Friction Model not working

[roman.gebhard5MV47]

Hello,

please can you help me with my model.

I want to look if the friction is enough (red marked) so left component dont fall down.

Left: Load

Right: Fixity

Im trying a nonlinear solution:

I also tried the following settings:

But it always comes the error:

I have mesh size of 2,5mm and the following setup for the contact:

The contact works perfect if i change it to BONDED.

Also the nonlinear solution works for the bonded contact.

So i dont think the problem is in the model itself.

I try it with a shell model, but i had the same problem before with a volume model.

I just switched to shell because i thought its easier to find the problem.

Thanks in advance.

[John_Holtz]

Hi @roman.gebhard5MV47 

The question is when did the E5000 error occur: at the very beginning of the analysis? After N steps and the part on the left slid off? If it occurs after N steps, you know that the part is trying slide off. You should support the part on the left with multiple grounded springs (with low stiffness) so that the part is statically stable and does not give the E5000 error, but the weak springs do not take up very much load before it slides off. (Springs are added using "Connectors > Springs".)

Also, be sure to output all the load increments so that there are results while the load is being increased.

Personally, I try to avoid friction because it greatly adds to the complexity of the convergence. In your situation, these are some alternatives:

• Do a hand calculation. All the forces are known, so it is easy to calculate if the friction is sufficient.
• Use bonded contact. Does the contact force exceed the force that can be developed by friction?

John

[roman_gebhard97]

Hello John,

thanks for your reply. I'll try this with the weak springs (i also know them from ansys).

A hand caluclation isn't possible here, because i only use 2 profiles to show you the problem.

In reality there are 40 of them in a row, and i my task is it to figure out if they can hold their own weight 

without any weldseams. So i think i need a nonlinear solution even without friction. 

But yes if the solution doesn't work with weak springs, it would be an idea to determine the bonded contact forces.

I will try to use smaller steps at beginning it also may help if it's a convergence problem.

Thanks in advance, i'll give an update.

Roman