Encyclopaedia Index

WALL

------ Command; group 13 --------------

This command is used to declare areas of wall-friction. Non-zero wall values are declared with the VALUE command. See the entry on VALUE for more information.

The syntax is : WALL(NAME,TYPE,IXF,IXL,IYF,IYL,IZF,IZL,ITF,ITL)

NAME is a unique identifier for the WALL, up to 8 characters in length. This name must be unique within the first five characters. TYPE can be NORTH,SOUTH,EAST,WEST,HIGH or LOW, and specifies which cell faces are in contact with the fluid.

IXF ... ITL specify the limits in space and time over which the wall is active. If the wall is internal to the flow domain, friction is applied to both sides, and flow is prevented from passing through the wall.

If the K-E turbulence model is active, the appropriate wall conditions are set.

If the wall is not adiabatic, the boundary condition for the enthalpy equation must be supplied as COVAL(NAME,H1,GRNDn, specific heat*wall temperature).

See the entry on WALLS for more information. Thus :

REAL(UWALL) ; UWALL = 3.0
WALL(WALL1,SOUTH,1,NX,1,1,1,NZ,1,LSTEP)
VALUE(WALL1,U1,UWALL)

specifies a WALL, extending from IX = 1 to NX, IZ = 1 to NZ at IY = 1, active during all time steps, on the SOUTH side of the cells specified, moving with a velocity of UWALL in the X direction, but with zero velocity in the Z direction. If enthalpy is solved, this example would be an adiabatic wall.

The function of the command is to generate a PATCH with the NAME specified over the set limits. The patch type is set to the appropriate wall type i.e.NORTH->NWALL etc. If the wall is internal, a CONPOR is generated to block flow normal to the wall, and another patch is generated for the other side of the wall. The name of the new patch is based on the first 5 characters of NAME.

COVALs are generated for velocities parallel to the wall, with GRND2 and 0.0 as the third and fourth arguments. If KE and/or EP are solved, their COVALs have GRND2 and GRND2 as arguments.

The GRND2 default may be changed by setting WALLCO=GRND1 or GRND3 as required ( see WALLCO ).


WALL DISTANCE

See PHENC entries: DISTANCE from the WALL, and DISWAL


WALL friction, declaring areas of

(see WALL command, Group 13)


WALL function parameter, variable for

(see WALLA real, Group 13)


WALL Functions

See Wall functions entry in Turbulence Models in PHOENICS


WALLA

----- PIL real; default= 0.0; group 13 -- -

WALLA....GREX variable for wall-function parameter.

For the GRND2 (i.e.LOGLAW) and GRND3 (i.e.GENLAW) turbulent wall- functions, it represents the absolute "sand-grain roughness" size for rough walls, according to the formula of Nikuradse.

For GRND5 wall functions, it is essential to set a roughness height. WALLA represents the effective roughness height. The default of 0.0 gives the smooth-wall value. It is specified in the same units as other lengths.


WALLB

----- PIL real; default= 0.0; group 13 -- -

Inactive.

WALLB....GREX variable for wall-function parameter.


WALLCO

----- PIL real; default=GRND2; Group 13 - -

WALLCO indicates whether the :

is to be used whenever wall- friction effects are generated automatically by Earth.

This occurs in two circumstances, namely:

  1. when EGWF = T, its default value; and
  2. when EGWF = F and the Satellite generates patches on the boundaries of solids by reason of the CONPOR(-1,... or CONPOR(0,... commands.

See the help and encyclopaedia entries for WALL and EGWF for further information.


WALLS

--------------------------------------

WALL....is not itself a TYPE. However, the appearance of WALL as part of the PATCH-type name implies that the specified value of the COefficient will be multiplied by the reference kinematic viscosity times the fluid density and divided by the distance of the grid node from the specified face. See the types: EWALL, WWALL, NWALL, SWALL, HWALL and LWALL.

As implied in the above paragraph, wall effects (friction and heat transfer, etc) are represented by sources (of U1, U2, V1, V2, W1 and W2 for wall friction, of H1 and H2 or TEM1 for wall heat transfer, etc). The PATCH command is used to locate the wall, and COVAL is used to activate the source for the variable indicated in the second argument of COVAL.



    The entries:

           PATCH(name,EWALL,.....

           COVAL(name,H1,1.0/PRNDTL(H1),-10.0)

    imply that there is a wall at an east cell face and that the

    transfer rate of heat from it is equal to:

           ((-10.0) - fluid enthalpy)*(east-face area)*

           (reference kinematic viscosity in the cell)*

           (fluid density prevailing in the cell)/

           ((distance of cell centre from the wall)*PRNDTL(H1))

    The additional command,

           COVAL(name,W1,1.0,30.0)

    sets the velocity of the wall equal to 30.0 m/s in the z direction,

    and supplies the following source of momentum to the W1 variable

           ((30.0) - W1 velocity)*(east-face area)*

           (reference kinematic viscosity in the cell)*

           (fluid density prevailing in the cell)/

           (distance of cell centre from the wall)

The above settings supply wall functions valid for laminar flow. If the third argument of COVAL is equal to GRND, GRND1, etc, the GROUND subroutine is entered for a further multiplier, which is ordinarily dependent on the local Reynolds number.

Four options are provided in subroutine GXWFUN called from GREX, namely:

  1. coefficient = GRND1 gives Blasius' law;
  2. coefficient = GRND2 gives the logarithmic law.
  3. coefficient = GRND3 gives the generalised wall function.
  4. coefficient = GRND5 gives the fully-rough logarithmic law.

Options (b) and (c) also allow the specification of a 'rough' wall by the setting of WALLA to a finite positive 'sand-grain' roughness height. Option (d) defines a 'fully-rough' wall specified in terms of the effective roughness height WALLA.

When the turbulent kinetic energy KE is solved, it is necessary to supply wall boundary conditions for it. The following command:

COVAL(name,KE,GRND2,GRND2)

fixes KE at the near-wall nodes identified by PATCH to equal:

(wall shear stress)/(sqrt(CMUCD)*density) .

The following command:

COVAL(name,EP,GRND2,GRND2)

fixes EP at the near-wall nodes identified by PATCH to equal:

(CD/AK)*KE**1.5/(cell-node-to-wall distance)

For the generalised wall function, for boundary conditions for KE and EP, set:

COVAL(name,KE,GRND3,GRND3)
COVAL(name,EP,GRND3,GRND3)

The turbulence- model constants CMUCD, CD and AK are set in group 1 of GREX thus: CMUCD=0.09, CD=0.1643 and AK=0.435.

For the fully-rough logarithmic wall functions, the boundary conditions are set as follows:

COVAL(name,W1,GRND5,0.)
COVAL(name,KE,GRND5,GRND5);COVAL(name,EP,GRND5,GRND5)

which arranges that the boundary conditions used for KE and EP are exactly the same as used for GRND2 wall functions.

See CONPOR for an automatic means of activating wall friction at fully-blocked cell faces.

See WALL for means of activating wall friction without the need to declare PATCHes and COVALs for all variables at all surfaces.


Walls, high faces

(see HWALL)


WALLs, low faces

(see LWALL)


WALLTY

-------------------------------------

Ground logical.

WALLTY....is .TRUE. when the current patch 'type' is one of the wall types (ie. EWALL, WWALL, NWALL, SWALL, HWALL or LWALL). See TYPE.


wbs