I have performed 3-point bending on steel beam with 20*20mm square hollow section. I had obtained the Force-displacement graph for the experiment. I am trying to simulate this experiment in Abaqus Dynamic EXPLICIT.
For material model, I have used yield stress-plastic strain data. Element type used is C3D8R and three elements are used per thickness. The loading rate is increased to 1 m/s within the first 10 millisecond and thereafter it is maintained constant. Kinematic contact method is used to define the interaction between roller and beam. Also, considering the symmetry of the model, I modelled only quarter section ( force was multiplied by 4 times ). Symmetric boundary conditions were applied
The values for Force Displacement curve in simulation comes considerably lower than actual values. What could be the possible reasons for this ? How can I rectify this error?
To get better simulation results, stress strain behaviour of material should be same as tested in the lab. Standard material models will lead to large variation in results. With your input for 'yield stress-plastic strain data', is it containing strain-hardening beyound yield ? That should be given as input from physical tests of the same materisl used for 3 point bending.
Second issue: 3 point bending is used for solid sections, where worst effects, say maximum bending with large shear force at the loading point occurs. This test may not give good results for hollow box section. Are you following any ASTM or other test method (as recommended) ?
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers