pin.constraintDynamics and data.lambda_c

[MR-KNV]

Hello!
I would like to thank you and your team for the Pinocchio library. I used Pinocchio 2.99.0 from this conda package: package
I have a model with closed kinematic loop and want to calculate dynamics of it.
I checked in Meshcat viewer that inverseConstraintKinematic works well but Im not sure that vector of data.lambda_c is calculated correctly. I know exactly, that vector of data.lambda_c should be about 3-5 N by X axis, but I got an array with values that are close to zero.

I use this command for calculations:

pin.initConstraintDynamics(model, data, constraint_model)
pin.rnea(model, data, q, vq, ddq) % to calculate data.tau
ddq= pin.constraintDynamics(model, data, q, vq, data.tau, constraint_model, constraint_data, prox_settings) % update ddq

Could you please tell me if my calculations are correct? Is there anywhere I can read the documentation for the pin.constraint Dynamics ()method?

Thank you!

[jorisv]

Hello @MR-KNV,

Can you give us a script to reproduce your issue ?

The method is described in this paper Justin Carpentier 2021

[MR-KNV]

Yes, my script is here (https://github.com/MR-KNV/pinocchio_test)