Force measurement on simulations

[andersonnardin]

Hello all,

I have a question about the forces on simulations. I will refer to a scene from the tutorial and it can clarify the general doubt. If we take as an example the cablegripper tutorial, the CableConstraint component force of the finger 3, for example, when we impose 3 displacements to the tendon, it is around 40. The default dt for this scene is 0.01.
If we change dt to 0.1, for the same displacement, the force is about 400. So, my questions is:

Forces obtained from fields called "force" (like those available in the mechanical object ) are, indeed, force times dt, is it correct?
If so, wouldn't be more accurate to call them impulse?
For the cited example, where the gravity is set to -981.0, what is the dimension of the measurement? N.s or N.e-2.s?

Thank you very much for your support.
Anderson

[EulalieCoevoet]

Hello,

Yes, indeed you are right. There's also a comment about that:

For your second question, I think the answer is here.

Eulalie

[andersonnardin]

Hi Eulalie,

Thank you for your reply.
I also asked about the field "force" on MechanicalObject nodes (or ConstantForceField), for example.
As you can see, such a comment is not shown:

To obtain the forces from there, should I divide it by dt too?

[andersonnardin]

The same question applies to the SurfacePressureConstraint

[hugtalbot]

Hey @andersonnardin which forces are you exactly interested in?

Note that in general the force data in the MechanicalObject contains all the terme included in the b vector of the linear system Ax=b. It therefore depends on your integration scheme.

[andersonnardin]

Hi @hugtalbot
I'll tell you what I expected to read in these fields.
CableConstraint force: the force applied to the cable
SurfacePressureConstraint force: the force applied to the inner walls of the chamber
MechanicalObject force: for each node, the force applied to it

Is this correct? I am asking because I don't see the same behavior for the 3 applications I mentioned. For example, the MechanicalObject force doesn't seem to depend on dt.

What is correct/wrong with my observations?

[hugtalbot]

Hey @andersonnardin

For CableConstraint and SurfacePressureConstraint, you can refer to this post: sofa-framework/sofa#3812

For the forces applied on each node it would be a bit more complicated, a first way would be to consider the force vector of the MechanicalObject but in fact I think this is not exactly what you are looking for. You would be interested in recovering all contributions from the ForceFields on your object using their function addForce() (explicit force contribution)

[andersonnardin]

Hi @hugtalbot

Because it has been a while since my question was not answered, I decided to reformulate it.

1. The code you provided in your post is valid for the case where the number of nodes in the collision mesh impacted by the constraint is 1, right? Otherwise, the constraints will not be composed of 6 elements on each line as you presumed (for example in index = index + 6)

2. The class you created provides the forces (later, it is used to draw the arrows on the scene). However, it seems to me that the total force was already present in the lambda field of the MechanicalOject of the Cube. As follows:
   
   Is that correct?