Learning Forum

[bayraktaroglu]

I am modeling an offshore jacket type gravity foundation using a 20 node user-defined coupled element. 

First, I have applied the gravitational load using ACEL command and calculated the displacements caused by the gravitational force. While doing that I have turned off the transient effects using TIMINT,OFF command so that I was able to find the consolidations under the self-weight of both structure and soil itself.

So far I have no problem. Since the model runs without transient effects no excess pore water accumulations are observed which makes sense. But the problem is immediately after continuing the analysis by turning on the transient effects (using TIMINT,ON) excess pore water accumulations starts to build up even though I do not apply any additional force to the model.

Normally what I expect is, since I have performed an analysis using TIMINT,OFF the acceleration and velocities at each node should be equal to the zero, in other words, the displacements that I get at the final step of the TIMINT,OFF should be the final displacements under gravitational loads (displacements should reach an equilibrium). 

I have attached my macro file, it would be great if you could help with this.

In advance thank you for your help.

[mrife]

bayraktaroglu

Please see the following help section:

Ansys Help -> Mechanical APDL -> Theory Reference -> Chapter 15.2 on Transient analysis.  

KBC (load application method) defaults to ramped when the analysis type is transient and the time effects are turned off.  This results in the nodes having a a non-zero 'velocity' at the end of the load step.  Choose a far larger time value for the first step to reduce this initial non-zero velocity. Or use step applied loading, kbc,1 , instead of ramped loading.

Mike   

[bayraktaroglu]

Mike,

Thank you so much for your reply.

I will try it. As you said the KBC defaults to zero (ramped).

The thing that I don't understand is while TIMINT is OFF (no transient effect) the deformation becomes independent of time. For example, when I model the displacements under self-weight (gravitational load) and perform two successive solving let say at TIME,1 and TIME,3 the displacements at time=3 becomes equal to displacements at time=1 (No change between time 1 and 3). What I am understanding from this is, since TIMINT is OFF at the end of the time=1 load step the velocity of each node becomes zero and that is why the system actually reaches static equilibrium at time=1 so that no change occurs at time=3. But if I add NSUBST for TIME,3 then result differs from time=1 which I don't understand reason at all. Same forces apply at the beginning of the TIME,3 and at the final substep so what kind of ramping occurs there?

Thanks

[mrife]

Bayraktaroglu

Please read that help section - I know that there should not really be nodal velocities in a static analysis.  But we use the static 'pseudo velocity' as an way to define the initial condition of the time transient.  Another way to have an initial displacement but zero initial velocity is to use two load steps with the time integration off.  Ramp the load in the first load step, then hold it constant in the second.  This will zero out the pseudo velocity and not take much more solve time.  

Mike

[bayraktaroglu]

Thank you so much for your time.

As you suggested I have selected a larger time value for the first step and the problem is solved.  I will read the manual carefully and try other options as well to further improve myself.