Welcome, Guest
Username: Password: Remember me
Forgot your password? Forgot your username? Create an account
Forum
System Modules
TELEMAC-3D
Unreasonable velocity results near boundary when add velocity On BC
Flat Threaded

TOPIC: Unreasonable velocity results near boundary when add velocity On BC

Unreasonable velocity results near boundary when add velocity On BC 6 years 8 months ago #29023

  • liuy
  • liuy's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 13
Hi,
I am running a telemac 3D test case in a rectangle domain,with the open boundary on right side, and add a velocity of U=-1,V=0 on it,but the results seems unreasonable,the velocity are positive near the boundary node,shows in attachments.
I also see the result of free surface, it keeps zero in the boundary node,it may be the reason why the velocity are positive near the boundary node,but i don't know how to set my steering file to get a reasonable results.Besides,I tried THOMPSON Boundary conditions,it doesn't work.

Attachments are my files used in this case.

Hope you can help me.thank you

liuy
Attachments:
The administrator has disabled public write access.

Unreasonable velocity results near boundary when add velocity On BC 6 years 8 months ago #29041

  • Leballeur
  • Leballeur's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 434
  • Thank you received: 163
Hi Liuy,

First can you explain what you want to do.
The settings of your model are strange, there are useless or unsuitable keywords. Why do you use binary boundary file for such a simple case ? The positive velocity near your boundary is due to your prescribed elevation to 0 combined with incoming velocities that are filling your domain.
You can start with a 2D configuration and have a look to the examples where you will find how the flow is prescribed for this kind of test case.

Regards,
Laurent
The administrator has disabled public write access.
The following user(s) said Thank You: liuy

Unreasonable velocity results near boundary when add velocity On BC 6 years 8 months ago #29050

  • liuy
  • liuy's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 13
Hi,Laurent
Great thanks for your reply, what I want to do are add ocean current(from HYCOM) to my Telemac 3D real case, I just simplified my case to idealized domain,so I use binary boundary file. In my Telemac 3D real case ,the velocity results near boundary shows the same as the idealized domain.
Can you share me a telemac3d cas file for add ocean current in the boundary? thank you so much.

Regards,
liuy
The administrator has disabled public write access.

Unreasonable velocity results near boundary when add velocity On BC 6 years 8 months ago #29055

  • Leballeur
  • Leballeur's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 434
  • Thank you received: 163
Hi Liuy,

The results you're getting are agree with your model's set-up.
If you want to avoid this behaviour, try not prescribing 0 elevation values in your binary boundary file, but rather values in agreement with the velocities. Or you can also try to let the elevation free, by prescribing only velocities (then you can play with Thompson option).
You may also have to wait for a spin-up period.

By the way, for your test case, no need using tidal flats and associated options, you can remove the PPE options as you don't use non-hydrostatic version, remove the bottom smoothing (even for your real case), turn on Coriolis effect for your real case, keep the default value for the implicitation of depth.

Regards,
Laurent
The administrator has disabled public write access.

Unreasonable velocity results near boundary when add velocity On BC 6 years 8 months ago #29061

  • liuy
  • liuy's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 13
Hi,Laurent
Thanks for your usefull advice.
From bord3d.f
!
!     PRESCRIBED ELEVATION GIVEN IN STEERING FILE (NCOTE<>0)
!     OR IN BINARY BOUNDARY DATA FILE
!     -------------------------------------------------------
!
      IF(LIHBOR%I(K).EQ.KENT.AND.NCOTE.NE.0) THEN
!
        IPOIN2 = NBOR2%I(K)
        ICOT=NUMLIQ%I(K)
        IF(STA_DIS_CURVES(ICOT).EQ.1) THEN
          HBOR%R(K) = STA_DIS_CUR(ICOT,FLUX_BOUNDARIES(ICOT),
     &                            PTS_CURVES(ICOT),QZ,NFRLIQ,
     &                            ZF%R(IPOIN2)+H%R(IPOIN2))
     &                - ZF%R(IPOIN2)
          HBOR%R(K) = MAX(0.D0,HBOR%R(K))
        ELSEIF(NCOTE.GE.NUMLIQ%I(K)) THEN
          N=IPOIN2
          IF(NCSIZE.GT.1) N=MESH2D%KNOLG%I(N)
          IF(READ_BIN_Z) THEN
            ! GET THE ELEVATION Z FROM THE LIQUID BOUNDARY BINARY FILE
            IF(INFOGR.AND..NOT.DEJA) THEN
              IF(LNG.EQ.1) WRITE(LU,*) 'COTES AUX FRONTIERES LUES DANS
     &                     LE FICHIER BINAIRE'
              IF(LNG.EQ.2) WRITE(LU,*) 'ELEVATION AT THE BOUNDARY READ
     &                     IN THE BINARY FILE'
              DEJA = .TRUE.
            ENDIF
            HBOR%R(K) = T3_01%R(N+(NPLAN-1)*NPOIN2)-ZF%R(IPOIN2)
          ELSE
            ! GET THE ELEVATION Z FROM THE ASCII FILE
            IF(INFOGR.AND..NOT.DEJA) THEN
              IF(LNG.EQ.1) WRITE(LU,*) 'COTES AUX FRONTIERES LUES DANS
     &                     LE FICHIER ASCII'
              IF(LNG.EQ.2) WRITE(LU,*) 'ELEVATION AT THE BOUNDARY READ
     &                     IN THE ASCII FILE'
              DEJA = .TRUE.
            ENDIF
            HBOR%R(K) = SL3(ICOT,AT,N,INFOGR)-ZF%R(IPOIN2)
          ENDIF
          HBOR%R(K) = MAX(0.D0,HBOR%R(K))
From the code ,If i didn't set prescribing elevation and velocity, it won't use the elevation and velocity read from binary file,and the Thompson option "However it is important to note that, given the two-dimensional aspect, the Thompson method can only be used in the case of a zero velocity gradient imposed on the vertical (uniform velocity along the vertical)."(From manual of telemac 3d), as the manual said ,i think the Thompson method may not suitable for read case.
I think I should change some code in bord3d.f to make my real case run.

Regards,
liuy
The administrator has disabled public write access.

Unreasonable velocity results near boundary when add velocity On BC 6 years 8 months ago #29062

  • Leballeur
  • Leballeur's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 434
  • Thank you received: 163
Hi Liuy,

In bord3d, there is a section after the one you mention that deals with the current too. So according to your boundary conditions file, you can prescribe either elevation alone, either current alone or both of them, and prescribed values will be read from your binary boundary data file. You can turn on thompson option for each setting, but some combinations may be unsuitable here.
(By the way, you're using an uniform velocity profile along the vertical, so you can use it).

For your test case, you're prescribing a 0 elevation value on the right boundary with an incoming current filling your domain. You can see on the left side of your domain, the elevation is increasing. On the right side, you request it remains to 0, so outgoing velocities are created to equilibrate the slope of free surface generated.

If you want that the elevation on the right side also increases, set it free on the boundary. I tested it without thompson option, and after some spin-up time, it works.

For your real case, it will be different as you'll have non-zero elevations, and time-varying speeds and elevations.

Regards,
Laurent
The administrator has disabled public write access.
The following user(s) said Thank You: liuy
Flat Threaded
Moderators: pham
Forum
System Modules
TELEMAC-3D
Unreasonable velocity results near boundary when add velocity On BC

The open TELEMAC-MASCARET template for Joomla!2.5, the HTML 4 version.