Using surface to surface, penalty method and hard contact - should penetration be zero?

[drennon236]

Say I have two cylindrical surfaces in contact (in reality there are three) and I am using surface-to-surface, penalty method and hard contact:

The cylinders are sticking, but when I check maximum penetration I see that it comes to 4.5e-09 m. Meaning there is very small amount of penetration.

However, if I use "hard contact" - shouldnt the pressure develop like down below? Meaning no penetration. Because right now I get a small amount of penetration, also normal pressure of 5 MPa develops due to the penetration that occurs. The compressive strength of my part is 120 MPa, so this gives an error of approximately 4% - Is this acceptable? Or do I have to use Lagrange method instead of penalty? My project manager said he wants a penetration of 0, but im not sure this is possible without significantly increasing computational time.

 

[FEA way]

It’s typical for penalty method that some small penetration occurs. Usually it’s not significant for results but you may try adjusting penalty stiffness or using Lagrange multipliers.

 

[drennon236]

I get this amount of stress due to penetration, would you say its acceptable? The part (concrete) has a concrete compressive strength of 120 MPa, its a high strength concrete. If its hard to read red is shown as 5MPa

 

[FEA way]

You should try reducing this penetration but if nothing helps, omit it. Proper value of penalty stiffness (maybe nonlinear function) can help. Also make sure that mesh of these two parts is good.