Learning Forum

[Bassemaa]

Hello Everyone,

For a simple cantilevered I-section beam with transverse force at its end, there were some differences identified between BEAM188 and BEAM189 elements when meshing with a single element for each beam.

When I keep element order to “Program controlled”, BEAM188 elements are applied automatically, and deflection result matches analytical calculations and everything is fine. However, when I apply BEAM188 elements with APDL commands under geometry (et,matid,beam188), the deflection values deviate by about 15%.. I was wondering what is the difference between BEAM188 elements applied with APDL commands and BEAM188 (same elements type) that Ansys applies automatically?

Applying BEAM189 either by APDL or by changing element order to quadratic is also fine and delivered matching results. Increasing elements count also removes the deviation, but we want to employ the single element practice.

I am using Workbench 18.1. Thank you

[peteroznewman]

BEAM188 is a 2-node beam, so is a linear element.
BEAM189 is a 3-node beam, so is a quadratic element.
When you keep the element order Program Controlled, that generally creates quadratic elements. Look in the Solution Output and you may find that BEAM189 elements are used. What evidence do you have that BEAM188 elements are used?
For a single element, the quadratic element is going to match the analytical calculation of a cantilever beam. When you have many elements, the linear elements will also match the analytical calculation.

[Bassemaa]

My evidence is that I do look in the solution output and it is BEAM188. This is also supported by what the Ansys Help documentation says:
"For surface bodies and beam models, Program Controlled is identical to the Linear option described below. For solid bodies and 2-D models, Program Controlled is identical to the Quadratic option described below."
also if program controlled assigns BEAM189, which is indeed recommended for single element analysis, then there should not be any deviation (which is actually the case, when I change to BEAM189 either way)
The deviation occurs only when I change to BEAM188 with APDL command

[peteroznewman]

In Mechanical, when you mesh a line body, you are meshing with either linear elements or quadratic elements, even if you are only meshing a single element.
If you meshed with linear elements, you can't assign BEAM189 elements to that mesh.
If you meshed with quadratic elements, you can't assign BEAM188 elements to that mesh.

[Bassemaa]

Thanks for your relpy. Yes sure, I always select linear elements when I assign BEAM188 elements, and vice versa.
My quesition in other words: shouldn't BEAM188 always deviate because they are linear and element count is too low (actually 1)? why do they work fine one way and deviate in the other
by the way, BEAM188 also support quadratic functions if you activate KO(3)=2, and this also delivers conforming results.