Quantifying the Impact of Porosity on Mode I Fracture Behavior in Advanced Composite Materials

Abstract

Advanced composite material applications are rapidly proliferating within the aerospace industry, but challenges remain to their more widespread adoption due to the influence of process-induced defects. The complex manufacturing efforts to combine multiple discrete materials inevitably result in porosity being a pervasive issue for composite materials. These process-induced defects impair material reliability and limit the application potential of advanced composite materials.

While the impacts of porosity on composite materials' elastic properties are widely researched numerically and analytically, gaps remain in quantifying porosity's effect on advanced composites' crack resistance. Another challenge within this study is that porosity is stochastic, which means that the material characterization tests for two specimens with the same porosity level will produce two different results.

This work will focus on the mode I delamination behavior of porous composites. A 'pristine' specimen and a purposefully engineered defected specimen validate the ability to accurately model a well-defined porous system within an established modeling procedure, the cohesive zone model. This process expands to evaluate stochastic porosity through random pore generation and random sampling. Despite the collapse of the self-similar crack assumption, this method proposes a way to calculate an effective mode I strain energy release rate within these frameworks. Further, this study investigates the impact of local pores on the local effective mode I strain energy release rate. Additionally, this study opens the door to leveraging a dataset from random sampling to train a machine learning (ML) algorithm to predict material behavior based on a given porous microstructure and quantify the impact of various pore parameters on the system.