Comparison of simply supported T-beam and beam attached to the shell plate is done. Simply supported T-beam calculations are performed analytically and numerically (ABAQUS software). The beam attached to the shell plate calculations is done numerically.
Maximum bending stress difference in the case of T-beam analtically and numerical is 2.93%
Maximum deflection difference in the case of T-beam analtically and numerical is 3.34%
Maximum bending stress difference in the case of beam attached to the plate is 0 %
Maximum deflection difference in the case of beam attached to the plate is 0.093 %
Can anybody explain to me , why there is more difference is the case of T-beam( Analytically and numerically) as compared to the beam attached to the shell plate ?
I have attached the tables.
abaqus
Supposing that mesh is fine, the 3% difference seemes to be too much. Check that position of beam neutral axis in Abaqus and in hand calculation is the same.
Victor J. Аdlucky Thank you for your reply. In the case of hand calculation neutral axis (Cy ) is 67.778 mm. I put the same Cy in Abaqus along with other dimensions.
I have attached the document which I made for this and Abaqus inp, cae, jnl, and odb files.
Hi Zuffain Hussan,
Yes, input data are correct. The reason of difference in results may lie in how Abaqus works with T-shaped cross section. When I replace it by rectangular cross-sections with the same moment of inertia (3142222.22), Abaqus gives max deflections coinsiding with analytical solution.
It is worth comparing integration schemes for I (and T) cross-section and rectangular cross-section in Abaqus. The latter always has integration points on the boundary which makes integration more exact. A little table for the case when I cross-section becomes square section with moment of inertia= 3142222.22 is presented. It can be seen that the accuracy of integration decreases as the I-beam flange thickness increases.
- Badges
- Science topic