transformation issue

bkosbab · Post by **bkosbab** » Thu May 15, 2008 7:17 am

I've built a 2D model, and am getting results using Linear transformations for all elements save 2, which are pinned-base columns, and use a PDelta transformation. When I try to replace the PDelta transformations for Corotational transformations, I get a repeating error message:

CorotCrdTransf2d::getGlobalResistingForce: affect of Po not implemented yet.using zero value

Any ideas/thoughts? Thanks in advance

pejman_opensees · Post by **pejman_opensees** » Thu May 15, 2008 10:09 am

What is the exact problem? plz write some codes.

bkosbab · Post by **bkosbab** » Fri May 16, 2008 4:34 am

unfortunately, my model has gotten a bit large and cumbersome to post here.

The issue is that, during each load step (static or transient) when using the Corotational transformation for ANY element in the model, I get the error/warning message posted above. This issue does not appear when using a PDelta transformation. The model is of a container crane, which is essentially a pinned-base portal frame supporting a large superstructure.

silvia · Post by **silvia** » Fri May 16, 2008 5:33 am

there is some element that does not accept the corotational transformation.

are you using opensees 2.0?

bkosbab · Post by **bkosbab** » Fri May 16, 2008 6:04 am

Indeed, OpenSees 2.0.

I am using nonlinearBeamColumn elements for the members to which I would like to apply a corotational tranformation. However, the same error/warning occurs when I attempt to use elasticBeamColumn elements as well.

silvia · Post by **silvia** » Fri May 16, 2008 8:11 am

while we wait for Frank to return and look at the OpenSees code, please double and triple-check your model.
start with something simple and see exactly what is causing the problem.

pejman_opensees · Post by **pejman_opensees** » Fri May 16, 2008 11:13 am

Is your model 3D and are u using equalDOF commands in your model? Did u take care of VECXZ direction?

bkosbab · Post by **bkosbab** » Fri May 16, 2008 12:37 pm

2D model, so I have no vecxz defined.

thanks for the suggestions. I'll take a closer look at this on Monday, and build a fresh, simple model to find the underlying issue. I'll post my findings, if I get any

silvia · Post by **silvia** » Fri May 16, 2008 12:51 pm

great,
i do think that it is something within the model.

bkosbab · Post by **bkosbab** » Mon May 19, 2008 9:50 am

Well, I have found the issue. Unfortunately, it isn't just a dumb mistake on my part.

It appears that the corotational transformation has issues when a uniform load is defined along the the axial direction of an element. This can easily be seen in Ex4.Portal2D.build.ElasticElement.tcl, if a uniform axial load is defined along one of the columns, and the corotational transformation is used for the columns.

Is there a way around this issue besides defining point loads at nodes along a column length? My model's columns are quite heavy.

silvia · Post by **silvia** » Mon May 19, 2008 11:19 am

interesting problem.
if it is such a big deal to have a distributed p-delta load, you might want to break the columne element down to 2-3 element (not more than that, though)

also, please contact Frank when he returns (after may 26th) with an example of your problem

ca493 · Post by **ca493** » Fri May 23, 2008 11:48 am

just as a note, i've experienced the same phenomenon; however, i was running into Corotational errors for distributed loads along beams as well as columns. They seem to wrok fine as concentrated loads now.