# DEM restitution coefficient

*v4.2*guide, "the restitution coefficient ije is taken as the average of the two materials coefficients, in what is the only calibration parameter of the model." So what is the function of friction coefficient? I tried to keep restitution coefficient as constant and change friction coefficient, but the movement of the object are almost the same. Does it mean the value of friction coefficient will never influence the result?

## Comments

Hey,

the same idea is applied to the friction coefficient, the tangential force at an interface is computed using the average of the two material coefficients.

What you are describing is one of the main reasons we decided to alternatively explore multibody models such as Chrono. Dealing with most frictional contacts on a DCDEM model is very complicated. If you imagine two planes sliding along one another, and if those planes are discretized as spheres, you can see that by simple geometrical effect what you'll end up having are mostly normal interactions, not tangential (frictional). Such effects can be seen in many types of interactions.

For high temperature lubricated granular dynamics (the focus of the PhD that implemented the DCDEM model) the model actually works quite well as the contacts are short-term and tendentially non sliding.

There are combinations of resolution, restitution and friction coefficients that should give you good results and will be sensitive to the frictional component, unfortunately it is a matter of calibration by trial and error for your specific case.

Thanks for your reply and I appreciate your help! Your words really made it clear for me to understand why DEM deal with the contact force in this way.

I will try to use the Chrono to carry out my simulation.