spikefer
Structural
- Jul 31, 2012
- 10
Hi,
I am trying to identify a structural instability with nonlinear solution (106) in a very simple model, a beam loaded by a force acting with a 2° angle of incidence (to generate the lateral instabililty). What happens is as follow:
- setting "ITER, 1" as "KMETHOD, KSTEP", the simulation, as expected, doesn't converge approximately approaching the buckling load.
- setting "ITER, 2 (or above)" or "AUTO, whatever KSTEP" or "SEMI, whatever KSTEP", the simulation doesn't find any instability and provide a result very close to the linear solution.
In case of "AUTO" method I can accept this results since the stiffness matrix update can also not take place. However, in the "SEMI" method the stifness matrix update is forced every load step increment (and I included 25 load steps!!!). On the other hand, "ITER, 1" settings are definitely too heavy!!
Does anybody have an answer to this problem??
Regards!
I am trying to identify a structural instability with nonlinear solution (106) in a very simple model, a beam loaded by a force acting with a 2° angle of incidence (to generate the lateral instabililty). What happens is as follow:
- setting "ITER, 1" as "KMETHOD, KSTEP", the simulation, as expected, doesn't converge approximately approaching the buckling load.
- setting "ITER, 2 (or above)" or "AUTO, whatever KSTEP" or "SEMI, whatever KSTEP", the simulation doesn't find any instability and provide a result very close to the linear solution.
In case of "AUTO" method I can accept this results since the stiffness matrix update can also not take place. However, in the "SEMI" method the stifness matrix update is forced every load step increment (and I included 25 load steps!!!). On the other hand, "ITER, 1" settings are definitely too heavy!!
Does anybody have an answer to this problem??
Regards!