time history analysis of water tank 1

[SAGG]

Hi all

I am doing seismic analysis of a water tank. I used fluid 80 and shell63 in ansys 11. i coupled nodes of fluid and shell elements at pheripery of tank. This is done by rotataing nodal coordinate system.
1. I have couples nodes at the bottom plate and fluid elements.
2. Now I want to do seismic analysis. please help me how to go about.
3. I have acceleration time history in x,y z direction. This I read in a table.
4. please guide me how to apply accelerations at bottom node and how to go about solution procedure.
5. whether giving acceleraion on entire structure and applying at nodes mean the same?

6. I have 3000 points of accelearation and time
whether following steps are right?

*do,i,1,3001,1
cmsel,s,bottom,,
d,bottom,accx,xacetl(i,1,3)
allsel
time,xacetl(i,0,3)
lswrite,i
*enddo

7. i get an error saying that acceleration is not applied at rotated nodes. how to solve it

please help me

with thanks
SAGG

 

[juzz]

Hi SAGG,

1. Block all nodes at bottom of tank (shell only)and try to use ACEL command instead of prescribing acceleration of nodes. This will apply linear acceleration to entire structure along directions of Global Cartesian System axes.
2. Be aware that doing seismic analysis of vertical tanks containing fluid could be a very difficult task. I strongly recommend to perform first a static analysis with gravity only (is mandatory to have this as initial step in a time-history analysis) and check if pressure distribution on the tank shell is ok. After that perform a modal analysis of the structure and check the results (frequencies, participating masses) against values estimated by analytical formulas. For reference for analytical estimations see following document:

http://www.iitk.ac.in/nicee/IITK-GSDMA/EQ04.pdf

Success!

Juzz

 

[Drej]

You can't apply acceleration as a time history variable. You'll need to convert these to displacements and apply using a table (there are other methods, but not as efficient) as you've stated with the 'd' command - you need to reference the table name in the 'd' command using %table_name%. See the manual for the 'd' command.

A word of warning: using ACEL is not the same as applying your acceleration time history - ACEL is a different loading altogether.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[juzz]

Hi, Drej!

Could you please detail a little bit your word of warning?
I mean why is ACEL not suitable to use (as applied acceleration at the base of the tank)? It is not the tank under the effect of inertial loading?

Thanxs!

Juzz

 

[Drej]

Juzz,

ACEL is a whole-body inertial load. If the structure is 'rigid' then this is fine. In the rigid body case, you can apply an acceleration time history at the base or you can apply ACEL as a time history and it will be the same, since the whole structure will see the same acceleration. If the structure responds elastically then ACEL will apply the inertial load to the entire body, which is not the same as applying a time history of local acceleration base motions.

Cheers.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com