Through-thickness dielectric sensing for thermoset composite cure analysis and process validation

Through-thickness dielectric sensing for thermoset composite cure analysis and process validation

Thermoset composites are an attractive material choice for high performance applications due to their superior properties and high level of tailorability compared with traditional homogeneous materials. However, implementation of these materials is challenging due to their high level of uncertainty within the manufacturing process. Sources of variation include slight differences between material batches and the natural temperature variation in manufacturing equipment causing a range of expected final part properties. Much research has been conducted on quantifying the expected variation and methods of monitoring the cure reaction to explicitly identify how the part cure progresses. The research presented in this thesis focuses on the use of dielectric sensors as a method for live-monitoring of the thermoset cure reaction, as a strategy for capturing and validating material cure state information. Dielectric sensors are an appealing option for in-situ monitoring of the thermoset cure reaction because they provide qualitative and quantitative information on the cure reaction progression. However, to date, there is no clear consensus on which dielectric parameter should be used for performing these analyses, or how they should be executed. In this research, a systematic approach to material cure state determination was completed and a new, comprehensive set of parameter-independent dielectric analysis techniques are presented. The results are compared with numerical simulations and analytical testing, demonstrating high accuracy and part-to-part repeatability. The newly proposed methods are comparable or better than existing techniques and allow for more analysis flexibility. The methods were further validated during a study on a novel dielectric sensor which is designed to monitor through the thickness of parts up to 20 mm. While the design of the prototype sensor influenced the signal reading, a correction factor was determined which allowed for successful implementation of the newly proposed dielectric analysis methods. The methods showed high accuracy and part-to-part repeatability for composite laminates of thicknesses between 2 and 20 mm. The culmination of this thesis is an exploration of the cure kinetics modelling variability that is expected for this material system. A stochastic approach with Monte Carlo methods was used to characterise the influence of cure kinetics modelling and oven temperature uncertainty on the polymer viscosity and cure reaction. A novel approach for quantifying cure kinetics uncertainty is provided and the results of the subsequent convergence analyses were validated with experimental trials. Results indicate that the time to fully cured has high amount of variability (upwards of 10%), and suggestions for ensuring process robustness are provided.

File reproduced in accordance with the copyright policy of the publisher/author/creator.