TELEMAC-MASCARET Forum: Problem with TOMAWAC-ARTEMIS nesting option 2 (trunk) (1/1)

Welcome, Guest

TOPIC: Problem with TOMAWAC-ARTEMIS nesting option 2 (trunk)

Problem with TOMAWAC-ARTEMIS nesting option 2 (trunk) 5 years 6 months ago #33561

cyamin
OFFLINE
openTELEMAC Guru
Posts: 997
Thank you received: 234

Hello,
I am trying out the TOMAWAC-ARTEMIS nesting (NESTING WITHIN TOMAWAC OUTER MODEL=2) that is available in the trunk.
In a particular wave condition, I encountered the following error:
After making the following modifications to subroutine STWC1 to make it more verbose:

      !     COMPUTES ENERGY DENSITY AT REQUIRED F AND DIR
!
      IF ((INDD*INDF).EQ.0) THEN
        WRITE(LU,*) 'D1, D2 =' ,D1,D2
        STWC1 = 0.D0
        WRITE(LU,*) '--------------------WARNING----------------------'
        WRITE(LU,*) 'SUBROUTINE STWC1: YOU ASK FOR A PERIOD/DIRECTION '
        WRITE(LU,*) 'OUTSIDE THE RANGE OF THE TOMAWAC SPECTRUM        '
        IF(INDF.EQ.0) THEN
          WRITE(LU,*) 'F = ',F
          WRITE(LU,*) 'FMIN, FMAX =' ,SPEC%FRE(1),SPEC%FRE(NF)
        ENDIF
        IF(INDD.EQ.0) THEN
          WRITE(LU,*) 'DIR = ',TET
          WRITE(LU,*) 'DMIN, DMAX =' ,SPEC%DIR(1),SPEC%DIR(NDIR)
        ENDIF
!
        CALL PLANTE(1)
        STOP

I get this error at some point:

D1, D2 =   0.0000000000000000        0.0000000000000000     
 --------------------WARNING----------------------
 SUBROUTINE STWC1: YOU ASK FOR A PERIOD/DIRECTION 
 OUTSIDE THE RANGE OF THE TOMAWAC SPECTRUM        
 DIR =    0.0000000000000000     
 DMIN, DMAX =   3.5082387819463303E-015   350.00000000000000

In the ARTEMIS steering file I ask for a full 360 deg directional discretization (as it is done in the relevant example 'beach'):

DIRECTION OF WAVE PROPAGATION =126.0
MINIMUM ANGLE OF PROPAGATION  =-54.0
MAXIMUM ANGLE OF PROPAGATION  =306.0

This issue appears to be related to both machine precision and/or the difficulties arising in the 0/360 deg crossing. I am trying to locate the culprit but so far my efforts have failed and I am reaching for some expert advice...
Best Regards,
Costas

The administrator has disabled public write access.

Problem with TOMAWAC-ARTEMIS nesting option 2 (trunk) 5 years 6 months ago #33566

tfouquet OFFLINE Expert Boarder Posts: 294 Thank you received: 112	Hello Costas, It looks like some case may have been forgotten concerning epsilon and modulo, did you try with eps=1e-15 ? May be that there should be 2 epsilons one to find the closest frequencies and one for modulo to find directions. Can you post your case so that i can test quickly this hypothesis ? Best regards Thierry
	The administrator has disabled public write access.

Problem with TOMAWAC-ARTEMIS nesting option 2 (trunk) 5 years 6 months ago #33567

cyamin
OFFLINE
openTELEMAC Guru
Posts: 997
Thank you received: 234

Hello Thierry,
I did try an eps down to e-10 and although it worked, it appeared that spectrum transfer was incorrect. However I am not convinced it is the addition of eps that is causing the problem.
First of all, crashes begun when I defined 360deg directional discretization in ARTEMIS. Narrowing it down (by quite a bit) stops causing errors. However I was not satisfied with the results and started expanding the discretization sector and the problems begun.
I was wondering if the TOMAWAC steering files for test case beach or westcoast are available so that I can test/reproduce it there?
Regards,
Costas

The administrator has disabled public write access.