Elastic Foundation

[matos11]

Hello.

Can anyone help me with my problem, please? I am working on my project and the task is modal analysis of tank on elastic foundation in Ansys APDL. Elastic foundation is represented by Winkler's model.
For elastic stiffness foundation I used element type SURF154 and add real constant.
This element I surfed to the bottom of the tank. Then I set boundary condition to the base of the tank - fixed displacement to all base without displacement in direction UZ.

Eigenfrequencies of this tank are much lower (1st freq. = 11 Hz) than the tank resting on the rigid foundation (1st freq.=120 Hz). I think that is not correct because I think that eigenfrequencies of the tank on elastic foundation should be higher than tanks on rigid foudnation. I think that I set wrong boundary conditions which should represent Winkler's model.

Could anyone help me?

Thanks.

 

[jhardy1]

I would expect lower frequencies for the tank on elastic foundation than for the rigid foundation:
Frequency = (k/m)^(0.5)
And when you use a finite (less than infinite) stiffness, you have reduced k, and therefore the Frequency is reduced.

As a check of sensible solution magnitude, you should see a simple global vertical "bounce" mode, with a frequency which correlates to (k/m)^(0.5) where k is the total vertical stiffness of the foundation, and m is the total mass of the tank and contents. This mode would not be evident in the rigid foundation model, because k is infinite for this case.

Conversely, if you use the Winkler model and set the coefficient of subgrade reaction to a very high value, it should trend back to the rigid foundation solution.

You might want to think about the other base constraints while you're looking at it - fully fixing all the other DOF at every node might over-restrain the base of the tank.

Hope this helps!

http://julianh72.blogspot.com