The quadruped robot walks unstable during physical simulation of gait planning, whether something wrong with the Sphere to Plane Force module？

Hi everyone,

When I 'm using Simscape Multibody Contact Forces Library for my simulation，I found that my quadruped robot walks unstable.

After a few test, I found that nothing wrong with my calculation results. The joint angle calculated by inverse kinematics is OK and the robot was relatively stable at the beginning, then, gradually the robot tend to fall...

I think maybe something wrong with the Sphere to Plane Force module, maybe there is an error when the sphere and the plane collide？

[zexian Lee]

I want to ask how this module judge the contact between the sphere and the plane？