How to replicate a rigid distributed load on a "flexible" beam?

[Michaelmcgo]

Hello,
First post here. I am using Inventor 2012 Professional with the FEA package. I am trying to model a frame with 800 lbm of load support across the faces of two angle rails (see screenshot). The load is extremely rigid (a stack of structural UHMW "pallets") and will not deform significantly under its own weight even when supported at just 2 points. Looking at my model, the load should be supported by "Points A" where the angle is welded to the frame, yet I am getting .045" of deflection on the angle at "Points B" (i.e. the angle is deforming more than the load placed on it will deform).

I have had FEA instruction in the past and as far as I understand, a distributed load behaves like bag of water, not an infinitely rigid board. Meaning the load moves with the supporting beam as it deforms. Is this correct? If so, it appears the only way to define a rigid distributed load would be by drawing in the structure of the load and giving it a weight (which adds a lot of complexity to the simulation and therefore introduces its own inaccuracies). I understand this may be the most accurate way of modeling a load, but I am hoping for a settings option that will allow me to define my load as rigid.

 

[rb1957]

are points B attached to the pallet ? ... the pallet can be rigid, and if the frame isn't attached to it, it can deflect differently.

if you want to constrain a bunch of points, use an RBE connection.

 

[Michaelmcgo]

The pallet rests on the two angles and is not attached to them in any way. The pallet's weight causes a vertical load (gravity) to be distributed along those two angles.

The solution shows the angles being overly stressed and deflected at the "Points B" locations. It would seem that the FEA simulation interprets a distributed load as a flexible "blob" that continues to load and deform the angles as the angles deflect. In actuality, the load will be rigid and will not continue to contact the angles in the areas of "Points B" if the angles deflect at all and therefore would not continue to to exert a force on the angles in that area.

I've run into this problem many times and I have a hard time believing the designers of the software would assume that all loads are completely flexible.

Does this make sense?

 

[rb1957]

if your pallet is rigid, then instead of a distributed load onto the frame underneath the pallet, maybe you should load the edges of the pallet (allowing that the frame will deflect away from the pallet).