This content has been marked as final. Show 2 replies
Regarding your boundary conditions (restraints). You should expect to see higher stresses if you model your restraints as perfectly rigid. This is true for areas closest to the restrained areas. As you get further away from your restrained areas, the way you model your restraint has less impact. As a general rule, locations greater than 3 characteristic lengths away are not siginificantly impacted (St Venant's principle). If you model your actual restraint (a floor) such as I think your beam is supposed to represent, then I would imagine your stress/strain values would be actually closer to reality provided the way you modeled your forces closely matches reality and that you structured your mesh suitably. You could always take the safer approach if you are unsure and make design decisions based on a rigid restraint, but this might lead to un-needed material use.
First, thank you for your comments and advice. I am encouraged by the willingness of others with more knowledge and experience willing to share with less exp. users on this format and hope someday I can return the favor. I owe a great deal to people like yourself, Vince Adams, and Basil Gello (on the fluids side).
For this problem, I think I agree, that the stress value in the shell is not much different assuming the induced moment created by the location of the stiff web of the beam is comparable with the rigid/fixed restraint option. I imagine the leg teetering on the beam slightly which I would assume increase reaction in the shell.
Others before me have modeled this scenario by fixing the shell (axially on top and bottom, w/ sym on the sides) and pushing up at the leg the equiv. mg. I think this is not a bad approach either, but again, my hunch is that results are slightly off since force vector would need to be updated with deflection of the pad in loading. (contact condition w/ traction).
Two points for clarity: I'm not sure I understand the comment about mesh structure... unless referring to sufficiently refined in critical areas without being too localized in level of refinement relative to global size? Also, I agree that conservative is best, but assumed that fixed restraint on some pad area would actually be less conservative than the actual conditions (with pad area actually rotating with beam rotation)?
Thanks again for the help.