Zero Pivot and Numerical singularity problems

[USBdroid]

Hi everybody,

I have problems in analyzing my model; I am trying to model a mast supported by cables. So I modelled every cable, the mast and spreaders as independent parts with beam elements. I connected everything using MPC constraints (pin for cable to mast intersection, and tie for spreaders to mast intersection). To model the cables I created a material tension only. I constrained all the cables and the bottom of the mast where they meet the ground (the boat in this case). In order to test the model, at the mast head I applied a load of 1000N horizontally in a static step (including large def and time step size I reduced it to as low as 0,05)
But then the analysis was aborted because the model does not converge. I tried already to remove the no compression feature to identify whether or not it was the cause but seemed that not.
Another question I wonder, might be stupid but well, when a beam is partitioned, ABAQUS still represents it (the original beam) as continuous? Beacuse the mast is tapered and to represent the taper I partitioned the mast and assigned discrete cross sections. But hopefully the connections from beam to beam are still alright? no?
I get the following warning messages, which from my little research on the net seem quite frequent;
-numerical singularities at nodes... 5 (or 6) DOF
-zero pivot when processing node...
-negative eigenvalues

The zero pivot nodes, some were identified as overconstraints, some not. All the nodes causing problems are located in the cables or at the intersections.
I did not find how to deal with the overconstraints.
Concerning the numerical singularities, by searching on the net, I found that it might be due to the model not being constrained enough... but well, it seems to me that it is!

So I attach here the python script (made in ABAQUS 6.10) (and the message file I can add if anyone needs) for anyone who can give a bit of his time for my problem and wants to have a look. What is in comment at the end of the script is what I changed at the last minute.

Thanks in advance

G

 

[USBdroid]

I add here a picture of the model to help.

I now solved the numerical singularities by replacing the beam element in the wires by trusses and the zero pivot warnings by being smarter when I specify MPC constraint. As well what help was to add stabilisation by damping factor of previous step.

But now I don't seem to manage to make the model converge;
I tried to go from the basics and remove first the diagonal arrangement, it did not work. I then tried the mast alone, and as long as I did not tick large deformation, it converged.

Any help?