University of Southern California Dissertations and Theses

Hybrid vat photopolymerization: methods and systems

(USC Thesis Other)

Hybrid vat photopolymerization: methods and systems

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

HYBRID VAT PHOTOPOLYMERIZATION:
METHODS AND SYSTEMS

by

Jie Jin

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)

December 2020

Copyright 2020 Jie Jin
ii

Acknowledgments
Eight years ago, I met the 3D printing technology for the first time when I was an
undergraduate student. I was so impressed by this amazing technology. After graduation, I joined
a 3D printing company for a year before I came to USC to pursue my Ph.D. degree in additive
manufacturing. Time flies, five years just passed for my Ph.D. study. When I look back on my past
eight years, it is all about a journey of exploring.
Pursing this Ph.D. degree has been a truly life-changing experience for me, and it would not
have been possible to do without the support and guidance that I received from many people.
Firstly, I would like to express my deepest appreciation to my advisor Prof. Yong Chen for
the continuous support of my Ph.D. study and related research for his patience, motivation, and
immense knowledge. His guidance helped me in all the time of research and writing of this
dissertation. The scheduled one-on-one weekly brainstorm meeting with Prof. Chen really helped
me work out all the challenges I have ever met in my projects. I could not have imagined having
a better advisor and mentor for my Ph.D. study. Without his guidance and constant feedback, this
Ph.D. would not have been achievable.
Besides my advisor, I would like to thank the rest of my screening committee, qualify
committee and dissertation committee: Prof. Berok Khoshnevis, Prof. Qiang Huang, Prof.
Satyandra K. Gupta, Prof. Qifa Zhou and Prof. Qiming Huang, for their insightful comments and
encouragement, but also for the hard question which incented me to widen my research from
various perspectives.
I cannot begin to express my thanks to my fellow labmates in the lab, Huachao Mao, Xiangjia
Li, Yang Yang, YuenShan Leung, Tsz-Ho Kwok, Yang Xu, Wenxuan Jia, Han Xu, Kai Xu,
Dongping Deng, and Xuan Song. The research life has never been boring because of them. In
iii

particular, I’m deeply indebted to Huachao for the stimulating discussions, for the sleepless nights
we were working together before deadlines, and for all the fun we have had in the last four years.
I believe that it will be an unforgettable time.
I also gratefully acknowledge the assistance of my master assistants, Jingfan Yang, Yulong
Yang, and Fangzhou Zhang. Sometimes they brought me some new fresh ideas based on their
knowledge background. Without their effort, I could not implement all the 3D printing experiments
by myself, and it significantly saved me a lot of time.
Special thanks to the Open Source community for high-quality resources that are developed
and maintained by people who do not ask for a payment, who respect my freedoms, and who are
generous with their time and energy. They provide me both the hardware and software solutions
to accelerate my research works.
My thanks also go out to the support I received from the USC Center for Advanced
Manufacturing. It provides us a comfortable working environment, which makes it possible to
construct various experimental setups at the same time, which, in a way, improves efficiency.
During this global COVID-19 pandemic, thanks should also go to those medical staff on the
front lines, which are protecting this world day and night so that I can safely write this dissertation
at home.
Last but not least, I would like to thank my family: my parents and my brother for supporting
me spiritually throughout writing this dissertation and my future wife for always believing in me
and encouraging me to follow my dreams.

iv

I dedicate this dissertation to
my parents, my brothers, sisters and my gf
for their constant support and unconditional love.
I love you all dearly.

v

TABLE OF CONTENTS
Acknowledgments........................................................................................................................... ii
List of Tables ................................................................................................................................ vii
List of Figures .............................................................................................................................. viii
Abstract ......................................................................................................................................... xv
Chapter 1: Introduction ................................................................................................................... 1
1.1 The definition of Hybrid Vat Photopolymerization ......................................................... 9
1.2 The overall structure of this dissertation ........................................................................ 10
Chapter 2: Literature Review ........................................................................................................ 11
2.1 Solutions for the multi-material 3D printing .................................................................. 11
2.2 Designs and materials for the support structures in 3D printing .................................... 16
2.3 Reduction of separation force in the SLA process ......................................................... 20
Chapter 3: Research Questions and Hypotheses ........................................................................... 23
3.1 Statement of Problems ................................................................................................... 23
3.2 Hypotheses ..................................................................................................................... 24
Chapter 4: Results and Evaluations .............................................................................................. 27
4.1 Hybrid Direct Ink Writing (DIW) and Stereolithography (SLA) for multi-material 3D
Printing ...................................................................................................................................... 27
4.1.1 Design of the hybrid multi-material printing process ............................................. 28
4.1.2 Multi-material diffusion and multi-exposure mechanism ....................................... 34
4.1.3 Modeling the hybrid 3D printing process ............................................................... 43
4.1.4 Results and Discussion ........................................................................................... 47
4.1.5 Experimental settings .............................................................................................. 61
4.1.6 Limitations and challenges ..................................................................................... 67
4.1.7 Summary of this work ............................................................................................. 68
4.2 Highly Removable Water Support for Stereolithography .............................................. 70
4.2.1 Water support building process study ..................................................................... 71
4.2.2 Ice surface level control .......................................................................................... 77
4.2.3 Experimental hardware and software systems ........................................................ 81
4.2.4 Results and discussion ............................................................................................ 84
4.2.5 Limitations and challenges ..................................................................................... 92
4.2.6 Summary of this work ............................................................................................. 93
4.3 A Vibration-assisted Method to Reduce Separation Force for Stereolithography ......... 95
4.3.1 Design of the vibration-assisted system .................................................................. 95
4.3.2 Analysis of experimental data and modeling ........................................................ 101
4.3.3 Modeling of the vibration-based separation process ............................................ 107
vi

4.3.4 Results and discussion .......................................................................................... 109
4.3.5 Limitations and challenges ................................................................................... 113
4.3.6 Summary of this work ........................................................................................... 114
Chapter 5: Conclusion and Recommendation for Future Research ............................................ 116
5.1 Answering the Research Questions/Testing Hypotheses ............................................. 116
5.2 Engineering Achievements and Scientific Contributions ............................................ 118
5.2.1 Engineering Achievements ................................................................................... 118
5.2.2 Scientific Contributions ........................................................................................ 120
5.3 Future work .................................................................................................................. 121
Bibliography ............................................................................................................................... 124

vii

List of Tables
Table 1: Factors and levels, study of ice impact to dimensional accuracy. ................................. 76
Table 2: Study of the impact of ice on dimensional accuracy. .................................................... 76
Table 3: The dimensions along the Z direction.......................................................................... 111

viii

List of Figures
Figure 1: Current popular additive manufacturing technologies and their corresponding
representatives................................................................................................................................. 2

Figure 2: Three main categories of the stereolithography process. ............................................... 4

Figure 3: An illustration of the cross-linking mechanism of photopolymers when exposed to UV
light. ................................................................................................................................................ 5

Figure 4: A DMD chipset with a micro-mirror array on it and the detailed structure of the
micromirrors. .................................................................................................................................. 6

Figure 5: An illustration of the classical MIP-SL process. a) a top-down MIP-SL process; b) a
bottom-up MIP-SL process. ............................................................................................................ 7

Figure 6: A big picture to illustrate the concept of Hybrid Vat Photopolymerization. ................. 9

Figure 7: Schematic of the 3D printer head used to produce continuous fibers using in-nozzle
impregnation based on FDM [89]. ................................................................................................ 11

Figure 8: Multimaterial multi-nozzle 3D printheads[95]. ........................................................... 12

Figure 9: Hardware system of the prototype multi-material MIP-SL system[69]. ...................... 12

Figure 10: A typical direct ink writing setup [103]. .................................................................... 14

Figure 11: Conductive ink is printed into an uncured elastomeric reservoir, which is capped by
filler fluid[19]................................................................................................................................ 15

Figure 12: The support material generated by the built-in 3D printing software for MakerBot
Replicator 2 [121]. ........................................................................................................................ 16

Figure 13: An illustration of adding necessary support structures for some overhanging structures.
....................................................................................................................................................... 17

Figure 14: Water Soluble Rinse-Away Material developed for the FDM process, by 3D
Systems[131]................................................................................................................................. 18

Figure 15: A schematic illustration of the CNC accumulation process[142]. ............................. 20

Figure 16: A two-channel system with PDMS[69]. .................................................................... 22

ix

Figure 17: Schematic of the CLIP printer where the part (gyroid) is produced continuously by
simultaneously elevating the build support plate while changing the 2D cross-sectional UV images
from the imaging unit. The oxygen-permeable window creates a dead zone (persistent liquid
interface) between the elevating part and the window[159]. ........................................................ 23

Figure 18: The overview of the designed experimental setup, including the multi-material DIW
printhead, which can move in the X-Y plane, the laser scanning based SLA module, which can
move in Z direction along with the tank above and the building platform which can move in X-Y-
Z direction. .................................................................................................................................... 28

Figure 19: Schematics of the hybrid DIW and SLA multi-material 3D printing system. a) First
step. A dispense syringe is moving horizontally towards the left to dispense the loaded material A
onto the bottom surface of a tank, which contains photocurable liquid base material. The syringe
nozzle tip is immersed into the base material and forms a layer-thickness gap to the bottom surface
of the tank. The scanning laser spot is tracing the dispensed material to solidify it while the syringe
is dispensing the newly coming out material ahead. b) Second step. Switch to another dispense
syringe to dispense material B. c) A region of interest (ROI) in b to illustrate the curing while
dispensing process in detail. d) Third step. Switch to the building platform to transfer the newly
built layer from the bottom surface of the tank to the previous fully cured layers with a second
light exposure. ............................................................................................................................... 29

Figure 20: Third step. Switch to the building platform to transfer the newly built layer from the
bottom surface of the tank to the previous fully cured layers with a second light exposure. An ROI
in d to illustrate how the current layer attaches to the last fully cured layers. .............................. 30

Figure 21: The schematic design of the pneumatic dispensing system, including connections of
the air tubes and signal wires. ....................................................................................................... 31

Figure 22: The corresponding logic control diagram of the designed low-cost pneumatic
dispensing system. Ax and Bx denote the electrical air valves, x=1,2,3…n. Dx denotes the
dispenser, x=1,2,3…n. The value n indicates the number of dispensers. ..................................... 32

Figure 23: Motion synchronization. a) Multiple syringe dispensers are mounted together on the
slider of the linear stage; b) Calibration of the relative distance between mounted syringe nozzle
tips in the X-Y plane. .................................................................................................................... 33

Figure 24: The workflow diagram of the printing process, from the digital CAD model to motion
control. .......................................................................................................................................... 34

Figure 25: Study the diffusion effect of the dispensed material trace in the liquid base material. a)
The width of the trace V.S. different delay time to cure and fix the material trace under different
printing settings, dispensing in resin / in air and applying 1 psi / 3 psi air pressure to extrude out
the material from the syringe. b) A region of interest in a, cure and fix the material trace within a
x

short delay time. c) The corresponding microscopic views from samples in a, group ④ and ⑥ are
top views of the dispensed traces under different printing settings, group ⑤ is the cross-sectional
views of the dispensed traces under different printing settings, all scale bars 0.1 mm. d) The
corresponding microscopic views from samples in b, group ①, and ③ are top views of the
dispensed traces under different printing settings, and group ② is the cross-sectional views of the
dispensed traces under different printing settings, all scale bars 0.1 mm. .................................... 37

Figure 26: Additional experimental results collected to study the diffusion effect. Four straight
lines, A, B, C, D, were printed using different printing parameters for comparison. A: the material
was dispensed in another resin, and the laser was used to cure it with a long delay time applied; B:
the material was dispensed in the air, and the laser was used to cure it with a long delay time
applied; C: the material was dispensed in another resin and the laser was used to cure it with a
short delay time applied; D: the material was dispensed in the air, and the laser was used to cure
it with a short delay time applied. The air pressure applied to the dispensers is 3 psi in the
experiments. .................................................................................................................................. 40

Figure 27: The multi-exposure mechanism to ensure the bonding between layers. a-c) Double
exposures are applied to the photocurable materials for multi-material 3D printing; a) First UV
light exposure is applied to cure the dispensed material trace into a gel-like state; b) Move down
the printing platform and apply a second exposure to fully cure the current layer and further attach
to the previous cured layers; c) Move up the printing platform to detach the part from the bottom
film; d-g) Triple exposures are applied to the process when non-photocurable materials are
involved; d) First exposure is applied to cure a thin layer of base material in the bottom of the
reservoir; e) The non-photocurable material is dispensed onto the previous cured thin layer and a
second exposure is applied to generate a gel-like boundary wall around the dispensed trace; f)
Move down the printing platform and apply a third exposure to fully cure the current layer and
further attach to the previous cured layers; g) Move up the printing platform to detach the part
from the bottom film. .................................................................................................................... 42

Figure 28: Modeling the hybrid 3D printing process. a) An inset of Gaussian function for the
energy distribution of the scanning laser spot; b) Illustration of the light-curing model for the
photopolymer; c) Region of interest in (d), indicating a scanning laser spot is curing the liquid
material while the syringe nozzle is dispensing the material; d) A laser spot is following the
material trace dispensed by the syringe dispenser. ....................................................................... 47

Figure 29: 3D printed multi-material crystal ant, amber. a) Input 3D digital CAD model. b) Sliced
layers of the digital model. c) A certain sliced layer picked from the sliced layers in (b). d) Region
of interest in (c), the concentric G-Code syringe toolpath, and the scanning laser path. ............. 48

Figure 30: Visualization of the G-Code toolpath......................................................................... 49

Figure 31: Run the G-Code toolpath to dispense the red resin in the transparent resin. The laser
spot follows the same toolpath to cure the red resin trace dispensed. .......................................... 49
xi

Figure 32: Transfer the newly semi-cured layer on the bottom surface of the tank to the previous
fully cured layers by scanning the surrounding transparent base material in a square shape for this
case. ............................................................................................................................................... 50

Figure 33: The imperfect flatness on the top surface of the current building layer. a) The material
dispenser is dispensing the red material into the liquid base material reservoir following by a
scanning laser; b) SEM image of the top surface of the current building layer with interlaced
highland and valley. ...................................................................................................................... 51

Figure 34: An illustration of the sandwiched liquid base material serving as a “glue” to tightly
bond the current layer with the previous layers when a second exposure is applied. a) A second
laser exposure is applied to transfer the current layer to the previous cured layers. b) A schematic
cross-sectional view of the sandwiched “glue” layer. .................................................................. 51

Figure 35: a) Cured layers stick on the building platform. b) The SEM image of the bottom surface
of the cured layers, scale bar 0.1 mm. ........................................................................................... 52

Figure 36: Printed multi-material crystal ant. a) Top view of the built part, a red ant with
surrounding cured transparent resin, scale bar 5 mm. b,d) Microscopic views of the corresponding
detailed features in a, scale bars 0.5 mm. c) Back view of the built part. e) Side view of the built
part. ............................................................................................................................................... 53

Figure 37: The cross-sectional view of the built part, scale bar 0.1 mm. .................................... 53

Figure 38: a) Printed ant picture from top view for distance measurements. b) lines were drawn
on the screenshot of the original ant STL file for distance measurements. c) The length obtained
from the measurements. ................................................................................................................ 54

Figure 39: 3D printed multi-material tensile bars. a) Designed and printed tensile bars of various
stiffness using interwoven structures of stiff material. Red color indicates the stiff material applied,
and the grey color indicates the soft/elastic material applied. 0%, 25%, 50%, 75%, and 100%
percent of the stiff material are applied correspondingly in the designed tensile bars. b)
Corresponding tensile test results for different combinations of elastic material and stiff material.
....................................................................................................................................................... 56

Figure 40: 3D printed multi-material wheel with different stiffness. a) Input 3D digital CAD
model. b) Sliced layers of the digital input model in (a). c) Multi-material G-Code toolpath. d)
Region of interest in c, the rim is applied with the stiff material, and the tire is applied with the soft
material. e) Printed multi-material wheel in a single part. f-h) The wheel under compression. ... 57

Figure 41: Schematic design of the wearable sensors on the human hand. ................................ 57

xii

Figure 42: a) The 3D CAD design of the strain sensor. b) The schematic diagram of the circuits
to measure the capacitance change of the designed capacitive buttons. c) Wires are inserted into
the printed part to connect with the conductive ink. ..................................................................... 58

Figure 43: a) The 3D CAD design of the capacitive sensor. b) The schematic diagram of the
circuits to measure the capacitance change of the designed capacitive buttons. .......................... 59

Figure 44: The corresponding conical nozzle tip used to dispense the conductive ink. .............. 60

Figure 45: A plot of the resistance change for a strain sensor fabricated by the process. ........... 60

Figure 46: A plot of digital capacitance changes for a capacitive sensor fabricated by the process.
....................................................................................................................................................... 61

Figure 47: Using the ImageJ line tool to measure the length of legend. a) Using a line measurement
tool to measure multiple material widths from the microscopic images for material printed in resin
and cured with a large delay. The white line indicates the measured length, the black line is the
current measuring line; b) Using line measurement tool to measure multiple material widths from
the microscopic images for material printed in resin and cured with a small delay; c) Line
measurement tool is used to measure the vertical height of a cross-section in the air; d) Line
measurement tool is used to measure the vertical height of a cross-section in resin; e) Setting up
the pixel scale before the measurement. ....................................................................................... 62

Figure 48: Selection of the substrate for the proposed hybrid 3D printing process. a) Testing the
stability when the target material is dispensed on the surface of different substrates, all the scale
bars are 3 mm; b) The corresponding data measurement, width ratio (maximal trace width/minimal
trace width) versus time; c) SEM images of the different substrate surfaces, scale bars are 100 μm.
....................................................................................................................................................... 64

Figure 49: Amber produced by nature. ........................................................................................ 70

Figure 50: The building platform for the process. ....................................................................... 72

Figure 51: A schematic illustration of the water support building process. ................................ 73

Figure 52: Flow chart description of the water support building process. .................................. 74

Figure 53: A proof of concept. (a) A designed CAD model; (b) and (c) support structures that are
required in the traditional SLA process; (d) printing in process using water support; (e) and (f)
melting the surrounded ice after the building process; and (g) - (i) the fabricated object in different
views. ............................................................................................................................................ 74

xiii

Figure 54: An illustration of the building height stretch problem. (a) Residual water droplet left
on the building surface; (b) Iceball pushes up tool B. .................................................................. 78

Figure 55: A water vacuum design. (a) The CAD model of the designed water vacuum; (b) the
fabricated water vacuum; and (c) the internal structure of the vacuum design. ........................... 79

Figure 56: Illustration of the efficiently designed water vacuum. ............................................... 80

Figure 57: (a) A test case of the designed CAD model to be built; (b) a top view of the built layers
with ice at the initial few layers; and (c) a top view of the built layers with ice in the middle of the
CAD model. .................................................................................................................................. 81

Figure 58: The difference between using a bounding wall and without using the bounding wall.
(a) Top view of the rectangular ice boundary when the bounding wall is applied; (b) top view of
the irregular ice boundary without using the bounding wall; and (c) - (d) cross-section views of
with and without a bounding wall, respectively. .......................................................................... 81

Figure 59: The prototype hardware system for the water support building process.................... 82

Figure 60: The electrical architecture of the hardware system. ................................................... 83

Figure 61: Software system user interface. .................................................................................. 83

Figure 62: The building process and result of the 3D ant. ........................................................... 84

Figure 63: The built 3D and after removing ice in a microscopic view. ..................................... 85

Figure 64: The built 3D bee after removing ice in the room environment. ................................. 86

Figure 65: (a) CAD model of a bee with traditional support structure; (b) printed part using 100
μm layer thickness; (c) printed part using 50 μm layer thickness. ............................................... 86

Figure 66: An empty channel fabricated by the water-support-based SLA process and its
microscopic cross-sectional views in different sections. .............................................................. 88

Figure 67: A Y-type channel fabricated by the water-support-based SLA process and its
microscopic cross-sectional views in different sections. .............................................................. 88

Figure 68: a) and b) Different views of the printed single spring; c) The printed single spring in
microscopic views. ........................................................................................................................ 89

Figure 69: The ice support for the single spring was melting in the room temperature (23 ° C) after
the completion of the printing. ...................................................................................................... 90

xiv

Figure 70: Description of the characteristics for the spring. ........................................................ 91

Figure 71: The double helix spring. ............................................................................................. 91

Figure 72: The inclusive spring. .................................................................................................. 92

Figure 73: The hardware setup of the vibration-assisted system. ................................................ 96

Figure 74: The illustration of the separation force measurement. ............................................... 98

Figure 75: A schematic of using a pressure force sensor. ........................................................... 98

Figure 76: (a) weights added to calibrate the sensor; (b) using screws as standard weights. ...... 99

Figure 77: Calibration results for the corresponding force sensors. (a) left force sensor calibrated;
(b) right force sensor calibrated. ................................................................................................. 100

Figure 78: CAD model used in our experiments. ...................................................................... 101

Figure 79: The flow chart of the designed printing process. ..................................................... 102

Figure 80: (a) the raw data obtained from the left sensor; (b) the raw data obtained from the right
sensor. ......................................................................................................................................... 103

Figure 81: Data visualization. (a) the calculated separation force over time; (b) a zoom-in view
of the turning point...................................................................................................................... 104

Figure 82: A comparison of the separation force using different printing processes. ............... 105

Figure 83: The screenshots of a recorded normal separation process. ...................................... 110

Figure 84: The screenshots of a recorded vibration-assisted separation process. ..................... 110

Figure 85: The microscopic views of the printed part in different sections. ............................. 112

Figure 86: (a) the designed CAD model for the test; (b) part printed by the vibration-assisted
printing process; (c) a close view of the small features in part (b); (d) part printed by the normal
printing process; (e) a close view of the small features in part (d). ............................................ 113

xv

Abstract
Stereolithography Apparatus (SLA) is an additive manufacturing (AM) process in which
liquid photopolymer resin is cross-linked and converted to hardened plastic. As a light-induced
layer-based manufacturing process, there are several fundamental problems we are facing now
limiting the potential of the SLA process due to its nature of fabrication principle, which seems
unavoidable in the traditional SLA process. Hybrid Vat Photopolymerization (Hybrid VPP) is the
smart integration of technologies or materials in other fields to address the current limitations in
the conventional Vat Photopolymerization process to make it more efficient and more widely used.
Firstly, SLA can achieve high resolution and fabricating intricate parts at a fast speed.
However, material contamination issues caused by the nature of the vat-based process hinder the
development of multi-material SLA 3D printing.
Secondly, the layer-based fabrication manner means the fabrication is a unidirectional process
during which the building part can only grow in a single direction. Thus, some overhanging
structures will occur in some layers due to the geometry complexity. Therefore, a significant
amount of additional support structures is required to be added to the fabricated structures for an
intricate computer-aided design (CAD) model to ensure it can be successfully fabricated. However,
these support structures may be difficult to remove. Even worse, the removal of the support
structures may cause unexpected damage to delicate features and undesired surface quality.
Thirdly, it is so essential to control the constant flatness of each layer in the layer-based SLA
process since the layer flatness will significantly affect the attachment between two adjacent layers
and the dimensional accuracy of the fabricated part. It has been a popular way to add a constrained
surface to ensure the surface flatness in the newly cured layer. However, for a constrained surface-
based top-down or bottom-up SLA process, it will be challenging to separate the building layers
xvi

from the constrained glass due to the near-vacuum environment between the built layer and the
constrained glass. The separation force will increase when the contact area increases.
In this research, we propose three hybrid 3D printing processes aiming to address the
aforementioned limitations in the current traditional SLA process. Hybrid 3D printing for
stereolithography is the smart introduction and integration of technologies or phenomena in other
fields to make current stereolithography more efficient and more widely used. It connects
something that seems unrelated together to achieve enhanced performance, which is impossible
for the conventional SLA process.
When it comes to multi-material 3D printing, a hybrid 3D printing method by incorporating
the advances from Direct Ink Writing (DIW) and SLA processes is presented. The proposed
method uses syringes to dispense the liquid-like materials inside another photocurable base
material. It exploits the Galvano mirror-controlled laser beam to cure it or its surrounding material
by following the dispensed material trace, thus retaining the shape after the material leaves the
nozzle. As exemplars, a 3D ant with intricate features inside the transparent polymer and a tire
with selectively printing stiff and soft materials are fabricated. The proposed method is versatile
to build heterogeneous materials with various functionalities for real-world application.
To address the limitation of the current support structures in the SLA process, a novel SLA
process using highly removable and widely available water as supports is presented. The process
uses solid ice to surround the built parts in the layer-by-layer fabrication process. A cooling device
is used to freeze the water into ice for each layer. The photocurable resin is spread on the ice
surface and then solidified by a projected mask image. Accordingly, a complex 3D object can be
fabricated without using traditional support structures. After the fabrication process has been
xvii

finished, the additional ice structure can be easily removed and leave no undesired marks on the
bottom surfaces. Several test cases are presented to show the effectiveness of the presented method.
As for the separation force problem in the constrained surface-based SLA process, a novel
separation approach for the SLA process utilizing vibration-assisted glass is presented. A pair of
general loudspeakers were used to provide a low-frequency vibration for the constrained glass
used in the SLA process. Two force sensors were used to measure the separation force in real-time.
Controlled experiments were implemented, and the corresponding data were collected and
analyzed. The analyzed results have demonstrated that the proposed method can significantly
reduce the separation force for the SLA process.
The future work includes: 1) explore the possibility of the presented hybrid multi-material 3D
printing method in the fabrication of advanced materials for biomedical applications. 2) investigate
the potential solutions to solve the issue of frost concentration happened on the building surface
during the ice support fabrication process, and explore more potential functional applications that
can benefit from the proposed ice support which are impossible for the traditional SLA process to
achieve in the past, such as the microfluidics. 3) develop an adaptive method to dynamically
change the optimal vibration frequency and other parameters that are associated with the change
of the cross-sectional area to protect the small features during the separation.
The proposed Hybrid VPP is a new systematic strategy to address the limitations in the
traditional SLA process, which opens a door towards the development of wearable electronics,
soft robotics, and tissue engineering.

1

Chapter 1: Introduction
Additive manufacturing (AM)[1-4] , commonly known as three-dimensional (3D) printing[5],
involves the whole process of making 3D solid objects from Computer-aided Design (CAD) files
(e.g., a .STL file)[6, 7]. The two terms, additive manufacturing, and 3D printing are usually
interchangeable these days. For the sake of simplicity, we will use 3D printing in the following
sections.
3D printing technology has been a revolutionary way to fabricate the entire physical parts in
the past several years, which enables us to produce complex functional shapes using less material
than traditional manufacturing methods. The objects that can be fabricated by 3D printing are
tremendous and continue to become more ambitious. Applications include rapid prototyping,
architectural scale models, 3D printed prosthetics, and movie props, which have been benefiting
many areas in recent years. It can print anything from simple toys in small scales to general tools
to architectures[8, 9] in large scales. People can also apply this technology to tissue
engineering[10-14], soft robotics[15-18], wearable devices[19-22], and biomedical devices, such
as hearing aids, dentures, and microfluidics for lab-on-a-chip purposes[23-26]. More recently, it
has been demonstrated that it is even possible to 3D print a living human organ[27].
3D printing is the opposite of a subtractive manufacturing process, which is cutting out a piece
of metal or plastic with a milling machine[28]. The creation of a 3D printed object is achieved
using additive processes. In an additive process, an object is created by laying down successive
layers of material until the object is created. Each of these layers can be seen as a thinly sliced
horizontal cross-section of the eventual object.
2

Figure 1: Current popular additive manufacturing technologies and their corresponding representatives.

3

According to the different processes used, 3D printing can be classified into seven main
categories: vat photopolymerization[29-32], material extrusion[33-37], material jetting[38-40],
binder jetting[41-43], powder bed fusion[44-47], direct energy deposition[48-50] and sheet
lamination[51]. Figure 1 shows the current several popular 3D printing technologies and their
corresponding representatives.
Among all the 3D printing processes, the SLA process[52-57] has been recognized as one
of the most accurate 3D printing technologies in the world so far. Recently, the standard term to
call the SLA process has been changed to Vat Photopolymerization. For the sake of simplicity, we
will use term SLA in the following sections. In the very early 1980s, people first used this initial
process as an affordable way to create prototypes for product development within specific
industries, and people usually call it Rapid Prototyping (RP) technologies[58]. The first patent of
stereolithography can be traced back to 1986 when an American inventor, Charles (Chuck) Hull,
invented the first SLA 3D printer[5]. As the first commercialized 3D printing technology, the SLA
process has attracted more and more attention as the current industry widely adopts it due to its
high geometry accuracy and low cost. Still, it continues to grow in potential markets such as
jewelry, dentistry, automobile, and biomedical devices. SLA process can be further classified into
three sub-categories according to the light source used in the process: laser beam scanning based
SLA process[59-61] and digital light processing (DLP) based SLA process[62-71] and liquid
crystal display (LCD) based SLA process[72-74] as shown in Figure 2. All of them use light to
induce polymerization of photocurable resins. But the way how they apply this principle is what
sets them apart.
4

Figure 2: Three main categories of the stereolithography process.
Over the past three decades, the SLA process evolves a lot along with the development of
technologies in other fields. A lot of research efforts have been made to improve the performance
of the SLA process, in terms of fabrication speed[68], building resolution[69], dimensional
accuracy[66], post-processing, material properties[75], etc.
The typical materials used in the SLA process are photopolymers, and the corresponding
chemical reaction happens in the SLA process is called photopolymerization. Photopolymers[76]
are light-sensitive polymer materials that change their properties when exposed to ultraviolet (UV)
light. The water-like liquid substance can be solidified to the solid plastic-like substance. Only the
area exposed to UV light hardens, whereas unexposed parts will remain like liquid. The schematic
illustration of the photopolymerization process is shown in Figure 3. Photopolymerization[77] is
defined as a synthesis of polymers by chain reactions that are initiated upon the absorption of light.
Light only serves as an initiating tool, and it does not interfere with the propagation and termination
stages of the chain process. Different combinations of the monomers, oligomers, photoinitiators,
and various other additives that form a polymer resin result in different material properties.
5

Figure 3: An illustration of the cross-linking mechanism of photopolymers when exposed to UV light.
The fundamental photo polymerization follows the Beer-Lambert law[78]. Given the light
intensity 𝐼 , and light exposure time 𝑡 , the curing depth 𝐶 𝑑 is determined as[79]:
C
d
= 𝐷 𝑝 ln ⁡(
𝐼 ⋅ 𝑡 𝐸 𝑐 ) (1)
𝐷 𝑝 is penetration depth, and this term is related to material property, and is derived as:
D
p
=
1
ℎ
(2)
Hence, the polymerization law can be rewritten as:
C
d
=
1
ℎ
ln ⁡(
𝐼 ⋅ 𝑡 𝐸 𝑐 ) (3)
However, the light-induced fabrication mechanism in the SLA process limits the selection of
available materials that can be applied in the current SLA process. The current photopolymer resin
can produce parts with a quite smooth surface finish and preserve the fine features which look like
the injection mold-like quality. Despite the impressive appearance and surface quality of the parts
that can be fabricated by the traditional SLA process, the printed parts may not be suitable for the
general use due to its low impact strength and the relatively brittle mechanical property.
Furthermore, the material properties may even change over time.
6

The recent development of Micro-Electro-Mechanical Systems (MEMS), such as Digital
Micromirror Device (DMD)[80], provides the capability of accurately and selectively controlling
the energy input of the target area which makes it possible to develop the novel mask-image-
projection-based stereolithography process (MIP-SL). As shown in Figure 4, a typical DMD chip
is a reflective spatial light modulator that can dynamically control the direction of the reflected
light on each micromirror, thus turning on/off the light in a single pixel in a macro view.

Figure 4: A DMD chipset with a micro-mirror array on it and the detailed structure of the micromirrors.
A typical DMD chip consists of a mirror array of up to 2 million, individually controlled,
highly reflective aluminum micromirrors, the status of which can be switched at a very high
frequency (can be up to 16,000 Hz)[80]. It enables end-users to program high-speed, efficient
patterns of light. The design of a single unit of the micromirror is shown in Figure 4. The
micromirror can be tilted ± 10°, which can reflect the exposed light either onto or away from the
screen. The DMD’s micro-mirror array is optically efficient from 350 nm – 2500 nm depending
on the differently designed capability of the DMD chipset. The light intensity can be controlled by
Pulse Width Modulation (PWM), which is the principle to realize the grayscale for a projection
image.
Powered by the advanced DMD technology, the basic idea of the MIP-SL process is using
mask images sliced from an input 3D CAD model to directly solidify the liquid photopolymer in
7

a desired 2D pattern in a layer-based manner. An illustration of the classical MIP-SL process is
shown in Figure 5. Basically, the conventional MIP-SL process can be further classified into two
main sub-categories: the top-down and bottom-up MIP-SL processes depending on where the
newly cured layer is or where the projection direction is.

Figure 5: An illustration of the classical MIP-SL process. a) a top-down MIP-SL process; b) a bottom-up MIP-SL
process.
When the building process begins, the DMD chip driven by the DMD controller will
project the mask image of a sliced layer onto the liquid resin interface to selectively solidify it at
the desired regions. After one layer is finished, the building platform will move to the next position
to form a thin gap between the newly solidified surface and the liquid resin surface where the next
layer will begin. The process repeats until the desired 3D object has been complete. Compared to
the laser-based SLA process, the whole layer of the selected liquid resin can be solidified by just
one single exposure shot in the MIP-SL process instead of scanning the tool path of the image
pattern by a laser beam. Hence, the entire building speed is relatively faster for a MIP-SL process
than the laser-based SLA process.
8

Due to the layer-based SLA fabrication process, it is so essential to control the constant
flatness of each layer since the layer flatness will significantly affect the attachment between two
adjacent layers and the dimensional accuracy of the fabricated part. It has been a popular way to
add a constrained surface to ensure the surface flatness in the newly cured layer, as shown in
Figure 5. However, for a constrained surface-based top-down or bottom-up SLA process, it will
be challenging to separate the building layers from the constrained glass due to the near-vacuum
environment between the built layer and the constrained glass. The separation force will increase
when the contact area increases.
On the other hand, the layer-based fabrication manner means the fabrication is a
unidirectional process during which the building part can only grow in a single direction. Thus,
some overhanging structures will occur in some layers due to the geometry complexity. Therefore,
a significant amount of additional support structures is required to be added to the fabricated
structures for an intricate computer-aided design (CAD) model to ensure it can be successfully
fabricated. However, these support structures may be difficult to remove. Even worse, the removal
of the support structures may cause unexpected damage to delicate features and undesired surface
quality.
Despite the advancement of the SLA process among other 3D printing processes, several
research problems remain for the SLA process over the past three decades which significantly
hinder the development of the SLA process, such as the problems caused by the material
contamination issue in the traditional multi-material SLA printing process, the conventional
support structures and separation problems for the large area constrained surface-based MIP-SL
process. It is urgent and desirable to address these problems and explore the possibility of the SLA
process in future applications. In this dissertation, several Hybrid VPP approaches based on the
9

traditional SLA process have been proposed to connect seemly unrelated things and address the
problems from different perspectives.
1.1 The definition of Hybrid Vat Photopolymerization
Hybrid Vat Photopolymerization (Hybrid VPP) is the smart integration of technologies and
materials in other fields to address the current limitations in the traditional Vat
Photopolymerization process to make it more efficient and more widely used as shown in Figure
6.

Figure 6: A big picture to illustrate the concept of Hybrid Vat Photopolymerization.
As we can see from Figure 6, there are three categories that can contribute to the Hybrid VPP.
It is known that the traditional Vat Photopolymerization or SLA has some limitations in dealing
with some specific problems due to its light-induced and layer-based natural fabrication principle.
By integrating the technologies or materials from other domains with the traditional Vat
Photopolymerization process, it opens the door towards the new possibilities. It could combine
10

multiple advantages together into one single process and make up each other’s shortages. It could
also potentially address the current limitations for the traditional SLA process, which is the
motivation of my research topic.
1.2 The overall structure of this dissertation
The following sections are organized as follows. Section 2 will briefly review the related
literature for the proposed research work. Several current research problems will be stated in
Section 3, and the corresponding hypotheses to the aforementioned research problems. Some
research results and evaluations will be presented in Section 4, and Section 5 concludes the
dissertation with some remaining future work.

11

Chapter 2:Literature Review
This section discusses state of the art in 3D printing processes, with emphasis on innovative
methods proposed by other researchers aiming to improve the performance of the current 3D
printing processes, in terms of the solutions for the multi-material 3D printing process, the various
designs and materials used for the support structures in the different 3D printing process and the
approaches adopted to reduce separation force in the MIP-SL process.
2.1 Solutions for the multi-material 3D printing
Although there is a growing trend for AM techniques to fabricate composite materials[81-85],
most current 3D printing processes can only deal with isotropic materials, which hinders the
development of advanced materials. As for each printing process, many previous attempts have
been made to achieve multi-material 3D printing aiming to fabricate composite materials for both
visual and functional purposes. For example, the FDM process can build multi-color parts by
extending extra printing nozzles and composite material with enhanced mechanical properties by
embedding continuous carbon fiber into the extruded filament[86-91] as shown in Figure 7. Like
the FDM process, DIW is an extrusion-based process to dispense materials from the nozzle tip,
which is ideal for processing liquid-like material[92-94].

Figure 7: Schematic of the 3D printer head used to produce continuous fibers using in-nozzle impregnation based
on FDM [89].

12

Y-type channels have been designed to mix different materials for the multi-material DIW process
by dynamically controlling the switching frequency of the channel[95], as shown in Figure 8.

Figure 8: Multimaterial multi-nozzle 3D printheads[95].
MJP process can print multi-material by jetting multiple photocurable materials in the micro-liter
level simultaneously onto the building surface layer by layer. In contrast, it is only limited to print
some materials with low viscosity due to the nature of the jetting mechanism[39, 96]. SLA process
can fabricate digital material by switching the building part between different material
reservoirs[97-99], as shown in Figure 9.

Figure 9: Hardware system of the prototype multi-material MIP-SL system[69].
13

As the first commercialized additive manufacturing technology, SLA has been recognized as
a high-resolution 3D printing method, among others, due to its fabrication nature. Currently, two
different general SLA technologies are existing in the industry, which are distinguished by the way
how the light is exposed to the photocurable material, including the laser-scanning based SLA and
mask-image-projection (MIP) based SLA[68, 100]. The laser-scanning based SLA was initially
invented as the sliced image pattern from an STL file when converted into scattering points where
the scanning laser spot will travel through[101]. The mask-image-projection based SLA appeared
along with the development of DMD technology in which the whole image pattern is projected to
the photocurable material in a single shot[102]. Compared to MIP SLA, laser-scanning based SLA
has more dynamic control flexibility in points-based toolpath, while MIP SLA can print parts at a
faster speed. When it comes to multi-material 3D printing, some previous works have been
reported by switching the building part between different material reservoirs[97-99]. Due to the
material contamination issue, multiple additional sequential procedures have to be involved during
the printing process to address this problem, which includes rinsing the uncured material attached
on the surface of the cured part and drying the surface afterward which makes the whole building
process quite tedious. Meanwhile, a significant amount of material will also be wasted during the
cleaning process in each layer.
DIW is an extrude-on-demand 3D printing method. A typical DIW dispenser includes a
syringe that contains the liquid-state material, a syringe nozzle where the loaded material will be
ejected from, and a pneumatic controller which controls the air pressure applied to the syringe to
push out the loaded material as shown in Figure 10.
14

Figure 10: A typical direct ink writing setup [103].
Currently, several techniques have been incorporated to advance the DIW to fabricate high-
performance parts using various materials such as bioinks[104, 105], conductive inks, and
magnetized fluids[19, 106-110]. Multiple individual syringes have been involved in a single setup
to realize selectively dispensing various materials in the desired position[111, 112]. In addition,
Y-type channels and fluidic mixing devices have been designed to mix various materials before
the mixed fluid exits the nozzle tip for multi-material 3D printing[95, 113, 114]. Although DIW is
capable of printing liquid-state materials with different viscosity, which makes it quite attractive
for multi-material 3D printing, there is a critical challenge that most DIW processes will meet.
Unlike the thermoplastic materials commonly used in the FDM process in which the molten
filament-state material exits the nozzle will keep its cylindrical shape due to a sudden temperature
change between the heated nozzle and the air to cause a quick morphological transition from the
molten state to the solid state, it is difficult for a DIW process to retain the shape of the liquid-state
material when it comes out of the nozzle tip since fluid is freeform and dynamic, especially when
the material is dispensing around the boundary of the building layers. As a result, material will
flow down to the previous layers which will eventually get an undesired geometric shape[114].
15

Some recently published works have studied the rheological behavior of the dispensed material to
retain the shape of the dispensed material and reported that the shape of the dispensed material can
be well maintained by using a syringe nozzle tip to dispense desired material inside another gel-
like base material[19, 115, 116]. The surrounding gel-like material can prevent the desired material
from collapsing due to the surface tension of the material exposed in the air. Meanwhile, the
surrounding gel-like material can work as supporting material to make it possible for a DIW
process to dispense material in a 3D space rather than just on a 2D plane[117, 118] as shown in
Figure 11.

Figure 11: Conductive ink is printed into an uncured elastomeric reservoir, which is capped by filler fluid[19].
However, despite its advances in dealing with shape-retaining and freestanding structures, the
success of the whole process highly relies on the material compatibility between the dispensed
material and the base material. For example, the density of the desired material must be close to
the density of the base matrix material. Otherwise, the dispensed material inside the base material
will float up or sink in the base material during the printing process, which will ruin the desired
structure. It is also noticed that it takes time for the entire dispensed material traces to be solidified
after exiting the nozzle tip, and they are still floating in the base material. Any disturbances caused
by the moving nozzle will eventually result in an inaccurate desired structure.
16

2.2 Designs and materials for the support structures in 3D printing
Layer-based AM processes can directly fabricate parts from CAD models. As a direct digital
manufacturing approach, current AM processes can effectively build extremely complex three-
dimensional (3D) shapes that used to be impossible to be made[4]. However, for most CAD models
with complex geometries that have a lot of overhanging features, plenty of extra support structures
are needed in the layer-based fabrication process[119, 120], as shown in Figure 12.

Figure 12: The support material generated by the built-in 3D printing software for MakerBot Replicator 2 [121].
It is well-known that the 3D printed parts have a rough surface with staircase defects due to
the layer-based printing mechanism. These staircase defects can significantly reduce bonding
strength and the fatigue life of the printed parts[122]. Also, this rough surface is unacceptable in
many applications, such as optics[123]. As shown in Figure 13, typically, where structures have
an overhanging angle that is bigger than 45 °, there it will generate some necessary support
structures underneath the overhanging features during the slicing stage in the software[121].
However, the support structures not only significantly increase the printing time (for specific 3D
printing processes) and the material waste but also need substantial post-processing time.
Moreover, removing the support structure will introduce defects and leave undesired marks on the
surface.
17

Figure 13: An illustration of adding necessary support structures for some overhanging structures.
Besides the above limitations, current layered 3D printing processes have difficulties in
printing 3D curves and shells. Recently, 3D curves have attracted much attention for its pivotal
applications but lack of effective fabrication methods. For example, 3D circuits[124-126] are a set
of 3D curves for the conductive traces. It is neither efficient nor effective using the current 3D
printing method to print these curves. The conventional 3D printing processes are invented for
printing 3D solid objects and behave poorly in printing 3D curves, especially when these 3D curves
require smooth surface quality. Another example of a 3D curve is the 3D antenna[127]. Shells are
another critical but challenging application for conventional layered 3D printing.
Since the first 3D printing process was invented in the 1980s, a lot of research effort has
been made trying to eliminate or reduce the support structures to save building time and cut down
the extra material cost. Different 3D printing processes have different innovative ways to deal with
this problem.
For the fused deposition modeling (FDM) process, Goyanes et.[128, 129] reported that
Polyvinyl alcohol (PVA), a water-soluble synthetic polymer with a molecular formula of
(C2H4O)n), as an extruded commercial filament from Makerbot Inc., can be used to print support
structures by choosing multi-material printing mode[130]. After the printing job is done, the
18

fabricated part along with the water-soluble support material will be put into the regular water for
a certain time until all the water-soluble material is dissolved in the water as shown in Figure 14.

Figure 14: Water Soluble Rinse-Away Material developed for the FDM process, by 3D Systems[131].
People also used wax material as the support structures in the multi-jet modeling (MJM)
process[39, 132, 133]. After the printing job is done, post-processing, mainly using a water gun,
will be carried out to remove wax-like supporting structures around the desired parts.
Consequently, removing support structures becomes easier, and it is less likely to damage the built
parts when removing the support structures. However, the wax is a kind of toxic material. It is also
an expensive method to use wax as a sacrificial material since the wax for 3D printing, especially
for the MJM process, is usually costly. Besides, removing wax for the post-processing is not an
environmentally friendly way, rather a messy approach.
From the design perspective, Paramita Das et. [134]proposed an optimization model trying
to identify the best building orientations for minimizing the support structures while meeting
design tolerance. On the other hand, the building orientation of a part is a crucial process parameter
which affects part quality. It will introduce undesired geometric dimensioning & tolerancing errors
to the part. It could be a tradeoff between the desired part quality and the reduction of the support
structures.
19

For the powder-based 3D printing processes, such as Selective Laser Sintering (SLS)[135]
and Selective Laser Melting (SLM)[136], the support structures are the powder particles
themselves in the powder bed since the condense powder particles always surround the building
part. However, this is not always a support-free process. Due to the relatively loose powder
particles surrounded and the gravity of the overhanging structures, the built overhanging structures
will have a chance to sink down to the powder bed if there are no any support structures below
them.
Nevertheless, when it comes to the SLA process, it seems there exists a research gap here.
For the SLA process, the building process happens in a reservoir which contains a certain amount
of liquid resin. Hence, it will be challenging to implement a multi-material process to involve
another sacrificial material as the support structures like what other methods did. However, some
previous efforts have been made to achieve a multi-material SLA process[69, 137-139].
As for the non-layer-based 3D printing processes, support structures can be eliminated due
to its printing nature. Dai et.[140] also presented a new method based on the FDM process to
fabricate 3D models on a robotic printing system equipped with multi-axis motion, which allows
us to use curved tool paths to fill the material inside the volume so that supporting structures can
be avoided. Gupta et al.[141] also presented a multi-axis printing process using 6-DOF robotic
arms. In our previous work, Chen et al.[142] presented a non-layer-based 3D printing process
called CNC accumulation (CNC-A) without using additional support structures, as shown in
Figure 15. Song et al.[143] proposed a multi-axis FDM printing process and printed a 3D curve
on a glass bottle’s surface. Mao et al.[144] further extends the point-based accumulation to a line-
based accumulation approach so that it can print 3D objects efficiently under the resin. Pan et
al.[145] also demonstrated that point-based accumulation tools could have various sizes and shapes.
20

Highley et al.[116] can extrude material in the slurry-like substrate, which acts as a support
material.

Figure 15: A schematic illustration of the CNC accumulation process[142].
In short, support structures do play an essential role in the success of 3D printing objects
with overhanging features. Without generating the necessary support structures for the desired
parts, the completion of the desired results cannot be guaranteed. On the other hand, the support
structures will also result in the flaws of the parts since they are just the sacrificial structures and
do not come with their original design. It takes significant time and effort to remove them in the
post-processing stage, in which three is a chance that removal of the support structures will cause
unexpected damage to the original parts resulting in an unsmooth surface.
2.3 Reduction of separation force in the SLA process
As the first commercialized AM technology, the SLA process is one of the most popular
AM technologies. In the SLA process, the liquid photosensitive polymer is solidified by a pattern
controllable irradiation light source such as a digital light processing (DLP) projector or a laser
beam[58, 146, 147]. Compared to other polymer AM techniques such as extrusion or jetting
processes, SLA produces parts with higher accuracy, the best surface quality, and faster building
speed[68, 69, 148-151]. Basically, there are two types of SLA processes according to different
resin filling mechanisms: free surface method and constrained surface method[148]. For the free
21

surface method, the polymer is directly exposed to irradiation light and solidified. As for the
constrained surface method, the top-down or bottom-up approach is used to build a part. The
photosensitive polymer is always sandwiched between the building surface and the constrained
surface. In contrast to the free surface method, the platform needs to move up and down to let the
polymer refill into the gap between the building surface and the constrained surface.
Compared to the free surface method, the constrained surface process has several advantages,
such as higher vertical dimensional accuracy, higher material filling rate, and faster building speed,
etc. More importantly, by controlling the moving distance of the platform, part with the thinner
layer can be achieved. However, there are still many problems existing for this constrained surface
method. It is challenging to separate the newly cured layer from the constrained surface. This
difficulty results from an adhesive bonding that is developed between the newly cured layer and
the constrained surface[152]. Thus, a separation force is needed to separate the built part from the
constrained surface. Such a separation force will significantly affect the printing speed, reliability
of the printing process, printable size, and life cycle of the constrained surface[153, 154].
The existing separation problem in constrained surface SLA process has blocked the
development of 3D printing on a large scale of building area. To solve this problem, many attempts
have been made[155-157]. The most common way to address this problem is making a non-sticky
and air-permeable coatings on a constrained surface, such as using Teflon films or
polydimethylsiloxane (PDMS)[158]. However, the force required to separate the cured layer from
the PDMS-coated constrained surface is still considerably large. EnvisionTEC Inc. developed
another approach based on a peeling mechanism, aiming to reduce the separation force as the force
required for peeling is much less than pulling. To facilitate the peeling process, a tilting motion
system was adopted by lifting or lowering one side of the platform slowly while pivoting it to the
22

other side. However, this mechanism performs poorly for a part with a large cross-section area as
the designed tilting angle is coupled with the maximum building area the process can deal with.
Furthermore, the additional tilting motion increases the fabrication time resulting in a
reduction of productivity. Another mechanism was developed based on the peeling approach using
a two-channel system[69], as shown in Figure 16. This method is aimed at changing the pull-up
force into shear force. During the horizontal translation, the built part gets separated from the resin
vat, facilitating the convenient vertical motion of the platform by a pulling-up action. However,
the disadvantage of this two-channel approach is that the area of the vat should always be designed
at least double the size of the maximum building area resulting in increased construction
complexity as the building area increases.

Figure 16: A two-channel system with PDMS[69].
A novel method based on the continuous liquid interface production process was
proposed[159], as shown in Figure 17. In this method, a highly air permeable coating was used to
increase the oxygen concentration below the constrained surface. Despite its advancement, it is
still a significant challenge when it comes to a part with a large contact surface area because the
separation force will be extremely large even though the separation force is inversely proportional
to the thickness of the dead zone.
23

Figure 17: Schematic of the CLIP printer where the part (gyroid) is produced continuously by simultaneously
elevating the build support plate while changing the 2D cross-sectional UV images from the imaging unit. The
oxygen-permeable window creates a dead zone (persistent liquid interface) between the elevating part and the
window[159].
Although various methods have been proposed based on the surface treatment or the
intelligent design of motions for the building vat, the barrier has never been completely solved.
Chapter 3: Research Questions and Hypotheses
3.1 Statement of Problems
In this dissertation, we try to develop hybrid 3D printing processes to address the current
limitations aforementioned in the layer-based SLA process. They are multi-material 3D printing
solutions for the SLA process, support structures for the overhanging features in the SLA process,
and significant separation force while breaking the adhesive bonding in the constrained-surface-
based SLA process.
As for the multi-material 3D printing solutions for the SLA process, some previous works
have been reported by switching the building part between different material reservoirs. Due to the
material contamination issue, multiple additional sequential procedures have to be involved during
the printing process to address this problem, which includes rinsing the uncured material attached
on the surface of the cured part and drying the surface afterward which makes the whole building
24

process quite tedious. Meanwhile, a significant amount of material will also be wasted during the
cleaning process in each layer.
On the other hand, the support structures in the SLA process not only mainly increase the
printing time and material waste but also need substantial post-processing time. Moreover,
removing the support structure can introduce defects on the surface. In some individual cases, it is
even impossible to reach and remove the built support structures when the support structures are
almost in the sealed empty chamber. There are two possible directions to address the problem
according to the previous literature review result: 1) selection of the suitable sacrificial material
for the support structures. 2) innovative multi-material 3D printing process to incorporate the
sacrificial material during the fabrication process.
Furthermore, as the building area increases, the separation force will significantly increase
as well in a constrained-surface-based SLA process. It is challenging to separate the newly cured
layer from the constrained surface. This difficulty results from an adhesive bonding that is
developed between the newly cured layer and the constrained surface. Thus, a separation force is
needed to separate the built part from the constrained surface. Such a separation force will
significantly affect the printing speed, reliability of the printing process, printable size, and life
cycle of the constrained surface.
3.2 Hypotheses
The primary research goal of this research is to address current limitations for the SLA
process, and it will mainly focus on the following three aspects: 1) how can we elegantly realize
multi-material 3D printing without having any material contamination issue? 2) how can we easily
remove the support structures without leaving unexpected marks on the surface of the original
25

parts with minimal post-processing effort required? 3) how can we reduce the separation force in
a constrained-surface-based SLA process for the large-scale 3D printing with the balance of the
printing speed, reliability of printing process, printable size, and life cycle of the constrained
surface?
To answer the above-proposed questions, we have the following hypotheses:

To answer this question, the following hypotheses are investigated:

To answer this question, the following hypotheses are investigated:
Q1: how can we elegantly realize multi-material 3D printing without having any material
contamination issue?
Hypothesis 1.1: An innovative material deposition method can be used to selectively deposit
various materials into desired places since the material switching process in the traditional
multi-material SLA process is the main problem to introduce the material contamination issue.
Hypothesis 1.2: The proposed material deposition method must be a very low-cost method
compared to the existing expensive MJP process.
Hypothesis 1.3: If there exists a low-cost material deposition method, it can be easily
integrated with the current SLA process.
Q2: how can we easily remove the support structures without leaving unexpected marks on the
surface of the original parts with minimal post-processing effort required?
26

To answer this question, the following hypotheses are investigated:

Hypothesis 2.1: A special low-cost sacrificial material can be used as the support material for
the SLA process.
Hypothesis 2.2: The proposed support material is phase changeable and can be incorporated
into a multi-material SLA printing process.
Hypothesis 2.3: The removal of the support material after the building process is an automated
process with less or no human interference.

Q3: how can we reduce the separation force in a constrained-surface-based SLA process for
the large-scale 3D printing with the balance of the printing speed, reliability of printing process,
printable size and life cycle of the constrained surface?
Hypothesis 3.1: Less force but multiple attempts may lead to a more efficient result.
Hypothesis 3.2: The reduction of the separation force for a constrained-surface-based SLA
process can be achieved by introducing a small force to induce the initial crack in the interface
between the newly cured layer and the constrained surface.
Hypothesis 3.3: The initial crack in the interface between the newly cured layer and the
constrained surface can be quickly propagated to the entire interface.
27

Chapter 4: Results and Evaluations
This chapter presents the current status of the research work, including some preliminary
results and evaluations that have partially validated our proposed hypotheses. Section 4.1 will
discuss the proposed hybrid DIW and SLA process to address the current limitation of the material
contamination issue for the current multi-material SLA 3D printing, including the integration
process development, investigation on the material diffusion effect, and some results fabricated by
this hybrid process. Section 4.2 will discuss the proposed innovative approach adopting the
water/ice as the sacrificial material to address the current limitation of the support structures for
the SLA process, including the process development, the challenging behind this approach, and
some results fabricated by this approach. Section 4.3 will discuss how the proposed vibration-
assisted method can address the limitation of the separation force in the constrained surface SLA
process, including the process development, analysis of millions of data got from the embedded
force sensors to measure the separation force and the side-by-side comparison of the performance
between the vibration-assisted method and the traditional SLA process.
4.1 Hybrid Direct Ink Writing (DIW) and Stereolithography (SLA) for multi-
material 3D Printing
In this research, we report the design and development of a hybrid 3D printing system that
integrates traditional direct ink writing with laser-based stereolithography printing processes for
multi-material 3D printing of heterogeneous materials without having material cross-
contamination issues. The laser spot tracks the syringe nozzle tip to cure the newly dispensed
material immediately. Using the proposed hybrid 3D printing method, we can utilize advantages
from both 3D processes to fabricate complex structures with various functionalities.
28

4.1.1 Design of the hybrid multi-material printing process
Multiple syringe dispensers with different materials loaded have been mounted together on
the linear stages, which can move in the X-Y plane. A low-cost pneumatically controlled
dispensing system has been developed to achieve fast response and accurate material deposition
without any lag or over-dose problem.
4.1.1.1 Design of the process
All dispensers are pneumatically driven by pre-configured air pressure, and each of them
can be selectively activated by programmed valves. Different dispensers may be required to
configure different nozzle diameters and air pressures due to different viscosity of loaded materials.
The nozzle tip is immersed in a material reservoir that contains the photocurable resin to dispense
the target materials inside the base material, as shown in Figure 18.

Figure 18: The overview of the designed experimental setup, including the multi-material DIW printhead, which
can move in the X-Y plane, the laser scanning based SLA module, which can move in Z direction along with the
tank above and the building platform which can move in X-Y-Z direction.
A UV laser scanning system is mounted right below the material reservoir as shown in the
Figure 18. The laser beam guided by the galvano mirror can pass through the transparent glass
and scan the image pattern on the bottom surface of the material reservoir. The moving dispense
29

nozzle tip immerses into the base material. It begins to dispense the primary material while the
laser is tracing the nozzle tip toolpath to cure or fix the newly dispensed material immediately, as
shown in Figure 19a-c.

Figure 19: Schematics of the hybrid DIW and SLA multi-material 3D printing system. a) First step. A dispense
syringe is moving horizontally towards the left to dispense the loaded material A onto the bottom surface of a tank,
which contains photocurable liquid base material. The syringe nozzle tip is immersed into the base material and
forms a layer-thickness gap to the bottom surface of the tank. The scanning laser spot is tracing the dispensed
material to solidify it while the syringe is dispensing the newly coming out material ahead. b) Second step. Switch
to another dispense syringe to dispense material B. c) A region of interest (ROI) in b to illustrate the curing while
dispensing process in detail. d) Third step. Switch to the building platform to transfer the newly built layer from
the bottom surface of the tank to the previous fully cured layers with a second light exposure.
Hence, after dispensing the current layer of materials, the dispensed material will no longer
float in the base material but stick to the bottom film of the material reservoir. Then a building
platform will be switched to the place above the current dispensed layer and lower down to the
position where a layer-thickness gap is formed between the previously built part surface and the
bottom surface of the material reservoir. After one second exposure of the current newly built layer
by the scanning laser spot, the current layer will be transferred to the previous built surface and
30

stack together, as shown in Figure 20. After repeating this building process layer by layer, the
final multi-material part will be finished. Unlike the traditional DIW process in which the
dispensed material will not be cured until the deposition of the whole layer of material is finished,
the proposed method is a curing-while-dispensing fabrication process that is capable of preventing
material collapsing and cross-contamination issue, thus building multi-material part with high
resolution. Most of all, our proposed hybrid 3D printing process could be easily extended to a large
number of photocurable and non-photocurable materials necessary for engineering highly complex
functional composites, which can benefit robotics, tissue engineering, and wearable devices in the
future.

Figure 20: Third step. Switch to the building platform to transfer the newly built layer from the bottom surface of
the tank to the previous fully cured layers with a second light exposure. An ROI in d to illustrate how the current
layer attaches to the last fully cured layers.
4.1.1.2 Design of the low-cost pneumatic dispensing system
A homemade pneumatic dispensing system was developed for our proposed hybrid 3D
printing process, which is cost-effective. Multiple components were involved in the designed
system, as shown in Figure 21. The air compressor with a container (Pancake Air Compressor,
BOSTITCH, MD, USA) is used to generate the compressed air, which is up to 150 psi. An air
31

pressure controller (YDL-983A, Taishi, China) is used to enable/disable the global air pressure
input to the system. Multiple air pressure regulators are used to individually adjust the air pressure
for different dispensers as different materials loaded in the dispensers may have different viscosity
that requires different air pressure to drive them. The electrical air valves (USS2-00081, U.S. Solid,
OH, USA) are used to control the airflow in the tubes.

Figure 21: The schematic design of the pneumatic dispensing system, including connections of the air tubes and
signal wires.
When the air valve is open, the corresponding air pressure in the dispenser will be equal to
the desired value, and the dispenser is supposed to start dispensing the loaded material. When the
air valve is closed, the dispenser is supposed to stop working with no extra material coming out
from the dispenser nozzle tip. However, after the Ax (x=1,2,3...n) valve is closed, the residual air
pressure left in the air tube that is connected to the Ax valve and the dispenser will not be decreased
to zero immediately due to the sealed environment resulting in the delay stopping issue. That is,
when the signal is sent to close the Ax valve, the material in the corresponding dispenser will still
come out from the nozzle tip until a pressure balance state is achieved, as shown in Figure 22. To
address this problem, we incorporate another valve Bx (x=1,2,3...n) in the design, which is
32

connected to the tube, which is in between the Ax valve and the dispenser, as shown in Figure 21.
The Bx valve is used as an air exhausting valve. When a signal is sent to close the Ax valve, a
delayed signal will also be triggered to open the corresponding Bx valve to exhaust the residual
air pressure in the tube, thus preventing the undesired extra material coming out from the dispenser
nozzle tip after a signal is sent to stop dispensing. An air filter is mounted on the other end of the
Bx valve to filter the air that is released to the environment and also to prevent dust and big particles
from the open environment entering into the dispenser to cause any undesired contamination.

Figure 22: The corresponding logic control diagram of the designed low-cost pneumatic dispensing system. Ax
and Bx denote the electrical air valves, x=1,2,3…n. Dx denotes the dispenser, x=1,2,3…n. The value n indicates
the number of dispensers.
The difference between using an additional exhausting air valve in the system and without
using it is also illustrated in the logic control diagram, as shown in Figure 22. It may take a much
longer time to reach a balanced state to stop the extra material coming out from the nozzle tip when
the exhausting air valve is not enabled in the system. A pulse signal is used to drive the exhausting
air valve with a pulse width of about 200 ms depending on the response time of selected electrical
air valves. The designed pneumatic dispensing system has been tested and verified that the material
dispensed can be well controlled without having the delay-stopping issue by applying the
exhausting air valves.
33

4.1.1.3 Motion synchronization
The proposed hybrid 3D printing process must synchronize the movement of the dispensing
system and the laser scanning system since they are initially two independent systems. Spatial
coordinates between the dispensing system and laser scanning system have to be calibrated before
printing, as well as the coordinates calibration between different dispensers to avoid misalignment
of the image pattern dispensed and solidified. A simple image recognition method has been
adopted to adjust the transformation matrix applied to the generated G-Code to align the
coordinates. Besides, centralized control commands from control software have been distributed
to the dispensing system and laser scanning system to ensure motion synchronization.

Figure 23: Motion synchronization. a) Multiple syringe dispensers are mounted together on the slider of the linear
stage; b) Calibration of the relative distance between mounted syringe nozzle tips in the X-Y plane.
To realize the motion synchronization between two independent systems, it is quite
important to calibrate the systems before starting a print. Due to the mechanical error during parts
assembly, it is difficult to achieve an accurate distance between syringe dispensers as they are
expected in the design stage, as shown in Figure 23. A simple image recognition method has been
introduced to help with calibration. Each dispenser nozzle tip will draw a square on a piece of
paper by running a pre-generated G-Code. We can get an image by taking a photo from a top view
of these drawn rectangles. A pre-programmed python script will be running to load the image and
detect all the squares on the paper and calculate the corresponding central distance between them
34

along the X and Y direction. A quarter coin is used as a measurement reference to map the pixel
distance to the real size. It is obvious that only linear translation needs to be calibrated for the
syringe nozzles since they are all mounted on the same slider on a linear stage. The calibrated
offset distances between syringe nozzles will then be input to the slicing software to slice the multi-
material part. More importantly, the laser scanning system must be calibrated to align the toolpath
of the dispenser nozzle tip. A similar method must be implemented for the calibration. A pre-
generated G-Code will be running to guide the laser spot to draw a square on a piece of paper. The
same G-Code will also be decoded by the motion controller to draw a square by the most left
dispenser nozzle. Besides the calibration of the linear translation, scale and rotation also have to
be calibrated since they are in two different coordinate systems. After we obtain all the information
about the transformation matrix, the calibration work is done. It will be used to process the input
G-Code and generate the corresponding updated ADC value to control the Galvano mirrors. The
workflow diagram of the printing process is shown in Figure 24.

Figure 24: The workflow diagram of the printing process, from the digital CAD model to motion control.
4.1.2 Multi-material diffusion and multi-exposure mechanism
A salient feature of our proposed hybrid 3D printing is curing while dispensing, in which the
newly dispensed material from the nozzle tip will soon be solidified by a scanning laser spot. The
35

nozzle tip of the dispenser has to be immersed in a liquid base material during the printing process.
The liquid material will be self-healed immediately after it is sheared by the dispensing nozzle
leaving no voids in the base material. There are several benefits we can have when the main
material is dispensed inside a liquid base material rather than dispensed in the air: (1) The liquid
base material can isolate the dispensed material from being exposed to the oxygen in the air to
ensure all the dispensed material can be cured. If the photopolymer material is dispensed in the air,
the oxygen will consume the free radical ion in the photopolymer near the region where the
material is contacted with the air to form a thin shell of the dead zone in which the chain
polymerization will be terminated even when it is exposed to the UV light[159]. Therefore, only a
portion of dispensed material will remain in a liquid state after exposure, and it will be even worse
when only a thin layer of material is dispensed on the substrate surface. (2) The base material can
work as a sacrificial material when the original desired part requires a significant amount of
supporting structures to preserve some delicate features. Thanks to the multi-material printing
capability, it is free to use a chemical soluble photopolymer as the base material which can be
chemically dissolved after printing without causing any damage to the original part. (3) Instead of
assigning the supporting structure to the base material, the base material can also be used to print
the outer shell of a part whose shape can be highly defined by the scanning laser-like the traditional
SLA process can achieve. Other functional materials can be dispensed in the inner region of the
part, which contributes to the functionalities of the design. (4) A low viscous material dispensed
on the surface of the substrate that is exposed in the air will collapse instantly due to the surface
tension, which will significantly affect the printing quality of the part. The boundary surface
property will be modified when the base material is filled on the surface of the substrate. The liquid
36

base material can prevent the dispensed material from collapsing, thus improving the building
quality of the final part.
4.1.2.1 Characterization of multi-material diffusion
The undesired diffusion effect will occur due to the fluidity of the material dispensed in the
base material reservoir, causing unavoidable defects for the designed geometry in each layer or
material cross-contamination issue resulting from the displacement of the dispensed material. If
the diffusion effect occurs, the width of the dispensing trace will expand shortly towards the
interface between the dispensed material and the base material along the horizontal direction, as
shown in Figure 25a. Different liquid materials may have different interactions due to the material
difference in terms of concentration, density, and viscosity, resulting in various diffusion rates.
Characterizing the material diffusion effect is vital for succeeding in the hybrid 3D printing process.
The dispensed material has to be cured or fixed by the scanning UV laser in the base material
reservoir as soon as possible to prevent it from diffusing towards the boundary of the dispensed
trace. However, there is a chance that the dispensing nozzle tip will be clogged if the scanning
laser spot is getting too close to it. Therefore, it is vital to characterize the diffusion effect for
determining a safe distance between the scanning laser spot and the dispensing nozzle tip. In our
designed experiments, the dispenser will dispense a straight line of the loaded material on the
substrate surface with a length of 50 mm using a constant moving speed, 10 mm/s. Meanwhile,
the scanning laser spot will follow the dispensing path with various scanning speeds.
37

Figure 25: Study the diffusion effect of the dispensed material trace in the liquid base material. a) The width of the
trace V.S. different delay time to cure and fix the material trace under different printing settings, dispensing in
resin / in air and applying 1 psi / 3 psi air pressure to extrude out the material from the syringe. b) A region of
interest in a, cure and fix the material trace within a short delay time. c) The corresponding microscopic views
from samples in a, group ④ and ⑥ are top views of the dispensed traces under different printing settings, group
⑤ is the cross-sectional views of the dispensed traces under different printing settings, all scale bars 0.1 mm. d)
The corresponding microscopic views from samples in b, group ①, and ③ are top views of the dispensed traces
under different printing settings, and group ② is the cross-sectional views of the dispensed traces under different
printing settings, all scale bars 0.1 mm.
The purpose of conducting this experiment is to study the relationship between the
diffusion effect and different delay time to cure and fix the material trace under different printing
settings in terms of air pressure applied to the dispenser and dispensing in the air/resin as shown
in Figure 25a,b. We assume that the dispensed material trace inside the liquid base material
exhibits 1D Fickian diffusion[160]:
𝐶 ( 𝑥 , 𝑡 ) ⁡ =
𝑁 /𝐴 √ 4 𝜋𝐷 𝑡 𝑒 𝑥𝑝 (
− ( 𝑥 − 𝑥 0
)
2
4 𝐷𝑡
) (4)
38

where C is the dispensed material concentration, N is the number of material molecules, A is the
cross-sectional area through which it diffuses, 𝑥 0
is the position of the line source along the x-axis,
perpendicular to the line length, D is the diffusion constant, and t is the time. The peak width,
measured as the temporal peak variance, is first converted to a spatial peak variance ( 𝜎 2
) and then
using the Einstein–Smoluchowski equation:
𝜎 2
= ⁡2 𝐷𝑡 (5)
The above equation can be rewritten as a standard Gaussian function:
𝐶 ( 𝑥 , 𝑡 ) ⁡ =
𝑁 /𝐴 √ 2 𝜋 𝜎 2
𝑒 𝑥𝑝 (
− ( 𝑥 − 𝑥 0
)
2
2 𝜎 2
) (6)
We measured the width of the dispensed material trace to indicate the degree of diffusion.
From the plotted data, we can learn: (1) After the material is dispensed in the liquid resin, the
diffusion will occur as time elapses, as shown in Figure 25a. The dash lines are the fitted trend of
the corresponding measured sparse dots. Top views and cross-sectional views of corresponding
dispensed traces at different moments were captured, as shown in Figure 25c,d. (2) When the air
pressure increases, the width of the dispensed material trace will increase correspondingly either
in the air or in the resin, which has a similar trend in a traditional DIW process, as shown in Figure
25a,b. (3) When the material is dispensed in the air, we can learn that the material trace collapses
as soon as the dispensed material exits from the dispenser nozzle tip, and the width of the dispensed
trace can be maintained quite stable in the air as shown in Figure 25. (4) In the early stage,
comparing the width of the dispensed material in the air and the liquid resin when the same air
pressure is applied to the dispenser, we noticed that the width of dispensed trace in the air is about
three times wider than the width of dispensed trace in the liquid resin as shown in Figure 25b,
39

regardless of the amplitude of the air pressure being applied. It shows the evidence that the
surrounding liquid base material in the material reservoir can prevent the dispensed material trace
from collapsing, thus improving the dispensing resolution compared to the traditional DIW process,
as shown in Figure 25d.
Our proposed method has the advantage of dealing with material with low viscosity
compared to the conventional DIW fabrication process in which the material is dispensed on a
substrate that will be exposed in the air during the dispensing process. The liquid-state material
will immediately collapse and spread on the substrate surface as soon as it exits from the dispenser.
It is difficult to maintain the desired thickness of the dispensed trace, which is critical in a layer-
based fabrication process. From our experimental tests, the thickness of the dispensed material
trace is about 20 μm after it exits from the dispenser tip and settles down on the substrate surface
when it is dispensed in the air. From the top views of the captured images shown in Figure 25d,
the width of the dispensed material trace remains quite steady overtime when it is dispensed in the
air, as shown in Figure 25a,b. However, we noticed it failed for a 500-mW scanning laser to
solidify this thin liquid photopolymer trace due to the oxygen inhibition effect[159].

The oxygen
in the air can penetrate from the interface between the air and the outer surface into the liquid
photopolymer to consume the free radicals contained in the liquid photopolymer whose role is to
trigger the chain polymerization that will convert the photopolymer from a liquid state to a solid-
state. On the other hand, our proposed hybrid 3D printing process immerses the dispenser nozzle
tip into a liquid base matrix material to dispense the target material which naturally prevents the
oxygen in air penetrating the target dispensed material to ensure the scanning laser can solidify the
small amount of dispensed material trace. Figure 26 shows the additional experimental results
collected in order to study the diffusion effect. From the comparison results, we can learn that our
40

proposed curing while dispensing mechanism has a significant advantage in maintaining the shape
of dispensed material, which is the fundamental criterion to fabricate more complex 3D objects
further.

Figure 26: Additional experimental results collected to study the diffusion effect. Four straight lines, A, B, C, D,
were printed using different printing parameters for comparison. A: the material was dispensed in another resin,
and the laser was used to cure it with a long delay time applied; B: the material was dispensed in the air, and the
laser was used to cure it with a long delay time applied; C: the material was dispensed in another resin and the
laser was used to cure it with a short delay time applied; D: the material was dispensed in the air, and the laser
was used to cure it with a short delay time applied. The air pressure applied to the dispensers is 3 psi in the
experiments.
4.1.2.2 The multi-exposure mechanism to address the bonding issue between layers
A key challenge of the designed hybrid 3D printing process is the bonding issue between
layers since the curing process of each layer is quite different from the DIW or SLA process. In
our proposed curing-while-dispensing process, only after the dispensed material in the current
layer is fixed and stuck on the film surface in the bottom of the reservoir, the current layer can then
be transferred to the previously cured layers successfully. Otherwise, in addition to the diffusion
effect, the current layer of the dispensed image pattern will be ruined by the disturbance when the
printing platform moves down to get close to the current layer. Therefore, to ensure the success of
41

the designed printing process, two critical requirements must be fulfilled during the fabrication of
each layer: (1) The dispensed material from the nozzle tip has to be fixed without further movement
even when disturbance happens in the liquid reservoir. (2) The current layer has to be able to attach
to the previously cured layers.
Here, we propose a multi-exposure method to address this critical problem. Two different
circumstances should be considered individually. When the material candidates all belong to the
photocurable category for the proposed multi-material 3D printing, a double-exposure mechanism
is applied to successfully transfer the current dispensed layer from the bottom film to the
previously cured layers as shown in Figure 27a,c. Firstly, low power is input to the laser scanning
system, and the dispensed material on the film surface in the base material reservoir is cured into
a gel-like state during the curing-while-dispensing process, as shown in Figure 27a. Secondly,
after the dispensed material is converted to a gel-like state, the printing platform moves down to
the position where the current layer is. A second exposure with a relatively high energy input is
applied to ensure the current gel-like layer will be fully cured and attached to the previously cured
layers, as shown in Figure 27b. The chemical bonding force between the current layer and
previous layers is larger than the mechanical bonding between the current layer and the surface of
the bottom film. Thus, the printing platform is able to move up and separate the current layer from
the bottom film, as shown in Figure 27c.
42

Figure 27: The multi-exposure mechanism to ensure the bonding between layers. a-c) Double exposures are
applied to the photocurable materials for multi-material 3D printing; a) First UV light exposure is applied to cure
the dispensed material trace into a gel-like state; b) Move down the printing platform and apply a second exposure
to fully cure the current layer and further attach to the previous cured layers; c) Move up the printing platform to
detach the part from the bottom film; d-g) Triple exposures are applied to the process when non-photocurable
materials are involved; d) First exposure is applied to cure a thin layer of base material in the bottom of the
reservoir; e) The non-photocurable material is dispensed onto the previous cured thin layer and a second exposure
is applied to generate a gel-like boundary wall around the dispensed trace; f) Move down the printing platform
and apply a third exposure to fully cure the current layer and further attach to the previous cured layers; g) Move
up the printing platform to detach the part from the bottom film.
When some functional but non-photocurable materials are involved in the material library,
such as electrically conductive ink, magnetic material and so on, a triple-exposure mechanism is
applied to successfully transfer the current dispensed layer from the bottom film to the previously
cured layers as shown in Figure 27d-g for 3D printing multi-functional materials. Firstly, a thin
layer of the base material is cured by the scanning laser, as shown in Figure 27d. This thin layer
works as a bed to hold the non-photocurable material. Then the non-photocurable material is
dispensed onto the surface of the previous cured thin layer along with a second laser exposure. The
43

second exposure is necessary to cure the base material abound the non-photocurable material trace
to build a boundary wall in a gel-like state. The boundary wall is used to wrap the non-photocurable
material so that the corresponding dispensed material is constrained within the boundary wall to
prevent it from further collapsing, as shown in Figure 27e. Therefore, when the printing platform
goes down to the position where the current layer is, the geometric shape of the non-photocurable
material can be retained well. A third exposure is applied to fully cure the current layer and attach
it to the previously cured layers, as shown in Figure 27f. Finally, the printing platform goes up to
separate the current layer from the bottom film, as shown in Figure 27g.
4.1.3 Modeling the hybrid 3D printing process
The laser spot irradiance distribution is Gaussian, as shown in Figure 28. The
photopolymer resin obeys the Beer-Lambert law of exponential absorption. When a UV laser spot
is focused at the surface of a polymer solution, the curing depth 𝐶 𝑑 can be calculated by the
“working curve equation” [57]:
𝐶 𝑑 = 𝐷 𝑝 𝑙𝑛 ( 𝐸 𝑚𝑎𝑥
/𝐸 𝑐 ) (7)
Where 𝐶 𝑑 is the depth of resin cured as a result of laser irradiation, 𝐷 𝑝 is the depth of penetration of
laser into a resin until a reduction in irradiance of
1
𝑒 is reached, 𝐸 𝑚𝑎𝑥
is the peak exposure of laser
shining on the resin surface (center of laser spot), 𝐸 𝑐 is the critical exposure at which resin
solidification starts to occur. The curing line width is determined by the following equation:
𝐿 𝑤 = 𝑊 0
√ 2 𝐶 𝑑 /𝐷 𝑝 (8)
44

Where 𝐿 𝑤 ⁡is the curing line width, 𝑊 0
⁡is the radius of the laser beam focused on the resin surface.
The line width is proportional to the beam spot size. If a greater cure depth is desired, the line
width must increase. The maximum laser exposure value is given by the following expression:
𝐸 ( 0 , 0 ) ≡ 𝐸 𝑚𝑎𝑥
=
√
2
𝜋 𝑃 𝐿 𝑊 0
𝑉 𝑠 𝑐𝑎 𝑛
(9)
Where 𝑃 𝐿 ⁡is the output power of the laser, 𝑉 𝑠 𝑐𝑎 𝑛 ⁡is the scanning speed of the laser spot. By
substitution of Equation (9) into Equation (7), it can be shown that
𝐶 𝑑 = 𝐷 𝑝 𝑙𝑛 (
√
2
𝜋 𝑃 𝐿 𝑊 0
𝑉 𝑠 𝑐𝑎 𝑛 𝐸 𝑐 )
(10)
Accordingly, we can get that
𝑉 𝑠𝑐 𝑎 𝑛 =
√
2
𝜋 𝑃 𝐿 𝑊 0
𝐸 𝑐 𝑒 − 𝐶 𝑑 / 𝐷 𝑝
(11)
We assume the cross section of the nozzle is a cylindrical shape. The ejection speed, 𝑉 𝑒 𝑗 𝑒 𝑐𝑡 ,
of the dispensed material at the nozzle tip can be described as
𝑉 𝑒 𝑗 𝑒 𝑐𝑡 =
𝑄 𝜋 (
𝑑 2
)
2

(12)
Where 𝑄 is the volumetric flow rate of the dispensed material from the nozzle tip, 𝑑 is the inner
diameter of the dispensing nozzle. The relationship between volumetric flow rate 𝑄 and pressure
drop across the nozzle channel is[113]:
45

𝑄 = 𝜂 ( 𝑃 𝑠 − 𝑃 0
) (13)
Where 𝑃 𝑠 and 𝑃 0
are the pneumatic pressure applied in the syringe and ambient pressure,
respectively, and 𝜂 is the hydraulic conductance. The ambient pressure 𝑃 0
is the pressure applied
to the nozzle tip from the liquid material in the reservoir, which is related to the depth of the base
material in the reservoir and its corresponding fluid density.
For a Newtonian fluid with viscosity 𝜇 , the hydraulic conductance of the nozzle channel is
given by
𝜂 =
𝜋 𝑟 4
8 𝜇𝑙
(14)
Where 𝑟 and 𝑙 are the radius and length of the cylindrical nozzle channel, respectively.
Let 𝑉 𝑑 𝑖𝑠 𝑝 𝑒 𝑛𝑠 𝑒 𝑟 denotes the constant moving speed of the syringe dispenser. We assume
𝑉 𝑒 𝑗 𝑒 𝑐𝑡 = 𝑉 𝑑 𝑖𝑠 𝑝 𝑒 𝑛𝑠 𝑒 𝑟 for a steady deposition of the material from the nozzle tip, thus maintaining a
good cylindrical shape of the dispensed trace in the base material reservoir. Therefore, we can get
the relationship between the pneumatic pressure applied to the syringe dispenser and the moving
speed of the syringe dispenser by combining Equation (12), Equation (13) and Equation (14).
𝑉 𝑑 𝑖𝑠 𝑝 𝑒 𝑛𝑠 𝑒 𝑟 =
𝑟 2
8 𝜇𝑙
( 𝑃 𝑠 − 𝑃 0
) (15)
Furthermore, it is vital to synchronize laser scanning with the motion of the syringe dispenser for
our proposed hybrid 3D printing process. The following critical condition must be fulfilled.
46

𝑉 𝑑 𝑖𝑠 𝑝 𝑒 𝑛𝑠 𝑒 𝑟 = 𝑉 𝑠 𝑐𝑎 𝑛 (16)
By combining Equation (10), Equation (15), and Equation (16), we can get the relationship
between the input pneumatic pressure and the curing depth, which is the layer thickness during
printing.
𝐶 𝑑 = 𝐷 𝑝 𝑙𝑛 (
√
2
𝜋 8 𝑃 𝐿 𝜇𝑙
𝑊 0
𝐸 𝑐 𝑟 2
( 𝑃 𝑠 − 𝑃 0
)
)
(17)
By substitution of Equation (17) into Equation (8), we can also get the relationship between the
input pneumatic pressure and the curing width.
𝐿 𝑤 = 𝑊 0
√
2 𝑙𝑛 (
√
2
𝜋 8 𝑃 𝐿 𝜇𝑙
𝑊 0
𝐸 𝑐 𝑟 2
( 𝑃 𝑠 − 𝑃 0
)
)
(18)
Without a priori knowledge of the designed printing system and material properties, one cannot
predict the optimal printing parameters since they are coupled together, so they must be tuned
and determined experimentally. The purpose of the theoretical model is, therefore, to provide a
framework for understanding the relationships between different parameters.
47

Figure 28: Modeling the hybrid 3D printing process. a) An inset of Gaussian function for the energy distribution
of the scanning laser spot; b) Illustration of the light-curing model for the photopolymer; c) Region of interest in
(d), indicating a scanning laser spot is curing the liquid material while the syringe nozzle is dispensing the material;
d) A laser spot is following the material trace dispensed by the syringe dispenser.
4.1.4 Results and Discussion
4.1.4.1 As a proof of concept - print a 3D ant inside a crystal box
To demonstrate the capability of the proposed hybrid multi-material 3D printing process
in dealing with multi-color objects without material contamination issue, we printed a 3D ant that
is sealed in a transparent cuboid as a proof of concept which is inspired by the amber in the natural
world.
Two materials with different colors are involved in the design. The stiff red material is
assigned to the ant body, and the stiff transparent material is assigned to the surrounding cuboid.
A 3D digital STL model, as shown in Figure 29a, is sliced into multiple 2D layers, as shown in
Figure 29b with a 100 μm layer thickness. Each layer is then converted to a G-Code toolpath, as
48

shown in Figure 29c. The syringe dispenser and the scanning laser spot share the same toolpath
with a different starting time, as shown in Figure 29d.

Figure 29: 3D printed multi-material crystal ant, amber. a) Input 3D digital CAD model. b) Sliced layers of the
digital model. c) A certain sliced layer picked from the sliced layers in (b). d) Region of interest in (c), the concentric
G-Code syringe toolpath, and the scanning laser path.
10-seconds delay time has been applied to the scanning laser path since we have
characterized the diffusion of the dispensing material trace. That is, the scanning laser spot is
following the dispenser toolpath but always behind it from a certain distance. Both the laser
scanning speed and the syringe dispensing speed is set to be 10 mm/s. The concentric toolpath
pattern has been chosen in this test case for the best contour result. Before the experiment is
implemented, a toolpath simulation is always performed to verify the correctness of the desired
target result, as shown in Figure 30.
Once everything is ready, a real experiment will be conducted to initiate the printing
process. The dispenser needle is inserted into a material reservoir that is full of the transparent base
resin leaving a layer-thickness gap between the dispenser tip and the bottom surface. A 2D sliced
layer will then be drawn on the substrate surface guided by the predefined G-Code toolpath, as
shown in Figure 31.
49

Figure 30: Visualization of the G-Code toolpath.
Meanwhile, a delayed scanning laser spot will follow the dispenser toolpath to cure and fix
the dispensed material trace in the right place to prevent the diffusion effect so that the desired
printing resolution and quality can be ensured.

Figure 31: Run the G-Code toolpath to dispense the red resin in the transparent resin. The laser spot follows the
same toolpath to cure the red resin trace dispensed.
As it is pre-calibrated, ~100 mW of the input power will be provided to the focused
scanning laser spot to ensure it will only cure and fix the right amount of the dispensed material
rather than over cure the base matrix material in the material reservoir. The input laser power is
dynamically adjusted by the Pulse Width Modulation (PWM) mechanism of a microcontroller.
The curing width and depth are highly related to the input power of the scanning laser spot and the
scanning speed. The ideal curing calibration is that the curing width and depth match with the
50

width and thickness of the dispensed cylindrical material trace on the substrate in the material
reservoir. After this first exposure, the current 2D layer on the substrate will be in a gel-like state
rather than a fully cured layer, but it is strong enough to stick on the substrate surface without
further diffusion. That is, the crosslinked polymer network is not fully established but waiting for
another exposure to be fully polymerized[161].
After the material is finished depositing in the desired place and cured by the scanning
laser, the whole current layer will then be transferred to the building platform or the previously
solidified layers with a second exposure as shown in Figure 32. ~400 mW of input power will be
provided to the focus scanning laser spot to ensure the current layer is fully solidified and attached
to the previous layers.

Figure 32: Transfer the newly semi-cured layer on the bottom surface of the tank to the previous fully cured layers
by scanning the surrounding transparent base material in a square shape for this case.
It is noticed that there is a little amount of base material left between the current layer and
the previously solidified layers when the building platform lowers down to contact the top surface
of the current layer due to the imperfect flatness on the top surface of the current layer as shown
in Figure 33b.
51

Figure 33: The imperfect flatness on the top surface of the current building layer. a) The material dispenser is
dispensing the red material into the liquid base material reservoir following by a scanning laser; b) SEM image of
the top surface of the current building layer with interlaced highland and valley.
This imperfection is caused by the cylindrical material exited from the dispenser nozzle tip
and the non-constrained freeform photopolymer growth. Because of the curvature of the
cylindrical material overlay on the substrate, it is difficult to form a flat plane on the top surface
when multiple cylindrical lines overlie one by one.

Figure 34: An illustration of the sandwiched liquid base material serving as a “glue” to tightly bond the current
layer with the previous layers when a second exposure is applied. a) A second laser exposure is applied to transfer
the current layer to the previous cured layers. b) A schematic cross-sectional view of the sandwiched “glue” layer.
Also, although the same type of photopolymer will have a constant curing depth when the
same amount of energy input is applied, curing depth variation still exists if there is no constrained
surface used. The sandwiched liquid base material also serves as a “glue” to tightly bond the
current layer with the previous layers when a second exposure is applied, as shown in Figure 34.
52

Due to the delicate features of our test case, we choose to cure a square shape of the base
material every layer that surrounds the target 3D ant to protect the delicate structures and support
overhanging structures in the original model, as shown in Figure 35. Figure 35b shows the
corresponding SEM image of the bottom surface, which contacts the substrate film in the material
reservoir, as shown in Figure 48.

Figure 35: a) Cured layers stick on the building platform. b) The SEM image of the bottom surface of the cured
layers, scale bar 0.1 mm.
Due to the much faster speed of a scanning laser spot than the motion of a dispensing nozzle,
compared to the traditional multi-material 3D printing process, which only involves multiple
dispensers, our proposed hybrid multi-material 3D printing process has its advantage for
fabrication speed. The whole fabrication process repeats layer by layer by alternatively switching
between dispensing the target material while curing the dispensed material and applying a second
exposure to transfer the current layer to the previously built layers until all the layers are fabricated
as shown in Figure 36a. As we can see from the close views of the fabricated result, some small
features are well kept which shows evidence that our proposed method can achieve multi-material
printing capability while keeping accurate printing quality as shown in Figure 36b,d.
53

Figure 36: Printed multi-material crystal ant. a) Top view of the built part, a red ant with surrounding cured
transparent resin, scale bar 5 mm. b,d) Microscopic views of the corresponding detailed features in a, scale bars
0.5 mm. c) Back view of the built part. e) Side view of the built part.
A cross-sectional view of the fabricated ant is shown in Figure 37.

Figure 37: The cross-sectional view of the built part, scale bar 0.1 mm.
Dimensional accuracy is another critical feature to consider. Due to geometry restrictions,
only 2D dimensional accuracy could be tested. Based on the image of the printed ant, which was
taken from the top view, the STL file of the original ant is zoomed into the same size and oriented
the same way in SolidWorks. Under the same resolution, a screenshot of the original ant STL file
is taken, as shown in Figure 38b.
54

Figure 38: a) Printed ant picture from top view for distance measurements. b) lines were drawn on the screenshot
of the original ant STL file for distance measurements. c) The length obtained from the measurements.
Using ImageJ, horizontal lines were drawn on both pictures at the same location to take
measurements between features. In total, forty lines were drawn, and the results were saved in
excel files. Figure 38a,b explains the way lines were drawn on the pictures. Differences between
two sets of measurements are calculated as how much the printed ant varies from the original ant.
The percentage of variation was calculated and graphed.
4.1.4.2 3D printing of heterogeneous materials
Heterogeneous materials have been widely adopted in soft robotics, wearable devices and
sensors. 3D printing of heterogeneous materials with various stiffness and functionalities has
attracted more and more attention. Even though the conventional SLA printing process can print
parts either with soft materials or stiff materials, it is still challenging for an SLA fabrication
process to print parts with various stiffness due to the nature of the printing process. By
incorporating the DIW process with the SLA process, we can selectively insert the stiff material
into the soft material, thus achieving 3D printing of heterogeneous materials. Specifically, our
55

proposed method fully utilizes the capability of the DIW process to deposit the stiff material on
demand to the desired places in the soft material based on a predefined filling pattern. The SLA
process is used to define the outer geometric shape of the built part with high accuracy. Multiple
dispensers containing materials with different properties can be involved together in the fabrication
process to enhance the functionalities of the designed part. In our experiments, several tensile bars
with a different percent of stiff material fillers are intended for testing the difference in terms of
mechanical stiffness of the designed material. Interwoven patterns with varying rates of filling,
ranging from 0% ~ 100%, are chosen in the design, as shown in Figure 39a. The stiff interwoven
structures embedded in the soft material are expected to strengthen the soft material after the part
is solidified. From the corresponding tensile test result for different combinations of soft elastic
material and stiff material, we can see that the stiffness/softness of the designed material can be
well-tuned by applying different percent of stiff material, as shown in Figure 39b. It is noticed
that the relative strain of the designed tensile bars reduces approximately by half when the
volumetric percentage of the stiff material increases by 75%, showing a nonlinear relationship.
Also, since different materials may have various material properties, calibration has to be
performed if new materials are added to the material library before programming the
heterogeneous distribution of the desired materials.
56

Figure 39: 3D printed multi-material tensile bars. a) Designed and printed tensile bars of various stiffness using
interwoven structures of stiff material. Red color indicates the stiff material applied, and the grey color indicates
the soft/elastic material applied. 0%, 25%, 50%, 75%, and 100% percent of the stiff material are applied
correspondingly in the designed tensile bars. b) Corresponding tensile test results for different combinations of
elastic material and stiff material.
To demonstrate the capability of our proposed hybrid 3D printing process in fabrication of
multi-material parts with heterogeneous materials, a 3D wheel was printed with different stiffness
of materials assigned, as shown in Figure 40e. Specifically, the red material shown in the figure
indicates the stiff material which is assigned to the rim and deposited by the G-Code toolpath
guided moving dispenser, as shown in Figure 40e. The semi-transparent material indicates the soft
material which is assigned to the tire. The soft material is filled in the material reservoir, and the
scanning laser solidifies the tire structure at the second exposure stage, which is also guided by the
G-Code toolpath. The software generates the multi-material G-Code toolpath in each sliced layer.
Different materials or components in the assembly part can be assigned with varying tool paths in
terms of their filling pattern and filling rate, as shown in Figure 40a-d. A compression test was
performed on the printed wheel, as shown in Figure 40f-h. It is clearly demonstrated that the
deformation of the tire made of soft material is significant when the compression is applied while
the stiff rim structure still maintains its shape with little deformation observed.
57

Figure 40: 3D printed multi-material wheel with different stiffness. a) Input 3D digital CAD model. b) Sliced layers
of the digital input model in (a). c) Multi-material G-Code toolpath. d) Region of interest in c, the rim is applied
with the stiff material, and the tire is applied with the soft material. e) Printed multi-material wheel in a single part.
f-h) The wheel under compression.
Besides, the proposed hybrid 3D printing process has successfully demonstrated the potential
of fabricating wearable sensors by involving electrically conductive ink, as shown in Figure 41.

Figure 41: Schematic design of the wearable sensors on the human hand.
Multiple resistance-based strain sensors can be mounted on each joint of the finger to measure
the bending angle of joints, and a capacitive sensor can be mounted on the back of the hand to
interact with users as responsive buttons.
58

Figure 42: a) The 3D CAD design of the strain sensor. b) The schematic diagram of the circuits to measure the
capacitance change of the designed capacitive buttons. c) Wires are inserted into the printed part to connect with
the conductive ink.
Figure 42a shows the 3D CAD model of the designed strain sensor based on the resistance
change. The total thickness of the design is 0.9 mm. The red material indicates the stiff material
used in this case, and the thickness of the stiff material is 0.4 mm. The purpose of applying the
stiff material in the designed strain sensor is to protect the contact point of the conductive ink when
a wire is inserted to connect with the conductive ink so that it will not damage the contact point
during the bending process as shown in Figure S6d. Another function of using stiff material is to
work as an anchor for the transparent soft material in the middle during the bending process to
ensure that the bending will only happen in the middle portion of the printed part, especially when
the strain sensor is mounted on the joint of the finger as shown in Figure 41. The thickness of the
conductive ink dispensed is 0.1 mm for the design, which is sandwiched in the designed part.
Figure 42b shows the schematic design of the testing circuits to measure the resistance change of
the designed strain sensor when there is a bending angle. Arduino is used to process the analog
59

raw data to the digital data. Wires are inserted to the printed part to connect with the conductive
ink, as shown in Figure 42c.
Figure 43a shows the 3D CAD model of the designed capacitive sensor. The thickness of the
designed conductive ink is 0.1 mm, which is also sandwiched in the middle of the part. The red
lines distributed in the center of the part are intentionally for adjusting the anisotropic stiffness of
the part. The four red dots scattered on the edges are purposely designed for fixing the inserted
wires. Figure 43b shows the schematic design of the testing circuits to measure the capacitive
change when a finger touches the individual button[162]. All the pins are connected to an Arduino
for data collecting and processing.

Figure 43: a) The 3D CAD design of the capacitive sensor. b) The schematic diagram of the circuits to measure
the capacitance change of the designed capacitive buttons.
In the experiments, a conical nozzle tip with end inner diameter 200 μm is used to dispense
the conductive ink, as shown in Figure 44.
60

Figure 44: The corresponding conical nozzle tip used to dispense the conductive ink.
Figure 45 shows a printed result of the designed strain sensor with its corresponding plot of
resistance change from an initial relaxing state to a bending state.

Figure 45: A plot of the resistance change for a strain sensor fabricated by the process.
Figure 46 shows a printed result of the designed capacitive sensor with its corresponding plot
of digital capacitance change from a normal state to a pressed state. There are four individual
buttons designed in the designed capacitive sensor.
61

Figure 46: A plot of digital capacitance changes for a capacitive sensor fabricated by the process.
4.1.5 Experimental settings
4.1.5.1 Materials
The UV curable base material, Anycubic clear, used in the diffusion test and building the
transparent cuboid of 3D ant test case was purchased from Anycubic (Shenzhen, China). The
apparent viscosity of Anycubic clear is 260 cps at room temperature. The UV curable red material
loaded in the dispenser, CPS PR57-M, used in all the experiments was purchased from Colorado
Photopolymer Solutions (CPS, CO, USA). The apparent viscosity of CPS PR57-M is 480 cps at
room temperature. The soft elastic material, Formlabs elastic, used in building the tensile bars, and
the tire was purchased from Formlabs (MA, USA). The apparent viscosity of Formlabs elastic
material is 3300 cps at room temperature. The electrically conductive ink, carbon conductive
grease, used in building the wearable sensors were purchased from MG Chemicals (B.C., Canada).
All the materials were used as they were received.
62

4.1.5.2 Material Diffusion Characterization
The top view and cross-sectional view of the dispensed traces were observed under the
microscope. The optical images were obtained on a Micro Vu Sol 161 microscope. After the
material was dispensed and cured in the air for comparison, we buried the cured material trace
with Anycubic clear resin and manually cured it with a UV flash in order to successfully transfer
the printed results from the substrate surface to the microscope for observation. After capturing
the microscopic images of different views of the test cases, we used the ImageJ line tool to measure
the width of different segments during diffusion stages, as shown in Figure 47.

Figure 47: Using the ImageJ line tool to measure the length of legend. a) Using a line measurement tool to measure
multiple material widths from the microscopic images for material printed in resin and cured with a large delay.
The white line indicates the measured length, the black line is the current measuring line; b) Using line
measurement tool to measure multiple material widths from the microscopic images for material printed in resin
and cured with a small delay; c) Line measurement tool is used to measure the vertical height of a cross-section in
the air; d) Line measurement tool is used to measure the vertical height of a cross-section in resin; e) Setting up
the pixel scale before the measurement.
To characterize the diffusion effect of the dispensed traces under different printing settings,
several straight lines were printed on the substrate surface, which was either exposed in the air or
immersed in the liquid resin pool. For each printed line with a total length of 50 mm, we sampled
nine different segments under the microscope with equally spaced distance along the dispensing
direction. The head and tail of the specimen are not taken into consideration, which allows us to
observe the diffusion behaviors as time elapses. Both top-view and the corresponding cross-
63

sectional microscopic images were taken. The digital image measurement software, ImageJ, was
used to collect the data in our experiments. To take measurements, 76 pixels to 0.1 mm scale, as
shown in Figure 47e was initially set up in ImageJ by measuring the legend provided from the
microscopic images with the line tool in the software. The advantage of ImageJ allowed us to take
the length of the drawn line directly and easily store them in the result table, as shown in Figure
47a-d.
All the data for each sampled segment was collected and converted into an excel file for
further analysis. An average was calculated to represent each sampled segment, and standard
deviations were calculated to represent the corresponding width variations of the dispensed traces.
4.1.5.3 Selection of the Substrate Film
The surface property of the selected substrate film is quite critical in our proposed printing
process. Several criteria must be fulfilled when we select the substrate film. Firstly, the substrate
film must have a reasonable light transmission rate, and it must be chemically compatible with the
liquid photopolymers used in our experiments. More importantly, the dispensed material trace
must stay on the substrate surface steadily without bulge due to the surface tension resulting in the
discontinued trace. In our experiments, we used 0.003" thickness TPX / (Polymethylpentene, or
PMP) film from CS Hyde (IL, USA) based on the tests before the experiments, as shown in Figure
48.
64

Figure 48: Selection of the substrate for the proposed hybrid 3D printing process. a) Testing the stability when the
target material is dispensed on the surface of different substrates, all the scale bars are 3 mm; b) The corresponding
data measurement, width ratio (maximal trace width/minimal trace width) versus time; c) SEM images of the
different substrate surfaces, scale bars are 100 μm.
The material, CPS PR57-M, was dispensed on the surface of different substrates. After that,
we took a video and selected multiple screenshots from the recorded video to determine which
substrate is most suitable for the proposed hybrid printing process. We can see from Figure 48a
that the Teflon film will cause a discontinuous trace as soon as the material is dispensed on its
surface. The bulged issue will happen on the PDMS substrate. Figure 48b shows the quantitative
measurement from the corresponding screenshots. The vertical axis indicates the width ratio
(maximal trace width/minimal trace width). We measured the width ten times along the dispensed
trace in each screenshot and calculated the corresponding width ratio by dividing the maximal
trace width measured in the current screenshot by the minimal trace width measured. Both the
TPX film and glass substrate show good stability when the target material is dispensed onto these
surfaces. However, the glass surface has much larger surface energy compared to the TPX film
resulting in a bigger detaching force when the solidified layer is separated from the substrate
65

surface during the printing process. In our experiments, TPX was finally selected. Figure 48c
shows the SEM images of the different substrate surfaces. We can learn that the chosen TPX film
has a textured surface compared to other substrates.
4.1.5.4 Hybrid Multi-material Printing System
A homemade hybrid multi-material 3D printing setup was built to fabricate parts in the
experiments, including a Galvano mirror-based laser scanning system, a 20KPPS laser scanning
kit, which was purchased from AliExpress (AliExpress.com, China) and a low-cost homemade
pneumatic dispensing system. The controller of the laser scanning system, Lasershark, was
purchased from Macpot (Macpot.com, USA). The detailed implementation of the homemade low-
cost pneumatic dispensing system without the delay-stop dispensing issue is illustrated in Figure
S1 (Supporting Information). The material reservoir tensed with the selected TPX film is placed
on a piece of the borosilicate glass plate (CG-1904-37, VWR International, USA). The thickness
of the chosen glass plate is 1.5 mm. The material dispensers are mounted on a linear X-Y stage
(OpenBuilds, NJ, USA), and the laser scanning system is installed on a linear Z stage (Sprintray,
CA, USA). A second linear Z stage (OpenBuilds, NJ, USA) mounted with the building platform
is also mounted on the linear X-Y stage along with the material dispensers. The linear stages are
controlled by the open-sourced hardware RAMPS 1.4 controller from Reprap with open-sourced
firmware Marlin 1.0 loaded. Customized modification has been made to the firmware for the main
controller to decode the generated G-Code and control the movement of the designed system,
which integrates the control of the pneumatic system with the motion movement.
4.1.5.5 Printing Parameters
Dispenser needle (32 gauge Yellow 0.5" length, VWR International, USA) with an inner
diameter 90 µ m was used in the experiments. The pneumatic pressure applied is 3 psi. The constant
66

moving speed of the dispenser in the X-Y directions is set to 10 mm s-1 in all the experiments.
The scanning speed of the laser spot in the X-Y plane is also set to 10 mm s-1 for synchronization
purposes to follow the material dispensing movement. A 405 nm UV laser with a full power 300
mW was used in the experiments. The diameter of the focused laser spot is about 150 µ m. The
dynamic laser power output is realized by the PWM mechanism of the main controller since
different photopolymers require different power inputs to have a certain curing depth. The printing
layer thickness in the experiments is 100 µ m. During the curing-while-dispensing stage, 25%
percent of the full laser power was used to cure the CPS PR57-M dispensed material when the
Anycubic clear resin was used as the base material in the material reservoir. 60% percent of the
full laser power was used to cure the CPS PR57-M dispensed material when the Formlabs elastic
resin was used as the base material in the material reservoir. 80% percent of the full laser power
was applied in the second exposure stage when the current layer was transferred from the substrate
surface to the previously solidified layers. The amount of energy input during different fabrication
stages was pre-calibrated to meet the requirement of the desired curing depth and width.
4.1.5.6 Multi-material Slicer
We used an open-sourced software, Sli3r, to slice the target designed multi-material 3D
digital model as it has the capability for users to assign different material profiles to various
individual components associated with different dispensers. The concentric filling pattern was
chosen in Sli3r to build the 3D ant model, as shown in Figure 29 and the 3D wheel model, as
shown in Figure 40. The linear filling pattern was chosen in Sli3r to build the tensile test cases, as
shown in Figure 39. A python script was written to modify the G-Code toolpath generated by Sli3r
to encode more customized actuation information for the low-level central controller to decode
67

and drive additional GPIO controlled actuators associated with the linear motion movement and
the dispensing events.
4.1.5.7 Scanning Electron Microscopy
SEM images of the dispensed layer surface, the cross-section of the printed part as shown in
Figure 33, and surface textures of the substrate films, as shown in Figure 48, were taken by using
JSM-6610LV scanning electron microscope.
4.1.6 Limitations and challenges
Currently, the presented hybrid DIW and SLA for multi-material 3D printing process has
been demonstrated for fabricating general multi-material parts, including some non-photocurable
functional material using our designed experimental setup, which opens the door towards 3D
printing wearable electronics, sensors, and soft robotics. However, several issues and potential
challenges have been identified during the experiments. They will be addressed in our future work.
➢ The building time of the proposed process. Since the proposed multi-material building
process utilizes syringes to dispense materials, the building time will increase a lot when
the area of the sliced layer pattern is big. There is also a balance in the selection of the
syringe nozzle tips. The smaller diameter of the syringe tips, the finer detailed features can
be printed. However, it will also increase the building time since the toolpath will be much
longer to fill the pattern in each layer due to the smaller width of the dispensed material.
➢ The calibration process. For traditional DIW and laser-based SLA processes, the
calibration process for each method is much accessible. However, when we integrated two
independent manufacturing process, the sensitivity of the position error will dramatically
increase. As we can imagine, if the laser scanning path is off aligned with the dispensing
path, the whole building process will fail, and material contamination issues will occur
68

accordingly since the laser miss curing portion of the dispensed material. Thus, it is quite
critical to have a reliable calibration process to make sure two subsystems are well aligned.
In reality, the laser system itself is also sensitive to the temperature change, and it will
cause a significant scanning position bias even the initial positioned angle of the Galvano
mirror slightly changes. Therefore, it is necessary to run a calibration process before each
printing to achieve the best performance.
➢ Material compatibility issue. The proposed hybrid multi-material 3D printing method
cannot deal with all the materials. One criterion to choose the potential materials is that the
dispensed material will be continued in the base material as soon as it is dispensed.
Sometimes, due to the material compatibility issue, the dispensed material in the base
material will quickly bulge up to form balls, e.g., dispensing the hydrophilic material into
the hydrophobic base material.
4.1.7 Summary of this work
In summary, we present an innovative 3D printing process that hybrids traditional deposit-
on-demand DIW process and laser-based SLA process that integrates both advantages of two
methods enabling the capability of multi-material 3D printing elegantly. Dynamically tuning the
material properties and multi-material printing has always been challenging for the SLA 3D
printing process previously due to the material contamination issue. Similarly, it is difficult for the
DIW process to stack material layer by layer due to the fluidity of the material and well maintain
the geometric shape. The curing-while-dispensing and multi-exposure mechanisms introduced in
the presented hybrid multi-material 3D printing process makes it possible to make up each
individual process’s shortage and thereby shows its versatility and reliability. A preliminary
diffusion test has been conducted to prove its advance for the immersed dispensing approach
69

compared with the traditional in-air dispensing method. At the same time, it is also useful for
figuring out the optimized printing parameters in terms of the safe distance between the dispensing
nozzle tip and the scanning laser spot. An experimental setup has been designed and built with its
corresponding software system, which seamlessly connects two independent manufacturing
processes. Various test cases have been fabricated to verify the feasibility and functionality of the
proposed hybrid multi-material 3D printing process. Some remaining work will be implemented
in the future including: (1) Developing an in-situ method to assist the motion synchronization of
two involved fabrication processes in real-time to reduce the toolpath alignment error and further
improve the dynamic reliability. (2) Exploring the possibility of the presented method in the
fabrication of advanced materials for biomedical applications.

70

4.2 Highly Removable Water Support for Stereolithography
1

In ancient times, some types of trees can produce resin. The abnormal development of resin
in living trees can result in the formation of amber, as shown in Figure 49. Amber is a fossilized
tree resin, which has been appreciated for its color and natural beauty for a long time. Amber
sometimes contains animal and plant material as inclusions, especially when the resin dropped
onto the ground. Hence an insect may be surrounded by this unexpected tree resin. Over time, the
resin may survive long enough to become amber, and the insect pose inside will last forever[163].
Inspired by this phenomenon in nature, we hope to apply the amber mechanism to additive
manufacturing (AM) processes for support structures.

Figure 49: Amber produced by nature.
In this research, we investigate the fabrication process based on water/ice for the mask-
image-projection-based stereolithography (MIP-SL). In the MIP-SL process, a CAD model is
sliced into a set of two-dimensional (2D) layers with given layer thickness. Each layer is prepared
individually by projecting the masked image of the layer onto the liquid resin surface. After the
UV light exposure, liquid resin is solidified into the sliced layer shape and attach to the previous

1
The full text of this section has been published on the journal of Manufacturing Process and featured as the
Outstanding Paper in 2017 N.A. Manufacturing Research Conference, as Jie Jin, and Yong Chen. "Highly removable
water support for Stereolithography." Journal of Manufacturing Processes 28 (2017): 541-549.
71

layers. To add a water support structure, we investigate the top-down projection system in the
MIP-SL process. A critical thermoelectric device[164, 165] is selected and applied in the MIP-SL
system to freeze the water into ice.
4.2.1 Water support building process study
In this research, a top-down MIP-SL process is applied for building water support
structures. However, unlike the traditional top-down SLA process based on the free surface, the
top resin surface is constrained by a Teflon-coated transparent glass in our approach. Besides,
instead of using a tank to store liquid resin, we adopt a material spreading and removing method
to achieve the desired water support building process. Consequently, the final building part will
be surrounded by solid ice instead of fixing by additional supports that are merged inside the liquid
resin tank.
4.2.1.1 Building platform design
As shown in Figure 50, a thermoelectric cooler is mounted on a linear Z-stage, which can
move up and down along the Z direction. The thermoelectric cooler in our setup is used as the
building platform where the built layers will be grown from. When a positive voltage is applied to
the thermoelectric cooler, the temperature on the top surface will dramatically decrease to below
zero degrees Celsius just in seconds. Thus, it can maintain a continued low-temperature
environment to ensure the success of the building process. Consequently, the thermoelectric cooler
that is used as a building platform will always be kept on during the entire building process until
all the layers have been built.
72

Figure 50: The building platform for the process.
4.2.1.2 Building process illustration
A water dispenser is mounted on the front of tool A as shown in Figure 51a. Adjusting the
linear Z-stage to form a 100 μm gap between the top and bottom cooler, then we move tool A to
spread one layer of water in a given thickness on the top surface of the bottom cooler. Due to the
surface tension, water will be constrained in the gap between the two coolers. After turning on
both of the two coolers, the water in the gap will be converted into ice in several seconds, as shown
in Figure 51b. The bottom surface of the top cooler is coated with a piece of Teflon film in order
to decrease its surface friction such that the top cooler can easily be separated from the ice below.
The reason for building the first ice layer is for the easy separation of the part after the entire
building process has been finished, i.e., the built object can easily be taken away from the building
platform by simply melting the base ice layer. No extra effort is required. It will significantly
prevent the built objects from being damaged by using a scraper in the traditional approach.
As shown in Figure 51c, another tool B will move towards the building platform right after
moving away from tool A. Similar to tool A, a resin dispenser is mounted on the front of tool B,
which will be used to spread one-layer liquid resin on previous solidified layer surface with a
73

desired layer thickness. After that, a pattern image is projected on the resin surface through a
transparent glass. The exposure time for each layer in our setup is ~20 seconds.

Figure 51: A schematic illustration of the water support building process.
A resin vacuum is mounted on the rear side of tool B. After the exposure time, as shown
in Figure 51d, tool B is moved away. During the movement of tool B, the mounted resin vacuum
will be turned on to suck out all the unsolidified residual liquid resin from the building platform,
leaving only one layer of solidified resin pattern. After that, tool A will move towards the building
platform again, as shown in Figure 51e. It will spread water to fill the empty slots in the previous
74

layer and freeze them to ice. Consequently, we will get a layer that consists of both the solidified
resin pattern and solidified ice. Then by repeating the process from Figure 51c - e, another
patterned layer, as shown in Figure 51g, can be fabricated.
The flow chart shown in Figure 52 briefly describes the water support building process.

Figure 52: Flow chart description of the water support building process.
4.2.1.3 A proof of concept

Figure 53: A proof of concept. (a) A designed CAD model; (b) and (c) support structures that are required in the
traditional SLA process; (d) printing in process using water support; (e) and (f) melting the surrounded ice after
the building process; and (g) - (i) the fabricated object in different views.
As shown in Figure 53a, a simple CAD model was designed to verify the concept of the
presented building process. Unlike the traditional SLA process, which will require a support
75

structure as shown in Figure 53b for this CAD model, to ensure the success of the building process,
we used the aforementioned process to directly fabricate the designed CAD model without adding
any support structures. Instead, ice support similar to amber is added in the building process, as
shown in Figure 53d. After finishing the building process, the surrounded ice support was then
melted away, as shown in Figure 53e - f. Consequently, the desired part can be fabricated without
any undesired surface marks, as shown in Figure 53g - i.
4.2.1.4 Study of the impact of ice on dimensional accuracy
The presented ice support building process requires the built part to be embedded in ice.
Since water expands when it freezes, a study of the dimensional accuracy of a built part has been
performed to understand the impact of ice and low temperature on the fabrication accuracy. The
CAD model used in the study is a simple cube whose dimension is 20mm × 20mm × 2.5mm. Two
experimental factors are studied, which are the use of ice and low temperature, as shown in Table
1. Each treatment has three runs, and the observation data are the three-dimensional sizes and the
weight of each built part. A digital caliper is used to measure the dimensions. We measure ten
times in different sections along the same dimension. Each dimensional data, as shown in Table
2, is the mean of the ten measurements. In this study, every part is built in the same position along
with the same bounding wall (introduced in Section 4.2.2.4), and the resin is applied in the first
layer for each part.
By comparing the treatment 1 (Ice - & Temperature -) with the treatment 2 (Ice - &
Temperature +), we can conclude that the temperature will not affect the dimensional accuracy
since the data between treatment1 and treatment 2 are approximately the same. By comparing the
treatment 2 (Ice - & Temperature +) with the treatment 3 (Ice + & Temperature +), we can conclude
that the ice will not affect the dimensional accuracy since the data between treatment1 and
76

treatment 2 are approximately the same. However, the particular photopolymer we investigate in
the study, MakerJuice G plus, is just one of the candidates. The ice and temperature impact on
other photopolymers will be studied in our future research.
Table 1: Factors and levels, the study of ice impact to dimensional accuracy.
Factors
Levels
- +
Ice Without ice With ice
Temperature Room temperature (~ 22 ℃) Low temperature (~ -5 ℃)

Table 2: Study of the impact of ice on dimensional accuracy.
Run
Factors Dimension
Weight (g)
Mean
Ice Temp. X (mm) Y (mm) Z (mm) X (mm) Y (mm) Z (mm) Weight (g)
1 - - 19.00 19.04 2.67 1.133
19.02 19.03 2.65

2 - - 19.03 18.99 2.62 1.116 1.125
3 - - 19.02 19.06 2.65 1.125
4 - + 19.05 19.00 2.64 1.126
19.05 19.02 2.63

5 - + 19.03 19.04 2.62 1.118 1.124
6 - + 19.06 19.01 2.64 1.129
7 + + 19.06 19.05 2.61 1.121
19.05 19.04 2.63

8 + + 19.03 19.04 2.62 1.127 1.124
9 + + 19.05 19.02 2.65 1.123

In the X and Y dimensions, there is about 1 mm dimensional error between the designed
model and the built part for all runs. This might be caused by the projector scale calibration error.
In the Z dimension, there is about 0.1 mm dimensional error between the designed model and the
built parts for all the runs. This error might be from the base layer because we calibrated the
horizontal bottom surface of the building glass window, as shown in Figure 51c. However, the
77

top surface of the building platform is not adjustable, as shown in Figure 50. Hence, the gap
between the bottom surface of the building glass window and the top surface of the building
platform at the very beginning will be a little bigger than the desired 100 μm layer thickness.
In short, we can conclude from the dimensional study that the use of ice and low
temperature shows an insignificant impact on the dimensional accuracy of the original part. This
is reasonable because we dispense the water after the liquid resin is solidified, as shown in Figure
51d and Figure 51e. The material used in this study is rigid after it is solidified. Assuming the ice
expansion force is homogeneous in each dimension, the ice expansion force applied to each
dimension of the part is symmetrical. Hence, the defects due to ice expansion will be insignificant
if the solidified material is reasonably rigid.
4.2.2 Ice surface level control
An interesting phenomenon observed in our initial experiments indicates that the actual
building layer thickness is much larger than what is expected for the layers. Even worse, some
cured resin layers cannot be attached well to the previously built layers. After analyzing the entire
building process, we determine that the water dispense process, as shown in Figure 51e, may cause
the problem. The detailed analysis will be discussed in Section 4.2.2.1, and the solution to address
the problem will be presented in Section 4.2.2.2. Consequently, the layer thickness can be under
control after applying the solution in the experimental cases. A strategy is also introduced in
Section 4.2.2.4 to enhance the maximum building Z height.
4.2.2.1 Analysis of the height stretch problem
As studied in physics, the volume of water will be 10% larger when liquid water is
converted into solid ice. However, it is obvious that the phenomenon observed from our initial
experiments is much worse than this value. One scenario, as shown in Figure 54a, is that some
78

unexpected residual water droplets will remain on the building surface when tool A moves away
from the building platform.

Figure 54: An illustration of the building height stretch problem. (a) Residual water droplet left on the building
surface; (b) Iceball pushes up tool B.
The residual water droplet may be taken from the rear edge of tool A. However, the water
droplet will soon be frozen to a solid ice ball due to the low-temperature environment. Such an ice
ball will stick on the top of the building surface. Hence when the next motion cycle begins, tool B
will be pushed up when it passes through these ice balls. If the height of these ice balls is much
bigger than the designed layer thickness, as shown in Figure 54b, the cured resin height on this
layer will be much larger than the given layer thickness, resulting in the accumulated height that
has a significant error when the process is repeated.
Another scenario is that when tool A is applied to the building surface as shown in Figure
51e, the bottom surface of the top cooler cannot guarantee seamless attachment with the previous
layers because the previous layer surface is formed by another tool. That is, it may leave a rather
thin ice layer on the top of the building surface after tool A moves away from the building platform.
So, when the next resin layer comes, it will not be able to attach to the previously built layers due
to the existing extra intermediate thin ice layer between them.
4.2.2.2 The solution developed to control the surface level
The two problems, as discussed in Section 4.2.2.1, are all caused by a common reason, i.e.,
the residual ice. A developed solution that is presented in this section will focus on how to
79

efficiently remove such residual ice on the top building surface, including both small ice balls and
residual thin ice layer. We propose a water vacuum design, as shown in Figure 55, to address the
problem. A thermal heater is mounted on the bottom surface that is used to preheat the residual ice
droplets or thin ice layer. Thus, the residual molten ice can be easily sucked out via the vacuum
slots of a water vacuum device. A temperature sensor is also applied to the bottom surface
generating a closed-loop control of the heating temperature. The appropriate temperature used in
our experiment is 40 degrees Celsius. A soft rubber blade is also integrated with the water vacuum,
which is embedded on the rear side of the vacuum. The rubber is about 0.5 mm higher than the
bottom surface of the vacuum. The purpose of the rubber blade is to clean all the residual molten
ice that is missed by the water vacuum head.

Figure 55: A water vacuum design. (a) The CAD model of the designed water vacuum; (b) the fabricated water
vacuum; and (c) the internal structure of the vacuum design.
As shown in Figure 55c, a parametric design model is developed, which allows us to use
experiments to determine the following critical parameters for the vacuum structure, including
vacuum slot width δ, slot angle θ, and moving speed Vvacuum.
➢ Vacuum slot width δ: This slot is designed to suck in the residual liquid water. The slot
width is a vital factor to affect the vacuum sucking force. In our experiment, we have tried
the width from 0.2 mm to 2 mm, and found that 0.2 mm is the most efficient one;
80

➢ Moving speed Vvacuum: In our experiments, the optimum Vvacuum is found to be 15mm/s. The
maximum speed that the rotary stage can achieve is 30 mm/s. Due to the time required for
heat transfer between the thermal heater and the ice residues, the optimum value is chosen
by balancing the maximum heat transferring time and the minimum building time;
➢ Slot angle 𝜃 : the designed slot angle 𝜃 is constrained to 𝜃 ≤ tan
− 1
𝑉 𝑠 𝑢𝑐𝑘 𝑉 𝑣 𝑎𝑐 𝑢𝑢𝑚 , where Vsuck
denotes the speed of sucking airflow and is determined by the air pump and the slot width
δ. The vacuum requires certain rigidity so that 𝜃 cannot be set to a very small value. In our
experiment, 𝜃 is set to 45 degrees.
4.2.2.3 Evaluation of the designed water vacuum
The improvement in the cleaning efficiency using the water vacuum, as discussed in
Section 4.2.2.2, is obvious. The height increase problem, as discussed in Section 4.2.2.1, has been
solved. The improvement is illustrated in Figure 56.

Figure 56: Illustration of the efficiently designed water vacuum.
4.2.2.4 The strategy to enhance the building height
In the building process, we introduce an artificial bounding wall to enhance the ability to
build relatively high parts in the Z direction, as shown in Figure 57.
81

Figure 57: (a) A test case of the designed CAD model to be built; (b) a top view of the built layers with ice at the
initial few layers; and (c) a top view of the built layers with ice in the middle of the CAD model.
The designed bounding wall will have a certain distance offset from the boundary of the
input CAD model. The bounding wall is built along with the model that is being built. The purpose
of the bounding wall is to constrain the deposited water to surround the sliced layers, preventing
it from spreading towards the boundary of the building platform.

Figure 58: The difference between using a bounding wall and without using the bounding wall. (a) Top view of the
rectangular ice boundary when the bounding wall is applied; (b) top view of the irregular ice boundary without
using the bounding wall; and (c) - (d) cross-section views of with and without a bounding wall, respectively.
In comparison, without the bounding wall, the cross-section shape of solidified ice will
have a pyramid effect that can be observed from the experiments, resulting in the limited printing
height that can be achieved, as shown in Figure 58d.
4.2.3 Experimental hardware and software systems
4.2.3.1 Hardware system
A prototype system has been built to verify the developed building process. The hardware
setup of the water-support-based SLA system is shown in Figure 59. In the designed system, an
82

ultraviolet (UV) light projector is used. The wavelength of the light source is 405 nm. The power-
consuming of the light source is ~7 watts. The projection resolution of the projector is 1280 × 800,
and the building platform size is 62 mm × 62 mm. A precise linear stage from Parker Inc. is used
as the elevator to drive the building platform in the Z-axis. A precise rotary stage from Parker Inc.
is used as the actuator for rotating the water dispensing tool and the resin dispensing tool. A high-
performance 4-axis motion controller with 28 GPIOs from Dynomotion Inc. (Calabasas, CA) is
used to drive the steppers on the stages. Two critical thermoelectric devices are used to control the
freezing of water to ice from the bottom and top. And the electrical architecture of the hardware
system is shown in Figure 60.

Figure 59: The prototype hardware system for the water support building process.
83

Figure 60: The electrical architecture of the hardware system.
4.2.3.2 Software system
The software system used in the experiments is developed using C ++ language with
Microsoft Visual C ++ compiler. The software integrates the geometry slicing and motion control
features. It also coordinates the projection of the sliced image with the motion movements as well
as the control of pumps and the thermoelectrical coolers. The graphical user interface of the
developed software is shown in Figure 61.

Figure 61: Software system user interface.
84

4.2.3.3 Material selection
MakerJuice G plus from Makerjuice labs is used in the experiments. This type of resin is
selected since it is not sensitive to the temperature change and can work on the temperature range
used in our study.
4.2.4 Results and discussion
Case studies based on the developed water-support-based SLA process have been
performed. The building results of the designed parts verified the developed prototyping system.
Also, they demonstrated the capabilities of the new method in fabricating challenging geometric
shapes that are difficult to be built by traditional support structures.
4.2.4.1 3D ant

Figure 62: The building process and result of the 3D ant.
A 3D ant, as shown in Figure 62, was built to verify the water-support-based SLA process
in building small features. In the building process, the building direction of the test case is from
legs to body. As shown in the building results in Figure 63, the surrounded ice support presents
two main benefits. Firstly, the ice support is highly removable that requires a minimum post-
85

processing effort. Secondly, and more importantly, the ice support can prevent delicate features
from being damaged by constraining them during the printing process while adding no force to
them when removing the ice support after the building process.

Figure 63: The built 3D and after removing ice in a microscopic view.
4.2.4.2 3D bee
We also investigated the possibility of the proposed water-support-based SLA process to
fabricate an even more intricate structure, such as a bee with two thin wings. The designed
thickness of the wing is about 0.5 mm, including textures on the wings whose thickness is about
0.2 mm. The digital CAD model and the fabricated part are shown in Figure 64. The ice used as
the support structures was easily removed after the printing was finished. Two different colors of
polymer resin from Makerjuice were used to demonstrate the capability of multi-material SLA 3D
printing in the designed experimental setup. Three different materials in total were involved in this
case, including water. Some textures on the wing were also printed and can be observed in Figure
64.
86

Figure 64: The built 3D bee after removing ice in the room environment.
Furthermore, we used a traditional commercial SLA 3D printer to do a comparison printing
to compare the performance of our proposed approach and the commercial one in terms of dealing
with some intricate features. The commercial printer we used in the test is the ulTra machine from
Envision Tech Inc., which is a quite expensive SLA 3D printer among other commercial SLA
printers.

Figure 65: (a) CAD model of a bee with traditional support structure; (b) printed part using 100 μm layer
thickness; (c) printed part using 50 μm layer thickness.
87

As shown in Figure 65, a lot of support structures have to be generated to ensure the
success of the printing in a traditional SLA printing process. Due to the tiny, fragile legs and wings
in the 3D bee CAD model, the dimension of the supports generated became much smaller than it
is in a regular printing job. Thus, the possibility of successfully printing this 3D bee is little as the
failure rate of building the supports increases when the supports are not rigid enough to grow to a
certain height.
This test case demonstrated that the ice supports in our proposed approach can protect the
fragile features during the fabrication process since the whole delicate features can be surrounded
by solid ice, which can address the limitations of the current traditional support structures.
4.2.4.3 Microfluidics
The current fabrication of the microfluidic device is dominated by manual molding
approaches based on polydimethylsiloxane (PDMS) and thermoplastics which has to involve a
significant amount of human interference during the fabrication process. 3D printing has recently
attracted more and more attention as a promising way to fabricate microfluidic systems due to its
customized, automated, assembly-free 3D fabrication, rapidly decreasing costs, and fast-
improving resolution and throughput.
The current barriers for the traditional SLA process to fabricate the microfluidic devices
are mainly about resolution and resin compatibility. As the microfluidics contain a lot of micro-
scale channels, it will cause the over-cure problem during the fabrication of the traditional SLA
process since the UV light will pass through the transparent material to cure the unsolidified resin
in the channels thus the empty channels will disappear which is undesired. In contrast, it reveals
the advantages of our proposed water-support-based SLA process.
88

As shown in Figure 66 and Figure 67, channels with about 100 μm height were fabricated
by our proposed approach. During the fabrication, the designed empty channels will firstly be filled
with liquid water, and then the liquid water will soon be converted to solid ice, which can prevent
the over-cure problem in the SLA process. After the fabrication is finished, the solid ice in the
empty channel can be melted in the room temperature and blew out by the compressed air.

Figure 66: An empty channel fabricated by the water-support-based SLA process and its microscopic cross-
sectional views in different sections.

Figure 67: A Y-type channel fabricated by the water-support-based SLA process and its microscopic cross-
sectional views in different sections.
89

4.2.4.4 Single spring
A single spring, as shown in Figure 68, was built to verify the water-support-based SLA
process in building some flexible structures.

Figure 68: a) and b) Different views of the printed single spring; c) The printed single spring in microscopic views.
A sequence of images, as shown in Figure 69, indicates the ice support was gradually
melted in the room temperature after printing. And finally, the surrounded spring appears without
any human interference and damage in post-processing. The only cost is a short period of time.
90

Figure 69: The ice support for the single spring was melting in the room temperature (23 ° C) after the completion
of the printing.
As shown in Figure 70, we define the aspect ratio d/p to measure the flexibility, where d
denotes the diameter of the spring, and p denotes the distance between the two neighboring rounds
of the spring. The big aspect ratio corresponds to the significant flexibility. Due to gravity, the
spring itself will collapse if there is no constrain put to the spring when the aspect ratio is big. For
the same material, there is a balance point of the aspect ratio. On the left side of the balance point,
the traditional SLA process may be able to fabricate the spring without adding any additional
supports to the spring due to its self-support structure. However, on the right side of the balance
point, gravity mainly dominates the spring’s height when the spring is in a relaxed mode. Thus, it
is quite challenging for the traditional SLA process to fabricate parts with such kind of structures.
A significant amount of the supports has to be added to the original parts to fix the flexible spring
during the printing in the traditional SLA process. Still, it takes time and effort to remove it in the
post-processing stage. What’s worse, removing the supports may cause some damage to the
delicate structures of the original spring, whereas it reveals the advantages of the proposed water
support.
91

Figure 70: Description of the characteristics for the spring.
4.2.4.5 Double helix spring
Figure 71 shows a double helix spring fabricated by the water-support-based SLA process
to demonstrate the capability of the process.

Figure 71: The double helix spring.

92

4.2.4.6 Inclusive spring
Figure 72 shows an inclusive spring fabricated by the water-support-based SLA process
to demonstrate the capability of the process in building complex structures. A small spring is
included in a big spring, as shown in Figure 72.

Figure 72: The inclusive spring.
4.2.5 Limitations and challenges
Currently, the presented water-support-based SLA process has been demonstrated for
fabricating general shapes and parts with delicate and complex features using our designed
experimental setup. However, several issues and potential challenges have been identified during
the experiments. They will be addressed in our future work.
➢ The bonding strength between layers. Since the building process is implemented in a
relatively low-temperature environment, we have not tested the bonding force between
different layers with varying temperature conditions. The temperature impact on the
bonding strength is still unknown (i.e., becoming stronger or weaker, or remaining the
same). The bonding strength may be a potential problem for building more complex
structures.
93

➢ Materials selection. Most liquid resins available in the current market have a high viscosity
as temperature decreases. It is difficult to use them in the experiments. Hence the choice
of appropriate resins for the presented process is limited. The modification of resin with a
chemical solution to address the low-temperature building environment needs to be
investigated.
➢ Frost concentration. The experimental setup was placed in a normal room temperature
environment. The moisture contained in the air will have liquefaction phenomena. They
will convert to the liquid water, finally turn to solid ice and stick on the building surface,
which will significantly affect the success of building the parts. To build a sealed low-
temperature building environment seems necessary in future work.
4.2.6 Summary of this work
A new stereolithography process based on a highly removable water support structure has
been presented. Thermoelectric devices have been incorporated in the building process so that
liquid water can be frozen to ice in the layer-based building process. Since water or ice residues
will significantly affect the curing process of spreading liquid resin, it is critical to control the ice
surface level to ensure the success of the printing process. Our study introduces a novel design of
a vacuum tool with a heater to sufficiently remove the residual ice that is left on the building
surface. An experimental prototype system has been built, which integrates various hardware and
software components. Some representative case results based on the prototyping system are
presented to verify the capabilities of the water-support-based SLA process in building various
challenging geometries. Potential applications for the new AM process, including microfluidic
devices, are under investigation.
94

Considerable work remains to mature the developed process and the correspondingly
developed system. Future work that we are investigating includes: (1) developing multi-layer
microfluidic devices with extremely complex channel paths. (2) exploring applications that are
enabled by our water-support-based SLA process.

95

4.3 A Vibration-assisted Method to Reduce Separation Force for
Stereolithography
2

4.3.1 Design of the vibration-assisted system
In this research, a top-down mask-image-projection based stereolithography (MIP-SL)
process is applied to build desired parts. However, unlike the traditional top-down SLA process in
which a fixed transparent glass constrains the top resin surface, a Teflon-coated transparent glass
is mounted on two aluminum bars, which can vibrate along with the vibration of the membrane in
the paper cone of the loudspeaker (MB42X, Micca Inc.). Instead of using a direct-pull method to
separate the built surface and the glass after a certain exposure time, we enable the vibration of the
glass to break the vacuum environment between the built surface and the glass, which can facilitate
the surface separation.
4.3.1.1 Experimental hardware system
A prototype system has been built to study the effect of the vibration to reduce the
separation force. The hardware setup of the vibration-assisted SLA system is shown in Figure 73.
In the designed system, an UV light projector is used. The wavelength of the light source is 405
nm. The power-consuming of the light source is ~7W. The projection resolution of the projector
is 1280 × 800. The material we used in our experiments is Makerjuice G+. As shown in Figure
73, two ends of the glass are mounted on two aluminum frames, respectively. The other end of the
aluminum frames is mounted on the paper cone of the loudspeaker. The paper cone of the
loudspeaker can deform within a certain degree. When there is a particular electrical signal
transmitted into the speaker, it will cause the paper cone to vibrate and finally transfer the vibration

2
The full text of this section has been published on the journal of Manufacturing Process and featured as the
Outstanding Paper in 2018 N.A. Manufacturing Research Conference, as Jie Jin, Jingfan Yang, Huachao Mao, and
Yong Chen. "A vibration-assisted method to reduce separation force for stereolithography." Journal of Manufacturing
Processes 34 (2018): 793-801.

96

to the glass window mounted on the aluminum frames. Different frequencies of the signal will
cause various amplitudes of the vibration due to resonance. In our experiments, we used a sine
wave as the input signal, and the most effective frequency we have tested in our trials is 60 Hz,
with 50% of the maximum volume. An artificial boundary wall was added to prevent resin filling
into the top surface of the glass window, as shown in Figure 73.

Figure 73: The hardware setup of the vibration-assisted system.
Two pressure sensors were fixed on the flat top surface of the loudspeakers, as shown in
Figure 73. Right above the pressure sensors, there are two vertical aluminum bars mounted on the
aluminum frames, which can vibrate along with the vibration of the paper cone of the loudspeaker.
Two fixtures were installed in the bottom end of the vertical aluminum bars. A steel spring was
used to connect around the flat pad, and the fixture mounted on the vertical aluminum bar. The
purpose of using a steel spring here is to store energy when there is a vibration caused by the paper
cone of the loudspeaker, and the spring will be compressed or released at that moment. But it will
not influence the maximum value of the pressure force applied to the pressure sensor from the
round flat pad above. It can also filter out the noise data caused by the vibration as a similar concept
we learned from the design of the capacitor in the electrical circuit.
97

As shown in Figure 74, the top building layer of liquid resin will be solidified after a
certain exposure time of a mask image. Thus, a vacuum environment is formed between the top
surface of the building part and the bottom surface of the constrained glass. If we try to move down
the building platform to separate the building surface and the glass, the separation force will be
extremely big when a large area of the resin is solidified. Let 𝒇 𝒔 𝒆𝒑𝒂𝒓 𝒂𝒕𝒊𝒐𝒏 ⁡denotes the separation
force as shown in Figure 74. Due to the deformation of the paper cone of the loudspeaker, the
constrained glass will be pulled down a certain distance until which moment the vacuum
environment breaks, and the top building surface detaches from the constrained glass. It will also
cause downward displacement of the vertical aluminum bar during this period, as shown in Figure
74. The paper cone here at this moment can be treated as a hinge for simplification. Meanwhile,
the spring mounted in the bottom end of the vertical aluminum bar will be gradually compressed,
and it will cause an increasing pressure force applied to the pressure force sensor below. As shown
in Figure 74, let 𝒇 𝑳 _ 𝑳𝒊 𝒇𝒕 ⁡ denotes the vertical lift force the aluminum bar A applies on the glass.
Likewise, 𝒇 𝑹 _ 𝑳𝒊 𝒇𝒕 ⁡denotes the vertical lift force the aluminum bar B applies on the glass. Let
𝒇 𝑳 _ 𝑷 𝒖𝒔 𝒉 ⁡ and 𝒇 𝑹 _ 𝑷 𝒖𝒔 𝒉 ⁡denote the horizontal force the aluminum bar A and B applied on the glass,
respectively. From the force balancing, we will have the following equations,
| 𝒇 𝒔 𝒆𝒑𝒂𝒓 𝒂𝒕𝒊𝒐𝒏 | = ⁡ | 𝒇 𝑳 _ 𝑳𝒊 𝒇𝒕 | + ⁡ | 𝒇 𝑹 _ 𝑳𝒊 𝒇𝒕 |
𝒇 𝑳 _ 𝑷 𝒖𝒔 𝒉 + ⁡ 𝒇 𝑹 _ 𝑷 𝒖𝒔 𝒉 = 0
(19)

98

Figure 74: The illustration of the separation force measurement.
4.3.1.2 Calibration for the pressure force sensor
Figure 75 shows the electrical schematics of the application of the pressure sensor (Pololu
Inc., Las Vegas) we used in our experiments. The chosen pressure sensor is a force-sensitive
resistor. The resistance will change along with the change of the pressure force applied to the
sensor.

Figure 75: A schematic of using a pressure force sensor.

99

As shown in Figure 75, the analog/digital converter in the microcontroller is using a 10-
bits register, which means the raw data obtained from the sensor ranges from 0~1024. Obviously,
the data we read from the sensor does not have a linear relationship with the resistance of the
sensor, and the relationship between the resistance of the sensor and the real pressure force sensed
is unknown.
To establish the relationship between the reading and actual force, the calibration, as shown
in Figure 76, was performed. We used screws as a standard weight. We assumed the weight should
always be consistent for the same type of screws. We added the weight at point A, where there
would be a vertical pulling force applied during the printing process, as shown in Figure 74. We
gradually added weight with a certain increment. The total weight and the corresponding data
value read from the sensor are recorded each time.

Figure 76: (a) weights added to calibrate the sensor; (b) using screws as standard weights.
By using the above calibration method, we obtained 16 pairs of data from the calibration
experiments. Then we utilized the relationship between the data read from the sensor and the real
force to plot two scatter diagrams, which were left sensor read vs. real force and right sensor read
vs. real force, as shown in Figure 77a and Figure 77b.
After analyzing the trend of these scatter points, we found that there could be a logarithmic
relationship between the data read from the sensor and the real force. Hence, we decided to use the
100

natural logarithmic function to fit these scatter points. We utilized the Curve Fitting Tool from
MATLAB R2016a to get our fitting curves, as shown in Figure 77. The fitted function is shown
as follows.
𝑓 𝑅 ( 𝑥 ) = ⁡283 . 5 𝑙 𝑜 𝑔 ( 𝑥 ) − 859 . 8 (20)
The R-square here is equal to 0.9942, which is acceptable, and it also means that the basic
equation we used can fit these scatter points perfectly.
Likewise, we got the fitted function for the left sensor as follows.
𝑓 𝐿 ( 𝑥 ) = 159 . 3 𝑙 𝑜 𝑔 ( 𝑥 − 23 . 15 ) − 49 . 08 (21)
The corresponding R-square is 0.9836.

Figure 77: Calibration results for the corresponding force sensors. (a) left force sensor calibrated; (b) right force
sensor calibrated.
101

4.3.2 Analysis of experimental data and modeling
In this section, we implemented the controlled experiments to study the effect of glass
vibration to the separation force. A CAD model with different sections of the area was designed
to investigate the relationship between the printing cross-sectional area and the separation force
and the difference between using glass vibration and without using glass vibration. The data from
the left and right pressure sensors were collected and analyzed in both macro and micro manners.
A mathematical model was established to illustrate the vibration-based separation process.
4.3.2.1 Design of the experimental CAD model
As shown in Figure 78, a CAD model with four different sections of the area was designed.
The dimensional size was labeled in Figure 78. Basically, the cross-sectional shape of each section
was a square, and the cross-sectional area was halved section by section. The reason we integrated
the four different areas in a single model rather than four different models with the different cross-
sectional area was that we wanted to eliminate the unexpected data error between experiments
caused by frequently adjust the hardware setup to achieve the exactly the same level of the flat pad
above the pressure sensor as shown in Figure 73. We assume the setup configuration would keep
consistent during a single experiment.

Figure 78: CAD model used in our experiments.
102

The designed layer thickness was 100 μm, and we had 60 layers for the base section and
40 layers for each of the other three sections. The extra 20 more layers for the base section was to
eliminate the influence of over-compression in the first several building layers.
4.3.2.2 Design of printing process

Figure 79: The flow chart of the designed printing process.
In our experiments, we designed the printing process in the following manner. First of all,
we tried to build 40 layers for the base section using a conventional method, which means we
disabled the vibration for the glass and just directly moved down the platform to separate the
building surface with the constrained glass after a certain exposure time. The exposure time in our
experiments was set as 40 seconds for the first three layers and 12 seconds for the other layers.
After that, we built another 20 layers for the base section with the glass vibration enabled. Then
the fabrication for the second section began. We disabled the glass vibration in the first 20 layers
for the second section and enabled it the latter 20 layers. Likewise, the fabrication processes used
103

for the third and fourth sections were the same as the second one’s. The flow chart of the designed
printing process is shown in Figure 79.
4.3.2.3 Acquisition and analysis of experimental data
We sampled the data at a constant rate of 200 Hz, which is sufficient to capture the force
change. As shown in Figure 80, the raw data read from the left and right sensor with the
corresponding time was recorded.

Figure 80: (a) the raw data obtained from the left sensor; (b) the raw data obtained from the right sensor.
By applying the equation aforementioned in Section 4.3.1.2, we obtained the separation
force by summing up the left and right force measured. The calculated separation force over time
is shown in Figure 81.
104

Figure 81: Data visualization. (a) the calculated separation force over time; (b) a zoom-in view of the turning
point.
As shown in Figure 81, we have labeled the data plotted into eight groups from a-h. The
data from groups a-b was sampled from the base section of the CAD model, as shown in Section
4.3.2.1. The data from groups c-d was sampled from the second section of the CAD model. The
data from groups e-f was sampled from the third section of the CAD model. The data from groups
g-h was sampled from the fourth section of the CAD model. Obviously, the data from groups a, c,
e, and g was obtained from a normal printing process without applying the vibration to the
constrained glass, and the data from groups b, d, f, and h was obtained from a vibration-assisted
printing process with using the vibration to the constrained glass.
By combining the data from groups a, c, e, and g, we can learn that there is a decreasing
trend of the separation force as the building area decreases in the normal printing process. However,
the data from the groups b, d, f, and h shows that the force here is independent with the building
105

area in the vibration-assisted printing process. Specifically, the force in the groups b, d, f, and h
was caused by the upward and downward movement during the vibration rather than a separation
force as it was measured in the normal printing process. It was determined by the amplitude of the
vibration on the glass, which was initially determined by the frequency of the input signal to the
loudspeaker. We have figured out the amplitude of the vibration on the glass is ~ 300 μm. The
method we used to measure the amplitude is described as follows. Firstly, we manually applied a
pressure force at the point where a pulling force would be applied during the vibration-assisted
printing process as shown in Figure 74-point A. Then we gradually increased the pressure force
until the force value we read from the pressure sensor reached the value shown in Figure 80a and
Figure 80b, respectively. The vertical displacement of point A we measured is treated as the
amplitude of the vibration on the glass.

Figure 82: A comparison of the separation force using different printing processes.
Figure 82 above shows a plot of the mean value of the data and standard deviation from
groups of using the normal printing process and vibration-assisted printing process, respectively.
106

Here, we only care about the peak values generated during the separation process, as shown in
Figure 81.
As shown in Figure 82, by comparing the data obtained from the normal printing process
and the vibration-assisted printing process, we can learn that as the building area increases, the
vibration applied on the constrained glass with a particular frequency will help to reduce the
separation force significantly. More importantly, unlike other methods aiming to reduce the
separation force mentioned in Section 2.3, the proposed vibration-assisted printing process will
not dramatically increase the mechanical construction complexity even if the required maximum
building area increases.
As shown in Figure 81b, we zoom in a representative place near the turning point between
the normal printing process and the vibration-assisted printing process in the base section, as
shown in Figure 81a. The polyline from point A to B in Figure 81b denotes an increasing
separation force, which means the separation process began at the point A and reached the
maximum separation force at point B at which moment the building surface was instantly separated
from the constrained glass during the normal printing process. It caused a vertical displacement of
the constrained glass, as shown in Figure 74. The separation force after point B did not disappear
immediately because of the damping effect of the spring, as shown in Figure 73 and mentioned in
Section 4.3.1.1. The spring was under a compressed status from point A to B. The entire separation
process lasted about 1 second in a normal separation process. After finishing building the first 40
layers in the base section, we enabled the vibration to help separate the building surface from the
glass. As shown in Figure 81b, the vibration compressed the spring from point F to G due to the
downward movement of the round flat pad, as shown in Figure 73, and then released the spring
from the point G to H due to the upward movement of the round flat pad. This repeated upward
107

and downward movement during the vibration caused a relatively dense wave pattern, as shown
in Figure 81b. We can learn that the amplitude of the upward and downward movement is quite
consistent from the overall view of the data sampled at different sections during the vibration-
assisted printing process, as shown in Figure 81a.
4.3.3 Modeling of the vibration-based separation process
The dynamics of the system can be modeled as a forced vibration with damping:
𝑚 𝑥 ̈ + 𝑐 𝑥 ̇ + 𝑘𝑢 = 𝐹 0
𝑠𝑖𝑛 ⁡ ( 2 𝜋𝑓𝑡 ) (22)
where m is the mass of the system, c is the damping coefficient, and k is the system’s stiffness.
The input vibration force is generated by a loudspeaker. F0 and f are the amplitude and frequency
of the loudspeaker’s signal, respectively. The steady-state solution[166] is
x = A ∙ sin ⁡ ( 2 𝜋𝑓𝑡 + 𝜙 )
=
1
2 𝜋 𝐹 0
√ 𝑘𝑚
√
( 𝑓 2
−
𝑘 4 𝜋 2
𝑚 )
2
+
𝑐 2
𝑘𝑚
𝑓 2
𝑠 in ⁡( 2 𝜋𝑓𝑡 + 𝜙 )
(23)
Where the vibration amplitude A is subject to the vibration frequency f. Notice that the
vibration approaches steady-state just after several cycles.
In the proposed vibration-assisted system, the separation of the printed part with the
constrained glass can be modeled as a fatigue crack propagation under a cyclical load. And P. Paris
and F. Erdogan[167] demonstrated that the range of the stress intensity factor is the dominant
parameter determining the crack propagation.
The stress intensity factor is modeled[168] as
𝐾 = 𝜎 √ 𝜋𝑎 ∙ 𝜃 ( 𝑎 ) (24)
108

Where 𝜃 ( 𝑎 ) is a geometry related coefficient, and 𝜎 is the stress load. In our setup, 𝜎 is the
overall force divided by the contact area:
𝜎 =
𝑘𝑥
( 𝑤 − 2 𝑎 )
2
(25)
where k is the system’s stiffness, x is the vibration displacement, w is the length of the
printed part, and a is the separation crack length,
According to Paris’ Law[167], the growth rate of the crack length 𝑎 w.r.t cycles 𝑁 is
predicted by

𝑑𝑎
𝑑𝑁
= 𝐶 ( ∆ 𝐾 )
𝑛 (26)
where C and n are material parameters, ∆ 𝐾 is the range of stress intensity factor applied on
the edge of the interface between the printed part and the constrained glass. This value is computed
as:
∆ 𝐾 = 𝐾 𝑚𝑎𝑥
− 𝐾 𝑚 𝑖𝑛
= ⁡ σ
m ax
⁡ √ π a ∙ 𝜃 ( 𝑎 ) − σ
m in
⁡ √ π a ∙ 𝜃 ( 𝑎 )
= ( σ
m ax
− σ
m in
)
√ π a ∙ 𝜃 ( 𝑎 )
=
kA
( w − 2a )
2
√ π a ∙ 𝜃 ( 𝑎 )
=
1
2 𝜋 𝐹 0
√ 𝑘 / 𝑚 √
( 𝑓 2
−
𝑘 4 𝜋 2
𝑚 )
2
+
𝑐 2
𝑘𝑚
𝑓 2
√ π a ∙ 𝜃 ( 𝑎 )
( w − 2a )
2

(27)
By definition, the vibration frequency is 𝑓 = 𝑑𝑁 /𝑑𝑡 . Hence, the separation crack growth
rate is

𝑑𝑎
𝑑𝑡
=
𝑑𝑎
𝑑𝑁
∙
𝑑𝑁
𝑑𝑡
= 𝐶 ( ∆ 𝐾 )
𝑛 𝑓 (28)
109

This is the governing equation that predicts how the separation propagates with respect to
time. This equation indicates that the separation speed is determined by the vibration signal’s
frequency f and amplitude F0, system mass m, system stiffness k, material property C and n, and
the geometry of the contact area. Also, from the equation, we can see that there exists an optimal
frequency that can maximize the separation speed. This optimal frequency can be derived by
differentiating the equation (13) w.r.t the frequency f. Any higher or lower frequency generates a
smaller range of stress intensity factor ∆ 𝐾 , and hence leads to a lower crack progradation rate. Our
experiments also verify this phenomenon.
Notice that, during the vibration, the displacement of the constrained glass was controlled
within the vibration amplitude, which was ~300 μm in our experiments. Hence, the separation
force maintains a small value regardless of the contact area size, and the experimental results are
shown in Figure 82 also verifies this. Such a small separation force enables our process to fabricate
large-area components efficiently.
In comparison, the maximal force in the conventional process without vibrations is largely
affected by the contact area size. Various researches have illustrated that the separation force is
highly determined by the contact area size[148, 157, 158]. Our experimental result in Figure 82
also reveals that the separation force increases rapidly with the increase of the contact area.
4.3.4 Results and discussion
4.3.4.1 The camera shot of the separation process
Multiple images, as shown in Figure 83 and Figure 84, were taken to record the surface
separation process that happened in a random layer chosen in the base section of the part from both
the normal printing process and vibration-assisted printing process. Noticed that the building
platform moved down after the building surface had separated from the constrained glass in a
110

vibration-assisted printing process as described in Figure 79. To capture these images, we firstly
recorded a video using an iPhone 8 Plus with a constant frame rate, which was 60 Hz. Then we
just extracted the frame images from the recorded video and got several pictures with a specific
time interval, which was 1/60 seconds. Although it will take ~1.67 s to separate the building
surface from the constrained glass, it is still competitive among the existing separation methods,
as mentioned in Section 2.3.

Figure 83: The screenshots of a recorded normal separation process.

Figure 84: The screenshots of a recorded vibration-assisted separation process.
4.3.4.2 The impact of the vibration on the dimensional accuracy
We have printed ten same parts as designed in Section 4.3.2.1 and measured the dimensions
along with the X and Y directions in different sections. The measurement results match well with
the designed dimensions. Since the vibration only causes the movement along the z-axis, the
proposed vibration-assisted printing process does not affect the dimensional accuracy in the X and
Y directions. The X-Y dimensional size of the printed part is only determined by the dimensional
111

size of the projection pattern image. As long as the focal length of the projector was kept the same
during the printing process, the dimensional accuracy in X-Y directions can be guaranteed. For
simplicity, we just care about the dimensional accuracy in the Z direction. As shown in Table 3,
we have measured the dimensions along the Z direction, including the height of different individual
sections of the printed parts. The measurement results are quite consistent with designed
dimensions with the designed dimensions except for the dimension in the base section. The
dimensional error in the base section along the Z direction is caused by the over-compression
between the building platform and the top constrained glass. The measurement result along the Z
direction also shows the fact that if the setup hardware is reliable, especially the membrane of the
paper cone in the loudspeaker is highly recoverable which will not cause a permanent deformation
during the vibration, the layer thickness in the Z direction for each layer printed can be guaranteed.
It is not limited to use a loudspeaker to provide the vibration for the designed system, but the
chosen method is low cost. Other alternative ways to generate more reliable vibration will be tried
in our future work.
Table 3: The dimensions along the Z direction.
Height in Z direction /mm
Designed size Mean
Base section 6 5.32
Second section 4 3.97
Third section 4 3.92
Fourth section 4 3.95
4.3.4.3 The impact of the vibration to the building surface quality
From the experiment results, we did not find any significant surface quality difference in
the top building surface between using the normal printing process and the proposed vibration-
assisted printing process. In this section, we focus on comparing the side surface quality of the part
printed by the normal printing process and by the proposed vibration-assisted printing process.
112

Several images were taken in a microscopic view from different sections of the designed part, as
shown in Figure 85. The layer stairs were obviously observed at the side surface, which was built
by the normal printing process while they gradually disappeared after the vibration-assisted
printing process began. This result proves that the proposed vibration-assisted printing process will
significantly enhance the side surface quality of the printed part.

Figure 85: The microscopic views of the printed part in different sections.
4.3.4.4 The impact of the vibration to the small features
As mentioned in Section 4.3.2.3, the amplitude of the vibration on the glass is much larger
than the designed layer thickness, which may potentially cause some unexpected damage to the
original printed part, especially when there are some small features existing in the designed part.
As shown in Figure 86, we have designed a special case with some small features to verify this
problem. The smallest diameter of the designed cylinders is 300 μm, and the smallest wall
thickness of the design is 300 μm. As shown in Figure 86, we have printed two parts with the
same CAD design for comparison. One of them was printed using the normal printing process all
the way, and another one was printed using the proposed vibration-assisted printing process all the
way. From the printed results, we can observe that both parts preserve most of the designed small
113

features. The only significant difference is about the printed column of cylinders, which have the
smallest designed diameter. The one printed by the normal printing process contains 7 of 10
designed cylinders, while the other one printed by the vibration-assisted printing process only
contains 2 of 10 designed cylinders. We have repeated the same comparison experiment 5 times,
and the comparison results were similar to the one, as shown in Figure 86. This suggests that the
proposed vibration-assisted printing process slightly affects some small features in the designed
part.

Figure 86: (a) the designed CAD model for the test; (b) part printed by the vibration-assisted printing process; (c)
a close view of the small features in part (b); (d) part printed by the normal printing process; (e) a close view of
the small features in part (d).
4.3.5 Limitations and challenges
The presented vibration-assisted printing process has been demonstrated to have the ability
to fabricate parts with a large area using our experimental setup. However, several issues and
114

potential challenges have also been identified during the experiments that will be addressed in our
future work.
• Damage to the small features. Since the vibration-assisted process will cause a relatively
big amplitude of the movement on the constrained glass. The unexpected damage caused by this
upward and downward movement to the small features on the designed part cannot be avoided so
far. To address this problem, we can adopt an adaptive vibration frequency during the printing
process instead of using a constant vibration frequency. Thus, for parts that have small features,
we can lower down the vibration frequency or disable the vibration to protect the small features.
If there are a large cross-sectional area and small features existing in the same building layer, it
will become a tradeoff problem. Another method to mitigate the damage is gradually increasing
the vibration amplitude during each separation.
• Reliability of the hardware setup. Currently, we are using a loudspeaker to generate
vibration during our experiments. We haven’t tested an even much larger printing area in our built
setup. The reliability of the current setup is unknown. If the designed building area is much larger
than the current one, the membrane of the paper cone on the loudspeaker may be permanently
damaged. Thus, to build a more reliable setup to do further tests is necessary for a larger building
area.
4.3.6 Summary of this work
A new stereolithography process based on vibration-assisted has been presented. Pressure
sensors have been used to measure the separation force during the printing process. Controlled
experiments have been implemented to make comparisons between using a normal printing
process and the proposed vibration-assisted printing process. The data obtained from the sensor,
115

which was read in real-time, has been visualized and analyzed. An experimental prototype system
has been built, which integrates various hardware and software components. Some representative
test cases based on the prototyping system are presented to verify the capabilities of the proposed
vibration-assisted printing process in significantly reducing the separation force and potentially
building parts with a large cross-sectional area. The designed experimental system is based on a
top-down SLA process, but it can also be applied to the bottom-up SLA process by simply modify
the mechanical design.
Considerable work remains to mature the developed process, and the correspondingly
developed 3D printing system. Some future work that we are investigating includes: (1) refining
the current setup to build parts with a much larger cross-sectional area. (2) developing an adaptive
method to dynamically change the vibration frequency, which is associated with the cross-
sectional area change to protect the small features.

116

Chapter 5: Conclusion and Recommendation for Future Research
5.1 Answering the Research Questions/Testing Hypotheses
As stated in Section 3, we have the following three research questions to answer in this
dissertation to address current limitations in Stereolithography and push it forward:

To answer the above-proposed questions, we have the following hypotheses:

Q1: how can we elegantly realize multi-material 3D printing without having any material
contamination issue?
Q2: how can we easily remove the support structures without leaving unexpected marks on the
surface of the original parts with minimal post-processing effort required?
Q3: how can we reduce the separation force in a constrained-surface-based SLA process for
the large-scale 3D printing with the balance of the printing speed, reliability of printing process,
printable size and life cycle of the constrained surface?
Hypothesis 1.1: An innovative material deposition method can be used to selectively deposit
various materials into desired places since the material switching process in the traditional
multi-material SLA process is the main problem to introduce the material contamination issue.
Hypothesis 1.2: The proposed material deposition method must be a very low-cost method
compared to the existing expensive MJP process.
Hypothesis 1.3: If there exists a low-cost material deposition method, it can be easily
integrated with the current SLA process.
117

Hypothesis 1.1 and Hypothesis 1.2 are tested in Section 4.1, in which we propose a novel
hybrid DIW and SLA process to 3D print multi-material parts without material contamination
issues. The curing while dispensing mechanism has been demonstrated as an elegant method to
deal with multi-material 3D printing, including some non-photocurable materials. Meanwhile, it
is a low-cost method since all the hardware components can be sourced easily. It could potentially
have an impact on the current multi-material 3D printing technologies on the market if it could be
successfully commercialized at an affordable price one day.
Hypothesis 2.1: A special low-cost sacrificial material can be used as the support material for
the SLA process.
Hypothesis 2.2: The proposed support material is phase changeable and can be incorporated
into a multi-material SLA printing process.
Hypothesis 2.3: The removal of the support material after the building process is an automated
process with less or no human interference.

Hypothesis 3.1: Less force but multiple attempts may lead to a more efficient result.
Hypothesis 3.2: The reduction of the separation force for a constrained-surface-based SLA
process can be achieved by introducing a small force to induce the initial crack in the interface
between the newly cured layer and the constrained surface.
Hypothesis 3.3: The initial crack in the interface between the newly cured layer and the
constrained surface can be quickly propagated to the entire interface.
118

Hypothesis 2.1, Hypothesis 2.2, and Hypothesis 2.3 are tested in Section 4.2, in which we
hybrid a widely used material, water, into the traditional SLA process as a sacrificial material.
Water is recyclable and environmentally friendly. It is phase-changeable and can be incorporated
into a multi-material SLA printing process. A thermoelectrical device is used in the experiments
to convert the liquid water into solid ice so that it is able to support the printing structures. A
customized header is used in the tests to convert the residual solid ice to liquid water so that the
surface level of the building layer can be guaranteed. A customized experimental setup has been
designed to automate the proposed process.
Hypothesis 3.1, Hypothesis 3.2, and Hypothesis 3.3 are tested in Section 4.3, in which we
intentionally hybrid a vibration source to the current SLA 3D printing process aiming to vibrate
the constrained building surface in a low frequency to help reduce the overall separation force in
each layer during the printing process. The amplitude of the introduced vibration is small, so that
the force applied to the constrained surface each time is small as well. After multiple attempts of
the vibration, an initial crack can be generated due to the fatigue effect in the interface between
the Teflon film and the newly built layer. During the experiments, we can observe that the initial
crack always happens in the boundary of the interface after vibration is enabled for several seconds
and can be propagated to the center of the building surface, thus finishing the separation. The
theoretical model has indicated that there is an optimal vibration frequency for the designed system,
which is also consistent with the experimental result.
5.2 Engineering Achievements and Scientific Contributions
5.2.1 Engineering Achievements
The proposed hybrid 3D printing processes can address the limitations for the current SLA
process, including hybrid DIW and SLA processes for multi-material SLA 3D printing, water/ice
119

support structure for the SLA process, and vibration-assisted method to reduce the separation force
in the constrained-surface-based SLA process. The engineering achievements of this dissertation
are listed as follow:
1. To address the limitation of material contamination issue in traditional multi-material SLA
3D printing, we propose an innovative 3D printing process which hybrids traditional
deposit-on-demand DIW process and laser-based SLA process that integrates both
advantages of two methods enabling the capability of multi-material 3D printing elegantly.
The curing, while dispensing mechanism introduced in the presented hybrid multi-
material 3D printing process, makes it possible to make up each process’s shortage and
thereby shows its versatility and reliability. An experimental hybrid system has been
designed and built with its corresponding software system, which seamlessly connects
two independent manufacturing processes.
2. To address the limitations of traditional support structures in the SLA process, we propose
a novel approach to build water support that will surround the built objects in each layer.
Our approach is a multi-material printing method based on a top-down projection system.
We address the related challenges in controlling the surface level of ice due to the
volumetric expansion when water converts to ice and other considerations. By optimizing
the process settings, we have designed test cases to illustrate that our approach can
fabricate critical components such as microfluidic devices and complex porous structures.
Consequently, the developed water-support-based MIP-SL process can build highly
sophisticated parts that are previously impossible for the SLA process.
3. To address the limitations of the separation force in the constrained surface SLA process,
we present a novel approach aiming to utilize the vibration in a low frequency to separate
120

the newly cured surface from the constrained surface with a minimum incremental
construction complexity as the building area increases. Our approach is a constrained
surface SLA process based on a top-down projection system. An experimental prototype
is developed. Experimental comparisons have verified the effectiveness of the proposed
approach in significantly reducing the separation force. Consequently, the developed
vibration-assisted SLA process is promising to fabricate parts with a large cross-sectional
area.
5.2.2 Scientific Contributions
When seeking answers to the question: how we can effectively and efficiently fabricate
multi-material and multi-functional objects by innovative hybrid vat photopolymerization, new
knowledge is discovered and listed as below:
1. Characterization of the material diffusion effect. This dissertation studies the material
diffusion effect in the air and the liquid matrix material and finds the significant
difference through fundamental control experiments. Besides, the relationship between
the diffusion rate and the elapsed time is also experimentally characterized.
2. Modeling of the hybrid DIW and SLA process. This dissertation couples the traditional
light-curing polymerization model with the DIW dispensing process. It theoretically
bridges the two independent additive manufacturing processes together and provides a
guideline for tuning control parameters of the proposed hybrid system.
3. Modeling of the vibration-based separation process. This dissertation models the
separation of the printed part with the constrained glass as a fatigue crack propagation
under a cyclical load in the proposed vibration-based separation process. The derived
governing equation that predicts how the separation propagates with respect to time.
121

5.3 Future work
Currently, we have developed three customized hybrid 3D printing processes aiming at
addressing the current limitations in the traditional stereolithography process due to its light-
induced and layer-based fabrication principle. We proposed three methods to address the existing
problems from different perspectives: 1) hybrid traditional DIW and SLA processes to address the
limitations of the current multi-material SLA 3D printing regarding the material contamination
issue. An experimental setup has been built to verify the proposed idea. The curing-while-
dispensing and multi-exposure mechanisms are the keys to the success of the proposed hybrid
multi-material 3D printing process. 2) incorporating water/ice as the sacrificial material into the
traditional SLA fabrication process to address the limitations of the conventional support structures
in the SLA process. An experimental prototype system based on a multi-material MIP-SL 3D
printing process has been built, which integrates various hardware and software components. Some
representative case results based on the prototyping system are presented to verify the capabilities
of the proposed water-support-based SLA process in building various challenging geometries. 3)
introducing the vibration with a certain frequency to help easily separate the newly cured layer
from the constrained surface in the traditional SLA process. Some representative test cases based
on the prototyping system with the analytical results are presented to verify the capabilities of the
proposed vibration-assisted printing process in significantly reducing the separation force and
potentially building parts with a large cross-sectional area.
However, there is still some work left to accomplish the proposed research goal. Firstly, it
is important to come up with an innovative calibration method to fully automate the calibration
between involved dispenser nozzles and the calibration between the laser scanning system and the
material dispensing system in real-time for the proposed hybrid DIW and SLA multi-material
122

printing process; Secondly, it is critical to investigate the potential solutions to solve the issue of
frost concentration happened on the building surface during the ice support fabrication processes.
Thirdly, we need to develop an adaptive method to dynamically change the optimal vibration
frequency and other parameters that are associated with the change of the cross-sectional area to
protect the small features during the separation. Finally, more effort should be put on exploring
the applications for the proposed hybrid 3D printing processes. For example, the proposed hybrid
DIW and SLA multi-material 3D printing process can be used to fabricate more complex
anisotropic composite materials, such as wearable anisotropic sensors, soft robotics, and
biomedical devices. The proposed ice supports may benefit the fabrication of the microfluidics
since it is quite challenging for the traditional SLA process to fabricate small empty channels.
Overall, the remaining work can be classified into two directions:
1. Refining the proposed fabrication processes to improve the performance, including:
1) propose an innovative calibration method to fully automate the calibration process
in real-time for the proposed hybrid DIW and SLA multi-material printing process
to improve the reliability of the overall system further.
2) investigate the potential solutions to solve the issue of frost concentration that
happened on the building surface during the ice support fabrication processes.
3) develop an adaptive method to dynamically change the optimal vibration frequency
and other parameters that are associated with the change of the cross-sectional area
to protect the small features during the separation.
2. Exploring potential applications:
1) the proposed hybrid DIW and SLA multi-material 3D printing process can be used
to fabricate more complex anisotropic composite materials, such as wearable
123

electronics, soft robotics, and biomedical devices. The advance of the proposed
method is to selectively insert various materials, even the materials with high
viscosity, in the photocurable matrix material layer by layer to form a multi-
functional object. The potential functional materials can be liquid metal, magnetized
particles, bio-inks, etc. Multiple combinations of the functional materials in the
library can be made to realize various purposes in a single process, which is
impossible for traditional manufacturing processes.
2) the proposed ice supports may benefit the fabrication of the microfluidics since it is
quite challenging for the traditional SLA process to fabricate small empty channels.
Besides, the water is a bio-compatible material that is potentially ideal for building
biomaterials containing living cells whose structures may require additional
supporting material.
3) the proposed vibration-assisted separation method may possibly be applied to the
current commercial constrained surface-based SLA 3D printers to reduce the
separation force when the building area is significantly large. Controllable vibration
device, such as a piezo actuator, can be used to generate the vibration with a
particular frequency.

124

Bibliography

[1] K. V. Wong, and Aldo Hernandez, "A review of additive manufacturing," ISRN
Mechanical Engineering, 2012 (2012).

[2] I. Gibson, David W. Rosen, and Brent Stucker, Additive manufacturing technologies. New
York: Springer, 2014.

[3] N. Guo, and Ming C. Leu, "Additive manufacturing: technology, applications and research
needs," Frontiers of Mechanical Engineering, vol. 8, no. 3, pp. 215-243, 2013.

[4] W. Gao, Yunbo Zhang, Devarajan Ramanujan, Karthik Ramani, Yong Chen, Christopher
B. Williams, Charlie CL Wang, Yung C. Shin, Song Zhang, and Pablo D. Zavattieri, "The
status, challenges, and future of additive manufacturing in engineering," Computer-Aided
Design, vol. 69, pp. 65-89, 2015.

[5] C. W. Hull, "Apparatus for production of three-dimensional objects by stereolithography.,"
U.S. Patent 4,575,330,, 1986.

[6] Y. H. Chen, C. T. Ng, and Y. Z. Wang, "Generation of an STL file from 3D measurement
data with user-controlled data reduction," The international journal of advanced
manufacturing technology, vol. 15, no. 2, pp. 127-131, 1999.

[7] M. Szilvśi-Nagy, and G. Y. Matyasi, "Analysis of STL files," Mathematical and Computer
Modelling, vol. 38, no. 7-9, pp. 945-960, 2003.

[8] B. Khoshnevis, "Automated construction by contour crafting—related robotics and
information technologies," Automation in construction, vol. 13, no. 1, pp. 5-19, 2004.

[9] B. Khoshnevis, Dooil Hwang, Ke-Thia Yao, and Zhenghao Yeh, "Mega-scale fabrication
by contour crafting," International Journal of Industrial and Systems Engineering, vol. 1,
no. 3, pp. 301-320, 2006.

[10] S. Bose, S. Vahabzadeh, and A. Bandyopadhyay, "Bone tissue engineering using 3D
printing," Materials today, vol. 16, no. 12, pp. 496-504, 2013.

[11] H. N. Chia and B. M. Wu, "Recent advances in 3D printing of biomaterials," Journal of
biological engineering, vol. 9, no. 1, p. 4, 2015.

[12] B. Leukers et al., "Hydroxyapatite scaffolds for bone tissue engineering made by 3D
printing," Journal of Materials Science: Materials in Medicine, vol. 16, no. 12, pp. 1121-
1124, 2005.

125

[13] W. Zhu, X. Ma, M. Gou, D. Mei, K. Zhang, and S. Chen, "3D printing of functional
biomaterials for tissue engineering," Current opinion in biotechnology, vol. 40, pp. 103-
112, 2016.

[14] A. Pfister, R. Landers, A. Laib, U. Hü bner, R. Schmelzeisen, and R. M ü lhaupt,
"Biofunctional rapid prototyping for tissue ‐ engineering applications: 3D bioplotting
versus 3D printing," Journal of Polymer Science Part A: Polymer Chemistry, vol. 42, no.
3, pp. 624-638, 2004.

[15] B. N. Peele, T. J. Wallin, H. Zhao, and R. F. Shepherd, "3D printing antagonistic systems
of artificial muscle using projection stereolithography," Bioinspiration & biomimetics, vol.
10, no. 5, p. 055003, 2015.

[16] M. Wehner et al., "An integrated design and fabrication strategy for entirely soft,
autonomous robots," Nature, vol. 536, no. 7617, pp. 451-455, 2016.

[17] T. Wallin, J. Pikul, and R. Shepherd, "3D printing of soft robotic systems," Nature Reviews
Materials, vol. 3, no. 6, pp. 84-100, 2018.

[18] J. Rossiter, P. Walters, and B. Stoimenov, "Printing 3D dielectric elastomer actuators for
soft robotics," in Electroactive Polymer Actuators and Devices (EAPAD) 2009, 2009, vol.
7287: International Society for Optics and Photonics, p. 72870H.

[19] J. T. Muth et al., "Embedded 3D printing of strain sensors within highly stretchable
elastomers," Advanced Materials, vol. 26, no. 36, pp. 6307-6312, 2014.

[20] A. D. Valentine et al., "Hybrid 3D printing of soft electronics," advanced Materials, vol.
29, no. 40, p. 1703817, 2017.

[21] S. Z. Guo, K. Qiu, F. Meng, S. H. Park, and M. C. McAlpine, "3D printed stretchable tactile
sensors," Advanced Materials, vol. 29, no. 27, p. 1701218, 2017.

[22] P. Maharjan, R. Toyabur, and J. Park, "A human locomotion inspired hybrid nanogenerator
for wrist-wearable electronic device and sensor applications," Nano Energy, vol. 46, pp.
383-395, 2018.

[23] N. Bhattacharjee, Arturo Urrios, Shawn Kang, and Albert Folch, "The upcoming 3D-
printing revolution in microfluidics," Lab on a Chip, vol. 16, no. 10, pp. 1720-1742, 2016.

[24] C. M. B. Ho, Sum Huan Ng, King Ho Holden Li, and Yong-Jin Yoon, "3D printed
microfluidics for biological applications," Lab on a Chip, vol. 15, no. 18, pp. 3627-3637,
2015.

[25] A. K. Au, Wilson Huynh, Lisa F. Horowitz, and Albert Folch, "3D‐printed microfluidics,"
Angewandte Chemie International Edition, vol. 55, no. 12, pp. 3862-3881, 2016.

126

[26] A. K. Au, Nirveek Bhattacharjee, Lisa F. Horowitz, Tim C. Chang, and Albert Folch, "3D-
printed microfluidic automation," Lab on a chip, vol. 15, no. 8, pp. 1934-1941, 2015.

[27] S. J. P. BY BAGRAT GRIGORYAN, DANIEL C. CORBETT, DANIEL W. SAZER,
CHELSEA L. FORTIN, ALEXANDER J. ZAITA, PAUL T. GREENFIELD, NICHOLAS
J. CALAFAT, JOHN P. GOUNLEY, ANDERSON H. TA, FREDRIK JOHANSSON,
AMANDA RANDLES, JESSICA E. ROSENKRANTZ, JESSE, "Multivascular networks
and functional intravascular topologies within biocompatible hydrogels," SCIENCE, vol.
03 May, pp. 458-464, 2019.

[28] A. Lasemi, Deyi Xue, and Peihua Gu, "Recent development in CNC machining of freeform
surfaces: a state-of-the-art review.," Computer-Aided Design, vol. 42, no. 7, pp. 641-654,
2010.

[29] G. I. Peterson, Johanna J. Schwartz, Di Zhang, Benjamin M. Weiss, Mark A. Ganter,
Duane W. Storti, and Andrew J. Boydston, "Production of materials with spatially-
controlled cross-link density via vat photopolymerization," ACS applied materials &
interfaces, vol. 8, no. 42, pp. 29037-29043, 2016.

[30] M. L. Griffith, and John W. Halloran, "Freeform fabrication of ceramics via
stereolithography," Journal of the American Ceramic Society, vol. 79, no. 10, pp. 2601-
2608, 1996.

[31] J. Stampfl, S. Baudis, C. Heller, R. Liska, A. Neumeister, R. Kling, A. Ostendorf, and M.
Spitzbart, "Photopolymers with tunable mechanical properties processed by laser-based
high-resolution stereolithography," Journal of Micromechanics and Microengineering, vol.
18, no. 12, p. 125014, 2008.

[32] B. Dhariwala, Elaine Hunt, and Thomas Boland, "Rapid prototyping of tissue-engineering
constructs, using photopolymerizable hydrogels and stereolithography," Tissue
engineering, vol. 10, no. 9-10, pp. 1316-1322, 2004.

[33] Z. Weng, Jianlei Wang, T. Senthil, and Lixin Wu, "Mechanical and thermal properties of
ABS/montmorillonite nanocomposites for fused deposition modeling 3D printing,"
Materials & Design, vol. 102, pp. 276-283, 2016.

[34] G. Siqueira, Dimitri Kokkinis, Rafael Libanori, Michael K. Hausmann, Amelia Sydney
Gladman, Antonia Neels, Philippe Tingaut, Tanja Zimmermann, Jennifer A. Lewis, and
André R. Studart, "Cellulose nanocrystal inks for 3D printing of textured cellular
architectures," Advanced Functional Materials, vol. 27, no. 12, p. 1604619, 2017.

[35] M. Vaezi, and Shoufeng Yang, "Extrusion-based additive manufacturing of PEEK for
biomedical applications," Virtual and Physical Prototyping, vol. 10, no. 3, pp. 123-135,
2015.

127

[36] R. Jones et al., "RepRap–the replicating rapid prototyper," Robotica, vol. 29, no. 1, pp.
177-191, 2011.

[37] B. N. Turner, R. Strong, and S. A. Gold, "A review of melt extrusion additive
manufacturing processes: I. Process design and modeling," Rapid Prototyping Journal,
2014.

[38] Y. L. Yap, Chengcheng Wang, Swee Leong Sing, Vishwesh Dikshit, Wai Yee Yeong, and
Jun Wei, "Material jetting additive manufacturing: an experimental study using designed
metrological benchmarks," Precision engineering, vol. 50, pp. 275-285, 2017.

[39] H. Yang, Jingying Charlotte Lim, Yuchan Liu, Xiaoying Qi, Yee Ling Yap, Vishwesh
Dikshit, Wai Yee Yeong, and Jun Wei, "Performance evaluation of projet multi-material
jetting 3D printer," Virtual and Physical Prototyping, vol. 12, no. 1, pp. 95-103, 2017.

[40] H. Sugitani, Hiroto Matsuda, and Masami Ikeda, "Liquid-jetting head," U.S. Patent
4,611,219, 1986.

[41] S. Meteyer, Xin Xu, Nicolas Perry, and Yaoyao Fiona Zhao, "Energy and material flow
analysis of binder-jetting additive manufacturing processes," Procedia CIRP, vol. 15, pp.
19-25, 2014.

[42] J. Deckers, Jef Vleugels, and Jean-Pierre Kruth, "Additive manufacturing of ceramics: a
review," Journal of Ceramic Science and Technology, vol. 5, no. 4, pp. 245-260, 2014.

[43] N. Afshar-Mohajer, C.-Y. Wu, T. Ladun, D. A. Rajon, and Y. Huang, "Characterization of
particulate matters and total VOC emissions from a binder jetting 3D printer," Building
and Environment, vol. 93, pp. 293-301, 2015.

[44] F. Fina, Alvaro Goyanes, Simon Gaisford, and Abdul W. Basit, "Selective laser sintering
(SLS) 3D printing of medicines," International journal of pharmaceutics, vol. 529, no. 1,
pp. 285-293, 2017.

[45] L. E. Murr, Edwin Martinez, Krista N. Amato, Sara M. Gaytan, Jennifer Hernandez, Diana
A. Ramirez, Patrick W. Shindo, Frank Medina, and Ryan B. Wicker, "Fabrication of metal
and alloy components by additive manufacturing: examples of 3D materials science,"
Journal of Materials Research and technology, vol. 1, no. 1, pp. 42-54, 2012.

[46] I. Yadroitsev, I. Shishkovsky, Ph Bertrand, and I. Smurov, "Manufacturing of fine-
structured 3D porous filter elements by selective laser melting," Applied Surface Science,
vol. 255, no. 10, pp. 5523-5527, 2009.

[47] A. Dawood, B. Marti Marti, V. Sauret-Jackson, and A. Darwood, "3D printing in
dentistry," British dental journal, vol. 219, no. 11, p. 521, 2015.

128

[48] S. M. Thompson, Linkan Bian, Nima Shamsaei, and Aref Yadollahi, "An overview of
Direct Laser Deposition for additive manufacturing; Part I: Transport phenomena,
modeling and diagnostics," Additive Manufacturing, vol. 8, pp. 36-62, 2015.

[49] J. M. Wilson, Cecil Piya, Yung C. Shin, Fu Zhao, and Karthik Ramani, "Remanufacturing
of turbine blades by laser direct deposition with its energy and environmental impact
analysis," Journal of Cleaner Production, vol. 80, pp. 170-178, 2014.

[50] N. Shamsaei, Aref Yadollahi, Linkan Bian, and Scott M. Thompson, "An overview of
Direct Laser Deposition for additive manufacturing; Part II: Mechanical behavior, process
parameter optimization and control," Additive Manufacturing, vol. 8, pp. 12-35, 2015.

[51] M. Vaezi, Srisit Chianrabutra, Brian Mellor, and Shoufeng Yang, "Multiple material
additive manufacturing–Part 1: a review: this review paper covers a decade of research on
multiple material additive manufacturing technologies which can produce complex
geometry parts with different materials," Virtual and Physical Prototyping, vol. 8, no. 1,
pp. 19-50, 2013.

[52] R. Gauvin, Ying-Chieh Chen, Jin Woo Lee, Pranav Soman, Pinar Zorlutuna, Jason W.
Nichol, Hojae Bae, Shaochen Chen, and Ali Khademhosseini, "Microfabrication of
complex porous tissue engineering scaffolds using 3D projection stereolithography,"
Biomaterials, vol. 33, no. 15, pp. 3824-3834, 2012.

[53] M. P. Lee, Geoffrey JT Cooper, Trevor Hinkley, Graham M. Gibson, Miles J. Padgett, and
Leroy Cronin, "Development of a 3D printer using scanning projection stereolithography,"
Scientific reports, vol. 5, p. 9875, 2015.

[54] Z. Wang, Raafa Abdulla, Benjamin Parker, Roya Samanipour, Sanjoy Ghosh, and
Keekyoung Kim, "A simple and high-resolution stereolithography-based 3D bioprinting
system using visible light crosslinkable bioinks," Biofabrication, vol. 7, no. 4, p. 045009,
2015.

[55] A. P. Zhang, Xin Qu, Pranav Soman, Kolin C. Hribar, Jin W. Lee, Shaochen Chen, and
Sailing He, "Rapid fabrication of complex 3D extracellular microenvironments by dynamic
optical projection stereolithography," Advanced materials, vol. 24, no. 31, pp. 4266-4270,
2012.

[56] M. N. Cooke, John P. Fisher, David Dean, Clare Rimnac, and Antonios G. Mikos, "Use of
stereolithography to manufacture critical ‐ sized 3D biodegradable scaffolds for bone
ingrowth," ournal of Biomedical Materials Research Part B: Applied Biomaterials: An
Official Journal of The Society for Biomaterials, The Japanese Society for Biomaterials,
and The Australian Society for Biomaterials and the Korean Society for Biomaterials, vol.
64, no. 2, pp. 65-69, 2003.

[57] P. Jacobs and C. Hull, "Introduction to Rapid Prototyping and Manufacturing," Rapid
Prototyping & Manufacturing: Fundamentals of Stereolithography, pp. 1-25, 1992.
129

[58] J.-P. Kruth, Ming-Chuan Leu, and Terunaga Nakagawa., "Progress in additive
manufacturing and rapid prototyping," Cirp Annals, vol. 47, no. 2, pp. 525-540, 1998.

[59] Z. Weng, Yu Zhou, Wenxiong Lin, T. Senthil, and Lixin Wu, "Structure-property
relationship of nano enhanced stereolithography resin for desktop SLA 3D printer,"
Composites Part A: Applied Science and Manufacturing, vol. 88, pp. 234-242, 2016.

[60] H. Xing, Bin Zou, Shasha Li, and Xiangsong Fu, "Study on surface quality, precision and
mechanical properties of 3D printed ZrO2 ceramic components by laser scanning
stereolithography," Ceramics International, vol. 43, no. 18, pp. 16340-16347, 2017.

[61] X. Luo, J. Li, and M. Lucas, "Galvanometer scanning technology for laser additive
manufacturing," in Laser 3D Manufacturing IV, 2017, vol. 10095: International Society
for Optics and Photonics, p. 1009512.

[62] R. Raman, Basanta Bhaduri, Mustafa Mir, Artem Shkumatov, Min Kyung Lee, Gabriel
Popescu, Hyunjoon Kong, and Rashid Bashir, "High ‐ resolution projection
microstereolithography for patterning of neovasculature," Advanced healthcare materials,
vol. 5, no. 5, pp. 610-619, 2016.

[63] A. S. Limaye, and David W. Rosen, "Process planning method for mask projection micro-
stereolithography," Rapid Prototyping Journal, vol. 13, no. 2, pp. 76-84, 2007.

[64] X. Song, Yong Chen, Tae Woo Lee, Shanghua Wu, and Lixia Cheng, "Ceramic fabrication
using mask-image-projection-based stereolithography integrated with tape-casting,"
Journal of Manufacturing Processes, vol. 20, pp. 456-464, 2015.

[65] C. Zhou, and Yong Chen, "Additive manufacturing based on optimized mask video
projection for improved accuracy and resolution," Journal of Manufacturing Processes,
vol. 14, no. 2, pp. 107-118, 2012.

[66] K. Xu, and Yong Chen, "Mask image planning for deformation control in projection-based
stereolithography process," Journal of Manufacturing Science and Engineering, vol. 137,
no. 3, p. 031014, 2015.

[67] Y. Pan, Xuejin Zhao, Chi Zhou, and Yong Chen, "Smooth surface fabrication in mask
projection based stereolithography," Journal of Manufacturing Processes, vol. 14, no. 4,
pp. 460-470, 2012.

[68] Y. Pan, Chi Zhou, and Yong Chen, "A fast mask projection stereolithography process for
fabricating digital models in minutes," Journal of Manufacturing Science and Engineering,
vol. 134, no. 12, p. 051011, 2012.

[69] C. Zhou, Yong Chen, Zhigang Yang, and Behrokh Khoshnevis, "Digital material
fabrication using mask-image-projection-based stereolithography," Rapid Prototyping
Journal, vol. 19, no. 3, pp. 153-165, 2013.
130

[70] Y. Pan, Chi Zhou, and Yong Chen. "Rapid manufacturing in minutes, "The development
of a mask projection stereolithography process for high-speed fabrication," in American
Society of Mechanical Engineers, 2012.

[71] C. Zhou, Hang Ye, and Feng Zhang, "A novel low-cost stereolithography process based on
vector scanning and mask projection for high-accuracy, high-speed, high-throughput, and
large-area fabrication," Journal of Computing and Information Science in Engineering, vol.
15, no. 1, p. 011003, 2015.

[72] R. C. Luo, Jyh Hwa Tzou, and Wei Zen Lee, "The development of LCD panel display
based rapid prototyping system for advanced manufacturing," in EEE International
Conference on Robotics and Automation. Symposia Proceedings, 2000.

[73] Z. Wang, Nicholas Martin, Delator Hini, Barry Mills, and Keekyoung Kim, "Rapid
fabrication of multilayer microfluidic devices using the Liquid Crystal Display-based
stereolithography 3D printing system," 3D Printing and Additive Manufacturing, vol. 4,
no. 3, pp. 156-164, 2017.

[74] R. C. Luo, and Jyh Hwa Tzou, "Development of an LCD-photomask-based desktop
manufacturing system," IEEE Transactions on Industrial Electronics, vol. 55, no. 10, pp.
3715-3723, 2008.

[75] Y. Yang, Xuan Song, Xiangjia Li, Zeyu Chen, Chi Zhou, Qifa Zhou, and Yong Chen,
"Recent progress in biomimetic additive manufacturing technology: from materials to
functional structures," Advanced Materials, vol. 30, no. 36, p. 1706539, 2018.

[76] R. Liska, M. Schuster, R. InfŘhr, C. Turecek, C. Fritscher, B. Seidl, V. Schmidt et al,
"Photopolymers for rapid prototyping," Journal of Coatings Technology and Research, vol.
4, no. 4, pp. 505-510, 2007.

[77] E. Andrzejewska, "Photopolymerization kinetics of multifunctional monomers," Progress
in polymer science vol. 26, no. 4, pp. 605-665, 2001.

[78] D. F. Swinehart, "The beer-lambert law," Journal of chemical education, vol. 39, no. 7, p.
333, 1962.

[79] J. H. Lee, Robert K. Prud'Homme, and Ilhan A. Aksay, "Cure depth in photopolymerization:
Experiments and theory," Journal of Materials Research, vol. 16, no. 12, pp. 3536-3544,
2001.

[80] "DMD chip," ed: Texas Instruments.

[81] Y. Yang, Zeyu Chen, Xuan Song, Zhuofeng Zhang, Jun Zhang, K. Kirk Shung, Qifa Zhou,
and Yong Chen, "Biomimetic anisotropic reinforcement architectures by electrically
assisted nanocomposite 3D printing," Advanced Materials, vol. 29, no. 11, p. 1605750,
2017.
131

[82] Y. Yang et al., "Electrically assisted 3D printing of nacre-inspired structures with self-
sensing capability," Science advances, vol. 5, no. 4, p. eaau9490, 2019.

[83] X. Wang, Man Jiang, Zuowan Zhou, Jihua Gou, and David Hui, "3D printing of polymer
matrix composites: A review and prospective," Composites Part B: Engineering, vol. 110,
pp. 442-458, 2017.

[84] S. J. Leigh, R. J. Bradley, C. P. Purssell, D. R. Billson, and D. A. Hutchins, "A simple,
low-cost conductive composite material for 3D printing of electronic sensors," PloS one,
vol. 7, no. 11, 2012.

[85] X. Wei et al., "3D printable graphene composite," Scientific reports, vol. 5, p. 11181, 2015.

[86] D. Espalin, J. A. Ramirez, F. Medina, and R. Wicker, "Multi-material, multi-technology
FDM: exploring build process variations," Rapid Prototyping Journal, 2014.

[87] H. L. Tekinalp, Vlastimil Kunc, Gregorio M. Velez-Garcia, Chad E. Duty, Lonnie J. Love,
Amit K. Naskar, Craig A. Blue, and Soydan Ozcan, "Highly oriented carbon fiber–polymer
composites via additive manufacturing," Composites Science and Technology vol. 105, pp.
144-150, 2014.

[88] P. Parandoush and D. Lin, "A review on additive manufacturing of polymer-fiber
composites," Composite Structures, vol. 182, pp. 36-53, 2017.

[89] R. Matsuzaki et al., "Three-dimensional printing of continuous-fiber composites by in-
nozzle impregnation," Scientific reports, vol. 6, p. 23058, 2016.

[90] F. Ning, W. Cong, J. Qiu, J. Wei, and S. Wang, "Additive manufacturing of carbon fiber
reinforced thermoplastic composites using fused deposition modeling," Composites Part
B: Engineering, vol. 80, pp. 369-378, 2015.

[91] G. D. Goh, Y. L. Yap, S. Agarwala, and W. Y. Yeong, "Recent progress in additive
manufacturing of fiber reinforced polymer composite," Advanced Materials Technologies,
vol. 4, no. 1, p. 1800271, 2019.

[92] Q. Fu, E. Saiz, and A. P. Tomsia, "Direct ink writing of highly porous and strong glass
scaffolds for load-bearing bone defects repair and regeneration," Acta biomaterialia, vol.
7, no. 10, pp. 3547-3554, 2011.

[93] J. L. Simon et al., "In vivo bone response to 3D periodic hydroxyapatite scaffolds
assembled by direct ink writing," Journal of Biomedical Materials Research Part A: An
Official Journal of The Society for Biomaterials, The Japanese Society for Biomaterials,
and The Australian Society for Biomaterials and the Korean Society for Biomaterials, vol.
83, no. 3, pp. 747-758, 2007.

132

[94] R. A. Barry III, R. F. Shepherd, J. N. Hanson, R. G. Nuzzo, P. Wiltzius, and J. A. Lewis,
"Direct ‐write assembly of 3D hydrogel scaffolds for guided cell growth," Advanced
materials, vol. 21, no. 23, pp. 2407-2410, 2009.

[95] M. A. Skylar-Scott, J. Mueller, C. W. Visser, and J. A. Lewis, "Voxelated soft matter via
multimaterial multinozzle 3D printing," Nature, vol. 575, no. 7782, pp. 330-335, 2019.

[96] S. C. Ligon, R. Liska, J. r. Stampfl, M. Gurr, and R. Mü lhaupt, "Polymers for 3D printing
and customized additive manufacturing," Chemical reviews, vol. 117, no. 15, pp. 10212-
10290, 2017.

[97] C. Zhou, Y. Chen, Z. Yang, and B. Khoshnevis, "Digital material fabrication using mask ‐
image ‐projection ‐based stereolithography," Rapid Prototyping Journal, 2013.

[98] D. Han, C. Yang, N. X. Fang, and H. Lee, "Rapid multi-material 3D printing with
projection micro-stereolithography using dynamic fluidic control," Additive
Manufacturing, vol. 27, pp. 606-615, 2019.

[99] C.-D. Matte, M. Pearson, F. Trottier-Cournoyer, A. Dafoe, and T. H. Kwok, "Automated
storage and active cleaning for multi-material digital-light-processing printer," Rapid
Prototyping Journal, 2019.

[100] H. C. W, "Apparatus For Production Of Three-dimensional Objects By Stereolithography,"
U.S. Patent Appl. 4575330, 1986.

[101] Y. Y. C. Choong, S. Maleksaeedi, H. Eng, P.-C. Su, and J. Wei, "Curing characteristics of
shape memory polymers in 3D projection and laser stereolithography," Virtual and
Physical Prototyping, vol. 12, no. 1, pp. 77-84, 2017.

[102] L. J. Hornbeck, "The DMD TM projection display chip: a MEMS-based technology," Mrs
Bulletin, vol. 26, no. 4, pp. 325-327, 2001.

[103] C. M. Larson et al., "Direct ink writing of silicon carbide for microwave optics," Advanced
Engineering Materials, vol. 18, no. 1, pp. 39-45, 2016.

[104] D. Chimene, K. K. Lennox, R. R. Kaunas, and A. K. Gaharwar, "Advanced bioinks for 3D
printing: a materials science perspective," Annals of biomedical engineering, vol. 44, no.
6, pp. 2090-2102, 2016.

[105] M. Guvendiren, J. Molde, R. M. Soares, and J. Kohn, "Designing biomaterials for 3D
printing," ACS biomaterials science & engineering, vol. 2, no. 10, pp. 1679-1693, 2016.

[106] J. W. Boley, E. L. White, G. T. C. Chiu, and R. K. Kramer, "Direct writing of gallium ‐
indium alloy for stretchable electronics," Advanced Functional Materials, vol. 24, no. 23,
pp. 3501-3507, 2014.

133

[107] M. A. Skylar-Scott, S. Gunasekaran, and J. A. Lewis, "Laser-assisted direct ink writing of
planar and 3D metal architectures," Proceedings of the National Academy of Sciences, vol.
113, no. 22, pp. 6137-6142, 2016.

[108] B. Chen, Y. Jiang, X. Tang, Y. Pan, and S. Hu, "Fully packaged carbon nanotube
supercapacitors by direct ink writing on flexible substrates," ACS applied materials &
interfaces, vol. 9, no. 34, pp. 28433-28440, 2017.

[109] D. Kokkinis, M. Schaffner, and A. R. Studart, "Multimaterial magnetically assisted 3D
printing of composite materials," Nature communications, vol. 6, no. 1, pp. 1-10, 2015.

[110] Y. Kim, H. Yuk, R. Zhao, S. A. Chester, and X. Zhao, "Printing ferromagnetic domains
for untethered fast-transforming soft materials," Nature, vol. 558, no. 7709, pp. 274-279,
2018.

[111] X. Liu et al., "3D printing of living responsive materials and devices," Advanced Materials,
vol. 30, no. 4, p. 1704821, 2018.

[112] C. Xu, B. Quinn, L. L. Lebel, D. Therriault, and G. L’Espé rance, "Multi-material direct
ink writing (DIW) for complex 3D metallic structures with removable supports," ACS
applied materials & interfaces, vol. 11, no. 8, pp. 8499-8506, 2019.

[113] J. O. Hardin, T. J. Ober, A. D. Valentine, and J. A. Lewis, "Microfluidic printheads for
multimaterial 3D printing of viscoelastic inks," Advanced materials, vol. 27, no. 21, pp.
3279-3284, 2015.

[114] W. Liu et al., "Rapid continuous multimaterial extrusion bioprinting," Advanced materials,
vol. 29, no. 3, p. 1604630, 2017.

[115] W. Wu, A. DeConinck, and J. A. Lewis, "Omnidirectional printing of 3D microvascular
networks," Advanced materials, vol. 23, no. 24, pp. H178-H183, 2011.

[116] C. B. Highley, Christopher B. Rodell, and Jason A. Burdick, "Direct 3D printing of shear ‐
thinning hydrogels into self ‐healing hydrogels," Advanced Materials, vol. 27, no. 34, pp.
5075-5079, 2015.

[117] T. Uchida and H. Onoe, "4D printing of multi-hydrogels using direct ink writing in a
supporting viscous liquid," Micromachines, vol. 10, no. 7, p. 433, 2019.

[118] T. Bhattacharjee et al., "Writing in the granular gel medium," Science advances, vol. 1, no.
8, p. e1500655, 2015.

[119] Y. Chen, Kang Li, and Xiaoping Qian, "Direct geometry processing for telefabrication,"
Journal of Computing and Information Science in Engineering, vol. 13, no. 4, p. 041002,
2013.

134

[120] P. Huang, Charlie CL Wang, and Yong Chen, Algorithms for layered manufacturing in
image space. ASME Press, 2014.

[121] J. anek, Jorge A. Garcia Galicia, and Bedrich Benes, "Clever support: Efficient support
structure generation for digital fabrication," In Computer graphics forum, vol. 33, no. 5,
pp. 117-125, 2014.

[122] B. Panda, Suvash Chandra Paul, Nisar Ahamed Noor Mohamed, Yi Wei Daniel Tay, and
Ming Jen Tan, "Measurement of tensile bond strength of 3D printed geopolymer mortar,"
Measurement vol. 113, pp. 108-116, 2018.

[123] K. Willis, Eric Brockmeyer, Scott Hudson, and Ivan Poupyrev, "Printed optics: 3D
printing of embedded optical elements for interactive devices," in n Proceedings of the
25th annual ACM symposium on User interface software and technology, 2012.

[124] E. Macdonald, Rudy Salas, David Espalin, Mireya Perez, Efrain Aguilera, Dan Muse, and
Ryan B. Wicker, "3D printing for the rapid prototyping of structural electronics," IEEE
access vol. 2, pp. 234-242, 2014.

[125] D. Espalin, Danny W. Muse, Eric MacDonald, and Ryan B. Wicker, "3D Printing
multifunctionality: structures with electronics," The International Journal of Advanced
Manufacturing Technology, vol. 72, no. 5-8, pp. 963-978, 2014.

[126] A. Joe Lopes, Eric MacDonald, and Ryan B. Wicker, "Integrating stereolithography and
direct print technologies for 3D structural electronics fabrication," Rapid Prototyping
Journal vol. 18, no. 2, pp. 129-143, 2012.

[127] M. Kong, G. Shin, S-H. Lee, and I-J. Yoon, "Electrically small folded spherical helix
antennas using copper strips and 3D printing technology," Electronics Letters, vol. 52, no.
12, pp. 994-996, 2016.

[128] A. Goyanes, Asma BM Buanz, Abdul W. Basit, and Simon Gaisford., "Fused-filament 3D
printing (3DP) for fabrication of tablets," International journal of pharmaceutics, vol. 476,
no. 1-2, pp. 88-92, 2014.

[129] A. Goyanes, Hanah Chang, Daniel Sedough, Grace B. Hatton, Jie Wang, Asma Buanz,
Simon Gaisford, and Abdul W. Basit, "Fabrication of controlled-release budesonide tablets
via desktop (FDM) 3D printing," International journal of pharmaceutics, vol. 496, no. 2,
pp. 414-420, 2015.

[130] T. Tagami, Noriko Nagata, Naomi Hayashi, Emi Ogawa, Kaori Fukushige, Norihito Sakai,
and Tetsuya Ozeki, "Defined drug release from 3D-printed composite tablets consisting of
drug-loaded polyvinylalcohol and a water-soluble or water-insoluble polymer filler.,"
International journal of pharmaceutics, vol. 543, no. 1-2, pp. 361-367, 2018.

135

[131] "3D Systems Unveils Water Soluble Rinse-Away Material " 3dprint.com. (accessed 06,
2020).

[132] N. A. Meisel, Amelia M. Elliott, and Christopher B. Williams, "A procedure for creating
actuated joints via embedding shape memory alloys in PolyJet 3D printing," Journal of
intelligent material systems and structures, vol. 26, no. 12, pp. 1498-1512, 2015.

[133] K. A. Seerden, Nuno Reis, Julian RG Evans, Patrick S. Grant, John W. Halloran, and Brian
Derby, "Ink ‐jet printing of wax ‐based alumina suspensions," Journal of the American
Ceramic Society, vol. 84, no. 11, pp. 2514-2520, 2001.

[134] P. Das, Ramya Chandran, Rutuja Samant, and Sam Anand, "Optimum part build
orientation in additive manufacturing for minimizing part errors and support structures,"
Procedia Manufacturing, vol. 1, pp. 343-354, 2015.

[135] F. Fina, Alvaro Goyanes, Simon Gaisford, and Abdul W. Basit, "Selective laser sintering
(SLS) 3D printing of medicines," International journal of pharmaceutics vol. 529, no. 1-
2, pp. 285-293, 2017.

[136] I. adroitsev, I. Shishkovsky, Ph Bertrand, and I. Smurov, "Manufacturing of fine-structured
3D porous filter elements by selective laser melting," Applied Surface Science, vol. 255,
no. 10, pp. 5523-5527, 2009.

[137] J.-W. Choi, Ho-Chan Kim, and Ryan Wicker, "Multi-material stereolithography," Journal
of Materials Processing Technology, vol. 211, no. 3, pp. 318-328, 2011.

[138] P. Sitthi-Amorn, Javier E. Ramos, Yuwang Wangy, Joyce Kwan, Justin Lan, Wenshou
Wang, and Wojciech Matusik, "MultiFab: a machine vision assisted platform for multi-
material 3D printing," ACM Transactions on Graphics (TOG), vol. 34, no. 4, p. 129, 2015.

[139] N. Meisel, and Christopher Williams, "An investigation of key design for additive
manufacturing constraints in multimaterial three-dimensional printing," Journal of
Mechanical Design, vol. 137, no. 11, p. 111406, 2015.

[140] C. Dai, Charlie CL Wang, Chenming Wu, Sylvain Lefebvre, Guoxin Fang, and Yong-Jin
Liu, "Support-free volume printing by multi-axis motion," ACM Transactions on Graphics
(TOG), vol. 37, no. 4, p. 134, 2018.

[141] P. M. Bhatt, Max Peralta, Hugh A. Bruck, and Satyandra K. Gupta, "Robot assisted
additive manufacturing of thin multifunctional structures," in ASME 2018 13th
International Manufacturing Science and Engineering Conference, 2018.

[142] Y. Chen, Chi Zhou, and Jingyuan Lao, "A layerless additive manufacturing process based
on CNC accumulation," Rapid Prototyping Journal, vol. 17, no. 3, pp. 218-227, 2011.

136

[143] X. Song, Yayue Pan, and Yong Chen, "Development of a low-cost parallel kinematic
machine for multidirectional additive manufacturing," Journal of Manufacturing Science
and Engineering, vol. 137, no. 2, p. 021005, 2015.

[144] H. Mao, Chi Zhou, and Yong Chen, "LISA: Linear immersed sweeping accumulation,"
Journal of Manufacturing Processes, vol. 24, pp. 406-415, 2016.

[145] Y. Pan, Chi Zhou, Yong Chen, and Jouni Partanen, "Multitool and multi-axis computer
numerically controlled accumulation for fabricating conformal features on curved
surfaces," Journal of Manufacturing Science and Engineering, vol. 136, no. 3, p. 031007,
2014.

[146] F. P. Melchels, Jan Feijen, and Dirk W. Grijpma, "A review on stereolithography and its
applications in biomedical engineering," Biomaterials, vol. 31, no. 24, pp. 6121-6130,
2010.

[147] H. Mao, Yuen-Shan Leung, Yuanrui Li, Pan Hu, Wei Wu, and Yong Chen, "Multiscale
Stereolithography Using Shaped Beams," Journal of Micro and Nano-Manufacturing vol.
5, no. 4, p. 040905, 2017.

[148] Y.-M. Huang, and Cho-Pei Jiang, "On-line force monitoring of platform ascending rapid
prototyping system," Journal of materials processing technology, vol. 159, no. 2, pp. 257-
264, 2005.

[149] Y. Pan, Yong Chen, and Chi Zhou, "Fast recoating methods for the projection-based
stereolithography process in micro-and macro-scales," in Proceeding of solid freeform
fabrication symposium, Austin, Texas, 2012.

[150] Y. Pan, Chi Zhou, and Yong Chen, "Rapid manufacturing in minutes: The development
of a mask projection stereolithography process for high-speed fabrication," in ASME 2012
International Manufacturing Science and Engineering Conference collocated with the 40th
North American Manufacturing Research Conference and in participation with the
International Conference on Tribology Materials and Processing, 2012.

[151] Y. Chen, Huachao Mao, and Xiangjia Li, "Mask Video Projection Based Stereolithography
With Continuous Resin Flow," U.S. Patent 15/187,713, 2016.

[152] F. Liravi, Sonjoy Das, and Chi Zhou, "Separation force analysis and prediction based on
cohesive element model for constrained-surface Stereolithography processes," Computer-
Aided Design, vol. 69, pp. 134-142, 2015.

[153] P. J. Bá rtolo, ed, Stereolithography: materials, processes and applications. Springer
Science & Business Media, 2011.

137

[154] I. Kovalenko, Maryna Garan, Andrii Shynkarenko, Petr Zelený, and Jiří Šafka,
"Examining the relationship between forces during stereolithography 3D printing and
geometric parameters of the model," in MATEC Web of Conferences, 2015.

[155] A. Shkolnik, and Ali El-Siblani, "Method of reducing the force required to separate a
solidified object from a substrate," U.S. Patent 8,326,024, 2012.

[156] K.-C. Syao, "Stereolithography apparatus," U.S. Patent 9,452,567, 2016.

[157] H. Ye, Abhishek Venketeswaran, Sonjoy Das, and Chi Zhou, "Investigation of separation
force for constrained-surface stereolithography process from mechanics perspective,"
Rapid Prototyping Journal, vol. 23, no. 4, pp. 696-710, 2017.

[158] Y. Pan, Haiyang He, Jie Xu, and Alan Feinerman, "Study of separation force in constrained
surface projection stereolithography," Rapid Prototyping Journal, vol. 23, no. 2, pp. 353-
361, 2017.

[159] J. R. Tumbleston, David Shirvanyants, Nikita Ermoshkin, Rima Janusziewicz, Ashley R.
Johnson, David Kelly, Kai Chen et al, "Continuous liquid interface production of 3D
objects," Science, vol. 347, no. 6228, pp. 1349-1352, 2015.

[160] C. T. Culbertson, S. C. Jacobson, and J. M. Ramsey, "Diffusion coefficient measurements
in microfluidic devices," Talanta, vol. 56, no. 2, pp. 365-373, 2002.

[161] M. Rubinstein and R. H. Colby, Polymer physics. Oxford university press New York, 2003.

[162] P. Badger. "Capacitive Sensing Library."
https://playground.arduino.cc/Main/CapacitiveSensor/ (accessed 05, 2020).

[163] G. O. Poinar, Life in amber. Stanford University Press, 1992.

[164] R. Venkatasubramanian, Edward Siivola, Thomas Colpitts, and Brooks O'quinn, "Thin-
film thermoelectric devices with high room-temperature figures of merit," Nature, vol. 413,
no. 6856, p. 597, 2001.

[165] R. E. Simons, and R. C. Chu, "Application of thermoelectric cooling to electronic
equipment: a review and analysis," in Sixteenth Annual IEEE Semiconductor Thermal
Measurement and Management Symposium, 2000.

[166] M. Gé radin, and Daniel J. Rixen, Mechanical vibrations: theory and application to
structural dynamics. John Wiley & Sons, 2014.

[167] P. Paris, and Fazil Erdogan, "A critical analysis of crack propagation laws," Journal of
basic engineering, vol. 85, no. 4, pp. 528-533, 1963.

138

[168] H. Tada, Paul C. Paris, and George R. Irwin, The stress analysis of cracks. Handbook, Del
Research Corporation, 1973.

Abstract (if available)

Abstract Stereolithography Apparatus (SLA) is an additive manufacturing (AM) process in which liquid photopolymer resin is cross-linked and converted to hardened plastic. As a light-induced layer-based manufacturing process, there are several fundamental problems we are facing now limiting the potential of the SLA process due to its nature of fabrication principle, which seems unavoidable in the traditional SLA process. Hybrid Vat Photopolymerization (Hybrid VPP) is the smart integration of technologies or materials in other fields to address the current limitations in the conventional Vat Photopolymerization process to make it more efficient and more widely used. ❧ Firstly, SLA can achieve high resolution and fabricating intricate parts at a fast speed. However, material contamination issues caused by the nature of the vat-based process hinder the development of multi-material SLA 3D printing. ❧ Secondly, the layer-based fabrication manner means the fabrication is a unidirectional process during which the building part can only grow in a single direction. Thus, some overhanging structures will occur in some layers due to the geometry complexity. Therefore, a significant amount of additional support structures is required to be added to the fabricated structures for an intricate computer-aided design (CAD) model to ensure it can be successfully fabricated. However, these support structures may be difficult to remove. Even worse, the removal of the support structures may cause unexpected damage to delicate features and undesired surface quality. ❧ Thirdly, it is so essential to control the constant flatness of each layer in the layer-based SLA process since the layer flatness will significantly affect the attachment between two adjacent layers and the dimensional accuracy of the fabricated part. It has been a popular way to add a constrained surface to ensure the surface flatness in the newly cured layer. However, for a constrained surface-based top-down or bottom-up SLA process, it will be challenging to separate the building layers from the constrained glass due to the near-vacuum environment between the built layer and the constrained glass. The separation force will increase when the contact area increases. ❧ In this research, we propose three hybrid 3D printing processes aiming to address the aforementioned limitations in the current traditional SLA process. Hybrid 3D printing for stereolithography is the smart introduction and integration of technologies or phenomena in other fields to make current stereolithography more efficient and more widely used. It connects something that seems unrelated together to achieve enhanced performance, which is impossible for the conventional SLA process. ❧ When it comes to multi-material 3D printing, a hybrid 3D printing method by incorporating the advances from Direct Ink Writing (DIW) and SLA processes is presented. The proposed method uses syringes to dispense the liquid-like materials inside another photocurable base material. It exploits the Galvano mirror-controlled laser beam to cure it or its surrounding material by following the dispensed material trace, thus retaining the shape after the material leaves the nozzle. As exemplars, a 3D ant with intricate features inside the transparent polymer and a tire with selectively printing stiff and soft materials are fabricated. The proposed method is versatile to build heterogeneous materials with various functionalities for real-world application. ❧ To address the limitation of the current support structures in the SLA process, a novel SLA process using highly removable and widely available water as supports is presented. The process uses solid ice to surround the built parts in the layer-by-layer fabrication process. A cooling device is used to freeze the water into ice for each layer. The photocurable resin is spread on the ice surface and then solidified by a projected mask image. Accordingly, a complex 3D object can be fabricated without using traditional support structures. After the fabrication process has been finished, the additional ice structure can be easily removed and leave no undesired marks on the bottom surfaces. Several test cases are presented to show the effectiveness of the presented method. ❧ As for the separation force problem in the constrained surface-based SLA process, a novel separation approach for the SLA process utilizing vibration-assisted glass is presented. A pair of general loudspeakers were used to provide a low-frequency vibration for the constrained glass used in the SLA process. Two force sensors were used to measure the separation force in real-time. Controlled experiments were implemented, and the corresponding data were collected and analyzed. The analyzed results have demonstrated that the proposed method can significantly reduce the separation force for the SLA process. ❧ The future work includes: 1) explore the possibility of the presented hybrid multi-material 3D printing method in the fabrication of advanced materials for biomedical applications. 2) investigate the potential solutions to solve the issue of frost concentration happened on the building surface during the ice support fabrication process, and explore more potential functional applications that can benefit from the proposed ice support which are impossible for the traditional SLA process to achieve in the past, such as the microfluidics. 3) develop an adaptive method to dynamically change the optimal vibration frequency and other parameters that are associated with the change of the cross-sectional area to protect the small features during the separation. ❧ The proposed Hybrid VPP is a new systematic strategy to address the limitations in the traditional SLA process, which opens a door towards the development of wearable electronics, soft robotics, and tissue engineering.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Hybrid vat photopolymerization processes for viscous photocurable and non-photocurable materials

PDF

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

PDF

Multi-scale biomimetic structure fabrication based on immersed surface accumulation

PDF

Scalable polymerization additive manufacturing: principle and optimization

PDF

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

PDF

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

PDF

Deformation control for mask image projection based stereolithography process

PDF

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

PDF

Fabrication-aware machine learning for accuracy control in additive manufacturing

PDF

Statistical modeling and machine learning for shape accuracy control in additive manufacturing

PDF

Fabrication of ultrasound transducer and 3D-prinitng ultrasonic device

$Metasurfaces in 3D applications: multiscale stereolithography and inverse design of diffractive optical elements for structured light$

PDF

Metasurfaces in 3D applications: multiscale stereolithography and inverse design of diffractive optical elements for structured light

PDF

Deformable geometry design with controlled mechanical property based on 3D printing

PDF

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

PDF

Machine learning-driven deformation prediction and compensation for additive manufacturing

PDF

2D ultrasonic transducer array’s design and fabrication with 3D printed interposer and applications

PDF

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

PDF

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

PDF

Experimental and numerical techniques to characterize structural properties of fresh concrete

PDF

Evaluation of the printing trueness of CAD-CAM maxillary complete denture bases fabricated by using two different DLP 3D printers

Asset Metadata

Creator Jin, Jie (author)

Core Title Hybrid vat photopolymerization: methods and systems

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Industrial and Systems Engineering

Publication Date 09/18/2020

Defense Date 07/03/2020

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

Tag additive manufacturing,direct ink writing,embedded 3D printing,hybrid vat photopolymerization,ice support,multi-material 3D printing,OAI-PMH Harvest,separation force,stereolithography,support structure,vibration,water support

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chen, Yong (committee chair), Huang, Qiang (committee member), Zhou, Qifa (committee member)

Creator Email gingergmxy@gmail.com,jiejin@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-374559

Unique identifier UC11666261

Identifier etd-JinJie-8968.pdf (filename),usctheses-c89-374559 (legacy record id)

Legacy Identifier etd-JinJie-8968.pdf

Dmrecord 374559

Document Type Dissertation

Rights Jin, Jie

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