[Getdp] Independend Solution of the same system

Tammo Heeren getdp at tammoheeren.com
Mon Jan 8 18:39:26 CET 2007


Bernhard,

thank you for your reply. I had seen you ChainedResolutions before and  
tried to adapt it to my case. However, I never got it to work  
properly. After you mentioned it, I looked at it again. What I was  
missing was a second FunctionSpace definition. I got it to work nicely  
now.

Tammo

----------------------------------------------------------------------

Maybe what one could do is to create two formulations (chained ones).
The first is your normal problem. The second one inherits all data
from the first formulation, but does the same calculations for a
renamed set of fields and the changed parameter.
This is quite similar to the chained resolution example: [1]

The postprocessing then should be able to access both datasets,
distinguished by the renaming.

nice greetins,
bernhard

----------------------------------------------------------------------

I am trying to solve one system of equations twice, with one
parameter changed. Both results should be independent of each other.
However, I would like to be able to access both results in
PostProcessing to compute specific values like "{ Name C ; Value {
Term { [ {Q}/{V} ] ; In Grouped; } } }".
I dont want to have to run the simulation manually twice. I also want
to have mixed results (from run 1 and run 2) included in one
PostOperation run. Is this possible, and does anybody have an idea of
how to do that?

Tammo

Links:
------
[1] http://www.geuz.org/getdp/wiki/ChainedResolutions