Contents

1. About PINTO
2. What PINTO does
3. What PINTO cannot do
4. The interaction between PINTO and other PHOENICS programs
5. About Q1PIN
6. How to use PINTO

1. About PINTO

PINTO, the PHOENICS INTerpolation prOgram, is a stand-alone code of the PHOENICS program that can be used to transfer the results calculated by EARTH on to finer or coarser grids.

PINTO proceeds by interpolation on to the new grid the values of the variables saved by EARTH in the PHI file for the old grid. These variables are those for which the user requested, in the Q1 file, a solution (through SOLVE) or the allocation of storage space (through STORE).

Thus, PINTO can be used to:

• Interpolate the results obtained for a coarse grid on a finer grid, so atht these can be used as initial guesses for the computation (PIL command RESTRT).
• Interpolate the results obtained for a grid on to one having a different dimensionality, as initial guesses for the new computation.
• Interpolate the results obtained for a grid on to a finer grid, prior to analysis through PHOTON or AUTOPLOT. This technique can be used to increase the resolution of the plot (eg. A denser vector field in PHOTON).
• Interpolate the results obtained for a grid on to a coarser grid, prior to analysis through PHOTON or AUTOPLOT. This 'thinning-out' of the results will decrease the processing time needed by the graphics program, and also the storage requirements.

PINTO takes its input from a special Q1 file (Q1PIN), which accepts a subset of instructions of the PHOENICS Input Language (PIL) with some extensions.

As an aid to the user, PINTO records its transaction with the command-file Q1PIN in a log file (PINLOG), in which syntactical errors made in the command statements are recorded. The coordinates of the new grid are also written to this file, both for checking and for copying into the Q1 file that will be used for the fine-grid calculation.

2. What PINTO does

• Interpolation: PINTO performs, according to the dimensionality of the problem, a linear, bi-linear or tri-linear interpolation of the results on the old grid on to the new grid. It is also equipped with an option that allows the assignment of the value at the nearest point in the old grid, without interpolation.
• Grid systems: PINTO can handle all the grid types available in PHOENICS viz Cartesian, cylindrical polar and curvilinear (body-fitted) coordinates in PHOENICS.
• Grid manipolation: PINTO provides simple commands for defining the grid by doubling, trippling, quadrupling, etc the number of cells in each direction. It also allows the repositioning of the origin, so that the new grid can, for instance, surround the old grid. PINTO will also allow the dimensionality of the new problem to differ from the dimensionality of the old one.
• File handling: PINTO can handle both sequential and direct-access PHI-files; and it can produce, as output, either type of file, regardless of the input type. PINTO can therefore be used for file-type conversion.

3. What PINTO cannot do

• Interpolation between different types of grid. For instance, you cannot use PINTO to interpolate results from a Cartesian grid on to a BFC grid.
• Generation of XYZ files for the new BFC grid; these have to be generated by the user, through the Q1 file for the new problem.
• Although PINTO interpolates BFC grids, the interpolation scheme is a simplified one, in that it assumes that the grid is a uniform one.
• Interpretation of some of PIL commands and special PATCH names in Q1 files. For example, PINTO does not recognise semi-colon, ';' as continuation sign, REGIONS and % sign in the PATCH command.

4. The interaction between PINTO and other PHOENICS programs

PINTO will normally be used in combination with other PHOENICS programs, as follows:

1. The user will have a Q1 file that performs a coarse-grid computation. After running the SATELLITE and EARTH, a PHI file with the result for the coarse grid will be generated.

2. The user will then write a Q1PIN file for the interpolation of the coarse-grid results into the fine grid. The Q1PIN file will have settings to define the new grid and to control the format and name of the output files.

3. PINTO will then be run by entering the command runpin (or runpins for Salford version). On execution, PINTO will read in the Q1PIN file and will generate a PHI-like file with the new file values for the variables on the refined grid. APINLOG file will also be generated, containing a description of the new grid and any error message.

   After a PINTO run, it is essential to inspect the PINLOG file for errors. PINTO does not write any error messages to the screen, and the contents of this is therefore the same for successful and erroneous runs.

4. The new PHI file can then be used by the user in several ways:
   • As the input file for the graphics program, PHOTON.
   • As a restart file for a computation on the finer grid. In this case, a Q1 file will have to provided by the user, describing the problem in the same new grid; and the SATELLITE and EARTH will have to be run.

A flow chart of PINTO is provided below:

                                                                                 Q1
                                                                                |
                                     SATELLITE  
                                         |
                                       EARDAT
                                         |
                    PREFIX              EARTH
                      |                  |                                 
          PINCON      |     Q1PIN       PHI (or PHIDA)
            |         Ý       |          |                                     |       
     ---------------------------------------------
     |                PINTO                      |
     ---------------------------------------------
          |                      Ý                                                    
      NPHI (or NPHIDA)        PINLOG       
          |
     EARTH (or PHOTON)

5. About Q1PIN

Q1PIN, input file for PINTO, is similar in concept to the Q1 file used by the SATELLITE, which the reader is assumed to be familiar with.

Q1PIN is divided into 3 groups, as follows:

Group 1: Grid specification
Group 2: Interpolation controls
Group 3: File handling

The set of commands used by PINTO comprises some PIL commands, and some additional ones which are unique to PINTO. As in the PIL, the group structure of PINTO has been conceived just as an aid-memory for the user: PINTO commands do not need to be in the group they belong to.

Unlike the Q1 file, the Q1PIN file does not need to have a special first line (TALK=T;RUN=1,1) of the Q1 file; but the end of the Q1PIN file must be flagged with a STOP command.

Finally, PINTO commands must start in the first or second column of the file; and all other lines are treated as comments.

PINTO language is much more limited than the PIL. Please refer to TR218 for what commands will be recognised by PINTO.

6. How to use PINTO

This section describes how to use PINTO to perform the file handling and the grid interpolation.

1. File handling.

The steps in the execution of the file handling are:

• to activate runsat (or runsats for Slaford version) in order to load a library case;
• to set Compress = off in the cham.ini file in the directory \phoenics\d_allpro;
• to activate runear (or runears for Salford version); (it is assumed that PHIDA=F in the prefix file );
• to prepare a Q1PIN file where LDA=T is set;
• to activate runpin (or runpins for Salford version) which will produce a direct-access file, NPHIDA;

Similarly, a PHIDA file produced by EARTH, when PHIDA=T is set in prefix, can be converted to a formatted sequential file, NPHI by modifying the Q1PIN file and replacing LDA=T by LFS=T.

2. Grid interpolation.

The steps in the execution of the grid interpolation are:

• to activate runsat (or runsats for Salford version) to load library case 251;
• to activate runear (or runears for Salford version) to produce the result, PHI file;
• to prepare a Q1PIN file as follows:
  • Start of the listing
  • This Q1PIN is to extend a 2D case to 3D case for case 251
  • The following grid settings must be used when RESTRT=T
  • Use GRDPWR for the x-direction grid specification. GRDPWR(X,,1.813E-02,1.0)
  • * Full length of channel = 0.762
  • Method of pairs is used to set the new y-grid.
    NY=20
    YFRAC(1)=-5.0
    YFRAC(2)=0.03
    YFRAC(3)=10
    YFRAC(4)=0.07
    YFRAC(5)=5.0
    YFRAC(6)=0.03
  • new z-grid is also set by the GRDPWR command. GRDPWR(Z,30,3.626E-02,1.0)
  • To ensure that field values retain the same boundary
  • values set:
    BNDSET(ALL)=T
    STOP
  • End of the listing

  The above Q1PIN will interpolate the values obtained for the two-dimensional grid on to the three-dimensional one.

• to activate runpin (or runpins for Salford version) which will produce a NPHI file;
• to rename NPHI to PHI;
• to load library case 278 and to modify this case so that it restarts from the PHI file generated by PINTO.