[warlic5]

I am working with an inner steel cylinder, outer steel cylinder and grout (concrete) inbetween. Grouted joint connection in wind turbines. The inner steel part is fixed, while the outer steel part and concrete has not been given any boundary condition. I have applied uniform pressure to activate coulomb friction, and the step is dynamic implicit with instantaneous load application. Gravity acts in the vertical direction. I dont understand why I get that jiggly acceleration in the second attached picture. Might it have something to do with the penalty method which requires an initial displacement of the springs before entering slipping behavior. Because in the classic coulomb friction model we instantaneously get the correct acceleration "a", while in abaqus it wiggles up and down and then converges at "a". I am trying to understand why it does that.