UC San Diego

Dynamic Mechanical Analysis (DMA) is an important experimental characterization technique for the mechanical properties of polymers. In this paper, we translate such a technique into a Molecular Dynamics (MD) simulation workflow capable of outputting properties from DMA within a much wider frequency and temperature range than what can be experimentally observed. We also introduce metrics for convergence of simulation runtime and DMA straining size that are not discussed in comparable computational literature but that drastically affect the accuracy of results. We validate the efficacy of the work by comparing our obtained storage and loss modulus data on Polyvinylidene Fluoride (PVDF) with previous computational literature and indicating performance and accuracy improvements for such study based on our convergence metrics. Finally, we perform a complete workflow run at GHz-range frequencies on PVDF utilizing the All-Atom Optimized Potential for Liquid Simulations (OPLS-AA). With this, we hope to validate and justify this forcefield as a good standard for high-throughput analyses on other polymers with the workflow and to compare our data with experimental results employing the Time-Temperature Superposition (TTS) method.