This document discusses research into reducing warping in phenolic resin/fiberglass composites by engineering voids in the material. Volatile analysis identified residual solvents from the initial formulation as potential sources of voids. Higher relative mass loss in prepreg samples correlated to lower warp, indicating solvent content affects void formation and warping. Adding non-solvents of the base polymer was found to induce void creation through phase separation during curing. Producing prepreg with various void-inducing additives and characterizing the resulting warp could help determine formulations that satisfy requirements for void content and mechanical properties while minimizing distortion.
1. Future Work/ReferencesAcknowledgements
Potential Void Mechanisms
Volatile Analysis
Void Content and Solvent Dynamics in Composite Structures
Andrew Hollcraft, Ryan Hackler, Tyler Kirkness, and David Rider, Department of Chemistry, Western Washington
University, Bellingham, WA
Introduction
Conclusions
Figure 1: Plot correlating average warp value
obtained from composite parts and average
void content found via SEM
Figure 5: IR spectra of A-stage resin (left) and B-stage prepreg (right) collected
every 48 minutes throughout a production cycle
Figure 6: IR spectra of phenol and
ethanol C-O stretches in B-stage prepreg
Figure 9: Thermograms of B stage-to-C stage (top)
and C stage-to-Char (bottom) transitions with
corresponding warp values
Figure 11: SEM micrograph of B-
stage prepreg and corresponding
void analysis image (via imageJ)
after binary filter and void outline
Void Investigation
Background
Sandwich-structured composites are widely used in industry due to being lightweight, impact
resistant, and inexpensive. They are composed of a honeycomb core sandwiched between two
sheets of prepreg1.
Resin
Modification
Impregnation
of Fibers
Metering of
Fibers
Curing
@ 124°C
Characterization
Scheme 1: Method used to produce micro-scale prepreg
to analyze the effects of various additives on void
content and volatiles during curing
Solvent content decreases throughout the production
cycle in both A-stage resin and corresponding B-stage
prepreg
Relative amounts of solvent changes throughout the
production cycle, with little change in water content
and a significant change in ethanol
The ratio of IR peaks attributable to phenol and
ethanol correspond to a higher warp when a larger
amount of ethanol is present
Three proposed void induction mechanisms include:
Boiling Point3
Phase Separation
Surface Tension4
Higher relative mass loss in low warp B-stage prepreg indicates
solvent content affects void content
Jim Del Pinto
Kevin Bussard
Sean Mitchell
Kalin Karich
Nicole Hoekstra
Nicole Larson
Cecile Grubb
Charles Wandler
Erin Macri
Figure 8: Correlation plots for void content with surface tension (left) and
boiling point (right) of additives used in prepreg
Void content plays a vital role in the resulting geometric distortion
Volatile analysis identified residual solvent concentration from the initial formulation as
potential sources of voids
Distribution of residual solvents was correlated to macroscopic distortion
Suggestion
Non-solvents of the base polymer with high surface tension phase separate and nucleate
void creation
Desirable in order to facilitate void production and reduce geometric
distortion
Produce Medium-scale C-stage prepreg samples with various void inducing additives and
characterize resulting warp
Pilot scale prepreg development
Determine best void inducing additive(s) and loading
such that void content satisfies warp and
mechanical requirements
Novel resin systems synthesized
at Western Washington University
1. Strek, T.; Jopek, H.; Maruszewski, B.T.; Nienartowicz, M. Phys. Status Solidi B 2013, 1-13.
2. Hackler, Ryan A., Andrew T. Hollcraft, Tyler A. Kirkness, Nicole S. Larson, Nicole K. Hoekstra, and
David A. Rider. "Relief of Cure Stress in Prepreg Composites with Engineered Voids: A Solution to
Warping in Composite Phenolic Resin/Fiberglass Laminates." Industrial and Engineering Chemistry
Research; In Press (March 6, 2016).
3. Naganuma, T.; Naito, K.; Kyono, J.; Kagawa, Y. Comp. Sci. Tech. 2009 69, 2428-33.
4. Liu, P.; Song, J.; He, L.; Liang, X.; Ding, H. J. Appl. Polym. Sci. 2009, 114, 811-7.
Figure 7: TGA-IR spectra of evolved
gases during B to C stage curing Figure 12: Void distribution in
resulting composites
Scheme 2: Representative phenolic prepreg polymerization
Table 1: DSC on B stage formulated prepreg with tensile testing performed on C stage laminate
SEM analysis revealed a relationship between
void content of the surface and average warp of
the composite, which may be due to the voids
acting as stress relievers during curing that
allows for localized deformation as opposed to
geometric distortion (known as warping)2.
Kevlar honeycomb
Phenolic fiberglass composite
Decorative face sheet
Additive = water
Additive = ethanol
Additive = none
Figure 13: Qualitative warpage observed
in bench-top produced prepreg
Figure 2: Cartoon example of warpage observed
in 2” by 4” industrially produced composites
Figure 3: Representative industrial production of prepreg
Figure 4: Top – Sequence of
composite design, middle and
bottom – aerospace applications of
these composites
Figure 10: Cartoon example of void creation; left – B stage prepreg with phase
separated non solvent, right – resulting C stage prepreg with voids