Instability on the top of the tower

[Anonymous]

Dear users,

I am working on a special tower. I have to use for the columns totally rigid links (Ux,y,z/ Rx,y,z are blocked) and on the top of the external and internal tower parts (see on the figure) I have instabilities (Instability 3) with the nodes (Rx, Ry, Rz).

I have tried to close the tops with steel plates (as shells) with a fine mesh. Also I have tried to apply releases on the instable end of the columns and fix Rx, Ry, Rz, but the same intabilities happened.

I dont really understand why it happens, especially when I fix these degree of freedoms of the columns on the top end, any idea?

The only thing that helped to run it, if I applied a support, in which Rx, Ry, Rz were fixed. But I think it is not appropriate.. Or is it acceptable (because these top parts will be "closed" with steel plates).

Thanks in advance! 

[Anonymous]

I can provide the load in pm.

[Artur.Kosakowski]

Mind that this type of instability may also be reported when you have large difference in the stiffness of the elements of a model for a perfectly stable structure. Try to apply a very small force or moment in the node and in the direction of the reported instability and check values of displacements and rotations there. 

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

Thank you Artur your response!

This is a quite slender structure, but I thought there is no big difference in the stiffness of the elements. I added a moment (because the instability came from it) at a node (Rx was unstable and I added Mx=1kNm) .

But what should I see from the displacement results? To make sure that I can ignore the instabilities?

These were the results:
Node:134/2 Ux=-0 Uy=-2 Uz=-0 [mm] | Rx=0.006 Ry=-0.000 Rz=0.000 [rad]

[Artur.Kosakowski · Accepted solution]

If there was a real lack of stability I'd expect RX to be much larger. I think that the instability is caused by having very short elements in the model.

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

However, so far I have merged these small elements and the instability is still the same. So it seems it was not 100% the reason. But in this case what else can be? I can send the model pm.

[Artur.Kosakowski]

If you want to play with your model try leave only the top part of your model and add supports at :cut" level. Calculate and see if you have any instability reported. If not, open the original model > cut it below the previous one and do the same. In this way you should be able to determine what causes the instability (you may also go from top to bottom if you like :-). ) 

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

I think now it isn't worth it, because this problem is relating to the solver. I have chaanged the solver to a multi-thread option and now, the place for the instability is just jumping within the structure. So I guess it relates to the stiffness differences because of the rigid links and somehow Robot has a problem to solve the matrices.