0

FOR parameter under PARAM.3

I recently changed the default value of the `FOR` parameter under PARAM.3 in TOUGH3 (EOS5) from 1.0 to 0.5.

Upon reviewing the output, I noticed the time values in the time-series data seem to be scaled by 2.0 (the inverse of the value I set).

Is this inverse scaling of the time output the intended behavior for the `FOR` parameter, or is this an issue/feature still in development?

Thanks for any insight.

3 replies

null
    • kenny
    • 8 days ago
    • Reported - view

    Hi Joseph,

    In TOUGH3, the "FOR" is weighing factor for determining the level of implicitness (TOUGH3 user manul have  it as "factor to change the size of the time step during the Newtonian iteration". It seems not correct .). Time step size (TimeStep) and the dt (TimeIncrement) used in Newton's iteration has following relation:

    TimeIncrement = For*TimeStep

      • Mathematical Modelling Consultant
      • Joseph_Elmes.1
      • 8 days ago
      • Reported - view

       Ah, and in MULTI.f90 with the source term,

      FAC=FORD/EVOL(J),

      where

      FORD=FOR*DELTEX.

      I have a bespoke source term coded in MUTLI.f90, and I know where there are peaks in pressure associated with the bespoke source term. Me setting FOR=0.50 has meant that the peaks in pressure arrive at 2*T, instead of T.

      I suspect this is fixed by using

      FAC=DELTEX/EVOL(J)

      instead?

      • kenny
      • 6 days ago
      • Reported - view

       I checked the source codes, it seems TOUGH3 always fully implict. The "For" should always be 1.0 (default). To be hornest, I did not pay much attend to this parameter and no one report this problem before.  We will fix it in future release. Thanks. 

Login to reply

Content aside

  • 6 days agoLast active
  • 3Replies
  • 18Views
  • 2 Following