Tune solref and solimp in equality constraint

[SteMuc91]

Intro

Hi!

I am a researcher, and I'm using MuJoCo to simulate a 5 bar linkage.

My question

MuJoCo does not solve correctly the closure of the mechanism. The angle J4 does not reache 0.105 rad, but it stops at 0.385 rad. If I disable the spring, it works correctly. I think that I don't tune correctly the solref and solimp parameters. Is there a guide to tune them? Thank you very much!

Confirmations

• I searched the latest documentation thoroughly before posting.
• I searched previous Issues and Discussions, I am certain this has not been raised before.

[ayers]

Hi,
It does sound like you’ve got things fighting against each other, but without seeing the MJCF I can’t help narrow it down. Could you paste in a MJCF of the 5-bar linkage you’re using? If it’s part of a larger project, you can just strip it down to just the 5-bar part. Just for context, is the 0.105 rad number a target you’re trying to reach, a limit, or a closed-pose value that you calculated? It sounds like you meant the latter but I wasn’t sure.

It’s also helpful to know what version of Mujoco you’re using.

Thanks!