PREDICTION OF HYSTERESIS OF A THERMOPLASTIC POLYURETHANE USING COARSE-GRAINED MOLECULAR DYNAMICS

We predict hysteresis of a thermoplastic polyurethane (TPU) varying the configurations and weight % of hard segments from 34.90% to 62.30% using coarse-grained molecular dynamics (CGMD) simulations. Rate-dependent stress-strain responses of the molecular models are justified between energy equivalence constitutive modeling and atomic viral stresses. Uniaxial cyclic loading (tension/compression) of the coarse-grained (CG) models are performed using NPT ensembles (isothermal-isobaric) at the atmospheric condition to ensure no stresses in the other two directions except the loading directions. Engineering stresses are estimated from atomic viral stresses at different frequencies and up to various strain levels, whereas areas under the stress-strain curves give the hysteresis loss under cyclic deformations. We correlate the hysteresis losses of all of the models with their bulk moduli and Poisson's ratios. By the end of the study, we may answer the following research questions: i) How much hysteresis loss increases due to increasing the weight% of hard segments from 34.90% to 62.30%? ii) How sensitive are the losses corresponding to strain amplitudes from 5% to 15% and frequencies from 1.67 x 10(11) Hz to 5.0 x 10(11) Hz? iii) In order to reduce the hysteresis loss, how much we have to compromise in bulk modulus and how much Poisson's ratio will be increased corresponding to that compensation. This molecular simulation tool can be used to design new rubber materials with better mechanical properties and lower hysteresis losses without the trial and error based experimental work.