Combining Molecular Dynamics and Machine Learning to Analyze Shear Thinning for Alkane and Globular Lubricants in the Low Shear Regime

• DOI: 10.1021/acsami.2c16366
• Publication Date: 2023-01-30T12:39:15Z
• Authors: Ikki Yasuda; Yusei Kobayashi; Katsuhiro Endo; Yoshihiro Hayakawa; Kazuhiko Fujiwara; Kuniaki Yajima; Noriyoshi Arai; Kenji Yasuoka
• Keywords: Shear thinning

Lubricants with desirable frictional properties are important in achieving an energy-saving society. at the interfaces of mechanical components confined under high shear rates and pressures behave quite differently from bulk material. Computational approaches such as nonequilibrium molecular dynamics (NEMD) simulations have been performed to probe behavior lubricants. However, low-shear-velocity regions materials rarely simulated owing expensive calculations necessary do so, velocities comparable that experiments not clearly understood. In this study, we NEMD extremely lubricants, i.e., two layers for four types lubricants mica walls, 0.001 1 m/s. While confirmed thinning, velocity profiles could show flow when was much slower than thermal fluctuations. Therefore, used unsupervised machine learning approach detect movements contribute thinning. First, extracted simple features large amounts MD data, which were found correlate effective viscosity. Subsequently, interpreted by examining trajectories contributing these features. The magnitude diffusion corresponded viscosity, location slips varied depending on spherical chain irrelevant. Finally, attempted apply a modified Stokes-Einstein relation equilibrium systems. systems low obeyed sufficiently, deviations observed rates.