Learning Forum

[rlolachi]

Hi All,

I have been lurking on this forum for a little while now and have found it an invaluable resource in solving my ANSYS problems, so far. Unfortunately, I have become stuck. I am very much a newbie and only a physicist rather than an engineer, but I hope someone could help me.

I am conducting a simulation of a silicon rubber probe tip pressing into a steel plate. I am interested in modelling the Contact Area vs Contact Force. I have been doing this with static structural and transient structural models. There are two basic scenarios: A) pushing a probe tip alone into the steel plate, or pushing the probe tip whilst in an aluminium holder. For my transient analysis the problem comes in scenario .

The probe tip radius is larger than the holder hole radius and so needs to be modelled as an interference fit. This is done geometrically in the SpaceClaim model. The first load step in the model is to allow ANSYS to sort out the interference fit, followed by pressing (through displacement), the probe tip and holder into the baseplate. 

I have been able to successfully model the interference fit in with a static structural model. However, when I try to do the same using the first loadstep of the the transient model I cannot get convergence and am issued with the following error in the Solver Output:

REASON FOR TERMINATION. . . . . . . . . .ERROR IN ELEMENT FORMULATION

I have tried increasing the number of substeps and reducing mesh element size. Neither of these have helped. Any help would be most gratefully received.

 

Thank you in advance!

 

Note on Attached Archive: The archive contains both models I mentioned above and a few extras. The transient model is called "Smaller Model with Holder Transient" and the static interference fit model mentioned is called: "Smaller Model With Holder". 

 

Model Setup Showing Displacement vs Time Step: Probe tip in holder at top pressing into circular baseplate

[jj77]

This is a quasi static problem thus inertia is not important since you are not doing any dynamics say studying an impact or a shock. So linear static is the correct analysis.

[rlolachi]

Thanks for your reply jj77. The process that I am basing it on is a bit more dynamic than the way the model is set. In that the probe tip comes down at 2.5 ms^-1 then goes straight back up. Is that impact speed low enough that it can still be treated as quasi-static for this case? Also, the idea was to get an Area vs Force plot as it pushed down during one run (i.e. push and recede). Would it be better then for me to try a number of static runs at predefined force values instead to represent the different stages of a run? 

[peteroznewman]

Chapter 4 in the ANSYS Help Explicit Dynamics Guide describes how to setup a Static Structural model to Prestress an Explicit Dynamics model.

 

[rlolachi]

Very sorry I did not get round to thanking you both sooner, but I was away for a while. Thank you both @peteroznewman @jj77. Your help is much appreciated.