I try to perform a nonlinear buckling analysis. Therefore I found one paper in internet where three possible ways are compared:
- load ramp step up using Newton’s method
- arc-length convergence method
- employing a nonlinear eigenvalue buckling analysis
Basically this is just a report of comparing the methods results, but not a proceeding how to perform a nonlinear buckling analysis.
Does anybody has some information or tutorial about this?
At the moment I try to use the first method, but I alredy have some points I am not sure about.
Basically in Sol 106 using subcases, the second subcase starts at the ending conditions of the first subcase, so e.g. I want to see the left plastic stresses after a deformation I clone the first subcase and neglect the forces. But how I apply a force when I want to rise the force from first subcase (e.g. 1000N) to 1500 N in the second subcase?
Would I set 1500 N in the second subcase or just the difference, means 500 N?
Is the bisectioning method used automatically or is it necessary to enable it somewhere?
Thanks for your help.
In SOL106, as you stated correctly, the second subcase starts at the stress/strain/stiffness state at the end of the first subcase. However, the loads in any subcase are total loads. If you want 1000 in the first and 1500 in the second, you apply 1000 in the first and 1500 in the second (not 500). Similarly, if you want to load to 1000 in the first, then unload in the second, you apply 1000 in the first and 0 in the second (not -1000). The solver automatically calculates the delta from the previous load vector.
NLPARM parameters control the loading and solution. The default for NINC is 10, which means that your loads will be applied in 10 increments. In the first subcase, this would be 100, 200, ... 1000. In the second subcase, the software calculates the delta P from the prescribed total loads, so the increments would be 1050, 1100, 1150, ..., 1500
Bisectioning is controlled by MAXBIS on NLPARM. it is set to 5 by default.
Arc length methods are activated/controlled by including NLPCI.