CAMX 2024
- September 9th-12th, 2024
- San Diego, California
Schrödinger is excited to be participating in the CAMX 2024 taking place on September 9th – 12th in San Diego, California. Join us for presentations by Andrea Browning, Director at Schrödinger. Stop by booth G33 to speak with Schrödinger Scientists.
A multi-scale framework to determine effects of moisture and temperature on composite parts using molecular dynamics and finite-element methods
Speaker:
Andrea Browning
Abstract:
Computational approaches help to improve the efficiency of materials product design by understanding the viability of a candidate material without requiring experimentation or testing. Herein, we describe a multiscale simulation framework for understanding how a material’s properties change as a result of its environment using molecular-level models, and then utilizing those properties to determine how composite materials behave using finite-element models. We applied this framework to assess the effects of moisture and temperature and on a model system based on Hexcel Hexply 8552. Using molecular dynamics (MD), along with specialized tools to model crosslinking and moisture uptake, atomistic systems were created and simulated at varying temperatures (hot, room and cold) and varying moisture content at high temperature (dry and wet). At each condition, elastic moduli and coefficients of thermal expansion were determined using dedicated workflows. These condition-dependent moduli and CTEs were then incorporated into representative volume element (RVE) models for a unidirectional (UD) prepreg fabric. An uncertainty analysis was performed to account for microstructural variations in carbon fiber content and alignment. The uncertainty analysis revealed that we were able to predict orthotropic materials properties with an accuracy similar to systematic experimental characterization for Hexcel Hexply 8552 and they lie within the range of variation obtained at room temperature. In addition, a high sensitivity of the transverse and shear moduli to the matrix uncertainty was observed, as would be expected for lamina properties. Effective elastic properties were further upscaled through composite layup modeling into a sub-component level model of a racecar nose. The multiscale solution methodology revealed changes in the structural response of the part that are expected to occur at varying temperature and humidity conditions.
Panel: AI-Driven Innovations: Shaping the Future of Materials Development
Speaker:
Andrea Browning
Live Demo
Speaker:
Andrea Browning