A number of files are used by WIND in the course of a solution. The script file you run to submit WIND jobs will assign all the necessary files to their appropriate Fortran unit numbers, so you should not need to do any of that yourself. Each of the support files is described briefly below.
The input data file is the primary control file for WIND and must be created for each case you want to run with the code. Input data and code options are entered through the use of descriptive keywords in this file. You should observe the following formatting rules in creating the input data file:
The following is an example of a simple input data file:
Geometric Title Flowfield Condition Title Optional Title Freestream static 0.9 14.7 530. 4. 0. Cycles 15 Inviscid End
The computational grid used by WIND for a particular case is stored in the grid file. In this file are stored the (x, y, z) coordinates of all computational grid points, zone coupling interpolation factors, and grid reference and scaling data.
This file was originally referred to as the common grid file,
so named because the file was formatted according to Boeing's Common
File Format (CFF).
Starting with version 4.0 however, WIND also supports grid files in
CGNS (CFD General Notation System) format.
[Detailed information on the CGNS standard may be found at the
CGNS web site
.]
Both common files and CGNS files are binary, and portable to virtually
every hardware platform, except Cray, with no need for explicit data
conversions.
Grid files for WIND may be created by several mesh-generation codes in either common file or CGNS format. If necessary, the cfcnvt utility may be used to convert a variety of other formats, including PLOT3D xyz format, to common file format.
Zone coupling, reference, and scaling data are added to common grid
files using the GMAN program.
Common grid files are also used in the
CFPOST post-processing package.
Neither GMAN nor CFPOST currently support CGNS files, however.
Keywords: CGNSBASE
The flow file contains the computed flow field. For Navier-Stokes and Euler solutions, the file contains density, momentum, and energy data, and, for viscous solutions, turbulence data. The flow file also records the current solution cycle number to allow the file to be used for solution restarts.
Like the grid file, the flow file was originally referred to as the common flow file, but may now be written in either common file or CGNS format. Several graphical post-processing programs are able to read both common files and CGNS files.
The CFPOST post-processing package
may be used with common flow files to produce other files for
post-processing and/or to create flowfield plots directly.
CFPOST does not currently support CGNS files, however.
Keywords: CGNSBASE
The Global Newton file is a copy of the flow file used for Global Newton time marching. It holds the current sub-iteration of the time step. It is only created when Global Newton time marching is active, and is a scratch file. If the code aborts, the solution restarts from the conditions in the flow file, at the beginning of the time level, so the Global Newton file does not need to be saved.
WIND's boundary data file is used during solution restarts, and results in smoother restarts, especially with higher-order boundary coupling. The file acts as a buffer for the transfer of zone coupling information and a holding bin for data needed throughout a WIND run but not stored in the grid or solution file.
The time history file is a common file which stores data
resulting from the use of WIND's history tracking capability.
The file contains a lookup table corresponding to the range of
computational indices tracked during the run, time stamp data, and
flowfield data for each time stamp.
Upon completion of a time history run, the auxiliary program
thplt may
be used to view the contents of the time history file.
Keywords: HISTORY
The purpose of WIND's list output file is to echo the input from the input data file, track convergence and integration results, record CPU/job statistics, and log code error messages. The auxiliary program resplt may be used to extract convergence and integration data from the list output file and create a GENPLOT-style data file, which may be plotted with the plot command in the CFPOST post-processing package. For parallel-processing runs, this file also contains messages relating to the PVM system and to the allocation and operation of slave processors. Convergence data is written to this file for each of the equation sets solved during the run, including the Euler, Navier-Stokes, and non-algebraic turbulence model sets.
The stop file is used to stop WIND execution cleanly in the middle of a CFD solution. Although a solution may be stopped simply by killing the WIND process from the system queue, doing so will not ensure a clean update of all zonal flowfield data to the flowfield file. The stop file - named NDSTOP - provides a means to cleanly shut down a running solution, completing the current cycle or zone, performing zone coupling, updating the flowfield file, and removing all symbolic links to Fortran file unit numbers.
To stop the code, you must place one of two words in the stop file in the directory in which the WIND job is running. The word STOP in the stop file signals WIND to complete the current cycle (at the end of the last zone) and exit. In single-processing mode, you may also use the word STOPZONE in the stop file to stop WIND after completing the zone currently in memory. [In multi-processing mode, STOPZONE does the same thing as STOP.] You may then restart the code in your next run, starting at the next zone or back at zone one. Regardless of the existence of this file, WIND will stop if the requested number of cycles has been computed, or if your solution has converged.
During the clean-up procedure at the end of the job, the NDSTOP file is automatically removed.
Note that the directory in which the WIND job is running, where the NDSTOP file must be, is not necessarily the one you were in when the wind command was issued. If you don't use the -runinplace option to the wind script, the job will be run in a remote directory. The "root name" of the remote run directory may be specified using the -runroot option, or in response to a command line prompt. The default for the "root name" of the remote directory is /tmp. The full name of the remote directory will be rootname/userid/basename.scr, where rootname is the "root name" described above, userid is your userid, and basename is the base name of your .dat file.
On a Unix system, you might submit a simple "at" job for a later time as follows:
at 0530 monday echo STOP > NDSTOP ^D
At 5:30 AM on the next Monday, the system would create the NDSTOP file
with the word "STOP" in it.
Note that this will not stop the run exactly at 5:30 AM; WIND must
still complete the current cycle, which may take an hour or more
for large cases.
Keywords: RESTART
WIND can read temperature and transition files, specifying the
temperature or transition to turbulence on any boundary surface,
that were created with the older tmptrn utility.
This option is included for backward compatibility with WIND 2.0,
and is only needed for an initial (i.e., non-restart) run.
The wind script
copies the files to the run directory, and WIND
opens the files directly (unit 45) using the file names specified
with the TTSPEC keyword.
New applications should use the latest
tmptrn utility
to write the temperature/transition data into the flow
(.cfl) file directly.
Keywords: TTSPEC
The generalized chemistry files contain all the information WIND requires to compute a general chemistry mixture. A chemistry file has a header line, followed by three sections containing thermodynamic coefficients, finite rate coefficients, and transport properties.
The header line must contain a single parameter, ispec, specifying the type of reaction rate and the format for the finite rate data. The format for the header line is:
ispec ISPECwhere
| ispec | Reaction Rate and Format | ||
|---|---|---|---|
| 1, 4, 100, 110 | Forward and backward rates from equilibrium constant polynomial, Format 1 | ||
| 3, 130 | Forward and backward rates, Format 2 | ||
| 120 | Forward rate only, Format 3 |
This section of the chemistry data file contains the information
necessary to compute the specific heat cp for each species.
The value of cp is defined by a series of polynomials, each valid
within a defined temperature range, of the form:
The format for this section is:
THERMODYNAMIC COEFFICIENTS Title, line 1 Title, line 2 ns NS Information defining species 1 ... Information defining species nswhere ns is the number of species.
For each species, the information defining the species and for computing cp, etc., is read as follows:
read (26,100) name,ncurves,ds1,hm1,ds2,hm2,unk1,unk2,con,molwt
do 10 ic = 1,ncurves
read (26,110) t1,t2,(a(i,ic),i=1,5),shf(ic),F(ic)
10 continue
100 format (a8,7x,i5,4x,2(a2,f3.0),16x,f10.1,5x,f10.3)
110 format (5e15.5)
where
| Variable | Definition | ||
|---|---|---|---|
| name | Name of species | ||
| ncurves | Number of temperature segments in cp vs. T curve | ||
| ds1 | Name of first constituent atomic species in name | ||
| hm1 | Number of atoms per molecule of ds1 | ||
| ds2 | Name of second constituent atomic species | ||
| hm2 | Number of atoms per molecule of ds2 | ||
| unk1 | Unknown (0.0) | ||
| unk2 | Unknown (0.0) | ||
| con | Coefficient defining number of degrees of freedom, defined as (translational degrees of freedom + rotational degrees of freedom)/2. | ||
| molwt | Molecular weight | ||
| t1,t2 | Beginning and end of temperature range for curve segment ic | ||
| a(1-5,ic) | Polynomial coefficients a1-a5 for curve segment ic | ||
| shf(ic) | Heat of formation of species, in J/kg | ||
| F(ic) | Gibb's free energy (for future use in gas constant) |
This section of the file contains the reaction rate information. There are three possible formats for this section, depending on the value of ispec.
This option may be used for any chemically reacting flow.
Information is specified defining the forward
reaction rate kf and the equilibrium constant K.
The rates themselves are computed using the equations
The various parameters and polynomial coefficients are read from the chemistry data file, in the following form:
FINITE RATE COEFFICIENTS Title, line 1 Title, line 2 nreq ndeq NREQ, NDEQ tfrmin TFRMIN Information defining dissociation reaction 1 ... Information defining dissociation reaction ndeq Information defining exchange reaction 1 ... Information defining exchange reaction nreq - ndeqwhere nreq is the total number of reactions (dissociation + exchange + ionization), ndeq is the number of dissociation reactions (i.e., have a third body), and tfrmin is the temperature in K below which no reactions occur.
For each reaction, the information defining the reaction and for computing the reaction rates is read as follows:
read (26,100) r1,r2,p1,p2,S,D,(a(i),i=1,5),nthd
do 10 i = 1,nthd
read (26,120) spec,C
10 continue
100 format (4(a5,3x),2e12.4/5e12.4,i4)
110 format (16x,a8,e12.4)
where
| Variable | Definition | ||
|---|---|---|---|
| r1 | Name of reactant 1 | ||
| r2 | Name of reactant 2. (Leave blank for dissociation reactions.) | ||
| p1 | Name of product 1 | ||
| p2 | Name of product 2 | ||
| S | Temperature exponent S in equation for kf | ||
| D | D/KB in equation for kf | ||
| a(1-5) | Polynomial coefficients a1-a5 in equation for kf | ||
| nthd | Number of third bodies, implies nthd reactions (one for each third body). Always set nthd = 0 for exchange and ionization reactions. | ||
| spec | Name of third body reactant (not needed if nthd = 0 or 1) | ||
| C | Third body coefficient for species spec as third body, or average if nthd = 0 or 1 |
This option is similar to the previous one, except that the
forward and backward reaction rates are computed separately, using the
equations
The various parameters are read from the chemistry data file, in the following form:
FINITE RATE COEFFICIENTS Title, line 1 Title, line 2 nreq ndeq NREQ, NDEQ tfrmin TFRMIN Information defining dissociation reaction 1 ... Information defining dissociation reaction ndeq Information defining exchange reaction 1 ... Information defining exchange reaction nreq - ndeqwhere nreq is the total number of reactions (dissociation + exchange + ionization), ndeq is the number of dissociation reactions (i.e., have a third body), and tfrmin is the temperature in K below which no reactions occur.
For each reaction, the information defining the reaction and for computing the reaction rates is read as follows:
read (26,100) r1,nr1,r2,nr2,p1,np1,p2,np2,Sf,Df,Cf
read (26,110) Sb,Db,Cb
100 format (4(a5,f3.1),3e12.4)
110 format (32x,3e12.4)
where
| Variable | Definition | ||
|---|---|---|---|
| r1,nr1 | Name and number of reactant 1 | ||
| r1,nr2 | Name and number of reactant 2. (Leave blank for dissociation reactions.) | ||
| p1,np1 | Name and number of product 1 | ||
| p2,np2 | Name and number of product 2 | ||
| Sf,Sb | Temperature exponents Sf and Sb in equations for kf and kb | ||
| Df,Db | Df/KB and Db/KB in equations for kf and kb | ||
| Cf,Cb | Third body coefficients Cf and Cb in equations for kf and kb |
This option is only for global 1- or 2-reaction chemistry, and only for
forward reactions.
It is a quick method for simulating detonation problems, for example, in
which reactions proceed only forward to completion.
The forward reaction rates are computed using the equation
The various parameters and polynomial coefficients are read from the chemistry data file, in the following form:
FINITE RATE COEFFICIENTS Title, line 1 Title, line 2 nreq ndeq NREQ, NDEQ tfrmin TFRMIN Information defining reaction 1 ... Information defining reaction nreqwhere nreq is the total number of reactions (dissociation + exchange + ionization), ndeq is the number of dissociation reactions (i.e., have a third body), and tfrmin is the temperature in K below which no reactions occur.
For each reaction, the information defining the reaction and for computing the reaction rates is read as follows:
read (26,100) r1,nr1,r2,nr2,p1,np1,p2,np2,S,D,(a(i),i=1,5),nm
read (26,110) Cb1,Cb2,C
100 format (4(a5,f3.1),2e12.4/5e12.4,i4)
110 format (2f8.3,8x,e12.4)
where
| Variable | Definition | ||
|---|---|---|---|
| r1,nr1 | Name and number of reactant 1 | ||
| r2,nr2 | Name and number of reactant 2 | ||
| p1,np1 | Name and number of product 1 | ||
| p2,np2 | Name and number of product 2 | ||
| S | Temperature exponent S in equation for k | ||
| D | D/KB in equation for k | ||
| a(1-5) | Place holder for use with future ispec options | ||
| nm | Place holder for use with future ispec options | ||
| Cb1,Cb2 | Exponents on concentrations of reactants 1 and 2 | ||
| C | Third body coefficient C in equation for k |
This section of the chemistry data file contains the coefficients used
to compute the laminar viscosity and thermal conductivity.
The formulation is based on Wilke's law which uses Sutherland's law
individually for each species.
Thus for each species, there are equations of the form:
The various parameters and polynomial coefficients are read from the chemistry data file, in the following form:
TRANSPORT COEFFICIENTS Title, line 1 Title, line 2 Information for species 1 ... Information for species ns
For each species, the information is read as follows:
read (26,100) name,ncurves,t1,t2,mu0,t0mu,smu
read (26,110) k0,t0k,sk
do 10 ic = 2,ncurves
read (26,120) t1,t2,mu0,t0mu,smu
read (26,110) k0,t0k,sk
10 continue
100 format (a8,2x,i5,2f10.3,1x,3e12.4)
110 format (36x,3e12.4)
120 format (15x,2f10.3,1x,3e12.4)
where
| Variable | Definition | ||
|---|---|---|---|
| name | Name of species | ||
| ncurves | Number of temperature segments in mu and k vs. T curve | ||
| t1,t2 | Beginning and end of temperature range for curve segment ic | ||
| mu0 | Reference viscosity | ||
| t0mu | Reference temperature for mu equation, K | ||
| smu | Reference temperature offset for mu equation, K | ||
| k0 | Reference conductivity | ||
| t0k | Reference temperature for k equation, K | ||
| sk | Reference temperature offset for k equation, K |
Unit numbers 15, 21, and 55 are reserved for proprietary features, and should not be used by WIND developers.