I have a cantilever beam, 8 in long, rectangular cross section W=0.75 in, H=1.5 in, E=29 msi, Poisson's=0.3, under uniform line load qy = -50 lbf/in (y direction) entire length, bending around strong axis (zz). Beam theory solution M*c/I gives max. stress = 5687 psi.
I model with ANSYS Workbench, uniform mesh Ex-3-6-3D-Normal-Stress.PNG attached.
If I refine the mesh I only get more and more stress concentration. Expected but how to avoid? I just want to get the same result as M*c/I to show how FEA can predict the same as beam theory.
The results seem to be wrong. Stress concentratio at corner I understand, but why stress drops so fast when moving away from the clamped end?
Is there a way to clamp the beam to avoid the stress concentration?
| Attachment | Size |
|---|---|
| Bending stress in clamped beam | 33.58 KB |
problem solved !
Do not use FIXED support for the face at the root of the beam (clamped end). That gives Ok result of max. deflection but terrible result for bending stress.
I fixed only UX at root face.
To hold the shear force, I fixed the line created by the intersection of the neutral surface and the end face of the beam. There is a stress concentration there but I do not care because it is far from the loaction of max. bending stress. If you fix the edge at the bottom of the beam, the stress concentration ruins your bending results (compresion side).
To avoid rigid body motion I fix UZ at the center of the face.
To get truly 2D results, I set Poisson's=0.003 but if I use 0.3, the only result affected is Szz which anyway I do not need.