more variance in horizontal
deformation than that of a 2-hour case.

Figure 6.12 Interaction plot of melting temperature against ash proportion

Figure 6.13 Interaction plot of melting time against ash proportion

100

B) Vertical deformation analysis
Similar to the horizontal deformation analysis, the vertical deformation experiment dataset is
listed in the following Table 6.9 by mean and log variance. During the width deformation
process, the height of the extruded sample also slumps. A regression analysis is conducted on the
data using R software, half-normal plots for location and dispersion and all factorial effects are
obtained in Figure 6.14 and Figure 6.15.
Table 6.9 Full factorial experiment vertical deformation dataset
Group A B C ybar ln(s2)
1 1 1 1 0.0606665 2.22E-05
2 1 1 -1 0.029833 5.00E-07
3 1 -1 1 -0.011165 3.47E-05
4 1 -1 -1 0.118 8.89E-05
5 -1 1 1 0.0181665 2.01E-05
6 -1 1 -1 0.0348335 2.01E-05
7 -1 -1 1 0.0253335 5.55E-06
8 -1 -1 -1 0.020167 4.50E-06

Figure 6.14 Half-normal distribution of location model in vertical deformation
101

Figure 6.15 Half-normal distribution of dispersion model in vertical deformation
The half-normal distribution of location model in vertical deformation does not identify any
significant factor. The half-normal distribution of dispersion model in vertical deformation also
does not identify any significant factor.
C) Deformation errors analysis
In previous two deformation experiments on horizontal deformation and vertical deformation,
only ash proportion is identified as a significant factor of horizontal deformation. The interaction
effect of melting temperature against ash proportion and the interaction effect of melting time
against ash proportion conduct very little influence on the deformation. Table 6.10 shows the
statistical analysis of deformation on both directions.
For all eight groups, the horizontal deformation mean error of 4 groups is less than 5% and the
vertical deformation mean error of 6 groups is less than 5%. Take the Group 5 for example, it has
the horizontal deformation mean error of 4.73% and standard deviation of 0.022, and the vertical
deformation mean error of 2.04% and standard deviation of 0.006. This deformation result meets
the Hypothesis.

102

Table 6.10 Horizontal and vertical data statistical analysis
Group
Horizontal
deformation
mean (in)
Horizontal
deformation
mean error
Horizontal
deformation
standard
deviation
Vertical
deformation
mean (in)
Vertical
deformation
mean error
Vertical
deformation
standard
deviation
1 0.115 4.50% 0.028 0.025 2.56% 0.006
2 0.617 24.24% 0.030 0.118 11.92% 0.025
3 0.147 5.75% 0.038 0.035 3.52% 0.011
4 0.508 19.95% 0.014 0.061 6.13% 0.010
5 0.089 3.50% 0.022 0.020 2.04% 0.006
6 0.590 23.18% 0.037 -0.011 -1.13% 0.014
7 0.120 4.73% 0.014 0.018 1.84% 0.006
8 0.346 13.58% 0.019 0.030 3.01% 0.011
6.5. The Extrudate Sample Cross-section Comparison
Statistically, the result previous section meets the Hypothesis 2. To further study the extrudate
experiment, the extrudate samples are sliced into sections and compared in Figure 6.16. All the
pictures are taken by one camera with the same configuration and plotted in the same scale. By
comparing the cross sections, the following conclusions can be reached:
A) By comparing the #1, 3, 5, 7 and #2, 4, 6, 8 samples, the samples built with ash proportion of
10% by weight have less horizontal deformation than those built with ash proportion of 20%.
Moreover, the deformation in width also causes cracks in the body of samples as marked by the
bold arrows.
B) By comparing the #1, 4, 6, 7 and # 2, 3, 5, 8 samples, the samples built with melting time of 2
hours has a darker color than those built with melting time of only 1 hour. Technically, with
longer melting time the bitumen modification process matures more completely. According to
Mohamed, the more maturely modified sulfur concrete has better mechanical strength and
thermal cycle resistance [57].
103

#1. 10%Ash_140 ℃_2hours #2. 20%Ash_140℃_1hours

#3. 10%Ash_155 ℃_1hours #4. 20%Ash_155℃_2hours

#5. 10%Ash_140 ℃_1hours #6. 20%Ash_140℃_2hours

#7. 10%Ash_155 ℃_2hours #8. 20%Ash_155℃_1hours
Figure 6.16 The cross section of sulfur concrete extrudate
C) Seeing the #1, 5, 7, both horizontal deformation mean error and vertical deformation mean
error are less than 5%, which achieves the goal of the stated hypothesis.
D) For all eight samples, the upper part is more porous than the lower part, which means the
sulfur sinks to the lower part during the solidification process. Further study should concentrate
on how to make sulfur concrete mixture uniform.
104

6.6. Summary
This chapter presented the results of fractional factorial experiment to study the influence of four
high impact factors. The result shows that the ash proportion, sulfur proportion and melting time
are the three main influential factors on the horizontal deformation. Considering the
experimental samples, those that were built with 30% sulfur proportion contain unacceptable
discontinuity. Consequently, a full factorial experiment is carried out based on the ash
proportion, melting time and melting temperature. This experiment evidently shows the
influential factors to both horizontal deformation and vertical deformation. The cross-section of
these sample are compared and the results of three groups meet the hypothesis at the beginning
of this dissertation.

105

7. Conclusion
A combination of developmental, theoretical, simulation and experimental activities constitute
the body of work that has been presented in this dissertation. Following are the research
contributions:
 After numerous trial designs, a heated and low pressure extruder system has been
especially developed for extrusion of molten sulfur concrete, and a systematic experiment
have been carried out to characterize and configure the material composition and
preparation and the process parameters. The result is proven to be a consistent and robust
process which has been confirmed through the following experiments.
 Fabrication of numerous specimens using a 6-axis robotic printing framework has been
carried out. The overall system is capable of creating complex structures, such as
protection walls and spacecraft shelters, as well as a variety of terrestrial structures.
 A theoretical model of sulfur concrete deformation has been created and a FEA modeling
framework has been developed to study the extrudate deformation phenomenon during
the extrusion process.
 A series of factorial experiments have been designed to study the influence of
experimental parameters. By analyzing the results optimized factors have been achieved
that can result in an extrudate deformation within 5% of total extrudate weight, as well as
reduced leakage between the successive layers which result in smooth surfaces.

106

7.1. Recommendations for future work
7.1.1. Complex Shape Construction
The conducted experiments show the success of single layer printing. For future work, multiple
layer printing with for slant structures as shown Figure 7.1 may be studied.

Figure 7.1 Modified sulfur concrete printing on different platforms
Construction of solid-core vertical walls shown in Figure 7.2 (a) and walls with corrugated
internal structure (b) as well as more complicated multi-segment structures (c) may be pursued in
future research.

(a) (b) (c)
Figure 7.2 0° printed structures
Free-form small-slope curved structures shown in Figure 7.3 (a) designed as blast protection
walls next to landing pads as shown Figure 7.3 (b) may also be studied.
107

(a) (b)
Figure 7.3 10° printed structures
Finally large-angle sloped curved structures shown in Figure 7.4 (a) designed for self-supported
roofed structures, for applications such as spacecraft shelters shown in Figure 7.4 (b) may also be
studied. For this type of structures, fast-curing and interlocking methods mentioned in Figure 4.5
can be utilized.

(a) (b)
Figure 7.4 40° sloped curved structures
7.1.2. Thermal Transfer Process
Thermal Transfer Process is another relevant topic in sulfur concrete Contour Crafting. In the
mini-size Martian/Lunar regolith simulant sulfur concrete experiment (Section 3.3), the
dimension of extrudate is 0.5×0.5-inch as it is shown in Figure 7.5 (a). In this case the quick heat
loss speeds up the solidification of sulfur concrete. So, the horizontal and vertical deformations
108

are relatively small. While, the heat loss becomes negligible in the full-size (2.5×1-inch) sulfur
concrete Contour Crafting experiment. This deformation of full-size extrudate is due to the
increased gravity and high heat energy concentration of molten sulfur concrete. Studying the
thermal transfer phenomenon can help to understand the molten sulfur concrete solidification
process as a result of which heat transfer solutions to speed up the curing of the material mix
may be developed.

(a) Dimension comparison of two extrudate (b) Section of flowing material in a nozzle
Figure 7.5 Two topics about thermal transfer
Figure 7.5 (b) shows a section of flowing material in a nozzle. The two black blocks represent
the extrusion nozzle body. The yellow block with an arrow represents the flowing sulfur concrete
in the nozzle. The heat transfers between the inner wall of nozzle and the sulfur concrete
influences the friction of flowing material. Since the viscosity of sulfur changes with its
temperature, the pumpability of the extruder can be improved by studying this heat transfer
phenomenon.

2.5×1-inch
0.5×0.5-inch
109

8. Bibliography
[1] astm, “Standard Terminology for Additive Manufacturing Technologies.” [Online].
Available: http://www.astm.org/FULL_TEXT/F2792/HTML/F2792.htm#s00200.
[2] I. Gibson, D. W. Rosen, B. Stucker, and others, Additive manufacturing technologies.
Springer, 2010.
[3] 3d printing Industry, “The free beginner’s guide to 3D printing.”
[4] R. A. Buswell, R. C. Soar, A. G. F. Gibb, and A. Thorpe, “Freeform Construction : Mega-
scale Rapid Manufacturing for construction,” vol. 16, pp. 224–231, 2007.
[5] B. Khoshnevis, G. Bekey, and C. Science, “Automated Construction using Contour
Crafting – Applications on Earth and Beyond,” pp. 1–6, 1997.
[6] H. Kwon, “Experimentation and Analysis of Contour Crafting (Cc) Process Using
Uncured Ceramic Materials,” no. August, 2002.
[7] D. Hwang and B. Khoshnevis, “CONCRETE WALL FABRICATION BY CONTOUR
CRAFTING Dooil Hwang and Behrokh Khoshnevis,” 2004.
[8] D. Hwang and B. Khoshnevis, “An Innovative Construction Process-Contour Crafting,”
22nd Int. Symp. Autom. Robot. Constr. ISARC, no. Cc, 2005.
[9] B. Khoshnevis, D. Hwang, K.-T. Yao, and Z. Yah, “Mega-scale fabrication by contour
crafting,” Int. J. Ind. Syst. Eng., vol. 1, no. 3, pp. 301–320, 2006.
[10] B. Khoshnevis and D. Hwang, “Contour crafting,” in Rapid Prototyping, Springer, 2006,
pp. 221–251.
[11] K. Behdinan, “METHODOLOGY FOR DESIGN OF A VIBRATION OPERATED
110

VALVE FOR ABRASIVE VISCOUS FLUIDS by,” Univ. South. Calif., vol. 53, no. 9, pp.
1689–1699, 2013.
[12] K. Behdinan and B. Khoshnevis, “Methodology for design of a vibration operated valve
for abrasive viscous fluids,” Mechatronics, vol. 23. pp. 1025–1036, 2013.
[13] B. Khoshnevis, M. Thangavelu, X. Yuan, and J. Zhang, “Advances in Contour Crafting
Technology for Extraterrestrial Settlement Infrastructure Buildup,” in AIAA SPACE 2013
Conference and Exposition, 2013.
[14] B. Khoshnevis, X. Yuan, B. Zahiri, and B. Xia, “Construction by Contour Crafting using
Sulfur Concrete with Planetary Applications.”
[15] B. Khoshnevis, X. YUAN, B. Zahiri, J. Zhang, and B. Xia, “Deformation Analysis of
Sulfur Concrete Structures Made by Contour Crafting,” in AIAA SPACE 2015 Conference
and Exposition, 2015.
[16] B. Khoshnevis, “Automated construction by contour crafting - Related robotics and
information technologies,” Autom. Constr., vol. 13, pp. 5–19, 2004.
[17] N. Leach, A. Carlson, B. Khoshnevis, and M. Thangavelu, “Robotic Construction by
Contour Crafting: The Case of Lunar Construction,” Int. J. Archit. Comput., vol. 10, no.
03, pp. 423–438, 2012.
[18] B. Khoshnevis, M. Thangavelu, X. Yuan, and J. Zhang, “Advances in Contour Crafting
Technology for Extraterrestrial Settlement Infrastructure Buildup,” AIAA, vol. 5438, pp.
10–12, 2013.
[19] CivilBlog, “Factors Affecting Workability of Concrete.” [Online]. Available:
civilblog.org/2014/04/04/what-are-the-factors-affecting-workability-of-concrete/.
111

[20] K. Saidi, J. O’Brien, and A. M. Lytle, “Robotics in construction,” Springer Handb.
Robot., pp. 1079–1099, 2008.
[21] J. G. Everett and H. Saito, “CONSTRUCTION AUTOMATION: DEMANDS AND
SATISFIERS IN THE UNITED STATES AND JAPAN.”
[22] P. Bosscher, R. L. Williams, L. S. Bryson, and D. Castro-Lacouture, “Cable-suspended
robotic contour crafting system,” Autom. Constr., vol. 17, no. 1, pp. 45–55, 2007.
[23] ExOne, “What is Binder Jetting?” [Online]. Available:
http://www.exone.com/Resources/Technology-Overview/What-is-Binder-Jetting.
[24] J. Zhang and B. Khoshnevis, “Selective Separation Sintering ( SSS ) A New Layer Based
Additive Manufacturing Approach for Metals and Ceramics,” pp. 71–79.
[25] N. Leach, E. Dini, and F. Partners, “3D PRINTING IN SPACE,” pp. 108–113.
[26] X. De Kestelier, “Lunar Outpost Design,” pp. 3–6, 2015.
[27] G. Cesaretti, E. Dini, X. De Kestelier, V. Colla, and L. Pambaguian, “Acta Astronautica
Building components for an outpost on the Lunar soil by means of a novel 3D printing
technology,” Acta Astronaut., vol. 93, pp. 430–450, 2014.
[28] S. Lim, T. Le, J. Webster, R. Buswell, S. Austin, A. Gibb, and T. Thorpe,
“FABRICATING CONSTRUCTION COMPONENTS USING LAYERED
MANUFACTURING TECHNOLOGY,” pp. 512–520, 2004.
[29] WIKI, “Lunarcrete.” [Online]. Available: https://en.wikipedia.org/wiki/Lunarcrete.
[30] C. Verseux, M. Baqué, K. Lehto, J.-P. P. de Vera, L. J. Rothschild, and D. Billi,
“Sustainable life support on Mars – the potential roles of cyanobacteria,” Int. J. Astrobiol.,
112

vol. 15, no. 01, pp. 65–92, Jan. 2016.
[31] WIKI, “Colonization of Mars.” [Online]. Available:
https://en.wikipedia.org/wiki/Colonization_of_Mars.
[32] Alex Kirk, “Statement of Mars Society President Robert Zubrin on the LCROSS Results.”
[Online]. Available:
https://web.archive.org/web/20091124042249/http://www.marssociety.org/portal/ZubrinSt
atementLCROSS/.
[33] CNN, “Moon hole might be suitable for colony.” [Online]. Available:
http://www.cnn.com/2010/TECH/space/01/01/moon.lava.hole/.
[34] “In-Situ Resource Utilization,” Ames Technology Capabilities and Facilities. [Online].
Available: http://www.nasa.gov/centers/ames/research/technology-onepagers/in-
situ_resource_Utiliza14.html.
[35] J. Mangels, “Coping with a lunar dust-up,” the seattle times. [Online]. Available:
http://www.seattletimes.com/nation-world/coping-with-a-lunar-dust-up/.
[36] E. Robens, a. Bischoff, a. Schreiber, a. Dabrowski, and K. K. Unger, “Investigation of
surface properties of lunar regolith: Part I,” Appl. Surf. Sci., vol. 253, pp. 5709–5714,
2007.
[37] E. Robens, a. Dąbrowski, E. Mendyk, J. Goworek, K. Skrzypiec, M. Drewniak, M.
Dumańska-Słowik, W. Gac, R. Dobrowolski, S. Pasieczna-Patkowska, M. Huber, M.
Iwan, K. J. Kurzydłowski, T. Płociński, J. Ryczkowski, Z. Rzączyńska, and J. W.
Sobczak, “Investigation of surface properties of lunar regolith - Part IV,” Ann. UMCS,
Chem., vol. 63, pp. 144–168, 2010.
113

[38] F. Ruess, J. Schaenzlin, and H. Benaroya, “Structural Design of a Lunar Habitat,” J.
Aerosp. Eng., vol. 19, no. July, pp. 133–157, 2006.
[39] C. M. Schrader, D. L. Rickman, C. a Mclemore, J. C. Fikes, and M. Space, “Lunar
Regolith Simulant User ’ s Guide,” no. September, 2010.
[40] C. Schrader, “Extant and Extinct Lunar Regolith Simulants : Modal Analyses of NU-LHT-
1M,” 2008.
[41] J. a Senseney, L. D. Arp, and C. Lindbergh, “Fig . 1 shows the proposed three story
concrete structure with a 24 ft story height and a 120 ft diameter . The reason for choosing
this particular height is to provide work space suitable to industrial operations and
flexibility of converting it into two,” vol. 2, no. 1, pp. 10–19, 1989.
[42] W. V. B. Street, “Material safety data sheet Material safety data sheet,” pp. 1–8, 2013.
[43] Casanova, “FEASIBILITY AND APPLICATIONS OF SULFUR CONCRETE FOR
LUNAR BASE DEVELOPMENT: A PRELIMINARY STUDY. I.”
[44] B. Khoshnevis, M. Bodiford, K. Burks, E. Ethridge, D. Tucker, W. Kim, H. Toutanji, and
M. Fiske, “Lunar contour crafting—a novel technique for ISRU-based habitat
development,” Autom. Constr., vol. 13(1), no. January, pp. 5–19, 2004.
[45] L. a Taylor and Y. Liu, “Earth and Space 2010: Engineering, Science, Construction, and
Operations in Challenging Environments © 2010 ASCE 106,” Environments, pp. 106–
118, 2010.
[46] S. Thickness and A. Rates, “The Lunar Soil Chapter 6,” 1968.
[47] D. Vaniman, D. Pettit, and G. Heiken, “Uses of Lunar Sulfur,” Most, pp. 429–435.
114

[48] M. M. Vlahović, M. M. Savić, S. P. Martinović, T. D. Boljanac, and T. D. Volkov-
Husović, “Use of image analysis for durability testing of sulfur concrete and Portland
cement concrete,” Mater. Des., vol. 34, pp. 346–354, 2012.
[49] A. H. Vroom and M. a Ward, “Sulfur Concrete-a New,” 1975.
[50] Y. Zheng, S. Wang, C. Li, Z. Ouyang, J. Liu, and Y. Zou, “The Development of CAS-1
Lunar Soil Simulant,” ILEWG Conf. Explor. Util. Moon (7th), Sept 18-23, pp. 0–3, 2005.
[51] G. R. Darnell, “Sulfur polymer cement, a new stabilization agent for mixed and low-level
radioactive waste,” 1991.
[52] M. Ai-qudah, “Materials Materials and Experimental,” Concrete, vol. 24, pp. 165–175,
1994.
[53] P. Bartos, Fresh concrete: properties and tests. Elsevier, 2013.
[54] H. Cement, M. Rooms, B. Statements, and D. Mass, “Standard Test Method for
Compressive Strength of Hydraulic Cement Mortars,” vol. 04, pp. 1–6, 2000.
[55] S. Specification, “Standard Specification for Surface-Applied Bonding Compounds for
Exterior,” vol. 04, no. August, pp. 1–5, 2003.
[56] C. Strengths and T. Strengths, “There are a variety of commonly used test procedures .”
[57] A.-M. O. Mohamed and M. El-Gamal, Sulfur concrete for the construction industry: a
sustainable development approach. J. Ross Publishing, 2010.
[58] B. Meyer, “Elemental sulfur,” Chem. Rev., vol. 76, no. 3, pp. 367–388, 1976.
[59] R. F. Bacon and R. Fanelli, “The Viscosity of Sulfur1,” J. Am. Chem. Soc., vol. 65, no. 4,
pp. 639–648, 1943.
115

[60] A. Hackl and R. Gregor, “NEW APPROACH TO SULFUR CONCRETE,” in
ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY, 1977, vol. 173,
no. MAR 20, p. 18.
[61] A. Bruggink, E. C. Roos, and E. de Vroom, “Penicillin acylase in the industrial production
of β-lactam antibiotics,” Org. Process Res. Dev., vol. 2, no. 2, pp. 128–133, 1998.
[62] W. A. Pryor, Mechanisms of sulfur reactions. McGraw-Hill, 1962.
[63] L. Blight, B. R. Currell, B. J. Nash, R. A. M. Scott, and C. Stillo, “Preparation and
properties of modified sulfur systems,” 1978.
[64] A.-M. O. Mohamed and M. El Gamal, “Sulfur based hazardous waste solidification,”
Environ. Geol., vol. 53, no. 1, pp. 159–175, 2007.
[65] E. L. Nimer and R. W. Campbell, “Sulfur cement-aggregate compositions and methods
for preparing.” Google Patents, 1985.
[66] G. G. Allan and A. N. Neogi, “Macromolecular organometallic catalysis: I. Epoxidation.
A comparative kinetic study of tungstate, molybdate, and vanadate ions as catalysts for the
epoxidation of maleic acid at pH 5,” J. Catal., vol. 16, no. 2, pp. 197–203, 1970.
[67] B. Khoshnevis, M. P. Bodiford, K. H. Burks, E. Ethridge, D. Tucker, W. Kim, H.
Toutanji, and M. R. Fiske, “Lunar contour crafting—a novel technique for ISRU-based
habitat development,” in 43rd AIAA Aerospace Sciences Meeting and Exhibit—Meeting
Papers, 2005, pp. 7397–7409.
[68] K. Chockalingam, N. Jawahar, and U. Chandrasekhar, “Influence of layer thickness on
mechanical properties in stereolithography,” Rapid Prototyp. J., vol. 12, pp. 106–113,
2006.
116

[69] J. W. Choi, F. Medina, C. Kim, D. Espalin, D. Rodriguez, B. Stucker, and R. Wicker,
“Development of a mobile fused deposition modeling system with enhanced
manufacturing flexibility,” J. Mater. Process. Technol., vol. 211, no. 3, pp. 424–432,
2011.
[70] R. R. Dehoff and S. S. Babu, “Characterization of interfacial microstructures in 3003
aluminum alloy blocks fabricated by ultrasonic additive manufacturing,” Acta Mater., vol.
58, no. 13, pp. 4305–4315, 2010.
[71] a. Hadjoui, H. Mebarek, and B. B. Bouiadjra, “Finite Element Method,” Int. J. Comput.
Methods Eng. Sci. Mech., vol. 12, no. 2, pp. 59–74, 2011.
[72] R. Hahnlen and M. J. Dapino, “NiTi-Al interface strength in ultrasonic additive
manufacturing composites,” Compos. Part B Eng., vol. 59, pp. 101–108, 2014.
[73] Y.-M. H. and C.-P. Jiang, “Curl Distortion Analysis During Photopolymerisation of,” Int
Adv Manuf Technol, vol. 21, pp. 586–595, 2003.
[74] D. Li and R. Soar, “Influence of sonotrode texture on the performance of an ultrasonic
consolidation machine and the interfacial bond strength,” J. Mater. Process. Technol., vol.
209, pp. 1627–1634, 2009.
[75] J. O. Obielodan, B. E. Stucker, E. Martinez, J. L. Martinez, D. H. Hernandez, D.a.ramirez,
and L. E. Murr, “Optimization of the shear strengths of ultrasonically consolidated Ti/Al
3003 dual-material structures,” J. Mater. Process. Technol., vol. 211, no. 6, pp. 988–995,
2011.
[76] M. R. Sriraman, S. S. Babu, and M. Short, “Bonding characteristics during very high
power ultrasonic additive manufacturing of copper,” Scr. Mater., vol. 62, no. 8, pp. 560–
117

563, 2010.
[77] M. Vatani, F. Barazandeh, A. Rahimi, and A. S. Nezhad, “Distortion modeling of SL parts
by classical lamination theory,” Rapid Prototyp. J., vol. 18, pp. 188–193, 2012.
[78] P. F. G. Banfill, “Rheology of fresh cement and concrete,” Rheol. Rev., vol. 2006, p. 61,
2006.
[79] G. Christensen, Modelling the flow of fresh concrete: the slump test. ProQuest, 1991.
[80] N. Pashias, D. V Boger, J. Summers, and D. J. Glenister, “A fifty cent rheometer for yield
stress measurement,” J. Rheol., vol. 40, no. 6, pp. 1179–1189, 1996.
[81] N. Roussel and P. Coussot, “Fifty-cent rheometer for yield stress measurements: from
slump to spreading flow,” J. Rheol., vol. 49, no. 3, pp. 705–718, 2005.
[82] N. Roussel, “Three-dimensional numerical simulations of slump tests,” Ann. Trans. Nord.
Rheol. Soc., vol. 12, pp. 55–62, 2004.
[83] N. Roussel, C. Stefani, and R. Leroy, “From mini-cone test to Abrams cone test:
measurement of cement-based materials yield stress using slump tests,” Cem. Concr. Res.,
vol. 35, no. 5, pp. 817–822, 2005.
[84] W. R. Schowalter and G. Christensen, “Toward a rationalization of the slump test for fresh
concrete: comparisons of calculations and experiments,” J. Rheol., vol. 42, no. 4, pp. 865–
870, 1998.
[85] X. Zhou and Z. Li, “Characterization of rheology of fresh fiber reinforced cementitious
composites through ram extrusion,” Mater. Struct., vol. 38, no. 1, pp. 17–24, 2005.
[86] N. M. Barkoula, B. Alcock, N. O. Cabrera, and T. Peijs, “Fatigue properties of highly
118

oriented polypropylene tapes and all-polypropylene composites,” Polym. Polym. Compos.,
vol. 16, pp. 101–113, 2008.
[87] M. Bengtsson, M. Le Baillif, and K. Oksman, “Extrusion and mechanical properties of
highly filled cellulose fibre-polypropylene composites,” Compos. Part A Appl. Sci.
Manuf., vol. 38, pp. 1922–1931, 2007.
[88] Y. Chen, L. J. Struble, and G. H. Paulino, “Fabrication of functionally graded-cellular
structures of cement-based materials by co-extrusion,” AIP Conf. Proc., vol. 973, pp. 532–
537, 2008.
[89] H.-S. Chu, K.-S. Liu, and J.-W. Yeh, “Aging behavior and tensile properties of 6061Al–
0.3 μm Al2O3p particle composites produced by reciprocating extrusion,” Scr. Mater.,
vol. 45, pp. 541–546, 2001.
[90] B. Equations and C. Relation, “ORIENTING FIBERS BY EXTRUSION IN
REINFORCED REACTIVE POWDER CONCRETE,” no. June, pp. 593–598, 2001.
[91] S. H. Hong, K. H. Chung, and C. H. Lee, “Effects of hot extrusion parameters on the
tensile properties and microstructures of SiCw-2124Al composites,” Mater. Sci. Eng. A,
vol. 206, pp. 225–232, 1996.
[92] J. Iba, A. Borrego, R. Ferna, and G. Gonza, “Influence of extrusion temperature on the
microstructure and the texture of 6061Al – 15 vol .% SiC w PM composites,” vol. 62, pp.
731–742, 2002.
[93] D. De Koker and G. Van Zijl, “Extrusion of engineered cement-based composite
material,” Proc. BEFIB, no. September, pp. 1301–1310, 2004.
[94] H. J. Li, L. H. Qi, H. M. Han, and L. J. Guo, “Neural network modeling and optimization
119

of semi-solid extrusion for aluminum matrix composites,” J. Mater. Process. Technol.,
vol. 151, pp. 126–132, 2004.
[95] Z. Li, Y. Zhang, and X. Zhou, “Short Fiber Reinforced Geopolymer Composites
Manufactured by Extrusion,” J. Mater. Civ. Eng., vol. 17, no. December, pp. 624–631,
2005.
[96] K. Oksman, M. Skrifvars, and J. F. Selin, “Natural fibres as reinforcement in polylactic
acid (PLA) composites,” Compos. Sci. Technol., vol. 63, pp. 1317–1324, 2003.
[97] E. Ploumis-Devick, “Characterization of sulfur-asphhalt-dune sand paving mixtures,” Soc.
Change, p. 125, 1983.
[98] Y. Shao, J. Qiu, and S. P. Shah, “Microstructure of extruded cement-bonded fiberboard,”
Cem. Concr. Res., vol. 31, pp. 1153–1161, 2001.
[99] I. T. O. Users, “Dune sand-aggregate mixes and dune sand-sulfur mixes for asphaltic
concrete pavements.”
[100] X. Zhou and Z. Li, “Characterizing Rheology of Fresh Short Fiber Reinforced
Cementitious Composite through Capillary Extrusion,” J. Mater. Civ. Eng., vol. 17, no.
February, pp. 28–35, 2005.
[101] M. Pastor, O. C. Zienkiewicz, and A. H. C. Chan, “Generalized plasticity and the
modelling of soil behaviour,” Int. J. Numer. Anal. Methods Geomech., vol. 14, no. 3, pp.
151–190, 1990.
[102] H. Nasa and H. Toutanji, “Generated by Foxit PDF Creator © Foxit Software
http://www.foxitsoftware.com For evaluation only . Archive for sulfur concrete Waterless
concrete to cut building costs on moon Generated by Foxit PDF Creator © Foxit Software
120

http://www.foxitsoftware.com Fo,” 2008.
[103] P. Menetrey and K. J. Willam, “Triaxial failure criterion for concrete and its
generalization,” ACI Struct. J., vol. 92, no. 3, pp. 311–318, 1995.
[104] D. Simulia, “ABAQUS 6.11 analysis user’s manual,” Abaqus, vol. 6, p. 22, 2011.
[105] W. W. Are, A. Information, O. Services, C. Us, O. Supplies, S. A. Articles, W. H.
Capacity, U. Hydraulic, S. W. Retention, S. W. Holding, S. T. Classification, S. Testing,
C. E. Capacity, and C. S. Testing, “Water Holding Capacity,” pp. 1–2, 2015.
[106] M. Arlery, M. Gardou, J. M. Fleureau, and C. Mariotti, “Dynamic behaviour of dry and
water-saturated sand under planar shock conditions,” Int. J. Impact Eng., vol. 37, no. 1,
pp. 1–10, 2010.
[107] eLibray, “Plant & Soil Sciences,” 2014, pp. 1–2, 2014.
[108] W. H. Gardner, “Water content,” Methods Soil Anal. Part 1. Monogr. 9, pp. 493–544,
1986.
[109] D. Hillel, “Soil and water,” p. 288, 1971.
[110] A. Kudrolli, “Granular matter: sticky sand.,” Nat. Mater., vol. 7, no. March, pp. 174–175,
2008.
[111] C. T. Me, M. a Nh, and N. Y. Ri, “Northeast Region Certified Crop Adviser ( NRCCA ),”
Management, pp. 2–4, 2007.
[112] S. Quality and I. Sheet, “Soil Quality Resource Concerns :,” Office, no. January, pp. 0–1,
1998.
[113] J. L. M. van Mechelen, “Strength of moist sand controlled by surface tension for tectonic
121

analogue modelling,” Tectonophysics, vol. 384, pp. 275–284, 2004.
[114] F. Wikipedia, “Soil water ( retention ),” pp. 28–30, 2015.
[115] C. F. J. WU, No TitleExperiments: Planning, Analysis, and Optimization (Second
Edition). Wiley Online Library, 2009.

122

9. Appendix: R Code and Result for Experimental Data Analysis
# R Code for fractional factorial experiment design
#R Codes for Data Analysis
# set working directory
setwd("~/Desktop/Xiao's Paper")

# 1.For Horizontal Deformation Analysis
#Design Matrix
print<-read.table('FFED.txt',h=F)
ybar<-apply(print[,5:6],1,mean)
s2<-apply(print[,5:6],1,var)
print<-print[,1:4]
print[,5]<-ybar
print[,6]<-s2
colnames(print)<-c('A','B','C','D', 'ybar','s2')

#Location
gy<-lm(ybar~A*B*C*D,print)
summary(gy)
gy$coef[-1]*2
install.packages('faraway')
library(faraway)
halfnorm(gy$coef[-1]*2,labs=names(gy$coef[-1]))

#Dispersion
gs<-lm(log(s2)~A*B*C*D,print)
summary(gs)
halfnorm(gs$coef[-1]*2,labs=names(gs$coef[-1]))
gs$coef[-1]*2
factorialeffects<-gs$coef[-1]*2
Factorial factor design in horizontal deformation:
123

124

Factorial factor design in vertical deformation

125

# 2.For Vertical Deformation Analysis
#Design Matrix
print<-read.table('FFED.txt',h=F)
ybar<-apply(print[,7:8],1,mean)
s2<-apply(print[,7:8],1,var)
print<-print[,1:4]
print[,5]<-ybar
print[,6]<-s2
colnames(print)<-c('A','B','C','D', 'ybar','s2')

#Location
gy<-lm(ybar~A*B*C*D,print)
summary(gy)
gy$coef[-1]*2
library(faraway)
halfnorm(gy$coef[-1]*2,labs=names(gy$coef[-1]))

126

#Dispersion
gs<-lm(log(s2)~A*B*C*D,print)
summary(gs)
halfnorm(gs$coef[-1]*2,labs=names(gs$coef[-1]))
gs$coef[-1]*2

# R Code for full factorial experiment design
#R Codes for Data Analysis
# set working directory
setwd("~/Desktop/Xiao's Paper")

#For Horizontal Deformation Analysis
#Design Matrix
print<-read.table('CC DOE.txt',h=F)
ybar<-apply(print[,4:5],1,mean)
s2<-apply(print[,4:5],1,var)
print<-print[,1:3]
print[,4]<-ybar
print[,5]<-s2
colnames(print)<-c('A','B','C','ybar','s2')

#Location
gy<-lm(ybar~A*B*C,print)
summary(gy)
gy$coef[-1]*2
install.packages('faraway')
library(faraway)
halfnorm(gy$coef[-1]*2,labs=names(gy$coef[-1]))

#Dispersion
gs<-lm(log(s2)~A*B*C,print)
summary(gs)
127

halfnorm(gs$coef[-1]*2,labs=names(gs$coef[-1]))
gs$coef[-1]*2

#Interaction Plot
interaction.plot(print$A,print$B,print$ybar, main="B against A Interaction
Plot",xlab='A',ylab='B',col=1:2)
interaction.plot(print$B,print$A,print$ybar, main="A against B Interaction
Plot",xlab='B',ylab='A',col=1:2)
interaction.plot(print$A,print$B,log(print$s2), main="B against A Interaction
Plot",xlab='A',ylab='B',col=1:2)
interaction.plot(print$B,print$A,log(print$s2), main="A against B Interaction
Plot",xlab='B',ylab='A',col=1:2)
interaction.plot(print$A,print$C,log(print$s2), main="C against A Interaction
Plot",xlab='A',ylab='C',col=1:2)
interaction.plot(print$C,print$A,log(print$s2), main="A against C Interaction
Plot",xlab='C',ylab='A',col=1:2)

#For Vertical Deformation Analysis
#Design Matrix
print<-read.table('CC DOE.txt',h=F)
ybar<-apply(print[,6:7],1,mean)
s2<-apply(print[,6:7],1,var)
print<-print[,1:3]
print[,4]<-ybar
print[,5]<-s2
colnames(print)<-c('A','B','C','ybar','s2')

#Location
gy<-lm(ybar~A*B*C,print)
summary(gy)
gy$coef[-1]*2
library(faraway)
halfnorm(gy$coef[-1]*2,labs=names(gy$coef[-1]))

#Dispersion
128

gs<-lm(log(s2)~A*B*C,print)
summary(gs)
halfnorm(gs$coef[-1]*2,labs=names(gs$coef[-1]))
gs$coef[-1]*2

#Interaction Plot
interaction.plot(print$A,print$B,print$ybar, main="B against A Interaction
Plot",xlab='A',ylab='B',col=1:2)
interaction.plot(print$B,print$A,print$ybar, main="A against B Interaction
Plot",xlab='B',ylab='A',col=1:2)
interaction.plot(print$A,print$C,print$ybar, main="C against A Interaction
Plot",xlab='A',ylab='C',col=1:2)
interaction.plot(print$C,print$A,print$ybar, main="A against C Interaction
Plot",xlab='C',ylab='A',col=1:2)
interaction.plot(print$B,print$C,print$ybar, main="C against B Interaction
Plot",xlab='B',ylab='C',col=1:2)
interaction.plot(print$C,print$B,print$ybar, main="B against C Interaction
Plot",xlab='C',ylab='B',col=1:2)

interaction.plot(print$A,print$B,log(print$s2), main="B against A Interaction
Plot",xlab='A',ylab='B',col=1:2)
interaction.plot(print$B,print$A,log(print$s2), main="A against B Interaction
Plot",xlab='B',ylab='A',col=1:2)
interaction.plot(print$A,print$C,log(print$s2), main="C against A Interaction
Plot",xlab='A',ylab='C',col=1:2)
interaction.plot(print$C,print$A,log(print$s2), main="A against C Interaction
Plot",xlab='C',ylab='A',col=1:2)
interaction.plot(print$B,print$C,log(print$s2), main="C against B Interaction
Plot",xlab='B',ylab='C',col=1:2)
interaction.plot(print$C,print$B,log(print$s2), main="B against C Interaction
Plot",xlab='C',ylab='B',col=1:2)
Full factor design in horizontal deformation:
129

130

Full factor design in vertical deformation:

131

Abstract (if available)

Abstract Over the past three decades, additive manufacturing (AM), also widely known as 3D printing, has impacted the traditional manufacturing industry for its specific capability in rapid prototyping, fabrication of complex geometries, creation of multi-material composites and customization. In contrast with most traditional manufacturing processes, which subtracts the unintended material from a blank piece, the AM process fabricates part by successively adding building material layer-by-layer based on a planned tool path generated from a CAD model. So far, various additive manufacturing technologies have been developed for applications in areas such as architecture, space, automotive, fashion, jewelry, energy and even biomedical. Classified by processes, additive manufacturing contains extrusion, powder bed, light polymerized, laminated and powder fed. ❧ Contour Crafting is an extrusion based additive manufacturing process invented by Dr. Behrokh Khoshnevis from University of Southern California, which can quickly build constructions such as houses out of Portland concrete. By depositing wet concrete through a hose against the side shaping trowel at the nozzle outlet, Contour Crafting can create a smooth surface finish over the accumulated successive layers. A special designed triple-nozzle assembly can construct hollow walls with increased strength and less amount of building material. This automatic construction technology aims to release construction workers from dangerous and tedious work. In the Mars and Lunar colonization missions led by NASA, contour crafting is considered to be the promising construction candidate for out space construction, especially in automatically structure protective hangars for equipment and shelter for astronauts. ❧ In this dissertation, the feasibility of contour crafting in space application is analyzed in consideration of the accessible local construction raw material, the adversary environmental conditions, the adjustment of existing Contour Crafting framework for space application and the optimization on special extruder for construction material. A mobile 6-axis robotic system is proposed which can fit in cargo area of launching system, deploy easily by itself after landing and move around to build structures without size limitation. To save the enormous shipping cost of building material from Earth and maximize the utilization of local resource, the sulfur concrete, composed of planetary regolith and elemental sulfur, is chosen as the main construction material for Contour Crafting space application. Unlike the chemical solidification process of Portland cement, in sulfur concrete, the liquid sulfur in the extruded paste solidifies when the temperature drops below 130 ℃ and acts as the binder of the aggregates, which is a physical thermal reaction. ❧ Accordingly, a novel lower pressure extruder is developed for this thermal extrusion process. Since the temperature of sulfur concrete should be kept higher than 130 ℃ before extrusion, the original extrusion process for Portland concrete of pumping building material from a terminal all the way to the nozzle through a hose, is inapplicable due to the multifarious and complex heating systems. This novel extruder is developed with an embedded sulfur concrete reservoir, gradient thermal control system and compliant extrusion blades for wide range properties of sulfur concrete. ❧ To further control the sulfur concrete extrusion flow rate, its workability, includes mobility, pumpability, formability and finishability, is studied. The mobility and pumpability represent the ability of material to be extruded, while the formability and finishability represent the ability of material to be shaped. Technically, the mobility and pumpability of a fluid material conflicts with its formability and finishability. Experiments were carried out to test the influence factors for both extrusion ability and shaping ability. ❧ Sulfur concrete is a Bingham plastic, which behaves as a rigid body at low stresses and flows as a viscous fluid at high stress. Without appropriate control, the solidifying sulfur concrete expands in width and shrinks in height, which causes the failure of construction or terrible finishing surface. A fraction factorial experiment and a full factorial experiment are carried out to study the influence factors and the interaction of these factors. Finally, the deformation of extrudate is controlled within the 5% by weight.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Thermal and deformation analysis of multiphase sulfur concrete extrusion for planetary construction

PDF

Contour Crafting process planning and optimization

PDF

Experimental and numerical techniques to characterize structural properties of fresh concrete

PDF

Evaluation of interlayer bonding strength in contour crafted structures: experimental and statistical approach

PDF

Analyses of strength of layered structures fabricated by Contour Crafting

PDF

Mixture characterization and real-time extrusion quality monitoring for construction-scale 3D printing (Contour Crafting)

PDF

A data driven software platform for process automation, planning and inspection of Contour Crafting large-scale robotic 3D printing system

PDF

Selective separation shaping: an additive manufacturing method for metals and ceramics

PDF

Selective Separation Shaping (SSS): large scale cementitious fabrication potentials

PDF

Energy control and material deposition methods for fast fabrication with high surface quality in additive manufacturing using photo-polymerization

PDF

Full scale contour crafting applications

PDF

Enhancing the surface quality and dimensional accuracy of SIS-metal parts with application to high temperature alloys

PDF

Slurry based stereolithography: a solid freeform fabrication method of ceramics and composites

PDF

Metallic part fabrication wiht selective inhibition sintering (SIS) based on microscopic mechanical inhibition

PDF

3D printing of polymeric parts using Selective Separation Shaping (SSS)

PDF

The extension of selective inhibition sintering (SIS) to high temperature alloys

PDF

Fabrication-aware machine learning for accuracy control in additive manufacturing

PDF

Motion-assisted vat photopolymerization: an approach to high-resolution additive manufacturing

PDF

Process planning for robotic additive manufacturing

PDF

4D printing of self-folding structures using polystyrene film

Asset Metadata

Creator Yuan, Xiao (author)

Core Title Contour crafting construction with sulfur concrete

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Industrial and Systems Engineering

Publication Date 08/03/2016

Defense Date 06/21/2016

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

Tag additive manufacturing construction,contour crafting,OAI-PMH Harvest,space colonization,sulfur concrete

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Khoshnevis, Behrokh (committee chair), Chen, Yong (committee member), Wang, Qiming (committee member)

Creator Email seanxyuan@gmail.com,xiaoyuan@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-293647

Unique identifier UC11280162

Identifier etd-YuanXiao-4723.pdf (filename),usctheses-c40-293647 (legacy record id)

Legacy Identifier etd-YuanXiao-4723.pdf

Dmrecord 293647

Document Type Dissertation

Format application/pdf (imt)

Rights Yuan, Xiao

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