dispbeamcolumn3D and co-rotational transformation

[pankotso]

Hi vesna and fmk, I would like to ask what is the exact reason why the 3D dispbeamcolumn element does not work with corotational transformation for element loads. Is it because of a follower force's problem? I can't see why it does not work since a distributed load just imposes fixed end forces at the element's nodes. Thanks in advance for your response, it will help clear things and find a way of overcoming this limitation. PK, UoE

[vesna]

In general, eleLoads do not work when the corotational formulation is used. It is just that the current formulation of corotational transformation in OpenSees does not account for eleLoad.

[pankotso]

Hi vesna, thanks for the comment. From some scripts that I 've checked it seems that corotational works ok for a 2D elemental load case but not for 3D ones. I have tried your answer and I can see that its working, the thing that I am not seeing is why is there such a problem from the beginning since an elemental load is just the same with equivalent nodal loads. This happens in the addload method of the dispbeamcolumn3D element.

if (type == LOAD_TAG_Beam3dUniformLoad) {
double wy = data(0)*loadFactor; // Transverse
double wz = data(1)*loadFactor; // Transverse
double wx = data(2)*loadFactor; // Axial (+ve from node I to J)

double Vy = 0.5*wy*L;
double Mz = Vy*L/6.0; // wy*L*L/12
double Vz = 0.5*wz*L;
double My = Vz*L/6.0; // wz*L*L/12
double P = wx*L;

// Reactions in basic system
p0[0] -= P;
p0[1] -= Vy;
p0[2] -= Vy;
p0[3] -= Vz;
p0[4] -= Vz;

// Fixed end forces in basic system
q0[0] -= 0.5*P;
q0[1] -= Mz;
q0[2] += Mz;
q0[3] += My;
q0[4] -= My;

Does it have to do with coordinate transformation of the elemental load when the element is not straight anymore? At that condition the elemental load wont be constant during the analysis. Thanks. PK

[vesna]

You are right. I will ask Frank to add few more lines to Corotational 3D formulation so it accounts for element loads.

[seu]

I intend to use a 3D model with geomTransf corotational and eleLoad on nonlinearBeamColumns, but the result is almost 0.
Must I manually calculate equivalent load at element ends??

Will the opensees source code solve this problem later? This is a big trouble indeed.

Thanks for your any reply.

[vesna]

We are working on this. In the meanwhile use Pdelta for 3D elements and element load on it.

We looked at the source code and realized that equivalent load at the nodes will not work, as 3D corotational formulation was coded differently from 2D. DO NOT use corotational with 3D elements that have element load.

[pankotso]

Hi vesna, are there any updates on this matter on the co-rotational formulation for 3D elements under elemental loads. Did you find where the problem actually comes from? Thanks! PK, UoE

[seu]

Up, also very concerned with this problem. Puts thanks to vesna.

[vesna]

This is not implemented yet. I will let you know after we add this to opensees.

[pankotso]

Hi Vesna, any news on this 3D corotational matter? this would be very useful!

[vesna]

It is not implemented yet.