which element should be used?

from tongji · Post by **from tongji** » Sat Jul 07, 2007 3:29 pm

1. i decretize a beam to 10 segments, and use the nonlinearBeamColumn element.but the error occur,i am not sure if the element is adequate?

2. i searched in the forum and found that dispBeamColumn element is recommended,then if the fiber section can be used in dispBeamColumn element?and how many segments is adequate for a beam?

3.when the dispBeamColumm element is used ,the following warning is got:
WARNING BandGenLinLapackSolver::solve<>-LAPACK routine returned 1
WARNING NewtonRaphson::solveCurrentStep<>-the LinearSysOfEqn failed in solve<>
StaticAnalysis::analyze<>-the Algorithm failed at iteration::0 with domain at load factor 0.1
Opensees >analyze failed,returned:-3 error flag

pejman_opensees · Post by **pejman_opensees** » Sat Jul 07, 2007 10:56 pm

If you study EERC report written by taucer & spacone and filippou, you will undrestand that 5 integration points are quite enough for capturing any distributed inelasticity and more thatn this is redundant. I think there is a controler in source code in which using more than certain amount of integration point is stopped and if there is, it seems to be quite right. the philosophy of force beam column existance is to model a structural member with only one element for which plasticity distribution and accurate definition of curvature is best maintained.

from tongji · Post by **from tongji** » Sun Jul 08, 2007 4:35 pm

in fact ,if i use i and j node to build a beam,Opensees works well.but when i use the nodes from i0,i1 ,......,i9 to build the same beam(decretize the beam to 10 segment), the error occured,why?i just use 10 nodes to build the same beam.THANK YOU!

pejman_opensees · Post by **pejman_opensees** » Sun Jul 08, 2007 10:19 pm

I have used very small force beam column element but I have not had any problem.I dont think it is dependent on length of the element and it definitely shouldnt be.
Send your problem to this address if u want: pejman_alanjari@yahoo.com

from tongji · Post by **from tongji** » Sun Jul 08, 2007 11:01 pm

thank you, Pejman! the problem have been resolved!

luoyunbiao · Post by **luoyunbiao** » Fri Jul 13, 2007 5:26 pm

pejman_opensees wrote:If you study EERC report written by taucer & spacone and filippou, you will undrestand that 5 integration points are quite enough for capturing any distributed inelasticity and more thatn this is redundant. I think there is a controler in source code in which using more than certain amount of integration point is stopped and if there is, it seems to be quite right. the philosophy of force beam column existance is to model a structural member with only one element for which plasticity distribution and accurate definition of curvature is best maintained.

I wanna study the EERC report too,but download the report require a EERC membership.

pejman_opensees · Post by **pejman_opensees** » Sat Jul 14, 2007 6:29 am

I will send it to you if you want. send me a blank email to following address I reply your email with an attachment.
pejman_alanjari@yahoo.com