Center for Composite Materials - University of Delaware
Visit UD's Coronavirus website for updated information, FAQs and guidance. More information.

Research Summary

Gas Transportation in Void Reduction of Thermoplastic Composites via Oven Vacuum Bag (OVB) Processing

Authors: D.Zhang (PhDMSEG), J.W. Gillespie Jr

INTRODUCTION AND MOTIVATION

High performance thermoplastic composites – Increasingly used as structural materials
• Enhanced impact resistance
• Recyclability
• New forming, assembly and joining methods

OVB processing for thermoplastic composites – A potential cost effective alternative of autoclave
• Vacuum bag based, no positive pressure applied
• Low cost

Void reduction – Major concern to provide equivalent properties of autoclave parts
• Severely degrade the mechanical properties
• Vv<1% required for aerospace applications

• Void dynamics and consolidation mechanisms are not fully understood!

Rod-like voids sealed in the prepreg – gas escaping through diffusion – Diffusion time for thick laminates will be extremely long with diffusion mechanism only

Gas escaping through interlayer region may provide another way for volatile removal.

OBJECTIVES

• Investigate the permeability of the interlayer gaps
• Investigate the characteristic volatile removal time for gas flow through interlayer region and diffusion mechanism.
• Investigate how the time changes as the intimate contact develops

IN-PLANE AIR PERMEABILITY OF TP PREPREG STACKS - EXPERIMENTAL SET UP

In-plane Air Permeability of TP Prepreg Stacks – Experimental Set-up
• 1D Darcy’s Flow Set-up
• INSTRON applies uniformly 1atm pressure to maintain even contact of layers
• Reliable and repeatable results
-Number of layers
-Sample size (W, L)
• Permeability with various stacking angles

In-plane Air Permeability of TP Prepreg Stacks
• 3 or 4 layers
• 2 inch ~4 inch along flow direction
• 8 inch ~12 inch in width direction

PERMEABILITY PARAMETER WITH VARIOUS STACKING ANGLES

• Permeability along the fiber direction is about 10 X of that along the transverse direction

AIR PERMEABILITY AT A CERTAIN LEVEL OF INTIMATE CONTACT

• Modeling method – FEA
-Polyflow for intimate contact developing
-Fluent for Air flow through channels
--Laminar flow
--Pressure drop: 1atm
--Isothermal
--Sample length: 500mm
• Surface geometries of Intermediate levels of intimate contact → permeability prediction

Initial permeability
Mass flow rate: 5.55E-08 Kg/s
Flow rate: 4.54E-08 m3/s
Pe: 4.15E-17 m3

Partial intimate contact
Sample width:621.08um
Contact length: 487.71
Dic = 78.5%
Mass flow rate: 2.79E-08 Kg/s
Flow rate: 2.32E-08 m3/s
Pe: 2.08E-17 m3

CHARACTERISTIC TIME

Air Flow through permeable interface
• Initial void gas pressure is Patm
• Steady state pressure dropping: at each time increment, Pv = Papply +surface tension(ignored)
• Total volume of gas increases at T>298 K

Pure gas diffusion
• Gas diffuse from the center of the thickness
• Symmetric diffusing to both tape surfaces
• Fick’s second law
• To evacuate 99.9% of the void gas

CHARACTERISTIC TIME - DIFFERENT LEVEL OF PERMEABILITY

Panel In-plane dimension L = 1m
• As permeability decreasing, air escaping time through interface getting closer to diffusion time
• Both mechanisms may dominate for air escaping in thin laminates

COMPARISON OF CHARACTERISTIC TIME

• Panel In-plane dimension L = 1m
• Isothermal condition, Pe = 1E-15 m3
• Gas diffusing through 1 layer and flowing out through interlayer region before full intimate contact could reduce significant time for void reduction

SUMMARY AND ONGOING WORK

In-plane Air permeability of prepreg stacks were measured
• Fiber angles affect the permeability
• Along the fiber direction ~1E-15 m3 (1E-9m2 assuming interface region h =100mm)
• Across fiber direction ~ 1 E-16 m3
• Measurement with conditioned surfaces
• Analytic model to predict permeability of off-axis fiber angles

FEA can help with intimate contact modeling with real tape surface, and also predict the interface permeability at different level of intimate contact
• Characteristic tape surface is needed
• Contact of two tape surface could be modeled

Air escaping time mainly dominated by diffusion; could be effectively reduced by diffusing through 1 layer and flow out through permeable interlayer region – controlling of intimate contact
• Transient condition for estimation of time escaping

ACKNOWLEDGEMENTS

Faculty Advisor:
• Prof. J. W. Gillespie, Jr.

Committee Members
• Prof. Suresh. G. Advani
• Prof. Dirk. Heider
• Prof. Michael Mackay

CCM Researchers and Technician
• Pavel Simacek
• John Tierney
• John Thiravong
• Funding
• EADS

302-831-8149 • info-ccm@udel.edu