Cancel
Showing results for 
Search instead for 
Did you mean: 

Contact in NX8

Experimenter
Experimenter

Hello,

 

I am student and have a propably short question (using NX8).

I want to create a simulation of a falling sphere on a thin plate. I use SOL 601, 129, but the "results"-button disappears after solving.

 

Unbenannt-2.jpg

 

I tried to create a Surface-to-Surface-Contact, the plate as the bottom-region and the sphere as the top-region. But no results after solving...

 

Unbenannt-1.jpg

 

I hope you can help me. Thank you very much!

Best regards from Germany

Robin

 

10 REPLIES

Re: Contact in NX8

Solution Partner Genius Solution Partner Genius
Solution Partner Genius

Hi Robin,

 

Here is the modified files. A lot of parameters changed.

The main reason why you are not getting your result is because you did not use Strategy Parameter at all. ATS (automatic time stepping) setting is required whenever Sol601 is used.

 

Regards,

Tuw

Re: Contact in NX8

Experimenter
Experimenter

Hello Tuw,

 

thank you very much for your reply.

I tried to rebuild it and tried to figure out, what your changes have been. I thought I found all changed paramters, but my solution stops exactly before contact. Do you have a final advice?

 

Thank you very much!

Many regards,
Robin

Re: Contact in NX8

Solution Partner Genius Solution Partner Genius
Solution Partner Genius

Hi,

 

The ground plane need to the 2D mesh.

 

Regards,

Tuw

Re: Contact in NX8

Experimenter
Experimenter

Thank you! It is working.

 

Many, many thanks!

Best regards Smiley Happy

Robin

Re: Contact in NX8

Creator
Creator

If the object being dropped is assumed to already have a max initial velocity (computed from hand-calcs), say in my case, 125 in/s from a 20" drop, do I need to still apply an enforced displacement boundary condition?     In which contexts would applying an enforced displacement necessary?   Just any where the objects are not yet touching each other?

Re: Contact in NX8

Solution Partner Genius Solution Partner Genius
Solution Partner Genius

Hello NXLover,

 

Provided that you are still referring to Sol601129, what you are describing is actually a good practice, where we no need to simulate from 20" drop height, but mere moment before the impact occurs.

 

For Sol601129, enforced displacement boundary condition is not required, but you need Transient Initial Condition definition. Enforced displacement boundary condition is more suitable when we try to simulate quasi-static analysis, where inertia effect is not involved in the simulation model. For example, we can use Enforced Displacement BC to gradually make the ball in contact with the floor 2D Mesh, without inertia or wave propagation of stress contour. Sol601129 take into consideration of inertia effect, thus enforced displacement does not make sense in this case..

 

Kind regards,

Tuw

 

Highlighted

Re: Contact in NX8

Creator
Creator

thank you!

Re: Contact in NX8

Creator
Creator

Ok, so just clarify, I'm trying to do impact between a shell ball and an aluminum cover (I thought this might save me a little bit of time not using any solid elements).    So theoretically, I've positioned the parts so that they would be physically touching, but model-wise, the models are separated by a 0.2875" distance (the wall thickness of the cover being 0.2" thick, and the ball being 0.375" thick).   Since these are both meshed at the midsurface, the physical gap should be (0.2+0.375)/2 = 0.2875".     Please see attachment for a picture.

 

Is using shell models a terrible idea for NX soln601/129?     I just started playing around with this yesterday as sort of a check solution to something else I'm running in Abaqus, but this NX model's having its own set of issues of not being able to converge and just erroring out. 

Re: Contact in NX8

Solution Partner Genius Solution Partner Genius
Solution Partner Genius

Hi NXLover,

 

If you are using 0.2875" exactly, between the ball and cylinder feature, then i would recommend you to slightly increase the gap, for example until 0.3", in order to avoid initial penetration between the two object. Due to discretization (or meshing process) of the non-planar surfaces (or the ball and cylinder), it is possible that at some point, the measured gap between the ball mesh to the cylinder mesh is lower then 0.2875", and initial penetration occurs. although I believe the solver will be able to solve for initial penetration as well, why dont make life of the solver simpler by avoiding it?

 

In Sol601129, the default contact option for OFFSET is 0mm. This setting is can be accessed from Surface-to-surface contact dialog, Local Contact Pair Parameters group, Advanced Nonlinear (BCTPARA) > click Create Modeling Object... , in Contact Parameters - Advanced Nonlinear Pair, find default offset (OFFSET). By having 0mm means, for example in your case, the ball 2d mesh would need to travel 0.2875" before making impact with the cylinder 2d mesh. In other words, the 2d mesh (generated from midsurface) ignores the mesh thickness.

 

If you think the mesh should not travel 0.2875" without contact happening and should directly assume impact, then you will need to adjust this parameter. If you are using midsurface, then there is a convenient option which is to change Type of Offset (OFFTYPE) = Half of Shell Thickness. By setting this, contact impact occurs without the need to travel 0.2875".

 

Using 2D mesh is definitely fine for advanced nonlinear solution, Sol601106, Sol601129 and Sol701. If you are using NX12 or Simcenter 3D version 12, please see documentation > Simcenter 3D (CAE) > NX Nastran 12 > NX Nastran Advanced Nonlinear Theory and Modeling Guide > page 46.

 

Kind regards,

Tuw