University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2022 Daniel Zebrine

IN SITU PROCESS MONITORING FOR MODELING AND DEFECT MITIGATION IN
COMPOSITES PROCESSING

by

Daniel Zebrine

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
(MATERIALS SCIENCE)

August 2022

ii
Acknowledgments
Thank you to my advisor, Prof. Steve Nutt, for all the guidance and support through my
time at USC. I’m very grateful you believed in a fellow physics major from a tiny school in the
middle of PA, and provided me this great opportunity to learn and grow both professionally and
personally.
I am incredibly grateful to have had the opportunity to work with Tim Centea. Your
mentorship throughout the years was invaluable, rivaled perhaps only by your insightful and
mostly unbiased hockey discussion.
Thank you to the whole M.C. Gill Composites Center; to Mark Anders for the mentorship
and collaboration on projects; and to Jonathan Lo, Trisha Palit, Elana Wadhwani, Nick Bortolon,
and Tilman Voorsanger. Thanks especially to Yunpeng Zhang for keeping the lab running
smoothly.
This work was funded by NASA Langley Research Center, Hexcel and M.C. Gill
Composites Center, and I would like to acknowledge project leads Roberto Cano (NASA) and
Olivia Niitsoo (Hexcel) as well as all those who provided helpful discussion throughout my
research. I am also grateful to Navid Niknafs Kermani, Pavel Simacek, Thomas Cender, and Prof.
Suresh Advani at the University of Delaware for their collaboration on the NASA project.
Finally, thank you to my family, and most importantly to Tricia Mundell, for all the love
and support, and for keeping me sane all these years. And of course, I must thank Matilda Zebrine,
who arrived just in time to make it into these acknowledgements.

iii
TABLE OF CONTENTS

Acknowledgments........................................................................................................................... ii
List of Tables .................................................................................................................................. v
List of Figures ................................................................................................................................ vi
Abstract .......................................................................................................................................... xi
Chapter 1: Introduction ................................................................................................................... 1
1.1 Motivation ...................................................................................................... 1
1.2 Honeycomb Core Sandwich Structures.......................................................... 2
1.3 Out-of-autoclave Processing .......................................................................... 5
1.4 In Situ Visualization Methods ........................................................................ 9
1.5 Dissertation Outline...................................................................................... 10
Chapter 2: Experimental Methods for In Situ Visualization ......................................................... 13
1. Introduction ....................................................................................................... 13
2. In Situ Co-Cure Fixture ..................................................................................... 14
3. Oven Window Frame ........................................................................................ 15
Chapter 3: Path-Dependent Bond-line Evolution in Equilibrated Core Honeycomb Sandwich
Structures .......................................................................................................................... 18
1. Introduction ....................................................................................................... 18
2. Experimental Methods ...................................................................................... 22
3. Results and Discussion ..................................................................................... 27
4. Void Growth Modeling ..................................................................................... 37
5. Conclusions ....................................................................................................... 44
Chapter 4: Process Mapping for Defect Control in the Adhesive Bond-line of Co-cured
Honeycomb Core Sandwich Structures ............................................................................ 47
1. Introduction ....................................................................................................... 47
2. Experimental Methods ...................................................................................... 48
3. Results and Discussion ..................................................................................... 50
4. Conclusions ....................................................................................................... 55
Chapter 5: Experimental Validation of Co-cure Process of Honeycomb Sandwich Structures
Simulation: Adhesive Fillet Shape and Bond-line Porosity ............................................. 58
1. Introduction ....................................................................................................... 58
2. Materials ........................................................................................................... 61
3. Adhesive Fillet Formation ................................................................................ 62
4. Bond-line Porosity Development ...................................................................... 66
5. Conclusions ....................................................................................................... 81
Chapter 6: Surface Porosity Development in Tool-side Facesheets of Honeycomb Core
Sandwich Structures during Co-cure ................................................................................ 84
iv
1. Introduction ....................................................................................................... 84
1.1 Objectives and Approach ................................................................................ 86
2. Materials and Methodology .............................................................................. 87
3. Results and Discussion ..................................................................................... 92
4. Conclusions ..................................................................................................... 104
Chapter 7: Mitigating Void Growth in Out-of-Autoclave Prepreg Processing Using a Semi-
Permeable Membrane to Maintain Resin Pressure ......................................................... 108
1. Introduction ..................................................................................................... 108
2. Experimental Methods .................................................................................... 112
3. Results and Discussion ................................................................................... 118
4. Conclusions ..................................................................................................... 125
Chapter 8: Conclusions ............................................................................................................... 128
1. Honeycomb Core Sandwich Structures .......................................................... 128
2. Out-of-autoclave processing with semi-permeable membranes ..................... 130
References ................................................................................................................................... 132

v
List of Tables
Table 1: Testing parameters for samples manufactured in the in situ co-cure fixture. All
pressures are absolute. ...................................................................................................... 24
Table 2: Parameters and corresponding levels used for cure cycle screening. ............................. 49
Table 3: Parameters used in the void growth model. .................................................................... 73
Table 4: Experimental and simulation results for bond-line porosity in each sample. ................. 80
Table 5: Testing conditions for autoclave processing. All pressures are absolute. For each
testing condition, two sandwich panel samples were fabricated. Additionally, a
comparison laminate was made for the baseline and vented bag conditions. ................... 89
Table 6: Testing conditions for lab-scale laminate fabrication. Two laminates were produced
for each condition. .......................................................................................................... 114
Table 7: Testing conditions for resin pressure measurements. Each test was repeated twice. ... 117

vi
List of Figures
Figure 1: a) Diagram showing composite usage in Boeing 787 Dreamliner. b) Projected
demand for various aircrafts between 2019 and 2040. Image source: The Boeing
Company (www.boeing.com). ............................................................................................ 2
Figure 2: Relative stiffness and weight of laminate compared to sandwich panels with
different core thickness. Image source: [2] ......................................................................... 3
Figure 3: Pressure transfer in a honeycomb sandwich panel. At the stiff core cell walls, applied
autoclave pressure is transferred to the facesheet and adhesive. Within the cells,
compaction pressure applied to the facesheet and adhesive is dependent on the gas
pressure within the cell. ...................................................................................................... 4
Figure 4: Top: schematic of out-of-autoclave prepreg design. Bottom: Scanning electron
microscope image of OoA prepreg showing dry fibers between resin and impregnated
fibers. Image source: [9] ..................................................................................................... 6
Figure 5: Typical layup with consumables for out-of-autoclave prepreg processing, including
breathable edge dams to enable in-plane air evacuation. Image source: [10] .................... 6
Figure 6: Top: Schematic of semipreg format, with through-thickness air removal enabled by
discontinuous resin patterns. Bottom: Examples of semipregs with different resin
patterns. Image source: [13] ................................................................................................ 7
Figure 7: Development of resin pressure during autoclave cure using a spring in a fluid-filled
piston as an analog. In the model, the spring represents the fiber bed, and the fluid
represents the resin. In Step (3), a valve in the piston is opened, allowing for fluid to
bleed out of the piston and transfer pressure to the spring. Image source: [14] ................. 9
Figure 8: Schematic of the in situ co-cure fixture, or “mini-autoclave”. Honeycomb core and
a single facesheet are laid on a glass window, enabling visualization of the bond-line
during cure. ....................................................................................................................... 13
Figure 9: a) Schematic view of camera perspective of a single honeycomb core cell via mini-
autoclave. b) Sample frame from a test run using the MA fixture. .................................. 14
Figure 10: Diagram of oven cure set-up for in situ visualization of the tool-side surface using
the oven window frame.. The use of the framed glass tool plate reduced thermal
gradients (compared to placing samples directly on the oven window). A digital
microscope is placed outside the oven to observe the tool-side facesheet through the
oven window and glass tool plate. .................................................................................... 16
Figure 11: Fillet measurement method for Test 1 (top) and Test 8 (bottom). .............................. 27
vii
Figure 12: Time-lapse video frames for aluminum bonding under vacuum (Test 1) and at
ambient pressure (Test 3). t1 is the initial state, t2 is halfway through the first
temperature ramp, and t3 is the end of the first temperature ramp. ................................... 28
Figure 13: Time-lapse video frames for co-cure tests, with the temperature profile and
modeled viscosities of both the prepreg and the adhesive. Times are denoted on the
temperature plot. ............................................................................................................... 29
Figure 14: Height (top) and void area (bottom) versus core pressure for aluminum bonding
and co-cure. ....................................................................................................................... 32
Figure 15: Micrographs for: a) Test 1 (aluminum bonding, vacuum), b) Test 3 (aluminum
bonding, ambient), c) Test 4 (co-cure, vacuum), d) Test 8 (co-cure, ambient), e) Test
11 (co-cure, super-ambient). ............................................................................................. 33
Figure 16: TGA Results for EA 9658 AERO NWG adhesive (top) and 8552S prepreg
(bottom). Data for the adhesive under vacuum has been smoothed. ................................ 35
Figure 17: FTIR spectroscopy data for the prepreg, with the thermal cycle (left) and mass loss
(from TGA, right) overlaid. The volatile observed beginning ~110 °C was identified
as methyl ethyl ketone. Image source: [48] ...................................................................... 36
Figure 18: Comparison of individual bubble (left) and “foaming” behavior (right). Individual
bubble growth was not considered when determining the onset of void growth. ............. 40
Figure 19: Experimental and model results for critical temperature as a function of pressure. ... 41
Figure 20: Two sample pressure cycles demonstrating path dependence of void growth. In
Cycle A, core pressure is maintained at a level greater than the critical pressure, thus
no void growth is expected. In Cycle B, the delayed increase in pressure causes the
critical pressure to increase beyond the core pressure, and void growth is predicted
during this period. ............................................................................................................. 43
Figure 21: Process mapping (left) and void growth factor (right). ............................................... 52
Figure 22: Cross-sections of samples made under each processing condition. A) Baseline B)
1 °C/min ramp rate C) 4 °C/min ramp rate D) 0 min dwell time E) 240 min dwell time
F) 90 °C dwell temperature G) 130 °C dwell temperature. .............................................. 54
Figure 23: Void fraction in fillets for each temperature cycle assessed. Error bars are ± one
standard deviation of average values for each section (four sections per condition). ...... 55
Figure 24: Geometry used in the model for fillet formation. a) Fillet geometry as modeled. b)
Fillet dimensions and contact angles. c) Parametrization of applied force. ...................... 62
Figure 25: a) Cross-section of adhesive fillet model validation sample. b) Sample of fillet
height measurement. ......................................................................................................... 65
viii
Figure 26: Model predictions compared to experimental data for fillet formation model. .......... 66
Figure 27: Coupling between physical phenomena that affect porosity development in the
adhesive bond-line. ........................................................................................................... 67
Figure 28: Left: Diagram showing possible motion of voids or dissolved volatiles, including
transfer of voids from the prepreg to bond-line region via resin bleed and escape of
voids from the bond-line into the core. Right: Schematic of void escape description in
the integrated co-cure model. If a void is located near the interface between adhesive
and core gas pocket (Zone 2) and its radius is large enough such that the void touches
this interface, the voids is defined to escape into the core and is removed from the
bond-line. Image source: [51] ........................................................................................... 68
Figure 29: Cure cycles I and II used for fabrication of validation samples for void growth
model. a) Cure cycle I for Sample A had the vacuum bag and core vented to ambient
pressure. b) Cure cycle II for Samples B and C had vacuum applied to the bag and
core. ................................................................................................................................... 70
Figure 30: In situ images of the bond-line during cure. Images a-e are from Sample A and f-j
from Sample B. Times, which correspond to cure cycles in Figure 5, are a) 0 min, b)
30 min, c) 60 min, d) 112 min, e) 150 min (cured state), f) 0 min, g) 2 min, h) 30 min,
i) 120 min, j) 150 min (cured state). ................................................................................. 71
Figure 31: Stability maps of each cure cycle used. a) Ambient pressure within the core (cure
cycle I). b) Vacuum pressure applied to the core (cure cycle II). The time and
temperature for the initial void growth are 112 min/124 °C and 2 min/24 °C
respectively, ...................................................................................................................... 72
Figure 32: a) Simulated fillet shape and b) simulated volume of prepreg resin squeezed out of
the facesheet for each sample. .......................................................................................... 74
Figure 33: Simulated time-dependent development of bond-line porosity for each sample. The
porosity considered is the effective porosity, which is the weighted average of the
porosity in the adhesive itself and bled prepreg resin. a) Sample A, 4-ply facesheet,
ambient core pressure (cure cycle I), b) Sample B, 4-ply facesheet, vacuum core
pressure (cure cycle II), and c) Sample C, 8-ply facesheet, vacuum core pressure (cure
cycle II). ............................................................................................................................ 77
Figure 34: Cross-sections for each sample fabricated. a) Sample A, 4-ply facesheet, ambient
core pressure, b) Sample B, 4-ply facesheet, vacuum core pressure, and c) Sample C,
8-ply facesheet, vacuum core pressure. ............................................................................ 79
Figure 35: Measured bond-line porosity and simulation results for each sample fabricated. ...... 80
Figure 36: Schematic of sandwich panel layup. Laminates were fabricated using the same
layup procedure, including consumables, but omitting the adhesive and core insert. ...... 88
ix
Figure 37: Sample measurements of tow shearing. To measure tow shearing, individual
sections created by the fabric weave were outlined and the circularity measured.
Values were then compared to a baseline established using the vacuum laminate
sample. .............................................................................................................................. 90
Figure 38: Surface images of autoclave-cured samples: A, baseline vacuum laminate; B,
baseline vacuum sandwich; C, vented bag laminate; D, vented bag sandwich; E, in-
bag pressurization sandwich; F, in-bag pressurization sandwich with crushed core; G,
long room-temperature vacuum hold sandwich; H, no intermediate dwell sandwich.
Voids were overlaid with black for better visibility. ........................................................ 92
Figure 39: Surface images of baseline vacuum panel under the honeycomb core (top) and in
the flange (bottom). ........................................................................................................... 93
Figure 40: Surface void area content (top) and circularity (as deviation from the baseline
laminate average, bottom) for autoclave-cured samples................................................... 95
Figure 41: Cross-sections showing internal structure of tool-side facesheet and bond-line of
autoclave-cured samples: A, baseline vacuum; B, vented bag; C, in-bag pressurization;
D, in-bag pressurization with crushed core; E, long room-temperature vacuum hold;
F, no intermediate dwell. .................................................................................................. 97
Figure 42: Surface vs. internal porosity (top) and circularity (as deviation from baseline
laminate) vs. maximum facesheet thickness (bottom). ..................................................... 98
Figure 43: Frames from time-lapse video for baseline oven window cure test, with times
marked along with measured temperature and modeled viscosity. At t1, entrapped air
was observed moving toward pinholes in the surface ply and disappearing from view.
At t2, new voids began forming, with some remaining trapped in the surface after cure.
......................................................................................................................................... 101
Figure 44: Surfaces of oven window cure parts: baseline vacuum (top) and no room-
temperature vacuum hold/no intermediate dwell (bottom). Surface void area content
for oven window cure parts compared to autoclave-cured equivalents. Note that, for
the autoclave-cured sample, the no-dwell case did include a room-temperature vacuum
hold. ................................................................................................................................ 102
Figure 45: Prepreg made with de-wetted resin film to create a discontinuous resin pattern,
providing through-thickness evacuation channels to increase gas permeability. The
resin appears as dark areas. The white squares are pinholes in the fabric weave. .......... 113
Figure 46: Example of an in-progress layup using pressure probe assembly incorporated into
the laminate for resin pressure measurements. The needle is inserted between plies,
and a ball of excess resin is placed at the tip of the needle to seal the interface upon
initial application of vacuum. The needle of the pressure probe is passed through the
bag sealant tape to avoid introducing bag leaks.............................................................. 115
x
Figure 47: Frames from in situ tests for Test C:B/B (left; continuous resin, breathing edge
dams, perforated film), Test D:B/B (center, discontinuous resin, breathing edge dams,
perforated release), and Test D:S/S (discontinuous resin, sealed edges, semi-
permeable membrane). Times correspond to those marked on the thermal cycle
(bottom)........................................................................................................................... 118
Figure 48: Internal images of cured in situ testing samples. The left column was produced
with continuous resin film, and the right with discontinuous film. Boundary conditions
were: a and e) breathing edge dams, perforated film; b and f) breathing edge dams,
semi-permeable membrane; c and g) sealed edges, perforated film; and d and h) sealed
edges, semi-permeable membrane. ................................................................................. 120
Figure 49: Comparison of internal void area content (%) for continuous and discontinuous
resin formats for each set of boundary conditions. Test labels correspond to those in
Table 1: either bleed (B) and sealed (S) conditions for the edges and for the bag-side
surface, respectively. ....................................................................................................... 121
Figure 50: Resin pressure measurements for commercial OoA prepreg under bleed (edge
breathing dams, perforated release film) and sealed (sealed edges, semi-permeable
membrane) conditions. Times t1 and t3 are the predicted onset and full impregnation
of the material, respectively [27]. At time t2, measured pressure deviates between
boundary conditions: the bleed test continues losing resin pressure, while pressure
equilibrates under sealed conditions. .............................................................................. 123
Figure 51: Resin pressure measurements for the non-commercial resin under bleed (edge
breathing dams, perforated release film) and sealed (sealed edges, semi-permeable
release membrane) conditions for both continuous and discontinuous resin formats.
Time t1 is the peak in resin pressure, independent of boundary condition. At time t2, a
second drop in resin pressure was observed under bleed conditions but not sealed
conditions. Time t3 is the end of the first temperature ramp, during which void growth
at the tool surface was observed in in situ tests. ............................................................. 124

xi
Abstract
Carbon fiber-reinforced polymer composites consist of a cross-linked polymer (e.g.,
epoxy) matrix supported by woven or unidirectional carbon fibers. Offering increased mechanical
performance metrics such as stiffness at low weight compared to conventional engineering
materials, such composites are used in weight-critical applications such as aerospace.
Manufacturing defects such as porosity in the polymer matrix, however, can reduce mechanical
properties by serving as crack propagation sites. Traditionally, investigations into causes of and
mitigation strategies for porosity in CFRP composites – as well as process optimization and
qualification in industry – have relied on the analysis of cured samples. This process can be time-
and resource-intensive, and only gives insight into final porosity as an output in response to
material (resin type, solvent content, flow characteristics, etc.) and process (temperature, pressure,
consumables, etc.) parameters. In this work, two manufacturing cases are investigated using in situ
analysis methods to track real-time porosity development to clarify the underlying physical
mechanisms by which voids form, grow, and evacuate during composites processing. For the co-
cure of honeycomb core sandwich panels, in which prepreg facesheets are consolidated and
adhered to a low-density core in a single thermal cycle, a custom instrumented fixture was utilized
to visualize the development of porosity in the adhesive bond-line during cure. Behavior was
dependent on gas pressure within the core as well as interactions between the prepreg resin and
adhesive. Insights led to the development of an integrated, physics-based model for bond-line
formation, which experimental validation showed could reliably capture trends in porosity
development in response to material and process inputs. In situ visualization was also used to
assess void growth during cure at the prepreg-tool interface. For sandwich panels, real-time
imaging of the surface demonstrated that non-uniform compaction conditions imposed by the
xii
honeycomb core geometry led to the evolution of residual solvent that remained trapped as
porosity. In a second manufacturing case, in situ visualization was used to analyze void growth at
the prepreg-tool interface during out-of-autoclave processing of composite laminates. Varying
consumables were used to impose different boundary conditions, and results indicated that the
combination of discontinuous resin film and a semi-permeable release film on the bag-side surface
could evacuate initially entrapped air while retaining resin pressure to mitigate voids due to volatile
evolution. Through this work, in situ process analysis including visualization techniques clarified
time-dependent behavior and underlying physics of void formation during composites processing.
Insights gained can inform manufacturing decisions and model development, reducing time and
material waste and leading to more efficient manufacturing of high-quality composite parts.

1

Chapter 1: Introduction
1.1 Motivation
Composites including carbon-fiber reinforced polymers are prevalent in weight-critical
applications such as the aerospace industry due to their relative high strength and low weight and
therefore replacing traditional materials such as metals. For such high-performance applications,
CFRP parts are typically manufactured using prepreg, which consists of a fiber bed pre-
impregnated with a partially cured resin. Prepreg affords advantages including more controlled
resin content and easier manipulation for layup, compared to other methods (e.g., a wet layup, in
which a two-part resin is mixed and applied to dry fibers during layup).
After layup, prepreg parts are typically placed within a vacuum bag and cured under a
thermal cycle. The applied vacuum facilitates removal of air and other volatiles that can result in
porosity if not properly managed, while the bag itself provides compaction pressure on the
laminate that can further inhibit void formation and growth. To achieve reliable and robust
processing of defect-free parts, CFRP composites are often cured in autoclaves to supply additional
compaction pressure (~700 kPa) [1]. While effective for laminates, autoclave processing is still
prone to void formation when producing complex structures such as honeycomb core sandwich
panels. In such parts, discontinuous material elements result portions of the CFRP components not
being compacted under the full applied autoclave pressure.
Autoclaves also present a challenge to industry growth, as they are a large capital
investment, inefficient and expensive to run, and often a bottleneck in production. To meet
projected demand (e.g., for composite material in aircraft, Figure 1), out-of-autoclave (OoA)
methods relying on vacuum bag-only (VBO) compaction pressure and cured in an oven have been
2
developed. Limited to VBO conditions (compaction pressure limited to atmospheric pressure,
~100 kPa), OoA processing offers limited robustness and is prone to porosity.

Figure 1: a) Diagram showing composite usage in Boeing 787 Dreamliner. b) Projected demand for various aircrafts
between 2019 and 2040. Image source: The Boeing Company (www.boeing.com).
Process analysis employing in situ visualization methods affords a distinct advantage in
addressing such challenges facing the composites industry. Traditionally, visual analysis of defects
such as porosity relies on post-cure analysis, providing information only on the final state of the
part given the input parameters. In situ methods yield time-dependent visual data tracking process
phenomena including defect formation, providing insight into how and when voids form during
processing. Such information can inform manufacturing decisions and process modeling to
increase robustness during composite fabrication.
1.2 Honeycomb Core Sandwich Structures
Honeycomb core sandwich structures (HCSS) consist of 3 primary components:
1. Facesheets, which serve as the primary load-bearing component, are often
aluminum or CFRP laminates.
3
2. The low-density honeycomb core, consisting of hexagonal cells and typically made
of Nomex-dipped Aramid paper or aluminum, increases stiffness of the overall
structure by increasing the distance between facesheets (Figure 2).
3. An adhesive film – typically a thermoset polymer compatible with the composite
facesheet – bond the structure.

Figure 2: Relative stiffness and weight of laminate compared to sandwich panels with different core thickness.
Image source: [2]
Because they offer increased stiffness with minimal increased weight compared even to
CFRP laminates, HCSS are especially desirable in weight-critical applications. However, the
complex geometry introduced by the core creates challenging processing conditions. Specifically,
the discontinuous core imposes non-uniform application of compaction pressure on the facesheet
and adhesive bond-line (Figure 3). Where the facesheet is supported by the stiff core cell wall, the
full compaction pressure is transferred to the facesheet and adhesive. However, between the walls,
there is not solid support for the facesheet, and instead the applied compaction pressure is
4
dependent on the core gas pressure Pcore. As the core gas volume is typically isolated from the
vacuum bag by the two facesheets, this gas pressure will evolve during cure as a function of several
parameters (temperature, time, facesheet permeability, etc.) Unless the core is intentionally
pressurized (e.g., [3,4]), Pcore typically begins the processing cycle at ambient pressure and
decreases as core gas is evacuated during a room-temperature vacuum hold prior to heating,
resulting in an overall reduced compaction pressure applied to the facesheet/adhesive compared to
a laminate under the same autoclave pressure.

Figure 3: Pressure transfer in a honeycomb sandwich panel. At the stiff core cell walls, applied autoclave pressure is
transferred to the facesheet and adhesive. Within the cells, compaction pressure applied to the facesheet and
adhesive is dependent on the gas pressure within the cell.
HCSS consisting of CFRP facesheets are manufacturing following two general steps: 1)
the facesheets must be cured from prepreg and 2) the facesheets must be bonded to the honeycomb
core [1]. In secondary bonding, these steps occur separately. First, prepreg laminates are laid up
and cured in one thermal cycle. Then, an adhesive film is applied to each side of the core along
with the cured facesheets, and the adhesive film is cured in a second thermal cycle. The primary
5
advantage of secondary bonding is the curing of facesheets as laminates, which eliminates the
concern of reduced compaction pressure (although this can still be a concern for the adhesive bond-
line). However, for complex shapes, pre-curing the facesheets introduces challenges with fitting
facesheet to core. Further, manufacturing time is increased by requiring two steps, especially if an
autoclave is used for both.
To expedite fabrication of HCSS, curing of facesheets and bonding them to the core can
be combined into a single thermal cycle in a process called co-cure. In this process, uncured
prepreg and a layer of adhesive is laid up directly onto the honeycomb core insert, which also
facilitates the production of complex shapes. However, because the facesheets are cured over
discontinuous core, they can experience reduced compaction pressure compared to laminate
processing. Further, co-cure couples the curing of two separate thermosets, introducing potential
interactions between the materials.
1.3 Out-of-autoclave Processing
To address limitations of autoclaves – primarily their high cost and low throughput –
prepreg processing can also be done out-of-autoclave. In such processes, thermal cycles are applied
using an oven, with no autoclave pressure applied. Because compaction pressure is therefore
limited to that which can be imposed by the vacuum bag itself (atmospheric pressure, ~100 kPa),
OoA processing is prone to porosity from air entrapment and volatile evolution. To facilitate gas
removal, prepregs have been designed specifically for OoA processing by sandwiching dry fibers
between layers of resin [5–9] (Figure 4).
6

Figure 4: Top: schematic of out-of-autoclave prepreg design. Bottom: Scanning electron microscope image of OoA
prepreg showing dry fibers between resin and impregnated fibers. Image source: [9]
While OoA prepregs can yield high-quality, defect-free parts, the process has limited
robustness due to the nature of the prepregs. Because the evacuation pathways are sandwiched
between continuous layers of resin, OoA processing relies on in-plane gas removal and breathable
edge dams (Figure 5). As part size increases so does breathe-out distance, thus increasing the
required duration of a room-temperature vacuum hold to fully evacuate air entrapped in the
laminate. Evacuation pathways can also be blocked by features such as embedded ply drops,
trapping air in the part.

Figure 5: Typical layup with consumables for out-of-autoclave prepreg processing, including breathable edge dams
to enable in-plane air evacuation. Image source: [10]
7
To address limitations of conventional OoA prepregs and improve process robustness, a
format was introduced that uses discontinuous resin layers rather than continuous films (Figure 6).
These prepregs, called semipregs, have increased through-thickness gas permeability compared to
conventional OoA prepregs. Through-thickness air evacuation reduces breathe-out lengths relative
to in-plane evacuation, as part thickness is generally much smaller than length and width.
Additionally, semipregs have been demonstrated to be less sensitive to complex processing
conditions including embedded ply drops, sealed edges, and humidity conditioning [11,12].

Figure 6: Top: Schematic of semipreg format, with through-thickness air removal enabled by discontinuous resin
patterns. Bottom: Examples of semipregs with different resin patterns. Image source: [13]
Semipregs, to-date, have only been investigated at lab scale, and research has primarily
focused on improving air evacuation. However, semipregs are also susceptible to voids because of
the limited compaction pressure in OoA processing. Compaction pressure directly impacts the
hydrostatic resin pressure, which in turn either suppress void growth (if void gas pressure is less
than resin pressure) or enable void growth (if void gas pressure is greater than resin pressure) [14].
Because compaction pressure is limited to a maximum of atmospheric pressure, any reduction in
resin pressure could result in porosity formation.
8
A potential cause of decreasing resin pressure during processing is resin bleed from the
laminate, which has been documented in autoclave cure [14]. A fully impregnated prepreg (as in
conventional autoclave prepregs) can be modeled as a spring (representing the fiber bed) in a piston
filled with fluid (representing the resin). Initially, because the fiber bed is fully saturated by resin,
the applied compaction pressure is carried fully by the resin. However, as temperature increases
and resin viscosity decreases, resin can flow out of the system (via perforated release film, edge
breathing dams, etc.) until enough resin is lost such that some portion of the pressure is transferred
to the fiber bed. The decrease in resin pressure is proportional to the amount of resin lost.
In autoclave processing, detrimental effects of resin bleed can be offset by autoclave
processing. Conversely, bleed can also potentially be minimized by careful selection of
consumables. The limited compaction pressure possible in OoA processing renders the process
more susceptible to porosity if resin bleeds out of the laminate. Further, the edge breathing dams
that conventional OoA prepregs require for air evacuation are also a potential pathway for resin
loss. Therefore, there are limited options for altering consumables with the goal of minimizing
resin bleed. By enabling through-thickness air evacuation in addition to in-plane, semipregs offer
more flexibility in consumable selection than conventional OoA prepregs. To ensure robust OoA
processing of semipregs can be maintained in a commercial setting, porosity from both entrapped
air and from insufficient compaction/resin pressure must be mitigated.
9

Figure 7: Development of resin pressure during autoclave cure using a spring in a fluid-filled piston as an analog. In
the model, the spring represents the fiber bed, and the fluid represents the resin. In Step (3), a valve in the piston is
opened, allowing for fluid to bleed out of the piston and transfer pressure to the spring. Image source: [14]
1.4 In Situ Visualization Methods
Porosity is a common defect and major metric by which the quality of composite samples
is determined, but measurement of void content is typically restricted to post-cure analysis. From
this, final quality can be determined based on process inputs, but no insight into how or when voids
formed is provided. At a commercial level, reliance on post-cure analysis can inhibit
implementation of new materials or processes, as qualification is time- and resource-intensive. At
the lab scale, in situ visualization techniques have been employed to elucidate physical
mechanisms by which voids form and develop during cure (e.g., curing laminates on a windowed
oven [15,16] or using a custom windowed injection molding tool [17]). Such techniques provide
time-dependent visual data tracking the development of physical features throughout cure and have
been used to identify timing of void growth, understand how voids move within a laminate, etc.
In this work, two fixtures affording in situ visualization capabilities were employed to
elucidate aspects of HCSS co-cure and OoA processing. An in situ co-cure fixture [4] was utilized
to investigate the development of the adhesive bond-line during co-cure, with the aim of
10
developing a physics-based model to predict bond-line quality. Time-dependent visual data was
vital in understanding underlying physics of bond-line defect formation to guide model
development. To assess void formation at the tool-side surface of HCSS, a framed glass tool plate
that could be inserted into the window of an oven was employed. This fixture was also
implemented for a study on OoA processing to investigate the effect of consumable selection on
void growth in semipregs. Visual data was correlated to resin pressure measurements to attribute
observed void growth to resin bleed.
1.5 Dissertation Outline
This dissertation spans two research projects: 1) a multi-year effort to develop a physics-
based model describing the evolution of the adhesive bond-line during co-cure of honeycomb
sandwich structures and 2) an investigation into improving process robustness of out-of-autoclave
prepreg processing by combining semipregs with resin-impermeable consumables. In both
projects, in situ visualization methods were used to clarify time-dependent void evolution to
establish relationships between process parameters and void behavior, provide insight into to guide
model development, and propose and verify mitigation strategies to produce low-porosity parts.
Chapter 1 details the in situ visualization test fixtures employed throughout this work. Chapters 2-
5 discuss co-cure of HCSS, from foundational experiments through model development and
validation, as well as a related study on surface quality. Chapter 6 describes the study on OoA
processing.
Chapter 1: The in situ visualization fixtures used are detailed. A co-cure fixture enabling
visualization of the adhesive bond-line in honeycomb core sandwich structures was employed in
Chapters 2-4. A framed glass tool plate that could be placed in a windowed oven to assess the
development of surface quality during cure was used in Chapters 4 and 5.
11
Chapter 2: We utilize a fixture enabling direct visualization of the adhesive bond-line
during co-cure to investigate the effect of the processing parameters on defect formation.
Specifically, constant core pressures were imposed to establish a relationship between applied
pressure and bond-line evolution. Results indicated a non-monotonic trend in porosity with
changing core pressure. A model was also developed to predict when void growth would initiate
for a given temperature and pressure cycle, and validation tests were conducted to confirm model
predictions and demonstrate the path-dependent nature of bond-line formation.
Chapter 3: Continuing the work detailed in Chapter 2, we apply the developed model to
screen varying temperature cycles in addition to pressure cycles. The model was expanded to
include a simple description of extent of void growth. Both model and experimental results
indicated that an extended intermediate dwell can eliminate porosity by gelling the adhesive prior
to the onset of void growth. Results also showed that increased extent of void growth does not
always yield increased porosity, as the model did not account for voids bursting.
Chapter 4: In collaboration with co-authors from the University of Delaware, we validated
an integrated co-cure model. Individual sub-models such as for fillet formation were described and
validated individually. Then, the integrated model was validated using tests to assess sensitivity to
core pressure and to facesheet thickness. While validation data agreed with individual sub-models,
the complexity and stochasticity of the co-cure process made precise numerical predictions of
bond-line quality (in terms of fillet size and void content) difficult. However, the integrated model
captured trends in porosity in response to changing variables and could be used to inform
processing decisions to mitigate porosity.
Chapter 5: We investigated the effect of varying pressure conditions on the development
of surface porosity on the tool-side facesheet of honeycomb core sandwich structures. Pressure
12
was controlled both explicitly (by changing applied bag pressure) and implicitly (by changing
duration of holds during the thermal cycle). Results indicated that porosity could be mitigated by
suppression (impose super-ambient bag pressure) or evacuation (extended duration of room-
temperature vacuum hold). Tests were also conducted using a glass tool plate to directly observe
the surface during cure, with results showing that voids generally evacuated into the core but could
be trapped if the adhesive gelled.
Chapter 6: We studied the effect of consumables on void development in OoA processing
of semipregs, employing the framed glass tool plate to observe porosity at the surface directly.
Specifically, a semi-permeable membrane was used in place of a conventional perforated release
film to retain resin content while allowing for gas evacuation, and such a configuration yielded
negligible porosity. To confirm that observed porosity was due to resin loss and not air entrapment,
a probe was used to measure resin pressure in situ. Data was correlated to visual data and confirmed
that, when consumable allowing for resin bleed to occur, void growth was observed in conjunction
with a measured decrease in resin pressure.

13
Chapter 2: Experimental Methods for In Situ Visualization
1. Introduction
In this work, in situ visualization methods were employed to provide real-time data
describing the evolution of physical features (e.g., voids, fillets, resin flow) during cure. Such
methods can be used to elucidate the underlying physics for key processing phenomena. In
contrast, post-cure analysis provides data only for the final state of the sample, but gives no
information regarding how that state was reached.

Figure 8: Schematic of the in situ co-cure fixture, or “mini-autoclave”. Honeycomb core and a single facesheet are
laid on a glass window, enabling visualization of the bond-line during cure.
We utilized two fixtures designed to collect in situ visualization data for different
applications related to CFRP productions. The in situ co-cure fixture was designed to enable direct
visualization of the adhesive bond-line during co-cure of honeycomb core sandwich panels,
providing insight into fillet formation and porosity evolution. An oven window frame tool was
used in conjunction with an oven with a window to observe the tool plate/surface ply interface,
which was used to investigate both sandwich structure surfaces and OoA prepreg laminates.
14
2. In Situ Co-Cure Fixture
The in situ co-cure fixture, or “mini autoclave” (MA, Figure 8) was designed to simulate
autoclave conditions typical of HCSS production while enabling direct visualization of the
adhesive bond-line during cure [18]. The bond-line region of a sandwich structure is normally
obscured during cure by the autoclave, consumables, and two facesheets. Samples fabricated using
the MA, however, consisted of a single facesheet adhered to the honeycomb core insert, thus
affording a direct view of the bond-line through the open side of the core (Figure 9).

Figure 9: a) Schematic view of camera perspective of a single honeycomb core cell via mini-autoclave. b) Sample
frame from a test run using the MA fixture.
The fixture consists of an aluminum base with a recessed pocket into which the honeycomb
core is placed. The bottom of the core pocket includes a glass window, and glass spacers can be
inserted into the pocket to accommodate various core thicknesses while enabling visualization of
the bond-line.
15
An aluminum lid can be bolted onto the base, and the enclosed cavity can be pressurized
to simulate autoclave conditions. Both the lid and the base feature integrated heating elements with
separate controls. Autoclave, vacuum bag, and core pressures (through ports in the core pocket)
can be controlled through separate ports and measured with separate pressure sensors. Temperature
is measured and recorded using four thermocouples within the fixture.
The MA accommodates samples with a core size of 76 mm × 76 mm (3 in × 3 in) and
thickness of up to 19 mm (3/4 in, with thinner cores requiring a glass insert). Reducing the length
or width of the core results in at least one edge not being flush with the pocket, introducing the
potential for edge effects (e.g., resin bleed, facesheet sagging) that can affect the sample as a whole.
The MA enables tests to be conducted under two different core pressure conditions that
simulate different relevant manufacturing cases. In the most general case, the core gas volume is
isolated from the vacuum bag by two uncured prepreg facesheets, with the gas pressure within the
core evolving throughout cure (as a function of facesheet permeability, time, temperature, pressure
gradient, etc.). This can be simulated in the MA by sealing the core pressure port. The core gas
pressure can also be controlled directly in the MA and typically was equilibrated with the vacuum
bag pressure to simulate cases in which a pre-cured and perforated tool-side facesheet is used, thus
providing a direct pathway for gas to travel between the vacuum bag and core. In this work, the
latter configuration was used primarily to establish direct relationships between imposed pressure
and bond-line development.
3. Oven Window Frame
The oven window frame (OWF) fixture (Figure 10) was designed to enable direct
visualization of the tool-side surface of samples (both HCSS panels and laminates) to assess void
development at the tool-prepreg interface. The OWF was used in conjunction with a windowed
16
oven (Blue M), with the frame offset from the oven’s window to enable air flow around the glass
tool plate and therefore minimize thermal gradients (as opposed to using the oven’s window as the
tool plate).

Figure 10: Diagram of oven cure set-up for in situ visualization of the tool-side surface using the oven window
frame.. The use of the framed glass tool plate reduced thermal gradients (compared to placing samples directly on
the oven window). A digital microscope is placed outside the oven to observe the tool-side facesheet through the
oven window and glass tool plate.
Procedures for the OWF fixture generally follow those for curing on a conventional tool
plate. The glass tool plate is 304.8 mm × 304.8 mm (12 in × 12 in), and part size is limited only
by the need to fit the part and any necessary consumables within the inside boundary of the frame.
Typically, the maximum part size used was 152.4 mm × 152.4 mm (6 in × 6 in). Conventional
17
consumables can be used, although any release medium should not obscure the view of the
composite surface through the glass tool plate. The frame is placed vertically into the oven door,
and while it features adjustable legs to secure the frame in place, care was taken when attaching
the vacuum hose to ensure it did not pull the frame out of place.
The fixture is limited to use in the windowed oven and therefore restricted to VBO
processing conditions. When employed for sandwich panel tests, varying pressure conditions that
could be applied in an autoclave were not available, and so the effect of pressure on void
development on the tool-side facesheet surface could not be studied. However, time-dependent
void development for a baseline OoA cycle was used to inform results of autoclave processing
trials that included varying pressure conditions.
18
Chapter 3: Path-Dependent Bond-line Evolution in Equilibrated Core
Honeycomb Sandwich Structures
1. Introduction
In this study, we clarify the physical mechanisms by which porosity forms in the adhesive
bond-line during co-cure of honeycomb sandwich structures, and assess the dependence of these
mechanisms on material and processing parameters. Additionally, we use the insights gained by
these observations to develop a simplified predictive model for defect formation.
Co-cure of honeycomb sandwich panels reduces manufacturing time for sandwich
structures by combining two processes – facesheet consolidation/cure and facesheet-core bonding
– into a single step [1]. However, by coupling facesheet consolidation with adhesive bond-line
formation introduces complex phenomena and material interactions (including multi-phase flow,
dynamic pressure gradients, dissimilar cure kinetics, and intrinsic material variations) that are not
well-understood. Developing an understanding of the physical mechanisms of defect formation
and creating predictive tools for co-cure phenomena can inform processing decisions to increase
manufacturing efficiency.
Prior research on the adhesive bond-line formation in sandwich structures has typically
focused on producing parts with superior mechanical properties. Grimes [19] identified fillet size
as the primary factor affecting bond-line performance in honeycomb core sandwich structures,
with larger adhesive fillets providing greater shear, peel, and flatwise tensile (FWT) strength.
These results were corroborated in studies that focused on determining the processing parameters
that influence bond-line formation, with different studies concluding that cure temperature [20],
compaction pressure and heating rate [21], and adhesive film thickness [22] strongly affect the
19
bond-line. Resin viscosity [23] and solvent content [24,25] were identified as important material
parameters in bond-line formation.
Several studies have investigated the effect of applied pressure on adhesive bond-line
behavior in various processing situations. Reducing the vacuum level (i.e., increasing the absolute
pressure) was shown to reduce bond-line porosity when bonding flat panels under vacuum [26,27].
For vacuum-bag-only (VBO) co-cure of sandwich structures, Nagarajan et al. [28] demonstrated
that vacuum level, as well as moisture content and core type, can have inconsistent effects on fillet
quality and mechanical performance (e.g., reducing the vacuum level increased peel strength for
standard aluminum core but decreased peel strength for vented core). Alteneder et al. [3] employed
a technique to impose super-ambient gas pressure within the core prior to cure, reducing facesheet
porosity and core crush during autoclave cure.
For the VBO co-cure of sandwich structures, Tavares et al. used an apparatus consisting of
a tool plate with a recessed pocket, enabling the measurement of core pressure during the cure of
“half sandwich” structures to characterize facesheet permeability [29,30]. Varying initial core
pressures were imposed using different perforation patterns in the adhesive and prepreg to vary
permeability [31], with intermediate average core pressures during processing yielding greater GIC
values. Kratz and Hubert [32–34] used a similar apparatus to characterize facesheet permeability
and develop a predictive model for core pressure during co-cure. Observations specific to the
adhesive primarily echoed two previous results: 1) large, void-free fillets were stronger than small
or porous fillets (characterized by peel strength [32]), and 2) at room temperature, the adhesive
film tended to act as a barrier to gas transport [34].
A common limitation in the works cited above is the reliance on post-processing analysis
of parts to optimize manufacturing methods, a practice which is resource-intensive and material-
20
specific. Further, by considering only the final state of the bond-line, the studies highlighted
provide limited insight into the underlying, time-dependent mechanisms by which bond-lines
form. To address these concerns, predictive, physics-based models requiring minimal material
characterization have been developed.
Rion et al. [35] developed a model for fillet height based on surface energy minimizations,
with contact angles for the adhesive-facesheet and adhesive-core interfaces as the primary inputs.
The model, however, is limited to secondary bonding, so prepreg resin bleed into the core can be
ignored, and void-free fillets are assumed. Similarly, Chen et al. [36] developed a predictive model
for the fillet size in sandwich structures produced with a self-adhesive prepreg, with prepreg
permeability and compaction as inputs (the characterization of which is detailed in a prior
publication by the same group [37]). However, the same assumptions as the Rion model still apply:
there is no interaction between prepreg and adhesive resin (here, because no separate adhesive is
used), and voids are not considered. These assumptions restrict the utility of the models for the
general co-cure case.
Likewise, void growth models are available, but assumptions limit the relevance to
sandwich structures. Kardos et al. [38] developed a diffusion-based model for the growth of a
water vapor void in a CFRP laminate, with others updating the model to improve accuracy [39–
41]. Préau and Hubert [42] adapted this diffusion-based model for an adhesive film used in repair.
However, factors such as geometric complexity and presence of two resins complicate similar
modeling for co-cure of sandwich panels.
We elucidate complex dynamic interactions between the physical phenomena during co-
cure by employing an in situ visualization tool [18]. This fixture affords direct observation of the
bond-line in a “half-sandwich” structure under conditions identical to processing, providing visual
21
data and insights that can be related to processing time and temperature, as well as autoclave, bag,
and core pressures. Parts were fabricated using an “equilibrated core” configuration (e.g., when
one facesheet is perforated [43,44]), in which core pressure could be controlled explicitly through
a direct path to the vacuum bag. Thus, observed time-dependent behavior could be linked to
specific pressures (as opposed to an initial or average value for an evolving core pressure, which
has typically been used in prior literature [3,31]). Results indicated a non-monotonic relationship
between void content and pressure, dependent on the adhesive viscosity during void growth for a
given pressure.
Mass loss as a function of both temperature and pressure was characterized for both the
adhesive and prepreg resin used in the study, and results informed in situ observations. Notably,
an increase in the rate of mass loss in the prepreg resin correlated well with observed void
formation. When considered together, in situ videos, polished cross-sections, and mass loss data
provided a consistent description of behavior for this material set during co-cure.
Finally, we used experimental results to guide model development to describe defect
formation in honeycomb sandwich panels. Both in situ visualization and mass loss data provided
necessary material parameters for a simplified model that could predict the onset temperature of
void growth (i.e., the model predicts when void growth begins, but does not describe time-
dependent evolution of void size) for a range of core pressures. The model was used to design two
staged pressure cycles, one predicted to suppress void growth and one to facilitate void growth.
Experimental validation of these cycles demonstrated the path-dependence of defect formation
mechanisms identified in this work. These results are a necessary first step towards further analysis
of the co-cure process through experimentation and model development.
22
2. Experimental Methods
Materials. The adhesive selected for this study was a modified epoxy (Henkel Loctite EA
9658 AERO) supplied as a film supported by a non-woven glass fiber (NWG), with an areal weight
of 320 g/m
2
. Models for the cure kinetics and viscosity were previously reported [45]. The
manufacturer-recommended cure cycle consists of a 60 min cure at 177 C, with a ramp of 2.2 –
4 C/min.
The prepreg used for facesheets in co-cured samples consisted of a plain-weave carbon
fiber fabric impregnated with a toughened epoxy designed for structural aerospace applications
(Hexcel Hexply 8552S), intended for autoclave cure. Several variants of the 8552 resin exist,
including the “S” version used in this work, which is manufactured using a solvent-based method
to achieve full impregnation. The standard resin (8552) has been characterized previously, and
published models for thermal properties, including cure kinetics and viscosity, are available
[46,47]. The models in this study were adapted from Hubert et. al [46] based on supplied 8552-1
resin film (non-solvated), which closely approximates the kinetic and rheological behavior of
8552S prepreg. A separate study on the solvated prepreg has confirmed the presence of residual
solvent (methyl ethyl ketone [MEK]) using Fourier-transform infrared spectroscopy [48].
The core used was a phenolic-coated Nomex honeycomb (Gill Corporation HD132), with
3.2 mm (1/8 in) hexagonal cells, 12.7 mm (1/2 in) thickness, and a density of 48 kg/m
3
(3 pcf).
These materials are typical of those used for autoclave cure of honeycomb sandwich panels.
Sample fabrication and in situ visualization. To monitor the formation of the bond-line in
situ, the mini autoclave fixture detailed in Chapter 1 was used. Samples generally consisted of a
honeycomb core (76 mm  76 mm) and a facesheet and adhesive film (both 102 mm  127 mm).
For samples with aluminum facesheets, the aluminum was abraded with 60-grit sandpaper and
23
cleaned with acetone. Co-cure samples were assembled with four plies of prepreg ([0°/90°]2s).
Before layup, a release agent was applied to the surface of the fixture base and recessed pocket.
The core was placed into the pocket, and the adhesive and facesheet were placed overtop, leaving
an edge-band over the tool surface. Two layers of sealant tape were placed around the edges of the
facesheet to restrict air transport in the through-thickness direction, simulating the center of a large
part. A perforated release film, layer of breather, and vacuum bag were overlaid on the prepreg
facesheets. This layup configuration simulated the bag-side facesheet of a typical autoclave-cured
sandwich structure.
The cure cycle for all tests consisted of a 60 min dwell at 110 C and a 120 min dwell at
177 C, with a 2 C/min heating rate, per the manufacturer-recommended cycle for the 8552S
prepreg. No room-temperature vacuum hold was included prior to heating. Prescribed vacuum bag,
core, and autoclave pressures were imposed before the temperature cycle was started and held for
the duration of cure. In every test, the vacuum bag and core pressures were equilibrated to eliminate
gas transport through the facesheet and adhesive, and to limit void content to initially-entrapped
and evolved gases. Dynamically-changing features (such as facesheet consolidation), however,
could not be controlled.
For this study, samples were fabricated for 11 different material and core pressure
conditions, as listed in Table 1. Pressures are given in absolute units, and the compaction pressure
is defined as the difference between the autoclave and bag pressures. In this study, we focused on
the effect of core pressure (i.e., the gas pressure to which the adhesive was exposed) on
development of the adhesive bond-line. The tests were divided into three categories depending on
the material and processing parameters: (1) bonding to an aluminum facesheet; (2) co-cure with
24
sub-ambient or ambient pressure in the vacuum bag and core; and (3) co-cure with super-ambient
pressure in the vacuum bag and core.

Table 1: Testing parameters for samples manufactured in the in situ co-cure fixture. All pressures are absolute.
Test Type Core/Bag
Pressure (kPa)
Autoclave
Pressure (kPa)
Compaction
Pressure (kPa)
1 Bonding 0 101.3 101.3
2 Bonding 50.7 101.3 50.7
3 Bonding 101.3 239.2 137.9
4 Co-Cure 0 101.3 101.3
5 Co-Cure 25.3 101.3 76
6 Co-Cure 50.7 101.3 50.7
7 Co-Cure 76 101.3 25.3
8 Co-Cure 101.3 239.2 137.9
9 Co-Cure 152 239.2 87.2
10 Co-Cure 202.6 239.2 36.6
11 Co-Cure 253.3 377.1 123.8

Aluminum-bonded samples were fabricated at full vacuum (< 5 kPa), half vacuum (50.7
kPa), and ambient (101.3 kPa) core pressures. Because an impermeable aluminum facesheet was
used, a vacuum bag was not necessary to eliminate gas transport through the facesheet. However,
for the vacuum and half-vacuum tests, a bag was used to provide compaction pressure to ensure
intimate contact between the adhesive film and the core. Autoclave pressure was set to 101.3 kPa
(i.e., vented to ambient pressure) for simplicity. For the ambient pressure test, there was no
compaction pressure applied as a result of bag pressure, so autoclave pressure was set to 239.2 kPa
25
(20 psig, typical of autoclave pressures during co-cure) to ensure contact between the adhesive and
the core.
The sub-ambient co-cure tests included five pressure levels, ranging from full vacuum to
ambient pressure at increments of ~ 25.3 kPa (0.25 atm). One sample was fabricated for each
testing condition. To assess reproducibility, two additional samples were produced for select
conditions: aluminum bonding at 0 and 101.3 kPa and co-cure at 0, 101.3, and 253.3 kPa. Core
pressure was the focus of this study, and previous literature has shown it to be a dominant
parameter for porosity formation in the adhesive. Therefore, differences in compaction pressure
were neglected and autoclave pressure was set to 101.3 kPa (i.e., vented to ambient) for simplicity.
However, to avoid a zero-compaction case arising from ambient pressure throughout (bag, core,
and autoclave), an autoclave pressure of 239.2 kPa was used when the bag and core were set to
ambient pressure.
For the super-ambient co-cure tests, pressure in the bag and core was supplied using
nitrogen gas. Tests were conducted at 152, 202.6, and 253.3 kPa (1.5, 2, and 2.5 atm). To prevent
the bag from detaching from the tool surface, autoclave pressure was maintained at a level greater
than the bag and core pressure, and so a pressure of 239.2 or 377.1 kPa (20 or 40 psig) was applied
before pressurizing the bag and core to the specified levels. Otherwise, procedures for these tests
were identical to those for the sub-ambient tests.
Microstructural analysis. Samples fabricated in the in situ co-cure fixture were also
analyzed following cure to quantitatively assess the bond-line quality. Two sections (50 mm  25
mm) were cut from each sample – perpendicular to the ribbon direction and at the center of the
cells – and polished using a grinder-polisher (Buehler MetaServ). Measured fillet size may change
dependent on the location within the cell, and a constant plane was chosen so that fillets could be
26
compared directly. The polished sections were then imaged using a video microscope (Keyence
VHX-5000) to inspect the bond-line microstructure. Each section contained ~ 20 fillets, and thus
~ 40 fillets were evaluated for each test.
Image processing software (Adobe Photoshop) was used to assess the fillet quality, as
diagramed in Figure 11. Fillets were bound by a square with opposite corners set at where the
adhesive reached a 0° contact angle with the cell wall and the facesheet (or, if the contact angle
does not reach 0°, where the adhesive thickness reaches 0). This approximation was selected for
its combination of robustness, repeatability, and ease of use in the software. Fillet height was the
height of the square plus an adjustment for dimpling, which was defined as the difference between
the bottom edge of the square and the adhesive thickness at the inside corner (Figure 11B). This
definition of height was chosen for consistency with previous literature (e.g., [22,31]). Dimpling
was used solely as an adjustment factor for consistent measurement of fillet height and was not
itself analyzed in this study. For aluminum-bonded samples, dimpling was 0. Void area was
measured within the area bounded by the top and inside edges of the square, the cell wall, and the
facesheet. Therefore, voids contained within the adhesive layer not factored into the height were
still counted.
Resin outgassing. Outgassing of the adhesive (a factor contributing to porosity) was
analyzed by mass loss using thermogravimetric analysis (TGA, TA Instruments Q5000 IR). Tests
were performed at ambient and vacuum pressures with a heat cycle consisting of a 2 C/min ramp
to 350 C. The heating rate was selected to be consistent with the cure cycle used to fabricate half-
sandwich samples. The prepreg was also tested to assess whether residual solvent could be a factor
in defect formation observed in the in situ time-lapse videos. Sample size was ~ 5 mg.

27

Figure 11: Fillet measurement method for Test 1 (top) and Test 8 (bottom).
3. Results and Discussion
Sample fabrication and in situ visualization. Figure 12 shows frames from time-lapse
videos for bonding with an aluminum facesheet, recorded at times t1, t2, and t3 (temperature and
28
viscosity for these times are shown in Figure 13). Under vacuum, voids formed in the adhesive
(Test 1, t2). However, these voids burst or collapsed prior to gelation of the adhesive, redistributing
the adhesive onto the cells walls and resulting in small but void-free fillets (Test 1, t3). At ambient
pressure, the adhesive displayed negligible void growth, and bond-line formation was dominated
by viscous flow (Test 3, t2).
The half-vacuum test behaved similarly to the ambient case: some minor bubbling of the adhesive
was observed, but not enough to redistribute the resin through bursting of bubbles. In all cases,
nearly all activity occurred during the first ramp in the temperature cycle, and the bond-line was
largely stagnant during the first dwell and through the remained of the cure.

Figure 12: Time-lapse video frames for aluminum bonding under vacuum (Test 1) and at ambient pressure (Test 3).
t 1 is the initial state, t 2 is halfway through the first temperature ramp, and t 3 is the end of the first temperature ramp.
29

Figure 13: Time-lapse video frames for co-cure tests, with the temperature profile and modeled viscosities of both
the prepreg and the adhesive. Times are denoted on the temperature plot.

30
For co-cure tests, frames from time-lapse videos were recorded at times t1, t2, t3, and t4, as
shown in Figure 13. During the first temperature ramp, co-cure samples behaved like the aluminum
facesheet counterparts. Under vacuum, the adhesive foamed and redistributed onto the cell walls
(Test 4, t2 and t3). At ambient core pressure, fillets were observed to form by adhesive flow (Test
8, t2). Some clear prepreg resin, differentiated from the grey adhesive, flowed into the fillets from
the facesheet.
Behavior of the co-cure samples diverged from that of the bonded samples beginning with
the second temperature ramp. In the sample under vacuum, after the initial bubbling of the adhesive
during the first ramp, further void growth was negligible (Test 4, t4)). However, at ambient
pressure, voids grew during the second ramp (Test 8, t4) and remained trapped in the bond-line
after cure.
Time-lapse videos were compared to modeled cure kinetics and viscosities for the adhesive
and prepreg resins, indicating that void growth observed during the second ramp in the ambient-
pressure case occurred as the prepreg resin reached its minimum viscosity and the adhesive
viscosity began to increase. Increasing temperature and decreasing prepreg resin viscosity
facilitated volatilization or growth and transport of entrapped voids, and these voids inflated the
already-gelled adhesive that remained in the final bond-line.
Tests at partial vacuum levels demonstrated an approximately linear effect of pressure on
defect behavior, with observed void growth decreasing as core pressure was increased. At ambient
core pressure, voids grew in the bond-line following the gelation of the adhesive. However, as
pressure was decreased, some bubbles burst prior to gelation. Under full vacuum, nearly every
void ruptured before the adhesive gelled, resulting in a bond-line with negligible apparent porosity.
Two potential explanations exist for this: 1) from Henry’s law, for a given temperature, gas
31
solubility in a liquid is proportional to its partial pressure in the atmosphere the liquid is exposed
to, so lower pressures could cause evolution of dissolved volatiles in the adhesive or prepreg resin
prior to the gelation of the adhesive, and 2) decreasing the core pressure increases the difference
between the void gas and the core pressures, increasing the rate of void growth and thus the
likelihood that bubbles can grow and burst prior to adhesive gelation.
The sub-ambient tests described above demonstrated the effectiveness of vacuum in
reducing void content in the bond-line through the evacuation of entrapped air, volatiles, and other
potential void sources. In contrast, increasing the bag and core pressures beyond ambient level was
expected to decrease void content by suppressing evolution of dissolved gases in solution and
providing sufficient resin pressure to overcome gas pressure within voids to restrict growth.
An applied pressure of 253.3 kPa suppressed most porosity during the second temperature
ramp, although voids were not entirely eliminated (Test 11, t2, Figure 13). Some voids grew during
the first temperature ramp, attributed to air entrapped between prepreg plies during layup or at the
prepreg/adhesive interface that could not be evacuated due to the absence of applied vacuum.
However, during the second temperature ramp (t4), only minimal void growth was observed, which
had little effect on the shape of the bond-line. At 151.9 and 202.6 kPa core pressures, void growth
during the second temperature ramp was reduced but not fully suppressed.
32

Figure 14: Height (top) and void area (bottom) versus core pressure for aluminum bonding and co-cure.
Microstructural analysis. Data for the fillet heights and void areas is presented in Figure
14, and selected micrographs are shown in Figure 15. For bonding to aluminum, void area was
relatively low (~ 3% to 50% of corresponding co-cure values) for all pressures. At ambient core
pressure, no significant void growth was observed. Under vacuum, voids formed but collapsed or
burst before the adhesive gelled, providing a nearly void-free final bond-line. The half-vacuum
33
tests had roughly double the void content than the other bonding samples, but the values were still
low compared to co-cured samples. Average height was ~ 25% greater at ambient core pressure,
as the bursting of voids and redistributing of adhesive at lower pressures reduced fillet height (e.g.,
in Figure 15A, adhesive can be seen on the cell wall above where the adhesive contact angle with
the wall reaches 0º). At all core pressures, fillet height was low (~ 50% to 70%) compared to the
respective co-cure test (in which resin from the prepreg could bleed into the bond-line and increase
fillet size).

Figure 15: Micrographs for: a) Test 1 (aluminum bonding, vacuum), b) Test 3 (aluminum bonding, ambient), c) Test
4 (co-cure, vacuum), d) Test 8 (co-cure, ambient), e) Test 11 (co-cure, super-ambient).
For co-cure tests, results fell into three regions of quality (determined based on cited
literature, e.g., [19,20,23], with quality increasing as height increases and void area decreases)
34
(Figure 14). Region 1 consisted of tests at 0 and 25.3 kPa core pressures. Porosity was relatively
low, as voids that grew tended to burst or collapse before the adhesive gelled. However, collapsing
voids also led to reduced fillet height. In Region 2 (core pressures of 50.7, 76.0, and 101.3 kPa),
adhesive bubbles did not grow (and thus did not burst), and so fillet height was roughly twice that
of Region 1 tests. However, during the second ramp, volatiles from the prepreg were trapped in
the adhesive, resulting in large and numerous voids. Thus, void areas for Region 2 tests were
roughly five to ten times larger than for Region 1 tests.
Region 3 (super-ambient tests) produced fillets that were both tall and void-free. Like in
Region 2, core pressure during the first temperature ramp was sufficiently high to prevent the
growth and collapse of bubbles that resulted in reduced height in Region 1. Further, the core
pressure was sufficient to suppress volatilization of the prepreg during the second temperature
ramp, reducing void content relative to Region 2. These three regions demonstrate the non-
monotonic behavior of bond-line quality, a phenomenon that has led to inconsistencies in reports
of quality versus core pressure in prior literature.
Variability between fillets in individual tests was large, with relative standard deviation
ranging from ~ 10% to 60% for height and ~ 100% to 375% for void area. Despite the large
variability, results were reproducible. Qualitatively, in situ videos for repeated tests at each
condition demonstrated similar behavior. Additionally, for height and void area, deviation from
fillet to fillet within each sample was greater than the deviation between different samples.
Resin outgassing. Results for the TGA tests are presented in Figure 16. Due to high noise
levels in the data for the adhesive under vacuum, this data set has been smoothed (using a moving
average via the “smoothdata” function in MathWorks MATLAB 2017) for cleaner visualization.
Mass loss in the adhesive was low, ~ 0.4% at ambient pressure and ~ 1.0% under vacuum at 177
35
°C. The rate of mass loss was relatively constant within the processing window, although a slight
increase was observed around 150 °C when under vacuum.

Figure 16: TGA Results for EA 9658 AERO NWG adhesive (top) and 8552S prepreg (bottom). Data for the
adhesive under vacuum has been smoothed.
Up to ~ 110 °C, the prepreg behaved like the adhesive: mass loss was low (~ 0.5% at 110
°C and ambient pressure) and increased slightly under vacuum (~ 0.8% at 110 °C), while the rate
of mass loss remained constant. As temperature increased beyond 110 °C, however, the rate of
mass loss increased, peaking at ~ 160 °C. Under vacuum, this increase began at ~ 110 °C, while
at ambient pressure, the increase did not begin until 125 °C. Additionally, the degree to which this
accelerated mass loss occurred was greater under vacuum: mass loss in this temperature range was
36
~ 1.5% under vacuum and 0.5% at ambient pressure. Decomposition, marked by a second increase
in the rate of mass loss beyond the processing window, was estimated to begin ~ 225 °C.

Figure 17: FTIR spectroscopy data for the prepreg, with the thermal cycle (left) and mass loss (from TGA, right)
overlaid. The volatile observed beginning ~110 °C was identified as methyl ethyl ketone. Image source: [48]
The temperature range (110 °C to 225 °C) corresponds to the first dwell and second
temperature ramp (110 °C to 177 °C) of the cure cycle used to fabricate samples in the co-cure
fixture, and the TGA data was consistent with observations from the in situ videos and fillet data
obtained from micrographs. Voids trapped in the final bond-line tended to grow during the second
temperature ramp, which corresponds to increased mass loss in the prepreg. For the sub-ambient
tests, porosity increased as core pressure increased. Additionally, this temperature range matched
temperatures at which MEK was identified in FTIR spectra of vapors from the prepreg (Figure 17,
[48]), indicating that the increased mass loss and void formation observed was due to volatilization
of residual solvent in the prepreg.
The TGA data confirmed that lower pressures resulted in greater mass loss and an earlier
onset of the increased mass loss rate. At the dwell temperature (110 °C), volatiles were able to
evolve and evacuate under vacuum, but not at ambient pressure. In the corresponding in situ
videos, bubble growth and bursting was observed throughout the dwell when under vacuum. The
37
extent of bubbling decreased as pressure increased until, at ambient pressure, void growth did not
begin until the second temperature ramp, when the adhesive viscosity also began to increase. The
increased capacity of the prepreg to evolve and evacuate volatiles under vacuum, especially during
the first temperature dwell, resulted in lower volatile content available to evolve after the adhesive
gelled, thus reducing porosity in the bond-line.
4. Void Growth Modeling
Model development. The model employed was adapted from one published by Préau and
Hubert [42] for diffusion-based void growth in an adhesive used for repair applications. The
representative element consists of a single, spherical void surrounded by an infinite pool of resin,
such that 1) there are no interactions between bubbles, 2) the bubble size is small relative to the
bulk resin volume and does not affect the shape of the entire structure, and 3) the concentration of
volatile in the bulk resin is constant and uniform. The radius r of the void is given by
𝝏𝒓
𝝏𝒕
=
𝑫 𝒓 𝜷 (𝟏 +
𝒓 √ 𝝅𝑫𝒕 )
(1)
where D is a diffusion term and β is a “driving force” related to the volatile concentration gradient:

𝜷 =
𝒄 𝒃𝒖𝒍𝒌 − 𝒄 𝒗𝒔
𝝆 𝒈
(2)
where cbulk is the volatile concentration in the resin, cvs is the volatile concentration at the void
surface, and ρg is the density of the gas in the void.
By considering only when void growth begins, and not the size of the void, predictions
using Eq. (1) can be simplified. The onset of void growth occurs when
𝜕𝑟
𝜕𝑡
changes from negative
or zero to positive, which is determined solely by β. All other terms in Eq. (1) are always positive.
The condition for void growth, therefore, is
38
𝜷 = 𝟎 ⇒ 𝒄 𝒃𝒖𝒍𝒌 = 𝒄 𝒗𝒔
(3)
The bulk resin concentration is assumed to be constant. For moisture-based voids, this
concentration has been represented as a function of the relative humidity ϕ in which the material
is conditioned [42]:

𝒄 𝒃𝒖𝒍𝒌 =
𝒌 𝝆 𝒓 𝝓 𝟐 𝟏𝟎𝟎

(4)
where k is a proportionality constant representing solubility of the volatile in the resin, and ρr is
the resin density.
Similarly, concentration at the void surface is a function of the instantaneous partial
pressure of the volatile(s) in the void, which varies with temperature and applied pressure:

𝒄 𝒗𝒔
=
𝒌 𝝆 𝒓 𝟏𝟎𝟎
(𝟏𝟎𝟎
𝑷 𝒗 𝑷 𝒗 ∗
)
𝟐
(5)
Pv and Pv
*
are, respectively, the partial pressure and the saturated vapor pressure of the volatile in
the void. If the void is assumed to be pure solvent, Pv equals Pcore. Temperature dependence of the
saturated vapor pressure is given by the Clausius-Clapeyron equation, so Eq. (5) can be written as

𝒄 𝒗𝒔
= 𝟏𝟎𝟎𝒌 𝝆 𝒓 𝑷 𝒄𝒐𝒓𝒆 𝟐 (𝑷 𝒓𝒆𝒇 ∗
𝒆𝒙𝒑 [−
∆𝑯 𝒗𝒂𝒑 𝑹 (
𝟏 𝑻 −
𝟏 𝑻 𝒓𝒆𝒇 )])
−𝟐
(6)
𝑃 𝑟𝑒𝑓
∗
is the vapor pressure of the volatile species at reference temperature Tref (e.g., the standard
boiling point), ΔHvap is the latent heat of vaporization, and R is the universal gas constant.
For water as a volatile, as in [42], k can be characterized through humidity conditioning
resin samples, which is restrictive for non-water volatiles. Alternatively, by treating cbulk as a
material constant independent of conditioning, Eqs. (3) and (6) can be combined and solved to
give k:
39

𝒌 =
𝒄 𝒃𝒖𝒍𝒌 𝟏𝟎𝟎 𝝆 𝒓 (
𝑷 𝒓𝒆𝒇 ∗
𝒆𝒙 𝒑 [
−∆𝑯 𝑹 (
𝟏 𝑻 𝒐𝒏𝒔𝒆𝒕 −
𝟏 𝑻 𝒓𝒆𝒇 )]
𝑷 𝒄𝒐𝒓𝒆 )
𝟐
(7)
where Tonset is the temperature at which void growth begins. Using Eq. (7), k can be computed if
cbulk, the volatile species, and the onset temperature for a single core pressure are known. The onset
temperature of void growth can then be predicted for any given cure cycle.
Determination of model parameters. Computing solubility parameter k using Eq. (7) first
requires characterization of the volatile species causing the voids, the concentration of volatile in
the resin, and the temperature at which void growth begins for a given core pressure. The volatile
species for the prepreg used in the study has been identified as MEK [48], and the latent heat of
vaporization ΔHvap and standard boiling point Tref (where 𝑃 𝑟𝑒𝑓
∗
is atmospheric pressure) have been
previously reported [49].
Volatile concentration in the resin was estimated using TGA data presented in Section 3.
As discussed previously, two regions of mass loss were identified: first, up to ~ 110 °C, mass loss
attributed to moisture occurred, followed by an increased rate of mass loss due to volatilization of
MEK. The concentration of MEK in the resin was determined as the percent mass loss under
vacuum in this temperature range (110 °C to decomposition at 225 °C), which was 1.5%.
Multiplying by the density of the resin yields the volatile concentration in the desired dimensions
of mass per unit volume. This definition requires two simplifications: 1) all the residual MEK was
able to be evacuated within the temperature range while under vacuum and 2) stages of mass loss
do not overlap (i.e., mass loss in this temperature region was solely due to volatilization of MEK).
In situ time-lapse videos presented in Section 3 were used to determine the onset
temperature of void growth for a given core pressure. As different void behaviors were identified,
growth of individual bubbles was ignored, and void growth was defined to begin when foaming
40
behavior attributed to the solvent was observed (Figure 18). Because temperature could not be
measured directly in the bond-line, the temperature reported here was the average of those
measured on the outside of the bag and of the tool plate. For the initial calculation of k, Test 8 (co-
cure at ambient core pressure) was selected, and void growth was observed beginning at 117.3 °C
(temperatures recorded outside the bag and at the tool plate were, respectively, 118.1 and 116.5
ºC).

Figure 18: Comparison of individual bubble (left) and “foaming” behavior (right). Individual bubble growth was not
considered when determining the onset of void growth.
Results and discussion. Using the parameters determined above, a value of 1.26 × 10
-3

(dimensionless) was computed for k, which is within roughly one order of magnitude of the value
reported for water in an epoxy adhesive [42]. Note that, for small changes in temperature, k does
not vary significantly (e.g., the ± 0.8 °C range measured corresponds to a change in k of less than
± 5%). This value of k was then used to predict void growth onset temperatures for the remainder
of the co-cure tests presented in Section 3, with results presented in Figure 19. No experimentally
observed temperature could be identified for the two extreme cases. For the full vacuum test, a
core pressure of 1 kPa was used in the model to ensure a solution, and the predicted onset
temperature was below room temperature. In the in situ video, void growth occurred immediately
as the viscosity decreased (i.e., resin viscosity and not temperature was the limiting factor), and so
no reliable onset temperature could be determined. At 253.3 kPa core pressure, the onset
41
temperature predicted occurred while the adhesive was gelling, and no foaming was observed. For
the remaining tests, the model showed agreement with experimental data.

Figure 19: Experimental and model results for critical temperature as a function of pressure.
Pressure cycle design and validation. The model can be used to design pressure cycles to
mitigate void growth for a given temperature cycle. With k known, Eq. (7) can be rewritten to give
Pcrit as a function of temperature:

𝑷 𝒄𝒓𝒊𝒕 =
𝑷 𝒓𝒆𝒇 ∗
𝟏𝟎
√
𝒄 𝒃𝒖𝒍𝒌 𝒌 𝝆 𝒓 𝒆𝒙𝒑 [−
𝚫 𝑯 𝑹 (
𝟏 𝑻 −
𝟏 𝑻 𝒓𝒆𝒇 )]
(8)
Pcrit is the minimum core pressure necessary to suppress void growth at the respective temperature,
and any pressure below this level may cause void growth. Figure 20 shows this critical pressure
during the temperature cycle used to fabricate samples in Section 3, along with two sample
pressure cycles demonstrating the path-dependence of void growth. Each cycle consists of two
identical pressure levels (101.3 kPa and 253.3 kPa), with the timing of the pressure increase offset
by ~ 40 min. While the model predicts a critical pressure of 407.9 kPa at the maximum processing
temperature of 177 °C, this prediction does not consider adhesive or prepreg resin viscosity. Recall
that in Section 3, a core pressure of 253.3 kPa sufficed to suppress void growth prior to gelation
0 50 100 150 200 250 300 350 400
P ressu re [kP a]
0
50
100
150
200
C ritical Tem perature [ °C ]
M odel
E xperim ent
42
of the adhesive. Here, in Cycle A, the core pressure is increased prior to the second temperature
ramp so that the core pressure always exceeds the critical pressure. In contrast, for Cycle B, the
pressure increase is delayed to the middle of the temperature ramp, thus critical pressure increases
beyond the core pressure. Consequently, void growth is expected during this portion of the cure
cycle.
To validate the model predictions, parts were produced using both staged Cycles A and B,
following the procedures for samples fabricated in the in situ co-cure fixture detailed in Section 2.
During the initial stage of ambient (101.3 kPa) pressure in the bag and core, an autoclave pressure
of 239.2 kPa was applied (as in Test 8). To maintain constant compaction pressure, autoclave
pressure was increased to 391.2 kPa prior to raising the bag and core pressure to 253.3 kPa. In
Cycle A, the second stage began prior to the second temperature ramp, while in Cycle B, this stage
began during the temperature ramp (~ 150 °C).
In situ observations for Cycle A were comparable to those of Test 11 (with 253.3 kPa core
pressure). During the first temperature ramp and intermediate dwell, critical pressure remained
below the 101.3 kPa core pressure, and void growth was suppressed. Initially, in the second stage
of the pressure cycle, core pressure was greater than the critical pressure. As temperature increased
during the second ramp, the critical pressure surpassed the core pressure, but the adhesive viscosity
had also increased. Some void growth was observed in regions where the prepreg resin (which
gels after the adhesive) was exposed through openings in the adhesive.
43

Figure 20: Two sample pressure cycles demonstrating path dependence of void growth. In Cycle A, core pressure is
maintained at a level greater than the critical pressure, thus no void growth is expected. In Cycle B, the delayed
increase in pressure causes the critical pressure to increase beyond the core pressure, and void growth is predicted
during this period.
By delaying the pressurization to the middle of the second temperature ramp, Cycle B
yielded behavior similar to Test 8 (with 101.3 kPa core pressure). Critical pressure surpassed core
pressure while adhesive viscosity was low (~ 100 Pa·s), and void growth occurred. The increase
in pressure briefly resulted in a core pressure greater than the critical pressure, collapsing some of
the voids and resulting in reduced porosity compared to Test 8.
Void area data reflected trends observed in in situ videos. Cycle A was within an order of
magnitude of Test 11 (0.0032 mm
2
and 0.0076 mm
2
, respectively). Average void area for Cycle B
(0.0520 mm
2
) increased nearly sevenfold compared to Test 11, as voids grew. However, because
some voids collapsed when the core pressure was raised, this void area was a quarter that of Test
8 (0.2118 mm
2
). Overall, these results are consistent with predictions and demonstrate the utility
0 30 60 90 120 150 180 210 240 270 300 330
Tim e [m inutes]
0
20
40
60
80
100
120
140
160
180
200
T em perature [ °C ]
0
50
100
150
200
250
300
350
400
P ressure [kP a]
V oid G row th
V oid C ollapse
Tem perature
C ritical P ressure
A dhesive gel tim e
C ycle A
C ycle B
44
of the model in determining whether void growth will occur, despite quantitative predictions not
being available.
5. Conclusions
The time-dependent formation of voids in the adhesive bond-line during co-cure of
honeycomb sandwich structures was investigated using an in situ co-cure fixture, and bond-line
quality was correlated to gas pressure in the core during processing. Notably, the greatest porosity
levels occurred when process and material parameters led to entrapment of prepreg volatiles in the
gelled bond-line, and pressurizing the core beyond ambient level suppressed volatile evolution and
produced tall, non-porous fillets. Mass loss behavior was consistent with visual observations, with
TGA measurements (along with FTIR spectra in a separate study) indicating a greater volatility of
the prepreg compared to the adhesive due to residual solvent in the former. These results aided in
the development of a simple void growth model to predict the presence of porosity for given
processing conditions, and validation tests using staged pressure cycles showed consistency with
trends predicted by the model.
The results provide a physical description of bond-line formation during co-cure as well as
insights into mechanisms responsible for the final morphology. In situ visualization was
particularly valuable, affording understanding of the time-dependent behavior of the adhesive
bond-line that could be correlated to measured or controlled processing parameters. These
observations clarified the path-dependence of defect formation, with certain behaviors linked not
just to core pressure, but to timing within the processing cycle. For example, applying vacuum
pressure in the core led to fillet disruption early during cure, which would result in short fillets in
the final morphology regardless of the core pressure throughout the rest of processing. Likewise,
validation studies demonstrated that void growth could be avoided as long as core pressure was
45
always sufficient to suppress void growth, but increasing the core pressure after void growth began
did not necessarily eliminate porosity in the final part.
This study also highlights the importance of material selection for honeycomb sandwich
structures. For example, the primary defect source identified was the volatilization of residual
solvent in the prepreg resin. While this prepreg may be suitable for monolithic parts, the inadequate
transfer of pressure intrinsic to the discontinuous honeycomb core makes suppression of volatiles
difficult, indicating that a solvent-free, hot-melt prepreg would be more appropriate. The
difference in viscosity profiles – specifically, the gel times – of the adhesive and prepreg resin was
also shown to be a factor in void growth and so must be chosen appropriately. As the temperature
at which the adhesive gels (and thus restricts further void growth) decreases, the pressure required
to suppress volatilization of the prepreg solvent prior to adhesive gelation is also decreased.
Therefore, the selection of an adhesive that cures at a temperature lower than the prepreg resin
could be another strategy to produce defect-free parts.
The void growth model developed, while simplified, was able to guide processing
decisions to suppress void growth. In a manufacturing setting, such a tool could aid in designing
processing procedures that lead to defect-free parts and reduced material waste. The model we
presented requires reduced material characterization that is not restricted to water as the volatile
species, facilitating the implementation of the model for different adhesives and prepreg resins.
However, this model does not include any dependence on time, which restricts applicability only
to determining if voids will grow. A prediction of the extent to which voids grow (e.g., in terms of
bubble diameter or fraction of void area) requires a more robust model that includes time-
dependent factors such as diffusion (Eq. (1)).
46
We have focused on porosity formation specifically in the adhesive bond-line, but defects
in the facesheet and core will also impact the overall performance of a sandwich structure.
Furthermore, we only considered cases in which core pressure was controlled directly and
equilibrated with the bag, conditions not representative of the general co-cure case in which core
pressure evolves as a function of temperature and facesheet permeability. Insights gained here can
be integrated with these other aspects to describe the co-cure of honeycomb core sandwich
structures and develop guidelines for robust and efficient manufacturing more fully.

47
Chapter 4: Process Mapping for Defect Control in the Adhesive Bond-
line of Co-cured Honeycomb Core Sandwich Structures
1. Introduction
This study, published as a technical paper and presented at the 2019 Composites and
Advanced Manufacturing Expo, continued work detailed in Chapter 2. Specifically, we utilized
the model for the onset of void growth developed in Chapter 2, Section 4 to screen cure cycles to
produce co-cured sandwich panels with defect-free bond-lines. As in previous work, parts were
fabricated under equilibrated-core conditions (i.e., Pb = Pc), with a single core pressure condition
used while varying several temperature parameters (heating rate, intermediate dwell temperature,
intermediate dwell duration) to assess the effect of the temperature cycle on both predicted and
observed defect behavior. For brevity, redundant aspects are omitted, with references to
appropriate sections of Chapter 2 provided.
Viability of the selected cure cycles was determined in two ways. Qualitatively, favorable
processes were those in which either the prepreg or adhesive viscosity reached a sufficiently high
level prior to the predicted onset of void growth. A quantitative “void growth factor” – a function
of time, viscosity, and volatile concentration gradient – was developed, with lower values
indicating a lesser degree of void growth expected.
To validate model predictions, sandwich panels were made with the mini autoclave detailed
in Chapter 1 under each of the assessed temperature cycles. Porosity levels were obtained from
polished cross-sections. Qualitatively, in situ visualization experiments showed good agreements
with predictions. However, in some cases, porosity remained low despite expected void growth,
indicating a need for more robust modeling capabilities.
48
2. Experimental Methods
Materials and methods used in this study were largely the same as those detailed in Chapter
2, Section 2. These aspects are summarized here, along with cure cycle screening methods unique
to this study described in detail.
Materials. Henkel Loctite EA 9658 AERO NWG, a modified epoxy supported by a non-
woven glass mat, was the adhesive used in this study. The prepreg used was Henkel HexPly
ACP193PW/8552S, a plain-weave carbon fiber fabric impregnated with a toughened epoxy resin.
The “S” in the product name denotes a solvated tower manufacturing process, and residual solvent
has been identified in the prepreg [48]. Thermal properties of both resins, including cure kinetics
and rheological behavior, have been previously characterized. The core consisted of an aramid
honeycomb (Gill Corporation HD132) with 3.2 mm (1/8 in) hexagonal cells, 12.7 mm (1/2 in)
thickness, and a density of 48 kg/m3 (3 pcf).
Model development. The model used in this study was presented in Chapter 2, Section 4,
with derivation and characterization of parameters detailed. To simplify characterization, the
model predicts only when (i.e., under what pressure and temperature conditions) void growth will
begin, but does not track the evolution of the size of bubbles. This simplification reduces the
relevant equations to

𝜷 =
𝒄 𝒃𝒖𝒍𝒌 − 𝒄 𝒗𝒔
𝝆 𝒈
(8)
𝛽 = 0 ⇒ 𝑐 𝑏𝑢𝑙𝑘 = 𝑐 𝑣𝑠
(9)
where β is a “driving force” for void growth, cbulk is the volatile concentration in the bulk resin, cvs
is the volatile concentration at the void source, and ρg is the gas density inside the void. Void
growth begins when β transitions from negative or 0 to positive, or when bulk concentration
49
becomes greater than the concentration at the void source. The critical pressure at which this
transition occurs is given as

𝑷 𝒄𝒓𝒊𝒕 =
𝑷 𝒓𝒆𝒇 ∗
𝟏𝟎
√
𝒄 𝒃𝒖𝒍𝒌 𝒌 𝝆 𝒓 𝒆𝒙𝒑 [−
𝚫 𝑯 𝑹 (
𝟏 𝑻 −
𝟏 𝑻 𝒓𝒆𝒇 )] (10)
where 𝑃 𝑟𝑒𝑓
∗
is the vapor pressure of the volatile species at reference temperature Tref (e.g., the
standard boiling point), ΔHvap is the latent heat of vaporization, and R is the universal gas constant,
ρr is the density of the resin, and k is a solubility constant.
Cure Cycle Screening. For this study, various temperature profiles were assessed for
predicted effectiveness at mitigating porosity in the bond-line. The baseline cure cycle consisted
of a 60 min dwell at 110 °C and a 120 min dwell at 177 °C, with a 2 °C/min heating rate. Two
levels each for heating rate, dwell time, and dwell temperature were assessed, as detailed in Table
2. Only one parameter was changed at a time, with the baseline values used for the other two
parameters. Pressures were kept constant for every test, with Pauto of 239.2 kPa (2.36 atm), and
Pbag = Pcore of 101.3 kPa (1 atm).
Table 2: Parameters and corresponding levels used for cure cycle screening.
Parameter Baseline Low Level High Level
Ramp Rate (°C/min) 2 1 5
Dwell Time (min) 60 0 240
Dwell Temperature (°C) 110 90 130

Ability to alleviate void growth was assessed in two ways. First, viscosity profiles of the
adhesive were overlaid on temperature-pressure contour plots, with contours indicating values of
50
β. A temperature cycle was deemed effective if the adhesive gelled (viscosity μ = 10
6
Pa·s) prior
to β becoming positive for P = 101.3 kPa (1 atm).
Extent of void growth was also predicted using a “void growth factor” defined as

𝑽𝑮𝑭 = ∫ 𝜷 𝟏 𝝁 𝒅𝒕
𝒕 𝒈𝒆𝒍 𝟎 (11)
where tgel is the time at which the adhesive gels and μ is the viscosity. The void growth factor,
therefore, increases with increasing β and tgel and decreasing viscosity. If the adhesive gels prior
to β becoming positive, the void growth factor is 0 and no porosity is expected. In addition to the
parameter levels in Table 2, additional levels were calculated for each parameter to better assess
trends in the void growth factor.
Experimental validation. Parts were fabricated using each temperature cycle evaluated
(Table 2) to determine the accuracy of the model predictions. The mini autoclave detailed in
Chapter 1 was used, with the same procedures followed as in Chapter 2. All pressures (autoclave,
vacuum bag, and core) were imposed prior to and held constant throughout cure. Pbag and Pcore
were equilibrated to eliminate gas flow through the facesheet due to pressure gradients, and were
vented to ambient pressure (101.3 kPa). Pauto was set to 239.2 kPa (20 psig).
In situ time lapse videos were used to assess qualitatively whether void growth occurred
during cure. Further, cross-sections were cut, polished used a grinder-polisher (Buehler
MetaServe), and imaged using a video microscope (Keyance VHX-5000) to quantitatively analyze
bond-line quality. Void area fraction was measured for each fillet. 4 sections were assessed for
every sample, with each section containing approximately 20 fillets.
3. Results and Discussion
Process screening. Temperature-pressure maps and void growth factor results are
displayed in Figure 21. The process maps include viscosity profiles for both resins. However,
51
because the adhesive gels prior to the prepreg for the selected materials, the adhesive was
considered the limiting factor and so void growth factors presented are computed based on the
adhesive viscosity. Void growth factors have been normalized to the value for the baseline cure
cycle.
Viscosity profiles for both resins showed negligible dependence on the heating rate used,
and so no difference in void growth was expected based on the temperature-pressure map. The
void growth factor corroborated this prediction. A slight increase in void growth factor with
increasing ramp rate was predicted, but the variation was low for the ramp rates assessed.
Increasing dwell time was expected to limit porosity by gelling the adhesive prior to
predicted void growth. With no intermediate dwell, adhesive viscosity remained low (under 10
2

Pa·s) when β became positive. However, a 240 min dwell was predicted to nearly gel the adhesive
at 110 °C, prior to β becoming positive. This trend is also reflected in the void growth factor: as
dwell time increases, the void growth factor decreases, by two orders of magnitude between the
baseline (60 min) and 240 min.
At ambient core pressure, dwell temperature was not predicted to have a significant effect
on void growth. For each of the three levels investigated, viscosity was low as β became positive.
At 130 °C, the adhesive gelled rapidly during the intermediate hold, requiring a core pressure of ~
152 kPa to suppress void growth (compared to greater than 304 kPa for the baseline temperature
cycle). Void growth factor was lowest for the baseline case, and increased with both increasing
and decreasing temperature.

52

Figure 21: Process mapping (left) and void growth factor (right).
Experimental validation. Cross-sections of samples for each testing condition are displayed
in Figure 22, with void area fractions in Figure 23. Model predictions suggested a minimal effect
of heating rate on void growth, with void growth factor decreasing slightly with increasing ramp
53
rate. This was true for the faster rate, with volatiles inflating the fillets as in the baseline case
(Figure 22C), although porosity was twice that of the baseline case. Similarly, in situ visualization
for the slower rate showed voids growing. However, void shrinkage and collapse was also
observed, and the cured parts contained low levels of porosity (Figure 22B). In this case, the
increased time spent with the adhesive at low viscosity allowed voids to grow large enough to
burst or tear, evacuating volatiles and shrinking voids prior to gelation.
Experimental results for dwell time matched model predictions. The test with no
intermediate dwell showed similar behavior to the baseline case in the in situ time-lapse videos.
Porosity for this test was approx. twice that of the baseline case, consistent with the increased void
growth factor (~ 15 times greater than the baseline case). The extended dwell, as predicted,
eliminated most void growth (Figure 22E). Some was observed early in the dwell, possibly due to
variations in local temperatures. Further void growth was seen early in the second temperature
ramp, but only in areas where gaps in the adhesive could be seen. Where the adhesive was
continuous, gelled resin prevented void growth, resulting in the lowest measured porosity levels.
The 90 °C dwell test was consistent with model predictions, behaving similarly to the
baseline test. In both cases, void growth was expected to begin during the second temperature
ramp, and so except for the viscosity profiles being shifted, these tests should not deviate
significantly. The increased time spent at low viscosity in the 90 °C test did result in a higher void
growth factor (~ 10 times that of the baseline), and accordingly this test had ~ 1.5 times the porosity
of the baseline case.

54

Figure 22: Cross-sections of samples made under each processing condition. A) Baseline B) 1 °C/min ramp rate C)
4 °C/min ramp rate D) 0 min dwell time E) 240 min dwell time F) 90 °C dwell temperature G) 130 °C dwell
temperature.
In the 130 °C test, however, the intermediate dwell occurred after the onset of void growth,
and a higher void growth factor was predicted. This test demonstrated behavior similar to the 1
°C/min heating rate test: voids grew, as predicted, but shrank or burst prior to the adhesive gelling.
55
In the case of the higher dwell temperature, void growth occurred leading into the intermediate
dwell. During the dwell, adhesive viscosity increased slowly enough for existing voids to have
time to burst and shrink but quickly enough to gel the adhesive and prevent further void growth
during the second ramp. Porosity in the cured part was relatively low, about half that of the baseline
case, despite having a void growth factor ~ 15 times greater than the baseline. This increased void
growth factor is consistent with the observed growth-and-bursting behavior observed, but the void
growth factor does not specifically account for this phenomenon and provides no indication a
priori whether voids will collapse or be trapped in the bond-line.

Figure 23: Void fraction in fillets for each temperature cycle assessed. Error bars are ± one standard deviation of
average values for each section (four sections per condition).
4. Conclusions
A model enabling the screening of co-cure processes was developed, based on existing
models for diffusion-based void growth in composite laminates or bonding applications. The
framework of the model allowed for simplified characterization of required constants, as well as
56
the consideration of non-water volatiles. The model, however, showed an inability to predict the
extent of void growth (i.e., no output such as bubble size or percent porosity is provided).
Predictions identified increasing the duration of the intermediate dwell as most effective in
inhibiting void growth, which experimental tests verified. In this case, the adhesive gelled during
the dwell and prior to predicted void growth, preventing porosity in the bond-line during the second
temperature ramp. All other temperature cycles assessed predicted void growth prior to the
gelation of either resin. In situ visualization of the bond-line during cure confirmed that bubbling
did occur, but some cases still resulted in relatively low porosity levels. With a ramp rate of 1
°C/min or a dwell temperature of 130 °C, voids grew but collapsed prior to gelation. These cases
highlight the importance of timing in mitigating porosity. Ideally, a processing cycle will suppress
void growth in general. However, porosity can also be limited by providing sufficient time with
the resins at a low viscosity after void growth begins, so that volatiles can evolve and escape. In
the rest of the cases (4 °C/min ramp rate, 0 min dwell time, 90 °C dwell temperature), the adhesive
gelled too quickly after void growth began, trapping voids in the bond-line.
The results described here represent preliminary work to predict porosity in the bond-line
of honeycomb core sandwich structures during co-cure, specifically with the goal of identifying
processing cycles that inhibit void growth. The simplified model facilitated characterization of
model parameters and could successfully determine viable processing cycles under certain
conditions, offering a sufficient initial screening method. The model deviates from experiment in
cases in which voids grow and collapse prior to gelation of the adhesive. Increased void growth
factors for cycles such as the 130 °C dwell, which still results in reduced porosity, may indicate a
threshold between growth-entrapment and growth-collapse behaviors. However, a robust model
that can track bubble size and predict bursting and collapse would be able to capture the time-
57
dependent phenomena in these scenarios, providing improved screening capabilities for process
cycles.

58
Chapter 5: Experimental Validation of Co-cure Process of Honeycomb
Sandwich Structures Simulation: Adhesive Fillet Shape and Bond-line
Porosity
1. Introduction
Fabrication of honeycomb sandwich structures requires bonding carbon fiber-reinforced
polymer facesheets, characterized by high stiffness and strength, to both sides of a low-density
core. Fibers within the skins carry loads, while the core structure primarily increases the bending
moment of inertia of the assembly by distancing the facesheets from the neutral axis and resisting
shear loads. The inclusion of the core increases the stiffness of the overall structure with minimal
additional weight, resulting in widespread use of sandwich composites in aerospace structures
[50]. In general, producing honeycomb sandwich structures requires two steps: 1) curing of the
prepreg facesheets and 2) bonding of the facesheets to honeycomb core inserts. During co-cure,
these steps are combined, with uncured prepreg facesheets placed on both sides of the core inserts
to be cured and bonded to the core in an autoclave under a prescribed cure cycle specifying the
process parameters such as applied pressure, vacuum pressure and temperature as a function of
time [1]. The co-cure process is preferred, as it reduces processing time and resources.
Additionally, the use of drapeable, partially-cured prepreg, rather than rigid cured facesheets,
facilitates the production of complex-shaped parts. However, during the co-cure process, adhesive
fillet formation and facesheet consolidation occur simultaneously and under limited compaction
pressure (due to the risk of core crush). Consequently, the process is susceptible to defects such as
poor bonding, poor consolidation across the facesheet, and porosity within the bond-line (the
region between the facesheets and the core) [4]. Therefore, a predictive model that simulates the
59
co-cure process as a function of process parameters is potentially useful in reducing or eliminating
such defects.
The processes of fillet formation, facesheet consolidation, and porosity development in
composites processing are well-understood separately. For example, predictive models for fillet
size exist for both an adhesive film [35] and a self-adhesive prepreg [36]. However, both models
assume only a single resin (either by the use of a pre-cured facesheet or self-adhesive prepreg) and
void-free fillets. Niknafs Kermani et al. [51] proposed a model for the adhesive fillet shape of the
honeycomb sandwich structures depending on material contact angles, surface tension, density,
and the size of the honeycomb cell. Also, facesheet compaction has been modeled for laminate
composites (e.g., [52–56]), although the model geometry is not directly transferrable to the co-cure
configuration. Finally, Simacek et al. [57] presented a model applicable to the co-cure process to
simulate spatial and transient development of reinforcement deformation (fiber volume fraction),
resin pressure, and the volume of bleeding resin. Likewise, existing models for diffusion-based
void growth (e.g., [38–41]) assume boundary conditions relevant to laminates. The above models
have been demonstrated to be viable under the relevant circumstances (e.g., monolithic laminates
for void growth models, secondary bonding of sandwich panels for fillet shape models) but were
not developed with assumptions and boundary conditions appropriate for application to the co-
cure process.
The bond-line porosity is a critical property of sandwich structures, as it affects facesheet
adhesion. Consequently, the effects of various material and processing parameters on porosity in
the bond-line have been reported [20,22,23,25,28,58,59]. These studies identify resin viscosity,
solvent content, cure temperature, and adhesive film thickness as influential factors. Studies have
also identified applied pressure as a key parameter affecting bond-line development, with
60
Nagarajan et al. demonstrating that vacuum level has an inconsistent influence on fillet quality
during vacuum bag-only co-cure [28]. Alteneder et al. used an in-bag pressurization technique to
impose super-ambient gas pressure in the honeycomb core cells, reporting reduced bond-line
porosity as a result [60]. Moreover, to investigate time-dependent void behavior in the bond-line,
an in situ visualization method was employed, identifying core pressure and differences in the
prepreg resin and adhesive viscosity profiles as key factors determining development of porosity
in the bond-line [4,61].
The combination of models proposed by Niknafs Kermani et al. [51,59], for the porosity
development within the bond-line and adhesive fillet shape, and Simacek [57] for the facesheet
consolidation process, into a simple analysis tool [51,62] provides a comprehensive simulation
tool for the co-cure process of honeycomb sandwich structures that accounts for interactions
between material and physical phenomena. However, while the proposed models capture the
correct physics [51,62], they have not been experimentally validated.
This work outlines the validation of the integrated co-cure simulation, accounting for both
adhesive and prepreg behavior, as well as geometric considerations, with the desired output being
the bond-line porosity. To date, there is no simulation model that integrates the entire physics of
the co-cure process of honeycomb sandwich structures to predict bond-line porosity. The
sensitivity of the model to process parameters and the number of prepreg plies in the facesheet are
studied. Key sub-models are presented and compared with experimental data. The integrated
model for porosity is then described and the simulated porosity in three cases of co-cure cycles is
compared to experimental results. To isolate fillet formation, validation tests consisted of bonding
honeycomb core to an aluminum facesheet using an adhesive film. Results show reasonable
agreement with bond-line shape predictions despite variability in the measured contact angle. To
61
validate the porosity model, tests are conducted in the equilibrated-core configuration, in which
core pressure is equilibrated with bag pressure (such as when a pre-cured, perforated tool-side
facesheet is used [63]), at ambient and vacuum pressures. A fixture allowing direct observation of
the bond-line is utilized for the experiments [18]. The porosity model captures major phenomena,
including the timing at which void growth begins, as well as the escape of voids from the bond-
line under vacuum. Precise quantitative porosity prediction is difficult due to variability in
experiment as well as simplifications in various sub-models. However, the model captures trends
in porosity in response to changing core pressure and facesheet thickness, which can guide
manufacturing decisions to reduce porosity in the bond-line.
2. Materials
The materials selected for this study - prepreg, adhesive, and core – are typical of those
used for aerospace structures. The prepreg consisted of a plain-weave carbon fiber fabric
impregnated with a modified epoxy resin (Hexcel HexPly AGP193PW/8552S, where the “S”
denotes a solvated tower manufacturing process). Previously, residual solvent was identified in the
prepreg using Fourier-transform infrared spectroscopy [64], and the solvent was identified as a
source of porosity in the adhesive bond-line [61]. Thermal properties, including cure kinetics and
viscosity, have been characterized and modeled for the neat resin (e.g., [46,47]). Models used were
adapted from those developed by Hubert et al. [46] based on a supplied 8552-1 resin film that
behaves similarly to the prepreg resin.
The adhesive used was a modified, flow-controlled epoxy supported by a non-woven glass
mat (Henkel Loctite EA 9658 NWG). Thermal properties for the adhesive were characterized and
reported previously [45]. The honeycomb core consisted of phenolic-dipped Nomex (The Gill
62
Corporation HD132) with 3.2 mm (1/8 in) hexagonal cells, 12.7 mm (1/2 in) thickness, and a
density of 48 kg/m
3
.
3. Adhesive Fillet Formation
Model development. The model used for the adhesive fillet shape is a steady-state model
based on the Young-Laplace equation, which accounts for surface tension, gravitational force, and
the cell geometry [51]. The steady-state condition is assumed due to the relatively short
characteristic time of fillet formation compared to the characteristic time of the overall co-cure
cycle (hours). As the actual shape of the hexagonal honeycomb cell would be difficult to model
numerically, the fillet shape for two simplified cases (linear and radial wall) were investigated
(Figure 24a).

Figure 24: Geometry used in the model for fillet formation. a) Fillet geometry as modeled. b) Fillet
dimensions and contact angles. c) Parametrization of applied force.
The schematic of the fillet geometry is shown in Figure 24a. The curve is determined by
hydrostatic and surface forces, the volume of the fluid (adhesive resin), and contact angles α and
β. As the adhesive does not de-wet the prepreg surface, the angle α between the adhesive and the
facesheet is 0˚ and the width W has a maximum value of half the cell size but ultimately depends
63
on the total volume of adhesive. The developed equations to predict the fillet shape in linear and
radial cases and the corresponding boundary conditions are as follows [51]:

𝜸 𝑾 𝟐 𝒚⃛ − 𝒚 ˙
𝟐 𝒚⃛ + 𝒚 ˙ 𝒚 ̈ 𝟐 ( 𝟏 − 𝒚 ˙
𝟐 )
𝟑 /𝟐 = 𝒚 ˙ (linear) (12)

𝛾 𝑊 2
(
𝑦⃛ − 𝑦 ˙
2
𝑦⃛ + 𝑦 ˙𝑦 ̈ 2
( 1 − 𝑦 ˙
2
)
3/2
+
𝑦 ̈ ( 𝑊 − 𝑟 )− 𝑦 ˙√1 − 𝑦 ˙
2
( 𝑊 − 𝑟 )
2
) = 𝑦 ˙ (radial) (13)
𝑥 ( 0) = 𝑊 ( 𝑎 )
𝑦 ( 0) = 0 ( 𝑏 )
𝑥 ( 𝐿 ) = 0 ( 𝑐 )
𝑦 ̇ ( 0) = 𝑠𝑖𝑛 ( 𝛼 ) ( 𝑑 )
𝑦 ̇ ( 𝐿 ) = 𝑐𝑜𝑠 ( 𝛽 ) ( 𝑒 )

(boundary
conditions)
(14)
Here, 𝛾 =
𝜎 𝜌𝑔 𝑊 2
, which is the reciprocal of the Bond number. The coordinates are
parametrized by arc (s) from 0 to unknown length L. Dots denote the derivative with respect to the
arc [51]. It is shown that by solving Eqns. 12 and 13 for the same 𝛾 and boundary conditions, the
predicted fillet shape in linear and radial cases are similar [6]. Therefore, it can be concluded that
the fillet shape for the hexagonal wall configuration, which is between the two extents of the linear
and radial cases, would be similar to the predictions in these cases as well.
According to the boundary conditions for the model and definition of 𝛾 , the simulated fillet
shape is dependent on several material parameters:
1) The surface tension σ, which determines the curvature of the fillet, was measured to be ~
40 mJ/m
2
. The model is not sensitive to small changes in σ, as even a change in value by
25% results in only a small shift in the modeled fillet shape.
2) The contact angle between the adhesive and the core cell wall β was 34.7˚, as measured
from micrographs of cross-sections. This property strongly impacts the fillet height H.
64
3) The width of the fillet W varied from fillet to fillet and from sample to sample, so 1 mm
was selected as an upper bound on all measured data. It was shown to provide a reasonable
fit. Note that the selected value for W is less than the cell half-width.
Experimental methods. Fillet formation tests consisted of bonding honeycomb core (76 mm ×
76 mm × 13 mm) to aluminum facesheets (102 mm × 102 mm) using a layer of film adhesive (76
mm × 76 mm). As the purpose of these tests was to generate validation data for the independent
fillet formation model, testing conditions were chosen to most accurately satisfy key assumptions
of the model: 1) the facesheet is rigid, 2) the bond-line contains a constant volume of resin, 3) fillet
shape is a function of applied pressure, 4) the bond-line is void-free, and 5) fillet formation occurs
counter to the influence of gravity. Aluminum facesheets were selected to satisfy the first two
requirements, as prepreg would be flexible and potentially transfer resin to the bond-line.
Because the model is dependent on applied pressure (in this case, the gas pressure within
the core cells), that pressure must be known. To achieve this, samples had only a single facesheet,
leaving the second side of the core exposed. For all tests, the core cavity was vented to atmospheric
pressure (~ 100 kPa), which produced bond-lines with minimal porosity. The aluminum facesheets
were placed on the tool side of the layup (i.e., underneath the core), so that fillet formation occurred
upward.
65

Figure 25: a) Cross-section of adhesive fillet model validation sample. b) Sample of fillet height measurement.
After cure, sections (38 mm in length) were cut from samples, polished (Buehler
MetaServe), and imaged (Keyance VHX-5000). Two samples were assessed, with ~20 fillets
measured for each sample. Fillet cross-sections were analyzed using image processing software
(Adobe Photoshop CC) to obtain fillet height as a function of distance from the cell wall. Fillet
height was measured from the surface of the adhesive to the adhesive/facesheet interface. At the
point at which contact angle between the adhesive and the facesheet goes to 0°, the measured
height was defined as the adhesive thickness and subtracted from measured heights (this volume
of adhesive is not accounted for in the model). Height was measured at eight points in addition to
the adhesive thickness, so that nine total points ranging from the cell wall (highest point along the
fillet) to the end of the fillet (0 height, by definition) for each fillet. A sample cross-section along
with height measurement method are shown in Figure 25.
66
Because experimental data were not easily measured at specific and regular points along
the fillet, data were interpolated quadratically using the three closest points before comparison to
model predictions (Figure 26). Results showed predicted height to be slightly higher (0.82%) than
experimental values when using the measured contact angle β = 34.7°. Error was greater in Panel
B than A, likely caused by variability inherent in the process (e.g., variation in substrate
smoothness, non-uniform temperature, etc.) Model predictions, however, still fell within the error
range.

Figure 26: Model predictions compared to experimental data for fillet formation model.
4. Bond-line Porosity Development
The bond-line region considered for the purpose of modeling and validation is created
during the bonding of the prepreg facesheets and the honeycomb core. Generally, the bond-line
can be formed by an adhesive film layer or by the prepreg resin (through the use of a self-adhesive
67
prepreg). In the former procedure, which is focus of this work, resin squeezed out from the
facesheet may also be present in the bond-line region, along with the adhesive. The presence of
the squeezed-out resin from the facesheet has been observed previously [18]. Previous studies have
demonstrated the possibility – depending on processing conditions – of voids in the bond-line
growing and escaping into the core cell [61,65]. Experimental observations and the proposed
stochastic models [51,59], which relate void growth and the escape process, highlight the
importance of capturing the escape process in the overall porosity development. Void escape is
determined primarily by the bond-line shape, as well as the porosity level within the bond-line. A
flow chart illustrates the interactions of different physics during the co-cure process (Figure 27)
[59].

Figure 27: Coupling between physical phenomena that affect porosity development in the adhesive bond-line.
As shown in Figure 27, the porosity development within the bond-line is affected by
multiple factors:
68
1) Facesheet consolidation can impact the bond-line porosity through mass transfer of
either resin (and therefore solvent mass) or existing voids.
2) Fillet formation determines the shape of the adhesive volume within which bond-line
porosity exists, thus directly affecting the void content as well as the growth, transport, and escape
of voids within the adhesive.
3) Gas pressure within the core cells is the pressure applied to the bond-line and therefore
influences void growth via the adhesive resin pressure. Due to the relatively thin bond-line, the
resin pressure is assumed to be equal to the core gas pressure. Therefore, voids within the bond-
line are assumed to be exposed to the core pressure.

Figure 28: Left: Diagram showing possible motion of voids or dissolved volatiles, including transfer of voids from
the prepreg to bond-line region via resin bleed and escape of voids from the bond-line into the core. Right:
Schematic of void escape description in the integrated co-cure model. If a void is located near the interface between
adhesive and core gas pocket (Zone 2) and its radius is large enough such that the void touches this interface, the
voids is defined to escape into the core and is removed from the bond-line. Image source: [51]
In this work, only the equilibrated-core process is considered, in which a direct connection
between the core cells and the vacuum bag enables equilibration of the two volumes of gases. This
69
simplifies modeling by eliminating the need for a core pressure model (typically a function of
several parameters including temperature, time, facesheet permeability, etc.) and replacing it with
an explicitly controlled core pressure processing parameter.
The mechanisms of void growth have been studied [38] and can be categorized as void size
evolution due to 1) gas expansion based on ideal gas law behavior, and 2) diffusion-induced
growth. In the latter case – which has a greater influence on void size than gas expansion –
dissolved volatiles in the liquid phase surrounding the voids diffuse into the gas bubble under
appropriate pressure and temperature conditions. A stability map is proposed to identify pressures
and temperatures at which the diffusion-induced growth occurs.
Experimental methods. Experiments related to void growth consisted of two parts. First, in
situ observations of void growth in the bond-line were used to compare with the stability maps and
estimate the volatile concentration. Second, bond-line porosity values were measured from sample
cross-sections and compared to model predictions. In both cases, samples were prepared using the
mini autoclave fixture detailed in Chapter 1.
Three tests with varying processing parameters were considered. The temperature cycle
was kept constant for all cases and consisted of a 60 min dwell at 110 °C and a 120 min dwell at
177 °C, with a 2 °C/min heating rate. Samples A and B were fabricated, respectively, with the
vacuum bag and core pocket vented to ambient pressure (~ 100 kPa) with vacuum applied (< 5
kPa). For both tests, an autoclave pressure of 377.1 kPa was applied. All pressures were imposed
prior to the start of the temperature cycle and held throughout the cure. Both cure cycles are shown
in Figure 29.
70

Figure 29: Cure cycles I and II used for fabrication of validation samples for void growth model. a) Cure cycle I for
Sample A had the vacuum bag and core vented to ambient pressure. b) Cure cycle II for Samples B and C had
vacuum applied to the bag and core.
Layup for Samples A and B were identical and consisted of a facesheet (102 mm × 127
mm) with 4 plies of prepreg ([0°/90°]2s), a layer of adhesive film (102 mm × 127 mm), and a
Nomex core insert (76 mm × 76 mm). A third sample, C, was fabricated under the same vacuum
pressure conditions as Sample B, but with a thicker facesheet (8 plies, [0°/90°]4s). Otherwise, the
layup was identical to that of Samples A and B. Following cure, samples were sectioned and
polished to measure porosity in the bond-line.
In situ void growth results. Selected frames from time-lapse videos for Samples A and B
are shown in Figure 30, with times corresponding to cure cycles in Figure 29. The adhesive film
can be seen in both tests in the initial states (a and f). During the first temperature ramp, adhesive
temperature increases, and therefore its viscosity drops, and it flows to form the fillet. For Sample
A, as the cure cycle progresses into the second temperature ramp (d, at 112 min), voids were
observed to grow and inflate the fillets. These voids remained trapped in the adhesive after gelation
(e).

71

Figure 30: In situ images of the bond-line during cure. Images a-e are from Sample A and f-j from Sample B. Times,
which correspond to cure cycles in Figure 5, are a) 0 min, b) 30 min, c) 60 min, d) 112 min, e) 150 min (cured state),
f) 0 min, g) 2 min, h) 30 min, i) 120 min, j) 150 min (cured state).
On the other hand, when vacuum was applied to Sample B, void growth started early in the
cure cycle (~ 2 min), as temperature increased and therefore viscosity of both the adhesive and the
prepreg resin decreased (g). These voids grew and escaped by bursting at the surface, which
continued through the initial stage of cure (h). After some time (~ 70 min through the cure cycle),
the number of voids growing and escaping reduced, indicating the mass of dissolved volatiles had
escaped through the adhesive bond-line (i). No further void growth was observed for the remainder
of the cure cycle, and the final state did not display any inflation of the fillets due to entrapped
porosity (j).
The utility of a void stability map in the co-cure of sandwich structures was discussed
previously [66]. Moisture was treated as the main volatile in the materials to simplify modeling,
and a stability map was created to predict stable and unstable pressure and temperature conditions
for a given volatile concentration. Stability maps for cure cycles used for both Samples A and B
are shown in Figure 31. The corresponding cure cycles were mapped for each sample. Volatile
72
concentrations were estimated based on the temperature and pressure at which void growth was
observed to be triggered from the in situ time-lapse videos.

Figure 31: Stability maps of each cure cycle used. a) Ambient pressure within the core (cure cycle I). b) Vacuum
pressure applied to the core (cure cycle II). The time and temperature for the initial void growth are 112 min/124 °C
and 2 min/24 °C respectively,
The stability maps provided critical information and insight into the cure cycles used and
the impact on bond-line porosity. By analyzing such figures, the bond-line porosity of the
fabricated part could be predicted qualitatively without further computation. It was shown
previously that the diffusion-induced bubble growth is more intensive if the cure cycle advances
further in the unstable region [59]. Figure 31 illustrates that, in the cure cycle used for Sample A,
the voids within the prepreg resin remained within the stable region for most of the cycle, and only
experienced critical growth conditions for a relatively short period of time (at 124 min) before the
adhesive gelled (gelation time of the adhesive was at 130 min). Therefore, the voids within the
prepreg resin which are growing at this stage may get trapped under the gelled adhesive until the
73
prepreg resin gelation at 143 min. This potentially leads to the higher level of porosity within the
bond-line. Conversely, in the cure cycle used for Sample B, both the prepreg resin and the adhesive
enters the unstable growth condition in the early stage (at 2 min) of the cure cycle due to the
reduced pressure, providing sufficient time for void growth and escape which may result in low
final porosity. This qualitative comparison of cure cycles with the stability maps indicates that the
final porosity of Sample A will be higher than Sample B. This hypothesis is validated by the
experimental results in the following section.
Porosity simulation. The bond-line porosity simulation is the integration of multiple sub-
models; (i) the bubble growth model, (ii) the stochastic model for the bubble escape process, (iii)
the model for the adhesive fillet shape, and (iv) the prediction of the prepreg resin volume, and the
corresponding bubbles, bleeding into the bond-line. The simulation for the bubble growth requires
assumptions about the initial size of voids and initial porosity in both the prepreg resin and
adhesive, as well as the initial volatile concentration of each material. The parameters used are
summarized in Table 3. One percent porosity represents the minimal void content of the aerospace
grade materials typically used in the co-cure process of honeycomb sandwich structures. The initial
radius of voids in the adhesive is assumed to be 10 μm, selected based on prior studies [41]. The
initial radius of voids entering the bond-line from the prepreg resin is assumed to be 5 μm, with
the smaller size relative to the adhesive voids reflecting the size of pinholes in the prepreg fabric
that voids must pass through to reach the bond-line.
Table 3: Parameters used in the void growth model.
𝑅 0,𝐴𝐷
= 1 ∗ 10
−5
𝑚 𝜑 0,𝐴𝐷
= 1% 𝑅𝐻 0
𝐴𝐷
= 20%
𝑅 0,𝑃𝑅
= 0.5 ∗ 10
−5
𝑚 𝜑 0,𝑃𝑅
= 1% 𝑅𝐻 0
𝑃𝑅
= 45%

74

Figure 32: a) Simulated fillet shape and b) simulated volume of prepreg resin squeezed out of the facesheet for each
sample.
Simulation results for fillet shape and volume of prepreg resin that bled from the facesheet
for the three samples are shown in Figure 32. The predicted fillet shape was constant across all
samples, as the same temperature cycle was used, and parameters affecting fillet shape (surface
tension, contact angle, and density) vary with temperature, but not pressure. The volume of prepreg
resin that bled from the facesheet varied due to differences in pressure gradients across the
facesheet as well as the total volume of resin. As autoclave pressure was the same for each sample,
the compaction pressure (defined as Pauto – Pcore) varied depending on the applied bag pressure.
Because the equilibrated-core configuration was used, the core pressure (Pcore) was equal to the
bag pressure (Pbag). In Sample A, the bag and core were vented to ambient pressure, and thus
compaction pressure was reduced compared to Samples B and C with vacuum pressure in the bag
and core. Sample A, therefore, had the smallest volume of resin that bled from the facesheet.
Sample C had double the number of prepreg plies and therefore double the initial volume of
prepreg resin compared to Sample B, and thus had the largest volume of resin that bled from the
75
facesheet. These results impact bond-line porosity through porosity transport from the facesheet
to the bond-line.
Figure 33 illustrates the development of bond-line porosity during the cure cycle for the
samples. Previously, it was discussed that three porosity values in the bond-line can be defined: (i)
the porosity within the adhesive, (ii) the porosity within the prepreg resin that bled into the bond-
line, and (iii) the effective porosity, which is the weighted average of the adhesive and prepreg
resin porosity values in the bond-line [51]. The influence of bled prepreg resin porosity on the
effective porosity depends on the bled prepreg resin volume. Because the volume of adhesive in
the bond-line is greater than the volume of prepreg resin transferred to this region, the adhesive
porosity is dominant in the calculation of effective porosity. In Figure 33a, corresponding to
Sample A, the porosity advances based on the ideal gas expansion until the last stage of the cure
cycle. At 112 minutes, the cure cycle enters the unstable bubble growth region of the stability map
and consequently, diffusion-induced growth is invoked within the bubbles in the bled prepreg
resin. As the cure cycle proceeds, the bubbles within the adhesive experience the diffusion-induced
growth at 126 minutes. However, because of the selection of the process parameters of the cure
cycle, the bubbles do not grow large enough to escape the bond-line before the material gels,
trapping them in the gelled adhesive.
Figure 33 b and c demonstrate the porosity development within Samples B and C. These
samples were fabricated using the same temperature cycle as Sample A but with vacuum pressure
applied to the bag and the core. In the simulations, the vacuum pressure is 1 kPa. According to the
stability map, the cure cycle enters the unstable growth region in the early stage of the process,
while the temperature continues to rise according to the prescribed temperature ramp. Therefore,
the diffusion-induced growth initiates early in the cure cycle, which accelerates the bubble growth
76
within the adhesive. The bubbles grow and escape once they reach the bond-line surface. At 11
minutes, the consolidation process drives the prepreg resin and the corresponding bubbles from
the facesheet into the bond-line, which we refer to as the bled prepreg resin. Therefore, at this time,
porosity within the bled prepreg resin starts increasing as well due to diffusion-induced bubble
growth. As the cure cycle proceeds and temperature increases, porosity of the bled prepreg resin
decreases and attains a constant final porosity value because of two phenomena that occur
simultaneously: (i) bubbles within the bled prepreg resin grow and escape at the bond-line and (ii)
the prepreg resin bleeding stops as the consolidation process approaches the steady-state condition.
The difference between Sample B and Sample C is the number of prepreg plies of the facesheet.
The porosity reported by the model is specific to the bond-line region (for both the adhesive and
bled prepreg resin), and the overall porosity (the effective bond-line porosity shown by red dashed
line) is a weighted average of each material. While the percent porosity in the bled prepreg resin
is the same for Samples B and C, the porosity remaining after the cure is primarily sourced from
the prepreg resin and therefore the weighted average increases with increased volume of prepreg
resin bleed. The findings show the sensitivity of the simulation model to the number of prepreg
layers. The findings show the sensitivity of the simulation model to the number of prepreg layers.
Cross-sections for each sample are shown in Figure 34, with measured and corresponding
simulated porosity values summarized in

Table 4 and Figure 35. To measure porosity, samples were sectioned (50 mm in length),
polished (Buehler MetaServe), and imaged (Keyence VHX-5000). Two sections for each sample
were analyzed, each containing ~ 20 fillets. Individual fillets and their contained voids were traced
77
using image processing software (Adobe Photoshop CC) to quantify porosity, taken as the ratio of
void area to total area. At the center of the cell, fillet area included the region up to the location
where the adhesive contact angle approached zero. In some instances, this boundary could not be
clearly determined (e.g., due to porosity in the center of the cell, as shown in Figure 34a), and these
fillets were not included in the measurements. The layer of adhesive between the facesheet and 0°
adhesive contact angle was not accounted for in the fillet formation sub-model, but is included in
the porosity model and hence is counted as part of the fillet region during porosity measurements.

Figure 33: Simulated time-dependent development of bond-line porosity for each sample. The porosity considered is
the effective porosity, which is the weighted average of the porosity in the adhesive itself and bled prepreg resin. a)
78
Sample A, 4-ply facesheet, ambient core pressure (cure cycle I), b) Sample B, 4-ply facesheet, vacuum core pressure
(cure cycle II), and c) Sample C, 8-ply facesheet, vacuum core pressure (cure cycle II).
The model predicted greatest porosity levels in Sample A and the lowest in Sample B.
Despite identical temperature and pressure cycles, the porosity of Sample C was predicted to be
greater than that of Sample B, due to the greater volume of bled resin from the facesheet
transporting more voids to the bond-line.
As demonstrated in Figure 35 and

Table 4, variation of the measured porosity is large, and is attributed to stochasticity in the
co-cure process and multi-scale geometries inherent in sandwich panels, which lead to
nonuniformities. For example, the distance between pinholes in the prepreg fabric compared to the
diameter of honeycomb cells results in different numbers of pinholes within each core cell. This
affects the volume of the bled resin and consequently, the number of voids migrating from the
facesheet to the bond-line.
79

Figure 34: Cross-sections for each sample fabricated. a) Sample A, 4-ply facesheet, ambient core pressure, b)
Sample B, 4-ply facesheet, vacuum core pressure, and c) Sample C, 8-ply facesheet, vacuum core pressure.
The measured porosity confirms that the model can predict the bond-line porosity of the
honeycomb sandwich structures fabricated by the co-cure process in the autoclave. However,
integration of multiple models, each with their specific assumptions and simplifications,
assumptions for material properties, and non-uniformities in fabrication, make accurate prediction
of bond-line porosity nonviable. Although the presented model yielded underestimates of the
porosity in the processed samples, trends were correctly predicted. The results correctly predicted
the effects of process parameters and the number of facesheet plies on bond-line porosity.

80

Figure 35: Measured bond-line porosity and simulation results for each sample fabricated.

Table 4: Experimental and simulation results for bond-line porosity in each sample.

Sample Experimental Data (%) Simulation Result (%)
A 41.1 ± 18.2 29.4
B 6.5 ± 10.4 5.1
C 23.5 ± 17.01 6.7

81
5. Conclusions
In this work, physics-based models related to the adhesive bond-line and porosity
development during co-cure of honeycomb sandwich structures were compared with experimental
results. First, a validation experiment was conducted to evaluate the model to predict adhesive
fillet shape. Model predictions agreed with experimental data, and the model was viable for
integration into the porosity development simulations in which multiple sub-models were
combined. The porosity model integrated diffusion-induced void growth with descriptions of void
transfer from the facesheet to the bond-line and void escape from the adhesive fillets, phenomena
which are not captured by existing models intended for porosity in monolithic composite structures
[38,41]. Stability maps of critical pressure and temperature conditions for unstable void growth
were used to identify when porosity would form, and three cure cycles under different pressure
and material conditions were simulated using the integrated porosity model. Comparison to
experimental results indicated that the model could describe trends in porosity as a function of
applied pressure and facesheet thickness despite the stochastic nature of the process.
Contrary to the common belief that decreased vacuum pressure increases the final porosity,
results showed that due to gas escape during co-cure, applying vacuum resulted in the lower bond-
line porosity. Thus, there are two plausible solutions to mitigate the bond-line porosity in the co-
cure process of sandwich structures: (i) devising a cure cycle that avoids the unstable bubble
growth region in the stability map, either by pressurizing the core (e.g., [61,65]) or deploying a
temperature cycle that achieves cure prior to the unstable growth region (e.g., [67]), or (ii) applying
high vacuum early in the cure cycle, before the unstable bubble growth region. The latter strategy
will cause intensive bubble growth and consequently, their escape from the bond-line to the core.
82
In the work described here, only the case of equilibrated-core co-cure was considered to
reduce the complexity of simulations. In the sealed-core configuration, the core cavity is isolated
from the vacuum bag via an adhesive layer and prepreg facesheet on both the bag and the tool
sides, and core pressure changes during cure based on the temperature, pressure gradient, and
facesheet permeability [34,68,69]. In future work, the simulation model presented for bond-line
porosity could be coupled with a model predicting core pressure evolution to describe this sealed-
core configuration. Sealing the core, however, also isolates individual core cells from each other,
and introduces cell-to-cell variation, making modeling more complicated and challenging.
Despite the success of the model in capturing trends in porosity in response to parameter
changes, quantitative predictions did not match experimental measurements. Experimentally,
scatter in measurements was large due to stochasticity of the co-cure process, as well as
uncontrolled variability in material parameters (such as locations of pinholes in the fabric relative
to the core cells, which affect resin bleed). More precise material characterization and smaller-
scale modeling (e.g., modeling on the scale of fiber tows) could improve model accuracy.
Combining the proposed model with a description of the stochastic behavior of the co-cure process
could also yield more accurate predictions. Variability in model predictions also stemmed from
the complexity of integration that compounds assumptions and simplifications of several sub-
models. The complex physics and interactions involved in co-cure, coupled with the need for
detailed characterization of multiple material parameters made precise quantitative predictions of
the bond-line porosity impractical. However, the ability to predict trends based on changes in
material and processing parameters provides a potentially valuable tool to guide material selection
and cure cycle design for more efficient and robust production of honeycomb sandwich panels
with void-free bond-lines [70].
83

84
Chapter 6: Surface Porosity Development in Tool-side Facesheets of
Honeycomb Core Sandwich Structures during Co-cure
1. Introduction
For sandwich panels in aerospace applications, smooth surfaces are required for
aerodynamic and aesthetic purposes. Often, remedying surface porosity requires either adding
resin to the tool-side surface (thus adding weight to the part) or filling in porosity post-cure
(requiring additional time and resources) [71]. Literature specific to the causes and mitigation of
surface porosity in honeycomb core sandwich panels, however, is limited. A challenge in
processing sandwich panels is the non-uniform compaction pressure applied to facesheets, which
is dependent on the gas pressure within the core [31,72]. Darrow, Poropatic, and Brayden observed
that difficulty in pressure transfer to the tool-side facesheet led to surface porosity, with void-free
laminates produced under the same conditions as porous sandwich panels, but none of the
parameters investigated besides the presence of the core was found to be significant [71]. The
inclusion of a permeable layer between the tool and facesheet improved surface quality, indicating
entrapped gases as the void source. Brayden and Darrow, in modifying a cure cycle to avoid core
crush, reported that continued application of vacuum (rather than venting the bag prior to heating)
yielded a more uniform surface [73].
Previous studies have investigated the effect of processing parameters on surface quality.
For example, Jouin, Pollock, and Rudisill assessed the effect of various processing and material
parameters on surface quality and other metrics in autoclave-cured sandwich panels [74]. Adding
an intermediate temperature dwell to the baseline cure cycle and imposing super-ambient pressure
in the bag reportedly improved tool-side surface quality based on a qualitative scale. Similarly,
Alteneder et al. reported improved tool-side surface quality when imposing super-ambient pressure
85
in the bag, but quantitative measures were not included [60]. Yuan et al. investigated the effect of
heating rate, applied pressure, and timing of applied pressure on sandwich panels fabricated using
a hot press [21]. They reported (but did not quantify) improved surface quality with increasing
pressure and delayed pressure application. Results of this study, however, are not necessarily
transferrable to autoclave processing. Additionally, the above studies were limited to post-cure
analysis to select optimal processing conditions.
Although not fully representative of sandwich structures, surface defects in composite
laminates have also been studied. In autoclave processes, both surface and internal porosity can be
mitigated by maintaining sufficiently high resin pressure through increased autoclave pressure and
avoiding resin loss [38,75]. For out-of-autoclave prepregs using vacuum bag-only oven cure,
Hamill et al. attributed surface porosity to air entrapped between the tool and the surface ply during
layup [76]. Porosity was reduced by increasing air evacuation (e.g., by increasing vacuum hold
time or perforating plies) or reducing air entrapment (by reducing resin tack). Bloom et al. also
found air entrapment to be a primary cause of surface porosity, investigating the effects of
parameters such as debulk time and surface topology on air entrapment [77]. Hu et al. employed
an in situ technique to simulate inter-ply conditions and observe void formation in a resin-rich
region on a glass tool, identifying stages of air evacuation, void expansion during heating, and
subsequent void shrinkage [15,16].
A key difference between sandwich panels and laminates, in addition to the honeycomb
core that supports the facesheets, is the presence of a second resin in the adhesive film. Such films
undergo flow during co-cure to form fillets and the overall bond-line. Previously, an in situ
visualization technique was employed to analyze defect formation in the adhesive bond-line during
co-cure, identifying interactions between the prepreg resin and adhesive as one cause of bond-line
86
porosity [4,61]. The effect of the adhesive film on facesheet surface quality, however, has not been
addressed in literature.
1.1 Objectives and Approach
This work aims to clarify the causes and mechanisms of defect formation at the tool-side
facesheet surface during autoclave co-cure of honeycomb core sandwich panels, specifically
addressing 1) the effect of varying pressure and temperature cycle conditions on surface quality
and 2) the time-dependent behavior of voids at the surface. Both autoclave fabrication and an in
situ visualization technique are employed to assess the development of surface porosity during
cure, as well as the cured surface morphology.
Lab-scale sandwich panels were fabricated in an autoclave under varying pressure
conditions to determine the effect of pressure on surface porosity. Surface void area content of
cured samples was measured and correlated to processing pressure. Imposing elevated pressure in
the bag has been shown to improve surface quality (e.g., [60,74]), but a relationship between
pressure and a metric of surface quality has not been established. Shearing of the surface tows was
observed in cured samples and was quantified. Samples were also sectioned so relationships
between internal and surface facesheet quality metrics were established.
Samples were then cured in a set-up enabling in situ visualization of the tool-facesheet
interface to directly observe dynamic growth and transport of voids during co-cure. The technique
required out-of-autoclave cure, so pressure conditions could not be altered. Instead, two different
temperature cycles were used to assess the development of voids at the surface. During cure, both
temperature and time-lapse images of the surface ply were recorded so that time-dependent void
behavior could be correlated to temperature, as well as to modeled resin properties. Insights from
in situ tests were applied to autoclave-cured samples to provide physical descriptions of
87
relationships between surface quality and process parameters. Altogether, results elucidated
physical mechanisms by which facesheet surface porosity forms and provided guidance for
designing mitigation strategies for surface porosity in sandwich panels.
2. Materials and Methodology
Materials. A prepreg designed for structural aerospace applications, consisting of a plain-
weave carbon fiber fabric and a toughened epoxy (Hexcel HexPly 8552S), was used for facesheets
in the sandwich structures. The “S” variant of the prepreg used in this study was manufactured
using a solvent-based method, and residual solvent was present in the prepreg [48]. Models for
thermal properties, including cure kinetics and viscosity, of the standard resin (8552) have been
previously reported [46,47]. Models used in this study were adapted from Hubert et. al [46] based
on non-solvated 8552-1 resin film, which exhibits similar kinetic and rheological behaviors of
8552S prepreg.
The adhesive selected was a modified epoxy (Henkel Loctite EA 9658 AERO) supplied as
a film with a non-woven glass fiber (NWG) support, with an area weight of 320 g/m
2
. Thermal
properties of the material, including cure kinetics and viscosity, were characterized and reported
previously [45].
A phenolic-coated Nomex honeycomb core was selected (Gill Corporation HD142), with
3.2 mm (1/8 in) width hexagonal cells, 12.7 mm (1/2 in) thickness, and a density of 64 kg/m
3
(4
pcf).
Part layup. A schematic of the sample layup, including part geometry and consumables
used, is shown in Figure 36. Samples were assembled using a prepreg facesheet and adhesive layer
on either side of a chamfered core insert (larger face 114 mm × 114 mm, 45° chamfer angle). Each
facesheet (152 mm × 152 mm) consisted of 4 plies of prepreg ([0°/90°]2s). Layup consisted of
88
assembling the tool-side facesheet and applying a layer of adhesive film (152 mm × 152 mm),
placing the larger face of the core centered on the adhesive, then laying the bag-side layer of
adhesive and facesheet over the core ply-by-ply. The facesheet flanges were trimmed so plies
would be flush, leaving final part dimensions approximately 140 mm × 140 mm.

Figure 36: Schematic of sandwich panel layup. Laminates were fabricated using the same layup procedure,
including consumables, but omitting the adhesive and core insert.
Autoclave processing. The selected sample size enabled two parts to be cured
simultaneously during autoclave processing trials. A non-perforated release film was used between
the tool plate and sandwich panel, and breathing dams were placed around the edges of the sample.
The bagging was completed using a perforated release film, cloth breather, and vacuum bag on top
of the sandwich panels. When laminates were produced for comparison, layups consisted of 4 plies
of prepreg only ([0°/90°]2s, 152 mm × 152 mm). Otherwise, the layup procedure including
consumables followed the procedure for sandwich panel samples. The baseline temperature cycle
used was adapted from manufacturer recommendations for the prepreg, and consisted of a 60 min
dwell at 110 °C, followed by a 120 min dwell at 177 °C, with a 2 °C/min heating rate. Prior to
heating, samples were held under vacuum at room temperature for 60 min. Prescribed vacuum bag
and autoclave pressures (Table 5) were imposed at the beginning of the first temperature ramp and
held constant throughout the rest of the cure cycle.
89
Table 5: Testing conditions for autoclave processing. All pressures are absolute. For each testing condition, two
sandwich panel samples were fabricated. Additionally, a comparison laminate was made for the baseline and vented
bag conditions.
Test Bag Pressure
[kPa]
Autoclave Pressure
[kPa]
RT Vacuum Hold
Duration [min]
Intermediate Dwell
Duration [min]
Baseline < 5 275.8 60 60
Vented Bag 101.3 377.1 60 60
In-bag
Pressurization
239.2 515.0 60 60
Long RT Hold < 5 275.8 960 60
No Intermediate
Dwell
< 5 275.8 60 0

Autoclave processing trials focused on the effect of pressure on surface quality, with tests
conducted at vacuum (< 5 kPa), ambient (101.3 kPa), and elevated (239.2) pressures applied to the
vacuum bag. Ambient pressure in the bag was applied by disconnecting the vacuum hose from the
vacuum pump and leaving it vented. Elevated pressure in the bag was supplied by connecting the
bag to a dry nitrogen tank in place of a vacuum pump. The autoclave pressure was adjusted to keep
the compaction pressure (defined as autoclave pressure minus vacuum bag pressure) constant
between tests. Other test variations included increasing the length of the room temperature vacuum
hold and removing the intermediate dwell. For each testing condition, two sandwich panel samples
were fabricated. Additionally, laminates were produced for the baseline vacuum and vented bag
conditions for comparison to sandwich samples.
90

Figure 37: Sample measurements of tow shearing. To measure tow shearing, individual sections created by the
fabric weave were outlined and the circularity measured. Values were then compared to a baseline established using
the vacuum laminate sample.
Following fabrication, samples were imaged using a video microscope (Keyence VHX-
5000). For each sample, a section (25 mm × 25 mm) of the tool-side surface was imaged at both
the center of the sandwich panel and at the flange. Panels were then cross-sectioned, polished, and
imaged to assess the internal quality of the tool-side facesheet. Surface porosity was measured
using image processing software (Adobe Photoshop CC) and reported as percent of void area
content to total area. Additionally, shearing of the tows observed in the surface ply was quantified
by the circularity (4𝜋 (
𝐴 𝑝 2
), where A is area and p is perimeter, as used by Adobe Photoshop CC,
so that circularity of a circle is 1) of individual “sections” of the tows (Figure 37). The average
circularity of the vacuum laminate part – which qualitatively displayed minimal deviation from
rectangular sections – was used as a baseline. Shearing was reported as deviation (as percent error)
from the baseline, computed for each individual tow section and then averaged for the entire part
to eliminate effects of direction of deviation (i.e., becoming more convex or more concave).
91
Internal quality was reported as facesheet porosity (as percent of void area content) and facesheet
pillowing (as thickness of facesheet at the cell center).
Oven window cure. To assess the time-dependent development of porosity at the tool-side
facesheet surface, sandwich panels were cured on a glass table in a windowed oven to enable direct
visualization of the facesheet surface. The test set-up is detailed in Figure 10. A single sample was
placed on the glass tool plate with a non-permeable release film. Bagging was completed following
the same procedure as in the autoclave processing trials, including the use of edge breathing dams.
The glass tool plate and frame were then placed in the door of the oven, and a digital microscope
(Dino-Lite Edge AM7815MZTL) was placed outside the oven to record time-lapse videos of the
tool-side facesheet during cure. The frame was used to enable air circulation at the tool surface
and decrease thermal gradients (compared to curing the sample directly on the oven window).
Possible pressure conditions for oven cure were limited to vacuum-bag only, so the effect
of varying pressure on surface quality could not be studied. The baseline case used was identical
to the baseline vacuum case in the autoclave processing trials, excepting the absence of autoclave
pressure. A second test was conducted with an altered temperature cycle, removing the room-
temperature vacuum hold and the intermediate dwell. These adjustments are expected to change
the development of core pressure during cure primarily by reducing time for gas to evacuate from
the core. While the material used was not intended for VBO cure, the different pressure conditions
experienced in the oven window cure tests were expected to affect the rate and extent of void
development but not the behavior (void source, flow direction, evacuation, etc.)
During cure, thermocouples were used to measure and record temperature at the bag-side
facesheet and on the tool surface. The digital microscope recorded images every 30 seconds, and
time-lapse videos were assembled tracking the development of surface morphology along with
92
temperature and modeled degree of cure and viscosity for both the adhesive and prepreg resin.
Cured samples were processed and imaged as in the autoclave processing trials to assess both
surface and internal facesheet quality.
3. Results and Discussion
Autoclave processing. Imaged tool-side surfaces are presented in Figure 38. For the
baseline vacuum and vented vacuum bag cases, a laminate is shown for comparison to the
sandwich panel. In each case, the laminate surface was void-free, while there were voids present
in the sandwich panel surfaces. Additionally, tow shearing was more pronounced in the vented
bag sandwich compared to the vacuum sandwich, while laminates did not display any significant
shearing, regardless of pressure.

Figure 38: Surface images of autoclave-cured samples: A, baseline vacuum laminate; B, baseline vacuum sandwich;
C, vented bag laminate; D, vented bag sandwich; E, in-bag pressurization sandwich; F, in-bag pressurization
sandwich with crushed core; G, long room-temperature vacuum hold sandwich; H, no intermediate dwell sandwich.
Voids were overlaid with black for better visibility.
93
Two different sandwich panel samples are shown for the in-bag pressurization case. In the
top sample (Figure 38e), porosity is visible, indicating that the increased pressure was not
sufficient to pressurize the core enough to completely suppress voids. In the bottom sample, core
crush was observed on the bag side of the sample, along with severe shearing on the tool-side
surface. Core crush has been linked to insufficient inter-ply friction to overcome lateral pressure
on the core (e.g., [78,79]), so the extreme shearing coupled with core crush indicates that the
shearing can be attributed to facesheet plies slipping and shifting laterally.

Figure 39: Surface images of baseline vacuum panel under the honeycomb core (top) and in the flange (bottom).
94
Flange surfaces were imaged to compare to surfaces under the core insert, and
representative images for the baseline vacuum case are presented in Figure 39. Regardless of
processing conditions, flange surfaces were comparable to laminates – void-free and with minimal
tow shearing. In contrast, within the same parts, sections under core displayed both porosity and
shearing. Some adhesive was present in the flange surfaces (visible as bright flecks clustered
around pinholes in the fabric weave, due to aluminum powder in the adhesive). Adhesive was not
observed at the surface under the core, regardless of processing conditions.
Measurements of surface porosity and circularity (as deviation from baseline laminate) are
presented in Figure 40. The three tests at varying bag pressure displayed a nonlinear relationship
between surface porosity and pressure. Increasing bag pressure by venting the bag had no
statistically significant effect on surface porosity. However, increasing pressure further through
in-bag pressurization yielded a reduction in porosity of more than half compared to the baseline
and vented bag samples. In a previous study with the same materials, the relationship between
bond-line porosity was non-monotonic: vacuum pressure in the core led to reduced porosity via
gas evacuation, in-bag pressurization led to reduced porosity through volatilization suppression,
and vented conditions yielded the greatest porosity [61]. The comparatively high surface porosity
under vacuum indicates that gas evacuation is not as effective at the tool-facesheet interface as it
is for the bag-side adhesive bond-line. Any gases located at the tool-side facesheet surface must
migrate through the facesheet and bond-line, into the core gas pocket, and through the bag-side
bond-line and facesheet before being evacuated into the breathing cloth. In contrast, the pathway
for gases in the bag-side bond-line and facesheet have a relatively direct evacuation pathway.
95

Figure 40: Surface void area content (top) and circularity (as deviation from the baseline laminate average, bottom)
for autoclave-cured samples.
Although gas evacuation was insufficient to mitigate porosity in the baseline case,
increasing the duration of the room-temperature vacuum hold reportedly decreases surface
porosity in laminates (e.g., [76]), and this same trend was observed for sandwich panels. Average
surface void area content for the long room-temperature hold samples was comparable to the in-
bag pressurization samples (respectively, 2.6% and 2.7%). However, in-bag pressurization yielded
greater consistency in surface quality, with individual tests ranging from ± 12.9% of the average
value, compared to ± 76.7% for the long room-temperature hold samples. The range in the long
96
room-temperature hold values was attributed to inconsistencies in core gas evacuation during
room-temperature vacuum holds, as reported by Kratz and Hubert [32]. When using facesheets
consisting of plain-weave fabric, the onset of core gas evacuation varied from ~ 60 min to 1020
min, and the core pressure did not always equilibrate to the applied vacuum pressure. Thus, for a
vacuum hold of 960 min, a range of post-vacuum-hold core pressures was expected, with pressure
also varying from cell to cell within the same part. In contrast, evacuation during the 60-min hold
used for the baseline condition was likely minimal, with core pressures post-vacuum-hold
relatively high (near ambient) but uniform.
In contrast to the increased gas evacuation during the long room-temperature hold case,
eliminating the intermediate temperature dwell reduced the time spent under vacuum with reduced
viscosity (of both prepreg resin and adhesive). A competing effect of increased core (and,
therefore, resin) pressure, however, is also expected with the reduced time to evacuate gas in the
core. The no-dwell samples resulted in the greatest surface porosity of 10.1%, indicating that the
decreased evacuation had a greater effect on surface quality than any potential suppression of voids
through increased core pressure.
Circularity showed no statistically significant trends regardless of processing conditions,
due primarily to large variation within certain cases. For samples subjected to in-bag
pressurization, shearing of the surface ply was linked to core crush, as one sandwich exhibited
distinctive crushing around the chamfered region of the core. This crushed sample also had the
greatest deviation of circularity from the baseline (17.6%, compared to 3.8% measured for the un-
crushed sample produced under identical conditions). Otherwise, circularity deviation from the
baseline for individual sandwich panels ranged from 1.3% to 6.4% regardless of conditions, and
no core crush was observed for any of these samples. These results indicate that surface ply
97
shearing is not directly related to any of the tested parameters, but may be indirectly related through
other phenomena (e.g., increased bag pressure leading to core crush, which leads to shearing).
Cross-sections of samples are shown in Figure 41. Internal porosity was observed in every
sample, regardless of processing conditions. Within each sample, voids were concentrated at cell
centers (where compaction pressure was dependent on gas pressure within the core), while no
porosity occurred under cell walls (where autoclave pressure could be transferred to the tool-side
facesheet). Compaction variations were observed in every sample, due to the non-uniform pressure
applied to the facesheet, but pillowing was most extensive in the crushed sample (Figure 41D).

Figure 41: Cross-sections showing internal structure of tool-side facesheet and bond-line of autoclave-cured
samples: A, baseline vacuum; B, vented bag; C, in-bag pressurization; D, in-bag pressurization with crushed core; E,
long room-temperature vacuum hold; F, no intermediate dwell.
Internal porosity and pillowing (as maximum facesheet thickness) were correlated to
surface quality metrics, as shown in Figure 42. Variability in internal void area content made it
impossible to identify any direct relationship between internal and surface porosity. Notably,
despite relatively low variation in surface porosity (± 5.93% from the average), the vented bag
case exhibited the greatest variation in internal porosity (± 85.2% from the average) and average
internal porosity values for all other testing conditions fell within this range. The absence of
98
correlation between internal and surface porosity indicates a possible difference in the
development of these two defects, either in mechanism or timing.

Figure 42: Surface vs. internal porosity (top) and circularity (as deviation from baseline laminate) vs. maximum
facesheet thickness (bottom).
Comparing circularity (as deviation from the baseline laminate) to pillowing (as maximum
facesheet thickness), excluding the in-bag pressurization data, showed circularity to be constant
across a range of facesheet thicknesses. When considered separately, the uncrushed in-bag
pressurization sample fell within the range of other testing conditions (3.83% circularity deviation,
1.04 mm maximum facesheet thickness), while the crushed sample was at the extremes of the plot
99
range (17.6% circularity deviation, 1.49 mm maximum facesheet thickness). These results
demonstrate a link between surface ply shearing, facesheet pillowing, and core crush: as the core
crushes, facesheet plies are pinched into the cell pockets, and the resulting ply slippage causes
shearing at the surface ply. In cases in which the core was not crushed, a moderate degree of
shearing was observed and was attributed to the presence of the honeycomb core, as supported by
the consistent deviation in circularity from the baseline laminate. A potential cause of this shearing
is the non-uniform compaction applied to the tool-side facesheet by the core walls and gas pressure
within the core cells.
Oven window cure. Frames for the baseline in situ visualization test are shown in Figure
43. Some air bubbles initially entrapped air between the tool and facesheet were observed in the
initial state. As resin viscosity decreased during the first temperature ramp, resin flowed to fill
gaps at the surface, and bubbles of gas were observed migrating toward pinholes in the fabric and
disappearing from view. These bubbles presumed to evacuate through the facesheet and into the
core. Then, toward the end of the intermediate dwell and into the second temperature ramp, new
voids formed. In contrast to the initial bubbles due to air, which were relatively small and spherical,
voids forming later in the cure cycle were larger (on the scale of fiber tows) and elongated to fit
within the surface contours of the fabric weave. As in the first temperature ramp, these gases
moved toward pinholes and evacuated into the core. However, as the adhesive viscosity increased,
gas transfer into the core was blocked, and porosity existing at the surface or developed after
adhesive gelation remained trapped at the surface. The voids forming during the intermediate dwell
and second temperature ramp were attributed to volatilization of residual solvent, as observed
previously for porosity in the bond-line using the same materials [61].
100
Because varying pressure conditions could not be imposed in the oven, a test was
conducted in which the room-temperature vacuum hold and the intermediate dwell were removed
to implicitly increase core pressure during cure by reducing time for core gas pressure to evacuate.
However, any increase in the core pressure that occurred did not have an observable effect on void
suppression. Instead, the behavior observed at the tool surface was comparable to the baseline
case, excepting that the reduced time spent at low resin viscosity (resulting from eliminating the
intermediate dwell) led to increase surface porosity.
Surface images for the two oven-window-cure parts are shown in Figure 44, along with
surface void area content compared to the respective autoclave-cured parts. Within the same
fabrication method, the baseline vacuum case showed lower porosity than the no-dwell case, a
finding attributed to the difference in times spent at low viscosity to evacuate gases trapped at the
tool surface. In general, the autoclave-cured parts exhibited lower surface porosity compared to
the respective oven-cured samples. This trend is attributed to the increased compaction pressure
enabled by the autoclave, which is expected to decrease facesheet permeability and therefore
increase average core pressure during cure, thus suppressing some void growth. Note that the no-
dwell case for autoclave cure still underwent a room-temperature vacuum hold that was not
included for oven window cure.

101

Figure 43: Frames from time-lapse video for baseline oven window cure test, with times marked along with
measured temperature and modeled viscosity. At t1, entrapped air was observed moving toward pinholes in the
surface ply and disappearing from view. At t2, new voids began forming, with some remaining trapped in the
surface after cure.
102

Figure 44: Surfaces of oven window cure parts: baseline vacuum (top) and no room-temperature vacuum hold/no
intermediate dwell (bottom). Surface void area content for oven window cure parts compared to autoclave-cured
equivalents. Note that, for the autoclave-cured sample, the no-dwell case did include a room-temperature vacuum
hold.
103
In situ observations clarify results of autoclave processing tests and provide valuable
insight. One result of note was the comparable levels of surface porosity in the baseline and vented
bag cases, despite the increased time under vacuum in the baseline cure cycle, which provided
more time to evacuate entrapped gas. Oven-cure window tests demonstrated that the baseline cure
cycle was effective in removing initially-entrapped air at the facesheet surface during the first
temperature ramp. Porosity remaining in the surface after cure, however, was due to residual
solvent and did not form until later in the cure cycle (during the intermediate dwell and second
temperature ramp, consistent with solvent-based voids observed previously [48,61]). There was,
therefore, insufficient time following the growth of these voids to evacuate the gas while adhesive
viscosity remained relatively low (adhesive viscosity, as modeled, reaches a local minimum ~ 15
min. after the second temperature ramp begins and increases until gelation after that point).
Porosity attributed to solvent volatilization due to the observed timing in situ presented as large,
oblong voids (~ 1-2 mm in length) that tended to follow fiber tow directions. This porosity
accounted for most of the void content seen in both autoclave and oven window samples, and
therefore porosity due to entrapped air was assumed to be negligible.
Although the baseline case was ineffective in evacuating volatilized solvent, increasing the
duration of the room-temperature vacuum hold resulted in decreased surface porosity. As revealed
by in situ videos, the baseline cure cycle evacuated initially-entrapped air. Thus, the reduction in
surface porosity was attributed to a greater mass of solvent volatilized and evacuated prior to the
adhesive viscosity increasing with the increased vacuum hold duration compared to the baseline.
Per Henry’s law, volatilization of the residual solvent in the prepreg resin is dictated by both
temperature and the gas pressure in the core. Increasing the duration of the vacuum hold decreased
the core pressure at the beginning of the temperature cycle, causing the solvent to volatilize at a
104
lower temperature than in the baseline case, and therefore increased time under sufficiently low
adhesive viscosity to evacuate evolved gases. Conversely, increasing the core pressure through in-
bag pressurization reduced surface porosity by increasing the temperature at which the solvent
volatilized, keeping more solvent mass in solution prior to the gelation of the prepreg resin.
No significant shearing was observed in either of the oven-cured samples, indicating that
shearing is caused by autoclave pressure. This finding is consistent with the observed link between
surface ply shearing, facesheet pillowing, and core crush. The vacuum bag-only curing conditions
used for oven cure was not sufficient to crush the core, regardless of the temperature cycle used.
4. Conclusions
Causes of surface porosity at the tool-side facesheet in honeycomb core sandwich
structures were investigated using both autoclave fabrication and an in situ oven cure technique.
Direct observation of the tool-facesheet interface during co-cure revealed voids were caused
primarily by the volatilization of residual solvent in the studied prepreg resin, rather than entrapped
air. Laminates produced under the same conditions were void-free, indicating that porosity in
sandwich panels was due to non-uniform compaction pressure applied to the facesheet, making it
difficult to maintain sufficient resin pressure. Increasing the core pressure (and therefore resin
pressure) by imposing super-ambient pressure in the vacuum bag was shown to be the most
effective and reliable strategy to reduce surface porosity.
Evacuation of evolved volatiles was ineffective, as surface porosity in samples with
vacuum pressure applied in the bag continuously was comparable to samples fabricated under a
vented bag. Challenges in gas evacuation were attributed to two main sources:
105
1) Gases at the tool-side facesheet surface had a tortuous path to escape, having first to
flow through the tool-side bond-line and core gas pocket, then through the bag-side
bond-line and facesheet.
2) Because surface porosity was due to residual solvent in the prepreg that evolved at
elevated temperature, there was a limited time window during which these gases could
evacuate while adhesive viscosity was sufficiently low. As adhesive viscosity
increased, gas flow through the bond-line was occluded, trapping any gas existing at or
evolved after that point at the facesheet surface.
Solvent removal was facilitated by increasing the duration of the room-temperature vacuum hold,
which reduced surface porosity compared to the baseline. These conditions reduced the core
pressure at the start of the temperature cycle and therefore the effective core pressure throughout
processing, reducing the temperature at which solvent volatilized, while increasing the time to
evacuate before adhesive viscosity increased. However, while resulting surface porosity was
comparable to levels for in-bag pressurization, greater variability was observed and attributed to
inconsistencies in room-temperature air evacuation in sandwich panels (e.g., [32]). For surface
porosity caused by volatiles dissolved in resin, maintaining resin pressure to keep volatiles in
solution was more reliable and effective at mitigating void growth than removing volatile mass
through vacuum application.
Although in-bag pressurization consistently yielded lowest surface porosity, the increased
pressure introduced the risk of core crush. In a sample with core crush, shearing of the surface
facesheet ply was observed. Comparing this shearing to facesheet pillowing showed that shearing
in the surface ply occurred because plies nearest the core were pinched and slid into the core cells
as they crushed. Core crush, however, had no impact on void formation, as the crushed sample
106
exhibited surface porosity comparable to the uncrushed in-bag pressurization sample. To a lesser
degree, shearing was observed in other sandwich panels fabricated under autoclave conditions, but
not in laminates nor in oven-cured sandwich panels. This finding was ascribed to the non-uniform
compaction pressure imposed on the facesheet by the honeycomb core (and exacerbated by the
autoclave pressure that was not applied in oven-cured samples). Full compaction pressure is
transferred at the cell walls, with reduced compaction pressure elsewhere that is dependent on core
gas pressure.
The results here indicate that surface porosity for the selected material set is due primarily
to residual solvent. Entrapped air had a negligible impact, as in situ visualization of the surface
morphology during cure showed effective evacuation of air prior to the volatilization of solvent.
Additionally, voids attributed to entrapped air were smaller and more spherical than the large and
oblong voids caused by solvent. These smaller voids were negligible compared to larger solvent-
based voids, or were absent entirely in cured surfaces. However, a non-solvated variant of the
prepreg was not available, so the effects of entrapped air and solvent on surface quality could not
be fully decoupled. In situ visualization results indicate, however, that entrapped air is relatively
easy to remove, as it is present in the initial layup and can therefore be evacuated at any point prior
to adhesive gelation (as opposed to solvent that must be first evolved and then evacuated). Using
a non-solvated prepreg in sandwich structures, therefore, is expected to alleviate surface porosity
without the need for elevated bag pressure that can crush the core.
This study focused on the tool-side facesheet surface quality, providing new insights into
the dependence of defects on process parameters. The bag-side facesheet has a unique set of
boundary conditions. Compared to the tool-side facesheet, the bag side has a more direct path to
evacuate gases into the breather. However, more complex compaction conditions arise from the
107
lack of rigid tool plate on the external surface (unless a caul plate is used). Additionally, basic
internal facesheet quality was assessed for correlations to surface quality, but bond-line quality
was not addressed in this study. Understanding the mechanisms of tool-side surface defect
formation and developing strategies to mitigate porosity is an important step in the robust
manufacture of honeycomb core sandwich panels, although these insights must be integrated with
other aspects of co-cure to guide processing decisions for co-cure of sandwich panels.

108
Chapter 7: Mitigating Void Growth in Out-of-Autoclave Prepreg
Processing Using a Semi-Permeable Membrane to Maintain Resin
Pressure
1. Introduction
Carbon fiber-reinforced composites are commonly cured in autoclaves, as the compaction
pressure provided by the autoclave suppresses void growth and reliably yields low-porosity parts
[1,14]. Autoclaves, however, represent a substantial capital and recurring expense, and restrict part
size and throughput. To overcome these limitations, prepregs designed for oven cure (also called
vacuum bag-only, or VBO prepregs) have been developed which rely on deliberate dry channels
of fiber beds to evacuate air and thus limit/eliminate porosity (e.g., [5–9]). Dry regions in these
out-of-autoclave (OoA) prepregs are achieved by sandwiching fiber beds between two layers of
resin, and deliberately achieving partial impregnation. Thus, these materials rely on in-plane
permeability and evacuation pathways to laminate edges to remove air. Studies have demonstrated
that in-plane permeability in such OoA prepregs are generally orders of magnitude greater than
through-thickness permeability, although the in-plane permeability approaches zero as resin flows
to fill dry areas [80,81].
Because of the reliance on air evacuation at part edges, OoA prepregs are susceptible to air
entrapment. For example, Arafath et al. characterized in-plane permeability of an OoA
unidirectional tape and modeled evacuation time, and predicted that a flow distance of 4 m would
require 7 days to remove air [82]. Even in lab-scale studies, debulk times of up to 24 hours were
required to mitigate both internal and surface porosity due to entrapped air [76,83,84]. In related
109
work, Kay and Fernlund also reported a gradient in porosity in laminates produced with OoA
prepregs, with void content increasing as distance from the vacuum source increased [83].
To address issues with air entrapment in OoA prepreg laminates, studies have investigated
methods to increase through-thickness air permeability in OoA prepregs. For example, prepregs
can be perforated using a simple spike roller, promoting air evacuation in both laminate processing
and honeycomb sandwich applications [31,85]. Grunenfelder et al. used discontinuous resin films
to create through-thickness evacuation channels on a 2 × 2 twill fabric, demonstrating that such
modified prepregs yielded laminates with negligible porosity, while commercial OoA prepregs
generally yielded laminates with much greater porosity (e.g., sealed edges and abbreviated room-
temperature vacuum hold) [11]. Edwards et al. presented a method to produce unidirectional
prepreg with through-thickness permeability, and similarly showed that such prepregs were less
susceptible to porosity than conventional prepreg formats under challenging process conditions,
such as humidity exposure and embedded ply drops [12].
While the studies cited above demonstrated that increased through-thickness permeability
reduces or eliminates porosity from air entrapment, voids also can arise when the pressure of gases
dissolved in the resin exceeds the resin pressure. In autoclave processing, resin pressure is
generally maintained by application of compaction pressure (typically ~ 700 kPa) [14]. VBO
processes, however, are limited to the compaction pressure applied by the vacuum bag, which is
equal to the ambient pressure (~ 100 kPa). Grunenfelder et al. demonstrated that porosity in VBO-
processed laminates increased when prepreg was humidity conditioned prior to layup, while the
same material processed under autoclave conditions yielded void-free parts, regardless of humidity
conditioning [86]. Similarly, Kay and Fernlund reported an increase in porosity with increasing
moisture content [83]. Other studies have investigated the effect of reduced compaction pressure
110
(either by reducing ambient pressure or increasing the absolute pressure within the vacuum bag),
reporting that porosity increased as applied pressure decreased [83,84,87].
Resin pressure during cure of prepreg laminates is typically maintained by relying on a
pressure gradient between atmosphere and vacuum. However, a loss of resin content in a laminate
during cure can result in a drop in resin pressure that is generally accompanied by a transfer in
load from the fluid resin to the fiber bed. Campbell et al. used a pressure sensor embedded in a
tool plate to measure resin pressure during cure of a fully impregnated prepreg [14]. They reported
that resin pressure remained near the applied autoclave pressure in a baseline case, but dropped
below vacuum when resin bleed was deliberately increased [14]. In related work, Lynch et al.
embedded sensors within laminates to measure through-thickness variations in resin pressure [88].
They reported an increase in pressure at the laminate-bleeder layer interface at the same time
pressure dropped throughout the laminate, which they attributed to resin bleeding out of the
laminate [88]. These studies were limited to autoclave prepregs. The evolution of resin pressure
during cure of OoA prepregs has not been addressed in literature.
Investigations into scaling of OoA prepreg processing – both in terms of size and geometric
complexity – have reported conflicting results. Ma et al. assessed the effect of corner geometry in
OoA processing, reporting that reduced compaction pressure due to consumable bridging at
concave corners led to increased porosity and laminate thickness [89]. Likewise, Levy et al.
developed and validated a model capturing laminate thickening at concave corners, but did not
discuss porosity. Hughes and Hubert reported porosity and laminate thickness variations in
demonstration parts due to geometric complexities, including corners and ply drops [90]. Other
studies claimed success in producing demonstration parts with acceptable quality (e.g., [91–93]).
111
While OoA prepregs can often yield high-quality parts, the process is not robust, largely because
of the absence of autoclave pressure.
This work combines prior advancements in prepreg design to increase through-thickness
air evacuation (e.g., [11,12]) with the use of a semi-permeable release film to maintain resin
pressure and therefore retain resin content. Laminates were fabricated using prepregs with through-
thickness-permeability (semipregs) and resin-impermeable edge boundaries, then compared to
similar laminates cured under conditions representative of conventional OoA prepreg processing.
Real-time evolution of voids at the laminate-tool interface during cure was captured using in situ
visualization, identifying porosity originating both from entrapped air and from volatile evolution.
Results demonstrated that through-thickness gas transport combined with a semi-permeable
release film could mitigate porosity from both sources.
A method to measure resin pressure during cure was employed, and results described trends
in resin pressure evolution specific to OoA/VBO prepregs. In particular, a decrease in resin
pressure was observed during cure, and was attributed to impregnation of dry fibers, a phenomenon
that does not occur in fully impregnated autoclave prepregs. Measurements further confirmed that
the use of a semi-permeable membrane maintained a higher resin pressure during cure than
conventional OoA consumables, while resin format had no clear effect on resin pressure. In
contrast, use of conventional consumables during cure resulted in a loss of resin pressure, which
was consistent with volatile-induced porosity identified via in situ visualization. Overall, results
of this work demonstrate additional measures that increase robustness of OoA processing by
maintaining resin pressure during cure. The work also demonstrates that judicious selection of
materials and consumables can mitigate porosity from both air entrapment and volatile evolution.
112
2. Experimental Methods
Materials. The resin used in this study was a non-commercial epoxy intended for out-of-
autoclave processing (Hexcel HCR140-15). Prepreg was produced by pressing resin films (55.5
g/m
2
) to either side of a plain-weave carbon fabric (Toray TORAYCA T300, 3k tows, 194 g/m
2
).
Two types of prepregs were fabricated – one with continuous resin (CR) film, similar to
commercial out-of-autoclave prepregs, and one with discontinuous resin (DR) film, intended to
increase through-thickness gas permeability.
To produce prepreg, resin films and fabric were cut to 305 mm × 305 mm (12 in × 12 in).
A resin film was placed on either side of the fabric (two layers of resin film per ply, yielding a
total areal density of 305 g/m
2
and a resin weight fraction of ~ 31%) and pressed onto the fabric
under 2 tons of force for 5 minutes at room temperature (Wabash Genesis). DR film was produced
through a de-wetting process in which films were subjected to a mild thermal cycle (10 minutes at
70 ˚C) that reduced viscosity and allowed resin to de-wet on the backing paper, yielding islands of
resin separated by dry spots. After de-wetting, prepreg was produced using the same process as
with the continuous film. Figure 45 shows the surface morphology of a prepreg ply with
discontinuous resin.
A second (commercial) prepreg was used to compare resin pressure measurements
obtained with the experimental prepreg and discontinuous resin to a commercial prepreg with
conventional continuous resin format. This prepreg also featured an epoxy resin and was designed
for VBO processing (Solvay Cycom 5320-1/T650-35, 3k tows, 8HS, 367 g/m
2
); it required no
further processing before testing.
Lab-scale Sample Fabrication. To assess the effect of processing parameters on void
behavior – both the time-dependent evolution of porosity and void content in the cured part – the
113
oven window frame fixture detailed in Chapter 1 was employed. The framed glass tool plate, which
was inserted in the door of a windowed oven during testing, enabled visualization while allowing
for air flow on both sides of the tool plate to minimize temperature gradients.

Figure 45: Prepreg made with de-wetted resin film to create a discontinuous resin pattern, providing through-
thickness evacuation channels to increase gas permeability. The resin appears as dark areas. The white squares are
pinholes in the fabric weave.
Two values each for three different processing variables were tested, as listed in
Table 6. The prepreg format (resin distribution) affects gas transport pathways, enabling
both in-plane and through-thickness transport (in the case of discontinuous resin) or restricting
transport to in-plane only (continuous resin). The edge conditions and the bag-side surface
conditions also affect resin bleed, in one case sealing the respective interfaces (sealed edges, semi-
permeable membrane), while in the other, allowing resin bleed at edges and through a perforated
release film.

114

Table 6: Testing conditions for lab-scale laminate fabrication. Two laminates were produced for each condition.
Test Resin Format Edge Condition Bag-side Surface Layer
C:B/B Continuous Bleed (breathing dams) Bleed (perforated film)
C:B/S Continuous Bleed (breathing dams) Sealed (semi-permeable membrane)
C:S/B Continuous Sealed (tacky tape) Bleed (perforated film)
C:S/S Continuous Sealed (tacky tape) Sealed (semi-permeable membrane)
D:B/B Discontinuous Bleed (breathing dams) Bleed (perforated film)
D:B/S Discontinuous Bleed (breathing dams) Sealed (semi-permeable membrane)
D:S/B Discontinuous Sealed (tacky tape) Bleed (perforated film)
D:S/S Discontinuous Sealed (tacky tape) Sealed (semi-permeable membrane)

Prepreg laminates (102 mm × 102 mm, [0˚/90˚/0˚]3s) were placed on a glass tool plate after
application of a liquid mold release agent. Edges were either sealed using vacuum sealant tape, or
configured to be permeable to gas and resin by constructing breathing dams (fiberglass cloth
wrapped around sealant tape). The bag-side release film – either a semi-permeable membrane for
the sealed condition (Airtech Dahltexx SP-2) or perforated film for the bleed condition (Airtech
A4000) – was then laid over the laminate. The bagging assembly was completed with a breather
cloth and vacuum bag.
After layup and bagging, the entire frame was placed in the oven with the bag-side inward,
and a digital microscope (Dino-Lite Edge AM7815MZTL) outside the oven was used to record
time-lapse videos of the tool-side laminate surface during cure. Images were collected every 30 s.
Thermocouples were also used to record temperature during cure. The applied thermal cycle
consisted of 120 minutes at 121 °C and 120 minutes at 177 °C, with a 2 °C/min heating rate. The
115
time-lapse videos captured during cure provided real-time data on void evolution at the tool-side
surface.
In addition, cured laminates were assessed for void content through polished sections. After
cure, sections of laminates (25 mm in length) were cut, polished, and imaged using a digital stereo
microscope (Keyence VHX-5000) to assess internal porosity. Void content was measured and
quantified (as percent of total area) using image processing software (Adobe Photoshop CC).

Figure 46: Example of an in-progress layup using pressure probe assembly incorporated into the laminate for resin
pressure measurements. The needle is inserted between plies, and a ball of excess resin is placed at the tip of the
needle to seal the interface upon initial application of vacuum. The needle of the pressure probe is passed through
the bag sealant tape to avoid introducing bag leaks.
Resin Pressure Measurement. Resin pressure was measured in situ using a probe inserted
between plies during cure, a method that was previously deployed to measure resin pressure
[88,94]. The resin pressure measurement assembly used is shown in Figure 46 and contains three
main components: 1) a probe inserted into the sample, 2) a reservoir situated outside the vacuum
bagging assembly, containing a transfer fluid, and 3) a pressure transducer. The probe was a 19-
gauge stainless steel needle with a 90° tip, which limited the influence of fibers on the pressure
116
measurement [88]. The pressure transducer (Honeywell Model S) had a pressure range of 689 kPa
(100 psi) and maximum operating temperature of 149 °C, with temperature compensation range
of 20-160 ˚C. A synthetic oil intended for hydraulic applications up to 204 °C (400 °F) was used
as the transfer fluid.
The layup for resin pressure measurements followed procedures for the lab-scale in situ
tests, albeit with a few key differences. Laminates were 102 mm × 102 mm and consisted of either
9 plies (commercial prepreg, [0˚/90˚/0˚]3s) or 18 plies (non-commercial prepregs, [0˚/90˚/0˚]6s) to
ensure sufficient thickness compared to the needle (final laminate thickness for both materials was
~ 4.5 mm, and the outer diameter of the needle was 1.07 mm). The laminate was placed close to
the vacuum bagging tape so that the probe could be placed through the bag and edge dams and
into the laminate. To keep the probe parallel to the tool plate and laminate, the needle was placed
between the first and second plies for tests with the commercial prepreg, and between the second
and third plies for tests with the non-commercial prepregs. Before placing the probe, a small ball
of excess resin was placed at the tip to ensure the tip remained sealed at room temperature. This
was especially a concern for prepreg with discontinuous resin films, as the probe tip would be
exposed to vacuum if placed in a resin-free region.
Testing conditions are listed in
Table 7. The conditions selected matched the extremes of the lab-scale tests, in which both
boundary conditions were either permeable to resin (breathing edge dams, perforated release film)
or sealed to resin flow (tacky tape edge dams, semi-permeable release film). Tests were conducted
using both CR and DR prepreg, as well as the commercial prepreg (Cycom 5320-1). Because of
the temperature range of the pressure transducer, tests were conducted only through the 120 min
dwell at 121 °C, and not through the rest of the thermal cycle. However, rheology tests showed
117
that resin gelation occurred during the intermediate dwell, so measuring resin pressure after this
point would not have yielded meaningful data.

Table 7: Testing conditions for resin pressure measurements. Each test was repeated twice.
Test Resin Format Plies Edge Condition Bag-side Surface Layer
Commercial Bleed Continuous 9 Bleed Bleed
Commercial Sealed Continuous 9 Sealed Sealed
CR Prepeg Bleed Continuous 18 Bleed Bleed
CR Prepreg Sealed Continuous 18 Sealed Sealed
DR Prepreg Bleed Discontinuous 18 Bleed Bleed
DR Prepreg Sealed Discontinuous 18 Sealed Sealed

The commercial prepreg was selected to provide a benchmark for results obtained with
experimental prepreg. The degree of impregnation (DoI) was expected to influence the
development of resin pressure in OoA prepregs, and a model describing the evolution of DoI for
the commercial prepreg was reported elsewhere [95]. Campbell et al. described the development
of resin pressure in a fully impregnated prepreg, noting that resin pressure was approximately
equal to applied compaction pressure until resin bleed out of the laminate caused a decrease in
resin pressure [14]. Because OoA prepregs are not fully impregnated, resin pressure is expected to
be always less than compaction pressure due to dry fibers carrying a portion of the applied load.
Additionally, resin can flow to fill dry tows regardless of boundary conditions, which is expected
to yield a reduction in pressure, with further development of resin pressure dependent on resin
format and consumables used.
118

Figure 47: Frames from in situ tests for Test C:B/B (left; continuous resin, breathing edge dams, perforated film),
Test D:B/B (center, discontinuous resin, breathing edge dams, perforated release), and Test D:S/S (discontinuous
resin, sealed edges, semi-permeable membrane). Times correspond to those marked on the thermal cycle (bottom).
3. Results and Discussion
Lab-scale samples. Time-lapse videos showed the evolution of porosity at the tool-side
surface for each testing condition, and frames for select tests are shown in Figure 47. Test C:B/B
(continuous film resin, edge breathing dams, perforated release film) is representative of
119
conventional OoA prepreg cure. At t1 (30 min, mid-way through the initial heating stage), porosity
due to entrapped air that could not be evacuated was observable. By t2 (75 min, beginning of the
intermediate dwell), these voids had migrated to pinholes in the fabric and increased in radius.
Porosity remained trapped at the surface after gelation at t3 (150 min, end of the intermediate
dwell).
In Test D:B/B the same consumables were used as with conventional OoA prepreg
processing (edge breathing dams, perforated release film). However, the prepregs in this test
featured increased through-thickness gas permeability because of the different format
(discontinuous resin). The increased permeability increased air evacuation to the surface, and no
porosity was observed at t1, after resin flowed to fill dry spots. However, new voids formed at t2,
which were attributed to evolution of volatiles as resin pressure dropped. These voids grew
throughout the intermediate dwell and remained trapped at the surface after gelation (t3).
Test D:S/S combined discontinuous resin with sealed boundary conditions (sealed edges,
semi-permeable membrane) and showed initial behavior similar to Test D:B/B, yielding a void-
free surface at t1 as resin flowed and air was evacuated. Because boundaries were resin-
impermeable, resin pressure did not drop, and no further void growth occurred through the
remaining cure. The surface of the cured laminate exhibited negligible porosity.
Cross-sections for each testing condition are shown in Figure 48, and porosity
measurements (average of two samples for each condition) are presented in Figure 49. Figure 48a
(Test C:B/B) followed conventional OoA processing and exhibited relatively minimal porosity
presenting as small (~ 0.2 mm), oblong inter-ply voids. Changing any of the consumables (Figure
48b-d) yielded an observable increase in porosity relative to Test C:B/B, as sealing laminate
boundaries limited both air and volatile removal. Voids were larger (up to ~ 1 mm in length) and
120
more numerous but retained the oblong shape. In the most extreme case, sealing all boundaries led
to pores that spanned multiple plies (Figure 48d). As DR prepreg was demonstrated to effectively
remove initially entrapped air, porosity in Test D:B/B (Figure 48e) was attributed to volatile
evolution and presented as oval or spherical voids (~ 0.1 to 0.5 mm in diameter). Sealing either
boundary independently (Figure 48f-g) yielded fewer and smaller voids (< 0.1 mm to ~ 0.3 mm in
diameter) by reducing resin bleed, and porosity in Test D:S/S contained small (< 0.1 mm in
diameter) inter-ply voids.

Figure 48: Internal images of cured in situ testing samples. The left column was produced with continuous resin
film, and the right with discontinuous film. Boundary conditions were: a and e) breathing edge dams, perforated
film; b and f) breathing edge dams, semi-permeable membrane; c and g) sealed edges, perforated film; and d and h)
sealed edges, semi-permeable membrane.
For laminates fabricated with continuous resin, void content was lowest when using
conventional VBO consumables (Test C:B/B, 0.99%). This result was not unexpected, as prepregs
121
with continuous resin layers rely solely on edge breathing to evacuate air. Thus, sealing the edges
(C:S/B) resulted in an increase in void content (4.35%). However, replacing the bag-side surface
boundary from the perforated release film to the semi-permeable membrane (C:B/S) also increased
porosity (3.47%). This finding was unexpected, since in-plane air evacuation was still enabled by
edge breathing dams. The finding indicates that a degree of air removal can occur through
perforated release films even without through-thickness gas permeability. The air removal occurs
via resin bleed that removes air bubbles and/or dissolved volatiles along with the resin itself.
Porosity was greatest when all boundaries were resin-impermeable (C:S/S), as no pathway for gas
removal existed.

Figure 49: Comparison of internal void area content (%) for continuous and discontinuous resin formats for each set
of boundary conditions. Test labels correspond to those in Table 1: either bleed (B) and sealed (S) conditions for the
edges and for the bag-side surface, respectively.
Void content for laminates produced with DR prepreg was equal to or less than the
respective CR laminates for each set of boundary conditions (Figure 49). Porosity was lowest
122
(0.23%) when all boundaries were resin-impermeable (D:S/S), a finding that was consistent with
in situ visualization tests. Through-thickness permeability enabled air evacuation through the
semi-permeable membrane, and resin pressure was maintained by restricting resin flow out of the
laminate. Varying boundary conditions demonstrated robustness imparted by discontinuous resin
patterns, as porosity remained low (< 2%) regardless of consumables used. Because the
discontinuous resin enabled air evacuation through both bag-side surface layers used (perforated
or semi-permeable) regardless of edge conditions, any porosity in these laminates was attributed
to evolution of volatiles as resin pressure dropped. Relative porosity differences between tests
would indicate which boundary accounted for more resin loss. However, except for the low-
porosity fully sealed case (D:S/S), there was no statistically significant difference in void content
between discontinuous resin laminates, and no conclusion could be drawn.
Resin pressure measurements. The commercial prepreg was used to assess and confirm the
effectiveness of the resin pressure measurement procedure. Results are shown in Figure 50, which
shows plots of temperature and resin pressure versus time. Times indicated in the plots correspond
to modeled onset of dry fiber impregnation (t1, ~ 45 ˚C), measured drop in pressure under bleeding
conditions (t2, ~ 60 ˚C), and approaching full impregnation (t3, ~ 90 ˚C) [95]. Pressure development
manifested three stages:
1) Initially (up to t1), no pressure change was detected as vacuum was applied, as resin
viscosity was high at room temperature. The measurement also confirmed that the excess
resin placed at the needle tip effectively sealed probe. As temperature increased and resin
viscosity decreased, resin pressure increased.
2) Between t1 and t2, resin pressure peaked at ~ 115 kPa abs and subsequently decreased. As
discussed in Section 2, resin pressure for fully impregnated prepregs typically peaks at a
123
value equal to the applied compaction pressure [14]. However, for semi-impregnated OoA
prepregs, the resin pressure was expected to peak at a value less than the compaction
pressure, because dry fibers share some of the applied load. The decrease in pressure
following the peak occurred regardless of boundary conditions used and was attributed to
resin flow to dry spots in the prepreg.
3) After t2 and through the remainder of the temperature cycle, resin pressure either
equilibrated at ~ 100 kPa abs (under sealed conditions) or continued to decrease (t2) before
equilibrating below ambient pressure (under high-bleed conditions).

Figure 50: Resin pressure measurements for commercial OoA prepreg under bleed (edge breathing dams, perforated
release film) and sealed (sealed edges, semi-permeable membrane) conditions. Times t 1 and t 3 are the predicted
onset and full impregnation of the material, respectively [27]. At time t 2, measured pressure deviates between
boundary conditions: the bleed test continues losing resin pressure, while pressure equilibrates under sealed
conditions.
Results from these experiments were compared to those from a model for impregnation [95]. The
peak in measured resin pressure occurred at ~ 45 ˚C, while the model predicted resin flow into the
124
fiber bed to begin at this temperature. The model also predicted that full impregnation would be
achieved at ~ 90 ˚C, and at this temperature, resin pressure had equilibrated.

Figure 51: Resin pressure measurements for the non-commercial resin under bleed (edge breathing dams, perforated
release film) and sealed (sealed edges, semi-permeable release membrane) conditions for both continuous and
discontinuous resin formats. Time t 1 is the peak in resin pressure, independent of boundary condition. At time t 2, a
second drop in resin pressure was observed under bleed conditions but not sealed conditions. Time t 3 is the end of
the first temperature ramp, during which void growth at the tool surface was observed in in situ tests.
Resin pressure measurements for the non-commercial prepregs are presented in Figure 51.
Times denoted represent key behaviors observed: t1 corresponds to the initial peak in pressure, t2
corresponds to a second drop in pressure that occurred only under bleed conditions, and t3
corresponds to observed void growth from in situ visualization tests. Trends in resin pressure
evolution were consistent with tests conducted with commercial prepreg, indicating that the initial
decrease in resin was due to impregnation of dry tows, while further pressure reduction occurred
if/when resin bled out of the laminate. Resin pressure depended strongly on boundary conditions
125
(i.e., whether resin content was retained), while no conclusive effect of resin distribution pattern
on pressure evolution was observed.
Porosity increase in laminates was caused by a drop in resin pressure during cure, which
began near the start of the intermediate dwell, at ~ 75 min. Under bleed conditions, resin pressure
equilibrated prior to the intermediate dwell, indicating that prior to this point, the reduced pressure
was not sufficient to trigger volatile evolution. However, after the intermediate dwell temperature
was reached (121 ˚C), volatile evolution caused porosity. Note that in situ tests afforded insights
to the tool-side surface only. Related studies have reported that resin pressure decreases nearer to
the site of resin bleed [14,88]. Therefore, pressure was expected to be lower near the bag-side
surface of the laminate, a location that was not observed directly.
4. Conclusions
A method was demonstrated to increase robustness of OoA prepreg processing by (a)
modifying the resin format and (b) using appropriate consumables to mitigate/eliminate porosity.
Prepregs featuring discontinuous resin provided through-thickness pathways for air and volatile
removal, reducing/eliminating laminate porosity. In situ visualization was used to assess porosity
evolution at the tool-side laminate surface during cure, identifying void formation during the
intermediate temperature dwell that was attributed to volatile evolution. To retain resin content
and therefore maintain resin pressure and suppress void growth, resin-impermeable boundary
conditions were used. Specifically, use of a semi-permeable membrane effectively mitigated void
growth by allowing for through-thickness evacuation of gas while maintaining resin content. A
method to measure resin pressure during cure was employed to test the hypothesis that resin-
impermeable boundaries (semi-permeable membrane, sealed edges) maintained higher resin
pressure than conventional OoA consumables. Results showed that regardless of whether
126
boundaries enabled resin flow out of the system, resin pressure in OoA prepregs decreased as resin
flowed to fill dry fiber tows. A subsequent observed decrease in resin pressure was unique to
conventional OoA consumables and attributed to resin flow out of the laminate, while resin
pressure equilibrated to higher value when resin-impermeable consumables were used. The
reduced resin pressure with conventional consumables was correlated to observed void growth
during in situ visualization tests. Semi-permeable membranes maintained a higher resin pressure,
thus mitigating this porosity.
Resin pressure measurements elucidated general behavior for OoA prepregs and
highlighted challenges faced in limited-compaction scenarios like VBO processing. While fully
impregnated prepregs used in autoclave processing can maintain resin pressure close to the applied
compaction pressure if resin content can be maintained, OoA prepregs experience a reduction in
the peak resin pressure from dry fibers carrying part of the applied load and resin flowing to fill
tows. Maintaining resin content in OoA prepreg processing is especially important, as the
maximum possible pressure can be only a fraction of the applied compaction pressure (already
limited to ambient pressure) and any further drop in pressure can lead to void growth.
This study investigated the utility of combining a semi-permeable release film with OoA
prepregs featuring discontinuous resin patterns to restore process robustness to an OoA process.
There was no attempt to optimize material or process parameters, yet despite this, and using
prepregs produced by hand in a lab setting, the process consistently yielded laminates with
negligible porosity (0.23%), without autoclave pressure. Insights into cure cycle optimization can
be extracted from in situ visualization of void growth combined with resin pressure measurements,
and thus guide process modifications. For example, for the non-commercial resin, the decrease in
resin pressure under bleed conditions occurred while temperature was increasing, prior to observed
127
void growth attributed to reduced resin pressure. Therefore, reducing the intermediate dwell
temperature so that the decrease in resin pressure occurs at a constant temperature could avoid
reaching conditions for void growth, even if resin bleed out of the laminate occurs. While there is
room for materials optimization, prior studies have investigated effectiveness of different resin
patterns to reduce air entrapment and minimize flow distances for discontinuous resin formats
(e.g., [13,96,97]).
This study represents an important step in the broader effort to impart process robustness
to OoA prepregs. VBO processing of prepregs is susceptible to voids both from entrapped air and
from the limited compaction pressure. The use of a semi-permeable membrane in conjunction with
discontinuous resin film effectively mitigates porosity from both sources, and resin pressure
measurements confirm that reduced resin pressure and subsequent void growth were caused by
resin bleed from the laminate. By restricting resin flow out of the laminate, semi-permeable
membranes and sealed edges were able to maintain higher resin pressure during cure than
traditional OoA consumables. The insights gained here can inform processing decisions intended
to reduce porosity in OoA processing, and the insights can be combined with material and process
optimization to further increase robustness of VBO processes.

128

Chapter 8: Conclusions
This work utilized in situ process monitoring techniques to address challenges facing
production of carbon fiber-reinforced, providing insight into the time-dependent development of
defects in different applications. This chapter summarizes outcomes of the included studies and
provides broader implications and areas for future work.
1. Honeycomb Core Sandwich Structures
1.1 Pressure dependence of bond-line development. The direct control of core pressure
afforded by the mini autoclave fixture enabled bond-line development to be correlated directly to
applied gas pressure. Results demonstrated non-monotonic behavior. At low (vacuum) pressure,
voids grew but burst, yielding a microstructure with low porosity but also poorly formed fillets.
At intermediate (ambient) pressure, voids grew but were trapped in the gelled adhesive, causing
large but very porous filets. At high (super-ambient) pressure, void growth was suppressed,
yielding void-free and well-formed fillets.
1.2 Coupling of prepreg and adhesive behavior. Results showed that, during co-cure, there
were interdependencies between the prepreg and adhesive. Particularly, relative timing of gelation
was found to be important. For the materials studied, the adhesive cured prior to the prepreg, which
created a barrier to gas evacuation. This yielded porosity in the bond-line, and was also observed
to trap voids at the tool-side facesheet surface. From this result, a mitigation strategy was
developed: if a cure cycle can be designed to gel the adhesive prior to volatilization, porosity in
the bond-line can be avoided.
1.3 Modeling of bond-line porosity. This work considered two modeling strategies. First,
a basic model was developed to predict only the onset of void growth, but not the extent. This
129
model was shown to be effective for the mitigation strategies noted above (increasing core pressure
or gelling the adhesive prior to volatilization). A particular advantage of this model is its relatively
simple parameter characterization, which can be done with data from the mini autoclave. The
second model was an integrated model coupling sub-models of various individual phenomena
related to co-cure. Validation experiments agreed numerically with sub-models, but not the
integrated model. Due to the stochasticity of co-cure, numerical prediction of bond-line porosity
is challenging. However, the model could accurately capture trends in porosity in response to
varying parameters.
1.4 Implications and future work. Results detailed here can inform processing decisions to
increase robustness of HCSS fabrication via co-cure. The time-dependent bond-line morphology
data provided by in situ visualization provided insight into mitigation strategies (increase core
pressure, gel the adhesive at a low temperature, etc.). Data also indicated that, for adhesive bond-
lines, porosity alone is not a good indication of quality, as vacuum core pressure was observed to
produce low-porosity bond-lines with poorly formed fillets.
In their current forms, the models presented can provide a quick method to screen
processing cycles but will not provide accurate numerical predictions for bond-line porosity.
Further work needs to be done to refine the integrated co-cure model, especially with consideration
for the stochasticity of the process. However, refining the model may require detailed material
characterization and be too computationally intensive to be feasible.
This work focused only on equilibrated-core co-cure, in which the core gas pressure was
controlled directly by equilibrating with the vacuum bag. Related work needs to be done on the
sealed core configuration to understand void evolution and extend the process model to such cases.
Such a model would also require integration with a model for core pressure (as a function of
130
facesheet permeability, temperature, pressure gradient, etc.) that was bypassed in the equilibrated-
core case by having the core pressure be a controlled parameter.
2. Out-of-autoclave processing with semi-permeable membranes
2.1 Utility of semi-permeable membranes. Semi-permeable membranes were demonstrated
to be effective in mitigating porosity in out-of-autoclave processing when used as a release film
for semipregs. The discontinuous resin pattern enables through-thickness permeability, while the
semi-permeable membrane allows evacuation through the surface of the laminate. This eliminated
the need for edge breathing dams, and by sealing the edges along with the resin-impermeable
release film, resin bleed and therefore resin pressure reduction was mitigated.
2.2 General resin pressure trends in OoA prepregs. Resin pressure had previously been
measured in situ for autoclave prepregs and was found to match predicted behavior based on
mechanical analogs. No such study had previously been reported for OoA prepregs, whether
conventional or semipregs. General evolution of pressure was demonstrated, which consisted of a
peak and subsequent decrease in pressure as resin flowed to fill dry areas. Notably, resin pressure
did not reach the theoretical maximum of atmospheric pressure (~100 kPa), highlighting the small
margin of error for resin pressure in OoA prepregs.
2.3 Resin pressure and in situ visualization. Resin pressure measurements demonstrated a
further decrease in pressure when using bleeding consumables that was not observed when sealed
boundary conditions were used. When synchronizing the separate data sets, comparing resin
pressure measurements to observed surface porosity evolution and measured temperature
confirmed that observed void growth with bleeding consumables was due to a decrease in resin
pressure. When boundaries were resin-impermeable, void growth did not occur.
131
2.4 Implications and future work. The semipreg format had been demonstrated previously
to increase robustness when processing challenges were related to air entrapment. Results detailed
here can be used in conjunction with prior studies to further increase robustness by mitigating
porosity due to resin pressure loss. This could reduce porosity in cases such as corners, where
consumable bridging can reduce compaction; further work is necessary to assess the application
of semi-permeable membranes to different challenging processing conditions.
This study was conducted with a non-commercial resin used for hand-made prepreg.
Despite the non-ideal material, low-porosity parts were reliably fabricated using semipregs and
semi-permeable membranes. Future work investigating semipreg optimization (resin formulation,
fiber bed selection, resin pattern design) could further increase robustness when using a semi-
permeable membrane.
With the ultimate goal being to increase OoA robustness in a commercial setting, future
work must be done to address scalability of semipreg production. For this lab-scale study, prepreg
was made by de-wetting individual squares of resin film, laying them up on dry fibers by hand,
then pressing the prepreg so the resin would stick to the fibers. To be feasible for industrial use,
semipreg production would ideally be a continuous process, e.g., a de-wetting module inserted in
a conventional prepregging machine, so that large quantities could be fabricated as conventional
OoA prepregs currently are. Scaling efforts must also take into account compatibility with optimal
material design, controllable and repeatable resin content, and economic factors such as initial
investment.

132
References
[1] Campbell FC. Manufacturing Technology for Aerospace Structural Materials. London:
Elsevier; 2006.

[2] HexCel Composites. Honeycomb sandwich design technology. HexWeb Honeycomb
Sandw Des Technol 2000:1–28.

[3] Alteneder AW, Renn DJ, Seferis JC, Curran RN. Processing and Characterization Studies
of Honeycomb Composite Structures. Proc. 38th Int. SAMPE Symp., 1993, p. 1034–47.

[4] Anders M, Zebrine D, Centea T, Nutt SR. Process Diagnostics for Co-Cure of Sandwich
Structures Using In Situ Visualization. Compos Part A Appl Sci Manuf 2018.
https://doi.org/10.1016/j.compositesa.2018.09.029.

[5] Repecka L, Boyd J. Vacuum-bag-only-curable prepregs that produce void-free parts. Int
SAMPE Symp Exhib 2002;47 II:1862–74.

[6] Xu GF, Repecka L, Mortimer S, Peake S, Boyd J. Manufacture of Void-free Laminates and
Use Thereof, 2002.

[7] Bond GG, Griffith JM, Hahn GL, Bongiovanni C, Boyd J. Non-autoclave prepreg
manufacturing technology. Int SAMPE Tech Conf 2008.

[8] Hartness JT, Xu GF. Resin Composition, A Fiber Reinforced Material Having a Partially
Impregnated Resin and Composites Made Therefrom, 2000.

[9] 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. https://doi.org/10.1016/j.compositesa.2014.09.029.

[10] Grunenfelder LK, Centea T, Hubert P, Nutt SR. Effect of room-temperature out-time on
tow impregnation in an out-of-autoclave prepreg. Compos Part A Appl Sci Manuf
2013;45:119–26. https://doi.org/10.1016/j.compositesa.2012.10.001.

[11] Grunenfelder LK, Dills A, Centea T, Nutt S. Effect of prepreg format on defect control in
out-of-autoclave processing. Compos Part A Appl Sci Manuf 2017;93:88–99.
https://doi.org/10.1016/j.compositesa.2016.10.027.

[12] Edwards WT, Martinez P, Nutt SR. Process robustness and defect formation mechanisms
in unidirectional semipreg. Adv Manuf Polym Compos Sci 2020;6:198–211.
https://doi.org/10.1080/20550340.2020.1834789.

[13] Schechter SGK, Centea T, Nutt S. Effects of resin distribution patterns on through-thickness
air removal in vacuum-bag-only prepregs. Compos Part A Appl Sci Manuf 2020;130.
https://doi.org/10.1016/j.compositesa.2019.105723.
133

[14] Campbell FC, Mallow AR, Browning CE. Porosity in carbon fiber composites an overview
of causes. J Adv Mater 1995;26:18–33.

[15] Hu W, Grunenfelder LK, Centea T, Nutt S. In situ monitoring and analysis of void evolution
in unidirectional prepreg. J Compos Mater 2018;52:2847–58.
https://doi.org/10.1177/0021998318759183.

[16] Hu W, Centea T, Nutt S. Mechanisms of inter-ply void formation during vacuum bag-only
cure of woven prepregs. Polym Compos 2020;41:1785–95.
https://doi.org/10.1002/pc.25497.

[17] Anders M, Lo J, Centea T, Nutt S. Development of a process window for minimizing
volatile-induced surface porosity in the resin transfer molding of a benzoxazine/epoxy
blend. SAMPE J 2016;52:44–55.

[18] Anders M, Zebrine D, Centea T, Nutt S. In Situ Observations and Pressure Measurements
for Autoclave Co-Cure of Honeycomb Core Sandwich Structures. J Manuf Sci Eng
2017;139:111012. https://doi.org/10.1115/1.4037432.

[19] Grimes GC. The Adhesive-Honeycomb Relationship. Appl Polym Symp 1965;3:154–90.

[20] Grove SM, Popham E, Miles ME. An investigation of the skin/core bond in honeycomb
sandwich structures using statistical experimentation techniques. Compos Part A Appl Sci
Manuf 2006;37:804–12. https://doi.org/10.1016/j.compositesa.2005.07.005.

[21] Yuan C, Li M, Zhang Z, Gu Y. Experimental Investigation on the Co-Cure Processing of
Honeycomb Structure with Self-Adhesive Prepreg. Appl Compos Mater 2008;15:47–59.
https://doi.org/10.1007/s10443-008-9056-4.

[22] Butukuri RR, Bheemreddy VP, Chandrashekhara K, Berkel TR, Rupel K. Evaluation of
skin-core adhesion bond of out-of-autoclave honeycomb sandwich structures. J Reinf Plast
Compos 2012;31:331–9. https://doi.org/10.1177/0731684412437267.

[23] Hou T-H, Baughman JM, Zimmerman TJ, Sutter JK, Gardner JM. Evaluation of Sandwich
Structure Bonding in Out-Of-Autoclave Processing. Sampe J 2011;47:32–9.

[24] Hayes BS, Seferis JC, Chen JS. Development and hot-melt impregnation of a model
controlled flow prepreg system. Polym Compos 1996;17:730–42.
https://doi.org/10.1002/pc.10665.

[25] Pearce PJ, Arnott DR, Camilleri A, Kindermann MR, Mathys GI, Wilson AR. Cause and
effect of void formation during vacuum bag curing of epoxy film adhesives. J Adhes Sci
Technol 1998;12:567–84. https://doi.org/10.1163/156856198X00795.

134
[26] da Silva LFM, Adams RD, Gibbs M. Manufacture of adhesive joints and bulk specimens
with high-temperature adhesives. Int J Adhes Adhes 2004;24:69–83.
https://doi.org/10.1016/S0143-7496(03)00101-5.

[27] Bascom WD, Cottington RL. Air Entrapment in the Use of Structural Adhesive Films. J
Adhes 1972;4:193–209. https://doi.org/10.1080/00218467208072223.

[28] Nagarajan S, Menta VGK, Chandrashekhara K, Berkel TR, Sha J, Wu P, et al. Out-of-
Autoclave Sandwich Structure: Processing Study. Sampe J 2012;48:24–31.

[29] Tavares SS, Michaud V, Månson J-AE. Through thickness air permeability of prepregs
during cure. Compos Part A Appl Sci Manuf 2009;40:1587–96.
https://doi.org/10.1016/j.compositesa.2009.07.004.

[30] Tavares SS, Caillet-Bois N, Michaud V, Månson JAE. Non-autoclave processing of
honeycomb sandwich structures: Skin through thickness air permeability during cure.
Compos Part A Appl Sci Manuf 2010;41:646–52.
https://doi.org/10.1016/j.compositesa.2010.01.013.

[31] Tavares SS, Caillet-Bois N, Michaud V, Månson J-AE. Vacuum-bag processing of
sandwich structures: Role of honeycomb pressure level on skin–core adhesion and skin
quality. Compos Sci Technol 2010;70:797–803.
https://doi.org/10.1016/j.compscitech.2010.01.015.

[32] Kratz J, Hubert P. Processing out-of-autoclave honeycomb structures: Internal core pressure
measurements. Compos Part A Appl Sci Manuf 2011;42:1060–5.
https://doi.org/10.1016/j.compositesa.2011.04.009.

[33] Kratz J, Hubert P. Vacuum bag only co-bonding prepreg skins to aramid honeycomb core.
Part I. Model and material properties for core pressure during processing. Compos Part A
Appl Sci Manuf 2015;72:228–38. https://doi.org/10.1016/j.compositesa.2014.11.026.

[34] Kratz J, Hubert P. Vacuum-bag-only co-bonding prepreg skins to aramid honeycomb core.
Part II. In-situ core pressure response using embedded sensors. Compos Part A Appl Sci
Manuf 2015;72:219–27. https://doi.org/10.1016/j.compositesa.2014.11.030.

[35] Rion J, Leterrier Y, Månson J-AE. Prediction of the adhesive fillet size for skin to
honeycomb core bonding in ultra-light sandwich structures. Compos Part A Appl Sci Manuf
2008;39:1547–55. https://doi.org/10.1016/j.compositesa.2008.05.022.

[36] Chen C, Li Y, Gu Y, Li M, Zhang Z. Prediction of the resin fillet size in honeycomb
sandwich composites with self-adhesive prepreg skin. J Reinf Plast Compos 2016;35:1566–
75. https://doi.org/10.1177/0731684416659932.

135
[37] Li M, Gu Y, Zhang Z, Sun Z. A simple method for the measurement of compaction and
corresponding transverse permeability of composite prepregs. Polym Compos 2007;28:61–
70. https://doi.org/10.1002/pc.20255.

[38] Kardos J, Duduković M, Dave R. Void growth and resin transport during processing of
thermosetting - Matrix composites. Adv Polym Sci 1986;80:101–23.
https://doi.org/10.1007/3-540-16423-5_13.

[39] Wood JR, Bader MG. Void control for polymer-matrix composites (1): Theoretical and
experimental methods for determining the growth and collapse of gas bubbles. Compos
Manuf 1994;5:139–47. https://doi.org/10.1016/0956-7143(94)90023-X.

[40] Wood JR, Bader MG. Void control for polymer-matrix composites (2): Experimental
evaluation of a diffusion model for the growth and collapse of gas bubbles. Compos Manuf
1994;5:149–58. https://doi.org/10.1016/0956-7143(94)90024-8.

[41] Ledru Y, Bernhart G, Piquet R, Schmidt F, Michel L. Coupled visco-mechanical and
diffusion void growth modelling during composite curing. Compos Sci Technol
2010;70:2139–45. https://doi.org/10.1016/j.compscitech.2010.08.013.

[42] Préau M, Hubert P. Effects of processing conditions on bondline void formation in vacuum
bag only adhesive bonding: Modelling, validation and guidelines. Int J Adhes Adhes
2018;80:43–51. https://doi.org/10.1016/j.ijadhadh.2017.10.004.

[43] Thomas C, Gerstner R. Sandwich Panel for Sound Absorption. US 7,743,884, 2010.

[44] Wilson RS. Noise Attenuation Panel. US 6,827,180, 2004.

[45] Centea T, Zebrine D, Anders M, Elkin C, Nutt SR. Manufacturing of Honeycomb Core
Sandwich Structures: Film Adhesive Behavior Versus Cure Pressure and Temperature.
Proc. Compos. Adv. Mater. Expo, 2016.

[46] Hubert P, Johnston A, Poursartip A, Nelson K. Cure kinetics and viscosity models for
Hexcel 8552 epoxy resin. Int SAMPE Symp Exhib 2001:2341–54.

[47] Van Ee D, Poursartip A. NCAMP Hexply Material Properties Database for use with
COMPRO CCA and Raven. Natl Cent Adv Mater Perform 2009:141.

[48] Lo J, Nutt S. Method for In Situ Analysis of Volatiles Generated during Cure of Composites.
Compos Part A Appl Sci Manuf 2019. https://doi.org/10.1016/j.compositesa.2019.05.013.

[49] Smallwood IM. Handbook of Organic Solvent Properties. Elsevier; 1996.
https://doi.org/10.1016/C2009-0-23646-4.

136
[50] Hermann AS, Zahlen PC, Zuardy I. Sandwich structures technology in commercial aviation.
Sandw. Struct. 7 Adv. with Sandw. Struct. Mater., Dordrecht, Netherlands: Springer; 2005,
p. 13–26.

[51] Niknafs Kermani N, Simacek P, Advani SG. A bond-line porosity model that integrates
fillet shape and prepreg facesheet consolidation during equilibrated co-cure of sandwich
composite structures. Compos Part A Appl Sci Manuf 2020;139:106071.
https://doi.org/10.1016/j.compositesa.2020.106071.

[52] Hubert P, Poursartip A. A Review of Flow and Compaction Modelling Relevant to
Thermoset Matrix Laminate Processing. J Reinf Plast Compos 1998;17:286–318.
https://doi.org/10.1177/073168449801700402.

[53] Loos AC, Springer GS. Curing of Epoxy Matrix Composites. J Compos Mater
1983;17:135–69. https://doi.org/10.1177/002199838301700204.

[54] Åström BT, Pipes RB, Advani SG. On Flow through Aligned Fiber Beds and Its Application
to Composites Processing. J Compos Mater 1992;26:1351–73.
https://doi.org/10.1177/002199839202600907.

[55] Gutowski TG, Morigaki T, Zhong Cai. The Consolidation of Laminate Composites. J
Compos Mater 1987;21:172–88. https://doi.org/10.1177/002199838702100207.

[56] Davé R. A Unified Approach to Modeling Resin Flow During Composite Processing. J
Compos Mater 1990;24:22–41. https://doi.org/10.1177/002199839002400102.

[57] Šimáček P, Advani SG. A continuum approach for consolidation modeling in composites
processing. Compos Sci Technol 2020;186.
https://doi.org/10.1016/j.compscitech.2019.107892.

[58] Hayes BS, Seferis JC, Edwards RR. Self-adhesive honeycomb prepreg systems for
secondary structural applications. Polym Compos 1998;19:54–64.
https://doi.org/10.1002/pc.10075.

[59] Niknafs Kermani N, Simacek P, Advani SG. Porosity predictions during co-cure of
honeycomb core prepreg sandwich structures. Compos Part A 2020;132:105824.
https://doi.org/10.1016/j.compositesa.2020.105824.

[60] Alteneder AW, Renn D. ., Seferis JC, Curran RN. Processing and Characterization Studies
of Honeycomb Composite Structures. 38th Internatonal SAMPE Symp., 1993.

[61] Zebrine D, Anders M, Centea T, Nutt S. Path-dependent bond-line evolution in equilibrated
core honeycomb sandwich structures. Adv Manuf Polym Compos Sci 2020;6:127–41.
https://doi.org/10.1080/20550340.2020.1800194.

137
[62] Niknafs Kermani N, Simacek P, Advani SG. A simple analysis tool to simulate the co-cure
of honeycomb core composite sandwich structures. Proc. SAMPE 2019 Tech. Conf.,
Society for the Advancement of Material and Process Engineering; 2019.
https://doi.org/10.33599/nasampe/s.19.1432.

[63] Cook BG. Sound Attenuating Structural Honeycomb Sandwich Material, 1977.

[64] Lo J, Nutt S. Method for in situ analysis of volatiles generated during cure of composites.
Compos Part A Appl Sci Manuf 2019;123:141–8.
https://doi.org/10.1016/j.compositesa.2019.05.013.

[65] Anders M, Zebrine D, Centea T, Nutt SR. Process diagnostics for co-cure of sandwich
structures using in situ visualization. Compos Part A Appl Sci Manuf 2019;116:24–35.
https://doi.org/10.1016/j.compositesa.2018.09.029.

[66] Niknafs Kermani N, Simacek P, Advani SG. A Model for the Equilibrated Co-Cure of
Honeycomb Core Sandwich Structures in Autoclave Processing. Proc. 34th Int. SAMPE
Symp. Tech. Conf. Am. Soc. Compos., American Society for Composites; 2019.

[67] Zebrine D, Centea T, Anders M, Nutt S. Process Mapping for Defect Control in the
Adhesive Bond-line of Co-cured Honeycomb Core Sandwich Structures. Proc. CAMX
2019 Conf., 2019.

[68] Kratz J, Hubert P. Vacuum bag only co-bonding prepreg skins to aramid honeycomb core.
Part I. Model and material properties for core pressure during processing. Compos Part A
Appl Sci Manuf 2015;72:228–38. https://doi.org/10.1016/j.compositesa.2014.11.026.

[69] Palit T, Centea T, Anders M, Zebrine D, Nutt S. Permeability of co-cured honeycomb
sandwich skins: effect of gas transport during processing. Adv Manuf Polym Compos Sci
2020;6:142–53. https://doi.org/10.1080/20550340.2020.1802685.

[70] Niknafs Kermani N, Simacek P, Advani S. Optimization of Process Parameters during Co-
Cure of Honeycomb Sandwich Structures. Proc. Am. Soc. Compos. Tech. Conf., 2021.

[71] Darrow DC, Propatic PA, Brayden TH. Elimination of mold surface porosity on composite
parts. J Adv Mater 1995;27:41–6.

[72] Kratz J, Hubert P. Evaluation of core material on skin quality for out-of-autoclave
honeycomb panels. Proc SAMPE 2012 Conf 2012.

[73] Brayden TH, Darrow DC. Effect of cure cycle parameters on 350°F cocured epoxy
honeycomb core panels. Proc. 34th Int. SAMPE Symp., 1989.

[74] Jouin P, Pollock D, Rudisill E. Effects of processing variables on the quality of co-cured
sandwich panels. ASTM Spec Tech Publ 1992:283–307. https://doi.org/10.1520/stp20164s.

138
[75] Tang JM, Lee WI, Springer GS. Effects of Cure Pressure on Resin Flow, Voids, and
Mechanical Properties. J Compos Mater 1987;21:421–40.
https://doi.org/10.1177/002199838702100502.

[76] Hamill L, Centea T, Nutt S. Surface porosity during vacuum bag-only prepreg processing:
Causes and mitigation strategies. Compos Part A Appl Sci Manuf 2015;75:1–10.
https://doi.org/10.1016/j.compositesa.2015.04.009.

[77] Bloom LD, Napper MA, Ward C, Potter K. On the evolution of the distribution of entrapped
air at the tool/first ply interf‘ace during lay-up and debulk. Adv Manuf Polym Compos Sci
2015;1:36–43. https://doi.org/10.1179/2055035914Y.0000000005.

[78] Martin J, Putnam JW, Hayes BS, Laborato PC, Composites C. Effect of Impregnation
Conditions on Prepreg Properties and Honeycomb Core Crush. Polym Compos 1997;18.

[79] Hsiao H-M, Lee S, Buyny R. Core Crush Problem in the Manufacturing of Composite
Sandwich Structures: Mechanisms and Solutions. AIAA J 2006;44:901–7.
https://doi.org/10.2514/1.18067.

[80] Louis BM, Hsiao K, Fernlund G. Gas permeability measurements of out of autoclave
prepreg MTM45-1/CF2426A. Int SAMPE Symp Exhib 2010.

[81] Farhang L, Fernlund G. Out-of-Autoclave Prepreg Laminates. 18Th Int Conf Compos
Mater 2011.

[82] Arafath ARA, Fernlund G, Poursartip A. Gas transport in prepregs: Model and permeability
experiments. ICCM Int Conf Compos Mater 2009.

[83] Kay J, Fernlund G. Processing conditions and voids in out of autoclave prepregs. Int
SAMPE Tech Conf 2012.

[84] Juska TD, Musser BS, Jordan BP, Hall JC. The new infusion: Oven vacuum bag prepreg
fabrication. Int SAMPE Symp Exhib 2009;54.

[85] Centea T, Hughes SM, Payette S, Kratz J, Hubert P. Scaling challenges encountered with
out-of-autoclave prepregs. 53rd AIAA/ASME/ASCE/AHS/ASC Struct Struct Dyn Mater
Conf 2012 2012:1–15. https://doi.org/10.2514/6.2012-1568.

[86] Grunenfelder LK, Nutt SR. Void Formation in Composite Prepregs-Effect of Dissolved
Moisture. Compos Sci Technol 2010;70:2304–2309.

[87] Centea T, Hubert P. Out-of-autoclave prepreg consolidation under deficient pressure
conditions. J Compos Mater 2014;48:2033–45.
https://doi.org/10.1177/0021998313494101.

139
[88] Lynch K, Hubert P, Poursartip A. Use of a simple, inexpensive pressure sensor to measure
hydrostatic resin pressure during processing of composite laminates. Polym Compos
1999;20:581–93. https://doi.org/10.1002/pc.10381.

[89] Ma Y, Centea T, Nutt SR. Defect reduction strategies for the manufacture of contoured
laminates using vacuum BAG-only prepregs. Polym Compos 2017;38:2016–25.
https://doi.org/10.1002/pc.23773.

[90] Hughes SM, Hubert P. Out-of-autoclave prepreg processing: Effect of integrated geometric
features on part quality. Int SAMPE Tech Conf 2013.

[91] Lane SA, Higgins J, Biskner A, Sanford G, Springer C, Berg J. Out-of-autoclave composite
fairing design, fabrication, and test. J Manuf Sci Eng Trans ASME 2011;133:1–10.
https://doi.org/10.1115/1.4004321.

[92] Bernetich KR. Evaluation of detail part fabrication using out-of-autoclave prepreg. Proc
SAMPE 2010 2010.

[93] Dang C, Bernetich K, Carter E, Butler G. Mechanical comparison of out-of-autoclave
prepreg part to conventional autoclve prepreg part. Am. Helicopter Soc. 67th Annu. Forum,
2011, p. 3–5.

[94] Ganapathi AS, Maheshwari M, Joshi SC, Chen Z, Asundi AK, Tjin SC. In-situ measurement
and numerical simulation of resin pressure during Glass/Epoxy prepreg composite
manufacturing. Meas J Int Meas Confed 2016;94:505–14.
https://doi.org/10.1016/j.measurement.2016.08.028.

[95] Centea T, Hubert P. Modelling the effect of material properties and process parameters on
tow impregnation in out-of-autoclave prepregs. Compos Part A Appl Sci Manuf
2012;43:1505–13. https://doi.org/10.1016/j.compositesa.2012.03.028.

[96] Schechter SGK, Grunenfelder LK, Nutt SR. Design and application of discontinuous resin
distribution patterns for semi-pregs. Adv Manuf Polym Compos Sci 2020;6:72–85.
https://doi.org/10.1080/20550340.2020.1736864.

[97] Schechter SGK, Grunenfelder LK, Nutt SR. Air evacuation and resin impregnation in semi-
pregs: effects of feature dimensions. Adv Manuf Polym Compos Sci 2020;6:101–14.
https://doi.org/10.1080/20550340.2020.1768348.

Abstract (if available)

Abstract Carbon fiber-reinforced polymer composites consist of a cross-linked polymer (e.g., epoxy) matrix supported by woven or unidirectional carbon fibers. Offering increased mechanical performance metrics such as stiffness at low weight compared to conventional engineering materials, such composites are used in weight-critical applications such as aerospace. Manufacturing defects such as porosity in the polymer matrix, however, can reduce mechanical properties by serving as crack propagation sites. Traditionally, investigations into causes of and mitigation strategies for porosity in CFRP composites – as well as process optimization and qualification in industry – have relied on the analysis of cured samples. This process can be time- and resource-intensive, and only gives insight into final porosity as an output in response to material (resin type, solvent content, flow characteristics, etc.) and process (temperature, pressure, consumables, etc.) parameters. In this work, two manufacturing cases are investigated using in situ analysis methods to track real-time porosity development to clarify the underlying physical mechanisms by which voids form, grow, and evacuate during composites processing. For the co-cure of honeycomb core sandwich panels, in which prepreg facesheets are consolidated and adhered to a low-density core in a single thermal cycle, a custom instrumented fixture was utilized to visualize the development of porosity in the adhesive bond-line during cure. Behavior was dependent on gas pressure within the core as well as interactions between the prepreg resin and adhesive. Insights led to the development of an integrated, physics-based model for bond-line formation, which experimental validation showed could reliably capture trends in porosity development in response to material and process inputs. In situ visualization was also used to assess void growth during cure at the prepreg-tool interface. For sandwich panels, real-time imaging of the surface demonstrated that non-uniform compaction conditions imposed by the honeycomb core geometry led to the evolution of residual solvent that remained trapped as porosity. In a second manufacturing case, in situ visualization was used to analyze void growth at the prepreg-tool interface during out-of-autoclave processing of composite laminates. Varying consumables were used to impose different boundary conditions, and results indicated that the combination of discontinuous resin film and a semi-permeable release film on the bag-side surface could evacuate initially entrapped air while retaining resin pressure to mitigate voids due to volatile evolution. Through this work, in situ process analysis including visualization techniques clarified time-dependent behavior and underlying physics of void formation during composites processing. Insights gained can inform manufacturing decisions and model development, reducing time and material waste and leading to more efficient manufacturing of high-quality composite parts.

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

In situ process analysis for defect control during composites manufacturing

PDF

Void evolution in vacuum bag-only prepregs

PDF

Developing efficient methods for the manufacture and analysis of composites structures

PDF

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

PDF

Material and process development and optimization for efficient manufacturing of 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

Polymer flow for manufacturing fiber reinforced polymer composites

PDF

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

PDF

Sound transmission through acoustic metamaterials and prepreg processing science

PDF

Ageing and mechanical failure of fiber reinforced polymers

PDF

Development of composite oriented strand board and structures

PDF

Chemical depolymerization of amine-epoxy cured carbon fiber reinforced polymer composites and their re-use

PDF

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

PDF

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

PDF

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

PDF

The effect of lattice structure and porosity on thermal conductivity of additively-manufactured porous materials

Asset Metadata

Creator Zebrine, Daniel (author)

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

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Materials Science

Degree Conferral Date 2022-08

Publication Date 08/01/2022

Defense Date 06/17/2022

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

Tag autoclave,carbon fiber,carbon fiber reinforced polymer composites,co-cure,composites,honeycomb core sandwich structures,in situ,Modeling,OAI-PMH Harvest,out of autoclave,porosity,process monitoring

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Nutt, Steven (committee chair), Grunenfelder, Lessa (committee member), Ravichandran, Jayakanth (committee member), Williams, Travis (committee member)

Creator Email dzebrine@gmail.com,zebrine@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111375861

Unique identifier UC111375861

Legacy Identifier etd-ZebrineDan-11049

Document Type Dissertation

Format application/pdf (imt)

Rights Zebrine, Daniel

Type texts

Source 20220801-usctheses-batch-964 (batch), 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