Learning Forum

[vkm120991]

Hello All,

In penalty based contact formulations (augmented lagrange and pure penalty), when should we use "program controlled" stiffness vs when to override using a stiffness factor?

Current forum answers/literature suggests to use lower stiffness factor values (0.1 to 1) to soften the contact and overcome convergence issues. The only rule of thumb that keeps repeating in all literature and ansys help is "use high enough stiffness so that no convergence issues happen".

So if there are no convergence issues with 'program controlled' stiffness, we should never override stiffness?

I also noticed that similar element size 'tetrahedral' elements at contact generate higher contact pressure compared to 'hexahedral' elements (in my trials, i noticed that hexahedral elements showed good correlation with theoretical contact pressure calculations based on hertz theory, with program controlled stiffness). Any thoughts on the same?

 

[Akshay Maniyar]

Hi,

The Normal Contact Stiffness is the most important parameter affecting both accuracy and convergence behavior. A large value of stiffness gives better accuracy, but the problem may become more difficult to converge. If the contact stiffness is too large, the model may oscillate, with contacting surfaces bouncing off each other. We need find a balance between accuracy and allowable penetration. When any model is facing the convergence issue, and if it is because of high residuals at contact location, so we reduce the contact stiffness to reduce the contact force(penetration will be increased) and achieve the convergence.

Linear Tet elements are stiffer, so may because of that you might be getting high contact pressure. You can try with higher order tet elements with fine mesh and see if you can get similar results as hex elements.

Thank you,

Akshay Maniyar

How to access Ansys help links

Guidelines for Posting on Ansys Learning Forum

 

[vkm120991]

So if there are no convergence issues with 'program controlled' stiffness, we should never override stiffness?

[peteroznewman]

One type of model that is sensitive to contact penetration is the interference fit problem of a shaft in a hole. Stress in the parts changes a lot with the amount of penetration. An increase in Normal Stiffness of the contact can reduce penetration, but that is not the only adjustment needed to get accurate results in this model. I found that I also had to adjust the Analysis Setting. Under Nonlinear Controls,  override the Program Controlled Displacement Convervgence Tolerance and type in a much smaller value than 0.5%.