1

Lower limit of mesh grid?

Hello,

I have the following problem by creating a mesh grid with Meshmaker:

XYZ

NX      20    0.0004
NY       2       1.0
NZ      20    0.0004

or

XYZ

NX      20     0.125
NY       2       1.0
NZ      20     0.125

In both cases in block ELEME the distances to each grid block is wrong:
ELEME----1----*----2----*----3----*----4----*----5----*----6----*----7----*----8
A11 1              10.1600E-060.4000E-03               0.000     0.500    -0.000
A21 1              10.1600E-060.0000E+00               0.000     0.500    -0.001
A31 1              10.1600E-060.0000E+00               0.000     0.500    -0.001
A41 1              10.1600E-060.0000E+00               0.000     0.500    -0.001
A51 1              10.1600E-060.0000E+00               0.000     0.500    -0.002

or

ELEME----1----*----2----*----3----*----4----*----5----*----6----*----7----*----8
A21 1          SAND11.5620e-020.0000E+00               0.062     0.500    -0.188
A31 1          SAND11.5620e-020.0000E+00               0.062     0.500    -0.312
A41 1          SAND11.5620e-020.0000E+00               0.062     0.500    -0.438

Is there a lower limit for a grid size or can I change the limit in the sourcecode?

2 replies

null
    • Finsterle GeoConsulting
    • Stefan_Finsterle
    • 9 yrs ago
    • Reported - view

    Sebastian,

    When you say "the distances are wrong", I presume you are referring to the coordinates printed in the ELEME block. While the printout of these coordinates indeed has insufficient significant digits, it is important to realize that these coordinates are not used in TOUGH2. The relevant distances are provided in the CONNE block (i.e., the sum of the two nodal distances), and I'm very confident that they are correct! Please check to convince yourself that this is true.

    So:

    (1) The coordinates printed in the ELEME block are actually not "wrong"; they are just rounded to the third digit after the decimal point (note that for Z, the center point starts at -0.5*dZ=-0.0625, so A21 1 is at -0.0625-0.125=-0.1875, which is rounded to -0.188, i.e., all is as expected!)

    (2) If you are bothered by this rounding issue (even though it does not affect your calculations!), you could go to meshm.f and change the appropriate FORMAT statement.

    (3) You could also create a mesh that is given in cm instead of m (i.e., dX=0.04 instead of 0.0004). The coordinates (and everything else in ELEME and CONNE) are then in cm (without round-off issues); since TOUGH2 expects m, you would need to set SCALE in PARAM.2 to 0.01 to compensate for it.

    Hope this helps!

    Stefan

    • Sebastian
    • 9 yrs ago
    • Reported - view

    Hi Stefan,

    thank you very much.  Yes, in the CONNE block the relevant distances are correct.

Content aside

  • 1 Likes
  • 9 yrs agoLast active
  • 2Replies
  • 903Views
  • 3 Following