Rectangular Conc. Tanks WWTP Boundary Conditions 3

[crozroz0000]

Designing a WWTP with rect conc tanks:
It basically is 4 tanks together (4 cells). Each tanks is 100' x 67' x 18' tall. How should I model the foundation boundary constraints. I did: Vertical soil springs on interior. Around perimeter of group of tanks, I also did vertical subgrade springs, but two other directions were fixed (rotations not fixed).

Am I doing this right?
How thick should walls be? My 200k-ft horiz. corner moment gives 30-36" thick.

Plan view:
_____________
| | | 67'
|_____|_____|
| | | 67'
|_____|_____|
100' 100'

Adam

 

[GregLocock]

Is there not an outward bursting tendency in the outer walls? If so you have suppressed it with your constraints.

In the absence of better data (and with a project this size you need better data), I'd use shear cosntraints on each external wall, not normal ones, in the horizontal plane.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[vonlueke]

carpenters1son:  Assuming these tanks have an integrally-connected floor, then rather than constraining all nodes around the exterior perimeter in two horizontal directions, instead constrain only the midpoint of one exterior wall in two horizontal directions, and constrain the midpoint of the opposite exterior wall in only one horizontal direction parallel to the wall.

 

[crozroz0000]

Greg
Yes there is a bursting tendency. I'm not sure what this means: "you have suppressed it with your constraints"
I put no boundary conditions on the wall itself, just at the base of the wall. That is, I modeled that slab as about 1000 plate elements. Then the walls are made up of thousands of 5ft*5ft plates. I only put boundary conditions on the nodes at the bottom of the tank.

Regarding the "shear constraints," again I am not putting any boundary conditions on the walls as they are not attached to anything. Just the foundation where it "attaches" to the soil.

Adam

 

[crozroz0000]

vonlueke,

I think I get what your saying: I dont need as many translational boundary restaints as I have on the exterior. Currently I have one every 5ft or so around the perimeter ((100ft+100ft+67ft+67ft)/5ft = total 70 perimeter restaints).

Your saying I really only need 2 nodes in the entire model to be horizonal restrained. One should be restained in both directions...the other should be only restained parralell to the wall.

Am I understanding you?

Another question:
I am beginning to think that there should be no rotational restraints at the base of the exterior walls. I am thinking this because the wall plates are fixed to the foundation plates. Therefore there is no need to rotationally fix them.

Am I going in the right direction?

Last question:
Do I have the right idea about every node at the foundation level having vertical subgrade springs (about 56kips/inch)?

 

[vonlueke]

carpenters1son: Answers to your three questions. (1) Yes, you seem to have a good understanding of what I wrote. (2) I agree, and I agreed with where you said in your original post "rotations not fixed." (3) Yes, that sounds like a good approach.

 

[crozroz0000]

Thanks. That helps
Adam

 

[eduardotano]

I agree with vonlueke in trying to avoid to suppress horizontal displacements of the nodes of the base plate applying unneccessary horizontal constraints. This could reduce the reported values of the tension membranal stresses in the base plate. And these stress values are important to check concrete cracking in the plate.
Regards.

 

[torpen]

Hello,

The tanks have 3 rigid body movements on the soil : 2 translational movements and one rotational movement.
To fix them you need at least 3 constraints as explained by vonlueke.
The recipe for 3D Solid elements without rotational stiffness is as follows :
- Fix two translational dofs of one node. It lacks the rotational movement to fix.
- Choose one another node. The two nodes make a line.
- Fix the translational dof normal to this line to fix the rotation.

It is just a method to fix rigid body movements without adding constraints.

Does it reflect the reality?

Regards,

Torpen.

 

[crozroz0000]

Thanks Torpen and eduardotano

I understand much better now.

-Adam