Encyclopaedia Index
Both PHOTON and AUTOPLOT start by looking for and (if they find it in the working directory) reading a so-called 'USE' file with the name U. This can also be called a 'macro', in that it contains a succession of commands which can be understood and acted on by an executable program.
It is often convenient to use such a file, especially when many similar runs are being performed with only slight variations of data.
Of course, if a U file is left in the directory from a previous PHOTON session which is not appropriate to the new session, PHOTON will read some commands which do not fit the current data. It is therefore a wise practice to delete U, or save it with a different name, after use.
Since many Input-Library files have been provided with embedded PHOTON or AUTOPLOT
commands, to have a U file consisting of the single line:
use q1
ensures that these embedded commands are automatically picked up. No harm is done (apart
from the waste of a second or two) if the the Q1 file does not contain any such commands.
Two further points should be noted:
See also USE.
Because PHOTON represents three-dimensional objects on a two-dimensional screen, it is necessary to specify the direction from which the object plotted is to be viewed, and its orientation. To do this, a 'view' direction, V, and an 'upright' direction, U are specified.
The view-direction vector, V, is a vector of arbitrary magnitude that points from the view-point, at infinity, to the cartesian-coordinate origin. You specify the orientation of the projection on the screen by the upright-plane vector U which lies in a vertical plane perpendicular to the screen. Both vectors are specified in terms of their resolutes along the cartesian axes X, Y and Z shown in the bottom right-hand corner of the VDU screen.
The U and V directions cannot be the same: the program issues an error message if an attempt is made to make them so.
The U and V directions are set by means of the commands UP and VIEW, the arguments of which specify the direction.
----------------------------------------- -
Integer flag; value=3.
U1....standard name used to denote the first-phase velocity component in the 1x-direction.
See PHI and NAME for further information.
(see IURINI)
------ Real; default= 0.0; group 8 --- -
U1AD....add extra velocity U1AD to U1 so that the convection flux across the east face of cells is computed using the velocity U1+U1AD. For BFCs, there are two kinds of velocity, the velocity resolutes which are the subject of the momentum equations, and the velocity components along grid lines from which the convection fluxes are computed. PHOENICS code adds U1AD to the latter (UCMP), which are use to compute the east-face convection fluxes.
Set U1AD=GRND to cause EARTH to visit group 8 of GROUND for an array of values to add to the field of U1 at each z slab.
This option may be used to introduce a moving or floating grid.
----------------------------------------- -
Integer flag; value=4.
U2....standard name used to denote the second-phase velocity component in the 1x-direction.
See U1, PHI and NAME for further information.
------ Real; default= 0.0; group 8 --- -
U2AD....add extra velocity U2AD to U2. See U1AD.
------ Character flag; group 7 -------- -
U2CR....integer name recognized by EARTH denoting second-phase cartesian XC-directed velocity resolute. See UCRT.
------------ Integer flag
UC1.... name used to denote the first-phase co-located velocity component in the x-direction.
---- Logical; default=F; group 8 , -
UCONNE....must be set to T if convection neighbours are to be accessed or altered in group 8 of GROUND.
----- Logical; default=F; group 8 , -
UCONV....must be set to T if convection fluxes are to be accessed or altered in group 8 of GROUND.
---- Logical; default=F; group 8 , -
UCORCO....must be set to T if correction coefficients are to be accessed or altered in group 8 of GROUND.
----- Logical; default=F; group 8 , -
UCORR....must be set to T if correction values are to be accessed or altered in group 8 of GROUND.
------ Character flag; group 7 -------- -
UCRT....integer name recognized by EARTH denoting first-phase velocity resolute in the direction XC of the cartesian reference system XC, YC and ZC. The command STORE(UCRT) selects the first unallocated store to be found, working from variable NPHI backwards. See also VCRT and WCRT.
UCRT, U2CR, VCRT, V2CR, WCRT, W2CR are for use only in BFC simulations.
Files having the extension .uda are compressed versions of ASCII text files which are readable by POLIS.
A text file of which the name has no extension can be converted to .uda format by issue of the command:
poluda filename
Issue of the command:
poluda filename.uda
results in the recovery of the original file.
The input-library files which are accessed by the PHOENICS Satellite have the extension .uda; but their names also contain the characters da ahead of the dot. Thus, the input-file library concerned with body-fitted coordinates is called bfclibda.uda.
These characters, which derive from "direct-access", are an indication that, in the course of the creation of the files by way of the 'boot' command, a set of numbers has been generated, and placed at the bottom of the file, which informs the Satellite on which line a particular library case is to be found.
It is therefore unwise to use the 'poluda' command on these files; only 'boot' should be used.
----- Logical; default=F; group 8 , -
UDIFF....must be set to T if diffusion fluxes are to be accessed or altered in group 8 of GROUND.
---- Logical; default=F; group 8 , -
UDIFNE....may be set to T if diffusion neighbours are to be accessed or altered in group 8 of GROUND.
(see TR304)
----- Logical; default=F; group 6 ---- -
UGEOM....when set to T permits current z-slab geometry to be changed in group 6 of GROUND at the beginning of each iz-slab visitation.
A common mistake is to set problems for which there is no unique answer. For example, an incompressible fluid may be caused by some momentum source to circulate steadily within a fixed-volume cavity; then, although the pressure DIFFERENCES which prevail within the cavity can be uniquely computed, the ABSOLUTE LEVEL of pressure can not.
PHOENICS will find this situation confusing, especially when the pressure level drifts to such large values that round-off errors result when it subtracts the pressure at one point from the pressure at a neighbouring point as it is required to do when solving for velocity.
In such circumstances, PHOENICS must be given more information. Thus the pressure might be arbitrarily fixed (to zero, say) at an 'anchor point' within the flow domain, by means of appropriate PATCH and COVAL commands.
Similar difficulties may arise with temperature, for example, when the flow is steady and all the surroundings are insulated. Specification of the temperature at one point, at least, is essential.
Another mistake of the same kind is to seek to simulate in a steady- flow manner a two-phase phenomenon occurring within a container, into which only one of the phases is continuously supplied, and when there is no transfer of mass between the phases. An example is the bubbling of a gas through a liquid held in an open-topped tank.
Such problems have no unique solution; for what can determine the amount of the non-supplied phase which will be present in the container? It is only by setting STEADY=F, and providing appropriate initial conditions, that the flow which is to be simulated can be uniquely determined. Alternatively, the boundary conditions must be changed in some way which makes the steady-flow formulation physically meaningful, for example by the provision of non-zero inflows for BOTH phases.
(see RELAX; LINRLX; FALSDT; DTFALS; EXPERT; and SARAH)
The PIL setting RELAX(P2,LINRLX,factor), where 'factor' is a real number, multiplies the continuity errors by 'factor' before they enter the pressure-correction sequence. This may be a more effective solution-smoothing sequence than under- relaxing the pressure correction itself, because it results in reduced (for 'factor' >1.0 ) velocity adjustments. RELAX(P2,... has no influence on the P2 values themselves.
(Text menu) ---------------------------------- Photon Help ----
[Underline] shows whether the current TEXT element is underlined. It can be changed here.
---- Autoplot Help ----
UNE[QUAL]
Opposite of EQUAL. Forces return to normal axis scaling.
See also HELP on : EQUAL
(Vector Edit Menu) -------------------------------------- Photon Help ----
[Units] set the current units of the vector. It is shown at the bottom of the colour bar and by the reference vector if the [Key] option is [Yes].
(Contour Edit Menu) -------------------------------------- Photon Help ----
[Units] set the current units of the variable to be contoured. It is shown at the bottom of the colour bar.
----------------------------------------- Photon Help ----
U[p] <direction>....changes the orientation of the plot. The 'up direction' is any vector is 3D space which appears vertical when projected onto the screen. See the help entry for VIEW for details on how to specify directions.
For example, UP Y sets the Y-axis as the up direction.
NOTE that the UP and VIEW directions cannot coincide, by definition.
See also : VIEW
The message 'Press RETURN to continue' will then appaear in the screen, and remain until RETURN is indeed pressed.
--------------------------------------- Photon Help ----
Sets the UP vector to (1,0,0)
--------------------------------------- Photon Help ----
Sets the UP vector to (0,1,0)
--------------------------------------- Photon Help ----
Sets the UP vector to (0,0,1)
------------------------------------- Photon Help ----
UPA[use] <nsecs>.... is used within USE files to perform a timed pause. It differs from PAUSE in that no user input is required; instead, a timed pause of length proportional to the nsecs parameter is performed. The value of nsecs needed to produce a given time interval will be machine dependent.
(Contour edit menu) -------------------------------- Photon Help ----
The maximum value of the current contour range if the RANGE option is on.
(stream set menu) ------------------------------------- Photon Help ----
Streamlines will be tracked upstream from the starting positions.
In the upwind-differencing scheme, the diffusive flux is present irrespective of the cell Peclet number; whereas in the hybrid scheme the diffusive flux is cut out for Peclet numbers in excess of 1.0/DIFCUT.
See the Encyclopaedia entry 'DIFCUT' for further information.
------------------------------------ Photon Help ----
URE[wind].... is used within USE files to rewind the file and restart processing from the start of the file. It should be used with extreme care, as it will put PHOTON into an endless loop. This facility can be used to generate a repetitive display of a set of pictures.
USEGRD is a logical PIL variable which, when set = T in the Q1 file, ensures that the open-source subroutine GROUND is called during execution of EARTH.
USEGRX is a logical PIL variable which, when set = T in the Q1 file, ensures that the open-source subroutine GREX3 is called during execution of EARTH.
A USER_DEFINED object is used to define the location of one or more PATCH commands, together with their COVALs. The action of the PATCH/COVAL settings is restricted to the cells which fall inside the scope of the object. See the description in the PHOENICS_VR Reference Guide, TR326
USEPIL is a logical variable which is to be seen in the open-source Fortran coding associated with material properties.
It is set within EARTH for each such property, and should not be changed by the user; it is explained here solely in order that the coding provided by CHAM, and which users are free to extend, can be properly understood.
USEPIL=.FALSE. is an indication that:
USEPIL=.TRUE. is an indication that, whether PRPS is stored or not, the properties must be computed for each cell in the different manner dictated by the values given to the PIL variables: RHO1, ENUL, CP1, etc.
This is the case for all those properties which are not deducible from the PROPS file.
Even for properties and circumstances for which USEPIL is .FALSE., PIL variables will still be used for those cells for which the PRPS value is less than or equal to -1.0.
The PHENC entry PROPERTIES should be consulted for further relevant information.
---- Logical; default=F; group 8, sec -
USOLVE....must be set to T for entry to GROUND sections which permit introduction of user's own linear-equation solver. Selective operation of this feature is brought about by means of the GROUND variables SLBSOL and USER: thus, one or more variables can be solved with the GROUND solver but the remaining variables can be solved with the standard in-EARTH solver.
For example, if USOLVE=T and CSG3=GAUS, GREX calls GXGAUS for Gauss-Seidel solution of
all variables that have been selected to be solved whole-field (ie. those for which the
4th argument of SOLUTN is Y). After the call to GXGAUS, the following statement appears in
GREX:USER=.NOT.SLBSOL
This ensures that on RETURN to EARTH, EARTH solves those variables identified to be solved
by the slab-by-slab linear solver.
---- Logical; default=F; group 8, sec -
USOURC....must be set to T for entry to GROUND group 8, section 12 for modifications of or substitutions for the sources of the finite-volume equations.
It resides in \phoenics\d_utils\d_win64\d_usp.
------ Logical; default=T; group 22 - -
USTEER = T, coupled with TSTSWP.GE.0 (which switches off the graphical monitor), allows EARTH operations to be interrupted by the user, as follows:
The relevant data entries should be placed in the q1 file,
between the lines:
USTRBEGIN
and
USTREND
Library case 249 provides an example.
----- PIL Real; value= 0.3; group 19 -
This sets the width (in terms of SURN value) over which the tension force is applied.
------- Logical; default=F; group 6 ---- -
UUP....should be set to T when boundaries are present which give high curvature of the grid lines of constant J and K, ie. the direction of the U velocity resolute is changing rapidly.
UUP equal to T causes an upwinding of the contributions of the lateral velocities V and W to be performed in the representation of the grid-line curvature term for U for those cells adjacent to boundaries ( both domain boundaries and internal blockage ).
When a boundary layer is thickening prior to separation on a curved boundary, where U is in the flow direction, UUP=T will permit streamline separation to be predicted for coarser grids than would otherwise be possible.
------ Logical; default=t; group 22 - -
UWATCH = T causes line-printer plots of monitor values to be printed to the VDU in place of tables of residuals; and, if LITER(variable) has a negative value, it also provides VDU plots of residuals from the linear-equation solver.
The questions which appear on the screen when UWATCH and USTEER are set = T in the Q1 file have been placed in the user- accessible Fortran file GXRDQ1. This enables a user to create his own form of interaction or interference with the execution of EARTH.
See also the Encyclopaedia entry 'READQ1'.
Combined with USTEER = T, it allows users to intervene in the solution process, at chosen intervals during the EARTH run.
wbs