Inventor FEA constraints

[sambuls_autodesk]

Hello,

Assume a beam like the one in the picture below.

How must I configure the constraints?

The left constraints should be a "Fix constraint" on the edge(not the face) I assume.

But the right constraints must not fix all DOF's.

How can I configure the right constraint?

Thanks

sam

[henderh]

Hi Sam,

  The Frictionless constraint command would be the best one to use.  However, we allow only face geometry selection for this constraint type (not edges, since direction can be ambiguous).

  The best way to set up the roller constraint is to use the Fixed Constraint with Vector Components option.  Uncheck the directions that need to remain 'free':

 

Of course this example assumes the beam centerline is oriented in the global x-direction.  The beam centerline direction orientation to the global coordinate directions may differ for your model.

Bob van Der Donk made a very nice blog about how to use FEA to compare with these types of theoretical 'textbook' beam theory / strength of materials calculations (Dated 12/15/2011) here: http://beinginventive.typepad.com/being-inventive/stress-analysis/

I hope this helps!  Please let us know if you have any additional questions, comments or suggestions.

Best regards, -Hugh

[plyanne]

I have the same issue with the difference that I need the constraints in faces not in edges.

How would you say I should use the frictionless constraint? How can I specify the program that I want a reaction in the "y" or upward direction and not on the horizontal direction? I already have the coordinare system in place to be able to do so. 

Also, I tried using the fixed constraint specifying the direction that I wanted to constraint but the constraint indicates moment reactions. So it does not represent the rolling constraint as I want and the results are not as they should.

What should I do in this case?

[JDMather]

Attach your *.ipt file here.

You will probably need to look into Split faces.

[plyanne]

Here it is.

Mine is not as the beam in the previous picture. I need just a x reaction
force in the upper constraint and an y reaction force in the lower
constraint. The middle constraint is correct as it is.

Thank You

[plyanne]

Sorry. Here.

[QuasiMojo]

I haven't used the FEA in Inventor much because I have others available, and I was not able to produce an explicit reacton force here. Typically this is done as a summation/integral over a surface and I couldn't find how to do that in the time I could spare.  I did notice that the default mesh size is very coarse and makes your stresses look like they're having a much greater effect than they really are.  I refined the mesh around your holes and lowered the growth rate to give a better mesh.  You'll see all the higher stresses are local to the holes.  I'll be interested to learn how a reaction force can be calculated in this program.  That being said, you should be able to figure it out with a calculator; it's a pretty straightforward statics problem.

[plyanne]

Hi

I haven't used Inventor much either but to see the reaction forces and moments I right-click on top of the constraint in the left side column.

I can do a hand calculation to obtain the reactions but I am mostly interested to see the type of constraint being in effect just to see if there was a way to simulate a roller constraint. Like I said, I need a constraint with a single direction force reaction only; no moment reactions in any direction or forces in other directions. I think that is the only way I can simulate a roller.

I am not sure if Inventor has a spring force option but if it does I think maybe its possible to simulate a rolling single-direction reaction with a very high damping coefficient axial spring. I will see if thats possible and let you guys know. Just a theory.

I purposely made a coarse mesh just to decrease the simulation time while trying to figure out the constraints issue. But thanks for helping me skip that step.

I will look at the files you sent me as soon as I can.

Thank you very much.

[henderh]

Hi plyanne,

  I may have a solution for you.  We can take an assembly approach and use sliding / no separation contacts to eliminate the reaction moments.  I created some pins and applied the fixed constraints to those, freeing up the single direction component as you did in the part simulation.

  It has only an x-reaction in the upper constraint, and y-reaction in the lower constraint.  The smaller values are numerical noise and can be ignored.

Hope this helps!

-Hugh

[jlkono]

I have a slightly different scenario with my ramp that is pinned one end and grounded with intent for freedom to slide along the x-axis. However, my ramp is not a symetrical beam and I therefore need to have the freedom of all moments. When modeled as fixed edges x free, a moment reaction appears about the x-axis. Any other thoughts about a constraint that will result in 0.0 reaction moments?