rigid floor

Post Reply
Yulu Liu
Posts: 22
Joined: Fri Nov 15, 2019 6:31 am

rigid floor

Post by Yulu Liu » Tue Dec 10, 2019 3:37 pm

I build a high frame structure with beam and column, but without floorslap. I need each floor to move only in x or y direction under horizontal load, without vertical moving. Maybe I need to set the floor rigid, so the floor can only move horizontally. Is there this configuration?
Attachments
269.png
269.png (610.18 KiB) Viewed 744 times

MassimoPetracca
Posts: 37
Joined: Wed Jul 24, 2019 11:10 am

Re: rigid floor

Post by MassimoPetracca » Tue Dec 10, 2019 6:11 pm

Of course you can do it, but you need to explain better what you want to do. From your explaination it seems you what your floors to move rigidly only in X or Y direction but be fixed in Z... But I think what you really want to do is the rigidDiaphragm (the floor keeps its shape in the plane, but the beams of the floor can bend in the out-of-plane direction).

If you want to do this in STKO you need to follow these steps for each floor:
  1. draw a master-node representing the center-of-mass of the floor (optional, you can also use an existing node if you have one more or less near the center-of-mass)
  2. fix it in the Uz-Rx-Ry DOFs (if you created a new hanging node as in step 1)
  3. crate an interaction (master = master-node, slaves = all nodes of the floor)
  4. assign a constraints.mp.rigidDiaphragm to the interactions, choosing $perpDirn = 3 (i.e. the normal to the plane is the Z axis)

Yulu Liu
Posts: 22
Joined: Fri Nov 15, 2019 6:31 am

Re: rigid floor

Post by Yulu Liu » Wed Dec 11, 2019 7:25 am

Yeah, you get my words. Thank you.
And after creating a new hanging master-node, and fixing it in the Uz-Rx-Ry DOFs, do I need to merge it with the existed merged model?

Yulu Liu
Posts: 22
Joined: Fri Nov 15, 2019 6:31 am

Re: rigid floor

Post by Yulu Liu » Wed Dec 11, 2019 8:35 am

I find that when the master-node is out of the model it can't run correctly, but when the master-node is one node of the model, it run well.
Attachments
d478637beca05c4f6fd96de9a7e58cb.png
d478637beca05c4f6fd96de9a7e58cb.png (191.97 KiB) Viewed 737 times

Yulu Liu
Posts: 22
Joined: Fri Nov 15, 2019 6:31 am

Re: rigid floor

Post by Yulu Liu » Thu Dec 12, 2019 3:12 pm

When I locate the the mass center, I find that it is out of the model.
So I create a new hanging master-node, and fix it in the Uz-Rx-Ry DOFs, then do I need to merge it with the existed merged model?

MassimoPetracca
Posts: 37
Joined: Wed Jul 24, 2019 11:10 am

Re: rigid floor

Post by MassimoPetracca » Thu Dec 12, 2019 5:00 pm

No you don't need to merge it, just connect it with the slave nodes with an interaction and assign a rigidDiaphragm to the interaction

Yulu Liu
Posts: 22
Joined: Fri Nov 15, 2019 6:31 am

Re: rigid floor

Post by Yulu Liu » Fri Dec 13, 2019 2:04 pm

When I set the mass center(master-node) of the rigid floor at the existed node of the model, the analysis result is shown as photo 1, which is correct.
But when I set the mass center(master-node) at the new added hanging node out of the model, it produces photo 2, where there are incorrect results and some errors.
What is the matter? It is strange.
Attachments
2.png
2.png (144.92 KiB) Viewed 721 times
1.png
1.png (89.24 KiB) Viewed 721 times

MassimoPetracca
Posts: 37
Joined: Wed Jul 24, 2019 11:10 am

Re: rigid floor

Post by MassimoPetracca » Fri Dec 13, 2019 6:41 pm

Looking at the warings
ProfileSPDLinDirectSolver::solve() - aii < 0
it means that some diagonal elements in the stiffness matrix are 0, so the matrix is not positive definite.
If you use the rigidDiaphragm with direction Z, then it links the dofs Ux,Uy,Rz to those of the slave nodes, but the other 3 dofs are free. You said you fixed them, but are you sure you added those fix constraints in a constraintPattern?

If this is not the problem send us the input file and we will have a look

Post Reply