In by admin

Name Mr. Dylan Anstine
Organization or Institution University of Florida
Presentation Type Poster
Topic Computational Chemistry

Computational mechanical testing of amorphous polymers


Dylan Anstine and Coray M. Colina

Author Institution(s)

1 Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611
2 Department of Chemistry, University of Florida, Gainesville, FL 32611


Amorphous polymers are materials that are found extensively in many industrial applications, including their use for structural and material coatings. However, for coatings applications usually thin films are required, making their characterization by experimental and computational means challenging. For atomistic simulations a lack of significant periodic order presents difficulties in generating the initial structure for computational testing and analysis. In this work, three different in silico polymerization methodologies were utilized to generate an array of different initial amorphous polymer structures. The polymer samples studied were poly(methyl methacrylate), poly(propylene), and polymer of intrinsic microporosity-1.The modeling analysis was further expanded through the application of two different common polymer force fields: Dreiding and GAFF. Large polymeric samples of ~50,000 atoms were required for this work in order to increase the accuracy of computationally determined elastic moduli. After validation of the thermodynamic and structural properties of the samples, non-equilibrium molecular dynamics simulations were employed to perform simulated mechanical testing. Stress-strain plots were generated from the result of tensile tests that were performed through the application of simulated uniaxial strain. The results of this work demonstrate the complexity involved in simulating amorphous polymeric materials, the advantages of different in silico polymerization approaches, and the ability of non-equilibrium molecular dynamics simulations to yield data for mechanical testing if sufficiently large samples are evaluated.