Learning Forum

[anushka.dongare]

I am running a Hypervelocity Impact simulation of a projectile impacting an assembly built with Constrained nodal rigid bodies and Beam elements to represent the bolts.

When I run the model without setting a Control_termination endtime (so just initialise the model with a velocity applied to the impactor) it will initalise normally and terminate normally. But when I include the Control_termination endtime keyword and run the model, it will initialise but then all the solid elements which are connected to my CNRB fail on the first timestep.

Could someone help explain why this might be the case please?

For context, the model is built from entirely solid meshes (quad and tet for different parts), and the elements which are failing are those connected to a second order solid tetrahedral mesh (the mesh was generated in hypermesh and imported to LS DYNA) - for this part I am using a Solid Section with ELFORM 16 following literature I found online. 

 

Thank you!

[ErKo]

 

Hello Anushka

Good to hear from you.

Ansys empl. are not allowed to help here  (projectile vel. impact related) – please wait for other forum members to help (Dennis Chen).

Thank you

Erko

---

 

[Dennis Chen]

Hi anushka.dongare,  if including control_termination caused solid element to fail, My suspcision is that you have erode=1 set in control_timestep.    This would cause the solver to delete highly skewed elements while allowing the analysis to continue up to a certain point and then it would terminate.  

If this is allowed to share, feel free to send to zchen6@gatech.edu and I am happy to take a look tonight (in like 12 hours).   

If you cannot share the file, I would consider trying the following:

1) first ascertain why the element is deleted and which keyword created this.    This could be control_timestep as mentioned above.

2) check your CNRB definition to see if it creates any unreasonably high stress/strain on the constrained elements.    

3) set database_binary_d3plot with a load curve and set the first few thousand cycles to have very very small dt in d3plot.   This allows you to visualize what's happening so you can find more clues.  terminate the job after a few thousand cycles too in control_termination. 

4) try to start with a very simple model of what you want to do and see if you can replicate the issue in a very simple model.   That's often the best way to troubleshoot in my opinion. 

Hope this helps.