FEA ACI 350 with non-conventional tank geometries

[jacasalr]

Hello everybody!

I did not write too much in the Civil Engineering forum since I am not a Civil Engineer, however, I am needing some help on the analysis of concrete (actually, it is polymer concrete, but let's imagine is conventional concrete to be able to use the standard) tanks.

I was doing some research and was able to find that ACI 350, especifically, ACI 350-06 and ACI 350.3, gives the procedure to analyze cylindrical and rectangular tanks in operating and seismic conditions, respectivelly. However, there are many doubts I have since my problem is different and I will try to summarize them all next:

1) The geometry of my tank is not conventional, but it is "kind of close" to a rectangular tank, as you can see in the next image. The problem here is that the analytical models within the standard are related to standard tank geometries so I should not take this model for my tank, since errors may arise.

https://files.engineering.com/getfi...-4544-9bd8-223cfd981eb4&file=Tank_Example.jpg

Maybe it is not properly shown in this image, but internally the tank has very big radius as the external ones. This and also the presence of the 4 supporting legs, make this geometry unconventional.

2) I am using a Finite Element software, Femap with NX Nastran, using solid elements (in this example I can use both tet and hex elements, but normally I am just able to use tet elements depending on the tank design). The problem here is that the ACI standard is based on an analytical method which relates everything to bending moments and shear forces; however, the FEA study takes as input loads and gives me stresses when using solid elements.

I have been thinking about it and I have come to my mind the next solutions:

1) Use the ACI standard as it was a conventional rectangular tank in order to estimate the loads, both operational and seismic loads. Then, apply these loads to the tank walls and bottom to finally obtain stresses and compare the maximum principal stresses with the tensile (bending) strength of the material and the minimum principal stresses with the compressive strength.

2) Use the ACI standard as it was a conventional rectangular tank in order to estimate the loads, both operational and seismic loads, but in this case just taking the operational loads as described in the standard and the seismic loads as follow:

For the hydrodynamic force, I choose "m" as the total mass of stored liquid and the acceleration as the seismic horizontal acceleration multiplied times the factors established in the standard (C,I,R,S,etc.). Then I estimate the hydrodynamic for by F=m.a and apply it to the walls.

Finally, apply these loads to the tank walls and bottom to finally obtain stresses and compare the maximum principal stresses with the tensile (bending) strength of the material and the minimum principal stresses with the compressive strength.

What could you guys recommend me on this case? What do you think about my possible ways to solve it.

Thanks a lot in advane.

 

[HTURKAK]

.......not properly shown in this image, but internally the tank has very big radius as the external ones. This and also the presence of the 4 supporting legs,...

What are the dimensions of the tank and how the legs connected on the supporting structure? Is this ground supported tank?. Pls provide more details to get more valuable comments..

 

[jacasalr]

Hi HTURKAK,

Thanks for your reply! The tank is simple supported to the ground at its four support legs. There is no other mechanism to hold the tank.

Inner dimensions of this tank are: 5 meters length; 2 meters heigth, 1 meter width; internal and external radious equal to 0,25 m. The distance from the bottom surface of the supporting legs to the external bottom of the tank is 0,3 m.

My question is more related to how to approach the analysis using a FEA software and 3D elements, since in this way I can get stress distributions such as the principal stresses. Normally, I set the operational loads such as hydrostatic and other vertical loads, and then analyze the model and get the stress distribution; then I take the maximum principal stress value and compare it with the flexural strength of my material (polymer concrete). I also did this comparison between the minimum principal stress and the compressive strength of the material.

However, when I look at the ACI 350 standard the approach is quite different and I have not been able to find a right approach to implement the ACI 350 criteria and using a FEA software.

Thanks!