Termination of Analysis-2

[taneryilmaz]

Hi to everyone,

I've been dealing with 3-D spine models of bridges which include fiber elements for columns, nonlinear soil springs, and a couple of nonlinear elements for abutment or hinge responses. In one of my models, I've continuously faced with analysis termination (it is the singular matrix error; the error message changes depending on the solver used) so I made the model in its simplest shape and step by step introduced new inelastic elements to find out where the problem is coming from. I previously thought the problem was all about my foundation elements (http://opensees.berkeley.edu/community/ ... =2&t=60518), however now I put the blame on the fiber elements I use for the columns.

For the case when I make all of the abutment and hinge elements as elastic, soil spring elements assigned with pySimple1 material (nonlinear), and when columns are ELASTIC, time-history analysis is completed without error. But as soon as I make the columns modeled with displacement-based fiber elements, the analysis gets terminated. Bent columns have rectangular (nearly wall-type) sections, and the confinement effect is too low making core concrete having approximately the same strength as the cover concrete. Within my trials, there happens no difference whether I use Concrete02 or Concrete07; and number of integration points and number of fibers made no effect.

I wonder if anyone would like to advise me in working out my problem.

Thanks in advance.
Taner.

[fmk]

if elastic elements work. switch to dispBeamColumn elements with elastic sections. check yu get the same resuilts. then go fiber using elastic materials. check you get similar results. then go nonlinear materials.

[taneryilmaz]

Dear FMK,
I proceeded as you recommended: first fully elastic elements for columns, then displacement based nonlinear elements with elastic sections, then fiber sections with elastic materials. At all of these steps, I haven't got any analysis termination. But at the moment I assign Concrete07 or Concrete02 to the concrete fibers, the analysis fails. Would you give me some more advice for helping me solve the problem?