AnsweredAssumed Answered

Is it possible to define no penetration contact between a beam and a solid in non lineal dynamic simulation?

Question asked by David Barbado on Apr 5, 2017
Latest reply on Apr 7, 2017 by Bill McEachern

Like • Show 0 Likes0
Comment • 6

I'm working on a project that consists on the modelling of a rigid overhead contact line ( rigid catenary).

I need to simulate the contact between the pantograph of the train and the contact line.

In order to decrease computational times ( more than 1 day) I want to consider the overhead contact line as a beam. According to the properties of the pieces, this is a good simplification.

I'm working on a non lineal dynamic analysis.

The problem appears when I define the no penetration contact between the pantograph (solid body) and the catenary (beam), which is not allowed.

I can select faces of the beam, but then I can´t confirm the operation.

I attach some images in order to clarify my doubt.

Big thanks.

Ryan Dark Apr 5, 2017 12:17 PM

Correct Answer

Hi David,

You are totally correct in what you have found. No Penetration contact is not allowed for beam elements in SolidWorks Simulation (it is a software limitation). No Penetration is allowed for shell elements though which could also serve to cut down on your solve time. You may want to turn the beam into a shell body and try that.

See the reply in context

No one else had this question

Outcomes

Ryan Dark Apr 5, 2017 12:17 PM

Hi David,

Actions

David Barbado @ Ryan Dark on Apr 6, 2017 2:08 PM

What do you think about this solution?

Do you think it will work?

Do you think it will cut down the solve time?

I thought that the problem was that beam doesn't allow non penetration contact but on the other hand it allows bonded contact. Solid allows no penetration contact.

Because of that it cames to my mind the idea to use this configuration.

Thanks.

Actions

Ryan Dark @ David Barbado on Apr 6, 2017 4:29 PM

Hi David,

Technically that setup would work. I have never considered that approach though so I have never done any sort of validation testing on whether the results remain accurate. You would have to make sure that the solid object was not so stiff as to disrupt the results of your beam body. That is about the only hiccup that comes to mind right now though.

Actions

Bill McEachern @ David Barbado on Apr 6, 2017 7:34 PM

Is this an improvement over modeling the beam with solids or shells? You could use a pretty course mesh on the beam and still get pretty good results from what you have displayed. You could also just model the section in contact with a solid mesh and the rest as beams - I would htink this would be a better way to go. Solids would be a better choice than shells - more robust contact.

Actions

David Barbado @ Bill McEachern on Apr 7, 2017 4:28 AM

I have to do tests but I think that it is an improvement over modeling with solids (simulation with solids are extremely long).

I've tried to model with shells but I think that the simplification its not accurate enough.

"You could also just model the section in contact with a solid mesh and the rest as beams". How can I mesh only one section of a beam as a solid and the rest as a beam? I still need two bodies (one as a solid an the other as a beam), isn't it?

Actions

Bill McEachern @ David Barbado on Apr 7, 2017 2:10 PM

you would need three bodies by the looks of it. Just split your beam and make the before and after bits beams and the center section solid. bond the beam nodes to the surface of the cross sections.

Actions

6 Replies

Ryan Dark Apr 5, 2017 12:17 PM
Unmark Correct
Correct Answer
Hi David,
You are totally correct in what you have found. No Penetration contact is not allowed for beam elements in SolidWorks Simulation (it is a software limitation). No Penetration is allowed for shell elements though which could also serve to cut down on your solve time. You may want to turn the beam into a shell body and try that.
Like • Show 0 Likes0
Actions
- David Barbado @ Ryan Dark on Apr 6, 2017 2:08 PM
  Mark Correct
  Correct Answer
  What do you think about this solution?
  
  Do you think it will work?
  
  Do you think it will cut down the solve time?
  
  I thought that the problem was that beam doesn't allow non penetration contact but on the other hand it allows bonded contact. Solid allows no penetration contact.
  Because of that it cames to my mind the idea to use this configuration.
  
  Thanks.
  Like • Show 0 Likes0
  Actions
  - Ryan Dark @ David Barbado on Apr 6, 2017 4:29 PM
    Mark Correct
    Correct Answer
    Hi David,
    Technically that setup would work. I have never considered that approach though so I have never done any sort of validation testing on whether the results remain accurate. You would have to make sure that the solid object was not so stiff as to disrupt the results of your beam body. That is about the only hiccup that comes to mind right now though.
    Like • Show 0 Likes0
    Actions
  - Bill McEachern @ David Barbado on Apr 6, 2017 7:34 PM
    Mark Correct
    Correct Answer
    Is this an improvement over modeling the beam with solids or shells? You could use a pretty course mesh on the beam and still get pretty good results from what you have displayed. You could also just model the section in contact with a solid mesh and the rest as beams - I would htink this would be a better way to go. Solids would be a better choice than shells - more robust contact.
    Like • Show 0 Likes0
    Actions
    - David Barbado @ Bill McEachern on Apr 7, 2017 4:28 AM
      Mark Correct
      Correct Answer
      I have to do tests but I think that it is an improvement over modeling with solids (simulation with solids are extremely long).
      
      I've tried to model with shells but I think that the simplification its not accurate enough.
      
      "You could also just model the section in contact with a solid mesh and the rest as beams". How can I mesh only one section of a beam as a solid and the rest as a beam? I still need two bodies (one as a solid an the other as a beam), isn't it?
      Like • Show 0 Likes0
      Actions
      - Bill McEachern @ David Barbado on Apr 7, 2017 2:10 PM
        Mark Correct
        Correct Answer
        you would need three bodies by the looks of it. Just split your beam and make the before and after bits beams and the center section solid. bond the beam nodes to the surface of the cross sections.
        Like • Show 0 Likes0
        Actions

Is it possible to define no penetration contact between a beam and a solid in non lineal dynamic simulation?

Outcomes

Related Content

Recommended Content