608559 General-Purpose Coarse-Grained Toughened Thermoset Model for Aerospace Composites: Workflows, Parameterization, and Validation

Wednesday, November 18, 2020
Computational Molecular Science and Engineering Forum (21) (Poster Gallery)
Michael Henry1, Stephen Thomas1, Jared Bates2, Brittan Farmer3, Carla Reynolds3, Alexis Lanier2, Jeffrey Wiggins2 and Eric Jankowski1, (1)Micron School of Materials Science and Engineering, Boise State University, Boise, ID, (2)School of Polymer Science and Engineering, University of Southern Mississippi, Hattiesburg, MS, (3)The Boeing Company, St. Louis, MO

The efficient manufacturing of composite aerospace materials can be improved with insights from molecular simulations when sufficiently large volumes and timescales can be accessed. Predicting the structure of aerospace composites is especially challenging because of the high computational cost of relaxation (scales cubically with polymer chain length), morphological features spanning dozens to millions of repeat units, and the nonequilibrium dynamics of reacting epoxy thermosets. Coarse-grained simulations can overcome these barriers by using fewer simulation elements to represent the same number of atoms and simplified reaction mechanics which both accelerate sampling dynamics. In this work we deploy coarse-grained simulations of reacting epoxy thermosets and validate them against two common chemistries with the longer term goal of improving the fuel efficiency of airplanes of flight and aerospace manufacturing.

We use a single site coarse-grain model to represent each epoxy, amine, and toughner repeat unit. We implement a dynamic bonding plugin with HOOMD-Blue to simulate the cross-linking of the epoxy and amine species. We present workflows for parameterizing models which include deriving Lennard-Jones parameters from cohesive energy data and empirically deriving the kinetic parameters of our model from experimental curing data.

We validate our model using data from curing experiments of the amine monomer 44DDS with both DGEBA and TGDDM epoxy in stoichiometric ratios, both toughened with PES. We compare predicted morphologies against experimental scattering and microscopy data, as well as glass transition data. We find our model correctly predicts glass transition temperatures as a function of degree of cure, the phase behavior, and toughener domain sizes in 73 nm x 73 nm x 73nm volumes. We also demonstrate our model’s sensitivity to cure profile which enables studies of how processing conditions affect cure time, mechanical, and thermodynamic properties. This work demonstrates tooling to better understand the relationship between structure, processing, and properties in toughened epoxy thermosets for improving the efficiency of composite aerospace part manufacturing.


Extended Abstract: File Not Uploaded