Accession Number : ADA306779
Title : Control of Polyimide Condensation Composite Processing.
Descriptive Note : Final rept. 1 Feb 89-30 Sep 93,
Corporate Author : WASHINGTON UNIV ST LOUIS MO
Personal Author(s) : Kardos, John L. ; Dudukovic, Milorad P. ; Joseph, Babu ; Vasat, Jiri L. ; Kim, D. H.
PDF Url : ADA306779
Report Date : MAR 1994
Pagination or Media Count : 122
Abstract : AFR700B is a high temperature polyimide resin having the advantages of a high glass transition temperature (750 deg F) and superior thermo-oxidative stability, but the disadvantage of a volatile management problem. This study explored a new process simulation model, new sensing technology, and new concepts in process control in order to advance the state of processing science and to make the processing of AFR700B composite parts cost effective. The new devolatilization model was experimentally verified by comparing its predictions for pressure and temperature distribution through the laminate thickness with experimentally measured pressure and temperature profiles. The verified model was used to successfully predict the concentration and rate of removal of volatiles during composite cure for several heating rates and vacuum bag pressures. Comparison of measured dielectric ionic viscosity and devolatilization model prediction for volatile mass fluxes during processing shows that microdielectrometry can indicate (a) the start of polycondensation reaction, (b) the temperature-time window in which maximum devolatilization occurs and voids develop between laminate plies, (c) the end of the polycondensation reaction, and (d) the effects of water absorbed during freezer storage on the polyimidization reaction and devolatilization process. Artificial neural network models proved superior to classical graduatic regression approaches in capturing the non-linear relationships that exist between product quality and processing variables. Intermediate secondary measurements, taken while the cure cycle is in progress, can be used to detect unmeasured disturbances; hence, providing some feedback correction to the on-line model used.
Descriptors : *PRODUCTION CONTROL, *CONDENSATION, *POLYIMIDE RESINS, SIMULATION, MEASUREMENT, THICKNESS, STABILITY, NEURAL NETS, DETECTION, MANAGEMENT, PREDICTIONS, MODELS, DISTRIBUTION, WATER, POLYMERS, PROCESSING, FLUX(RATE), HIGH TEMPERATURE, LAMINATES, SECONDARY, VACUUM, THERMAL STABILITY, GLASS, REGRESSION ANALYSIS, MASS FLOW, TRANSITION TEMPERATURE, NONLINEAR SYSTEMS, PROFILES, FREEZING, OXIDATION, FEEDBACK, HEATING, PRESSURE, VOLATILITY, STORAGE, CURING, ONLINE SYSTEMS.
Subject Categories : Plastics
Distribution Statement : APPROVED FOR PUBLIC RELEASE