University of Southern California Dissertations and Theses

Material and process development and optimization for efficient manufacturing of polymer composites

(USC Thesis Other)

Material and process development and optimization for efficient manufacturing of polymer composites

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
Material and Process Development and Optimization for
Efficient Manufacturing of Polymer Composites
by
Jung Hwan Shin
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
(Chemical Engineering)
August 2021
ii
Acknowledgements
I first would like to thank my research advisor, Professor Steven Nutt, for all his support
and guidance throughout my Ph.D. degree. He has taught me to become an independent researcher
who knows how to define a scientific problem, investigate the problem to arrive at a solution, and
communicate my research deliverables effectively. Earning a Ph.D. degree in engineering has
challenged me to overcome moments of uncertainty and frustration, but Professor Nutt has always
put me back on the right track, encouraging me to regain my self-confidence and positive attitude.
I also would like to acknowledge my colleagues at M.C. Gill Composites Center.
Throughout my Ph.D., I have received tremendous amount of help and support from all my fellow
Ph.D. students and post-doctorate scholars here at the Composites Center. I would like to especially
thank Dr. Daniel Kim, Dr. Timotei Centea, and Dr. Mark Anders for always imbuing fresh
perspectives, providing passionate guidance, and being my career role models. I also had the
luxury of working with our lab manager Yunpeng Zhang, who I could always rely on when it came
to equipment handling or maintenance. And lastly, I want to thank our design team members,
Mikhail Cabillan and Jinkun Liu, for their efforts in building composite manufacturing tools.
The financial support for this work were provided by Airbus Institute for Engineering
Research (AIER, Airbus and Korean Air), M.C. Gill Composites Center, and NASA Johnson Space
Center. Materials used in this work were donated from several companies, including Hexcel, Cerex
Advanced Fabrics, Airtech International, and Solvay.
Finally, I cannot imagine how I would have completed my Ph.D. degree without the
endless support and love from my family—my wife Seung Hee Han, my son Aiden Jaebin Shin,
my parents, and my mother-in-law—and friends. Thank you everyone, and I love you all!
iii
Table of Contents
Acknowledgements ......................................................................................................................... ii
List of Tables ................................................................................................................................. vii
List of Figures .............................................................................................................................. viii
Abstract ........................................................................................................................................ xiii
Chapter 1. Introduction ................................................................................................................... 1
1.1 Motivation ............................................................................................................................. 1
1.2 CFRP Manufacturing Methods .............................................................................................. 3
1.2.1 Liquid Composite Molding (LCM) ................................................................................. 3
1.2.2 Prepreg Processing .......................................................................................................... 7
1.3 Approach .............................................................................................................................. 10
1.3.1 Material Characterization .............................................................................................. 10
1.3.2 In Situ Process Diagnostics ........................................................................................... 13
1.3.3 Process Modeling .......................................................................................................... 15
1.3.4 Process Simulation ........................................................................................................ 17
1.4 Objectives & Scope ............................................................................................................. 18
Chapter 2. Effects of Post-Infusion Dwell on Vacuum Infusion of Thermoset Composites
iv
Toughened by Thermoplastic Interlaminar Veils .......................................................................... 21
2.1 Abstract ................................................................................................................................ 21
2.2 Introduction ......................................................................................................................... 21
2.3 Experimental ........................................................................................................................ 25
2.3.1 Materials ........................................................................................................................ 25
2.3.2 Cure Kinetics & Rheology Characterization ................................................................. 26
2.3.3 Permeability Measurement & Infusion Simulation ....................................................... 29
2.3.4 Part Manufacture & Quality Analysis ........................................................................... 30
2.4 Results & Discussion ........................................................................................................... 32
2.4.1 Cure Kinetics & Rheology Characterization ................................................................. 32
2.4.2 Permeability Characterization & Infusion Simulation .................................................. 40
2.4.3 Cure Cycle Modification & Part Manufacture .............................................................. 42
2.4.4 Part Quality Analysis ..................................................................................................... 47
2.5 Conclusions ......................................................................................................................... 50
Chapter 3. In Situ Resin Age Assessment Using Dielectric Analysis and Resin Cure Map for
Efficient Vacuum Infusion ............................................................................................................ 52
3.1 Abstract ................................................................................................................................ 52
3.2 Introduction ......................................................................................................................... 52
v
3.3 Experimental ........................................................................................................................ 56
3.3.1 Materials ........................................................................................................................ 56
3.3.2 Cure Kinetics & Rheology ............................................................................................ 56
3.3.3 Resin Aging ................................................................................................................... 58
3.3.4 Filling Simulation .......................................................................................................... 58
3.4 Results & Discussion ........................................................................................................... 59
3.4.1 Cure Kinetics, Rheology, and Dielectric Analysis ........................................................ 59
3.4.2 Resin Aging ................................................................................................................... 66
3.4.3 Infusion Process Map .................................................................................................... 68
3.4.4 Process Validation & Simulation Refinement ............................................................... 72
3.5 Conclusions ......................................................................................................................... 78
Chapter 4. Thermoplastic Prepreg with Partially Polymerized Matrix: Material and Process
Development for Efficient Part Manufacturing ............................................................................ 81
4.1 Abstract ................................................................................................................................ 81
4.2 Introduction ......................................................................................................................... 81
4.3 Experimental ........................................................................................................................ 84
4.3.1 Materials ........................................................................................................................ 84
4.3.2 Polymerization & Resin Characterization ..................................................................... 85
vi
4.3.3 Prepreg Laminate Fabrication & Characterization ........................................................ 87
4.3.4 Laminate Thermoforming & Analysis .......................................................................... 89
4.3.5 Fully Polymerized Prepreg Laminate ............................................................................ 90
4.4 Results & Discussion ........................................................................................................... 91
4.4.1 Polymerization & Resin Characterization ..................................................................... 91
4.4.2 Prepreg Laminate Fabrication & Characterization ........................................................ 97
4.4.3 Laminate Thermoforming & Analysis ........................................................................ 103
4.5 Conclusions ....................................................................................................................... 108
Chapter 5. Conclusions and Future Work .................................................................................... 111
5.1 Conclusions & Contributions ............................................................................................. 111
5.1.1 Vacuum Infusion Projects ............................................................................................. 111
5.1.2 CFRTP Prepreg Processing Project ..............................................................................114
5.2 Broader Implications ..........................................................................................................115
5.3 Recommendations for Future Work ....................................................................................116
References ....................................................................................................................................119

List of Tables
Table 2-1. Values of the cure kinetics model parameters for RTM6 epoxy resin ............................. 34
Table 2-2. Values of the mechanical viscosity model parameters for RTM6 epoxy resin .............. 36
Table 2-3. Gelation times of RTM6 epoxy resin at different reaction temperatures ....................... 37
Table 2-4. Values of the ion viscosity model parameters for RTM6 epoxy resin ............................. 39
Table 3-1. Values of the cure kinetics model parameters for RTM6 epoxy resin ............................. 61
Table 3-2. Values of the mechanical viscosity model parameters for RTM6 epoxy resin .............. 63
Table 3-3. Gelation times of RTM6 epoxy resin at different reaction temperatures ....................... 63
Table 3-4. Values of the ion viscosity model parameters for RTM6 epoxy resin ............................. 65
Table 3-5. Heated filling simulation results for infusion lengths of 530 and 600 mm (α0 = 0.10)
........................................................................................................................................................................... 73
Table 3-6. Two-dimensional polynomial fitting model parameters for the flow contour maps .... 74
Table 3-7. Design guidelines for VI heated filling process ................................................................... 77
Table 4-1. Test matrix for two-ply PMMA pre-polymer prepreg laminate fabrication ................... 88
Table 4-2. Values of the fitting parameters & derivatives for MMA polymerization at 90 °C ...... 94

List of Figures
Figure 1-1. Global composites market share by end use, 2019 (%) [4] ............................................... 1
Figure 1-2. Autoclave system manufactured by ASC Process Systems [5] ........................................ 2
Figure 1-3. Resin Transfer Molding (RTM) setup & process cycle [10] ............................................. 4
Figure 1-4. Vacuum infusion (VI) process setup [11] .............................................................................. 5
Figure 1-5. Vacuum infusion process used to manufacture a yacht hull [13] ..................................... 6
Figure 1-6. Conventional thermoset and thermoplastic prepregs [18,19] ............................................ 7
Figure 1-7. Automated tape laying machine [22]...................................................................................... 8
Figure 1-8. Schematic of a vacuum bag assembly [23] ........................................................................... 9
Figure 1-9. TGA instrument and sample TGA thermal stability curve [24] ...................................... 10
Figure 1-10. DSC instrument and sample DSC thermoset cure curve [25] ....................................... 11
Figure 1-11. Rheometer and sample epoxy cure rheology curve [26,27] .......................................... 12
Figure 1-12. Dielectric cure monitoring (DCM) system with both reusable & disposable
interdigitated dielectric sensors [29] .......................................................................................................... 13
Figure 1-13. Interdigitated dielectric sensor on polyimide substrate [30] ......................................... 14
ix
Figure 1-14. Ex situ material characterization & in situ process monitoring for process modeling
........................................................................................................................................................................... 15
Figure 1-15. Example PAM-RTM simulation results for wind blade infusion [35] ........................ 17
Figure 2-1. Heat flow and reversing heat flow profiles during the residual cure of the epoxy resin
after isothermal dwell at 140 °C before and after annealing peak removal....................................... 27
Figure 2-2. Experimental setup of vacuum-driven constant injection pressure radial flow
permeability measurement ........................................................................................................................... 30
Figure 2-3. MDSC data showing heat flow profiles measured during (a) dynamic ramp tests of
RTM6 resin, (b) isothermal dwell tests at filling temperatures (80-120 °C), (c) isothermal dwell
tests at curing temperatures (120-180 °C), and (d) cure kinetics model fitting results, expressed in
terms of resin degree of cure vs. time (experimental: solid line, model-prediction: dotted line) . 32
Figure 2-4. Rheology data showing mechanical viscosity profiles of RTM6 resin during (a)
isothermal dwell tests at filling temperatures (80-120 °C), (b) isothermal dwell tests at curing
temperatures (120-180 °C) and viscosity model fitting results (experimental: scatter, model-
prediction: line), and (c) storage and loss modulus profiles measured during resin cure at 160 °C
........................................................................................................................................................................... 35
Figure 2-5. (a) DEA data showing ion viscosity profiles of the epoxy resin during isothermal dwell
tests at curing temperatures (120-180 °C), (b) α-based resin cure map, in which the dashed lines
represent degree of cure isolines, and (c) η-based resin cure map, in which the dashed lines are
mechanical viscosity isolines ...................................................................................................................... 37
Figure 2-6. (a) Progressive images of flow fronts captured during the plain weave carbon fiber
fabric permeability measurements, and (b) N term (terms in the bracket of Equation 2-9) plotted
against flow time for permeability measurements of plain weave carbon fiber fabric and non-
woven polyamide veil ................................................................................................................................... 41
Figure 2-7. Infusion simulation results for a multilayer laminate containing 16 plies of plain weave
carbon fiber fabric and 7 plies of polyamide veils, expressed in filling factor (degree of saturation)
gradient ............................................................................................................................................................ 42
x
Figure 2-8. Profiles of the MRCC (dashed) and the modified cure cycle (solid) showing
temperature and predicted profiles of (a) degree of cure and (b) mechanical viscosity, and (c) ion
viscosity measured during Case C (modified cure cycle) part manufacture ..................................... 44
Figure 2-9. Cross-sectional micrographs of the vacuum-infused laminates (Case A: containing no
veils and manufactured using the MRCC, Case B: toughened by veils and manufactured using the
MRCC, and Case C: also toughened by veils but manufactured using the modified cure cycle) 47
Figure 2-10. (a) Images and (b) ultrasonic C-scan results of the impacted laminates after drop tower
impact tests ...................................................................................................................................................... 48
Figure 2-11. Force-time profiles of Case A, B, and C laminates measured during 30-J drop tower
impact tests ...................................................................................................................................................... 49
Figure 3-1. (a) MDSC data showing heat flow profiles measured during isothermal dwell tests,
and (b) experimental (solid line) and cure kinetics model-predicted (dotted line) degree of cure
profiles for the isothermal dwell tests ....................................................................................................... 60
Figure 3-2. (a) Rheology data showing experimental (dotted) and viscosity model-predicted (solid
line) mechanical viscosity profiles for isothermal dwell tests, and (b) storage and loss modulus
profiles measured at 160 °C ........................................................................................................................ 62
Figure 3-3. (a) DEA data showing ion viscosity profiles during isothermal dwell tests, (b) resin
cure map, in which filled squares (connected by solid isoline) and unfilled circles (connected by
dashed isoline) represent experimental and model-predicted data points respectively, (c) ion
viscosity model parameters plotted against resin degree of cure, exhibiting a linear relationship,
and (d) model-predicted and experimental ion viscosity values at the resin pre-heating temperature
of 80 °C ............................................................................................................................................................ 64
Figure 3-4. (a) Degree of cure accrued during aging (αaged) plotted against room temperature out-
time, and (b) degree of cure profile at 80 °C, predicted by the cure kinetics model ....................... 67
Figure 3-5. (a) Infusion process map for aged resin with α0 of 0.05, infused at 120 °C, and (b) more
comprehensive process map for infusion temperatures of 120, 130, and 140 °C, using aged resin
with α0 of 0.10 ................................................................................................................................................ 68
xi
Figure 3-6. Time required for aged resins (α0 = 0.01, short dashed; 0.05, dashed; 0.10, solid line)
to reach different levels of viscosity, plotted against viscosity at infusion temperatures of (a) 120
and 130 °C, and (b) 120 and 140 °C .......................................................................................................... 70
Figure 3-7. Representative simulation results of VI heated filling of aged resin (α0 = 0.10) for (a)
infusion length of 530 mm and infusion temperature of 120 °C, expressed in terms of fill time
gradient, and infusion length of 670 mm and infusion temperatures of (b) 130 and (c) 140 °C,
expressed in terms of filling factor gradient ............................................................................................ 72
Figure 3-8. (a) 3D and (b) 2D flow contour maps that show the maximum possible flow distance
as a function of both resin age (α0) and infusion temperature (Tinfusion) ............................................. 75
Figure 3-9. Methodology of material characterization and infusion process adjustment for VI
process .............................................................................................................................................................. 76
Figure 4-1. (a) Aluminum tool for prepreg laminate fabrication (top cover & bottom container),
and (b) sample two-ply PMMA pre-polymer prepreg laminate (P-1) ................................................ 87
Figure 4-2. MDSC data showing (a) heat flow & temperature profiles during the isothermal dwell
(70 °C) and dynamic ramp steps of MMA free-radical bulk polymerization, and (b) reversing &
non-reversing heat flow profiles during the dynamic ramp step ......................................................... 91
Figure 4-3. (a) Measured degree of monomer conversion profile and mathematical kinetics model
fitting, and (b) viscosity profile of MMA polymerization at 90 °C .................................................... 93
Figure 4-4. Degree of monomer conversion profiles as a function of aging time for three different
storage temperatures, with sample XAged calculation .............................................................................. 96
Figure 4-5. (a) Final resin contents of two-ply prepreg laminates with varying polymerization times
at 90 °C, (b) experimental prepreg fabrication map at 90 °C for two-ply prepreg laminate, and (c)
temperature and time dependent prepreg process map, showing empirical (90 °C) and estimated
(70 & 80 °C) process windows ................................................................................................................... 98
Figure 4-6. Methodology of material and process development for the design of reactive CFRTP
prepreg .......................................................................................................................................................... 100
xii
Figure 4-7. (a) Normal pressure against the top plate retraction time for prepregs with varying out-
times, and (b) maximum force required to detach the top plate, or the sample tack .................... 101
Figure 4-8. (a) Partially polymerized thermoplastic prepreg draped onto 75° corner mold right after
fabrication, and (b) conventional thermoplastic prepreg with fully polymerized matrix ............ 102
Figure 4-9. Cross-sectional micrographs of thermoformed (a) 2-ply (P-1) and (b) 8-ply (TH-1)
laminates, free of voids .............................................................................................................................. 103
Figure 4-10. FTIR absorbance spectrum collected from the final thermoformed PMMA matrix
........................................................................................................................................................................ 105
Figure 4-11. Conventional PMMA prepreg with fully polymerized matrix, before and after
thermoforming at different temperatures—90, 150, and 200 °C ...................................................... 106

Abstract
Advanced composite materials based on continuous fiber-reinforced polymers (CFRPs)
feature high specific strength and stiffness. CFRPs are increasingly used across multiple industries,
including aerospace, automotive, energy, and sporting goods to manufacture structural parts.
CFRPs must be consolidated at high temperature and pressure to facilitate fiber bed saturation and
void removal. Thus, high-performance composite structures are traditionally produced using an
autoclave, which is a pressurized vessel.
Yet, the use of autoclave incurs high capital and operating costs and imposes inflexible
manufacturing environment because size and availability of autoclave is often limited. The
predicted increase in market demand for CFRPs is driving exploration of faster and more cost-
efficient out-of-autoclave (OoA) composite processing methods. Yet, OoA processes are more
likely to yield less robust parts, and thus requires further process developments and adjustments
for more extensive and efficient applications.
Vacuum infusion (VI) is a promising alternative to conventional autoclave prepreg process,
especially for manufacture of large and complex unitized composite structures. In VI, a dry fiber
preform is placed on a one-sided rigid mold and sealed with a flexible vacuum bag. Then, resin is
infused into the vacuum bag under vacuum (negative) pressure, then heated and cured. Interest in
VI has grown rapidly in the recent years, particularly in the aerospace industry, which seeks to
reduce the manufacturing costs associated with prepreg processing.
However, because only vacuum pressure is applied during infusion, VI requires use of
low-viscosity thermoset resin, which is inherently brittle and susceptible to impact damage. To
enhance impact performance, we introduced non-woven thermoplastic veils at interlaminar regions
xiv
of thermoset composites and addressed the associated process challenges (Chapter 2). The use of
thermoplastic veils in VI preforms increased impact resistance but also induced non-uniform flow
fronts during infusion, leading to high porosity in the final laminate. We modified the conventional
VI cure cycle to include a low-temperature post-infusion (LTPI) dwell to allow more time for resin
to equilibrate pressure and redistribute during the VI post-filling stage, achieving full saturation of
dry interlaminar regions.
Also, VI endures slow infusion rate and long fill time because only limited pressure is
available to drive resin flow during infusion. To mitigate this issue, resin is often heated to
accelerate infusion, although doing so can increase resin degree of cure and viscosity with flow
time and distance, requiring careful selection of process parameters. In Chapter 3, we assessed the
physical state of VI resin using in situ dielectric process diagnostics coupled with cure modeling.
Then, the heated filling process was simulated to guide accurate on-the-fly process adjustments
for more efficient use of aged resin. The results demonstrated that the VI process parameters must
be adjusted not only for part size and geometry but also for resin age.
Continuous fiber-reinforced thermoplastics (CFRTPs) provide key intrinsic advantages
over thermoset composites, including high impact resistance, short process cycle, unlimited shelf-
life, recyclability, and weldability. However, despite these advantages, melt processing of CFRTP
retains one critical challenge. Because of high melt viscosity and melting temperature, high-
performance CFRTPs must be consolidated at high temperature and high pressure, requiring use
of autoclave or other energy-intensive manufacturing processes. Therefore, extensive use of
CFRTPs in aerospace industry has been restricted by lack of technically mature and robust OoA
manufacturing processes.
To enable efficient OoA manufacturing of CFRTP, we conducted a proof-of-concept case
xv
study to demonstrate the feasibility of thermoplastic prepreg with partially polymerized matrix
(Chapter 4). As opposed to conventional thermoplastic prepreg, which is characterized by fully
polymerized matrix, our model pre-polymer prepreg contained low-viscosity pre-polymer resin.
The pre-polymer resin facilitated part consolidation at low temperature and pressure, even below
the glass transition of the final polymer matrix, while providing provisional tack and drape at room
temperature for improved material handleability. Overall, we proposed a pathway to reduce major
OoA process challenges associated with CFRTPs, including thermal and pressure conditions
required for part consolidation as well as material conformability of CFRTP prepregs.
Overall, this work addresses some of the major process challenges in OoA composites
manufacturing using various approaches. First, effective process guide and optimization tools for
OoA processes were developed applying material characterization results, polymerization process
models, in situ dielectric process diagnostics, and process simulation. Second, the conventional VI
process cycle was modified by introducing an additional post-infusion dwell step to extend the VI
post-filling stage and to improve part quality of multi-component laminate. Third, an efficient
methodology was developed to accurately assess the physical state of resin in situ, increase VI
process efficiency using process map and process simulation, and consequently, reduce liquid resin
material waste. Finally, a new material/product form of thermoplastic prepreg, which contain
partially polymerized matrix, was developed through a proof-of-concept case study to significantly
enhance material processability and conformability.
1
Chapter 1. Introduction
1.1 Motivation
Continuous fiber-reinforced polymers (CFRPs) consist of polymer matrix reinforced with
fibers. The commonly used fiber reinforcement materials include glass, carbon, aramid, and basalt.
The polymer matrix can be either thermoset (e.g., epoxy, bismaleimide, benzoxazine, polyester)
or thermoplastic (e.g., PEEK, PAEK, PEI, polycarbonate). CFRPs provide high specific strength
and stiffness, excellent fatigue properties, corrosion resistance, and thermal and chemical stability
[1–3]. In addition, the reinforcement plies of various architectures can be tailored and oriented in
the primary loading direction for greater design flexibility and additional weight savings [3].

Figure 1-1. Global composites market share by end use, 2019 (%) [4]
CFRPs have been increasingly used as lightweight structural parts in multiple industries,
including automotive, aerospace, defense, energy, and construction (Figure 1-1). The increasing
demand for composites is primarily driven by the automotive and transportation sector, which
strives to enhance fuel economy and minimize carbon emission to comply with the strict
government environmental regulations. The aerospace and defense industry seeks to manufacture
2
lightweight and rigid structural components using high-performance composite materials. In 2020,
the global composites market size was estimated to be over USD 95 billion, and by year 2027, the
market is projected to grow over USD 160 billion, at a compound annual growth rate of 7.6 % [4].

Figure 1-2. Autoclave system manufactured by ASC Process Systems [5]
Most high-performance composite structures for aerospace applications are traditionally
manufactured using an autoclave, which is a heated and pressurized vessel (Figure 1-2). Layers of
prepreg—an intermediate CFRP product form in which fiber reinforcement is pre-impregnated
with polymer matrix—are stacked into a form of laminate onto a mold, sealed in a vacuum bag
assembly, and placed in an autoclave for consolidation [6,7]. Then, vacuum is drawn in the bag,
and the laminate is consolidated under pressure at an elevated temperature. The applied pressure
ensures that the laminate conforms to the shape of the mold and suppresses porosity by facilitating
resin flow into dry areas and collapsing bubbles of entrapped air or volatiles [7].
Conventional autoclave processing is relatively well-established and yields high-quality
laminates [2,7]. However, autoclave incurs significant equipment acquisition and operation costs,
increasing the unit production cost per part. Furthermore, the part design and manufacture are often
constrained by the dimension and availability of the equipment, imposing a relatively inflexible
3
manufacturing environment and slow production rate. To meet the increasing demands, CFRP
manufacturers are thus seeking alternative out-of-autoclave (OoA) manufacturing methods that
can yield robust composite parts in a time- and cost-efficient manner [8].
Therefore, the work presented in this dissertation aims to develop and optimize materials
and processes for more efficient and robust OoA manufacturing of fiber-reinforced polymer
composites. Using various material characterization techniques, process modeling, in situ process
diagnostics, and process simulation, we address the major process challenges encountered in OoA
manufacturing to improve CFRP part quality, performance, and processability.
1.2 CFRP Manufacturing Methods
CFRP composites can be manufactured using diverse processing methods, but the majority
of high-performance CFRP parts are typically produced using either liquid composite molding
(LCM) or prepreg processing [9]. To determine the optimal manufacturing method, various process
factors should be considered carefully, including part design and processing requirements,
production speed and volume, and manufacturing flexibility. This work will address the major
process challenges encountered in both prepreg processing and LCM process, and a brief overview
of the two processes will be explained.
1.2.1 Liquid Composite Molding (LCM)
In LCM process, a dry fiber preform is placed in a mold, and liquid resin is injected into
the mold, impregnating the fiber preform. Once the resin completely saturates the fiber preform,
the resin is then heated and cured under pressure to produce a consolidated laminate. Because the
liquid resin must travel a long distance from injection pot to fully impregnate the preform, the resin
4
viscosity must remain low throughout the injection process to facilitate flow. Hence, LCM resin
cannot tolerate high additive or toughener content, exhibiting reduced mechanical performance
(compared to parts produced from prepregs). Depending on the type of mold used and the source
of pressure driving the resin flow, the LCM process can largely be divided into resin transfer
molding (RTM) and vacuum infusion (VI or V ARTM, vacuum-assisted RTM).

Figure 1-3. Resin Transfer Molding (RTM) setup & process cycle [10]
RTM uses a two-sided rigid mold, and resin is injected under positive pressure. A
representative RTM process setup and cycle are shown in Figure 1-3. The fiber reinforcement plies
pre-coated with binder material are first preformed into the desired geometry using a heated press,
and the stabilized preform is inserted into the mold. Once the mold is closed, the resin is injected
under pressure and cured at elevated temperature. Finally, the produced laminate can be demolded.
The major advantages of RTM process include fast process cycle, robust and consistent part quality,
smooth surface finish, and high degree of process automation. However, the RTM process requires
expensive equipment and tooling for the double-sided rigid mold, and thus can only be used to
manufacture small- to medium-sized parts.
5
In contrast, in vacuum infusion (VI), the top tool surface is replaced with a flexible
vacuum bag, which is an inexpensive consumable material (Figure 1-4). The resin is infused into
the bag assembly using atmospheric pressure differential and is subsequently cured at elevated
temperature. Using a one-sided rigid mold, VI process enables low-cost manufacture of large and
complex unitized composite structures without the need for secondary bonding (Figure 1-5).
Interest in VI has grown rapidly in the recent years, particularly in the aerospace industry, which
seeks to reduce the manufacturing costs associated with conventional prepreg processing.

Figure 1-4. Vacuum infusion (VI) process setup [11]
However, because only atmospheric pressure is available to drive resin flow from vacuum-
only consolidation, VI requires use of low-viscosity thermoset resin, which is inherently brittle
and susceptible to impact damage. The impact performance of vacuum-infused thermoset
composites can be enhanced by introducing thermoplastic content through bulk resin modification
or interlaminar toughening [12]. Yet, bulk matrix toughening is not favored for LCM processes,
6
because mixing rubber/thermoplastic liquid or particles into thermoset resin can result in
significant increase in resin viscosity. In contrast, interlaminar toughening can improve impact
properties without elevating the resin viscosity.

Figure 1-5. Vacuum infusion process used to manufacture a yacht hull [13]
In addition, the limited pressure available during infusion also causes VI to endure slower
infusion rate and longer fill time compared to RTM. Inserting a high-permeability flow distribution
medium into the vacuum bag assembly can enhance resin flow and reduce fill time. However, the
insertion of flow media can also induce unacceptable void contents and may not be applicable to
parts with complex geometries [14,15]. To mitigate this issue, resin is often heated during filling
to reduce viscosity and facilitate flow, but only at the cost of increasing resin degree of cure and
viscosity with flow time and distance [14,16]. Thus, the process parameters for VI heated filling
must be selected carefully to prevent premature resin gelation and dry spot formation. Even with
flow media, the problem of evolving resin state can persist, particularly as part size increases.
7
1.2.2 Prepreg Processing
Prepreg is an intermediate product form of CFRP in which continuous fiber reinforcement
is pre-impregnated with polymer matrix. Prepregs are designed to minimize the required resin flow
distance for complete impregnation of fiber reinforcement with polymer resin [17]. Because the
resin needs to travel only a short distance throughout the process cycle, prepreg resin can contain
high additive or toughener content, which can lead to enhanced part mechanical properties. Other
major advantages of prepreg processing include high degrees of process control, relatively
consistent part quality, reduced material waste, and improved material handleability (exhibiting
tack with B-staged thermoset prepreg or pre-heated thermoplastic prepreg vs. dry fiber plies).

Figure 1-6. Conventional thermoset and thermoplastic prepregs [18,19]
Depending on the type of polymer matrix used, prepreg can largely be divided into
thermoset prepreg and thermoplastic prepreg (Figure 1-6). Once fully cured and crosslinked,
thermosets cannot be melted and re-processed upon heating for further fiber impregnation and
consolidation. Hence, conventional thermoset prepregs contain partially cured (B-staged) resin to
ensure that the resin stays in place throughout the material handling process and provides sufficient
tack for laminate lay-up. Compared to thermoplastics, thermosets feature greater material strength
8
and resistance to high temperature. However, thermoset prepreg contains pre-catalyzed resin,
which can age quickly at ambient condition. Therefore, the material must be stored in a freezer to
prevent excessive resin aging; the manufacturer-recommended shelf-life and out-life of high-
performance thermoset prepreg generally span 6-12 months and 10-30 days respectively [20,21].
In contrast, thermoplastics do not form crosslinks upon polymerization, and polymer
chains associate by relatively weak intermolecular forces, allowing them to be melted and re-
processed upon heating. Thus, conventional thermoplastic prepregs contain fully polymerized
polymer, which does not age during room temperature storage. As a result, continuous fiber-
reinforced thermoplastics (CFRTPs) feature unlimited shelf-life, short process cycle (because part
consolidation involves simple physical changes to the matrix, with no time-consuming chemical
reactions), recyclability, weldability, and high impact resistance. However, melt processing of
CFRTP retains one critical challenge, especially for high-performance thermoplastics, which
exhibit high melt viscosity and melting temperature. To ensure proper fiber bed saturation and void
removal, CFRTPs must be processed at high temperature and pressure, requiring use of energy-
intensive and costly manufacturing processes.

Figure 1-7. Automated tape laying machine [22]
9
First, prepregs are cut into a desired dimension, and plies of cut prepregs are stacked into
a laminate on a one-sided tool. Having B-staged resin matrix, thermoset prepregs exhibit tack and
drape at ambient condition, enabling a simple room temperature lay-up. In contrast, thermoplastic
prepregs contain fully polymerized matrix and thus are dry, tack-less, and stiff at room temperature.
Therefore, CFRTP prepregs must be heated (e.g., pre-heating or spot welding) to high temperature
to exhibit sufficient tack and drape for practical lay-up. The relatively narrow prepreg tapes can be
laid up using automated tape layup (ATL, 12 inches or wider tapes, Figure 1-7) or automated fiber
placement (AFP, 4-6 inches of tapes) processes. ATL process is faster and is suitable for
manufacturing parts with flat or gently curved geometries, while AFP process can be used to
produce parts of much more complex geometries at the expense of longer lay-up time.

Figure 1-8. Schematic of a vacuum bag assembly [23]
Once the lay-up process is complete, the setup is vacuum bagged with other essential
consumable materials such as release film, breather, silicon tape, and vacuum bag (Figure 1-8).
Then, the vacuum bag assembly is placed in an autoclave or oven, and vacuum is applied within
the bag. Finally, the laminate is consolidated under heat and pressure to facilitate fiber bed
10
saturation and void removal. Apart from autoclave and vacuum bag-only (VBO) prepreg processes
that are generally used for manufacture of high-performance composites, prepregs can also be
consolidated using other more time- and cost-efficient processes such as thermoforming.
1.3 Approach
In this dissertation, we apply material characterization, in situ process diagnostics, process
modeling, and process simulation to develop and optimize materials and processes for efficient
out-of-autoclave (OoA) composite manufacturing. Material characterization technique includes
thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), and rheology. To
enable in situ monitoring of resin physical state during actual part manufacture, dielectric analysis
(DEA) is employed. Then, polymerization kinetics, rheology, and dielectric data are modeled to
predict evolution of process metrics and to provide effective process guide tools for process
optimization. Finally, a finite element analysis (FEA) simulation tool is used to validate and refine
the process parameters determined from material characterization results and process models.
1.3.1 Material Characterization

Figure 1-9. TGA instrument and sample TGA thermal stability curve [24]
11
In thermogravimetric analysis (TGA, Figure 1-9), change in sample mass is measured over
time as temperature changes. TGA can be used to detect physical (e.g., phase transition, absorption,
and adsorption) and chemical (e.g., chemisorption and thermal decomposition) phenomena. For
analysis of composites, TGA is often used to track volatile content or thermal degradation
temperature. During testing, composite or polymer samples are often purged with nitrogen (or
other inert gases) to ensure accurate measurement and to prevent sample oxidation. The analysis
results can be used to determine the operating temperatures of a material or the temperature ranges
for further material characterization tests (e.g., DSC or rheology).

Figure 1-10. DSC instrument and sample DSC thermoset cure curve [25]
Differential scanning calorimetry (DSC, Figure 1-10) measures the difference in the
amount of heat required to change the temperature of a sample and a reference (empty test pan) as
a function of temperature. In standard DSC, temperature changes linearly as a function of time,
and heat capacity of the reference must be predefined. DSC can be used to detect phase transition
and chemical reaction. If more heat is required to flow to the sample than the reference to maintain
both at the same temperature, then the process is endothermic (e.g., melting and evaporation). If
less heat is required for the same, then the process is exothermic (e.g., crystallization and epoxy
12
cure reaction). DSC can also be employed to detect glass transition phenomena, which appear as
a step in the baseline as sample experiences a change in heat capacity.
Modulated DSC (MDSC) uses two simultaneous heating rates—one linear and one
sinusoidal—to measure the total heat flow as well as its heat capacity component. The kinetic
component of the total heat flow can be obtained by subtracting the two. MDSC technique yields
reversing heat flow, which corresponds to heat capacity component (e.g., heat capacity, glass
transition, and melting), and non-reversing heat flow, which represents kinetic component (e.g.,
crystallization, enthalpy recovery, and cure reaction). Overall, MDSC enables separation of
complex and overlapping transitions, with increased sensitivity and resolution.

Figure 1-11. Rheometer and sample epoxy cure rheology curve [26,27]
Rheology is used to study flow and deformation of materials. In dynamic rheology testing,
an oscillatory shear stress (or strain) is applied by rheometer (Figure 1-11), and the material
response is measured in strain (or stress). The phase angle (𝛿 ) between the sinusoidal stress and
strain is also measured. For a purely elastic material, the oscillatory stress and strain will be
perfectly in phase (𝛿 = 0°), while for a purely viscous material, the strain (or stress) will lag stress
13
(or strain) by 90 degrees (𝛿 = 90°). Most polymers are viscoelastic, exhibiting both elastic and
viscous characteristics under deformation ( 0°< 𝛿 < 90°). The rheology data can be used to
characterize polymer resin viscosity evolution or to identify resin gel point.
1.3.2 In Situ Process Diagnostics
Dielectric analysis (DEA) or dielectric cure monitoring (DCM) system (Figure 1-12)
applies sinusoidal excitation voltage to a sample material (e.g., curing polymer resin) and measures
the resulting sinusoidal current and phase shift responses. Once the excitation voltage is applied,
dipoles within the sample align with the applied electric field, and charge carriers move toward
electrodes of opposite charges. DEA tracks dielectric properties such as ionic conductivity (σ) and
complex permittivity (ε*), which arise from ionic current and dipole rotation [28].

Figure 1-12. Dielectric cure monitoring (DCM) system with both reusable & disposable
interdigitated dielectric sensors [29]
The complex permittivity or ε* consists of real and imaginary components. The relative
dielectric constant (ε’) is a measure of alignment or orientation and number of dipolar groups in a
material. On the other hand, the loss factor (ε” = ε”ion + ε”dipole) is a measure of total energy lost
from aligning dipoles and moving ions in a material. As the state of resin cure or polymerization
14
progresses and mechanical viscosity increases, the degrees of both ion mobility and dipole rotation
subside, and the loss factor ε” decreases. As a result, with increasing degree of cure and mechanical
viscosity, ionic conductivity (σ) decreases, and ionic viscosity, which is a reciprocal of σ, increases.
Conventional parallel plate electrodes for dielectric measurements imposes a major
challenge for in situ measurements. In such bulk measurement, inaccurate results can occur
because of constantly changing distance between the parallel electrodes, arising from changing
process pressure or expansion/contraction of a sample material during actual composite processing.
To address this issue, DEA system instead employs two planar comb electrodes, which are
interdigitated on a polyimide film substrate (Figure 1-13). The interdigitated dielectric sensor
creates a fringing field, in which electrical field does not penetrate the bulk material. The depth of
the electrical field penetration can easily be controlled by changing the spacing between the two
electrodes, enabling surface measurement that is largely independent of process conditions.

Figure 1-13. Interdigitated dielectric sensor on polyimide substrate [30]
Ex situ material characterization techniques explained in Section 1.3.1 (TGA, DSC, and
rheology) can only be performed on pure polymer resins (as opposed to composites) and cannot
be incorporated into an actual manufacturing process setup. In contrast, using DEA, we can easily
insert a non-invasive dielectric sensor into a process setup to monitor the physical state of polymer
15
resin in situ during actual part manufacture in a straightforward, dynamic, and repeatable manner.
DEA technique can be applied to various materials (e.g., epoxies, acrylics, polyimides, polyamides)
and processes (e.g., autoclave processing, vacuum bagging, thermoforming, injection molding) for
formulation and kinetics study or online process monitoring applications. In this work, we correlate
thermal and rheological analysis results (and corresponding models) of polymerization reaction to
dielectric measurement data to develop effective process guide tools such as resin cure maps.
1.3.3 Process Modeling
In this work, we use chemical kinetics, rheology, and dielectric models—both semi-
empirical and theoretical—to predict thermoset cure and thermoplastic polymerization behaviors.
Such models enable accurate prediction of polymerization process metrics (e.g., degree of cure or
degree of monomer conversion, mechanical viscosity, and ionic viscosity) under different process
conditions. The models can also be used to correlate various process parameters to build effective
online process guide tools for process design or optimization (Figure 1-14).

Figure 1-14. Ex situ material characterization & in situ process monitoring for process modeling
For thermoset cure kinetics analysis, we apply a phenomenological model developed by
Kratz et al. [31], who combined (1) Cole model [32] to account for transitioning from kinetics-
16
controlled to diffusion-controlled cure reaction as degree of cure (α) exceeds αgel (α at resin
gelation) and (2) Kamal and Sourour model [33] to reflect a faster cure rate existent in experimental
data (vs. Cole model) at low degree of cure (α < 0.1):
d𝛼 d𝑡 = 𝐾 1
𝛼 𝑚 1
( 1 − 𝛼 )
𝑛 1
+
𝐾 2
𝛼 𝑚 2
( 1 − 𝛼 )
𝑛 2
1 + exp (𝐷 ( 𝛼 − ( 𝛼 𝐶 0
+ 𝛼 𝐶𝑇
𝑇 ) ) )

(1-1)
𝐾 𝑖 = 𝐴 𝑖 ∙ exp (−
𝐸 𝐴𝑖
𝑅𝑇
), where 𝑖 = 1, 2 (1-2)
where Ki is the Arrhenius temperature dependent term, m1, m2, n1, and n2 are the reaction order-
based fitting constants, D is the diffusion constant, αC0 is the critical α at absolute zero, αCT
accounts for the increase in critical α with temperature, Ai is the pre-exponential factor, EAi is the
activation energy, R is the universal gas constant, and T is the temperature.
For the mechanical viscosity model, we use a model developed by Khoun et al. [34] to fit
the rheology data and predict viscosity evolution during thermoset cure:
𝜂 = 𝜂 1
+ 𝜂 2
(
𝛼 𝑔𝑒𝑙 𝛼 𝑔𝑒𝑙 − 𝛼 )
𝐴 +𝐵𝛼 +𝐶 𝛼 2

(1-3)
𝜂 𝑖 = 𝐴 𝜂𝑖
∙ exp (−
𝐸 𝜂𝑖
𝑅𝑇
), where 𝑖 = 1, 2
(1-4)
where η is the viscosity, η1 and η2 are the Arrhenius dependent viscosity component, αgel is the
degree of cure at gelation, A, B, and C are the fitting constants, Aηi is the Arrhenius constant, and
Eηi is the viscosity activation energy.
Finally, resin ion viscosity can be expressed using the following theoretical model [28]:
17
log( 𝐼𝑉 ) = log (
𝑘 𝑞 2
𝑛 𝐷 0
) + log( 𝑇 )+
𝑄 𝑘𝑇 ln( 10)
(1-5)
where k is Boltzmann’s constant, q is the magnitude of electronic charge, n is the free ion
concentration, D0 is the maximum value of diffusion coefficient, and Q is the activation energy for
diffusion. In Equation 1-5, the parameters on the right side (except T itself) are independent of
temperature, and the equation can be re-written as:
log( 𝐼𝑉 ) = 𝐴 + log( 𝑇 )+
𝐵 𝑇
(1-6)
where both model coefficients A and B depend on α, because D0 and Q can vary with degree of
cure. At a given α, both model parameters A and B will remain constant. Between the two
temperature terms, log(T) and 1/T, the latter term dominates the temperature dependence of ion
viscosity, and thus ion viscosity decreases as temperature increases for a given α.
1.3.4 Process Simulation

Figure 1-15. Example PAM-RTM simulation results for wind blade infusion [35]
18
For the VI projects, we used a finite element analysis (FEA) resin infusion and cure
simulation software, PAM-RTM (ESI Group), to validate and refine the model-predicted process
maps and process metrics. For accurate results, PAM-RTM requires basic material properties or
models as inputs, including preform permeability as well as resin cure kinetics and viscosity
profiles. In addition, process conditions (e.g., process temperature and pressure profiles, infusion
pressure or rate, resin inlet and outlet locations, flow media insertion) must be determined based
on the type of process. Then, the simulation yields prediction results for filling time, dry spot
formation, flow velocity, local pressure variation, and temperature and degree of cure profiles. The
simulation results can be used to guide efficient selection of process parameters or strategies and
to enable effective process adjustments.
1.4 Objectives & Scope
This dissertation addresses the major manufacturing challenges commonly encountered in
(1) thermoset VI process (Chapters 2 and 3) and (2) thermoplastic prepreg processing (Chapter 4).
Because only limited atmospheric pressure is available to drive resin flow, VI process requires use
of low-viscosity, brittle thermoset resin and endures slow infusion rate. On the other hand,
thermoplastics exhibit high melt viscosity and melting temperature, and thus CFRTP prepregs
require application of high temperature and pressure for part consolidation, leading to energy-
intensive and costly manufacturing. In this work, we propose potential solutions for the identified
composites manufacturing concerns applying the approaches described in the earlier section:
material characterizations, in situ process diagnostics, process modeling, and process simulation.
In Chapter 2, we inserted non-woven thermoplastic veils at interlaminar regions of
thermoset composites to improve impact performance. However, permeability characterization
19
and infusion process simulation results demonstrated that the degree of permeability difference
between the fiber reinforcement and veil toughener is sufficiently large to induce non-uniform
flow fronts during infusion and high porosity in the final part. To address this issue, we introduced
an additional low-temperature post-infusion (LTPI) dwell to the conventional epoxy cure cycle to
extend the VI post-filling stage. Cure kinetics and viscosity model guided selection of LTPI dwell
process parameters, while in situ dielectric process diagnostics and resin cure map enabled real-
time process adjustments. Porosity analysis and impact test results showed that the added LTPI
dwell promoted resin redistribution and saturation of partially filled regions, resulting in much
reduced porosity and greater peak load during impact testing.
Chapter 3 focuses on the problem of slow infusion rate and resin waste in VI process. VI
resin is often heated during filling to promote infusion flow, but resin can age rapidly with flow
time and distance at elevated temperature, resulting in premature resin gelation. Also, use of
expired or aged resin can result in significantly reduced infusion window, requiring more careful
process design. For more efficient use of aged resin in VI process, we developed a resin cure map
by correlating cure models to DEA data to enable simple in situ measurement of VI resin age. An
age-adjusted infusion process map yielded the nominal infusion process window and key process
metrics. We conducted infusion simulation using PAM-RTM to validate and refine the process map
and the associated process metrics. The results demonstrated that the VI process parameters must
be adjusted for part size, geometry, and resin age. Finally, a flow contour map was constructed
through a parametric simulation study to guide effective selection of VI process parameters.
Chapter 4 addresses the main processing challenge of thermoplastic prepreg processing:
high matrix melt viscosity and melting temperature, requiring application of high temperature and
pressure. To improve processability and conformability of CFRTP prepreg, we investigated the
20
feasibility of developing a thermoplastic prepreg with partially polymerized matrix. To determine
the extent of polymerization for initial trials of model prepreg fabrication, the polymerization
kinetics and viscosity evolution of poly(methyl methacrylate) (PMMA) resin was characterized.
The aging study on PMMA pre-polymer resin was conducted to determine material stability under
different temperatures and to show that refrigeration can delay pre-polymer out-time accrual. The
manufactured prepregs were tested for room temperature tack and drapability to prove enhanced
material handleability. Overall, the low-viscosity pre-polymer matrix allowed part consolidation
at significantly reduced temperature and pressure and yielded near-zero porosity laminates.
The last chapter of this dissertation (Chapter 5) summarizes the key conclusions and
technical implications of each project and emphasizes the significance of our findings. We also
explain some of the limitations of the proposed solutions and suggest possible future work to
address the identified concerns.
21
Chapter 2. Effects of Post-Infusion Dwell on Vacuum Infusion of Thermoset
Composites Toughened by Thermoplastic Interlaminar Veils
2.1 Abstract
The effects of post-infusion dwell on vacuum infusion of thermoset composites toughened
by non-woven thermoplastic interlaminar veils were investigated. Permeability measurements and
simulation of the resin infusion process demonstrated that the toughening interlayers can
effectively act as interlaminar flow distribution media. Local variations in permeability induced
non-uniform flow fronts, resulting in high porosity. However, introduction of a low-temperature
post-infusion dwell allowed more time for the resin to equilibrate pressure and redistribute during
the post-filling stage, achieving full saturation of dry regions. The process parameters of the post-
infusion dwell were determined using cure kinetics and viscosity models, while in situ process
adjustments were implemented using dielectric cure monitoring system in conjunction with resin
cure maps. Laminates fabricated with the modified cycle exhibited reduced porosity and greater
peak load during impact testing. This work highlights potential advantages of the post-infusion
dwell, which can similarly be applied to other vacuum infusion processes requiring a protracted
post-filling stage.
2.2 Introduction
In this work, we investigate the effects of post-infusion dwell on part quality for vacuum
infusion (VI) of thermoset composites toughened by thermoplastic interlaminar veils. Previous
studies have demonstrated that insertion of non-woven thermoplastic veils increases impact
strength of thermoset composites [12,36–41], yet can also cause increased porosity because of
22
local variations in preform permeability created by the highly permeable veils [36,42]. One of
these reports speculated that modifying the manufacturing process to provide more time for resin
to saturate dry regions may reduce laminate void content [36], although no such studies have been
reported. To promote impregnation of partially saturated interlaminar regions, a low-temperature
post-infusion dwell was introduced to allow greater time for low-viscosity resin to equilibrate
pressure and redistribute during the post-filling stage. The process parameters of the post-infusion
dwell were determined using resin cure kinetics and viscosity model predictions. Dielectric cure
monitoring (DCM) system permitted in situ measurement of resin ion viscosity, while resin cure
maps and ion viscosity model were employed to convert the collected ion viscosity data into useful
cure process metrics and to adjust the cure cycle accordingly.
Conventional autoclave cure of prepreg composites is robust, yet involves high capital and
operating costs, driving the exploration of faster and more cost-efficient out-of-autoclave
processes [7,43]. A promising alternative to autoclave processing is liquid composite molding
(LCM), which involves lay-up of a dry fiber preform in a mold, infusion of liquid resin through
the preform, and resin cure [15,44–46]. Based on tool design, LCM processes are generally divided
into two—resin transfer molding (RTM) and VI. The RTM process features a two-part rigid mold
for top and bottom tool surfaces, and resin is injected under positive pressure (greater than ambient
pressure); in contrast, in the VI process, the top tool surface is replaced by a flexible vacuum bag,
and resin is infused using atmospheric pressure differential [15,45,47–49].
Using a one-sided rigid mold, the VI process can be employed to produce large, complex,
and unitized (without secondary bonding) parts at relatively low cost [15,49–52]. However,
because only atmospheric pressure is applied from vacuum-only consolidation, VI requires use of
low-viscosity thermoset resins, which in general are brittle and thus susceptible to impact damage
23
[12,42,52,53]. Introducing thermoplastic content into thermoset composites through bulk resin
modification or interlaminar toughening can mitigate the inherently low toughness of thermosets
[12,37]. Bulk matrix toughening generally is achieved by mixing rubber/thermoplastic liquid or
particles into the thermoset resin. Yet this approach is often incompatible with liquid molding
because of the associated increase in resin viscosity [12,36–38]. In contrast, interlaminar
toughening, which includes co-mingled fibers, thermoplastic films, and non-woven fiber veil
additions, can potentially impart toughening to weak interlaminar regions of thermoset composites
while simultaneously mitigating the processing concerns arising from elevated resin viscosity for
LCM manufacturing [12,37]. Note that the two methods (bulk resin modification and interlaminar
toughening) are not necessarily mutually exclusive, but potentially can be used together.
Inserting non-woven thermoplastic veils at interlaminar regions is a cost-effective
toughening method that can be applied to any composite manufacturing processes, but is relatively
less technologically mature than alternative approaches [12,36,37]. In hand lay-up of dry preforms,
the veils can be inserted between reinforcement plies. For automated tape laying (ATL) or
automated fiber placement (AFP) processes, the automated lay-up of preforms must be paused
intermittently for veil insertion between layers. However, the addition of thermoplastic veils can
be automated potentially, because the veils can be produced in tape form and heated to promote
tack. When inserted at interlaminar regions, the veils embed into adjacent fabric plies during
infusion and consolidation stages, resulting in a fiber bridging effect during interlaminar fracture
[12,39]. This interleaving mechanism can effectively resist delamination and crack propagation by
arresting damage growth and localizing damage [12,36]. Therefore, adding thermoplastic veils of
low areal weight to thermoset composites can enhance resistance to impact damage, while
increasing Modes I & II interlaminar fracture toughness and compression-after-impact (CAI)
24
strength [36,37,39–41]. Exhibiting greater permeability than reinforcement fabrics, the veils also
can serve as interlaminar flow distribution media, through which resin can flow more rapidly
during the infusion process [36,42]. However, the local variations in preform permeability can
induce non-uniform flow fronts and high porosity [36,42].
In the VI process, local compaction pressure exerted on the preform varies as the resin
flow front progresses during the infusion (from vacuum to atmospheric pressure), and the top
flexible vacuum bag surface allows preform thickness to adjust accordingly [45,54]. The
continuously evolving preform deformation and pressure gradient result in a unique processing
stage in VI, called “post-filling” stage, which refers to the period between the completion of
infusion and resin gelation [44,54,55]. Once infusion is complete and the resin inlet is closed, the
resin pressure field within the part is initially not uniform, and resin continues to flow and
redistribute to reach an equilibrium state, impregnating partially saturated regions [44,54,55].
Therefore, insufficient post-filling time in VI can lead to incomplete saturation of dry areas,
leading to greater void content in cured laminates toughened by non-woven thermoplastic veils.
Typically, the manufacturer-recommended cure cycle (MRCC) of high-performance
infusion-grade thermoset resin is comprised of a single isothermal dwell [56], during which the
resin degree of cure and viscosity rise sharply after a brief reaction induction period [44,56]. For
the epoxy resin used in this study (Hexcel HexFlow
®
RTM6), the MRCC features a single dwell
for 120 minutes at 180 °C, and the resin gels fast within 20 minutes at the prescribed temperature
[56]. However, such a brief post-filling period may not provide sufficient time to complete
saturation of partially filled regions, especially for preforms containing multiple plies of
thermoplastic veils.
The addition of interlaminar thermoplastic veils enhances impact resistance of thermoset
25
composites, but the relatively high permeability of the veils also introduces disparities in flow rates
in the layers and thus can result in unacceptable levels of porosity. To address this issue, we
modified the MRCC for the resin (RTM6) by adding a low-temperature post-infusion dwell to
provide additional time for the resin to fully saturate dry interlaminar regions before gelation. The
process parameters of the post-infusion dwell were initially determined using resin cure kinetics
and viscosity models. Resin cure maps (degree of cure and mechanical viscosity isolines) provided
correlations between the isoline metrics and ion viscosity, and were employed to identify the gel
point during the cure cycle. A DCM system was used to monitor the cure state in situ, to adjust the
cure cycle, and to validate and refine the cure maps accordingly.
Three different VI process cases were examined, demonstrating that the post-infusion
dwell step effectively achieved full saturation and void removal by extending the post-filling stage.
Overall, this work describes a potential pathway to address the problem of high porosity commonly
encountered in VI processes. The pathway includes a low-temperature post-infusion dwell
informed by material characterization, the use of resin cure maps, and in situ process diagnostics.
The post-infusion dwell can also be extended to other VI process variants that require a protracted
post-filling period for full impregnation of dry regions, including those that feature multi-
component or complex-shaped preforms.
2.3 Experimental
2.3.1 Materials
An aerospace-grade epoxy resin designed for infusion processes was selected and acquired
(HexFlow
®
RTM6, Hexcel [56]). The resin was stored under refrigeration at -18 °C to prevent
undesired curing of the material before use. For the reinforcement, a plain weave carbon fiber
26
fabric with an areal weight of 193 g/m
2
and 3000 fiber/tow count was used (part #1530, Fibre Glast
[57]), while for the thermoplastic interlayer, a non-woven PBN-II
®
polyamide (nylon 66) veil with
an areal weight of 34 g/m
2
was supplied by Cerex Advanced Fabrics [58]. Before part lay-up, the
polyamide veils were dried in a vacuum oven at 90 °C for two hours at 0.1 MPa (absolute pressure)
to remove any absorbed moisture during storage.
2.3.2 Cure Kinetics & Rheology Characterization
2.3.2.1 Modulated Differential Scanning Calorimetry (MDSC)
The cure kinetics of the epoxy resin was characterized using DSC (TA Instruments, Q2000)
under nitrogen purge (50 cm
3
/min). Resin samples (8-10 mg) were sealed in aluminum hermetic
pans with lids, and exposed to a dynamic ramp at four heating rates (5, 10, 15, and 20 °C/min) to
determine the total heat of the cure reaction. The temperature range of the dynamic runs spanned
from -50 to 340 °C. Isothermal dwell measurements were conducted at eight temperatures: from
80 to 120 °C in 10 °C increments to monitor advances in degree of cure under filling conditions,
and from 120 to 180 °C in 20 °C increments to examine and model resin cure kinetics. Following
the dwell, the samples were heated to 300 °C at 5 °C/min to measure the residual heat of cure.
However, during the temperature ramp to 300 °C, the samples exhibited an endothermic annealing
peak near the residual cure exotherm due to physical aging of the resin [59]. To eliminate this
annealing peak, the specimens were quenched rapidly at the endotherm spike to -50 °C and then
reheated during re-scan (Figure 2-1), using the method developed by Karkanas et al [59]. For the
ramping step, a sinusoidal temperature modulation of ±0.5 °C/min was applied to distinguish
reversing and non-reversing heat flow signals. Reversing heat flow depends on heat capacity and
the rate of temperature change, while non-reversing heat flow is associated with kinetic
27
components of the reaction [60,61]. Glass transition can be examined from the reversing heat flow
signal because it is a heating rate dependent transition, while the curing reaction is a temperature
dependent transition and thus appears in the non-reversing signal [60–62].

Figure 2-1. Heat flow and reversing heat flow profiles during the residual cure of the epoxy resin
after isothermal dwell at 140 °C before and after annealing peak removal
2.3.2.2 Rheometry
Viscosity evolution during epoxy resin cure was characterized using a rheometer (TA
28
Instruments, AR2000ex). The resin samples were exposed to an isothermal dwell at eight
temperatures, analogous to MDSC isothermal dwell measurements. Two different rheometer
geometry fixtures were applied for different dwell temperature ranges. A cone-and-plate geometry
fixture (cone angle 2 °, cone diameter 40 mm, and truncation 53 µm), which provides uniform
shear rate and high rheological accuracy, was equipped with a Peltier plate to accurately measure
low initial resin viscosity (below 0.1 Pa ⋅s) during the filling condition rheology tests (80-120 °C).
For high temperature cure condition tests (120-180 °C), disposable parallel-plate geometry fixture
(diameter 25 mm) and environmental test chamber setup were used instead to measure the wide
range of evolving resin viscosity throughout the entire cure span. For the rheology tests, cell
temperature was increased from the pre-heating temperature (either 80 or 100 °C) to the dwell
temperature at approximately 30 °C/min (the maximum ramp rate of the rheometer) and held
constant for either 3 hours (for filling condition tests to prevent resin full cure) or prescribed
durations (for curing condition tests), under oscillatory shear at 1 Hz frequency and 5% strain
(within the linear viscoelastic region).
2.3.2.3 Dielectric Analysis (DEA)
DEA measurements were conducted using a dielectric cure monitoring system (Netzsch,
DEA 288 Epsilon) in frequency intervals from 1 Hz to 1 kHz at four cure temperatures (120, 140,
160, and 180 °C). A dielectric sensor (Netzsch, Mini-IDEX 100/35) with 33 mm
2
sensing area and
100 μm electrode spacing, along with thermocouple, was mounted on the rheometer Peltier plate,
which served as the heated tool. The resin sample was placed on the dielectric sensor, and cell
temperature was rapidly increased from 80 °C to the dwell temperature at ~ 30 °C/min and held
constant until the resin fully cured. Once a sinusoidal excitation voltage is applied to the resin
29
sample, dipoles align with the applied electric field and charge carriers move toward electrodes of
opposite charge, resulting in sinusoidal current and phase shift responses, which can be converted
into useful dielectric properties such as complex permittivity (ε*) and ionic conductivity (σ)
[30,63,64]. The reciprocal of σ is called ion viscosity and strongly correlates to mechanical
dynamic viscosity (η) [30,65,66].
2.3.3 Permeability Measurement & Infusion Simulation
The permeabilities of plain weave carbon fiber fabric and non-woven polyamide veil were
assessed using unsaturated radial flow test. The isotropy of the polyamide veil was unknown prior
to testing, and the radial test configuration permitted simultaneous measurement of the two
principal permeability values of an anisotropic material [67,68]. A vacuum-driven constant
injection pressure experimental setup (Figure 2) was used, as demonstrated by Pierce et al [69,70].
A single-ply sample was placed between a bottom tool plate and an upper polycarbonate caul plate
(measured cavity thickness of 1.50 ± 0.05 mm), while breather cloth surrounded the sample
periphery to establish an even pressure gradient and to maintain cavity thickness. Instead of epoxy
resin, olive oil (Kirkland Signature) was infused as a facsimile resin to allow permeability
measurement at ambient temperature, as the viscosity of olive oil at room temperature (measured,
0.07 Pa ⋅s) was consistent with that of the pre-heated RTM6 resin (< 0.10 Pa ⋅s). The test fluid was
infused into the center of each sample through 6 mm ID inlet tubing, and the resulting radial flow
pattern was recorded using a digital video camera. Through visual image processing, flow front
position against flow time data were obtained, which were used to calculate material permeabilities.
For each material, the permeability measurements were conducted three times.
30

Figure 2-2. Experimental setup of vacuum-driven constant injection pressure radial flow
permeability measurement
With the measured permeability values, resin infusion through a multilayer laminate (355
mm × 255 mm) comprising 16 plies of plain weave carbon fiber fabric and 7 plies of polyamide
veil was simulated using a commercial finite element analysis software (PAM-RTM, ESI Group).
For the preform stacking sequence, one layer of veil was inserted after every two layers of carbon
fiber fabric. The infusion simulation was performed to demonstrate that the permeability of
thermoplastic interlayers was greater than that of carbon fiber fabric, such that the veils effectively
functioned as interlaminar flow distribution media.
2.3.4 Part Manufacture & Quality Analysis
2.3.4.1 Laminate Manufacture
Carbon fiber-epoxy laminates (355 mm × 255 mm) were produced by VI. A reference
sample (Case A), consisting of 16 plies of carbon fiber fabric with no veils, was fabricated first,
31
followed by toughened samples (Case B and C), which included the same number of carbon fiber
reinforcement plies with an additional 7 plies of polyamide veils. For Case B and C, one ply of
thermoplastic interlayer was inserted after every two plies of carbon fiber fabric. Although the
resin sample was already degassed, the resin was still degassed once more prior to infusion for 30
minutes to remove air entrapped during resin transfer. Following the manufacturer process
specifications [56], the degassed resin and the mold were pre-heated to 80 and 120 °C respectively.
The cure cycle was initiated once infusion completed: Case A and B were cured using the MRCC
(120 minutes at 180 °C), while Case C was cured using the modified cure cycle (170 min at 129 °C,
followed by 60 min at 180 °C). During Case C sample manufacture, a dielectric sensor (Netzsch,
Mini-IDEX 100/35) was embedded inside the vacuum bag to monitor ion viscosity throughout the
cure process. The manufactured laminates were cut into test specimens using an abrasive waterjet
cutting system.
2.3.4.2 Porosity Analysis & Impact Testing
For microstructural porosity analysis, laminates were sectioned, ground, and polished.
Images of each cross-section were recorded using a digital stereo microscope (VHX-5000,
Keyence). Impact resistance of the manufactured laminates (101 mm × 152 mm) was assessed
using a drop tower impact device (9250HV , Instron) in accordance with ASTM D7136 [71]. Four
samples for each case were tested at 30 J impact energy level, with a 7.8 kg impactor, a 16 mm
diameter hemispherical striker tip, and a pneumatic rebound brake system. The force-time profile
was recorded during the impact, and the damage area from impact was measured using a non-
destructive testing (NDT) ultrasound system with a 20 MHz transducer (Mistras NDT) and ImageJ
software.
32
2.4 Results & Discussion
2.4.1 Cure Kinetics & Rheology Characterization
2.4.1.1 Cure Kinetics

Figure 2-3. MDSC data showing heat flow profiles measured during (a) dynamic ramp tests of
RTM6 resin, (b) isothermal dwell tests at filling temperatures (80-120 °C), (c) isothermal dwell
tests at curing temperatures (120-180 °C), and (d) cure kinetics model fitting results, expressed
in terms of resin degree of cure vs. time (experimental: solid line, model-prediction: dotted line)
Figure 2-3a shows MDSC heat flow data measured during the dynamic ramp tests. The
total heat of cure reaction (Hr) was independent of the applied heating rate, and was determined to
be 429.9 (± 4.6) J/g, similar to previously reported values [59,72]. Heat flow data were also
33
obtained during the isothermal dwell tests performed at filling and curing temperatures. Under the
assumption that the rate of cure reaction is proportional to the rate of heat flow, the heat flow data
can be converted into cure rate data using the following equation [31,34]:
d𝛼 d𝑡 =
1
𝐻 𝑟 (
d𝐻 d𝑡 ) (2-1)
where α is the resin degree of cure, t is time, and H is the heat flow measured from the MDSC.
The resin degree of cure can be calculated by integrating Equation 2-1. At 80 and 90 °C, α
advanced negligibly, while at 110 and 120 °C, α evolved slightly faster, and the resin reached full
cure toward the end of the 8-hour dwell step (Figure 2-3b). Thus, the resin and mold pre-heating
temperatures of 80 and 120 °C were deemed appropriate [56]. As expected, the heat flow curves
evolved much faster at higher curing temperatures (120-180 °C), with much shorter reaction
induction periods (Figure 2-3c). The final degree of cure (αf) of the resin samples reacted at 120,
140, 160, and 180 °C were 0.83, 0.88, 0.93, and 0.97 respectively.
The MDSC data for resin cure were fit to a phenomenological cure kinetics model
developed by Kratz et al. [31], which accounts for transitioning from kinetics-controlled to
diffusion-controlled cure reaction as α rises above αgel (α at resin gelation), and a faster cure rate
present in the low-α region (α < 0.1):
d𝛼 d𝑡 = 𝐾 1
𝛼 𝑚 1
( 1 − 𝛼 )
𝑛 1
+
𝐾 2
𝛼 𝑚 2
( 1 − 𝛼 )
𝑛 2
1 + exp (𝐷 ( 𝛼 − ( 𝛼 𝐶 0
+ 𝛼 𝐶𝑇
𝑇 ) ) )

(2-2)
𝐾 𝑖 = 𝐴 𝑖 ∙ exp (−
𝐸 𝐴𝑖
𝑅𝑇
), where 𝑖 = 1, 2 (2-3)
where Ki is the Arrhenius temperature dependent term, m1, m2, n1, and n2 are the reaction order-
based fitting constants, D is the diffusion constant, αC0 is the critical α at absolute zero, αCT
34
accounts for the increase in critical α with temperature, Ai is the pre-exponential factor, EAi is the
activation energy, R is the universal gas constant, and T is the temperature [K].
Parameter Value Parameter Value
A1 [s
-1
] 1.70 × 10
6
A2 [s
-1
] 1.63 × 10
4

EA1 [J/mol] 8.22 × 10
4
EA2 [J/mol] 5.72 × 10
4

m1 0.27 m2 1.15
n1 10.65 n2 1.22
D 43.34 αC0 -0.11
αCT [K
-1
] 2.23 × 10
-3

Table 2-1. Values of the cure kinetics model parameters for RTM6 epoxy resin
Figure 2-3d shows both experimental (solid line) and model-predicted (dotted line) degree
of cure for reactions conducted at 120, 140, 160 and 180 °C. The model predicted resin degree of
cure with high precision, showing only minor fit deviations (from the experimental data) near the
onsets of the final plateaus, toward the end of the cure reaction. The values of the cure kinetics
model parameters are shown in Table 2-1. Karkanas et al [59]. studied the cure kinetics of epoxy
resin (RTM6) and developed a modified autocatalytic cure kinetics model (Equation 2-4):
d𝛼 d𝑡 = 𝐾 1
( 1 − 𝛼 )
𝑛 1
+ 𝐾 2
𝛼 𝑚 ( 1 − 𝛼 )
𝑛 2

(2-4)
Predictions from both Kratz and Karkanas models (Equations 2-2 and 2-4) resulted in agreement
with the experimental results. However, the Kratz model (Equation 2-2) achieved a slightly closer
fit, because the model included more fitting parameters that accounted for the shift from a kinetics-
controlled to a diffusion-controlled cure reaction.
35
2.4.1.2 Rheology

Figure 2-4. Rheology data showing mechanical viscosity profiles of RTM6 resin during (a)
isothermal dwell tests at filling temperatures (80-120 °C), (b) isothermal dwell tests at curing
temperatures (120-180 °C) and viscosity model fitting results (experimental: scatter, model-
prediction: line), and (c) storage and loss modulus profiles measured during resin cure at 160 °C
Viscosity evolution data were obtained during the isothermal scans conducted at low
filling temperatures between 80 and 120 °C, shown in Figure 2-4a. The use of a cone-and-plate
fixture yielded virtually noise-free rheological measurements, even at low viscosity values (below
0.1 Pa ⋅s). During the 3-hour isothermal dwell, the resin viscosity curves remained nearly flat at
80 and 90 °C, while increasing notably more at 110 and 120 °C. Figure 2-4b (scatter) shows the
viscosity measurements performed at higher curing temperatures (120-180 °C). After a brief
36
induction period, resin viscosity evolved in a sigmoidal shape, analogous to the degree of cure
profiles presented in Figure 2-3d.
A phenomenological viscosity model developed by Khoun et al. was used to fit rheology
data [34]:
𝜂 = 𝜂 1
+ 𝜂 2
(
𝛼 𝑔𝑒𝑙 𝛼 𝑔𝑒𝑙 − 𝛼 )
𝐴 +𝐵𝛼 +𝐶 𝛼 2

(2-5)
𝜂 𝑖 = 𝐴 𝜂𝑖
∙ exp (−
𝐸 𝜂𝑖
𝑅𝑇
), where 𝑖 = 1, 2
(2-6)
where η is the viscosity, η1 and η2 are the Arrhenius dependent viscosity component, αgel is the
degree of cure at gelation, A, B, and C are the fitting constants, Aηi is the Arrhenius constant, Eηi is
the viscosity activation energy, and T is the temperature [K]. In Figure 2-4b, the experimental data
(scatter) and model fitting results (line) of viscosity evolution at four cure temperatures are shown.
The values of the viscosity model parameters are shown in Table 2-2. Figure 2-4c shows
representative storage (G’) and loss (G”) modulus profiles measured during resin cure at 160 °C.
The gelation point (G’ = G”), which marks the phase transition from liquid state (G” > G’) to
rubbery state (G’ > G”) [73], was determined to be near η of 10
3
Pa ⋅s with αgel of 0.63. Gelation
time is an important metric in composites manufacturing, because resin flow ceases at gelation,
thus preventing further impregnation [65]. The gel times of the resin at different cure temperatures
were measured three times at each cure temperature and are summarized in Table 2-3.
Aη1
[Pa⋅s]
Eη1
[J/mol]
Aη2
[Pa⋅s]
Eη2
[J/mol]
αgel
[-]
A
[-]
B
[-]
C
[-]
1.54 × 10
-5
3.10 × 10
3
1.02 × 10
-8
4.90 × 10
4
0.63 3.72 5.69 × 10
-3
0.03
Table 2-2. Values of the mechanical viscosity model parameters for RTM6 epoxy resin
37
Temperature [°C] 120 140 160 180
Gelation Time [min] 270.6 (± 1.9) 101.5 (± 0.6) 47.1 (± 1.3) 21.5 (± 0.3)
Table 2-3. Gelation times of RTM6 epoxy resin at different reaction temperatures
2.4.1.3 Dielectric Analysis & Resin Cure Map

Figure 2-5. (a) DEA data showing ion viscosity profiles of the epoxy resin during isothermal
dwell tests at curing temperatures (120-180 °C), (b) α-based resin cure map, in which the dashed
lines represent degree of cure isolines, and (c) η-based resin cure map, in which the dashed lines
are mechanical viscosity isolines
Previous studies have explored the use of dielectric cure monitoring of epoxy (RTM6) in
various aspects. For example, Kazilas et al. conducted a comprehensive DCM study on the resin,
ranging from impedance curve modeling to temperature modulated dielectric analysis [74]. In
38
related work, Skordos et al. used imaginary impedance maximum to accurately monitor cure
reaction progress under both isothermal and dynamic curing conditions [75]. Finally, Karkanas et
al. demonstrated that the inflection point of the conductivity curve can be used to identify the point
of resin gelation [76]. However, the principle of determining gel point from the conductivity
inflection point was valid only for selected resin systems (e.g., autocatalytic resins, including
RTM6), and thus was not suitable for universal application [76]. In this study, we developed resin
cure maps to identify the cure state of interest (e.g., gel point) during actual part manufacture and
to enable process adjustments in situ.
Figure 2-5a shows the ion viscosity data measured during the isothermal dwell tests
conducted at curing temperatures (120-180 °C). The sigmoidal shape of the ion viscosity curves
resembles the profiles for degree of cure and mechanical viscosity, indicating that the three
properties are correlated. As the state of cure progresses and mechanical viscosity increases, the
degrees of both ion mobility and dipole rotation decrease, leading to a rise in ion viscosity [30]. In
addition, similar to mechanical viscosity, which exhibits an Arrhenius temperature dependence,
ion viscosity decreases with increasing temperature at a given cure state [28]. The ion viscosity
(IV) depends strongly on temperature, and can be expressed using the following equation [28]:
log( 𝐼𝑉 ) = log (
𝑘 𝑞 2
𝑛 𝐷 0
) + log( 𝑇 )+
𝑄 𝑘𝑇 ln( 10)
(2-7)
where k is Boltzmann’s constant, q is the magnitude of electronic charge, n is the free ion
concentration, D0 is the pre-exponential factor for the diffusion coefficient, Q is the activation
energy for diffusion, and T is the temperature [K]. In Equation 2-7, all parameters on the right side
are independent of temperature, and the equation can be re-written as:
39
log( 𝐼𝑉 ) = 𝐴 + log( 𝑇 )+
𝐵 𝑇
(2-8)
where both coefficients A and B depend on α and thus η, because D0 and Q vary with degree of
cure [28]. Between the two temperature terms, log(T) and 1/T, the latter dominates the temperature
dependence of ion viscosity, and thus ion viscosity decreases as temperature increases for a given
α.
α-based Cure Map η-based Cure Map
α A B η A B
0.1 -1.10 2.48 × 10
3
1 -0.62 2.70 × 10
3

0.2 -2.32 3.10 × 10
3
10 -3.49 4.10 × 10
3

0.3 -3.51 3.73 × 10
3
10
2
-5.71 5.15 × 10
3

0.4 -4.96 4.48 × 10
3
10
3
-6.79 5.69 × 10
3

0.5 -6.09 5.13 × 10
3
-
0.6 -6.95 5.69 × 10
3
-
Table 2-4. Values of the ion viscosity model parameters for RTM6 epoxy resin
Based on these correlations, resin cure maps (for RTM6), plotting ion viscosity along
either α or η isolines as a function of temperature¸ were constructed by correlating cure kinetics
and rheology data to dielectric measurements (Figure 2-5b and 2-5c, scatter). The resin cure map
data were fit using the ion viscosity model (Equation 2-8), and the results are presented as dashed
lines in Figures 2-5b and 2-5c. The values of the ion viscosity model parameters are shown in
Table 2-4. Near the gel point, the parameters of the α- and η-based ion viscosity models are similar
(at αgel ~ 0.63 and η ~ 10
3
Pa ⋅s). The model predicts ion viscosity at the specified level of α or η
as well as temperature. The DCM system permits in situ monitoring of resin ion viscosity, and the
40
resin cure maps and ion viscosity model can convert the collected ion viscosity data into degree of
cure or mechanical viscosity metrics. Thus, during actual part manufacture, the resin cure maps
can be employed to identify the gel point (or other cure state of interest) at any given cure
temperature for real-time process adjustments. The methodology deployed can be applied to any
epoxy resin system, in principle.
2.4.2 Permeability Characterization & Infusion Simulation
Both plain weave carbon fiber fabric and polyamide veil exhibited 2D isotropic flow
during radial flow permeability measurements. In each experimental run, a radial flow pattern was
recorded (1920 × 1080 px resolution), which was subsequently processed to yield flow front
position against flow time data (Figure 2-6a). The permeability value can be obtained using the
following isotropic permeability model for constant injection pressure, which is derived from the
Laplace equation in polar coordinates, combined with Darcy’s law and the continuity equation
[70,77]:
𝐾 = [𝑟 𝑓 2
(2 ln (
𝑟 𝑓 𝑟 0
) − 1) + 𝑟 0
2
]
1
𝑡 𝜂𝜖
4Δ𝑃 = 𝐹 𝑖 𝜂 𝜖 4Δ𝑃 (2-9)
where K is the permeability [m
2
], rf is the flow front radius, r0 is the inlet port radius, t is the time,
η is the fluid viscosity, ε is the porosity, ΔP is the inlet and outlet pressure difference [Pa]. In the
experiment, rf was measured against t, and the terms in the bracket of Equation 2-9 (N) was plotted
against time at different flow front positions (Figure 2-6b). Then, the slope of the resulting linear
regression line, Fi, was multiplied by ηε/4ΔP to calculate material permeability. The resulting
permeability of the polyamide veil, 2.3 × 10
-10
(± 2.6 × 10
-12
) m
2
, was an order of magnitude greater
than that of the carbon fiber fabric, 2.5 × 10
-11
(± 4.6 × 10
-13
) m
2
. The permeability value obtained
41
for 3K plain weave carbon fiber fabric was similar to a previously reported value [70]. The goal
of this permeability analysis was to verify that the relative permeability difference between the
carbon fiber fabric and the thermoplastic veil was large enough to create non-uniform flow fronts
during the infusion process.

Figure 2-6. (a) Progressive images of flow fronts captured during the plain weave carbon fiber
fabric permeability measurements, and (b) N term (terms in the bracket of Equation 2-9) plotted
against flow time for permeability measurements of plain weave carbon fiber fabric and non-
42
woven polyamide veil

Figure 2-7. Infusion simulation results for a multilayer laminate containing 16 plies of plain
weave carbon fiber fabric and 7 plies of polyamide veils, expressed in filling factor (degree of
saturation) gradient
Using the measured permeability values, resin infusion through a multilayer preform
containing 16 plies of carbon fiber fabric and 7 plies of thermoplastic interlayer was simulated
using a commercial software (PAM-RTM). Figure 2-7 shows the simulated resin filling factor
(degree of saturation) gradient during infusion. As expected, the simulation predicted more rapid
saturation of the veils than the carbon fiber fabrics, creating non-uniform flow fronts. The infusion
simulation demonstrates that the degree of permeability difference between the reinforcement and
thermoplastic interlayer is sufficiently large to expect the veils to function as effective interlaminar
flow distribution media, as first suggested by Nash et al [36].
2.4.3 Cure Cycle Modification & Part Manufacture
For thermoset composites toughened by thermoplastic interlaminar veils, results in the
previous section demonstrated that the interlayer veils can act as interlaminar flow distribution
media. During infusion, resin flows more rapidly through the highly permeable veils than through
43
the reinforcement fabrics and accrues in interlaminar regions, creating resin-rich interlaminar
regions and increasing inter-ply thickness [36,42]. Moreover, local variations in preform
permeability induce distorted flow fronts, resulting in air entrapment and dry spot formation,
particularly within fiber tows [45,78–80]. The formed intra-tow voids tend to migrate to resin-rich
regions such as veil interfaces, as formation of voids in resin-rich areas is energetically more
favorable than formation of voids within fiber bundles [78,81,82]. Hence, for laminates containing
inter-ply thermoplastic veils, porosity eventually concentrates in resin-rich interlaminar areas.
Because the top mold surface in VI is a flexible vacuum bag, preform thickness can
accommodate and adapt to the continuously evolving compaction pressure gradient during
infusion and post-filling stages. During the post-filling stage (i.e., the period between infusion
completion and resin gelation), the resin pressure field gradually becomes homogeneous, while
the compaction load on preform increases [44]. As a result, laminate thickness (in this case,
especially the inter-ply thickness, which increased during the infusion stage) decreases, and resin
continues to flow and redistribute until reaching an equilibrium state, facilitating impregnation of
unsaturated regions including macro-pores [36,44,54,55]. Insufficient post-filling time can thus
result in porosity in interlaminar regions of cured laminates containing thermoplastic interlayers.
Nash et al. suggested that increasing resin flow time may reduce porosity [36], although the
hypothesis was not tested.
The rapidly advancing degree of cure and viscosity observed during part production using
the MRCC (120 minutes at 180 °C) may not always suffice to fully saturate dry spots during the
post-filling stage. The time to attain the equilibrium state of resin is inversely proportional to resin
viscosity, while resin gel time limits the time window available for resin redistribution [54]. To
address this potential problem, an additional post-infusion dwell step at low temperature (129 °C)
44
was introduced in the process cycle to prolong the period of low-viscosity prior to gelation. The
additional dwell was intended to promote resin flow into partially saturated interlaminar regions
that remain after infusion completion. Figures 2-8a and 2-8b show the temperature profiles of the
MRCC (dashed) and modified cure cycle (solid), as well as the degree of cure and viscosity profiles
predicted by the cure models. In the modified cure cycle, the cure temperature was raised to 129 °C
from the fill temperature (120 °C) and held constant until gelation, followed by a one-hour final
cure dwell at 180 °C. The viscosity model predicted resin gelation after 160 minutes of the 129 °C
dwell, while the cure kinetics model predicted that both MRCC and modified cycle would yield a
final degree of cure (αf) of 0.97.

Figure 2-8. Profiles of the MRCC (dashed) and the modified cure cycle (solid) showing
temperature and predicted profiles of (a) degree of cure and (b) mechanical viscosity, and (c) ion
viscosity measured during Case C (modified cure cycle) part manufacture
45
The mechanical viscosity model provided an estimate of gel time at the desired cure
temperature. However, the actual duration of the post-infusion dwell (in this case, the gel time)
was determined using the DCM system in conjunction with the resin cure maps. The DCM system
provided in situ measurements of ion viscosity during the cure process, while the cure maps were
used to identify specific cure states of interest (in this case, the gel point) and adjust process time
in situ. The ion viscosity model and resin cure maps predict that at 129 °C, the ion viscosity reaches
9.95 at the gel point (η ~ 10
3
Pa ⋅s or α ~ 0.63). While producing Case C sample with the modified
cure cycle, this ion viscosity value of 9.95 was attained 170 minutes into the post-infusion dwell
(Figure 2-8c), which was similar to the mechanical viscosity model prediction of 160 minutes. The
slight difference between the predicted and measured gel times (~ 10 minutes) was attributed to
the difference in sample sizes used for rheology tests and for part manufacture.
The main objective of this study was to determine and demonstrate the effectiveness of a
post-infusion dwell in reducing porosity in vacuum-infused laminates containing thermoplastic
interlaminar veils. During the post-infusion dwell, the resin remained in the low-viscosity region
longer, allowing more time for dry spot saturation throughout the post-filling stage. The addition
of the post-infusion dwell can facilitate manufacture of autoclave-quality part using VI at much
lower cost. Furthermore, the effectiveness of the post-infusion dwell may not be limited to the
specific VI process case discussed here (toughening with non-woven thermoplastic veil
interlayers). In fact, other VI processes may also benefit from longer post-filling times, particularly
in cases involving multi-component preforms comprised of different reinforcement materials,
and/or three-dimensional preforms featuring complex shapes. In such preforms, local variations in
permeability can generate distorted flow fronts, causing air entrapments and dry spots.
Note that the addition of the post-infusion dwell reduced porosity in the toughened
46
laminate at the cost of increased cure cycle time (from 2.5 to 4 hours). However, in practice, the
entire process of VI generally consists of multiple stages, including preform fabrication and lay-
up, vacuum bagging, resin infusion and cure, and post-cure machining. Thus, the cure cycle itself
may represent only a minor fraction of the net manufacturing time, and extending the duration of
the cure cycle may have negligible effect on the total process time. Admittedly, further process
optimization may trim the duration of the cure cycle, possibly by increasing the post-infusion dwell
temperature. In this study, the dwell temperature (129 °C) was selected based on resin gelation
data (Table 2-3) to ensure sufficiently long post-filling duration for dry spot saturation. (At 120
and 140 °C, the resin gelled after 270 and 100 minutes.) In practice, the post-infusion dwell
temperature can be adjusted to optimize the cure cycle. An increase in dwell temperature (e.g.,
140 °C or higher) can reduce both the minimum viscosity level and the processing time, but only
at the cost of accelerating the evolution of resin viscosity and abbreviating the post-filling stage,
which can increase porosity. Additional work is required to the explore trade-offs between reducing
the overall process time (by decreasing the post-infusion dwell period) at the expense of increasing
risk of porosity.
47
2.4.4 Part Quality Analysis
2.4.4.1 Porosity Analysis

Figure 2-9. Cross-sectional micrographs of the vacuum-infused laminates (Case A: containing no
veils and manufactured using the MRCC, Case B: toughened by veils and manufactured using
the MRCC, and Case C: also toughened by veils but manufactured using the modified cure
cycle)
Cross-sectional images of the vacuum-infused 16-ply laminates are shown in Figure 2-9.
Case A samples contained no thermoplastic interlayers and were nearly void-free. Case B and C
samples, which were toughened by 7 plies of thermoplastic veils, showed substantial increases (~
0.1 mm per one ply of veil) in interlaminar thickness, with veil fibers embedded between plies.
The toughened laminates produced using the MRCC (Case B) exhibited high internal porosity (~
6.5 % by volume). As reported elsewhere [36], voids were elongated and aligned along fiber
48
directions in inter-ply regions, indicating that the voids were originally located within fiber bundles
and later migrated to energetically favorable resin-rich interlaminar areas. High laminate porosity
can degrade mechanical properties of cured parts, and thus should be minimized or eliminated
[83,84]. Case C samples, fabricated using the modified cure cycle, showed a marked decrease in
porosity. A few elongated voids still appeared between plies, but these were much narrower and
smaller, demonstrating that the introduced post-infusion dwell effectively increased resin
redistribution and saturation of partially filled regions during the post-filling stage. The overall
void content of Case C remained less than 0.5 %.
2.4.4.2 Impact Testing

Figure 2-10. (a) Images and (b) ultrasonic C-scan results of the impacted laminates after drop
tower impact tests
49
The key function of the interlaminar veils was to promote resistance to impact damage, a
common concern in structural thermoset composites. Figures 2-10a and 2-10b show images and
C-scan results of the laminates impact tested at 30 J energy level. Visual inspection and C-scan
analysis of the impacted laminates demonstrated that Case A, which was not toughened by the
thermoplastic interlayers, exhibited 30 and 50 % greater damage area than Cases B and C,
respectively.

Figure 2-11. Force-time profiles of Case A, B, and C laminates measured during 30-J drop tower
impact tests
The force-time history curves obtained during the 30-J impact test were characterized by
two distinctive peaks (F1 and Fmax), as shown in Figure 2-11. The first load drop (F1) was
associated with initial damage, specifically with the onset of delamination from indentation and
local matrix cracking near the impacted region. The peak force (Fmax) was ascribed to the
maximum tolerable load before extensive global delamination to the laminate [85–87]. The F1
50
values for Case B and C (4790 ± 120 N) were approximately 9 % greater than that of Case A (4410
± 120 N), indicating that the toughened samples exhibited greater resistance to initial local impact
damage. The Fmax values of Case A, B, and C were 4890 (± 120), 5790 (± 190), and 6280 (± 90)
N respectively, and the toughened laminates also showed greater resistance to global delamination,
which arose from extensive matrix cracking. In addition, Case C exhibited much lower void
content than Case B (< 0.5 % vs. 6.5 %), resulting in 8 % greater peak load during impact.
2.5 Conclusions
The use of thermoplastic interlaminar veils in VI preforms increases impact resistance but
also induces non-uniform flow fronts, leading to porosity. Hence, the conventional cure cycle was
modified to include a low-temperature post-infusion (LTPI) dwell to ensure full saturation of dry
interlaminar regions. Process parameters for the LTPI dwell were informed by predictions from
models for the cure kinetics and viscosity, and in situ process adjustments were implemented using
the DCM system in conjunction with the resin cure maps. Laminates produced using the modified
process cycle exhibited lower porosity and greater peak load during impact testing.
This study explores the effectiveness of a post-infusion dwell on part quality in vacuum-
infused laminates with interlaminar veils. Resin cure maps can be employed to identify key
junctures in the physical state of the resin, while the use of DCM permits in situ monitoring of the
resin state during VI. Thus, DCM can identify key junctures in resin state, particularly the gel point,
which can be used to guide in situ process adjustments. Interlaminar thermoplastic veils serve two
functions: (a) enhanced impact resistance and (b) inter-ply flow distribution media. However, the
latter function also increases the risk of porosity. The veils introduce local variations in
permeability and give rise to non-uniform flow fronts during infusion. To mitigate porosity that
51
can arise from the distorted flow fronts, the LTPI dwell can be introduced to extend the post-filling
stage. During the dwell, resin can redistribute and saturate interlaminar regions prior to gelation,
reducing porosity in cured laminates.
Overall, this work describes a potential pathway to address high porosity commonly
encountered in various VI processes, arising from local variations in preform permeability,
disparity in flow rates in different layers, and non-uniform or distorted flow fronts. Composites
manufacturing is gradually transitioning from expensive conventional methods such as autoclave
processing to simpler and more cost-effective yet relatively less robust (thus far) processes
including VI. Introducing a LTPI dwell based on material characterization, resin cure map
development, and in situ cure monitoring may provide a viable solution toward increasing part
quality of vacuum-infused laminates.
52
Chapter 3. In Situ Resin Age Assessment Using Dielectric Analysis and
Resin Cure Map for Efficient Vacuum Infusion
3.1 Abstract
The physical state of epoxy resin designed for vacuum infusion was assessed in situ by
immersing a dielectric sensor into the resin pot. The measured ion viscosity of aged resin was
directly converted to a degree-of-cure metric using a resin cure map constructed by correlating
cure kinetics and dielectric analysis data. Next, an age-adjusted infusion process map was
employed to define a nominal infusion window and to identify key process metrics. Finally,
process simulations and flow contour maps were used to validate and refine the process map, and
to guide adjustment of infusion process parameters based on resin age and part size/geometry. The
study describes a pathway to more efficient use of aged resin using in situ process diagnostics,
cure map design, and process simulation. The methodology employed to evaluate resin age and to
adjust process parameters accordingly can be extended to other composite manufacturing
processes, including conventional prepreg processing.
3.2 Introduction
In this work, we evaluate the physical state (age or life) of resin in situ using dielectric
analysis (DEA) in conjunction with a resin cure map. Based on the assessed resin age, we use
process simulations and flow contour maps to adjust process parameters for vacuum infusion (VI).
The study is motivated by a need to increase VI process efficiency. High-performance thermoset
resins designed for VI have relatively short shelf-lives and out-lives, and require multiple
protracted steps to pre-condition and pre-heat prior to infusion. In general, excess resin is prepared
53
for each infusion trial to ensure uninterrupted and complete saturation of dry preforms, inevitably
leaving unused aged resin as waste. Use of aged resin must be accompanied by accurate and
meticulous tracking of resin working time and thermal history, which is often difficult in practice.
Otherwise, additional thermal analysis must be conducted to assess resin life prior to each infusion
trial, and thus excess or expired resin is often discarded for convenience.
VI can be a cost-effective alternative to conventional autoclave prepreg process, especially
for the manufacture of large and complex unitized composite structures [14,15,49,51,88–91]. In
VI, a dry fiber preform is placed on a one-sided rigid mold and sealed with a flexible vacuum bag,
and resin is infused into the preform under vacuum pressure, then heated and cured [15,45,47–
49,91,92]. Interest in VI has grown rapidly in recent years, particularly in the aerospace industry,
which seeks to reduce the manufacturing costs associated with prepreg processing. However,
because vacuum pressure alone is applied during infusion, VI involves much slower infusion rates
and thus longer fill times relative to resin transfer molding (RTM), in which resin is injected under
positive pressure [14,15]. Inserting a high-permeability flow distribution medium on top of the
preform (e.g., SCRIMP, Seemann Composites Resin Infusion Molding Process) can enhance resin
flow and reduce fill times, but also can induce unacceptable void contents, particularly for large
parts with complex geometries [14,15]. To mitigate this issue, resin is often heated during filling,
although doing so increases resin degree of cure and viscosity with flow time and distance [14,16].
Thus, the process parameters for heated filling during VI must be selected carefully; even for
SCRIMP, the problem of evolving resin state can persist, particularly as part size increases.
Aerospace-grade thermoset resins designed for VI processes (e.g., Hexcel HexFlow
®

RTM6) have relatively short shelf-lives (months) and out-lives (weeks or days), and require
protracted pre-processing and pre-heating [56,93,94]. In general, a VI resin must be pre-
54
conditioned at room temperature (from cold storage), then pre-heated to reduce viscosity for
transfer to an infusion pot. The resin is maintained at the pre-heating temperature until degassing
and infusion steps are completed. Depending on the size and geometry of the part being infused,
resin will age (i.e., resin degree of cure will advance) throughout the protracted pre-conditioning
and pre-heating steps, reducing resin life. Furthermore, each infusion run generates waste resin,
because excess resin generally is prepared to ensure uninterrupted infusion of preforms. Unused
resin cannot be reused unless the thermal history—including shelf-life, out-life, and working
time—of the material has been tracked rigorously [95]. Even for previously unused resin, accurate
evaluation of resin life can be challenging, as resin ages even during freezer storage, albeit slowly.
Hence, expired resin is generally discarded without further material assessment, resulting in
economic loss and environmental hazard [95].
Previous studies have explored the use of in situ process diagnostics in liquid molding for
on-line monitoring of resin state, and investigated the effects of curing dependent viscosity on
heated infusion processes. For example, Pantelelis et al. developed a DC-based process monitoring
system to track the electrical resistance of resin during cure, and modeled resin viscosity as a
function of resistance using an empirical power law [95]. The same work also identified the
common manufacturing problem of aged resin exhibiting greater initial viscosity and more rapid
viscosity increase during infusion (compared to fresh resin). However, no remedies were offered
to guide process adjustments to mitigate the problem aside from mixing aged resin with fresh resin
at an arbitrary ratio [95,96]. For dielectric cure monitoring (DCM) of epoxy resin (RTM6),
Karkanas et al. related the dielectric response of the resin to chemical and physical changes (e.g.,
gel point and vitrification) using a principle specific to selected resin systems (autocatalytic resins
including RTM6) [76]. Kazilas et al. conducted impedance curve modeling and temperature
55
modulated dielectric analysis on the same resin (RTM6) [74]. Finally, Grujicic et al. [97] and Wu
et al. [98] developed mold-filling models to account for cure-state-dependent viscosity, and
performed heated filling simulations to minimize filling time for non-isothermal VI.
In this work, we demonstrate a new method to more accurately assess the physical state
of resin, using in situ process diagnostics coupled with cure modeling. To accurately monitor and
analyze the resin state, we develop a new resin cure map by correlating a cure kinetics model to
DEA ion viscosity measurements. We then plot degree-of-cure isolines across different
temperatures using an ion viscosity model. The straightforward linear relationships between the
ion viscosity model parameters and the resin degree-of-cure lead to accurate predictions of ion
viscosity across a wide span of resin age and temperature. Such predictions minimize the need for
extensive material characterization generally required for DEA modeling. Using these tools, we
can convert the ion viscosity data for an aged resin directly into a metric for resin life. We
demonstrate how to use this metric to guide process adjustments “on-the-fly” to compensate for
resin age. In addition, we conduct heated infusion simulations to determine the effects of process
temperature and resin age on the maximum resin flow distance and to develop flow contour maps.
The flow maps are intended to guide selection of resin-age-adjusted process parameters (infusion
temperature or number/placement of resin inlets).
The work described outlines a pathway to reduce the waste of aged resin, a widespread
problem in VI, and to efficiently adjust heated infusion process parameters based on resin age and
part geometry. First, the cure kinetics, rheology, and aging behavior of epoxy resin are examined
using differential scanning calorimetry (DSC), rheometry, and DEA. Next, a resin cure map is
constructed and deployed in conjunction with an on-line DCM system to assess the life of aged
resin in situ. The nominal infusion window and key process metrics are determined using an age-
56
adjusted infusion process map, while heated filling VI simulations demonstrate how resin age and
infusion temperature (combined) can affect the maximum resin flow distance. Finally, we conduct
a parametric filling simulation study and develop flow contour maps to guide selection of infusion
parameters. The methods allow more efficient use of aged resin and informed process adjustments.
Overall, the methodology employed in this work provides a blueprint to implement VI process
diagnostics and to guide adjustments that will reduce material waste for VI, as well as for other
composite manufacturing processes.
3.3 Experimental
3.3.1 Materials
An aerospace qualified epoxy resin was selected and acquired (HexFlow
®
RTM6, Hexcel).
The mono-component epoxy resin system (with curing agent pre-mixed) was developed for LCM
processes and featured a shelf-life of 9 months (at -18 °C) and an out-life of 15 days (at ambient
conditions) [56]. After freezer storage, the resin must be pre-conditioned at room temperature for
24 hours, followed by pre-heating to 60-80 °C for infusion pot transfer. In the pot, the resin is
maintained at 80 °C throughout degassing and infusion. The manufacturer-recommended infusion
temperature is 120-140 °C.
3.3.2 Cure Kinetics & Rheology
3.3.2.1 Modulated Differential Scanning Calorimetry (MDSC)
Resin cure kinetics was characterized using DSC (TA Instruments, Q2000) under a
nitrogen purge (50 cm
3
/min). Resin samples (8-10 mg) were first exposed to a dynamic ramp at
57
four heating rates (5, 10, 15, and 20 °C/min) from -50 to 340 °C to determine the total heat of
reaction. Isothermal dwell measurements were conducted at four temperatures (120, 140, 160, and
180 °C) to analyze the cure kinetics, followed by a ramp to 300 °C at 5 °C/min to measure the
residual heat of cure. During the subsequent ramp, the samples exhibited an endothermic annealing
peak near the residual cure exotherm because of physical aging of the resin [59]. To eliminate the
annealing peak, the specimens were quenched rapidly at the endotherm spike to -50 °C and
reheated during re-scan. During the ramp, a sinusoidal temperature modulation of ±0.5 °C/min
was applied to distinguish reversing and non-reversing heat flow signals. Reversing heat flow is
associated with heat capacity and rate of temperature change, while non-reversing heat flow is
dependent on kinetic component of reaction [53,60]. Glass transition can be examined from the
reversing heat flow signal, while the curing reaction appears in the non-reversing signal.
3.3.2.2 Rheometry
Viscosity evolution during cure was measured using a rheometer (TA Instruments,
AR2000ex). Resin samples were subjected to an isothermal dwell at four temperatures (120-
180 °C), analogous to MDSC isothermal dwell measurements. A disposable parallel-plate
geometry fixture (diameter 25 mm) was used to measure the evolving viscosity throughout the
cure cycle. Cell temperature was raised from the pre-heating temperature (80 °C) to the dwell
temperature at approximately 30 °C/min (the maximum ramp rate of the rheometer) and held
constant for prescribed durations, under oscillatory shear at 1 Hz frequency and 5 % strain (within
the linear viscoelastic region).
3.3.2.3 Dielectric Analysis (DEA)
A dielectric cure monitoring (DCM) system (Netzsch, DEA 288 Epsilon) was used to
58
perform DEA measurements in frequency intervals from 1 Hz to 1 kHz at four temperatures (120,
140, 160, and 180 °C). An interdigitated electrode sensor (Netzsch, Mini-IDEX 100/35) with 33
mm
2
sensing area and 100 μm electrode spacing was mounted on a rheometer Peltier plate, along
with a thermocouple. A resin sample was placed on the sensor, and cell temperature was increased
from 80 °C to the dwell temperature at ~ 30 °C/min, then held constant until completion of cure.
Once a sinusoidal excitation voltage is applied, dipoles and charge carriers within resin align with
the applied electric field and move toward electrodes of opposite charge. The resulting sinusoidal
current and phase shift response yields dielectric properties, including complex permittivity (ε*)
and ionic conductivity (σ). The reciprocal of σ is the ion viscosity, and exhibits a strong correlation
to mechanical dynamic viscosity (η) [63,65,66].
3.3.3 Resin Aging
Resin samples were aged by storing at room temperature (25 °C) for 1, 2, 4, 7, 10 days
and 2, 3, 4, 5, 6 weeks (two samples tested for each out-time period). Once aging was complete,
each sample was placed in the DSC and heated to 340 °C at 5 °C/min to determine the degree of
cure (α) accrued during aging. All samples were weighed before and after aging to confirm that no
weight loss had occurred. The purpose of this task was to monitor how fast the resin aged at
ambient condition, and thus to enable prediction of α advancement with respect to out-time.
3.3.4 Filling Simulation
Simulations of VI heated filling were conducted using commercial software (PAM-RTM,
ESI Group). The simulations assumed a preform comprised of 20 plies of plain weave carbon fiber
fabric (areal weight of 193 g/m
2
and 3000 fiber/tow count, part #1530, Fibre Glast [57]). The in-
plane permeability of the fabric was measured previously (2.5 × 10
-11
m
2
, isotropic) using an
59
unsaturated radial flow configuration [99].
Three VI cases were simulated, each with different infusion length: 530 mm (Case A), 600
mm (Case B), and 670 mm (Case C). The width of the preform was constant (200 mm) in all cases.
In each case, aged resin with initial degree of cure (α0) of 0.10 was infused, and three different
initial mold temperatures were employed (120, 130, and 140 °C). The process pressures were
specified across the linear width boundaries: atmospheric pressure (1.01 × 10
5
Pa) was applied at
the resin inlet boundary on one end, while vacuum (0 Pa) was applied at the vent boundary on the
opposite end. For each simulation run, a total of 9,558 elements were meshed (triangle type and
linear order). The objective of the filling simulation was to validate and refine the infusion process
map and to inform subsequent adjustment of process parameters. A second purpose was to
demonstrate the need to adjust the infusion process parameters not only for part size and geometry,
but particularly for resin age.
3.4 Results & Discussion
3.4.1 Cure Kinetics, Rheology, and Dielectric Analysis
3.4.1.1 Cure Kinetics
The total heat of cure reaction (Hr,total) measured was 430 J/g, similar to a previous report
(436 J/g [59]). Figure 3-1a shows the MDSC heat flow data measured via isothermal dwell tests.
The heat flow curves evolved more rapidly as the cure temperature increased, with shorter reaction
induction periods. Assuming that the rate of cure reaction was proportional to the rate of heat flow,
the heat flow data was converted to cure rate data using the equation [31,34]:
60
d𝛼 d𝑡 =
1
𝐻 𝑟 (
d𝐻 d𝑡 ) (3-1)
where α is the resin degree of cure, t is time, and H is the heat flow measured from the MDSC.
The resin degree of cure can be calculated by integrating Equation 3-1. The final degree of cure
(αf) values for samples reacted at 120, 140, 160, and 180 °C were 0.83, 0.88, 0.93, and 0.97
respectively.

Figure 3-1. (a) MDSC data showing heat flow profiles measured during isothermal dwell tests,
and (b) experimental (solid line) and cure kinetics model-predicted (dotted line) degree of cure
profiles for the isothermal dwell tests
The MDSC data for resin cure were fit to a phenomenological cure kinetics model
developed by Kratz et al. [31], which accounts for transitioning from kinetics-controlled to
diffusion-controlled cure reaction as α exceeds αgel (α at resin gelation), and a rapid rate of cure
present at low-α (α < 0.1):
d𝛼 d𝑡 = 𝐾 1
𝛼 𝑚 1
( 1 − 𝛼 )
𝑛 1
+
𝐾 2
𝛼 𝑚 2
( 1 − 𝛼 )
𝑛 2
1 + exp (𝐷 ( 𝛼 − ( 𝛼 𝐶 0
+ 𝛼 𝐶𝑇
𝑇 ) ) )

(3-2)
61
𝐾 𝑖 = 𝐴 𝑖 ∙ exp (−
𝐸 𝐴𝑖
𝑅𝑇
), where 𝑖 = 1, 2 (3-3)
where Ki is the Arrhenius temperature dependent term, m1, m2, n1, and n2 are the reaction order-
based fitting constants, D is the diffusion constant, αC0 is the critical α at absolute zero, αCT
accounts for the increase in critical α with temperature, Ai is the pre-exponential factor, EAi is the
activation energy, R is the universal gas constant, and T is the temperature. The measured (solid
line) and predicted (dotted line) degree of cure profiles for isothermal reactions are shown in Figure
3-1b. The results demonstrate that the model accurately predicted degree of cure at all four cure
temperatures. The values of the cure kinetics parameters are shown in Table 3-1.
Parameter Value Parameter Value
A1 [s
-1
] 1.70 × 10
6
A2 [s
-1
] 1.63 × 10
4

EA1 [J/mol] 8.22 × 10
4
EA2 [J/mol] 5.72 × 10
4

m1 0.27 m2 1.15
n1 10.65 n2 1.22
D 43.34 αC0 -0.11
αCT [K
-1
] 2.23 × 10
-3

Table 3-1. Values of the cure kinetics model parameters for RTM6 epoxy resin
62
3.4.1.2 Rheology

Figure 3-2. (a) Rheology data showing experimental (dotted) and viscosity model-predicted
(solid line) mechanical viscosity profiles for isothermal dwell tests, and (b) storage and loss
modulus profiles measured at 160 °C
Viscosity profile data were acquired from isothermal scans conducted at temperatures
between 120 and 180 °C. A phenomenological viscosity model developed by Khoun et al. was
used to fit rheology data [34]:
𝜂 = 𝜂 1
+ 𝜂 2
(
𝛼 𝑔𝑒𝑙 𝛼 𝑔𝑒𝑙 − 𝛼 )
𝐴 +𝐵𝛼 +𝐶 𝛼 2

(3-4)
𝜂 𝑖 = 𝐴 𝜂𝑖
∙ exp (−
𝐸 𝜂𝑖
𝑅𝑇
), where 𝑖 = 1, 2
(3-5)
where η is the viscosity, η1 and η2 are the Arrhenius dependent viscosity components, αgel is the
degree of cure at gelation, A, B, and C are the fitting constants, Aηi is the Arrhenius constant, and
Eηi is the viscosity activation energy. Figure 3-2a shows the measured data (dotted) and model
fitting results (solid line) for resin viscosity evolution during cure. Parameter values for the
viscosity model are shown in Table 3-2. The viscosity profile exhibited a sigmoidal shape after a
63
brief induction period, much like the degree of cure profiles shown in Figure 3-1b. Values
measured below 0.1 Pa ⋅s exhibited greater variance because of the rheometer geometry constraints.
However, the initial resin viscosity values for different temperatures were accurately predicted by
the model, yielding values similar to those reported in the material datasheet [56].
Aη1
[Pa⋅s]
Eη1
[J/mol]
Aη2
[Pa⋅s]
Eη2
[J/mol]
αgel
[-]
A
[-]
B
[-]
C
[-]
1.54 × 10
-5
3.10 × 10
3
1.02 × 10
-8
4.90 × 10
4
0.63 3.72 5.69 × 10
-3
0.03
Table 3-2. Values of the mechanical viscosity model parameters for RTM6 epoxy resin
Temperature [°C] 120 140 160 180
Gelation Time [min] 270.6 (± 1.9) 101.5 (± 0.6) 47.1 (± 1.3) 21.5 (± 0.3)
Table 3-3. Gelation times of RTM6 epoxy resin at different reaction temperatures
Figure 3-2b shows representative storage modulus (G’) and loss modulus (G”) profiles
measured during cure at 160 °C. The gelation point (G’ = G”), which marks the phase transition
from liquid state (G” > G’) to rubbery state (G’ > G”) [73], was determined to be roughly η = 10
3

Pa ⋅s and αgel = 0.63. Gelation time is an important process metric, because flow ceases at gelation,
thus arresting impregnation [73]. The gel times of the resin at different cure temperatures were
measured three times at each cure temperature and are summarized in Table 3-3, and range from
20 to 270 minutes.
64
3.4.1.3 DEA & Resin Cure Map

Figure 3-3. (a) DEA data showing ion viscosity profiles during isothermal dwell tests, (b) resin
cure map, in which filled squares (connected by solid isoline) and unfilled circles (connected by
dashed isoline) represent experimental and model-predicted data points respectively, (c) ion
viscosity model parameters plotted against resin degree of cure, exhibiting a linear relationship,
and (d) model-predicted and experimental ion viscosity values at the resin pre-heating
temperature of 80 °C
Figure 3-3a shows DEA data measured during the isothermal dwell tests at 120, 140, 160,
and 180 °C. The sigmoidal shape of the ion viscosity curves resembles the profiles for degree of
cure and mechanical viscosity, indicating that the three properties are correlated. As the state of
cure advances and mechanical viscosity increases, the degrees of both ion mobility and dipole
rotation subside, leading to an increase in ion viscosity [30]. Like mechanical viscosity, ion
65
viscosity (IV) depends strongly on temperature, and can be expressed using the equation [28]:
log( 𝐼𝑉 ) = log (
𝑘 𝑞 2
𝑛 𝐷 0
) + log( 𝑇 )+
𝑄 𝑘𝑇 ln( 10)
(3-6)
where k is Boltzmann’s constant, q is the magnitude of electronic charge, n is the free ion
concentration, D0 is the pre-exponential factor for the diffusion coefficient, and Q is the activation
energy for diffusion. In Equation 3-6, all parameters on the right side are independent of
temperature, and the equation can be re-written as:
log( 𝐼𝑉 ) = 𝐴 + log( 𝑇 )+
𝐵 𝑇
(3-7)
where both coefficients A and B depend on α and thus η, because D0 and Q vary with degree of
cure [28]. Of the two temperature terms, log(T) and 1/T, the latter term dominates the temperature
dependence of ion viscosity, and ion viscosity decreases as temperature increases.
α 0.1 0.2 0.3 0.4 0.5 0.6
A -1.10 -2.32 -3.51 -4.96 -6.09 -6.95
B 2.48 × 10
3
3.10 × 10
3
3.73 × 10
3
4.48 × 10
3
5.13 × 10
3
5.69 × 10
3

Table 3-4. Values of the ion viscosity model parameters for RTM6 epoxy resin
Based on the correlations described, a resin cure map was constructed (for RTM6) by
plotting ion viscosity for α isolines (lines that connect the same level of α) as a function of cure
temperature (Figure 3-3b). The filled squares (from 393 to 453 K in the α range between 0.1 and
0.6) were plotted by comparing cure kinetics data to DEA measurements, and the α isolines (solid
curves) were mapped by fitting the data points to the ion viscosity model (Equation 3-7). The
values of the IV model parameters are shown in Table 3-4. Both parameters A and B exhibited a
linear dependence on resin degree of cure (Figure 3-3c), indicating that the α isolines can be
66
constructed for any α values of interest (e.g., α = 0.01 and 0.05 isolines drawn in Figure 3-3b).
Using the IV model, the α isolines were extended to a pre-heating temperature of 80 °C
(unfilled circle points along 353 K), which was deemed the most practical temperature at which
resin life can be monitored in situ using a DCM system. At 80 °C, the ion viscosity values were
measured at different levels of α and were compared against the model-predicted values (Figure
3-3d). The comparison showed that the measured and predicted values differed by only 0.08 on
average (below 1.0 %), demonstrating the accuracy of the ion viscosity model. Overall, the DCM
system enabled in situ measurement of resin ion viscosity, while the resin cure map and ion
viscosity model were deployed to convert the measured ion viscosity straight into a metric that
reflected resin life. Using these tools, we can develop a program that yields resin degree of cure at
the specified temperature and ion viscosity, or one that predicts resin ion viscosity evolution during
a process cycle.
3.4.2 Resin Aging
Resin samples were aged at ambient conditions for up to 6 weeks. The degree of cure
accrued during aging (αaged) was calculated using the equation:
𝛼 𝑎𝑔𝑒𝑑 ( fraction of aged resin) =
𝐻 𝑟 ,𝑡𝑜𝑡𝑎𝑙 ( J/g)− 𝐻 𝑟 ,𝑎𝑔𝑒𝑑 ( J/g)
𝐻 𝑟 ,𝑡𝑜𝑡𝑎𝑙 ( J/g)
(3-8)
where Hr,total is the total heat of cure reaction (430 J/g) and Hr,aged is the heat of reaction of the aged
resin. The underlying assumption was that the degree of cure increased as a function of aging time.
The data from the aging study, shown in Figure 4-4a, demonstrated that αaged increased linearly
with out-time (R
2
= 0.98). According to the material datasheet [56], the resin shelf-life at room
temperature was 15 days. The linear regression line (Figure 4-4a) yielded αaged of 0.05 for resin
67
aged 15 days. After 30 days of out-time (twice the manufacturer’s out-time specification), αaged
reached 0.10.

Figure 3-4. (a) Degree of cure accrued during aging (αaged) plotted against room temperature out-
time, and (b) degree of cure profile at 80 °C, predicted by the cure kinetics model
Figure 4-4b shows the predicted degree of cure profile for a pre-heating temperature of
80 °C, obtained using the cure kinetics model. At this temperature, α is expected to reach 0.05 and
0.10 after 16 and 21 hours of dwell, respectively. The manufacturer’s processing guideline [93]
states that the maximum available working time of the resin is 12 hours at 80 °C. Combining this
specification with results from thermal analysis, the resin will be suitable for infusion, provided α
remains below ~ 0.05. The purpose of the aging study was to monitor and predict resin aging at
ambient and pre-heating conditions, and to determine the degree of cure (α) considered suitable
for infusion without requiring further material assessment.
68
3.4.3 Infusion Process Map

Figure 3-5. (a) Infusion process map for aged resin with α0 of 0.05, infused at 120 °C, and (b)
more comprehensive process map for infusion temperatures of 120, 130, and 140 °C, using aged
resin with α0 of 0.10
69
The age-adjusted nominal infusion process window, and key process metrics for infusion
of aged resin at different temperatures, can be determined by constructing an infusion process map.
In principle, infusion must be completed before the resin gels, because resin flow ceases at gelation
(αgel = 0.63) [65]. In practice, however, resin viscosity continuously rises with increasing degree
of cure, and once η exceeds a threshold value, resin velocity becomes impractically low for further
impregnation of dry preforms. This critical η value (ηcritical), which can vary with the reinforcement
type and resin system used, can be readily adjusted as needed. In this work, a value of 1 Pa ⋅s for
ηcritical was applied, which is widely regarded as the maximum resin viscosity suitable for infusion
[100–102]. The degree of cure at the threshold (αcritical) changes with process temperature because
resin viscosity exhibits an Arrhenius temperature dependence.
Figure 3-5a shows an example of a process map for infusion with aged resin (initial degree
of cure (α0) = 0.05). At an infusion temperature of 120 °C, αcritical (α at ηcritical of 1 Pa ⋅s) is 0.38.
The cure kinetics model predicts that for previously unused fresh resin, the degree of cure becomes
α0 (0.05) and αcritical (0.38) after 85 and 210 minutes at 120 °C, resulting in a nominal infusion
window of 125 minutes (210 less 85) for aged resin, depicted in the red shaded region. A more
comprehensive infusion map and its application are illustrated in Figure 3-5b. In this case, α0 is
increased to 0.10, the value for resin corresponding to 30 days of out-time (twice the manufacturer-
specified room temperature shelf-life). With increasing infusion temperature (indicated by red,
green, and blue curves in Figure 3-5b), the initial resin viscosity (η0 or η at α0) decreases, while
αcritical increases slightly because of the Arrhenius temperature dependence of resin viscosity. For
resin with α0 = 0.10, the nominal infusion windows, indicated by red, green, and blue shaded areas
in Figure 3-5b, are 89, 61, and 42 minutes at 120, 130, and 140 °C, respectively.
70

Figure 3-6. Time required for aged resins (α0 = 0.01, short dashed; 0.05, dashed; 0.10, solid line)
to reach different levels of viscosity, plotted against viscosity at infusion temperatures of (a) 120
and 130 °C, and (b) 120 and 140 °C
The competing effects of temperature on resin viscosity are depicted in Figures 3-6a and
3-6b, which show the time required for aged resin (α0 = 0.01, short dashed; 0.05, dashed; 0.10,
71
solid line) to reach different viscosity levels (η) at different temperatures. As infusion temperature
is increased, η0 (x-intercept) decreases, expediting infusion flow. Yet, higher temperature also
accelerates the evolution of viscosity, and consequently resin impregnation decelerates more
rapidly. For example, when α0 = 0.10, the initial viscosity values of the aged resin are 0.06, 0.04,
and 0.03 Pa ⋅s at 120, 130, and 140 °C, respectively. However, after 28 and 22 minutes at the dwell
temperature, the resin viscosity at 130 (0.10 Pa ⋅s) and 140 °C (0.09 Pa ⋅s) eventually exceeds the
resin viscosity at 120 °C.
Beyond these η cross-over points, the benefits of higher infusion temperature vanish,
because higher temperature causes η to increase more rapidly, decelerating the infusion process.
For an infusion process predicted to saturate before or shortly after reaching tcross-over (time required
to reach the η cross-over point), infusing at higher temperature would expedite infusion. Otherwise,
infusing at lower temperature would help prevent premature resin gelation before infusion
completion. Also, with increasing resin age (α0), tcross-over decreases, indicating that the advantages
of higher infusion temperature disappear earlier in the infusion process.
72
3.4.4 Process Validation & Simulation Refinement

Figure 3-7. Representative simulation results of VI heated filling of aged resin (α0 = 0.10) for (a)
infusion length of 530 mm and infusion temperature of 120 °C, expressed in terms of fill time
gradient, and infusion length of 670 mm and infusion temperatures of (b) 130 and (c) 140 °C,
expressed in terms of filling factor gradient
73
The cure reaction for epoxy is exothermic, and the heat generated during cure can increase
the effective process temperature. The increased temperature can accelerate the cure reaction,
producing a greater exotherm and further increasing the process temperature. The thermal runaway
effect can become severe at higher infusion temperatures, as resin degree of cure and viscosity
increase more rapidly, driving the resin to gelation more quickly. The nominal infusion process
window estimated from the infusion process map does not account for the reaction exotherm, and
thus is likely to be wider than the infusion window in practice. Thus, to validate and refine the
process map and associated process parameters, the heated filling process was simulated using
FEA software (PAM-RTM).
Infusion
Length
530 mm 600 mm
Tinfusion tfill αfinal Tmax tfill αfinal Tmax
120 °C 33 min 0.188 128 °C 44 min 0.242 133 °C
130 °C 23 min 0.197 139 °C 30 min 0.264 150 °C
140 °C 16 min 0.205 150 °C 22 min 0.294 159 °C
Table 3-5. Heated filling simulation results for infusion lengths of 530 and 600 mm (α0 = 0.10)
Simulation of a representative heated filling process yielded the result shown in Figure 3-
7a. The simulation assumed an infusion length of 530 mm, α0 = 0.10, and infusion temperature
(Tinfusion) = 120 °C. Under these conditions, the infusion was completed in 33 minutes (fill time or
tfill), and the final degree of cure (αfinal) and maximum temperature (Tmax) at the outlet were ~ 0.2
and 128 °C, respectively. The simulation results for infusion lengths of 530 and 600 mm are
summarized in Table 3-5. For both infusion lengths, higher infusion temperature resulted in shorter
fill times, as expected. However, for 600 mm infusion length, both αfinal and Tmax increased more
74
with increasing Tinfusion, demonstrating that α (and thus η) rise much more rapidly at higher
temperature. A 670 mm infusion length was also analyzed. At 120 °C, the infusion process
completed after 65 minutes, whereas at 130 and 140 °C, the resin gelled before the infusion
completed, leaving unfilled dry regions near the outlet (Figures 3-7b and 3-7c). The unsaturated
area was larger at 140 °C (vs. 130 °C), indicating that the degree of cure advanced more rapidly at
higher temperature, leading to premature resin gelation.
For small parts of simple geometry, infusion is likely to be completed before the degree
of cure and viscosity evolve significantly, favoring higher infusion temperatures (to reduce fill
times). On the other hand, larger, more complex parts (i.e., parts requiring greater fill time) can
benefit from lower infusion temperature. Once the resin viscosity exceeds the η value at the cross-
over points (determined from Figure 3-6), lower infusion temperature becomes advantageous,
because η increases more slowly, providing a wider process window. Use of flow distribution
media can mitigate the problem of long fill times (slow infusion) encountered in VI. However, as
infusion length increases and part geometry becomes more complicated, the problem of premature
resin gelation during high temperature infusion is increasingly likely.
Parameter C0 C1 C2 C3 C4 C5
Value 2.55 -22.1 -9.13 × 10
-3
57.7 3.81 × 10
-6
6.78 × 10
-2

Table 3-6. Two-dimensional polynomial fitting model parameters for the flow contour maps
A parametric infusion simulation study was conducted to construct 3D and 2D flow
contour maps (Figures 3-8a and 3-8b) that show the maximum flow distance varying with resin
age (0 < α0 < 0.10) and infusion temperature (110 °C < Tinfusion < 150 °C). The flow surface maps
were fit using a two-dimensional polynomial model to enable parametric analysis and process
optimization for select temperatures and part geometries (R
2
= 0.99):
75
𝐷 𝑖𝑛𝑓𝑢𝑠𝑖𝑜𝑛 = 𝐶 0
+ 𝐶 1
𝛼 0
+ 𝐶 2
𝑇 𝑖𝑛𝑓𝑢𝑠𝑖𝑜𝑛 + 𝐶 3
𝛼 0
2
+ 𝐶 4
𝑇 𝑖𝑛𝑓𝑢𝑠𝑖𝑜𝑛 2
+ 𝐶 5
𝛼 0
𝑇 𝑖𝑛𝑓𝑢𝑠𝑖𝑜𝑛 (3-9)
where Dinfusion is the maximum flow distance and Ci (i = 0 ~ 5) are the model fitting parameters.
The values of the model parameters are summarized in Table 3-6.

Figure 3-8. (a) 3D and (b) 2D flow contour maps that show the maximum possible flow distance
as a function of both resin age (α0) and infusion temperature (Tinfusion)
76
Note that with increasing α0 and Tinfusion, the maximum flow distance decreases, yielding
a narrower infusion process window and requiring more careful process design. The flow contour
maps can be used to guide selection of process parameters (e.g., infusion temperature, as well as
the number and locations of resin inlets) for parts of different size and geometry. For example,
when infusing resin into a preform of 1.1 m in length, any parameter set falling left of the red
dashed line in Figure 3-8b (maximum flow distance iso-contour of 1.1 m) would assure complete
saturation. Aged resin with life that exceeds the x-intercept value of the 1.1 m iso-contour will
require the use of multiple resin inlets to saturate the preform. When an infusion process requires
multiple resin inlets, the same contour maps can be used to guide the placement and number of
resin inlets. For SCRIMP, similar flow contour maps can be constructed using three-dimensional
flow simulations that account for flow distribution media.

Figure 3-9. Methodology of material characterization and infusion process adjustment for VI
process
77
VI Design
Guidelines
The competing effects of temperature on resin viscosity during heated
infusion:
• As Tinfusion increases, η0 decreases, expediting the infusion flow.
• Yet, higher Tinfusion also accelerates increase in η, resulting in more rapid
deceleration of resin infiltration.
• For an infusion process predicted to saturate before or shortly after
reaching the η cross-over points, infusing at higher temperature would
expedite infusion.
• Otherwise, infusing at lower temperature would help prevent premature
resin gelation before infusion completion.
Selection of infusion process parameters based on resin age and part
geometry:
• For small parts of simple geometry, infusion is likely to be completed
before η evolves significantly, favoring higher Tinfusion to reduce fill times.
• Larger, more complex parts (requiring greater fill time) can benefit from
lower Tinfusion to decelerate η evolution and prevent premature resin
gelation.
• With increasing α0, the infusion process window narrows, requiring more
careful process design. The flow contour maps can be used to guide
selection of infusion process parameters (e.g., temperature and
number/placement of resin inlets), especially for an infusion process
requiring multiple resin inlets.
Table 3-7. Design guidelines for VI heated filling process
The methodology employed in this work can be used to guide effective VI process
diagnostics and process adjustments, and is depicted in Figure 3-9. First, coupling resin cure
models to in situ dielectric analysis yields a resin cure map (shown earlier in Figure 3-3b). The
cure map is used in conjunction with DCM to assess the physical state or age of resin, which
provides the basis for process adjustments. Using material characterization results, an age-adjusted
infusion process map is constructed (Figure 3-5) to define a nominal process window and process
metrics. Then, the heated filling simulation (Figure 3-7) and flow contour maps (Figure 3-8) are
used to validate and refine the process map, and guide selection of key process parameters for VI
78
of aged resin, based on part size and geometry. Overall, deployment of these tools in conjunction
(in situ process diagnostics, a resin cure map, an infusion process map, process simulation, and
flow contour maps) provides a blueprint for more efficient use of aged resin and intelligent
adjustment of process parameters for VI. The same approach and blueprint can be implemented
for other composites manufacturing processes, including RTM or prepreg-based processes. The
design guidelines for VI heated filling are summarized in Table 3-7.
3.5 Conclusions
This study outlines material characterization and infusion guidelines for more efficient use
of aged resin and corresponding process adjustments. The life of VI resin can be monitored
dynamically and assessed using DCM and a resin cure map. DCM enables in situ measurement of
resin ion viscosity, which can directly be converted into a measure of resin life (or cure
advancement) using the cure map and ion viscosity model. The infusion process map provides a
nominal infusion window and key process metrics adjusted for infusion of aged resin, while the
filling simulation is used to validate and refine the process map and adjust process parameters
accordingly. Furthermore, process simulations and the flow contour maps can be used to determine
VI process parameters (e.g., temperature or resin inlet) for parts of different size and geometry.
Use of aged resin for VI requires careful selection of process parameters because of often
lengthy infusion times. Thus far, DEA in composites manufacturing has been employed primarily
for in situ process monitoring or validation of ex situ thermal analysis results, but less often for
guiding process adjustments. In this work, in situ process diagnostics coupled with cure modeling
yielded a tool for resin life assessment, as well as a process map that together accounted for resin
age. Deployment of these tools enables more efficient use of aged resin and effective process
79
adjustments for VI. For infusion of aged resin, small and simple parts can benefit from higher
infusion temperatures, because infusion can be expected to complete before the resin viscosity
approaches the critical threshold value. In contrast, parts requiring longer fill times can experience
resin gelation before completing infusion due to the fast-evolving degree of cure at higher
temperature. Based on these principles, the filling process possibly can be adjusted either by
reducing the process temperature after reaching the η cross-over point, or by imposing a
temperature gradient (or at least multiple temperature zones) across the infusion length. However,
lowering temperature can also cause resin viscosity to increase, and thus requires balancing the
trade-off between increases in infusion window and viscosity.
In principle, the methodology described here can be extended to other composites
manufacturing processes, including RTM or even conventional prepreg processing. Prepregs
consist of B-staged resin pre-impregnated into fiber beds, and generally feature longer shelf-life
and out-life compared to liquid molding resins. However, lay-up process often exposes prepregs
to substantially longer out-time, causing advances in resin degree of cure [103,104]. Using the
approaches described here, resin cure maps and process maps can be developed for prepreg-based
processes, which can be adjusted by modifying cure cycles based on process simulations or
analysis of key process metrics (e.g., effective flow number [105]). In addition, use of aged resin
(α0 > 0.10 or even greater) may become viable for RTM, in which resin is injected more rapidly
(compared to VI) under positive pressure.
Overall, this work describes a pathway to mitigate the problem of material waste in
composites manufacturing. Material waste can be reduced by monitoring resin life, constructing a
cure and process map, then employing filling simulations and flow contour maps to adjust key
process parameters. However, the current infusion map assumes that the process temperature
80
remains consistent throughout infusion, yielding a nominal process window that is wider than a
practical or effective window. For more accurate prediction of the infusion window, a process map
that accounts for the reaction exotherm must be developed. Moreover, the effects of η0 and overall
η profile on impregnation or part quality should also be assessed. This study demonstrates how to
evaluate and modify process conditions primarily in terms of process (or fill) time, yet different η
history during infusion can also affect dry spot or void formation. Further work is required to
address other potential processing concerns, including cure exotherm-adjusted process map and
effects of η history on part quality, and to refine the process guidelines described here.
81
Chapter 4. Thermoplastic Prepreg with Partially Polymerized Matrix:
Material and Process Development for Efficient Part Manufacturing
4.1 Abstract
Extensive use of thermoplastic composites has been restricted by processing challenges
emerging from high melt viscosities. We demonstrate the feasibility of thermoplastic prepreg with
partially polymerized poly(methyl methacrylate) (PMMA) matrix and carbon fiber reinforcement.
The low viscosity pre-polymer resin allowed part consolidation at low temperature and pressure.
The chemical kinetics and rheology of PMMA polymerization were characterized, and aging study
was conducted to assess pre-polymer stability. Prepregs were fabricated using lab-scale methods,
and fabrication map was constructed to determine the optimal extent of polymerization for the
prepreg. The prepregs were tested for tack and drape at ambient temperature, and thermoformed
for microstructural and chemical analysis. The results show that auto-acceleration drives both rate
of polymerization and viscosity evolution, while refrigeration delays pre-polymer out-time effects.
The thermoformed laminates exhibited near-zero porosity. This work establishes material and
process development guidelines for reactive thermoplastic prepreg, and highlights potential
advantages of the proposed prepreg.
4.2 Introduction
Continuous fiber reinforced thermoplastics (CFRTPs) are less technically mature and less
widely used than thermoset composites [106]. However, interest in CFRTPs has grown rapidly
across multiple industries including automotive, aerospace, and sporting goods, because of
intrinsic advantages over thermoset counterparts [107–110]. For one, thermoplastics are inherently
82
tougher and less brittle than thermosets, resulting in composites with greater impact resistance
[111]. Most CFRTP product forms have unlimited shelf-life, because the matrices are fully
polymerized before part consolidation, allowing room temperature storage [112]. Melt processing
of thermoplastic composites consists of imposing only physical changes to the matrix (melting and
solidifying) rather than time-consuming chemical reactions. Hence, production cycle times for
CFRTPs can be much shorter than those for thermoset composites [111,113]. Finally, thermoplastic
matrices can be re-processed and re-formed upon heating, which allows CFRTPs to be recycled
[114] and welded [113,115,116].
Despite these advantages, melt processing of CFRTP retains one critical challenge. The
manufacture of CFRTP structural parts with competitive mechanical and physical properties
requires the thermoplastic matrix to possess high molecular weight prior to processing, which
increases both the melt viscosity and the melting temperature of the matrix [117–119]. To ensure
proper impregnation of fiber beds and void removal, high-performance CFRTP parts must
therefore be processed at high pressure and temperature [108,111]. Typically, the processing
temperature of CFRTP must exceed the melting (for semi-crystalline polymers) or glass transition
(for amorphous polymers) temperatures of the matrix to supply the polymer chains with enough
energy for sufficient matrix flow [62,108,115]. Consequently, costly investments in infrastructure
and operations are required for CFRTP melt processing [113,120].
Recently, reactive resin transfer molding of CFRTP has been developed as an alternative
to melt processing [110,111,121]. During reactive liquid molding, low viscosity polymeric
precursor (monomer or pre-polymer) is injected into the tool instead of fully polymerized high
viscosity thermoplastic melt, thereby facilitating complete saturation of the fiber preform.
Subsequently, the precursor is polymerized in situ in the mold. This process eliminates the need
83
for high processing temperature and pressure, because the low viscosity pre-polymer can readily
impregnate the fiber bed [111,120]. In addition, the processing temperature of reactive CFRTP
processing is typically less than that of melt processing, because the temperature required to
polymerize the resin is generally less than the melting temperature of the resulting polymer
[111,122]. Engineering thermoplastics with well-developed reactive liquid molding schemes
include polyamide 6 [120–123], polyamide 12 [109,119,124,125], poly(butylene terephthalate)
[110,118], and poly(methyl methacrylate) (PMMA) [126,127]. However, the thermoplastic species
that can be liquid molded are limited by the crucial constraints imposed by the nature of in situ
polymerization and liquid molding itself [109,110,118].
Here, we investigate and demonstrate the feasibility of reactive processing of CFRTP
prepregs. A conventional thermoplastic prepreg includes a fully polymerized matrix before part
consolidation, whereas resin in a typical thermoset prepreg is B-staged (partially cured) to facilitate
prepreg handling and prevent excessive resin flow during part consolidation. Fully polymerized
CFRTP prepregs can be stored indefinitely at ambient conditions, but consolidation normally
requires high processing temperature and pressure. Moreover, CFRTP prepregs lack tack and
drapability at room temperature, both of which are important material characteristics. To address
these issues, we fabricated a thermoplastic prepreg with partially polymerized matrix by fully
impregnating plies of carbon fiber reinforcement with low viscosity monomer (methyl
methacrylate, MMA), then polymerizing the resin to an intermediate molecular mass state. The
objective of this case study was to create a CFRTP prepreg that not only provides tack and drape
at room temperature for greater conformability, but also facilitates resin flow during consolidation
below the final glass transition temperature of the amorphous PMMA thermoplastic matrix.
PMMA was used in this study as a technical pathfinder to identify opportunities and challenges
84
associated with reactive processing of thermoplastic prepregs.
First, the chemical kinetics and rheology of neat PMMA polymerization were examined
using differential scanning calorimetry (DSC) and rheometry. The aging of PMMA pre-polymer
was investigated for three different storage temperatures to gain insights into the effects of out-
time accrual and of refrigeration on pre-polymer resin aging. Two-ply PMMA pre-polymer prepreg
laminates were fabricated to construct a fabrication map to determine the optimal extent of
polymerization and monomer/fabric weight ratio for the prepreg. The prepregs were evaluated for
tack and drape at ambient temperature, and thermoformed at low pressure, below the final glass
transition temperature of PMMA, for porosity analysis. The chemical composition of the final
matrix of the consolidated laminate was analyzed using Fourier-transform infrared spectroscopy
(FTIR). To address scale-up challenges, an eight-ply prepreg laminate was produced and
thermoformed for microstructural analysis. Finally, a conventional thermoplastic prepreg with
fully polymerized matrix was also fabricated to explore matrix flow under various thermoforming
conditions and to demonstrate that forming of conventional CFRTP prepreg comprised of the same
material is substantially more resource-intensive.
4.3 Experimental
4.3.1 Materials
Methyl methacrylate (MMA) monomer and benzoyl peroxide free-radical initiator
(Luperox
®
A98) were acquired from Sigma-Aldrich. Under ambient conditions, MMA is a
colorless transparent liquid with low viscosity (0.53 mPa ⋅s) and 100 °C boiling point [128]. For
both polymerization characterization and prepreg laminate fabrication, 3.0 weight % of benzoyl
peroxide was dissolved in MMA, and the resulting monomer/initiator mixture was stored under
85
refrigeration (4 °C) to prevent polymerization of the material before application. For the prepreg
reinforcement, a plain weave carbon fiber fabric with an areal weight of 193 g/m
2
and 3000
fiber/tow count was obtained from Fibre Glast (Part # 1530 [57]).
4.3.2 Polymerization & Resin Characterization
4.3.2.1 Modulated Differential Scanning Calorimetry (MDSC)
The reaction kinetics of PMMA polymerization was characterized using DSC (TA
Instruments, Q2000) under nitrogen purge (50 cm
3
/min). Monomer samples (7-10 mg) were sealed
in aluminum hermetic pans with lids and exposed to an isothermal dwell at four different
temperatures (60, 70, 80, and 90 °C) for prescribed durations (1-4 hours). The samples were
subsequently heated to 200 °C at 2 °C/min to measure the remaining heat of polymerization and
determine the amount of residual monomer. For the ramping step, sinusoidal temperature
modulation of ±0.5 °C/min was applied over the linear temperature ramp to distinguish reversing
and non-reversing heat flow signals. Reversing heat flow is related to heat capacity and rate of
temperature change, while non-reversing heat flow depends on kinetic components of the reaction
[60,61,129]. Glass transition is a heating rate dependent transition and can be examined from the
reversing heat flow signal, while polymerization is a temperature dependent transition and thus
appears in the non-reversing signal [60,129].
4.3.2.2 Rheology
The viscosity evolution during PMMA bulk polymerization was measured using a
rheometer (TA Instruments, AR2000ex), equipped with a Peltier plate and cone-and-plate fixture
(cone angle 2:0:5, cone diameter 40 mm, and truncation 53 µm). The cone-and-plate geometry,
86
which provides uniform shear rate and high rheological accuracy, was selected because the
viscosity profile of PMMA polymerization was expected to span a wide range, from low viscosity
monomer to high viscosity polymer. A solvent trap cover was used to suppress evaporation of
highly volatile MMA monomer. To confine enough MMA at the center of the Peltier plate without
leakage, a circular side wall was fabricated with silicone vacuum bag sealant (A-800, General
Sealants). For the rheology test, cell temperature was increased from room temperature to 90 °C
at approximately 32.5 °C/min (the maximum ramp rate of the rheometer) and held constant for 1
h, under oscillatory shear at 1 Hz frequency and 0.01 % strain (within the linear viscoelastic region).
4.3.2.3 Pre-Polymer Aging Study
PMMA pre-polymer resin samples of known initial degree of monomer conversion were
prepared by reacting MMA monomer at 90 °C for 10 minutes using DSC. The samples were then
removed from the DSC and stored at three different temperatures: 25 °C, 4 °C, and -18 °C. For
each storage temperature, the samples were aged for 1, 2, 4, 8, 12 hours and 1, 2, 7, 14, 28 days.
Once aging was complete, each sample was placed back in the DSC and heated to 220 °C at
2 °C/min to determine the extent of additional polymerization accrued during aging. All samples
were weighed before and after aging to confirm that no weight loss had occurred. The purpose of
this study was to monitor accrual of pre-polymer resin out-time and determine if refrigeration is
required to prevent or delay PMMA pre-polymer aging.
87
4.3.3 Prepreg Laminate Fabrication & Characterization
4.3.3.1 Prepreg Laminate Fabrication

Figure 4-1. (a) Aluminum tool for prepreg laminate fabrication (top cover & bottom container),
and (b) sample two-ply PMMA pre-polymer prepreg laminate (P-1)
For prepreg fabrication, an aluminum tool (50 mm × 50 mm) consisting of top cover and
bottom container was designed (Figure 4-1a). The top cover was intended to prevent excessive
monomer vaporization at elevated temperature during pre-polymerization. Mold release agent (20-
8185, Buehler) was first applied to the tool interior and allowed to dry for 5 minutes. Then, the
mold was preheated to 90 °C using a hot plate, and two plies of carbon fiber fabric were placed in
the tool with the bottom side covered by non-perforated release film (A4000, Airtech).
For this study, two-ply prepreg laminates were assembled to facilitate handling of thin
samples. Once the tool temperature reached 90 °C, a known amount of MMA was injected into the
tool using a pipette, based on initial trials, and the top surface of the fabric was immediately
enclosed by another ply of release film. The low viscosity of MMA ensured complete saturation
of the fabric, which could be readily observed. Then, the polymerization proceeded for a prescribed
amount of time, determined from DSC and rheology analysis, yielding a partially polymerized
88
prepreg laminate. The test matrix is shown in Table 4-1.
Sample Monomer/Fabric Weight Ratio Polymerization Time
1
4.7
15 min
2 16 min
3
5.2
12 min
4 13 min
5 14 min
6 15 min
7 15.5 min
8 16 min
9 16.5 min
10
5.6
15 min
11 15.5 min
12 16 min
Table 4-1. Test matrix for two-ply PMMA pre-polymer prepreg laminate fabrication
Using similar fabrication methods, an eight-ply prepreg laminate was also fabricated to
explore the possibility that the approach could be scaled up. Eight plies of carbon fiber fabric were
fully impregnated by MMA, which was partially polymerized to pre-polymer state. The image of
a sample two-ply partially polymerized prepreg (P-1) is shown in Figure 4-1b. The proposed
processing technique can be used to fabricate both single- and multi-ply prepregs depending on
the end use. For manufacturing parts with simple geometries, a single multi-ply prepreg laminate
can be prepared for easy and fast lay-up. Conversely, for complex-shaped parts that cannot be
formed from a single thick prepreg laminate, multiple single-ply prepregs can be produced for lay-
up.
89
4.3.3.2 Prepreg Conformability Evaluation
Prepreg tack was measured using a rheometer (TA Instruments, AR2000ex) with parallel-
plate geometry. A Peltier plate was used as the bottom plate, and a 25 mm disposable disk as the
top plate. The prepreg samples of various out-times (from 0 to 60 min, in increments of 10 min)
were placed on the bottom plate at ambient temperature, and compressed at 0.10 N/mm
2
load for
10 s to ensure enough contact time between the samples and the top plate. After equilibration, the
top plate was pulled away at 0.1 mm/s, and the sample tack was defined, in first approximation, as
the maximum normal pressure measured by the load cell during top plate retraction [130]. For
comparison, the tack of toughened epoxy resin prepreg (CYCOM
®
5320-1, Cytec) was also
measured. The goal of the tack tests was not to determine the absolute values of tack for the prepreg,
but to provide a metric for comparing different prepreg embodiments and tracking with out-time.
To assess the drapability, the prepregs were laid up onto 75° sharp corner mold at room
temperature. Two prepreg samples with partially polymerized matrix (D-1 and D-2) and one with
fully polymerized matrix (F-1) were prepared. D-1 was draped over the mold right after fabrication,
while the D-2 was draped after 1 hour out-time. The objective of the drape evaluation was to
compare the room temperature pliability and drapability of partially polymerized prepreg to that
of a fully polymerized one.
4.3.4 Laminate Thermoforming & Analysis
The two-ply and eight-ply PMMA pre-polymer prepreg laminates were thermoformed
with hydraulic press (Genesis G30, Wabash) at 90 °C and 0.38 MPa for 5 min. For microstructural
analysis, the center of each laminate was sectioned, encapsulated with epoxy mounting resin, and
ground and polished. Images of each cross-section were recorded using a digital stereo microscope
90
(VHX-5000, Keyence).
The chemical composition of the matrix of the thermoformed laminate was investigated
with Fourier-transform infrared spectroscopy (FTIR, Thermo Electron, Nicolet 4700). The
collected FTIR spectrum of the final matrix was interpreted and compared to PMMA reference IR
spectra [131–134] to demonstrate clean polymerization of PMMA resin throughout the prepreg
fabrication and thermoforming.
4.3.5 Fully Polymerized Prepreg Laminate
To fabricate a fully polymerized prepreg laminate for control, two plies of carbon fiber
fabric were fully impregnated with MMA, then polymerized in situ at 90 °C for 30 minutes. The
prepared prepreg (F-1) was thermoformed multiple times using a hot press at different
temperatures (90, 150, and 200 °C) under 0.38 MPa for 5 min, and once more at 200 °C with
tenfold increase in pressure (3.8 MPa) for longer duration (15 min). The processing temperatures
were chosen with respect to the final glass transition temperature of PMMA (115 °C) and typical
forming temperature of conventional PMMA prepreg (200 °C).
91
4.4 Results & Discussion
4.4.1 Polymerization & Resin Characterization
4.4.1.1 Chemical Kinetics

Figure 4-2. MDSC data showing (a) heat flow & temperature profiles during the isothermal
dwell (70 °C) and dynamic ramp steps of MMA free-radical bulk polymerization, and (b)
reversing & non-reversing heat flow profiles during the dynamic ramp step
92
Figure 4-2a shows representative MDSC data of heat flow and temperature profiles during
MMA free-radical bulk polymerization at 70 °C. Under the isothermal dwell, MMA
polymerization first went through a brief induction period, after which the rate of polymerization
slightly decreased until it reached a minimum, as anticipated by the classical free-radical
polymerization theory [135,136]. After this local minimum, the polymerization rate deviated from
the classical theory by rapidly accelerating and then decelerating until the polymerization attained
completion. This auto-acceleration phenomenon is known as the Trommsdorff effect or gel effect
[127,136]. As the polymerization progressed, an increase in local resin viscosity led to a rapid drop
in termination rate and consequently to a surge in overall polymerization rate [127]. In essence,
the produced polymers generated a catalytic effect on the polymerization [135].
During the subsequent dynamic ramp, which was programmed to drive residual monomer
to complete polymerization, two overlapping peaks were detected. To identify the peaks, reversing
and non-reversing heat flow signals from MDSC were analyzed (Figure 4-2b). The glass transition
region in the reversing heat flow signal coincided with the second peak of the non-reversing heat
flow signal, which was therefore associated with enthalpic relaxation during glass transition. As a
result, the first peak corresponded to the heat of residual monomer polymerization.
The DSC data of MMA polymerization were fit to a mathematical kinetics model
developed by Jašo et al [135], which takes into account both classical free-radical polymerization
theory and auto-acceleration:
𝑋 ( 𝜏 ) = ( 𝑋 𝐾 − 𝑎 )∙ ( 1 − e
−𝑘 1
𝜏 )+
𝑎 1 + e
−𝑘 2(
𝜏 −𝜏 2𝑚𝑎𝑥 )
(4-1)
In this model, X (degree of monomer conversion), XK (final degree of monomer conversion), and
τ (time) were obtained from the measured data, while k1 (reaction rate constant for classical free-
93
radical polymerization), k2 (rate constant for auto-acceleration reaction), a (fraction of monomer
polymerized by auto-acceleration), and τ2max (time at maximum rate of reaction) were calculated
using the method of least squares. To calculate the degree of monomer conversion from DSC heat
flow data, the following equation was used [135]:
𝑋 ( 𝜏 ) =
∫ ( 𝑑𝐻 /𝑑𝜏 ) 𝑑𝜏 𝜏 0
∫ ( 𝑑𝐻 /𝑑𝜏 ) 𝑑𝜏 𝜏 𝐾 0
+ 𝐻 𝐷 (4-2)
where H is the heat evolved during polymerization, τK the time at XK, and HD the total heat evolved
during residual monomer polymerization.

Figure 4-3. (a) Measured degree of monomer conversion profile and mathematical kinetics
model fitting, and (b) viscosity profile of MMA polymerization at 90 °C
94
k1
[min
-1
]
k2
[min
-1
]
a
[-]
XK-a
[-]
τ2max
[min]
Auto-Accel. Onset
[min]
0.167 0.911 0.632 0.362 12.2 9.9
Table 4-2. Values of the fitting parameters & derivatives for MMA polymerization at 90 °C
The polymerization temperature of 90 °C was of particular interest, being the highest
chosen temperature that did not exceed the boiling point of the monomer (100 °C at ambient
pressure). At this level, the data showed that the final degree of monomer conversion (XK) was
0.994, and the polymerization essentially ended 19.1 min after the reaction started, when X reached
99% of XK. Figure 4-3a shows the degree of monomer conversion curve and its model fitting for
polymerization conducted at 90 °C. The values of the calculated fitting parameters are shown in
Table 4-2. Although not presented here, the model yielded similar quality of fit with experimental
data at all four polymerization temperatures when appropriate parameters (k1, k2, a, and τ2max) were
determined. Both reaction rate constants, k1 and k2, exhibited strong Arrhenius temperature
dependence, where the coefficients of determination (R
2
) for the Arrhenius linear regressions were
0.99 and 0.98 respectively.
4.4.1.2 Rheology
The viscosity profile for PMMA polymerization at 90 °C was divided into three sections
(Figure 4-3b). The first section (Zone I) corresponded to the first ten minutes of the polymerization,
as the initial resin viscosity (~0.01 Pa ⋅s) increased linearly on a semi-log scale. In Zone II, which
represented the next ten minutes of the reaction, the resin viscosity evolved in a sigmoidal shape.
During this phase, the viscosity increased sharply within a short period, driven by the steep rise in
the rate of polymerization. In the last region (Zone III), the viscosity curve reached a plateau near
95
10
6
Pa ⋅s, marking the end of polymerization.
Based on the model fitting using 90 °C polymerization DSC data, the onset of auto-
acceleration occurred at 9.9 minutes, which coincided with the viscosity phase transition from
Zone I to Zone II. The correlation indicated that Zone I represented the classical free-radical
polymerization phase, while Zone II corresponded to the auto-acceleration reaction step. Moreover,
the DSC analysis showed that MMA polymerization was complete at 19.1 min, when the viscosity
curve became flat. In short, the degree of monomer conversion and viscosity evolution profiles
exhibited similar shapes and trends, especially with respect to the auto-acceleration phenomenon.
The similarity indicates that it would be possible to model evolution of PMMA resin viscosity as
a function of degree of conversion as well as temperature, analogous to some well-established
thermoset cure viscosity models (e.g., Castro-Macosko model [137]).
4.4.1.3 Pre-Polymer Aging Study
The degree of monomer conversion of pre-polymer resin samples polymerized at 90 °C
for 10 min was 0.42. From this state, the samples were aged at room temperature or refrigerated
conditions (4 °C or -18 °C) for periods up to 4 weeks. Finally, the aged samples were heated to
220 °C in the DSC to determine the extent of additional polymerization accrued during aging,
which was calculated using the following equation:
𝑋 𝐴𝑔𝑒𝑑 ( fraction of aged resin) =
𝐻 𝑟 ,𝑟𝑒𝑓𝑒𝑟𝑒𝑛𝑐𝑒 ( J/g)− 𝐻 𝑟 ,𝑎𝑔𝑒𝑑 ( J/g)
𝐻 𝑟 ,𝑡𝑜𝑡𝑎𝑙 ( J/g)
(4-3)
The heat of reaction of the aged pre-polymer resin (Hr,aged) was subtracted from that of the unaged
reference pre-polymer resin (Hr,reference), for which the initial degree of conversion was 0.42
(Hr,reference equal to 0.58×Hr,total). The subtracted value was divided by the total heat of
96
polymerization of PMMA (Hr,total) to determine XAged. The underlying assumption here was that
the degree of conversion was expected to increase as a function of aging time. The data obtained
during the aging study are shown in Figure 4-4.

Figure 4-4. Degree of monomer conversion profiles as a function of aging time for three different
storage temperatures, with sample XAged calculation
The resin samples stored at ambient temperature exhibited rapid aging. The degree of
monomer conversion reached 0.86 after 1 day, advancing toward 0.91 and 0.93 after 1 and 4 weeks
of aging respectively, leaving little room for further polymerization. In contrast, the samples stored
at -18 °C (or 0 °F, the recommended storage temperature for typical thermoset prepregs) showed
much slower aging. The degree of conversion remained below 0.48 after 1 week and 0.57 after 4
weeks, indicating that less than 15% of the pre-polymer resin aged at this temperature. Storing at
4 °C seemed to delay polymerization for the first few days, but the degrees of conversion after 1
week eventually approached those of the samples stored at room temperature. The results of the
97
aging study demonstrate that refrigeration at -18 °C effectively delays out-time accrual of PMMA
pre-polymer resin, but also indicate that this case study material is not as shelf-stable as
commercial fully polymerized CFRTP prepregs.
4.4.2 Prepreg Laminate Fabrication & Characterization
4.4.2.1 Prepreg Laminate Fabrication
After examining polymerization kinetics and pre-polymer aging, prepreg laminates with
partially polymerized matrix were fabricated using lab-scale methods. One technical challenge
arose during fabrication: control of pre-polymer resin content through selection of an appropriate
monomer/fabric weight ratio and pre-polymerization time (in the presence of monomer
vaporization and bleed). Applying the insights gained from analysis of the chemical kinetics and
rheology, we predicted that polymerizing the resin to the onset of auto-acceleration would yield an
extent of polymerization acceptable for initial prepreg fabrication trials. This point, obtained from
model fitting with DSC data, served as an acceleration onset for both degree of monomer
conversion and resin viscosity.
When early fabrication attempts near the onset did not produce sufficient conversion and
resin viscosity for the model prepreg, the pre-polymerization time was gradually advanced for
subsequent trials. The kinetics and rheology analysis permitted analytical prediction of the pre-
polymerization time for prepreg fabrication at 90 °C, as well as at other reaction temperatures. For
instance, at 80 °C, the polymerization kinetics model indicated that the onset of auto-acceleration
occurred at 24.1 min, and hence that initial prepreg fabrication trials should be conducted after
24.1 min of pre-polymerization.
98

Figure 4-5. (a) Final resin contents of two-ply prepreg laminates with varying polymerization
times at 90 °C, (b) experimental prepreg fabrication map at 90 °C for two-ply prepreg laminate,
and (c) temperature and time dependent prepreg process map, showing empirical (90 °C) and
estimated (70 & 80 °C) process windows
99
Figure 4-5a shows the final resin contents of two-ply prepreg laminates with varying
polymerization times at 90 °C, where the monomer/fabric ratio was 5.2, a value determined from
initial trials. With increasing reaction time, the degree of conversion and resin viscosity increased
as well, leading to greater amounts of resin retained on the prepreg. Based on the target resin
content for the proposed prepreg, which was 38-42 weight %, an experimental prepreg fabrication
map at 90 °C was constructed (Figure 4-5b) to determine the optimal resin polymerization time
and monomer/fabric weight ratio for the two-ply prepreg laminate.
On the map, the samples on the red points were deemed impractical because they resulted
in either too little or too much resin. Samples on the blue triangles were acceptable, and a single
optimal fabrication point was eventually determined, represented by the green circle near the center.
At this point, the monomer/fabric weight ratio was 5.2, and the polymerization time was 15.5 min.
Higher or lower weight ratios resulted in excess or insufficient resin, respectively. When the
reaction time was less than 15 min, the resin viscosity was unacceptably low, and too much resin
was lost while handling the prepreg. For reaction times exceeding 16 min, the pre-polymer resin
was excessively converted, reducing the benefits of reactive processing.
The chemical kinetics model (Equation 4-1) can be used to predict degree of monomer
conversion (X) as a function of reaction time at a given polymerization temperature. Using this
model and prepreg fabrication trial results at 90 °C (Figure 4-5b), approximate process time
windows at other reaction temperatures can be predicted as well. At 80 °C, for instance, the time
required to achieve the same degree of conversion obtained at 90 °C during 15 and 16 minutes of
reaction is expected to be between 30 and 33 minutes.
Under the assumption that the ideal initial monomer/fabric weight ratio of 5.2 does not
vary greatly with respect to temperature, a prepreg process map that depends both on temperature
100
and time was constructed to show empirical (90 °C) and estimated (70 & 80 °C) fabrication time
windows (Figure 4-5c). These results, along with the onset of auto-acceleration data, can provide
reasonable time frames for initial prepreg fabrication trials at different temperatures. The optimal
processing time span widens with decreasing polymerization temperature as X increases more
slowly. Therefore, degree of conversion and thus viscosity of pre-polymer resin can be more
precisely controlled at lower temperatures. On the other hand, as reaction temperature rises, the
required polymerization time becomes much shorter, yet controlling X significantly more
challenging.

Figure 4-6. Methodology of material and process development for the design of reactive CFRTP
prepreg
The methodology employed here for material characterization and process development
of PMMA pre-polymer prepreg, can similarly be applied to design other reactive CFRTP prepregs
(Figure 4-6). For a given prepreg system, kinetics and viscosity evolution models for thermoplastic
polymerization can be used to predict an extent of polymerization acceptable for initial prepreg
fabrication trials, and to establish a pre-polymerization temperature profile for materials that
require complex reaction schemes. Finally, a prepreg fabrication map can be constructed to identify
101
suitable fabrication conditions for the prepreg.
4.4.2.2 Prepreg Tack Characterization

Figure 4-7. (a) Normal pressure against the top plate retraction time for prepregs with varying
out-times, and (b) maximum force required to detach the top plate, or the sample tack
To evaluate tack, we investigated the normal force required to detach the top plate from
the prepreg sample at ambient temperature. Tack tests yielded force-per-area peaks (Figure 4-7a),
where the normal pressure returned to zero level within 1-2 seconds at the given retraction speed.
The evolution of normal force indicated that the prepreg failed adhesively [130]. On the other hand,
epoxy prepreg (CYCOM
®
5320-1) exhibited more of a cohesive failure, where the normal pressure
degraded gradually over a longer period upon hitting the apex.
The maximum pressure measured during the top plate retraction, against out-time, is
shown in Figure 4-7b. With increasing out-time, the maximum pressure or sample tack decreased
linearly, and the fabricated prepreg completely lost tack after 1 hour of ambient exposure. The goal
of tack characterization was to verify that the prepreg possessed tack after fabrication, and to assess
the evolution of tack with increasing out-time. The results showed that the prepreg exhibited tack
at room temperature, which could potentially improve handling of the prepreg during laminate lay-
102
up, but that for the current embodiment, the evolution of the PMMA pre-polymer resin was
appreciably rapid.
4.4.2.3 Prepreg Drape Characterization

Figure 4-8. (a) Partially polymerized thermoplastic prepreg draped onto 75° corner mold right
after fabrication, and (b) conventional thermoplastic prepreg with fully polymerized matrix
The lay-up of the prepregs with partially polymerized matrix (D-1 and D-2) was
straightforward, because the samples were highly pliable and shearable under ambient condition
(Figure 4-8a). Still, D-2, the laminate with longer ambient exposure, exhibited reduced flexibility,
indicating that the sample stiffness increased with advancing out-time. The same sample also lost
most of its tack after 1 h of out-time (Figure 4-7b), but applying a small amount of liquid MMA to
the prepreg surface restored sample tack to sustain its adhesion to the mold surface.
In contrast, F-1, in which the matrix was fully polymerized, was rigid at room temperature,
and was impossible to drape onto the corner mold without pre-heating (Figure 4-8b). The results
of the comparative analysis demonstrate that, unlike conventional thermoplastic prepreg, partially
polymerized prepreg is pliable and can be readily draped onto complex geometries under ambient
103
conditions. While the drape evaluation was qualitative in nature, the results were sufficiently
different to indicate that partially polymerized thermoplastic prepreg offers handleability
advantages.
4.4.3 Laminate Thermoforming & Analysis
4.4.3.1 Microstructural Analysis

Figure 4-9. Cross-sectional micrographs of thermoformed (a) 2-ply (P-1) and (b) 8-ply (TH-1)
laminates, free of voids
The prepregs with partially polymerized matrix were thermoformed at 90 °C and 0.38
MPa for 5 min. The processing temperature (90 °C) was lower than the typical forming temperature
104
of conventional PMMA prepreg, which is near 200 °C [138,139], and was even below the glass
transition temperature of fully polymerized PMMA, which is 115 °C. Here, the applied pressure
of 0.38 MPa was the minimum applicable pressure of the hydraulic press, but remained well below
the recommended thermoforming pressure for commercial PMMA prepreg (0.5-2.0 MPa [139]a).
The cross-sectional images of the thermoformed two-ply laminate (P-1) are shown in
Figure 4-9a. The micrographs demonstrate that the laminate is essentially void-free. During
prepreg fabrication, the fiber bed was completely impregnated with pre-polymer matrix because
low viscosity monomer fully saturated both macro- and micro-pores in the dual-scale fabric, and
afterward polymerized in situ. Moreover, the pre-polymer matrix, with melt viscosity much lower
than that of completely polymerized polymer, enhanced resin flow during consolidation, even at
low temperature and pressure. Hence, the resin effectively impregnated all dry fiber tows, leaving
no resin-deprived regions.
The free-radical bulk polymerization of PMMA is highly exothermic—the heat of
polymerization is nearly three times greater than that of epoxy resin [127]. This high reaction
exotherm, combined with the auto-acceleration effect, can lead to thermal runaway and increase
the reaction temperature above the boiling point of MMA [111]. Thus, reactive processing of
PMMA at high temperature can result in porosity in the final part due to monomer vaporization
during polymerization [111,127]. However, for fabrication of the PMMA pre-polymer prepreg,
where the polymerization temporarily ceases at an intermediate stage before the reaction
temperature rises sharply, the number of bubbles entrapped in the matrix is greatly reduced. If the
prepreg contains porosity, the low viscosity of the pre-polymer allows removal of gas-induced
voids during part consolidation by in-plane resin bleed-out, even under low pressure. Within the
test samples, nearly all bubbles were situated at laminate edges, which are typically trimmed.
105
To verify validity of these results for thicker parts, the eight-ply prepreg laminate (TH-1)
was thermoformed under identical forming conditions (90 °C and 0.38 MPa for 5 min). Cross-
sectional images in Figure 4-9b show that this thicker laminate is also free of resin-deprived
regions or entrapped bubbles, indicating complete fiber bed saturation and void removal.
4.4.3.2 Chemical Composition Analysis

Figure 4-10. FTIR absorbance spectrum collected from the final thermoformed PMMA matrix
The chemical composition of the thermoformed PMMA matrix was analyzed using FTIR
to confirm that the interrupted polymerization process did not cause adverse effects. The
absorbance spectrum collected from the final matrix, shown in Figure 4-10, exhibited bands
characteristic of PMMA. Some example bands include methyl C-H stretch at 2993 & 2949 cm
-1
,
C=O stretch at 1722 cm
-1
, methylene C-H bend at 1479 cm
-1
, methyl C-H bend at 1435 & 1387
106
cm
-1
, and C-O-C stretch at 1190 & 1142 cm
-1
. In addition, the absence of C=C stretch near 1620–
1680 cm
-1
and alkenyl C-H stretch at 3000–3150 cm
-1
—two characteristic bands of the monomer,
methyl methacrylate—indicate that the matrix resin underwent complete and clean PMMA
polymerization.
4.4.3.3 Fully Polymerized Prepreg Laminate

Figure 4-11. Conventional PMMA prepreg with fully polymerized matrix, before and after
thermoforming at different temperatures—90, 150, and 200 °C
107
A conventional PMMA prepreg with fully polymerized matrix was fabricated to examine
matrix flow under selected thermoforming conditions. Sample (F-1) was not fabricated using the
standard industrial manufacturing protocol of CFRTP. Rather, it was intentionally fabricated to
produce high porosity to better observe matrix flows at different temperatures. Again, the final
glass transition and forming temperatures of PMMA were 115 and 200 °C, while the pre-polymer
prepreg was thermoformable at 90 °C.
Figure 4-11 shows images of sample F-1 before and after multiple thermoforming under
various conditions. When thermoformed at 90 °C, the fully polymerized prepreg exhibited no
matrix flow. At 150 °C, which is greater than Tg but less than Tforming, minimal matrix bleed-out
was observed at laminate edges, which drove some of the entrapped bubbles out. Full matrix flow
was observed at 200 °C and 3.8 MPa, conditions that were much more resource-intensive than
those used to form the pre-polymer prepregs. The purpose of this comparative analysis was to
demonstrate that, unlike partially polymerized prepreg, conventional thermoplastic prepreg must
be heated well above the melting or glass transition temperature of the matrix to achieve sufficient
flow, and become flexible and thermoformable.
In this work, laminates of relatively small lab-scale size (0.003 m
2
area) were
manufactured to demonstrate the feasibility of partially polymerized thermoplastic prepreg.
Consolidating parts with greater in-plane dimensions may require more energy-intensive
processing conditions for effective removal of voids near the laminate center if the principal void
removal mechanism remains in-plane resin bleed-out. Conversely, if consolidation phenomena
(flow and compaction) can be designed to favor some fluid pressure retention during molding,
voids could be suppressed by conventional application of pressure. Generally, the low viscosity of
pre-polymer resin will allow part consolidation under moderate processing temperature and
108
pressure, well below those required for producing laminates from fully polymerized prepreg of
comparable size.
4.5 Conclusions
The chemical kinetics and rheology of PMMA polymerization were analyzed to design
thermoplastic prepreg laminate with partially polymerized matrix. The auto-acceleration
phenomenon from the Trommsdorff effect governed the rate of polymerization and viscosity
evolutions of PMMA free-radical polymerization. Aging tests demonstrated that refrigerated
storage of PMMA pre-polymer delays conversion due to out-time. Prepreg laminates were
fabricated using lab-scale methods, and a fabrication map was constructed for two-ply prepreg
laminates to determine the optimal extent of polymerization. The prepregs were characterized for
tack and drape at room temperature, and then thermoformed below the final glass transition
temperature of PMMA for microstructural and chemical analysis. Both two- and eight-ply
laminates showed no sign of porosity.
The objective of this case study was to establish material characterization and process
development guidelines for reactive processing of CFRTP prepreg, and to demonstrate potential
advantages of thermoplastic prepreg with a partially polymerized matrix. Conventional
thermoplastic prepreg generally features a fully polymerized matrix with high melt viscosity, and
processing requires high pressure and temperature to ensure proper fiber bed impregnation and
void removal. On the other hand, the prototype prepreg laminate described here consists of
multiple plies of fabric fully impregnated with a pre-polymer matrix, whose melt viscosity is much
less than that of the fully polymerized material. The pre-polymer matrix facilitates resin flow
during part consolidation even at low temperature and pressure. Therefore, any voids created
109
throughout part processing are readily eliminated, with no visible flow- or gas-induced voids
remaining in the final consolidated laminate. Moreover, the pre-polymer resin also provides tack
and drape, which can accommodate prepreg conformability.
In this study, we focused on material characterization and process development for the
most basic MMA polymerization case to determine the feasibility of partially polymerized
thermoplastic prepreg. However, varying initiator weight content or mixing additional ingredients
such as PMMA polymer or comonomers into the monomer/initiator mixture can significantly alter
reaction kinetics and ultimately final degree of polymerization, both of which can affect
consolidation phenomena and mechanical properties of fabricated parts. Flow, compaction, and
(after gelation) residual stress formation can also influence process-induced deformation, which
were not studied here but can affect the shape conformity of the part.
Further work will be required to examine the effects of various reactant components on
polymerization kinetics and part properties, and to address other processing challenges identified
during this study. For example, storage of thermoplastic pre-polymer prepreg requires refrigeration
to delay out-time, similar to thermoset prepreg. In addition, the duration of tack and drape of the
model PMMA prepreg, while superior to that of conventional CFRTP prepreg, which possesses
neither tack nor drape, may be insufficient for complex part fabrication. These limitations can be
addressed, in principle, by adopting alternative aerospace-grade thermoplastic matrices with
higher melting temperatures, such as polyetheretherketone (PEEK), polyetherimide (PEI), or
polycarbonate (PC). Such polymers are expected to extend out-life, and tack and drape lives to
more practical levels.
Overall, this work describes a pathway to reduce major processing challenges associated
with CFRTPs, including the handleability of thermoplastic prepregs, and the thermal and pressure
110
conditions required for part consolidation. Composites manufacturing is undergoing a gradual shift
from legacy methods (e.g., autoclave cure of thermoset prepregs) to processes and materials that
enable simpler and more cost-effective fabrication of structural parts. This pathfinder study
indicates that partially polymerized thermoplastic prepregs may offer a viable solution toward such
advances in CFRTPs.
111
Chapter 5. Conclusions and Future Work
In this dissertation, the major processing challenges in OoA composites manufacturing
were addressed using material characterization, in situ process diagnostics, process modeling, and
process simulation. This chapter summarizes the main conclusions and findings obtained from
both VI process projects (Chapters 2 and 3) and CFRTP prepreg process project (Chapter 4). We
also provide broader technical implications associated with the presented work. Finally, we
recommend additional future work to potentially enhance and refine the composites process
guidelines and methodologies suggested in this work.
5.1 Conclusions & Contributions
5.1.1 Vacuum Infusion Projects
First, the following process guide and optimization tools for the VI process were
developed and deployed effectively:
1. Resin cure maps (Figures 2-4 and 3-3) enabled simple and accurate in situ monitoring of
the physical state of resin using dielectric analysis (DEA) coupled with cure modeling.
Online DEA measurement data (in ion viscosity or ion conductivity) were first correlated
with phenomenological cure kinetics and mechanical viscosity models. Then, using an ion
viscosity model, degree-of-cure and mechanical viscosity isolines were constructed as a
function of temperature. The straightforward linear relationship between the ion viscosity
model parameters and resin degree of cure allowed accurate prediction of ion viscosity
across a wide span of resin age and temperature, while minimizing the need for extensive
112
material characterization generally required for DEA modeling. Using this tool, we can
identify the specific cure state of interest during actual part manufacture (e.g., gel point)
or develop a program that yields resin degree of cure at the specified temperature or level
of ion viscosity. The resulting metrics for resin life could then be employed for further
process adjustments for VI process.
2. Infusion process maps (Figures 3-5 and 3-6) were constructed using cure kinetics and
mechanical viscosity models. The resulting process maps yielded age-adjusted nominal
infusion process windows as well as the key process metrics for the VI heated filling
process. The competing effects of process temperature on infusion flow were demonstrated
using the process maps, and infusion process simulation was conducted to validate and
refine the infusion maps. Through a parametric simulation study, flow contour maps
(Figure 3-8) were designed, which showed the maximum resin flow distance with varying
resin age and infusion temperature. With increasing resin age and infusion temperature,
the maximum resin flow distance decreased, resulting in a narrower process window and
requiring a more careful process design. The flow maps could also be used to guide
selection of VI process parameters (e.g., infusion temperature and the number or locations
of resin inlets) for parts of different size and geometry.
Second, in Chapter 2, a low-temperature post-infusion (LTPI) dwell was added to the
conventional VI cure cycle to address the problem of high void content in thermoplastic veil-
toughened thermoset composites. The study determined and demonstrated the effectiveness of a
post-infusion dwell in enhancing part quality of a multi-component laminate produced using VI
process. Interlaminar thermoplastic veils basically served two functions: (1) enhanced impact
resistance and (b) inter-ply flow distribution media. Yet, the latter function induced local variations
113
in preform permeability and non-uniform flow fronts during resin infusion, resulting in high
porosity in the final part. The addition of LTPI dwell effectively postponed resin gelation and
extended VI post-filling stage, providing greater time for the liquid resin to reach an equilibrium
state and redistribute to saturate any existing partially filled regions. The process parameters for
the LTPI dwell were initially informed by the material characterization results and cure process
model, which aided resin gel point identification. However, the actual duration of the post-infusion
dwell was determined and adjusted in situ using a dielectric cure monitoring (DCM) system and
the epoxy cure map we developed earlier. Laminated produced using the modified cure cycle
resulted in lower internal porosity and exhibited greater peak load during impact testing.
Finally, the previously explained VI process guide tools were employed to enable more
efficient use of aged resin (Chapter 3). Thus far, DEA in composites manufacturing has been used
primarily for validation of ex situ thermal or rheological analysis results and for simple in situ
process monitoring, but less often for effective online process adjustments. In this study, we
deployed DEA in conjunction with resin cure map, infusion process map, and process simulation
for effective VI process adjustments. The competing effects of infusion temperature—decreasing
initial resin viscosity vs. more rapidly increasing resin viscosity with increasing temperature—on
infusion flow were investigated using infusion process maps. The infusion process simulation
results (PAM-RTM) demonstrated how the various infusion process parameters must be adjusted
based on part size and geometry as well as resin age. For infusion of aged resin, small parts of
simple geometry can benefit from higher infusion temperatures to reduce fill times. In contrast,
large and complex-shaped parts require use of lower infusion temperatures to prevent premature
resin gelation and incomplete preform saturation.
114
5.1.2 CFRTP Prepreg Processing Project
In Chapter 4, a proof-of-concept case study was conducted to demonstrate the feasibility
of reactive processing of CFRTP prepreg. Conventional thermoplastic prepregs consist of fully
polymerized matrix, which exhibits high melt viscosity and melting temperature. To overcome
these characteristics, CFRTP prepreg must be consolidated at high temperature and high pressure
to promote fiber bed saturation and void removal. On the other hand, the model CFRTP prepreg
proposed in this work contained low-viscosity partially polymerized matrix, facilitating resin flow
during part consolidation even at low temperature and pressure. The pre-polymer resin also
provided provisional tack and drape at ambient condition without any pre-heating measures,
enhancing CFRTP prepreg handleability and conformability.
This work proposed the material characterization and process development guidelines
(Figure 4-6) for developing a reactive CFRTP prepreg system. First, the chemical kinetics and
viscosity evolution during thermoplastic polymerization were analyzed and modeled to determine
the extent of polymerization and pre-polymerization process parameters (e.g., temperature and
time) required for initial prepreg fabrication trials. For PMMA polymerization, a kinetics model
accounting for auto-acceleration phenomenon (from the Trommsdorff effect) was employed. An
aging study on thermoplastic pre-polymer resin was conducted to determine if refrigerated storage
could postpone out-time aging. Based on the material characterization results, prepreg samples
were produced using a lab-scale method, and a prepreg fabrication map was constructed to identify
the optimal fabrication condition for the model prepreg. Finally, the manufactured prepregs can be
tested for room temperature tack and drape and thermoformed for quality analysis.
115
5.2 Broader Implications
Traditionally, high-performance composite parts are manufactured using an autoclave
process, which is robust and well-understood. However, composites manufacturing is gradually
shifting from expensive and inflexible conventional methods to simpler and more cost-efficient
OoA manufacturing processes (e.g., VI and VBO prepreg processing). In this work, we applied
material characterization, process modeling, in situ process diagnostics, and process simulation to
enhance the processability and performance of CFRP parts produced using OoA methods.
Various material characterization techniques can provide the critical information on the
fundamental material behaviors and lay the basis for detailed process modeling. The developed
process models in turn can accurately predict how the process metrics may evolve throughout the
actual manufacturing process. In situ process diagnostics can be used for online process monitoring,
validation of ex situ material characterization results, and real-time process guide and adjustment.
And finally, process simulation can validate and refine the predictions made from the process
models and process guide tools.
The general approaches and methodologies employed in this dissertation to improve
processability and performance of OoA composite manufacturing could potentially be refined and
extended to almost any other composite processing methods:
1. The effectiveness of the additional VI post-infusion dwell, described in Chapter 2, can be
applied to other VI manufacturing cases that can benefit from longer post-filling times.
Examples include cases involving multi-component preforms comprised of different
reinforcement materials (e.g., carbon fiber-glass fiber hybrid reinforcement, a preform
consisting of various fabric architectures, use of interlaminar flow distribution media)
and/or three-dimensional preforms featuring complex shapes.
116
2. The methodology of material characterization and process adjustment for more efficient
use of aged VI resin, presented in Chapter 3, can possibly be extended to other composites
manufacturing processes, including RTM or conventional prepreg processing. Prepreg
resins generally feature longer storage life and available working time than liquid molding
resins. Yet, the time-consuming lay-up process, especially for large and complex-shaped
part geometries, can often expose prepregs to long out-times, substantially advancing resin
degree of cure in the process. A similar resin cure map and process map can be developed
for prepreg-based processes, and the prepreg process cycle can be modified based on
process modeling or simulation analysis results for more efficient manufacturing.
3. Using the methodology we applied for PMMA pre-polymer prepreg development (Chapter
4), we can design a similar thermoplastic prepreg containing pre-polymer resin using other
thermoplastic matrices. For a given CFRTP prepreg system, polymerization reaction can
first be examined and modeled to predict an extent of polymerization for initial prepreg
fabrication trials and to establish a pre-polymerization temperature profile for materials
that require complex reaction schemes. Finally, we can build a prepreg fabrication map to
identify the optimal fabrication condition for the proposed prepreg.
5.3 Recommendations for Future Work
Recommended future work include:
1. In this work, the addition of post-infusion dwell enhanced part quality of thermoplastic
veil-toughened laminate only at the cost of increased process time (Chapter 2). Using
the modified cure cycle, the overall cure cycle time increased from 2.5 hours to 4 hours
117
(vs. manufacturer-recommended cure cycle). Although the cure cycle itself may
represent only a minor fraction of the total process time, further process optimization
can be possible by increasing the post-infusion dwell temperature. An increase in
dwell temperature will expedite the cure reaction and result in shorter cycle time, At
the same time, however, the duration of VI post-filling stage will become shorter,
which can lead to increase in porosity. Therefore, the trade-off between reducing the
overall process time at the expense of increasing risk of void content should be
investigated in future work.
2. The infusion process map presented in Chapter 3 can further be refined and enhanced.
The current infusion map assumes that the process temperature remains constant
throughout the entire infusion process. In practice, however, the epoxy cure reaction
is highly exothermic, and the effective process temperature is likely to increase as resin
degree of cure advances. Therefore, the nominal infusion window estimated from the
infusion map is likely to be much wider than the real infusion window in practice. For
more accurate prediction of infusion window, a process map that accounts for epoxy
cure exotherm and increasing process temperature should be developed.
3. In Chapter 3, the VI process parameters and conditions were evaluated and modified
primarily in terms of VI process time (fill time); however, having a different initial
resin viscosity value or overall viscosity profile can also affect preform impregnation
or part quality as well as dry spot formation. Therefore, further work is required to
examine how the part quality can change with varying resin viscosity history
throughout the infusion stage.
118
4. For reactive processing of CFRTP prepreg (Chapter 4), the effects of varying process
parameters and conditions on consolidation phenomena and mechanical properties of
final composite parts should be studied. Varying initiator weight content, mixing of
additional ingredients (e.g., PMMA polymer or comonomer into monomer/initiator
mixture), and interrupted polymerization can affect the key fundamental polymer
molecular properties (i.e., final degree of polymerization and polydispersity) as well
as flow, compaction, and residual stress formation during part consolidation, which
can result in process-induced deformation.
5. The proposed model thermoplastic prepreg with pre-polymer matrix exhibited limited
shelf-life, out-life and tack and drape lives for practical part manufacturing, especially
for complex-shaped parts (Chapter 4). In principle, these limitations can possibly be
addressed by adopting alternative aerospace-grade thermoplastic matrices with higher
melting temperatures like polyetheretherketone (PEEK), polyetherimide (PEI), and
polycarbonate (PC). Use of such polymers may extend out-life and tack and drape
lives to more practical levels, and thus should be examined.
119
References
[1] Peters ST, editor. Handbook of Composites. Boston, MA: Springer US; 1998.
[2] Campbell Jr FC. Manufacturing Technology for Aerospace Structural Materials. London,
UK: Elsevier; 2006.
[3] Esp B. Practical Analysis of Aircraft Composites. Atlanta, GA: Grand Oak Publishing;
2017.
[4] Composites Market Size, Share & Trends Analysis Report By Product (Carbon, Glass),
By Resin (Thermosetting, Thermoplastics), By Manufacturing Process, By End Use, And
Segment Forecasts, 2020 - 2027 2020. https://www.grandviewresearch.com/industry-
analysis/composites-market (accessed April 23, 2021).
[5] ASC Process Systems Manufactured the World’s Largest Autoclave System n.d.
http://www.aschome.com/index.php/en/products/control-systems/2-uncategorised/327-
world-s-largest-composites-autoclave (accessed April 25, 2021).
[6] Tavares SS, Michaud V, Månson JAE. Assessment of semi-impregnated fabrics in
honeycomb sandwich structures. Compos Part A Appl Sci Manuf 2010;41:8–15.
doi:10.1016/j.compositesa.2009.09.005.
[7] Centea T, Grunenfelder LK, Nutt SR. A review of out-of-autoclave prepregs - Material
properties, process phenomena, and manufacturing considerations. Compos Part A Appl
Sci Manuf 2015;70:132–54. doi:10.1016/j.compositesa.2014.09.029.
[8] Tong R, Hoa S V., Chen M. Cost analysis on L-shape composite component
manufacturing. 18th Int. Conf. Compos. Mater., Jeju Island, Korea: 2011, p. 1–5.
[9] Anders M. In Situ Process Analysis for Defect Control During Composites
Manufacturing. University of Southern California, 2019.
[10] Ahmadova A. Numerical modelling of porosity generation, movement, and compaction
during the RTM process. University of Stuttgart, 2018.
[11] Resin Infusion n.d. https://www.airtechintl.com/resin-infusion (accessed April 26, 2021).
120
[12] Nash NH, Young TM, McGrail PT, Stanley WF. Inclusion of a thermoplastic phase to
improve impact and post-impact performances of carbon fibre reinforced thermosetting
composites - A review. Mater Des 2015;85:582–97. doi:10.1016/j.matdes.2015.07.001.
[13] The Benefits of Vacuum-Infused Resin Technology 2018.
https://www.carveryachts.com/upper-navigation-pages/the-carver-life/content-
container/the-benefits-of-vacuum-infused-resin-technology/ (accessed April 28, 2021).
[14] Yalcinkaya MA, Sozer EM, Altan MC. Fabrication of high quality composite laminates
by pressurized and heated-VARTM. Compos Part A Appl Sci Manuf 2017;102:336–46.
doi:10.1016/j.compositesa.2017.08.017.
[15] Sas HS, Šimáček P, Advani SG. A methodology to reduce variability during vacuum
infusion with optimized design of distribution media. Compos Part A Appl Sci Manuf
2015;78:223–33. doi:10.1016/j.compositesa.2015.08.011.
[16] Yi XS, Du S, Zhang L. Composite materials engineering. vol. 2. 2017. doi:10.1007/978-
981-10-5690-1.
[17] Update T. Technology update: prepregs. Reinf Plast 2003;47:20–1. doi:10.1016/S0034-
3617(03)00627-1.
[18] FibreGlast Prepreg 3K, 2x2 Twill Weave Carbon n.d.
https://www.fibreglast.com/product/Prepreg_3K_2x2_Twill_Weave_Carbon_03101/PrePr
eg_Fabrics (accessed April 28, 2021).
[19] Toray Cetex® TC960 n.d. https://www.toraytac.com/product-
explorer/products/TEW6/Toray-Cetex-TC960 (accessed April 28, 2021).
[20] Solvay. Technical Data Sheet Cycom® 5320-1 Prepreg. 2020.
[21] Hexcel. HexPly F155 Resin System Datasheet. 2016.
[22] TORRESLAYUP-Automatic Tape Layer Machine n.d.
https://www.mtorres.es/en/aeronautics/products/carbon-fiber/torreslayup (accessed April
29, 2021).
[23] Mallick PK. Thermoset–matrix composites for lightweight automotive structures. In:
121
Mallick PK, editor. Mater. Des. Manuf. Light. Veh., Elsevier; 2010, p. 208–31.
doi:10.1533/9781845697822.1.208.
[24] TA Instruments Q500 / Q50 Technology. 2014.
[25] TA Instruments DSC Q Series Specifications. 2012.
[26] TA Instruments Rheometer AR Series Brochure. 2006.
[27] TA Instruments ARES-G2 n.d. https://www.tainstruments.com/ares-g2/ (accessed April
30, 2021).
[28] Lee HL. The handbook of dielectric analysis and cure monitoring. 2nd ed. Boston:
Lambient Technologies; 2018.
[29] DEA 288 Ionic - Dielectric Analyzer n.d. https://www.netzsch-thermal-
analysis.com/en/products-solutions/dielectric-analysis-incl-sensxpert-solution/dea-
(accessed May 3, 2021).
[30] Park H. Dielectric cure determination of a thermosetting epoxy composite prepreg. J Appl
Polym Sci 2017;134:1–9. doi:10.1002/app.44707.
[31] Kratz J, Hsiao K, Fernlund G, Hubert P. Thermal models for MTM45-1 and Cycom 5320
out-of-autoclave prepreg resins. J Compos Mater 2013;47:341–52.
doi:10.1177/0021998312440131.
[32] Cole KC. A New Approach to Modeling the Cure Kinetics of Epoxy Amine
Thermosetting Resins. 1. Mathematical Development. Macromolecules 1991;24:3093–7.
doi:10.1021/ma00011a011.
[33] Kamal MR, Sourour S. Kinetics and thermal characterization of thermoset cure. Polym
Eng Sci 1973;13:59–64. doi:10.1002/pen.760130110.
[34] Khoun L, Centea T, Hubert P. Characterization methodology of thermoset resins for the
processing of composite materials -Case study. J Compos Mater 2010;44:1397–415.
doi:10.1177/0021998309353960.
[35] The industrial Liquid Composite Molding (LCM) simulation tool for large and complex
122
components n.d. https://www.esi-group.com/company/news/esi-releases-pam-rtm-2010
(accessed May 6, 2021).
[36] Nash NH, Young TM, Stanley WF. An investigation of the damage tolerance of
carbon/Benzoxazine composites with a thermoplastic toughening interlayer. Compos
Struct 2016;147:25–32. doi:10.1016/j.compstruct.2016.03.015.
[37] Yuan B, Tan B, Hu Y, Shaw J, Hu X. Improving impact resistance and residual
compressive strength of carbon fibre composites using un-bonded non-woven short
aramid fibre veil. Compos Part A Appl Sci Manuf 2019;121:439–48.
doi:10.1016/j.compositesa.2019.04.006.
[38] Tsotsis TK. Interlayer toughening of composite materials. Polym Compos 2009;30:70–86.
doi:10.1002/pc.20535.
[39] Kuwata M, Hogg PJ. Interlaminar toughness of interleaved CFRP using non-woven veils:
Part 1. Mode-I testing. Compos Part A Appl Sci Manuf 2011;42:1551–9.
doi:10.1016/j.compositesa.2011.07.016.
[40] Kuwata M, Hogg PJ. Interlaminar toughness of interleaved CFRP using non-woven veils:
Part 2. Mode-II testing. Compos Part A Appl Sci Manuf 2011;42:1560–70.
doi:10.1016/j.compositesa.2011.07.017.
[41] Beylergil B, Tanoğlu M, Aktaş E. Effect of polyamide-6,6 (PA 66) nonwoven veils on the
mechanical performance of carbon fiber/epoxy composites. Compos Struct 2018;194:21–
35. doi:10.1016/j.compstruct.2018.03.097.
[42] Garcí a-Rodrí guez SM, Costa J, Singery V, Boada I, Mayugo JA. The effect interleaving
has on thin-ply non-crimp fabric laminate impact response: X-ray tomography
investigation. Compos Part A Appl Sci Manuf 2018;107:409–20.
doi:10.1016/j.compositesa.2018.01.023.
[43] Hamidi YK, Altan MC. Process induced defects in liquid molding processes of
composites. Int Polym Process 2017;32:527–44. doi:10.3139/217.3444.
[44] Simacek P, Advani SG. Modeling resin flow and fiber tow saturation induced by
distribution media collapse in VARTM. Compos Sci Technol 2007;67:2757–69.
doi:10.1016/j.compscitech.2007.02.008.
123
[45] Acheson JA, Simacek P, Advani SG. The implications of fiber compaction and saturation
on fully coupled VARTM simulation. Compos Part A Appl Sci Manuf 2004;35:159–69.
[46] Song YS, Youn JR. Modeling of resin infusion in vacuum assisted resin transfer molding.
Polym Compos 2008;29:390–5. doi:10.1002/pc.20326.
[47] Hindersmann A. Confusion about infusion: An overview of infusion processes. Compos
Part A Appl Sci Manuf 2019;126:105583. doi:10.1016/j.compositesa.2019.105583.
[48] Correia NC, Robitaille F, Long AC, Rudd CD, Šimáček P, Advani SG. Analysis of the
vacuum infusion moulding process: I. Analytical formulation. Compos Part A Appl Sci
Manuf 2005;36:1645–56. doi:10.1016/j.compositesa.2005.03.019.
[49] Kedari VR, Farah BI, Hsiao KT. Effects of vacuum pressure, inlet pressure, and mold
temperature on the void content, volume fraction of polyester/e-glass fiber composites
manufactured with VARTM process. J Compos Mater 2011;45:2727–42.
doi:10.1177/0021998311415442.
[50] van Oosterom S, Allen T, Battley M, Bickerton S. An objective comparison of common
vacuum assisted resin infusion processes. Compos Part A Appl Sci Manuf
2019;125:105528. doi:10.1016/j.compositesa.2019.105528.
[51] Fracassi FT, Donadon M V. Simulation of vacuum assisted resin transfer molding process
through dynamic system analysis. J Compos Mater 2018;52:3759–71.
doi:10.1177/0021998318770000.
[52] Garschke C, Weimer C, Parlevliet PP, Fox BL. Out-of-autoclave cure cycle study of a
resin film infusion process using in situ process monitoring. Compos Part A Appl Sci
Manuf 2012;43:935–44. doi:10.1016/j.compositesa.2012.01.003.
[53] Shin JH, Kim D, Centea T, Nutt SR. Thermoplastic prepreg with partially polymerized
matrix: Material and process development for efficient part manufacturing. Compos Part
A Appl Sci Manuf 2019;119. doi:10.1016/j.compositesa.2019.01.009.
[54] Simacek P, Heider D, Gillespie JW, Advani S. Post-filling flow in vacuum assisted resin
transfer molding processes: Theoretical analysis. Compos Part A Appl Sci Manuf
2009;40:913–24. doi:10.1016/j.compositesa.2009.04.018.
[55] Simacek P, Eksik Ö , Heider D, Gillespie JW, Advani S. Experimental validation of post-
124
filling flow in vacuum assisted resin transfer molding processes. Compos Part A Appl Sci
Manuf 2012;43:370–80. doi:10.1016/j.compositesa.2011.10.002.
[56] Hexcel. Product Data Sheet (Hexcel HexFlow® RTM6) 2018:1–4.
[57] Product data sheet (Fibre Glast part #1530). 2020.
[58] Product data sheet (Cerex Advanced Fabrics PBN-II® ) 2000.
[59] Karkanas PI, Partridge IK. Cure modeling and monitoring of epoxy/amine resin systems.
I. Cure kinetics modeling. J Appl Polym Sci 2000;77:1419–31. doi:10.1002/1097-
4628(20000815)77:7<1419::AID-APP3>3.0.CO;2-N.
[60] Grunenfelder LK, Nutt SR. Prepreg age monitoring via differential scanning calorimetry. J
Reinf Plast Compos 2012;31:295–302. doi:10.1177/0731684411431020.
[61] Kim D, Centea T, Nutt SR. In-situ cure monitoring of an out-of-autoclave prepreg: Effects
of out-time on viscosity, gelation and vitrification. Compos Sci Technol 2014;102:132–8.
doi:10.1016/j.compscitech.2014.07.027.
[62] Shin JH, Kim D, Centea T, Nutt SR. Thermoplastic Prepreg with a Partially Polymerized
Matrix. SAMPE Conf., Long Beach: 2018, p. 1102–15.
[63] Kim D, Centea T, Nutt SR. Modelling and monitoring of out-time and moisture absorption
effects on cure kinetics and viscosity for an out-of-autoclave (OoA) prepreg. Compos Sci
Technol 2017;138:201–8. doi:10.1016/j.compscitech.2016.11.023.
[64] Maistros GM, Partridge IK. Dielectric monitoring of cure in a commercial carbon-fibre
composite. Compos Sci Technol 1995;53:355–9. doi:10.1016/0266-3538(95)00007-0.
[65] Chaloupka A, Pflock T, Horny R, Rudolph N, Horn SR. Dielectric and rheological study
of the molecular dynamics during the cure of an epoxy resin. J Polym Sci Part B Polym
Phys 2018;56:907–13. doi:10.1002/polb.24604.
[66] Müller U, Pretschuh C, Zikulnig-Rusch E, Dolezel-Horwath E, Reiner M, Knappe S.
Dielectric analysis as cure monitoring system for melamine-formaldehyde laminates. Prog
Org Coatings 2016;90:277–83. doi:10.1016/j.porgcoat.2015.10.019.
125
[67] Naik NK, Sirisha M, Inani A. Permeability characterization of polymer matrix composites
by RTM/VARTM. Prog Aerosp Sci 2014;65:22–40. doi:10.1016/j.paerosci.2013.09.002.
[68] Weitzenböck JR, Shenoi RA, Wilson PA. Radial flow permeability measurement. Part A:
theory. Compos Part A Appl Sci Manuf 1999;30:781–96. doi:10.1016/S1359-
835X(98)00183-3.
[69] Pierce RS, Falzon BG, Thompson MC. A multi-physics process model for simulating the
manufacture of resin-infused composite aerostructures. Compos Sci Technol
2017;149:269–79. doi:10.1016/j.compscitech.2017.07.003.
[70] Pierce RS, Falzon BG, Thompson MC. Permeability characterization of sheared carbon
fiber textile preform. Polym Compos 2018;39:2287–98. doi:10.1002/pc.24206.
[71] ASTM International. D7136/D7136M-15 Standard test method for measuring the damage
resistance of a fiber-renforced polymer matrix composite to a drop-weight impact. West
Conshocken, PA: ASTM International; 2015.
doi:https://doi.org/10.1520/D7136_D7136M-15.
[72] Navabpour P, Nesbitt A, Mann T, Day RJ. Comparison of the curing kinetics of a
DGEBA/acid anhydride epoxy resin system using differential scanning calorimetry and a
microwave-heated calorimeter. J Appl Polym Sci 2007;104:2054–63.
doi:10.1002/app.25501.
[73] Karkanas PI, Partridge IK. Cure modeling and monitoring of epoxy/amine resin systems.
II. Network formation and chemoviscosity modeling. J Appl Polym Sci 2000;77:2178–88.
doi:10.1002/1097-4628(20000906)77:10<2178::AID-APP11>3.0.CO;2-0.
[74] Kazilas MC. Acquisition and Interpretation of Dielectric Data For Thermoset Cure
Monitoring. Cranfield University, 2003.
[75] Skordos AA, Partridge IK. Determination of the degree of cure under dynamic and
isothermal curing conditions with electrical impedance spectroscopy. J Polym Sci Part B
Polym Phys 2004;42:146–54. doi:10.1002/polb.10676.
[76] Karkanas PI. Cure modelling and monitoring of epoxy/amine resin systems. Cranfield
University, 1998.
[77] Arbter R, Beraud JM, Binetruy C, Bizet L, Bréard J, Comas-Cardona S, et al.
126
Experimental determination of the permeability of textiles: A benchmark exercise.
Compos Part A Appl Sci Manuf 2011;42:1157–68.
doi:10.1016/j.compositesa.2011.04.021.
[78] Mehdikhani M, Gorbatikh L, Verpoest I, Lomov S V. Voids in fiber-reinforced polymer
composites: A review on their formation, characteristics, and effects on mechanical
performance. J Compos Mater 2019;53:1579–669. doi:10.1177/0021998318772152.
[79] Villière M, Guéroult S, Sobotka V, Boyard N, Bréard J, Delaunay D. Dynamic saturation
curve measurement in liquid composite molding by heat transfer analysis. Compos Part A
Appl Sci Manuf 2015;69:255–65. doi:10.1016/j.compositesa.2014.11.024.
[80] Lee DH, Lee W Il, Kang MK. Analysis and minimization of void formation during resin
transfer molding process. Compos Sci Technol 2006;66:3281–9.
doi:10.1016/j.compscitech.2005.07.008.
[81] Staffan Lundström T, Frishfelds V, Jakovics A. Bubble formation and motion in non-
crimp fabrics with perturbed bundle geometry. Compos Part A Appl Sci Manuf
2010;41:83–92. doi:10.1016/j.compositesa.2009.05.012.
[82] Matsuzaki R, Seto D, Todoroki A, Mizutani Y. In situ void content measurements during
resin transfer molding. Adv Compos Mater 2013;22:239–54.
doi:10.1080/09243046.2013.801822.
[83] Jeong H. Effects of Voids on the Mechanical Strength and Ultrasonic Attenuation of
Laminated Composites. J Compos Mater 1997;31:276–92.
doi:10.1177/002199839703100303.
[84] Kalantari M, Dong C, Davies IJ. Effect of matrix voids, fibre misalignment and thickness
variation on multi-objective robust optimization of carbon/glass fibre-reinforced hybrid
composites under flexural loading. Compos Part B Eng 2017;123:136–47.
doi:10.1016/j.compositesb.2017.05.022.
[85] Belingardi G, Cavatorta MP, Salvatore Paolino D. Repeated impact response of hand lay-
up and vacuum infusion thick glass reinforced laminates. Int J Impact Eng 2008;35:609–
19. doi:10.1016/j.ijimpeng.2007.02.005.
[86] Scarponi C, Sarasini F, Tirillò J, Lampani L, Valente T, Gaudenzi P. Low-velocity impact
behaviour of hemp fibre reinforced bio-based epoxy laminates. Compos Part B Eng
2016;91:162–8. doi:10.1016/j.compositesb.2016.01.048.
127
[87] Koricho EG, Khomenko A, Haq M, Drzal LT, Belingardi G, Martorana B. Effect of
hybrid (micro- and nano-) fillers on impact response of GFRP composite. Compos Struct
2015;134:789–98. doi:10.1016/j.compstruct.2015.08.106.
[88] Yalcinkaya MA, Sozer EM, Altan MC. Effect of external pressure and resin flushing on
reduction of process-induced voids and enhancement of laminate quality in heated-
VARTM. Compos Part A Appl Sci Manuf 2019;121:353–64.
doi:10.1016/j.compositesa.2019.03.040.
[89] Pierce RS, Falzon BG. Simulating Resin Infusion through Textile Reinforcement
Materials for the Manufacture of Complex Composite Structures. Engineering
2017;3:596–607. doi:10.1016/J.ENG.2017.04.006.
[90] Poorzeinolabedin M, Parnas L, Dashatan SH. Resin infusion under flexible tooling
process and structural design optimization of the complex composite part. Mater Des
2014;64:450–5. doi:10.1016/j.matdes.2014.08.008.
[91] Loudad R, Saouab A, Beauchene P, Agogue R, Desjoyeaux B. Numerical modeling of
vacuum-assisted resin transfer molding using multilayer approach. J Compos Mater
2017;51:3441–52. doi:10.1177/0021998316687145.
[92] Govignon Q, Bickerton S, Kelly PA. Simulation of the reinforcement compaction and
resin flow during the complete resin infusion process. Compos Part A Appl Sci Manuf
2010;41:45–57. doi:10.1016/j.compositesa.2009.07.007.
[93] Hexcel. HexFlow RTM6/ RTM6-2: Safety and processing recomendations for injection
and infusion 2014.
[94] Solvay. Technical Data Sheet. PRISM EP2400, Resin for Infusion 2018.
[95] Pantelelis NG, Bistekos E. Process monitoring and control for the production of CFRP
components. Int. SAMPE Symp. Exhib., Seattle: 2010, p. 5–9.
[96] Pantelelis N, Millar T. In-situ process monitoring and quality control for advanced liquid
composite moulding. Int. Conf. Manuf. Adv. Compos., Belfast: 2011, p. 2–5.
[97] Grujicic M, Chittajallu KM, Walsh S. Non-isothermal preform infiltration during the
vacuum-assisted resin transfer molding (VARTM) process. Appl Surf Sci 2005;245:51–
64. doi:10.1016/j.apsusc.2004.09.123.
128
[98] Wu D, Larsson R, Blinzler B. A preform deformation and resin flow coupled model
including the cure kinetics and chemo-rheology for the VARTM process. Int J Mater
Form 2020. doi:10.1007/s12289-020-01570-z.
[99] Shin JH, Anders M, Kim D, Jin BC, Nutt S. Effects of post-infusion dwell on vacuum
infusion of thermoset composites toughened by thermoplastic interlaminar veils. J
Compos Mater 2020. doi:10.1177/0021998320971342.
[100] Trevino L, Rupel K, Young WB, Liou MJ, Lee LJ. Analysis of resin injection molding in
molds with preplaced fiber mats. I: Permeability and compressibility measurements.
Polym Compos 1991;12:20–9. doi:10.1002/pc.750120105.
[101] Thostenson ET, Ziaee S, Chou TW. Processing and electrical properties of carbon
nanotube/vinyl ester nanocomposites. Compos Sci Technol 2009;69:801–4.
doi:10.1016/j.compscitech.2008.06.023.
[102] Sadler JM, Nguyen APT, Toulan FR, Szabo JP, Palmese GR, Scheck C, et al. Isosorbide-
methacrylate as a bio-based low viscosity resin for high performance thermosetting
applications. J Mater Chem A 2013;1:12579–86. doi:10.1039/c3ta12918g.
[103] Raponi O de A, Souza BR de, Baldo Junior JE, Ancelotti Junior AC, Guimarães A.
Dielectric analysis as a low-complexity methodology for tracking prepreg out-time and its
effects on the curing cycle. J Compos Mater 2019;53:4035–42.
doi:10.1177/0021998319853325.
[104] Gu Y, Li M, Zhang Z, Li Y. Effects of resin storage aging on rheological property and
consolidation of composite laminates. Polym Compos 2009;30:1081–90.
doi:10.1002/pc.20659.
[105] Kim D, Nutt SR. Processability of DDS isomers-cured epoxy resin: Effects of
amine/epoxy ratio, humidity, and out-time. Polym Eng Sci 2018;58:1530–8.
doi:10.1002/pen.24738.
[106] Parlevliet PP, Bersee HEN, Beukers A. Residual stresses in thermoplastic composites-A
study of the literature-Part II: Experimental techniques. Compos Part A Appl Sci Manuf
2007;38:651–65. doi:10.1016/j.compositesa.2006.07.002.
[107] Vaidya UK, Chawla KK. Processing of fibre reinforced thermoplastic composites. Int
Mater Rev 2008;53:185–218. doi:10.1179/174328008X325223.
129
[108] Hou M. Stamp Forming of Continuous Glass Fibre Reinforced Polypropylene. Compos -
Part A Appl Sci Manuf 1997;28A:695–702. doi:10.1016/S1359-835X(97)00013-4.
[109] Ó Máirtí n P, McDonnell P, Connor MT, Eder R, Ó Brádaigh CM. Process investigation of
a liquid PA-12/carbon fibre moulding system. Compos - Part A Appl Sci Manuf
2001;32:915–23. doi:10.1016/S1359-835X(01)00005-7.
[110] Parton H, Baets J, Lipnik P, Goderis B, Devaux J, Verpoest I. Properties of poly(butylene
terephthatlate) polymerized from cyclic oligomers and its composites. Polymer (Guildf)
2005;46:9871–80. doi:10.1016/j.polymer.2005.07.082.
[111] van Rijswijk K, Bersee HEN. Reactive processing of textile fiber-reinforced thermoplastic
composites - An overview. Compos Part A Appl Sci Manuf 2007;38:666–81.
doi:10.1016/j.compositesa.2006.05.007.
[112] Chang IY, Lees JK. Recent Development in Thermoplastic Composites: A Review of
Matrix Systems and Processing Methods. J Thermoplast Compos Mater 1988;1:277–96.
doi:10.1177/089270578800100305.
[113] Yousefpour A, Hojjati M, Immarigeon JP. Fusion bonding/welding of thermoplastic
composites. J Thermoplast Compos Mater 2004;17:303–41.
doi:10.1177/0892705704045187.
[114] Yang Y, Boom R, Irion B, van Heerden DJ, Kuiper P, de Wit H. Recycling of composite
materials. Chem Eng Process Process Intensif 2012;51:53–68.
doi:10.1016/j.cep.2011.09.007.
[115] Stavrov D, Bersee HEN. Resistance welding of thermoplastic composites-an overview.
Compos Part A Appl Sci Manuf 2005;36:39–54. doi:10.1016/j.compositesa.2004.06.030.
[116] Ageorges C, Ye L, Hou M. Advances in fusion bonding techniques for joining
thermoplastic matrix composites: A review. Compos - Part A Appl Sci Manuf
2001;32:839–57. doi:10.1016/S1359-835X(00)00166-4.
[117] Strong AB. Fundamentals of Composites Manufacturing. 2nd ed. Dearborn: Society of
Manufacturing Engineers; 2008. doi:10.1016/0956-7143(92)90210-L.
[118] Parton H, Verpoest I. In situ polymerization of thermoplastic composites based on cyclic
oligomers. Polym Compos 2005;26:60–5. doi:10.1002/pc.20074.
130
[119] Luisier A, Bourban P-E, Månson J-AE. Time-temperature-transformation diagram for
reactive processing of polyamide 12. J Appl Polym Sci 2001;81:963–72.
doi:10.1002/app.1518.
[120] Humphry J, Vandi L-J, Martin D, Heitzmann M. Time-Temperature-Transformation
Modelling for the in situ Processing of Thermoplastics as a Composite Matrix. PFAM-25
Conf., Auckland: 2017, p. 792–803.
[121] van Rijswijk K, Bersee HEN, Jager WF, Picken SJ. Optimisation of anionic polyamide-6
for vacuum infusion of thermoplastic composites: choice of activator and initiator.
Compos Part A Appl Sci Manuf 2006;37:949–56. doi:10.1016/j.compositesa.2005.01.023.
[122] Kim SW, Seong DG, Yi JW, Um MK. Development of high-speed reactive processing
system for carbon fiber-reinforced polyamide-6 composite: In-situ anionic ring-opening
polymerization. Times Polym. Compos. Conf., Naples: 2016, p. 020022 1-4.
doi:10.1063/1.4949597.
[123] Pillay S, Vaidya UK, Janowski GM. Liquid molding of carbon fabric-reinforced nylon
matrix composite laminates. J Thermoplast Compos Mater 2005;18:509–27.
doi:10.1177/0892705705054412.
[124] Luisier A, Bourban P, Månson J. In Situ Polymerization of Polyamide 12 for
Thermoplastic Composites. ICCM-12 Conf., Paris: 1999, p. 33–45.
[125] Rosso P, Friedrich K, Wollny A, Mülhaupt R. A novel polyamide 12 polymerization
system and its use for a LCM-process to produce CFRP. J Thermoplast Compos Mater
2005;18:77–90. doi:10.1177/0892705705041987.
[126] Ma CM, Chen C. Pultruded fiber reinforced thermoplastic poly(methyl methacrylate)
composites. Part I. Correlation of processing parameters for optimizing the process. Polym
Eng Sci 1991;31:1086–93. doi:10.1002/pen.760311503.
[127] Suzuki Y, Cousins D, Wassgren J, Kappes BB, Dorgan J, Stebner AP. Kinetics and
temperature evolution during the bulk polymerization of methyl methacrylate for vacuum-
assisted resin transfer molding. Compos Part A Appl Sci Manuf 2018;104:60–7.
doi:10.1016/j.compositesa.2017.10.022.
[128] Sigma-Aldrich. Safety data sheet, Methyl methacrylate 2017:1–10.
131
[129] Gill PS, Sauerbrunn SR, Reading M. Modulated differential scanning calorimetry. J
Therm Anal 1993;40:931–9. doi:10.1007/BF02546852.
[130] Chen TT. Strategies for Rheological Evaluation of Pressure Sensitive Adhesives. PSTC-
40 Conf., Las Vegas: 2017, p. 13–6.
[131] Coates J. Interpretation of Infrared Spectra, A Practical Approach. Encycl. Anal. Chem.,
Chichester, UK: John Wiley & Sons, Ltd; 2006, p. 10815–37.
doi:10.1002/9780470027318.a5606.
[132] Dazzi A, Deniset-Besseau A, Lasch P. Minimising contributions from scattering in
infrared spectra by means of an integrating sphere. Analyst 2013;138:4191.
doi:10.1039/c3an00381g.
[133] Duan G, Zhang C, Li A, Yang X, Lu L, Wang X. Preparation and characterization of
mesoporous zirconia made by using a poly (methyl methacrylate) template. Nanoscale Res
Lett 2008;3:118–22. doi:10.1007/s11671-008-9123-7.
[134] Dirlikov S, Koenig JL. Infrared Spectra of Poly(Methyl Methacrylate) Labeled with
Oxygen-18. Appl Spectrosc 1979;33:551–5. doi:10.1366/0003702794925002.
[135] Jašo V, Stoiljković D, Radičević R, Bera O. Kinetic modeling of bulk free-radical
polymerization of methyl methacrylate. Polym J 2013;45:631–6. doi:10.1038/pj.2013.6.
[136] Radičević RŽ, Stoiljković DM, Budinski-Simendić J. Characteristic events in free radical
polymerization of lower n-alkyl methacrylates. J Therm Anal Calorim 2000;62:237–49.
doi:10.1023/A:1010135400465.
[137] Castro JM, Macosko CW. Kinetics & Rheology of Typical Polyurethane Reaction
Injection Molding Systems. SPE-38 Conf., New York: 1980, p. 434–8.
[138] TenCate. Technical product data sheet (TenCate TW-2000) 2000:93012.
[139] Arkema. Technical product data sheet (Arkema Elium® ) 2017.

Abstract (if available)

Abstract Advanced composite materials based on continuous fiber-reinforced polymers (CFRPs) feature high specific strength and stiffness. CFRPs are increasingly used across multiple industries, including aerospace, automotive, energy, and sporting goods to manufacture structural parts. CFRPs must be consolidated at high temperature and pressure to facilitate fiber bed saturation and void removal. Thus, high-performance composite structures are traditionally produced using an autoclave, which is a pressurized vessel.❧ Yet, the use of autoclave incurs high capital and operating costs and imposes inflexible manufacturing environment because size and availability of autoclave is often limited. The predicted increase in market demand for CFRPs is driving exploration of faster and more cost-efficient out-of-autoclave (OoA) composite processing methods. Yet, OoA processes are more likely to yield less robust parts, and thus requires further process developments and adjustments for more extensive and efficient applications. Vacuum infusion (VI) is a promising alternative to conventional autoclave prepreg process, especially for manufacture of large and complex unitized composite structures. In VI, a dry fiber preform is placed on a one-sided rigid mold and sealed with a flexible vacuum bag. Then, resin is infused into the vacuum bag under vacuum (negative) pressure, then heated and cured. Interest in VI has grown rapidly in the recent years, particularly in the aerospace industry, which seeks to reduce the manufacturing costs associated with prepreg processing. However, because only vacuum pressure is applied during infusion, VI requires use of low-viscosity thermoset resin, which is inherently brittle and susceptible to impact damage. To enhance impact performance, we introduced non-woven thermoplastic veils at interlaminar regions of thermoset composites and addressed the associated process challenges (Chapter 2). The use of thermoplastic veils in VI preforms increased impact resistance but also induced non-uniform flow fronts during infusion, leading to high porosity in the final laminate. We modified the conventional VI cure cycle to include a low-temperature post-infusion (LTPI) dwell to allow more time for resin to equilibrate pressure and redistribute during the VI post-filling stage, achieving full saturation of dry interlaminar regions. Also, VI endures slow infusion rate and long fill time because only limited pressure is available to drive resin flow during infusion. To mitigate this issue, resin is often heated to accelerate infusion, although doing so can increase resin degree of cure and viscosity with flow time and distance, requiring careful selection of process parameters. In Chapter 3, we assessed the physical state of VI resin using in situ dielectric process diagnostics coupled with cure modeling. Then, the heated filling process was simulated to guide accurate on-the-fly process adjustments for more efficient use of aged resin. The results demonstrated that the VI process parameters must be adjusted not only for part size and geometry but also for resin age. Continuous fiber-reinforced thermoplastics (CFRTPs) provide key intrinsic advantages over thermoset composites, including high impact resistance, short process cycle, unlimited shelf-life, recyclability, and weldability. However, despite these advantages, melt processing of CFRTP retains one critical challenge. Because of high melt viscosity and melting temperature, high-performance CFRTPs must be consolidated at high temperature and high pressure, requiring use of autoclave or other energy-intensive manufacturing processes. Therefore, extensive use of CFRTPs in aerospace industry has been restricted by lack of technically mature and robust OoA manufacturing processes. To enable efficient OoA manufacturing of CFRTP, we conducted a proof-of-concept case study to demonstrate the feasibility of thermoplastic prepreg with partially polymerized matrix (Chapter 4). As opposed to conventional thermoplastic prepreg, which is characterized by fully polymerized matrix, our model pre-polymer prepreg contained low-viscosity pre-polymer resin. The pre-polymer resin facilitated part consolidation at low temperature and pressure, even below the glass transition of the final polymer matrix, while providing provisional tack and drape at room temperature for improved material handleability. Overall, we proposed a pathway to reduce major OoA process challenges associated with CFRTPs, including thermal and pressure conditions required for part consolidation as well as material conformability of CFRTP prepregs. Overall, this work addresses some of the major process challenges in OoA composites manufacturing using various approaches. First, effective process guide and optimization tools for OoA processes were developed applying material characterization results, polymerization process models, in situ dielectric process diagnostics, and process simulation. Second, the conventional VI process cycle was modified by introducing an additional post-infusion dwell step to extend the VI post-filling stage and to improve part quality of multi-component laminate. Third, an efficient methodology was developed to accurately assess the physical state of resin in situ, increase VI process efficiency using process map and process simulation, and consequently, reduce liquid resin material waste. Finally, a new material/product form of thermoplastic prepreg, which contain partially polymerized matrix, was developed through a proof-of-concept case study to significantly enhance material processability and conformability.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Efficient manufacturing and repair of out-of-autoclave prepreg composites

PDF

Fabrication and analysis of prepregs with discontinuous resin patterning for robust out-of-autoclave manufacturing

PDF

Defect control in vacuum bag only processing of composite prepregs

PDF

Void evolution in vacuum bag-only prepregs

PDF

Sustainable manufacturing of out-of-autoclave (OoA) prepreg: processing challenges

PDF

In situ process monitoring for modeling and defect mitigation in composites processing

PDF

Developing efficient methods for the manufacture and analysis of composites structures

PDF

Processing and properties of phenylethynyl-terminated PMDA-type asymmetric polyimide and composites

PDF

Polymer flow for manufacturing fiber reinforced polymer composites

PDF

Characterization, process analysis, and recycling of a benzoxazine-epoxy resin for structural composites

PDF

Chemical recycling of amine/epoxy composites at atmospheric pressure

PDF

In situ process analysis for defect control during composites manufacturing

PDF

Sound transmission through acoustic metamaterials and prepreg processing science

PDF

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

PDF

Processing, thermal-oxidative stability and thermal cyclic fatigue of phenylethynyl-terminated polyimides

PDF

Process planning for robotic additive manufacturing

PDF

Vacuum-bag-only processing of composites

PDF

AI-driven experimental design for learning of process parameter models for robotic processing applications

PDF

Development of composite oriented strand board and structures

PDF

Fabrication of polymer films on liquid substrates via initiated chemical vapor deposition: controlling morphology and composition

Asset Metadata

Creator Shin, Jung Hwan (author)

Core Title Material and process development and optimization for efficient manufacturing of polymer composites

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Degree Conferral Date 2021-08

Publication Date 07/16/2021

Defense Date 06/02/2021

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

Tag composites,manufacturing,material development,OAI-PMH Harvest,process design,process optimization

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Nutt, Steven (committee chair), Grunenfelder, Lessa (committee member), Gupta, Satyandra (committee member), Sharada, Shaama (committee member)

Creator Email jhshin892@gmail.com,shinjung@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15595904

Unique identifier UC15595904

Legacy Identifier etd-ShinJungHw-9755

Document Type Dissertation

Format application/pdf (imt)

Rights Shin, Jung Hwan

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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

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

Repository Email cisadmin@lib.usc.edu