Poroelastic Modelling

Numerical methods and mathematical models of Elmer
Matthew
Posts: 31
Joined: 19 Jun 2014, 18:22
Antispam: Yes

Re: Poroelastic Modelling

Post by Matthew »

Hi Peter,

My idea to solve for the equation, roughly, is as follows (.sif file):

Code: Select all

...
Solver 1
 Equation = "Heat Equation"
 Variable = Temperature1
End

Solver 2
 Equation = "Heat Equation"
 Variable = Temperature2
End

Solver 3
 Equation = "Heat Equation"
 Variable = Temperature3
End
...
Let say the above .sif file will able to calculate the three different temperatures but using the same Heat Equation (because I have define three different Variable Temperature), how can I add all these three temperatures to allow elastic solver to use it and compute the heat expansion?

The idea of using different variable name came from your answer in previous post below:

viewtopic.php?f=3&t=3510&p=11649&sid=4e ... 528#p11649

Hope you can justify my idea whether it is feasible or not. Thank you very much.

regards
Jamil
raback
Site Admin
Posts: 4832
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Poroelastic Modelling

Post by raback »

Hi

You could do three different solvers, for TempA, TempB and TempC. Then in the thrird one you could have an "Exported Variable 1" called "Temperature" such that would be the sum of the three. Then no changes is needed in the elasticity solver.

-Peter
Matthew
Posts: 31
Joined: 19 Jun 2014, 18:22
Antispam: Yes

Re: Poroelastic Modelling

Post by Matthew »

Thanks Peter for the clarifications!! :D

One more question. Regarding the exported variables and summing the temperatures, I am not very sure how to do it, although I got your idea. How can I add the temperatures and then represent it as exported variable? Do I need to use MATC or to write a user define function?

Thanks,
Jamil
Matthew
Posts: 31
Joined: 19 Jun 2014, 18:22
Antispam: Yes

Re: Poroelastic Modelling

Post by Matthew »

Hi Peter,

I just not really understand how elastic solver will read the variable temperature. I presume that, when using only heat equation and elastic solvers, the elastic solver will automatically read the variable temperature defined in the heat equation solver. However, if I am going to use three modified heat equation solvers, how the elastic solver knows which variable from which solver it needs to compute the heat expansion? Please clarify for me, cause I am quite confuse.

Thanks,
Jamil
raback
Site Admin
Posts: 4832
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Poroelastic Modelling

Post by raback »

Hi Jamil

Do something like in

Code: Select all

Solver 1
  Name = "SolverA"
  Variable = TempA
  ...

Solver 2
  Name = "SolverB"
  Variable = TempB
  ...

Solver 3
  Name = "SolverC"
  Variable = TempC
...
! create an additional variable
  Exported Variable 1 = Temperature
  Update Exported Variables = Logical True
End 

Solver 4 
  Name = "StressSolver"
  ...
End 

! sum up, this is activiated with "update exported variables"
Body Force 1 
  Temperature = Variable TempA, TempB, TempC
     Real MATC "tx(0)+tx(1)+tx(2)"
End 
Have fun!

Peter
Matthew
Posts: 31
Joined: 19 Jun 2014, 18:22
Antispam: Yes

Re: Poroelastic Modelling

Post by Matthew »

Hi Peter,

It took me quite awhile to try the method that has been discussed in this post.

I just want to ask, if I name the variable temperature in the solver section below:

Variable = Temperature

as something else, say TempA, will I need to make some changes in the solver as well, ie the .F90 code?

Because when I run a simulation to test the Heat Solver equation by naming the variable as TempA, I could not solve the problem properly.

Thanks,
Jamil
Post Reply