Learning Forum

[Pritish Sadhale]

Greetings all! I am having a problem related to the deactivation of displacement BC in my analysis. 

I have two identical and simple models as follows:

         

                                        Force Model                                                                                     Acceleration Model

Both the models have two remote displacement BCs at both ends, with rotation around Z axis free. The only difference is that the force model has a 15 pound force acting and the acceleration model has a 15 pound point mass, and an acceleration BC. I am more interested in the point mass and acceleration model, but I created the force model just for a comparison.

I want to ramp the load in the first load step, and then free one of the ends in the second load step. This is easily achievable and is successful if I set up both the loadsteps in a single analysis. However, it is my intention to run the models upto the 1st loadstep, save the result files and then, use these result files as a starting point for my next loadstep in an independent project. So it won't be required to solve the first load step again and again. My motivation behind doing this is that I have a bigger and more complex model that has a several load cases. But in all the load cases, my first load step is the same. The loads start to vary only after the second load step and thus, it is computationally efficient to solve the first load step once, and then using it as a starting point for all other load cases. There will also be of a great benefit in debugging as there will be no need to wait till the first load step solves everytime when a new load case is being solved. As my bigger model takes 3-4 hours to solve this first load step, this approach is valuable in terms of computational effort and time. But, I want to get it right in the smaller and simplified model first.

Moving to the issue, I ran both the models with the only difference being the force BC and acceleration BC. I had result files upto the first load step saved separately, which I then used as starting points for the second load steps. In the model with the force BC, the displacement BC is getting deactivated as intended. But in the acceleration model, it is not. I checked both ds.dat and solve.out files for both the models. In ds.dat file, the command "ddel" can be seen written in the same manner, indicating that the solver is trying to delete the BC at the required load step. However, in the solve.out files, the BC is getting deactivated in the force model but remains active in the acceleration model. The screenshots are as follows:

............ds.dat - force model

 ..........................solve.out - force model


.............ds.dat - acceleration model

 

............................solve.out - acceleration model



As it can be clearly seen, the "Number of Specified Constraints Deleted" line in the solve.out differs for both models. My main model contains same point mass and acceleration system and I faced the same issue when I tried to solve it there. The BC is not getting deactivated for some reason after restarting. 

I would appreciate it if somebody could guide me on what may be going wrong, or any command that I need to modify in order to make this work. I am happy to share more details if required. 

Thank you for taking the time to read this. 

Regards,
Pritish

[Gary Stofan]

 

Hi Pritish

 Please show the details of your “Acceleration” Boundary Condition.

Gary

 

[Pritish Sadhale]

Hello Gary,
Thank you for your reply. Please see the details as follows.

............... Original setup (Base case which is used as a restart)


.............Load case setup (Case where one of the remote displacements are to be deactivated after 2nd loadstep)


..... Remote displacement is deactivated for UY after 2nd loadstep.

I also tried deactivating all DOFs for that remote displacement, but the result does not change.

Thank you,
Pritish

[Gary Stofan]

 

Hi Pritish

Look in your ds.dat for a Tabular Load similar to below.

/com,*********** Create Acceleration ***********

*DIM,_acelx,TABLE,3,1,1,TIME
! Time values
*TAXIS,_acelx(1),1,0.,1.,2.
! Load values
_acelx(1,1,1) = 0.
_acelx(2,1,1) = 0.
_acelx(3,1,1) = 0.

 

*DIM,_acely,TABLE,3,1,1,TIME
! Time values
*TAXIS,_acely(1),1,0.,1.,2.
! Load values
_acely(1,1,1) = 10.
_acely(2,1,1) = 25.
_acely(3,1,1) = 0.

 

*DIM,_acelz,TABLE,3,1,1,TIME
! Time values
*TAXIS,_acelz(1),1,0.,1.,2.
! Load values
_acelz(1,1,1) = 0.
_acelz(2,1,1) = 0.
_acelz(3,1,1) = 0.

acel,%_acelx%,%_acely%,%_acelz%

[Pritish Sadhale]

Hi Gary,

The acceleration in ds.dat file looks like this.

/com,*********** Create Acceleration ***********

*DIM,_acelx,TABLE,4,1,1,TIME
! Time values
*TAXIS,_acelx(1),1,0.,1.,2.,3.
! Load values
_acelx(1,1,1) = 0.
_acelx(2,1,1) = 0.
_acelx(3,1,1) = 0.
_acelx(4,1,1) = 0.

 

*DIM,_acely,TABLE,4,1,1,TIME
! Time values
*TAXIS,_acely(1),1,0.,1.,2.,3.
! Load values
_acely(1,1,1) = 0.
_acely(2,1,1) = 386.
_acely(3,1,1) = 386.
_acely(4,1,1) = 386.

 

*DIM,_acelz,TABLE,4,1,1,TIME
! Time values
*TAXIS,_acelz(1),1,0.,1.,2.,3.
! Load values
_acelz(1,1,1) = 0.
_acelz(2,1,1) = 0.
_acelz(3,1,1) = 0.
_acelz(4,1,1) = 0.

[Gary Stofan]

Hi Pritish

 Restarts aside for a moment. 

Q: Does the Specified Constraint get deleted properly in a normal run?

[Pritish Sadhale]

Hi Gary,
Yes, in a normal run, the constraints do get deleted. 

[Gary Stofan]

 

 

 

Hi Pritish

 In my test case, I can see the constraint deleted message. 

 ******************* SOLVE FOR LS 3 OF 3 ****************
 DELETE SPECIFIED CONSTRAINT UX   FOR SELECTED NODES
  RANGE    1318 TO    1318 BY       1
 NUMBER OF SPECIFIED CONSTRAINTS DELETED=       1

It appears you are a commercial Ansys customer.
I would suggest opening a technical support ticket so that we may obtain your actual test files.