Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations GregLocock on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Roller Sliding Restraint

Status
Not open for further replies.

Adalius

Mechanical
Feb 13, 2009
57
This is probably a basic question but my experience lies outside of FEA. In its most basic form, the object we're analyzing is effectively a table. 4 legs, object weight roughly centered between them. In real life, the object will sit on the floor, not anchored or otherwise restrained. Is it better to have the bottom faces of all 4 legs be restrained via roller/sliding so they can 'flare' out, or should one be fixed for reference and the others sliding, or is there an even better way?
 
Replies continue below

Recommended for you

Adalius,

Is that even possible to model is Cosmos? I have a similar question a few threads down, which no one has answered.

For deflection and stress analysis, Cosmos requires your part to be fully constrained. You will have to be creative about your force and restraint model.

How about fully constraining a patch at the centre of your table top, and applying forces on the ends of your legs? Your legs ought to buckle at some point, but the corners of the table should be rotating a bit too.

Critter.gif
JHG
 
It's possible to do it either of the ways I described, it doesn't throw an error, blow the model apart with huge displacement, or produce numbers that are that far from what I'm getting with pen & paper. I'm just wondering which is the superior way of doing it.

I know it seems odd since it's not fully constrained if all 4 legs can slide around but I think the reason it works is that they're all sliding away from the center in different directions so the model is staying stationary and not 'squirting' out sideways or something.
 
Try modelling and solving the simulation with contact conditions between the legs and the floor and setting the appropriate coefficient of friction. I can't tell you what the appropriate coefficient would be, but maybe by successive runs with coefficients you could see if it even affects the solution. My gues is that once you find the threshold at where sliding would occur, any coefficient over that value should produce very similar results...
 
Use symmetry to provide restraint in two directions. The third restraint will be in the vertical direction, and the legs can flare out.

corus
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor