AnsweredAssumed Answered

Analysis of a spherical bearing

Question asked by Jahee Campbell on Mar 16, 2012
Latest reply on Apr 12, 2013 by Jared Conway

Like • Show 0 Likes0
Comment • 17

I'm trying to model a spherical bearing for a project I'm doing at university, I've drawn the model and it behaves as it should when I manually manipulate the model, rotating freely in the housing, however when I try and run any analysis on it in simulation to ensure it behaves as it should when it's used in the assembly it is to be inserted in, the bearing becomes rigid and I get some very strange and inaccurate results..

Does anyone have any experience with designing these, I'm sure it's a simple problem but it's had me baffled for a couple of weeks now.

I can provide pictures and models along with any other information anyone needs if requested.

Thanks in advance

Edit: There are no interferences, i've set the global contact to 'no penetration' and defined the surfaces that shouldn't be penetrated using the component contact definitions under connections drop down

First picture uploaded is the bearing assembly

Second picture is the analysis, with a simple bolt inserted to act as a lever.

Attachments

1.jpg261.8 KB
2.jpg293.4 KB

No one else has this question

Outcomes

17 Replies

John Burrill Mar 16, 2012 9:18 AM
Mark Correct
Correct Answer
Jahee, can you post a screen capture of the bearing and answer the question: what type of simulation are you doing? (motion, theremal, structural)
Like • Show 0 Likes0
Actions
- Jahee Campbell @ John Burrill on Mar 16, 2012 9:27 AM
  Mark Correct
  Correct Answer
  I've uploaded a couple of pictures, and to answer your question it's a static study. I wonder if this is the correct choice but the assembly that the bearing will be used in will be a static study so i wanted to ensure they are compatible.
  Like • Show 0 Likes0
  Actions
  - John Burrill @ Jahee Campbell on Mar 16, 2012 9:40 AM
    Mark Correct
    Correct Answer
    Try making your bearing subassembly flexible.
    To make a subassembly flexible: select it in the feature manager tree and pick the 'component properties' icon.
    In the component properties dialog, change the 'solve as' radio column to 'flexible' and click OK
    A flexible subassembly can be moved through it's remaining degrees of freedom. I'm not sure if it interfaces with simulation in this manner, but give it a shot.
    Like • Show 0 Likes0
    Actions
    - Jahee Campbell @ John Burrill on Mar 19, 2012 2:45 PM
      Mark Correct
      Correct Answer
      I had a try at your suggestion today, and found that the solve as rigid/flexible option is greyed out and i'm unable to click anything?
      Like • Show 0 Likes0
      Actions
      - Bill McEachern @ Jahee Campbell on Mar 19, 2012 2:55 PM
        Mark Correct
        Correct Answer
        mates have nothing to do with an FEA study or set up.
        Like • Show 0 Likes0
        Actions
Bill McEachern Mar 16, 2012 11:00 AM
Mark Correct
Correct Answer
I took a look at the pictures. The load case does not make sense from casual inspection. What are you trying to do? How are you trying to load the bearing? The shaft has a load but no restraint. the restraint on the bearing face is non-sensical. Isn't the shaft connected to something else?
Like • Show 0 Likes0
Actions
- Jahee Campbell @ Bill McEachern on Mar 18, 2012 4:10 PM
  Mark Correct
  Correct Answer
  Bill, admittedly my knowledge of solidworks is self taught so i may be approaching this the wrong way, but what i'm trying to do with this study is just to ensure the model acts as it should during analysis, my expected outcome is that the bearing and pin would simply rotate until they reached the housing, and stop at that point with some stresses concentrating at the point at which the two surfaces interact.
  
  The bearing is to be a part of another assembly, so i want to see that it moves freely and represents the real component correctly in this isolated simulation before creating the next assembly. That's the only aim of this test.
  
  The reasoning behind the restraint on the bearing face is that without it, the bearing dislocated from the housing,
  
  It might be of use for you to know how the components are mated? I created a reference point at the 'centre of face' of the bearing and the inner housing and mated them coincidently..
  
  John, i am unable to put this to the test until tomorrow but i'll let you know how i get on with that.
  Like • Show 0 Likes0
  Actions
  - Bill McEachern @ Jahee Campbell on Mar 19, 2012 2:59 PM
    Mark Correct
    Correct Answer
    You don't need simulation for what you want to know - just drag it around in SWX with collission detection on. Remember sigmaF=0 (must) in a static analysis. You can simulate the contact described but why would you? Wouldn't your design insure that never happens? You need toi set up contact pair on the surface and restrain the shaft with some small spring whose magnitude is imaterial in the big scheme of things to ensure the contact is caputred - shoudl work any way.....
    Like • Show 0 Likes0
    Actions
    - Jahee Campbell @ Bill McEachern on Mar 19, 2012 3:29 PM
      Mark Correct
      Correct Answer
      My design should ensure that never happens yes but in previous experience, parts that i have designed have not moved freely when run in a simulation for numerous reasons that took me days to get to the bottom of. I want to see that this compenent moves as i would expect in this isolated case first, which it is not doing thus far.
      I'll try and run it in my main assembly though to see if it does allow the movement i need and hope that it works. Thanks
      Like • Show 0 Likes0
      Actions
    - Jahee Campbell @ Bill McEachern on Mar 23, 2012 11:41 AM
      Mark Correct
      Correct Answer
      As i expected, i set up the assembly, including the spherical bearings, when the analysis was run, the bearings are rigid, i'm at a complete loss..
      
      Has anyone successfully modelled and simulated them before?
      
      There are pictures i can add if anyone would find them useful?
      Like • Show 0 Likes0
      Actions
      - Bill McEachern @ Jahee Campbell on Mar 23, 2012 12:01 PM
        Mark Correct
        Correct Answer
        post the bearing dialog box settings you used. This stuff is easy and it works - used it recently.
        Like • Show 0 Likes0
        Actions
        Jahee Campbell @ Bill McEachern on Mar 23, 2012 1:01 PM
        Mark Correct
        Correct Answer
        Sorry, what do u mean by bearing dialog box?
        Like • Show 0 Likes0
        Actions
        Bill McEachern @ Jahee Campbell on Mar 23, 2012 1:09 PM
        Mark Correct
        Correct Answer
        try usinghte bearing connector between the two parts........if the surfaces are coincedent the software by default bonds the surfaces. Your options are to set that contact to no penetration or to use teh bearing conector to manage the interaction of the parts - you don't even need the bearing parts inteh model when you use this but you will need appropraite selectable surfaces on the remaining parts (areas where thebearings are seated) to put the bearing connector between them.
        Like • Show 0 Likes0
        Actions
        Jahee Campbell @ Bill McEachern on Mar 23, 2012 1:42 PM
        Mark Correct
        Correct Answer
        Ah yea i get you, i did initially try this, but the program seems to only allow the bearing connector to be used between two cylindrical faces, when i attempt to select the bearing and the housing, two spherical faces, i get the following message "Please select cylindrical face or circular edge on shell", neither of which i have..
        
        The global contact is set to no penetration but i manually set no penetration sets to the two bearing parts and ran the simulation and it's now telling me the restraints are inadequate and the model is unstable, i can't think what other restraints are needed..
        Like • Show 0 Likes0
        Actions
        Bill McEachern @ Jahee Campbell on Mar 23, 2012 2:36 PM
        Mark Correct
        Correct Answer
        delete out the bearing, split the face on the shaft. If there is nothing on the other side of the bearing use the restraint/fixture bearing. Use the cylindiral face from the outer bearing housing if there are compoents there. Try taking a training course.
        Like • Show 0 Likes0
        Actions
Unspecified Unspecified Apr 12, 2013 5:26 AM
Mark Correct
Correct Answer
Hello

I think I have the same problem as the one you mentioned here. I've spent the last few weeks dealing with this but it just won't work, whatever I try. I have tried using the bearing connector but the results aren't right, the deformations are opposite of expected. If you somehow managed to solve it, please let me know (my problem is the main problem of my thesis). I tried connecting the inner face of the outter ring with a point in the middle of the section of the shaft using trusses but I just couldn't connect the joints to the outter ring. Now I'm trying with a concave circular ring and it still doesn't work.
Like • Show 0 Likes0
Actions
- Jared Conway @ Unspecified Unspecified on Apr 12, 2013 8:04 PM
  Mark Correct
  Correct Answer
  "results aren't right"
  "deformations are opposite of expected"
  
  you'll have to elaborate on what that means, and show some pictures and post your setups. last I looked at your problems, the issue was that you didn't have a static problem which was resulting in large displacement warnings or similar or the need for soft springs which invalidate the results. when the appropriate restrains were added, things moved as expected. until you find a static problem and add the appropriate restraints you will continue to have issues.
  Like • Show 0 Likes0
  Actions

Analysis of a spherical bearing

Attachments

Outcomes

Related Content

Recommended Content