Plastic (LLDPE)......FEA

[hello1581]

Hello all,

I am conducting FEA analysis for tank (LMDPE material) with the help of ansys software. Dimension of my tank is app. 5ft X 4ft. I am applying pressure gredient (0 to 6 psi linearly from top of tank to bottom) form the inside of the tank it shows me 10 inch displacement. However, we alredy installed this tank on our machine but we didn't have displacement more than 2". I don't know what is wrong on my analysis and why i am getting to much displacement.

I am trying NLGEOM option ON any OFF (for small tank or less pressure gradient value) and it will give huge difference in displacement. I know for big tank i should NLGEOM option ON so it will give me lesser value of displacement but when i am putting this option ON, ansys will not converge this and stop in between the analysis.

Some experts gave me advice that you need exact stree-strain curve values and enter in to ansys then it will converge and try this way also but it is not converge.

Can anyone give me some advice to me. My approach seems right on not.

One more thing do need put young's modulus value in ansys or Flexural modulus because i hard that if you get lot of bending then use flexural modulus value.

Please help me or suggest me how to approch to solve this problem. I am new to that analysis. My boundary condition is right.


Thanks
Sanjay

 

[Transient1]

You state your boundary conditions are correct. What are they?

Can you describe the tank better, what is the thickness of the walls?

Have you tried refining your mesh?

Are you sure your pressure gradient is correct?

Is this simply hydrostatic pressure?

Is the liquid adding structural support somehow (i.e. is it in a situation where it cannot move away from the tank deformation)?

 

[hello1581]

1.I have four mounting plate in four corner of the tank and i am constraining (fixing) all of them.

2. Tank is 62 inch X 60 inch and wall thickness is 0.3 inch. It is just like cuboid or square.

3. We alredy refined our mesh.

4.I calclate my pressure gradient theoretically as well it is right.

5. Yup

6. I am not sure about last question.

I alredy run this analysis.
1. Linear Geom* and Linear material- To much deflection.
2. Linear Geom and Non-linear* material- To much deflection
3. Non-linear Geom* and Linear material- not converging
4. Non-linear Geom and Non-linear material- not converging

Liner Geom - NOLGEOM, Off
Non Liner Gemo- NOLGEOM,On
Non-liner material - Enter the exact stress-strain data.

 

[btrueblood]

Sounds like you should increase your modulus in the analysis by a factor of 5 (which would worry me too), and/or double check your wall thickness in the model (are you using shell elements?)