contact between a ball and a plate

[shkoko]

hello everybody,

if someone couls help me pls.
I´m simulating a shock between a ball of 1 mm diameter and a plate of glass, both ball and plate are deformable,the ball has an initial velocity of 50 m/s. I simulate the shock by dynamic transient in à time so short ( one handret microsecond).
the probleme is : after some increment the jub stop, and it give me a error file which say " Error in 2d contact code. It should not occur."
what is the meening of this error?
Thank you in advance for any help.

shkoko

 

[Jordonlaw]

shkoko,

I love these messages such as " Error in 2d contact code. It should not occur." I've had similar ones for 3D contact and when I ask the support people they say that this message is included so that problems of this type are sent to a particular person within MSC. Simply, its a bug that they haven't fixed yet.

Ways around it include the usual things, change the mesh, change the timestep, move one part slightly with respect to the other. It might help if you can post pictures of the undeformed mesh and the deformed mesh when the job stops.

Does the problem occur if the impact velocity is small, ie a quasi-static loading case ?

 

[shkoko]

thank you Jordonlaw,

this problem occurs only in the case of dynamic loading (in my case, I use dynamic transient loadcase), I have already done the same mesh with a static case but there was not a problem, it is only for the case of dynamic load specially for comparatively big velocity,
in fact, when the velocity is big, I have progresly an interaction entre the mesh of ball and the mesh of plate (deformation of element in the mesh) until this deformation arrives at critical point , at this time it done me this error. therefore more the velocity is big more this error display quickly.
please what do you mean by "post pictures of the undeformed mesh and the deformed mesh when the job stops"?

 

[Jordonlaw]

My question is, can you provide an image of the mesh at the start of the solution and at the end of the high velocity analysis so we can see what mesh you have and how it deforms. This may help you get a solution.

 

[shkoko]

ok, i send you 2 photos for mesh before deformation and after deformation.
the plate is fix, and the ball has a velocity

 

[shkoko]

this is the photos for before deformation

 

[Jordonlaw]

OK, so it is clearer now. Your defomed mesh shows considerable penetration of the ball elements which is why you're getting the error. Looking at your undeformed mesh, the element size of the plate is 1/2 that of the ball; they should be practically the same if possible. Make these changes then look at reducing the timestep as well. Hopefully you'll get a better result.

One other thing. Is this a 2D or 3D analysis ? If 3D, try a 2D job first to get the job working and then move to 3D. The 2D job will run much quicker.

 

[shkoko]

ok thank you for your advice, I will try that then tell you about it.
by the moment it is a 2D analysis