open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

i'm modelling some erosion processes at a small test-channel(40x1m) the sediment is separated in 6 fractions and so i want to represent the development of the percentages of each fraction along the channel. Therefore i use a constant active layer with a thickness i estimated as three times the maximum sediment diameter. However i receive always an error after a short time: "Error in layer case 4". I found in the forum a similar topic, but the solution of varying the ZERO value does not help.
If i use the default value for the active layer (1000) the simulation works quite well and does not stop, but so the percentages of the fractions does not change. I guess this has something to do that a in the active layer a fraction vanished, and this is not implemented so far?

Maybe some one has some hints for this problem. I attached the sisyphe-steering file. If you need more information feel free to ask.

Best regards
Markus

Hi,

Markus, your active layer thickness is two high, usually it is define as 2D50, or 3D50, but in your case, you could define as D90. because it is straight canal.

Another thing, the following key words should be set in TELEMAC2d, not in SISYPHE except you use uncoupled model

FRICTION COEFFICIENT =0.008

RATIO BETWEEN SKIN FRICTION AND MEAN DIAMETER =3.0

LAW OF BOTTOM FRICTION =5

Hi

I recommend you to comment the message errors in layer.f.
There are sometimes small error in the algorithm which allows the fractions to be less than 0 or the sum to become slightly different than 1.

But it should not have much effect on the results.

Good luck, cath

Hi, cath,

I also have a question, if we use the roughness on sisyphe, do we need to define the keywords LAW OF BOTTOM FRICTION and FRICTION COEFFICIENT in sisphe.

Hi cath,
thanks for your reply.
actual i'm not sure what you mean with "comment" the message errors? Did you mean it in that way:

ELSE
DO I=1,NSICLA
AVAIL(J,1,I) = ( AVAIL(J,1,I) * ELAY0
& + ZFCL_W%ADR(I)%P%R(J)
& - EVOL*AVAIL(J,2,I) )/ELAY0
IF(AVAIL(J,1,I).GT.1.D0+ZERO.OR.
& AVAIL(J,1,I).LT.-ZERO) THEN
WRITE(LU,*) 'ERROR IN LAYER CASE 4'
! CALL PLANTE(1)
! STOP
ENDIF
ENDDO
ELAY%R(J) = ELAY0
ESTRATNEW(J) = ESTRAT%R(J) + EVOL
ENDIF

if i did it so the simulation still broke down but now i receive a different error message, see the attached figure.

Interesting was the fact that if i run the same simulation on telemac version v6p3 it works quite well...

Markus