open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello everyone,

I managed to generate the output for my culvert by implementing your suggestions into buse.f! I´m just wondering if the maximum output of parameters is limited to 6?
I would like to get the following parameters (attached my modified buse line):
AT,V2DPAR%R(I1),QMAX1,Q,DBUS(N),S1,S2,FTYP

The results file only prints until S1 and misses S2 and FTYP.
Is the file somehow limited in columns or any other reason?

Further the "VD2PAR%R" is confusing me. In scalar mode of my 1x1m regular grid it shows different areas, depending on the direction of the edges (see pictures).How is the area calculated exactly? In parallel mode it even drops to 6,7E-270. Any ideas, where this comes from?



Thanks
Best regards
Chris

ps: I´m running on V7p3r1, parallel

Hello,

The fact that variables after S1 are not printed might be due to the fact that your line 543 exceeds the maximum number of characters allowed per line. By default it is 73 or 74 but this can be tuned in the config file (via compilator option). I would advise you to divide your WRITE statement on 2 lines.

The fact that V2DPAR is returning 0 might be due to a wrong handling of node IDs in parallel (just a first idea, I havent looked at it in depth). Test running it in serial, it should work. Then you can remove V2DPAR from the WRITE statement in your parallel runs as this value is constant in time.

Hope it helps!

Kind regards
PL

Hi,

I agree with PL.

For V2DPAR, there is no reason that in parallel mode, your culvert is located in the MPI subdomain 0, computed by the processor 0 (IF IPID==0), but it certainly is in another subdomain.

About your mesh and the directions of the edges, it's normal that the area surrounding a point is not the same according to the edge orientation (see attached figures).




I hope it helps.

Regards,
Laurent