Finite Element Analysis of Collision Impact in MD Patran/Nastran

[abkdt41]

I am trying to run a finite element analysis for a collision impact. The program I am using is MD Patran/Nastran. My model consists of moving object colliding with a stationary object. I have created the mesh, applied the appropriate material, applied the boundary conditions (fixed for the stationary object, and velocity for the moving object) Is there anything that I am missing, since the solver does not show the results.

I have done some static analysis prior to this and have always gotten results.

 

[droc22]

You may be better off calculating what the total impact force of the object is, and then doing static analysis on what's being hit.

 

[Zachary Liu]

How do you define contact in the simulation?

I am trying to run a finite element analysis for a collision impact. The program I am using is MD Patran/Nastran. My model consists of moving object colliding with a stationary object. I have created the mesh, applied the appropriate material, applied the boundary conditions (fixed for the stationary object, and velocity for the moving object) Is there anything that I am missing, since the solver does not show the results.

I have done some static analysis prior to this and have always gotten results.

 

[Mech_Engineer]

I am trying to run a finite element analysis for a collision impact. The program I am using is MD Patran/Nastran. My model consists of moving object colliding with a stationary object. I have created the mesh, applied the appropriate material, applied the boundary conditions (fixed for the stationary object, and velocity for the moving object) Is there anything that I am missing, since the solver does not show the results.

I have done some static analysis prior to this and have always gotten results.

A dynamic collision model is a LOT more complicated than a steady-state linear analysis. You didn't mention anything about contact conditions or how you're detecting contact, so I'm going to guess that in the model the components fly right through each other and don't collide.

Even if you were able to add proper contact conditions, I think it's pretty unlikely you'll be able to get useful results out of the model. Dynamic analyses are a fine art, and I think you're probably in over your head.