Defect Reduction Strategies for the Manufacture

of Contoured Laminates Using Vacuum BAG-Only

Prepregs

Yijia Ma, Timotei Centea, Steven R. Nutt

Department of Chemical Engineering and Materials Science, University of Southern California,
Los Angeles, California 90089-0241, USA

Complex structures manufactured using low-pressure autoclave-level microstructural and geometric quality
vacuum bag-only (VBO) prepreg processing are more without difficulty. However, the limited compaction pres-
susceptible to defects than flat laminates because of
complex compaction conditions present at sharp cor-
sure available during manufacturing and the distinctive
ners. Consequently, effective defect mitigation strat- characteristics of VBO prepregs render the production of
egies are required to produce structural parts. In this parts with complex geometries challenging.
study, we investigated the relationships between lami- The two defects most commonly observed during com-
nate properties, processing conditions, mold designs, posites manufacturing are microstructural voids and
and part quality in order to develop science-based
guidelines for the manufacture of complex parts.
dimensional nonuniformity. Void content is a primary
Generic laminates consisting of a central corner and concern because of its effect on the mechanical properties
two flanges were fabricated in a multipart study that [2–4], and the mechanisms of void formation in prepreg
considered variation in corner angle and local curva- laminates have been studied extensively [4]. It is now
ture radius, the applied pressure during layup and widely accepted that gas-induced voids can be caused by
cure, and the prepreg material and laminate thickness.
The manufactured parts were analyzed in terms of absorbed moisture, residual solvents within the resin, and
microstructural fiber bed and resin distribution, thick- air entrapped between and within prepreg plies [4, 5].
ness variation, and void content. The results indicated Furthermore, flow-induced voids can arise if the resin
that defects observed in corner laminates were influ- matrix does not fully saturate the fiber bed during cure.
enced by both mold design and processing conditions However, in situ monitoring of void formation within
and those optimal combinations of these factors can
mitigate the defects and improve quality. POLYM. COM- laminates remains difficult to perform, and the exact
POS., 00:000–000, 2015. V
C 2015 Society of Plastics Engineers mechanisms that govern void nucleation, growth, and
migration are not entirely understood. Process-induced
deformations, such as thickness, variation, and laminate
warpage, are similarly complex as they can be caused by
INTRODUCTION microstructural features, process conditions, and phenom-
Out-of-autoclave (OoA) prepreg processing has been ena such as cure shrinkage, anisotropic thermal expan-
developed as an effective alternative to autoclave cure. sion/contraction, and tool-part interactions.
Although autoclaves use a pressurized environment to During autoclave cure, defects are typically suppressed
suppress voids and mitigate geometric variability [1], by increasing the compaction pressure. Conversely, for
OoA processing consists of curing specially designed pre- low-pressure VBO prepreg cure, the compaction is primar-
pregs using vacuum bag-only (VBO) method, and can ily governed by a partially impregnated prepreg micro-
provide technical, economic, and environmental benefits structure that features pervasive and highly permeable dry
when compared with conventional manufacturing. OoA regions within the fiber tows [6]. These “evacuation
processing can be used to produce flat laminates with channels” allow entrapped air or vaporized moisture to
migrate toward the laminate boundaries and escape into
Correspondence to: Yijia Ma; e-mail: [email protected] the bag during the early stages of cure process. They are
Contract grant sponsor: National Science Foundation G8 Research Pro- subsequently infiltrated by nearby resin to form a fully sat-
ject of Interdisciplinary Program on Material Efficiency; contract grant urated microstructure. Despite these characteristics, higher
sponsor: Contract grant number: CMMI-1229011; Contract grant spon- void contents have been associated with elevated levels of
sors: Cytec Industries; Airtech International.
DOI 10.1002/pc.23773
dissolved moisture [5, 7], excessive out-time [8], and low-
Published online in Wiley Online Library (wileyonlinelibrary.com). temperature cure cycles [4] that inhibit resin flow into the
C 2015 Society of Plastics Engineers
V dry evacuation channels. Such defect sources are relevant

POLYMER COMPOSITES—2015
to both flat and complex geometries. Fortunately, their kling formation in L-shape laminates governed by the
effects may be minimized by rigorous process control dur- shear force generated by mismatches of coefficient of
ing material handling, layup, and cure. Owing to their par- thermal expansion between tool and laminate, as well as
tial impregnation, VBO prepreg plies have a higher bulk ply slippage during compaction, and experiments based
factor (ratio of initial thickness to final thickness) than on this mechanism were able to minimize wrinkles and
comparable autoclave prepregs [9]. Consequently, they severe in-plane misalignments. Brillant and Hubert [10,
must undergo greater compaction during cure, and may be 19] studied the effect of different bagging configurations
more difficult to conform to nonflat geometries. on thickness uniformity and void content in L-shaped
Investigators have identified several issues that arise spe- concave and convex laminates with various laminate
cifically during the manufacturing of complex shape lami- thicknesses and corner radii. Their results showed that
nates. Ply layup over curved surfaces can pose difficulties void content and thickness variation could be greatly
due to the limited drapeability of the fiber bed. Furthermore, influenced by lay-up strategies. Levy et al. [20] proposed
at corners, the pressure distribution between the mold-side an analytical corner compaction model for L-shaped lami-
and bag-side surfaces of thick laminates can be uneven, cre- nates that accounted for local pressure difference and
ating nonuniform compaction conditions [10]. This effect is interplay friction, and data from the model agreed with
particularly pronounced for prepreg product forms with high experiments. Cauberghs and Hubert [21] focused on the
bulk factors. Prepreg and consumable materials can bridge effect of Z-shaped tight corner geometry with one convex
over concave molds or wrinkle over convex tooling [11], corner between two concave corners and the levels of
introducing in-plane stresses and adversely reducing the connectivity required between ply drop-offs and the vac-
compaction pressure over mold corners. Finally, large defor- uum source. They showed that resin accumulation in con-
mations can affect the initial prepreg microstructure, and cave corner regions was associated with consumable
hence, alter its original permeability. Although void content bridging, and low void content could be achieved by con-
remains a major concern, commonly observed microstruc- necting the dropped plies to the vacuum source.
tural defects in complex shape laminates also include ply More complex OoA/VBO composite structures were
wrinkling, corner thickening or thinning for parts manufac- also evaluated in previous studies. For example, Grunen-
tured on concave or convex tools, respectively, and, in felder et al. [22] studied the void content in embedded
extreme cases, delamination [11]. Moreover, resin cure doublers that consisted of a local build-up of additional
shrinkage and anisotropic thermal expansion can induce plies in regions of laminates and hat-stiffeners. Results
residual stresses in angle laminates during autoclave proc- showed that voids within the embedded plies could be
essing, leading to dimensional changes such as spring-in at reduced by fostering air evacuation through prepreg plies
corners and warpage in flat sections [12, 13]. impregnated on only one side and longer vacuum hold
Resin flow and fiber bed compaction are critical gov- times before cure. Finally, Hughes and Hubert [23] inves-
erning factors for the defect formation in complex shape tigated how quality changes with part size and complexity
laminates. Percolation flow during which the resin trans- scale-up. They observed that although the microstructural
port occurs relative to a compacting fiber bed, and shear quality of small-scale samples was indicative of what
flow during which the fiber bed and resin are considered were observed in samples extracted from a large-scale
an inextensible viscous fluid, are the two major mecha- part, the physical and manufacturing parameters govern-
nisms of flow behavior during cure [14]. The compaction ing part quality could interact to increase void content
of angle laminates involves the interaction of both mecha- and render industrial-scale manufacturing challenging.
nisms [11, 15]. In addition to resin flow and fiber bed The literature cited above clarified several phenomena
compaction, studies also show that during autoclave proc- governing the manufacturing of complex shapes, and con-
essing, the lay-up method has a significant effect on the firmed that deviations from a flat geometry introduce
thickness uniformity of cured laminates, whereas the cor- additional defect sources during VBO processing. How-
ner curvature radius, flange length, and laminate thickness ever, the fundamental aspects of VBO prepreg compac-
have less effect on the thickness uniformity of the cured tion, which are affected by local regions with complex
quasi-isotropic laminates than on the [908]n laminates geometry, remain unclear. Moreover, most studies have
[16]. Within this context, the low compaction pressure focused on L-shaped or C-shaped laminates with 908
available during VBO cure and the distinctive flow and angles, and have not analyzed the effects of other geome-
compaction characteristics of VBO prepregs are likely to tries, or of increasing geometric complexity. Finally,
further complicate the fabrication of complex parts. effective strategies to limit defect levels and ensure part
Studies on the OoA/VBO manufacturing of laminates quality remain elusive, but highly sought after.
with corners or other sharp curvatures have been reported.
Laminates manufactured over concave (female) tools
Objectives and Structure
exhibited resin accumulation and corner thickening,
whereas corner thinning was observed for convex (male) Several factors are typically considered during studies
tooling, in agreement with their autoclave counterparts that seek to optimize processing. Some, including the pre-
[17]. Lightfoot et al. [18] suggested a mechanism of wrin- preg format and ply stacking sequence, are “intrinsic” to

2 POLYMER COMPOSITES—2015 DOI 10.1002/pc

FIG. 1. Schematic illustration of mold design and bagging configuration. [Color figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]

the part and cannot be easily adjusted due to material angles (308, 458, 608) were chosen to evaluate the effect
availability, design constraints, and desired mechanical of increasing geometric complexity on laminate quality,
properties. Others, such as processing conditions and tool while various corner radii (0 mm, 6.35 mm, 9.53 mm,
design, are “extrinsic” and can be altered to maximize 12.7 mm) were selected to verify the effect of local geo-
quality if the relevant scientific relationships are under- metric discontinuity. The tools were machined from a sin-
stood. In this work, we seek to identify the effect of key gle billet of aluminum to avoid leaks, and surfaces were
processing conditions and tool characteristics (extrinsic fine-polished.
factors) on part quality for several prepreg and laminate
combinations (intrinsic factors). We focused on a generic Materials
part geometry consisting of a corner located between two
flanges. First, we studied the effect of increasing geome- Experiments were carried out using two kinds of pre-
try complexity by varying tool corner angles for two pre- pregs designed for OoA processing. The first, denoted
preg materials and various thicknesses. Then, we “prepreg A,” consisted of a toughened epoxy resin (Cycom
investigated three feasible process modifications: variation 5320) and a five-harness satin (5HS) carbon fiber fabric
to local corner radii, the use of intermediate vacuum hold (T650-35 3K). The second, “prepreg B,” was comprised of
before oven cure, and the application of pressure intensi- a later-generation OoA resin (Cycom 5320-1) and an
fiers at corners. This study aims to expand the existing eight-harness satin (8HS) carbon fiber fabric (T650-35
knowledge of OoA processing by clarifying the compac- 3K). Both fabrics possessed the same areal weight (375 g/
tion mechanisms that lead to defects in complex shape m2). The two resins can, and were, cured at the same tem-
laminates, and form the basis for practical guidelines for perature. However, the resin of prepreg B was less reac-
the selection of mold designs and processing parameters tive, exhibiting a longer out-life (30 days at ambient
that mitigate defects and improve part quality. conditions, versus 21 days for prepreg A), thus requiring a
longer cure at a given temperature. Note that the freezer
METHODS storage time of prepreg A exceeded the 12 months freezer
storage time stated by the manufacturer, resulting in
Corner laminates were manufactured using VBO pre- reduced tack and potentially other resin degradation phe-
preg processing. Two extrinsic factors, mold design and nomena, whereas prepreg B underwent negligible storage
pressure application methodology, were considered in time, and was therefore comparatively pristine. Both mate-
detail. These factors are specifically relevant to nonflat rials were used within the study because, as described in
geometries, and can be adjusted feasibly within a produc- the following, prepreg A was more prone to voids and con-
tion environment. In addition, several intrinsic factors, sequently highlighted several trends.
including prepreg product and laminate thickness, were
also varied.
Cure
The laminates were manufactured using traditional
Mold Design
OoA/VBO prepreg processing protocols. Figure 1 shows
The typical tooling used in this study is shown sche- details of the layup configuration. Prepreg plies meas-
matically in Fig. 1. A total of six tools were used to man- uring 76.2 mm (width) by 63.5 mm (flange) by 63.5 mm
ufacture parts. Each corner angle was associated with two (flange) were laid up on nonperforated release film in the
tools, which featured sharp corners or corners with finite 08 direction (along the laminate length). Edge-breathing
radii of curvature. Each tool included both concave and dams, which provided paths for air evacuation while pre-
convex corner regions, allowing concave and corner lami- venting resin bleed, were placed along the laminate perim-
nates to be manufactured simultaneously. Specific corner eter. A second nonperforated release film was placed on

DOI 10.1002/pc POLYMER COMPOSITES—2015 3

 TABLE 1. Test matrix.

Set Material Debulk Pressure strip Tool shape Corner radius/mm Corner angle No. of plies

I Prepreg A None None Concave 0 308, 458, 608 4, 8

None None Convex 0 308, 458, 608 4, 8
II Prepreg B None None Concave 0 308, 458, 608 4, 8
None None Convex 0 308, 458, 608 4, 8
III Prepreg B Yes None Concave 0 308, 458, 608 8
Yes None Convex 0 308, 458, 608 8
IV Prepreg B None None Concave 9.53, 12.7 308, 458, 608 8
None None Convex 6.35, 9.53 308, 458, 608 8
V Prepreg B None Yes Concave 0, 9.53, 12.7 308, 458, 608 8

top of the laminate. Two layers of breather were applied inward-facing cross-sectional areas were polished to 2400
evenly on the top of the parts within the bag. In some grit abrasive using a metallographic grinder/polisher
cases, pressure strips were used to intensify the compaction (Buehler MetaServ). A digital stereo microscope (Key-
pressure locally, at the corners. The vacuum bag was ence VHX-600) was used to capture micrographs over the
finally set, with excess bag material located at the corner entire cross-sectional area at a 1003 magnification. These
region to prevent bridging. After a vacuum bag was individual micrographs were stitched into a complete
assembled and vacuum was drawn with a stand-alone vac- image of the cross-section. For convenience, the complete
uum pump (Busch R5), a leak test was performed by image was divided into a corner region and two flange
measuring the pressure inside the vacuum bag using a regions.
pressure gauge to ensure that sufficient vacuum was drawn Laminate quality was analyzed on the basis of void
in the bag and no leak was present. First, the vacuum content and thickness variation, two frequently used
pump was disconnected from the bag. Then, the gauge was defect metrics that can be accurately quantified via light
used to ensure that the vacuum level within the bag did microscopy. The void content was measured using image
not decrease by more than 1” Hg (3.3 kPa) in 5 minutes, analysis software (ImageJ) by converting the stitched
as recommended by the manufacturer. The laminates were color image to 8-bit binary image, selecting the void
cured in a convection oven (Thermal Product Solutions regions, applying a binary threshold, and calculating the
Blue M). Manufacturer recommended cure cycles were ratio of void area to total cross-sectional area within the
used in all laminates, consisting of a 4-hour room tempera- region of interest. The void contents within the flange
ture vacuum hold before cure, a 1.58C/min ramp rate, and and corner regions were calculated individually for com-
a 1218C dwell of 2 hr for prepreg A or 3 hr for prepreg B. parison. To quantify the thickness variability, nine mea-
surement locations were selected, with three located in
Test Matrix the left flange, three in the corner, and three in the right
flange (Fig. 2). The thickness at each individual location
The five sets of manufactured parts are summarized in (xi) was then measured from the micrographs. The coeffi-
Table 1. First, Sets I and II studied the effect of laminate cient of variation (CoV) of the thickness, shown in equa-
thickness (four and eight plies), geometric complexity tion 2, was used to compute a single numerical metric for
(corner angles of 308, 458, and 608) and corner type (con- the thickness nonuniformity of the entire laminate (n is
cave and convex). These laminates comprised a baseline the number of measurement locations). Other microstruc-
dataset that was used to evaluate the effect of further pro- tural features, such as resin accumulation, were analyzed
cess modifications. In Set III, the influence of intermedi- individually as described in the following section.
ate debulking was assessed by applying a five-minute
vacuum hold after laying down every two plies. In Set x5ðx1 1x2 1x3 1x4 1x5 1x6 1x7 1x8 1x9 Þ=9 (1)
IV, the effect of mold corner curvature was analyzed, sX
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
using corner radii of 9.53 mm and 12.7 mm for concave 1 ðx2 x Þ2
corners and 6.35 mm and 9.53 mm for convex corners. Coefficient of Variation5 (2)
x ðn21Þ
Finally, in Set V, the local pressures at the concave cor-
ners were intensified by the application of pressure strips.
One laminate was manufactured for each test point. RESULTS AND DISCUSSION

Quality Analysis Sets I and II–General Defects in Sharp Corner Laminates

Two samples were cut along the laminate length from The two main defects observed in laminates were
each laminate for analysis. To avoid edge effects, samples thickness variation and void content, both of which were
were taken no closer than 10 mm from the end. The two concentrated in corner regions. Figure 3a shows void

4 POLYMER COMPOSITES—2015 DOI 10.1002/pc

FIG. 2. Thickness measurement locations for (a) concave corner laminate, (b) convex corner laminate, (c) concave corner laminate manufactured
with a pressure strip. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

contents in laminates made of prepreg A as a function of trend was not consistent in convex corner parts. For a
laminate thickness in flange regions, concave corner given corner angle, the CoV decreased as the laminate
regions and convex corner regions, respectively. The thickness increases in both mold types. Similar thickness
average void contents over the entire cross-sections were- variation behavior was observed in prepreg B laminates,
< 2% in all prepreg A laminates. However, void contents as shown in Fig. 4, indicating that prepreg age affected
were < 0.5% in laminates made of prepreg B, which was the amount of voids but not the variation in laminate
much less than those in the corresponding prepreg A sam- thickness, and that mechanisms governing laminate
ples. The relatively high void content in prepreg A lami- dimensional uniformity were therefore generally applica-
nate was attributed to the prepreg age, which resulted in ble. Laminates with a wider flange on one side (127 mm
reduced tack and potentially compromised the ability for [flange 1] versus 63.5 mm [flange 2]) were also manufac-
resin to flow and eliminate entrapped air. For prepreg A, tured to evaluate the effect of flange width, and results
the void content increased with laminate thickness in a showed that flange width had no significant effect on
quasi-linear manner in all three regions, but the defect thickness variation of laminate or on fiber distribution.
level increased more markedly in the concave corner Figure 5 shows representative micrographs of lami-
regions than in the convex corner regions, where behavior nates manufactured over concave and convex molds.
was similar to that of the flange areas. In contrast, the Microstructural defects were observed, including voids,
void content in prepreg B was negligible, and no clear corner thickening and corner thinning, and resin accumu-
trend was observed. lation. Generally, more defects were observed in the con-
Figure 3b shows the coefficient of variation (CoV) for cave molded laminates than in their convex counterparts.
thickness of concave and convex laminates made of pre- In the concave corner, resin was accumulated beyond the
preg A. The thickness variation in concave parts was gen- fiber bed, resulting in greater thickness changes in corner
erally greater than those in convex parts. Furthermore, in regions, whereas curvature radii of the fiber bed were
the concave laminates, for a given number of plies, the consistent through the thickness. In the convex corner, no
CoV increased linearly with corner angle. However, this resin accumulation was observed, although the radius of

FIG. 3. Laminates made of prepreg A. (a) void content in flange region, concave corner region and convex corner region, (b) thickness variation.
[Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

DOI 10.1002/pc POLYMER COMPOSITES—2015 5

 ally comparable. Locally, the fiber bundles nearest the
mold had a curvature radius of 0.8 mm, which was much
less than that of the far-field plies (4 mm). This specific
difference implied that in convex laminates, corner thin-
ning arose because of increased compaction pressure at
the sharp corner region but was limited by the average
stiffness of the fiber bed during low-pressure processing.
To support this assertion, the length of the accumu-
lated resin region in concave corners was measured
directly from micrographs (as shown previously in Fig.
5a). Figure 7 shows the concave corner resin length as a
function of corner angle in prepreg B laminates. The resin
length generally increased with corner angle, just as the
CoV of the overall thickness increased, confirming that
FIG. 4. Thickness variation of laminates made of prepreg B. [Color thickness variation in concave laminates was generated
figure can be viewed in the online issue, which is available at wileyonli- primarily by resin accumulation.
nelibrary.com.] From the results described above, we concluded that
defect formation mechanisms in concave corner laminates
curvature of the fiber on the mold side was significantly were more prone than their convex counterparts to parame-
less than others on the bag side, leading to convex corner ters such as corner angle and laminate thickness. In con-
thinning and increased thickness variation. cave parts, both the inability of the fiber bed to conform to
To identify the fundamental causes of thickness nonun- the tool under low compaction pressure, and the conse-
iformity, microscopic inspection was used to measure the quent resin accumulation, adversely influenced the lami-
CoV of the fiber bed thickness, independent of the overall nate dimensional uniformity. Furthermore, low compaction
laminate thickness. Figure 5a shows how the measure- force also increased the likelihood of void formation. Con-
ment of both thickness values was performed at a given versely, the stiffness of fiber beds in convex corners pre-
location. Figure 6 compares the overall thickness varia- vented corner thinning during low-pressure processing,
tion measured from such micrographs to that of the fiber thus reducing thickness variation. The dataset presented
bed thickness in prepreg B laminates. Figures 6a and b above forms a benchmark against which further process
display graphs of four-ply and eight-ply data, respec- modifications were evaluated.
tively. In concave corner laminates, the CoV of the fiber
bed was significantly less than that of the laminate thick-
ness, but remained greater than that of the flanges. This Set III–Effect of Intermediate Debulk Method
difference, along with observations of consistent fiber Intermediate debulking can improve compaction and
curvature radius through the thickness, suggested that reduce void content in laminates by ensuring that individ-
thickness variation in concave laminates was caused ual prepreg plies conform to substrate layers during layup,
mainly by resin accumulation at the concave corner. The and by allowing entrapped air pockets to escape more
stiffness of the fiber bed and its inability to conform to readily. Laminates consisting of eight plies of prepreg B
the sharp mold radius during low pressure processing was were cured on both concave and convex tools, with vac-
a contributing but minor cause. Conversely, in convex uum compaction applied every two plies for 5 minutes
corner laminates, the two thickness data sets were gener- during layup. Microscopic inspection of polished sections

FIG. 5. Defects in corner laminates. (a) concave corner, (b) convex corner. [Color figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]

6 POLYMER COMPOSITES—2015 DOI 10.1002/pc

FIG. 6. Overall thickness variation and fiber bed thickness variation in prepreg B laminates. (a) four-ply laminates, (b) eight-ply laminates. [Color
figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

revealed that most of these laminates were void-free. This and 12.7 mm) were selected for the concave tooling than
represented a marked improvement from laminates manu- for the convex tooling (6.35 mm and 9.53 mm). No
factured without intermediate debulking, which exhibited debulking was carried out.
voids. Inspection of polished sections revealed that resin
Figure 8 shows a comparison of thickness variation in accumulation within the corner regions was almost elimi-
laminates with select corner angles processed with and nated in all cases. Figure 9a and b show the CoVs in con-
without debulking. The CoV was not strongly affected by cave and convex laminates versus mold radius. In
debulking, and no clear trend was observed across all cor- concave laminates, the CoV decreased with increasing
ner angles. The inconsistent and negligible nature of the corner radius. Moreover, the difference in thickness uni-
effect was confirmed by comparing the corner resin formity between corner angles was also greatly reduced.
length in concave laminate, which was also insensitive to Molds with large corner radii required the fiber bed to
debulking. These results indicated that the debulking steps deform less, and therefore prevented resin accumulation.
used in this study contributed to the removal of entrapped In concave laminates fabricated on molds with radii of
air and reduced the void content, but did not affect the 12.7 mm, the CoV decreased close to that of the flanges
conformation of the prepreg laminate to the tool. (about 0.02). In convex laminates, larger corner radii also
decreased the thickness variation, but the effect was less
Set IV–Effect of Mold Corner Radius pronounced than in concave corner parts, as only the pre-
preg plies closest to the tool were affected.
Eight-ply laminates of prepreg were manufactured on These observations indicated that increasing the radius
molds with varying corner radii. Because concave corner of curvature of the mold was an effective and practical
laminates were shown to be more prone to defects than approach to reduce (or control) thickness variation at cor-
convex corner laminates, larger corner radii (9.53 mm ners in low-pressure VBO processing, especially when

FIG. 7. Concave corner resin length in laminates made of prepreg B. FIG. 8. Thickness variation of laminates manufactured with and with-
[Color figure can be viewed in the online issue, which is available at out debulking. [Color figure can be viewed in the online issue, which is
wileyonlinelibrary.com.] available at wileyonlinelibrary.com.]

DOI 10.1002/pc POLYMER COMPOSITES—2015 7

FIG. 9. Thickness variation of laminates manufactured with various tool corner curvatures. (a) concave corner laminates, (b) convex corner laminates.
[Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

concave tooling was used. In addition, increasing the cor- Figure 10 presents the thickness variation of laminates
ner curvature could also counteract the detrimental effects produced with and without pressure strips. For molds
of increasing geometric complexity, allowing parts with with corner radii of 0 mm and 9.53 mm, thickness varia-
large corner angles to attain the same dimensional consis- tion was greatly reduced by deployment of pressures
tency as those with smaller angles. strips, indicating that the insert redistributed pressure and
concentrated compaction force at corner regions, increas-
ing the extent of conformation of the fiber bed to the tool
Set V–Pressure Strip Application
surface. However, the pressure strip did not affect lami-
The pressure distribution conditions at a concave cor- nates produced using the mold with a corner radius of
ner can be improved by using a pressure strip (a silicone 12.7 mm, as the tool geometry already reduced the CoV
insert) placed between the bag-side release film and the close to flange levels.
breather. The strip concentrates the pressure at the corner These results indicated that the application of pressure
region, whereas potentially mitigating the effects of con- strips was an effective method for reducing thickness var-
sumable bridging. Laminates comprised of eight plies of iation in laminates produced over concave molds with rel-
prepreg B prepreg were fabricated using pressure strips atively low corner curvatures (0 mm and 9.53 mm). They
over molds with sharp corner radii (as in Sets I and II) are therefore well suited to situations in which the part
and corner radii of 9.53 mm and 12.7 mm (as in Set IV). geometry is tightly prescribed and cannot be locally
During layup, the pressure strip material was carefully adjusted by changing the mold geometry.
shaped into the form of the concave mold. Debulking was Although the use of pressure strips effectively reduced
not performed. CoVs in concave laminates, poorly compacted areas (or
laminate fillets) were sometimes observed at the boundary
of the region of application, as shown in Fig. 11a. These
defects were sensitive to the manner in which the pressure
strip was placed, and could be eliminated by carefully
designed layup. However, they reduced the robustness of
the process. In addition, the use of pressure strips was also
associated with concentrated void distributions along the
pressure strip, as shown in Fig. 11b. These voids were
larger than those observed in previous laminates, and
exhibited elongated ellipsoidal shapes. Increasing the
room-temperature vacuum hold time from 4 hours to 8
hours eliminated the voids. These defects can be attributed
to the modified pressure distribution caused by the pressure
strip. The fillets corresponded to the edge of the strip,
which experienced a local and steep reduction in compac-
tion pressure. The concentrated pressure caused by the strip
FIG. 10. Thickness variation of laminates manufactured without and
could also possibly decrease the air permeability of the
with a pressure strip. [Color figure can be viewed in the online issue, prepreg, reducing the rate of air evacuation and, for an
which is available at wileyonlinelibrary.com.] equivalent ambient evacuation time, causing higher void

8 POLYMER COMPOSITES—2015 DOI 10.1002/pc

 tion. The pressure strip effectively concentrated pressure in
concave corners, and improved the quality of laminates fabri-
cated on molds with tight corner radii, although air evacuation
was partially impeded.
The results presented here can be used as guidelines
for reducing defects and optimizing a manufacturing
cycle. For example, for a given laminate layup and geom-
etry (corner angle), a concave molding set-up may be
converted to a convex one to reduce defect levels. The
results further indicated that this conversion is particularly
important when transitioning from an autoclave to an
OoA environment (e.g., vacuum bag). If concave tooling
is specifically required, the local mold geometry can be
modified to increase the corner radius and reduce resin
accumulation. Moreover, a pressure strip can be used to
further control local defects. Interestingly, the effects of
FIG. 11. Typical microscopic images of laminates manufactured with a increasing geometric complexity can be counteracted by
pressure strip. (a) laminate fillet at the end of the pressure strip, (b) void relatively simple modifications to the mold corner radius
distribution. [Color figure can be viewed in the online issue, which is
or by using a pressure strip, enabling the fabrication of
available at wileyonlinelibrary.com.]
more aggressive geometries. Finally, for a generic set of
prepreg, laminate, and molding conditions, these results
contents. The longer room temperature vacuum hold
provide approximate predictions for laminate quality.
allowed more time for the evacuation of entrapped air, and
The shift from autoclave to non-autoclave manufactur-
consequently decreased the void content.
ing methods necessarily eliminates the safeguards associ-
ated with elevated pressure. As a result, the successful
CONCLUSIONS
expansion of OoA manufacturing requires the develop-
We investigated several aspects of the OoA manufac- ment of optimized and robust processes, as well as more
ture of VBO prepreg laminates with corner geometries. careful attention to protocols. This study reveals some of
First, concave and convex corner parts were manufactured the fundamental phenomena associated with low-pressure
for various corner angles and laminate thickness levels processing of corner laminates, and provides a basis for
and from two prepreg materials. Polished sections developing effective manufacturing guidelines. The exist-
revealed that for both concave and convex parts, defects ing literature suggests that size and complexity scale-up
were concentrated in corner regions, and increasing geo- may also complicate other phenomena that govern qual-
metric complexity led to a reduction in quality. However, ity, including air evacuation. Furthermore, such scale-up
defects levels were always greater in concave laminates may also lead to nonideal temperature and pressure distri-
than in convex laminates. Microstructural data indicated butions within the vacuum-bagged part. As a result, the
that two different mechanisms caused the dimensional insights developed here must be combined with further
nonuniformities. For concave laminates, corner thickening research on factors affecting prepreg compaction, and
was primarily associated with resin accumulation because with the development of more accurate predictive simula-
of reduced compaction pressure and the inability of the tions for compaction, cure, and microstructural quality.
fiber bed to conform to the geometry of the tooling. Con-
versely, in convex laminates, corner thinning was caused ACKNOWLEDGMENTS
by uneven pressure distribution but counteracted by the The authors thank Garrett Peters, an undergraduate
stiffness of the fiber bed. The void content was primarily research assistant, for his sample analysis work and useful
affected by the prepreg age, with the aged material exhib- suggestions. Finally, the helpful contributions and sugges-
iting greater void content. However, void content was tions of the G8 project’s academic partners and industrial
also associated with the compaction quality, with greater advisory board are gratefully acknowledged. Cytec Indus-
laminate thicknesses leading to higher voids levels. tries generously donated the prepreg materials used in this
On the basis of these observations, three defect reduction study. Airtech International donated the consumables.
strategies were considered: intermediate debulking, variation in
the mold corner radius, and the use of a pressure strip intensifier. REFERENCES
Debulking was an effective means of reducing void content in
laminates, providing evidence that air entrapped between plies 1. F.C. Campbell, Manufacturing Technology for Aerospace
during layup was a key source of voids. Increasing the mold Structural Materials, Elsevier, London (2006).
corner radius improved quality by allowing the fiber bed to con- 2. S.F.M. de Almeida and Z.D.S.N. Neto, Compos. Struct., 28,
form to the tool and reducing the potential for resin accumula- 139 (1994).

DOI 10.1002/pc POLYMER COMPOSITES—2015 9

 3. K.J. Bowles and S. Frimpong, J. Compos. Mater., 26, 1487 16. Y. Li, M. Li, Y. Gu, and Z. Zhang, Appl. Compos. Mater.,
(1992). 16, 101 (2009).
4. T. Centea, L.K. Grunenfelder, and S.R. Nutt, Compos. Part 17. Y. Ma, T. Centea, G. Nilakantan, and S. R. Nutt, “Vacuum
A Appl. Sci., 70, 132 (2015). Bag Only Processing of Complex Shapes: Effect of Corner
5. J.P. Anderson and M.C. Altan, Polym. Compos., 36, 376 Angle, Material Properties and Process Conditions,” in 29th
(2015). Tech. Conf. Am. Soc. Compos. US-Japan Conf. Compos.
6. T. Centea and P. Hubert, Compos. Sci. Technol., 71, 593 Mater. D-30 Meeting, La Jolla, (2014).
(2011). 18. J.S. Lightfoot, M.R. Wisnom, and K. Potter, Compos. Part
7. L.K. Grunenfelder and S.R. Nutt, Compos. Sci. Technol., A Appl. Sci., 49, 139 (2013).
70, 2304 (2010). 19. M. Brillant and P. Hubert, “Out-of-Autoclave Processing of
8. L.K. Grunenfelder, T. Centea, P. Hubert, and S.R. Nutt, Complex Shape Laminates,” SAMPE 2010 Conference of
Compos. Part A Appl. Sci., 45, 119 (2013). the Society for the Advancement of Materials and Process
Engineering, Seattle, (2010).
9. T. Centea and P. Hubert, J. Compos. Mater, 48, 2033
20. A. Levy, J. Stadlin, and H. Pascal, “Corner Consolidation
(2014).
in Vacuum Bag Only Processing of Out-of-Autoclave Com-
10. M. Brillant and P. Hubert, “Out-of-Autoclave Manufacturing posite Prepregs Laminates,” SAMPE 2014 Technical Con-
of Complex Shaped Composite Laminate,” eThesis, McGill ference, Seattle, (2014).
University, Department of Mechanical Engineering, (2011).
21. J. Cauberghs and P. Hubert. “Effect of Tight Corners and Ply
11. P. Hubert and A. Poursartip, J. Compos. Mater., 35, 2 (2001). Terminations on Quality in Out-of-Autoclave Parts,” SAMPE
12. C. Albert and G. Fernlund, Compos. Sci. Technol., 62, 1895 2011 Conference of the Society for the Advancement of
(2002). Materials and Process Engineering, Long Beach, (2011).
13. G. Fernlund, N. Rahman, R. Courdji, M. Bresslauer, A. 22. L.K. Grunenfelder, C. Fisher, C. Cabble, S. Thomas, and
Poursartip, K. Willden, and K. Nelson, Compos. Part A S.R. Nutt, “Defect Control in Out-of-Autoclave Manufactur-
Appl. Sci., 33, 341 (2002). ing of Structural Elements,” SAMPE 2012 Technical Con-
14. P. Hubert and A. Poursartip, J. Reinf. Plast. Compos., 17, ference, Charleston, (2012).
286 (1998). 23. S. M. Hughes, and P. Hubert, “Out-of-Autoclave Prepreg Proc-
15. P. Hubert, R. Vaziri, and A. Poursartip, Int. J. Numer. essing: Effect of Integrated Geometric Effects on Part Quality,”
Meth. Eng., 44, 1 (1999). SAMPE 2013 Technical Conference, Wichita, (2013).

10 POLYMER COMPOSITES—2015 DOI 10.1002/pc