Spatial Truss - Instability

[t.sautierr]

When I change the structure from 2D frame to 3D I get instabilities and Ican't find why .... please help.

I must be a silly thing but, I don't find ...

[Mirko.Jurcevic]

I don't get any instabilities, but your "Appui Uy" has "fff" definition. Like there is no support.

[t.sautierr]

Mirko,

switch the structure type of the model to "spatial frame" or "spatial truss" I don't what is the label in english and then you will have the instability.

Appui Uy is fff in the current model because it set for structure type 2D frame so no possibility to go out of the plane and on ly 3 dof, this support is useful when it is studied as a 3D frame and then the code support will be ffffbf, just Uy blocked.

thx for your help.

[Artur.Kosakowski]

I can't see any issue with the model after switching to the 3D frame either. Change to 3D truss has got little sense as all connections among bars are then pinned.

[Mirko.Jurcevic]

Yes, I got some warnings and an error no. 3060

But that 3D truss has free rotation, so it's mechanism. Artur is right.

[t.sautierr]

Actually i wanted to make a buckling analysis to have a critical load factor for buckling that i obviously expect to be for an out of plane shape, that's why i shifted to 3d frame in order to consider the third direction. But even in 3d i don't see where there is a mechanism, no posible translation, no rotation in plan and rotation out of plane is blocked by the pines on the central bar???

[Mirko.Jurcevic]

And what about torsion? Rotation around the longitudinal axis.

I agree with everything you said but you forgot about 6th degree of freedom  😉

[Rafal.Gaweda]

3D Frame - no instability

[t.sautierr]

I took 3D truss and got the instability -> pictures.

With 3D portal, like you Rafal and don't get any instability.

Actually I've never payed attention to the difference between 3D truss and 3D portal but it seems that 3D Truss, you don't have all the dof? for example you cannot specify anything for rotation RX, RY, RZ, but the instability is on dof Uz ...?

Mirko in the last model, you are right in 3D I forgot torsion,the ends bars where free to rotate along their axis, thanks 😉

but know in 3D truss, I can't tick rotation dof maybe this means this ignore the problem in this structure type.

thanks for your help guys. 

[Rafal.Gaweda]

I took 3D truss and got the instability -> pictures.

It must be like that.

Actually I've never payed attention to the difference between 3D truss and 3D portal but it seems that 3D Truss, you don't have all the dof? 

Every node has only 3 dof (UX UY UZ) in 3D Truss structure type that's why the model after changing structure type to 3D Trussn becomes a mechanism.

 this ignore the problem in this structure type.

You can not ignore isntabilities: 1. wrong results , 2. error 3060 in modal \ buckling case.

[t.sautierr]

3D portal = 6 dof and 3D truss = 3dof, right?

3D portal = no instability and 3D truss = instability -> with less dof I have got instability? it sounds weird to me.

If you have got 5mn in the day, can you just show me where is the mecanism for 3D truss? It still don't see what is unstable in Uz direction.

thx

[Rafal.Gaweda]

3D portal = 6 dof and 3D truss = 3dof, right?

Right

3D portal = no instability and 3D truss = instability -> with less dof I have got instability? it sounds weird to me.

See below

[Artur.Kosakowski]

---

3D portal = 6 dof and 3D truss = 3dof, right?

3D portal = no instability and 3D truss = instability -> with less dof I have got instability? it sounds weird to me.

---

Please check the messages 4 and 5 from this subject.

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.

[t.sautierr]

Rafal and Artur;

I've seen your explanations but maybe I'm a bit stubborn but I don't see why there are instabilities in the attached model :

- I tried but 3D Portal : no instability as long as I block UX UY,UZ + Rz at one of the two supports 

- 3D Frame : instability 1 Uz at the node connecting the inclined bar to the horizontal ones, why? no release, nothing else than 3 bars and 2 supports.

where is this dam...!%{ mechanism? 🙂

[Artur.Kosakowski · Accepted solution]

The model you attached is actually the 3D Truss (visible on the top of the attached picture). At the bottom you can see the equivalent model defined as a frame. Mind the releases. I hope this is easier to understand now.

In other words for trusses you need to think about a structure built from number of triangles to avoid instabilities.

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.

[t.sautierr]

gotcha! I was suspecting that ->  tricky because you cannot see the "hidden" release or so to say that the connection is assumed to be free in rotation at all nodes except that it called a truss ... and that it is the definition but for big model if someone change the structure type it can be diffucult to see it.... any warning could be welcome 🙂

[Artur.Kosakowski]

Out of curiosity: I can understand the reason why to change the model type from a truss to a frame but what is the reason to go in the 'opposite' direction? You need to consider also the fact that for a truss model you can apply loads to nodes only.

[t.sautierr]

Actually, no special purpose, all started because I first modeled this using a 2D frame, then I realized that I had to study out of plane buckling -> switch to 3D and I first changed to 3D truss (mistake) and not 3D frame, I got the instabilities and you know what follows ...

 You need to consider also the fact that for a truss model you can apply loads to nodes only

Why ? I agree taht this type of structure is "shaped" for this king of loading but it is not always the case ... for example Own Weight of the truss is a uniform loading and not a point load

[Artur.Kosakowski]

Why ? I agree taht this type of structure is "shaped" for this king of loading but it is not always the case ... for example Own Weight of the truss is a uniform loading and not a point load

Because this is the 'nature' of such structure type. As there is no rotational d.o.fs a load applied along a bar cannot 'generate' bending. For trusses self weight is applied as nodal forces at the origin and end nodes of a bar. Similar situation is when you compare a plate model with a shell one. For the prior you cannot have any horizontal loads.

[t.sautierr]

Oki doki.