Designing anisotropic friction through limit curve analysis

Abstract

Friction is an essential component of robotic manipulation which is highly dependent on contact surfaces. In practical applications, these surfaces are often anisotropic, a property that has been known to produce interesting movements in nature and uncertainty in human applications. Therefore, control of anisotropic frictional surfaces could result in more precise movement in manipulation, locomotion, and other facets touched by frictional contact. To arrive at such controllability, frictional force was collected across a spectrum of anisotropic micro-textures, and a limit curve was generated. Experimental data was analyzed in accordance to friction laws such as limit curve and maximum-inequality principle (MPI). Qualitative observation and residual sum of squares (RSS) was used to detect lack of normality and non-convexity within each limit curve. This lack of both normality and convexity contradicts MPI and suggests that an alternative model is necessary. Additionally, the anisotropic frictional behaviors observed advances the feasibility of "designing" micro-textures capable of controllable anisotropic friction.