Hello all,
I am trying to model a 3D cantilever beam in Finite Element Software. I am away from my campus and I do not have access to the computer lab so I am using Calculix, a free finite element software, and besides being a free software it takes similar input file as Abaqus. When I model it I have 1 unit force in y direction in all the nodes on one end and on the other all nodes in the surface are fixed and I am using hex elements. I get good results for stress but the displacement I get a bogus answer. And when I animate it I get a plastic deformation rather than a beam vibrating. My guess is fixing the boundary is giving me the bogus answer. My question is do I need to fix all the nodes in the elements at the end or just the nodes in the surface which is at the end?

Not sure about Abaqus or Calculix, but in some other programs I have used, usually have to fix all the nodes of element.

BBC news, I am not the FEA guru but I have done several FEA models in few FEA programs.

Your model is very simple, just constrain one end of the beam (depends if it is shell or solide model - solid model gives good answer , I thnink hex is solid element). Put force on other end. You will get decent stress and displacement. There may be so many java template for cantilever beam, you can check your answer or you use hand calculation to check the answer. Instead of putting 1lb on each node, try to put 1 lb only in one node. Or try putting 0.0001 lb. If still you see large displacement, check your nodes, there may be dis-connection or it may not be constrained properly. Try merge nodes, it may work. Magnify your model and check how nodes are connecting. How elements are laying. Try to make very very simple beam lets say 1/2 inch wide and 1 inch long. and put force at the end. The best way to defug FEA model is make it simple and simple and try to sort out the issue.

I dont have time otherwise I would go to your so called free internet program. But i guess you may be able to run this model.

Good luck.

Nepalean

Bythe way what is the input file for calculix? May be I can make a model but I dont like to do that since you are a student and thts not ethical.

Constraint all one end. put force on other end. Try to use with small load. May be put a point load, put only one end.

Khai ta hamra yi biddarthi le prasna rakhe, k bhayo patto nai diyenan?

I am not sure what you mean by vibration as an expected result instead of displacement..as long as it is a steady state static load on one end and the fixed nodes on the other end. I would normally model solids with tet10 elements,,could do it with tet4's as well,,but tet4 are stiffer than 10. I guess for a uniform cross sectioned cantilever beam hex elements should be ok. I would have rather modeled it with 1D beam elements since it is just a simple cantilever.

 As far as plastic deformation is concerned..are u using a non linear analysis?? Plastic deformations from linear static solutions does not mean anything