open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi Developers and users,

---

1)For 2D flows, if one needs to use FINITE VOLUME SCHEME, it is well understood from the reference manual that one should make use of the options
EQUATIONS : "SAINT-VENANT VF"
FINITE VOLUME SCHEME : 0 or 1 or 2
DESIRED COURANT NUMBER :1

In such case of FINITE VOLUME SCHEME, is it mandatory to select ONLY option 'Upwind explicit finite VOLUMES' for TYPE OF ADVECTION ?
OR
One can use any one of the following options?
1: Method of characteristics,
2: Centred semi-implicit scheme + SUPG upwinding,
3: Upwind explicit finite volumes,
4: N distributive scheme, mass-conservative,
5: PSI distributive scheme, mass-conservative (mandatory for H),
6 : PSI scheme on non-conservative equation,
7 : Implicit N scheme on non-conservative equation,
13: Edge by edge implementation of scheme 3,
14: Edge by edge implementation of scheme 4

---

---

2) Can we choose the option "Upwind explicit finite volumes" for TYPE OF ADVECTION in case if we want to use
EQUATIONS : "SAINT-VENANT EF" (Finite Element solution)





Thanks in advance

Hello,

With finite volumes the keyword TYpe OF ADVECTION is not looked at. With finite elements you can choose option 3 but it is equal to option 4 (only in 3D the two methods are different).

With best regards,

Jean-Michel Hervouet

Hi,
Thanks for the reply, i want to make sure that while we use finite volume method, even if the option ADVECTION TERMS is set TRUE, then also TYPE OF ADVECTION is not considered for the computation?

Thanks in advance

Hello,

when using the Finite Volume kernel a lot of Finite Element related keywords are disregarded. As Jean-Michel said also TYPE OF ADVECTION is not considered and the Finite Volume kernel computes the SWE equations. With the keyword FINITE VOLUME SCHEME you can choose between the following schemes:
'0="Roe scheme"';
'1="kinetic order 1"';
'2="kinetic order 2"';
'3="Zokagoa scheme order 1"';
'4="Tchamen scheme order 1"';
'5="HLLC scheme order 1"';
'6="WAF scheme order 2"'

Example for Finite Volume steering file:
EQUATIONS ='SAINT-VENANT VF'
FINITE VOLUME SCHEME =5 / 5=HLLC
VARIABLE TIME-STEP =true
DURATION =7200
DESIRED COURANT NUMBER =0.95

Have a look in the Telemac-2D User Manual for more informations:
www.opentelemac.org/index.php/manuals/vi...tegory/13-telemac-2d

Hope this helps,
Clemens

Hi Clemens,
Thanks for your help, i referred the manual. What i understood is for FINITE VOLUME KERNEL we should avoid using the options such as
ADVECTION OF H
ADVECTION OF U AND V
ADVECTION OF K AND EPSILON
ADVECTION OF TRACERS
should not be used in case of Finite-volume kernel. Right?

However, i wish to know whether the following can be mandatory for Finite volume kernal as they are related to diffusion terms, implicitation and diagonalization of the matrix [A] of any of the momentum equation?

DIFFUSION OF VELOCITY
SUPG OPTION
IMPLICITATION FOR DEPTH
IMPLICITATION FOR DIFFUSION OF VELOCITY
IMPLICITATION FOR VELOCITY
MASS-LUMPING ON H
MASS-LUMPING ON VELOCITY

Thanks in advance

Hello,
as far as I know in Telemac-2D with Finite Volume schemes you cannot turn off specific advection terms in the SW equations. And be aware that not all of the FV schemes account for the diffusion terms (eddy viscosity). Since the FV schemes use an explicit time discretization all the keywords related to implicitation coefficients don't matter. The SUPG and the Mass Lumping options don't matter within the FV solution method. Actually the matrix to be solved in FV is similar to a mass lumped matrix in FE.
In short:
if you have all these keywords in your steering file and somewhere further down you have the keyword EQUATIONS ='SAINT-VENANT VF' you will activate the FV kernel and disregard all the FE related keywords since Damocles reads the steering file sequentially.


Hope this helps,
Clemens

Hi Clemens,
Thanks for the information.