CFD Utilities Files

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

   program capsule_spokes

!  Description:
!
!     This utility is intended to complement the POLAR_INTERP scheme by defining
!     body points along capsule heat shield (or aft body) spokes at which the
!     NEQAIR radiation solver is intended to be run.  POLAR_INTERP turns the
!     radiative heat fluxes at those body points into a structured surface data-
!     set by interpolating nonlinearly across spokes in the azimuthal direction
!     after regularizing the spoke point counts.  This in turn may be interp-
!     olated to a desired mesh (SURFACE_INTERP) or contour-plotted directly.
!     [The same scheme is also applicable to mass loss data gathered with
!     material ablation solver FIAT in place of NEQAIR.]
!
!  Assumptions:
!
!     >  The capsule geometry smooth outer mold line is axisymmetric.
!     >  The analysis is to be at nonzero angle of attack (NOT axisymmetric).
!     >  The supplied portion of a generatrix consists of the points desired
!        along the vertical symmetry plane slice.  It is expected to be confined
!        to either the forebody or to the aft body--not both, in order to avoid
!        complications involving duplicate end points common to all spokes.
!     >  Further spokes are defined simply by rotating this generatrix.
!     >  The first spoke (input generatrix fragment) should be at the 12 o'clock
!        position, with the centerline point first        (as for POLAR_INTERP).
!     >  The spokes may proceed either clockwise or anticlockwise ("    "    ").
!        Angles 0 through +180 degrees in the one-line file mentioned below
!        produce the right-hand half of a whole body or forebody surface.
!     >  If the generatrix represents the aft body (only), the outputs will
!        start with the centerline point.  I.e., the input order along the
!        generatrix may be reversed if necessary.  Positive angles produce the
!        right hand half as for the forebody.
!     >  The spokes are traditionally NOT uniformly spaced.  Therefore, an
!        angular distribution defining the further spokes is read as input.
!
!  Inputs:
!
!     A few prompts suffice for file names and padded mesh dimensions.
!
!     The generatrix should look something like this (read to EOF):
!
!        x1  z1  [0.]  or  x1  0.  z1  ! 2 or 3 columns; 
!        x2  z2  [0.]      x2  0.  z2  ! if 3 columns, the column with smallest
!        x3  z3  [0.]      x3  0.  z3  ! data range is suppressed and the other
!         :   :  [ :]       :   :   :  ! two are treated as x and y.
!
!     If the first point is not at y = 0, it is reversed in place to make the
!     first point the centerline point.  If the resulting first point is still
!     not at y = 0, this is a fatal error because the spokes produced here are
!     supposed to have a common center point (that is suppressed in all but the
!     first spoke to avoid unnecessary NEQAIR runs; POLAR_INTERP handles the
!     suppressed points).
!
!     The angular spoke coordinates should appear in a one-line file something
!     like this (read as reals):
!
!        0 5 10 15 25 35 ... 155 165 170 175 180
!
!     See assumptions above for the meaning of positive angles here, in degrees.
!
!  Output Format:
!
!     Achieving compatibility with the PREPARE_NEQAIR_DATA scheme that will
!     probably be employed prior to the POLAR_INTERP step is a little awkward.
!     The solution is to number the points of all spokes from 1 through N where
!     N is the total number of points from all spokes.  The main output of
!     'x  y  z' coordinates produced here then serves as the body point list
!     expected by PREPARE_NEQAIR_DATA.  Then NEQAIR results will appear in
!     directories /LINE-1, /LINE-2, ..., /LINE-N.
!
!     Using the command "grep Total LINE-*/neqair.out > total.txt" (say) will
!     produce a total.txt file that starts like this:
!
!        LINE-1/neqair.out:Total radiative heating ... =     2.468401 W/cm2
!        LINE-11/neqair.out:Total radiative heating ... =     3.170634 W/cm2
!        LINE-12/...
!        ...
!        LINE-19/...
!        LINE-2/...
!        LINE-20/...
!
!     In order to obtain the desired order, LINE-1, LINE-2, ..., another utility
!     by the present author, named SORT_ROWS, can be used with 'LINE-' as the
!     prefix prompted for.  This will order the heat fluxes to match the list
!     of body point coordinates (for all spokes) written here.  Then some other
!     utility such as COLUMNEDIT (same author) can append the heat fluxes to the
!     table of coordinates.
!
!     This program also writes the POLAR_INTERP-type header that needs to be
!     inserted at the top of the combined "x y z f" table after the function
!     data (NEQAIR or maybe FIAT) have been appended to the body point coords.
!     generated here.  The file names are hard-coded as 'header.dat' and
!     'body-points.dat'.
!
!     The gyrations between running NEQAIR and running POLAR_INTERP may easily
!     be scripted.
!
!     Advantage is taken of POLAR_INTERP's option to avoid repeating NEQAIR runs
!     at the origin of all spokes.
!
!  History:
!
!     08/09/2017  D.A.Saunders  Initial design.
!     08/18/2017    "      "    Initial coding and testing after a hiatus.
!     08/22/2017    "      "    Diagnostics on input data are now written.
!                               Subroutine RVERSE has 3 arguments, not 2!
!     04/04/2019    "      "    Application to a forebody with a 3-column
!                               generatrix uncovered glitches that should
!                               have been found with more thorough testing.
!                               For the forebody/right-half case, positive
!                               input angles actually produced spokes on the
!                               left half.  Positive input angles need to be
!                               negated to produce the spokes on the right
!                               half for both forebody and aft body.  If
!                               negative angles are input, they are still
!                               negated and should produce left-half spokes.
!
!  Author:  David Saunders, AMA, Inc. at NASA Ames Research Center, CA
!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!