Hi Ivan,

I see a Ansys  tutorial for case contact Hertz for sphere, maybe you can try this to consult your problem.

https://confluence.cornell.edu/pages/viewpage.action?pageId=238913625

Unfortunately in Ansys tutorial I still can not find the solution.

And analysis is made using LS-Dyna , which is different from Ansys native modules.

Thanks

Best Regards

which are you use effect (for examle hourglas effect etc)

I can not understand. It seems i do not use any effects. (I think so and I hope so). I am using plane162 elements  (axi ximmetric, Lagrange, volume weighted ). Material density 7.85e-3, Young's modulus 2.1E+005, elastic material. 2D surface to surface contact.

In fact Ansys pre processor to LS- Dyna  allows only self contact in 2D. But self contact does not produce contact force output. I manually edit input file generated by Ansys to change contact type from 2d self contact to 2d surface to surface.

Is it possible that people from Ansys know about this problem with contact force and that is why they did not implement 2d surface to surface contact in Ansys pre processor for LS-Dyna?

Thanks

Best regards

you must select rigid  for plate , if shepher imact then sprint.. or you must select hourglas effect for plat If the sphere will pass through the plate

Unfortunately Hourglass control did not help.  Even turning hourglass control coefficient to the recommended maximum of 0.15 also did not help. Contact force is still more than it should be.

Increasing bulk viscosity also did not help.

So, should i stop using LS-Dyna, or should i publish incorrect results?

Or what can i do in this situation?

To LS-Dyna or not to LS-Dyna.

That is the question...

Thanks

Best regards

I understand you're modeling a drop test. and plate should  rigid so plate is not deformed.

You need to look at the time steps. for example  I divided contact time into fifty equal parts. So you can check the results by time in this way

Plate must be elastic, and sphere too. In reality all the bodies are elastic.

Time increment is 1.5e-6. I think it is small enough. This implies that elements are also small enough. So 25% error in contact force  is far too much.

All the problem seem lies in spherical form.

Thanks

Best regards.