Cancel
Showing results for
Did you mean:
Highlighted

# LMS Imagine.Lab Amesim - How are contact forces calculated

Experimenter

Hi everyone,

I'm having issues understanding how contact forces are calculated in the component"PLMCONT0 - contact between 2 contours".

In 3D,we usually calculate a force using F=P*S where S is the contact surface between two bodies.

I don't get how AMESim can compute a contact force in 2D. We define two shapes in the x y plan but there is no parameter refering to the depth of parts.

For exemple, if two cylinders are in contact, the resulting contact force will depend on the length of cylinders (the contact force will be greater if the length of the cylinder is 1cm and smaller if the lenth is 1m).

I can't find a way to set the depth of my parts. Or maybe the force is independant of the depth according to equations used in the model? I feel like i have something missing in my model.

* Contact force module F0: F0 = E1.a1 + E2.a2

* Linear penetration p : p = 2.(E1.a1 +E2.a2)/( (E1+E2).L )

Areas (a1 & a2) used in equations are the "intersection area between the shape and contact segmen"

Any ideas on how to take the depth of parts into account?

4 REPLIES 4

# Re: LMS Imagine.Lab Amesim - How are contact forces calculated

Experimenter

Hi DelphinO

You well understood the 2D contact. This component don't take into account the depth of parts.
The 2D contact is a good way to find reaction force but not to predict the linear penetration length because this last one depends on depth of parts as you mention.
If you want to take into account the depth you can increase Young modulus and the maximum stiffness by twice to reduce the linear penetration length by arround twice.

# Re: LMS Imagine.Lab Amesim - How are contact forces calculated

Experimenter

Hi Pschyt,

Correct me if I'm wrong, the unit of the Young modulus is N/m in AMESim. Does it mean that the contact force is calculated as if parts were 1m depth?

If it's the case, if my part is 0,5m depth, I should divide the Young modulus and the solid stiffness by 2, right?

(In 3D: As the depth is reduced, the contact force will be greater so the deformation. To take this into account in my model, the only way would be to reduce the Young modulus and the solid stiffness.)

And by the way, why would we change those two parameters at the same time? Shouldn't it be one or the other?

# Re: LMS Imagine.Lab Amesim - How are contact forces calculated

Valued Contributor

Hello Delphine,

I also had some difficulties to understand this submodel, but after using it a few time, I came up with a method to parameter it properly:

• The "young modulus" parameter is not the most usefull one to set up the global contact stiffness, I always set it to the value of the material I use, then to get a proper dynamic behavior and simulation time, I play with:
• solid stiffness
• contact damping coefficient
• limit penetration for full damping
• If your model simulation time is too long too launch a lot of runs, you can create a small model with just the 2D contact and few of its neighbours, then launch a lot of simulation on it to find the best parameters.

I had great results with this method, the mechanism I simulated always provided results close to tests, and I could give design insghts to the conception team.

Then to answer your questions, I would say the "young modulus" parameter is not truly the young modulus, I don't really know why it is named that way, but for me it is not that a big of a deal as you can still get very good results.

Also I use AMESim v16 and the parameter unit is properly set to N/m², maybe it is a bug the dev team corrected.

Finally I would say do not be concerned too much with the deformation of the shapes, it is not finite element analysis, your main focus shlould be: proper dynamic behavior, and fast simulation time.

# Re: LMS Imagine.Lab Amesim - How are contact forces calculated

Experimenter

Hi Hazzar,

As I can't conduct test on my mechanism, it is quite hard to predict its dynamic behavior. As you suggested, I kept on running simulations using different parameters to select the ones that seem to give a realistic behavior. I should be quite close to what we can expect now. I wanted to be sure that there wasn't another way to set the parameters and that I didn't misunderstand something throughout the modeling process.

I also asked the development team how to include the third dimension in my model and they simply said that it is not possible. There is the "3D mechanical" library but it can't model contact between two cylinders.

Thanks a lot for your help!