Nonlinear problem - Unable to apply

[rgduarte]

Yes, i'm developing a sma damping system using tensegrity proprieties, I've made the model of the bracing element and it was successful, but when i try to insert it in the benchmark structure, the program doesn't run.

The element is this one:
https://www.dropbox.com/s/5pmtbzjk1cugx9l/TENSE%5E.spf

Four elements pinned at both ends with two SMA cables (tensegrity elements).

Thank you

[huffte]

rgduarte, I modified the end releases at the joints framing into n1tes and n4tes so that only one elfrm element has moment releases at each of those nodes. I believe this is needed for stability and since all other elements framing into those nodes have releases or are truss elements, the moment at the ends will still be zero. I also modified restraints at ntes2 and ntes3 to include rx restraint, which has the same effect at those nodes as does the removal of releases at the other two nodes.

Before the changes, the model would not run. After the changes, the model ran fine in 6 minutes 51 seconds.

I wonder if this modification or something similar might do the trick for you. I really believe the issue is that you have zero rotational restraint at the 4 nodes in question.

[fakharifar.mostafa]

I looked at your models. Why not instead of releasing members moment at both ends, inserting a link element at members intersection with low rotational strength?! May solve the error. In reality some friction exists even in pin joints. Just a thought though. Not an expert in SMA modelling.

[rgduarte]

huffte, thank you very much for your help. I've restrained the n1tes, n2tes, n3tes and n4tes at x+rx+rz+ry and maintained the moment releases at both ends. The program runs perfectly and i get null moment at both ends of every element of the bracing system.

fakharifar.mostafa, thank you, i'll insert link elements to check the results and compare to the ones reached with the model i have.

Thank you both for your time and helpful tips .

[rgduarte]

Sorry to bother you again , but I've made a model with bracing elements (tensegrity with sma cables) on the three floors of the structure and when i change the diameter of the SMA cables i get some strange (rather impossible) deformed structure, besides that, when the program actually runs (e.g. with diameter of 0.025m on the sma cables) i get random peaks of Force and displacement which contaminate the results i want.
This is Happening (picture):
https://www.dropbox.com/s/c4nhm696753ssxb/defestr.png

I'm using d=0.045m on the longer lenght SMA (SMAx) and d=0.030m on the shorter lenght SMA (SMALTx)


And the model i'm currently using:
https://www.dropbox.com/s/rkoru0o3yfbylry/SMA-3.spf

I think it must be something with the program settings (I've watched and applied the settings suggested on this video: https://www.youtube.com/watch?v=p6tAyoofDog

[seismosoft]

The convergence criteria that you are using are extremely loose. Change them to 1e-4 or 1e-5 both for translation and rotation.

SeismoSoft Support

[fakharifar.mostafa]

While altering the convergence criteria offered by SSoft solved the problem, however, your applied Time-History contains values as large as 8 and the curve Multiplier is 1 !? Is this frame supposed to be under such excitation, just asking! Changing the curve multiplier to 0.1, deformations fall under a reasonable range and analysis completed. To me, when checked your models so far, it is a numerical divergency associated with frame instability problem rather than SStruct bugs.

BTW, it is really an interesting problem, GOOD LUCK.