AnsweredAssumed Answered

Joint dimensions for friction of coincident planar mate?

Question asked by Dave Kolacz on Mar 17, 2009
Latest reply on Mar 18, 2009 by Ian Hogg

Like • Show 0 Likes0
Comment • 6

Hello,
I made a simple model to try and understand the joint dimensions for friction in a coincident planar mate. It's left me more confused.

My model is just a .10 inch thick by 1.00 x 2.00 inch plate of aluminum (density of .10 lbf/in^3 so total weight is .020 lbf). I applied a .001 lb force to the .10 x 2.00 inch face. Gravity set to load the 1.00 x 2.00 inch face. Initially I set the friction to zero and got perfect results for accelleration, velocity, and displacement (as compared to hand calculations).

When I set the friction coefficient to .025 (to give half the acceleration as the frictionless case) I got strange results.
With coincident mate friction dimensions set to L=1, W=2, and R =2 the plate only moved about .050 inch and then stopped. I would expect some
"tipping" of the plate to cause less than ideal case results, but with the plate only being .10 thick this should not have had much influence.
I ended up increasing L and W to several hundred times thier actual size before getting reasonable results.

Can any body tell me what I am doing wrong?

How do L, W, and R relate to each other? Is L and W used (while ignoring R) for a rectangular surface (such as mine)? Is R used (while ignoring L and W) for a round surface?

Please help. This has really shaken my confidence in my ability to get useful results from any model.

No one else has this question

Outcomes

Ian Hogg

Mar 17, 2009 8:26 PM

Hi Dave,

Take a look at this thread:
https://forum.solidworks.com/forum/messageview.cfm?catid=82&threadid=7288&highlight_key=y&keyword1=friction

I believe the aspect you are neglecting is the effect of bending moment on the joint and the added friction this generates. Hence when you increase the joint dimensions, the bending moment is transfered to very small force couple and has negligble impact.

Let me know if you have any further questions.

Cheers,

Ian

Actions

Dave Kolacz @ Ian Hogg on Mar 18, 2009 12:26 PM

Hi Ian,
Thanks for the response. I am hoping to gain an understanding of what is happening. I could not get to the link suggested in your last reply. How do I get to it?
The .001 lb load on the .1 x 2.00 inch face would give a moment of .001 lb x .05 inch (as I'm sure you already knew).
The normal force between the 1.00 x 2.00 inch face is .020 lb (the weight of the plate.
The moment due to the .001 lb load would cause a redistribution of the .020 lb normal load. In effect the centroid of the normal force is offest .0125 inch from the center of the plate. This should not increase the friction since the net normal force is still .020 lb. -or even if there was some sceondary effect it would have to be small.
By the way: I reran this using a 3D contact (instead of the conincident mate with friction) and got the expected results. Something strange appears to be going on with the coincident mate with friction (I'll probably give SolidWorks Tech support a call about it).

Also what is the "R" in the parameters for friction in the coincident mate menu (how does R relate to L and W -I can't tell from the diagram)?

Actions

Ian Hogg

@ Dave Kolacz on Mar 18, 2009 12:40 PM

Hi Dave,

Do a search in the forums for "Friction in CMotion". For some reason this is placed under the Flow Simulation category (maybe they moved incorrectly when restructured the forums).

The most common root cause for different results is joint position vs cg. If you can post the model, I can verify for sure.

Care must be taken if utilizing joint friction to a plate on ground, since this does not really represent the true situation. As you said, 3D contact gave you the correct result since it cannot have tensile contact forces between the parts whereas the joint friction just deals with a force couple equally balanced between the ends of the plate (tensile & compressive forces are treated the same).

If the plate in a guide (ie it is restricted in movement on both the top and bottom surfaces, then joint friction more accurately the situation.

Cheers,

Ian

Actions

Ian Hogg

@ Ian Hogg on Mar 18, 2009 12:43 PM

I forgot to mention. The R term is the "effective radius". Then the plate is rotating, this is assumed to be the radius at which the frictional force is reacting the rotation.

Ian

Actions

Dave Kolacz @ Ian Hogg on Mar 18, 2009 3:23 PM

Ian, thanks for your quick and informative responses. you should be on the SolidWorks payroll

I contacted SolidWorks customer support about this issue. They agree it seems to be giving bad results and that the model was set up correctly. They are investigating further.

Actions

Ian Hogg

@ Dave Kolacz on Mar 18, 2009 3:37 PM

Hi Dave,

I actually do work for SW in product marketing.

Motion Simulation is one of my pet areas of interest since I started out in that industry many years ago.

Let me know how you get along with the investigation of the issue.

Cheers,

Ian

Actions

6 Replies

Ian Hogg Mar 17, 2009 8:26 PM
Mark Correct
Correct Answer
Hi Dave,

Take a look at this thread:
https://forum.solidworks.com/forum/messageview.cfm?catid=82&threadid=7288&highlight_key=y&keyword1=friction

I believe the aspect you are neglecting is the effect of bending moment on the joint and the added friction this generates. Hence when you increase the joint dimensions, the bending moment is transfered to very small force couple and has negligble impact.

Let me know if you have any further questions.

Cheers,

Ian
Like • Show 0 Likes0
Actions
- Dave Kolacz @ Ian Hogg on Mar 18, 2009 12:26 PM
  Mark Correct
  Correct Answer
  Hi Ian,
  Thanks for the response. I am hoping to gain an understanding of what is happening. I could not get to the link suggested in your last reply. How do I get to it?
  The .001 lb load on the .1 x 2.00 inch face would give a moment of .001 lb x .05 inch (as I'm sure you already knew).
  The normal force between the 1.00 x 2.00 inch face is .020 lb (the weight of the plate.
  The moment due to the .001 lb load would cause a redistribution of the .020 lb normal load. In effect the centroid of the normal force is offest .0125 inch from the center of the plate. This should not increase the friction since the net normal force is still .020 lb. -or even if there was some sceondary effect it would have to be small.
  By the way: I reran this using a 3D contact (instead of the conincident mate with friction) and got the expected results. Something strange appears to be going on with the coincident mate with friction (I'll probably give SolidWorks Tech support a call about it).
  
  Also what is the "R" in the parameters for friction in the coincident mate menu (how does R relate to L and W -I can't tell from the diagram)?
  Like • Show 0 Likes0
  Actions
  - Ian Hogg @ Dave Kolacz on Mar 18, 2009 12:40 PM
    Mark Correct
    Correct Answer
    Hi Dave,
    
    Do a search in the forums for "Friction in CMotion". For some reason this is placed under the Flow Simulation category (maybe they moved incorrectly when restructured the forums).
    
    The most common root cause for different results is joint position vs cg. If you can post the model, I can verify for sure.
    
    Care must be taken if utilizing joint friction to a plate on ground, since this does not really represent the true situation. As you said, 3D contact gave you the correct result since it cannot have tensile contact forces between the parts whereas the joint friction just deals with a force couple equally balanced between the ends of the plate (tensile & compressive forces are treated the same).
    
    If the plate in a guide (ie it is restricted in movement on both the top and bottom surfaces, then joint friction more accurately the situation.
    
    Cheers,
    
    Ian
    Like • Show 0 Likes0
    Actions
    - Ian Hogg @ Ian Hogg on Mar 18, 2009 12:43 PM
      Mark Correct
      Correct Answer
      I forgot to mention. The R term is the "effective radius". Then the plate is rotating, this is assumed to be the radius at which the frictional force is reacting the rotation.
      
      Ian
      Like • Show 0 Likes0
      Actions
      - Dave Kolacz @ Ian Hogg on Mar 18, 2009 3:23 PM
        Mark Correct
        Correct Answer
        Ian, thanks for your quick and informative responses. you should be on the SolidWorks payroll
        
        I contacted SolidWorks customer support about this issue. They agree it seems to be giving bad results and that the model was set up correctly. They are investigating further.
        Like • Show 0 Likes0
        Actions
        Ian Hogg @ Dave Kolacz on Mar 18, 2009 3:37 PM
        Mark Correct
        Correct Answer
        Hi Dave,
        
        I actually do work for SW in product marketing.
        
        Motion Simulation is one of my pet areas of interest since I started out in that industry many years ago.
        
        Let me know how you get along with the investigation of the issue.
        
        Cheers,
        
        Ian
        Like • Show 0 Likes0
        Actions

Joint dimensions for friction of coincident planar mate?

Outcomes

Related Content

Recommended Content