Inverse Design and Mechanics of Hybrid Filler Composites with Solid and Liquid Inclusions

University Of Washington, Seattle WA

Investigators

Abstract

Polymer composites with hybrid fillers are produced by a complex synthesis process involving two or more inclusions with different morpholoies and physical properties that often leads to laborious, time-consuming, low-yield, and expensive efforts. Furthermore, the uncertainty regarding their mechanical load capacity and susceptibility to premature failure hinders widespread adoption. This award supports fundamental research to inversely design composite materials with solid and liquid phase fillers while investigating their failure under large deformations. By building a comprehensive and interpretable machine learning model, this project will enable efficient and reliable synthesis of composites with targeted properties. The new knowledge will promote the utilization of soft multifunctional composites in emerging applications, such as self-powered wearable electronics, biomonitoring systems, and soft robotics. This award will also support the development of a diverse workforce through outreach programs and creation of free educational online content on topics of advanced mechanics and artificial intelligence. The objective of this research is to perform inverse design of “hybrid filler composites” with solid-liquid fillers and investigate their failure under large deformations. Embedded liquid-phase fillers add complexity to the mechanics of multiphase composites while offering unique advantages such as enhanced toughness and conductivity. To enable the rational design of these multifunctional materials, data-driven models will be formulated, incorporating a wide range of composite descriptors including filler composition, shape, volume fraction, solid to liquid filler ratio, and polymer matrix. In the modeling framework, data-driven embeddings will be utilized to reduce the design search space and expedite the discovery of optimal parameters. Predicting the failure of these heterogeneous materials is extremely challenging due to the presence of dissimilar filler phases and their complex microstructures. Therefore, sparse feature selection will be employed to identify the most dominant factors contributing to failure. The new insights will be applied to synthesize composites with designed properties, and the resulting data will be used for model validation and iterative feedback. This research will create a comprehensive data library for soft multifunctional materials, revealing the relationship between their descriptors and mechanical failure while establishing a universal framework for the efficient design of multiphase materials with engineered properties and failure characteristics. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.