Probing Shear Thinning of Liquids at High Strain Rates using Simulations and Machine Learning

The extraction of accurate rheological properties using nonequilibrium molecular dynamics simulations has improved our understanding of liquid flows under high shear strain rate conditions. However, the high dimensionality of the simulation output data makes a deeper interrogation of the link between molecular-scale features and rheological properties challenging. We demonstrate that dimension reduction methods can expedite the analysis of data generated by nonequilibrium molecular dynamics simulations of small-molecular liquids such as squalane. Dimension reduction of the orientation tensors of all atom pairs associated with squalane molecules enabled a clear visualization of the transition from Newtonian to non-Newtonian shear thinning with increasing shear rate. The end-to-end atom pairs dominate the largest variations in tensor components at low pressures, and provide the clearest separation of the orientation tensors with rate. However, the side atom pairs dominate the largest variations in tensor components at high pressures (>400 MPa). Dimension reduction shows that the decrease in viscosity with rate for low pressures is strongly correlated with changes in molecular alignment. However, for high pressures, shear thinning occurs at saturated orientational order.