2009/02/18 * establish git repository for tracking changes (repository available upon request) 2008/10/30 * incremented version to 2.43 * added more thorough Cobalt bc reader * added ability to test for single vs. double precision in unformatted Cobalt grid files * corrected patch number flood fill operation for Cobalt grids 2008/10/26 * minor revision to screen output for zero triangles 2008/07/15 * needed to move ierr outside of the ifdef USE_TECPLOT in TecplotNPHASE.F90 2008/07/12 * incremented version to 2.42 * add DiRTlib calls to read DCI file for NPHASE 2008/06/01 * incremented version to 2.41 * integrate logic to read full grid, bc, DCI data for NPHASE * modify TecplotNPHASE.F90 to allow for binary Tecplot format 2008/05/30 * added color map logic to NPHASE Tecplot file 2008/05/18 * incremented version to 2.40 * update logic to handle Cobalt files for NPHASE 2008/04/01 * add linear_solver,verbosity,nstage, and restart_from_itm variables to OVER-REL namelist 2007/07/18 * add freeze_distances,time_start,inertial_frame_velocity, and unsteady variables to SolverControls section of OVER-REL namelist 2007/06/19 * incremented version to 2.39 * add uncle.lst as a possible input file format for UNCLE-M 2007/01/30 * correct CalculateForcesAndMoments so that momentum flux is calculated even for rel <= 0 (for OVERFLOW) * allow for -v, -w, -vw, etc. options in CommandLine 2007/01/24 * ifdef out the "flush" command in ProcessPairProcedures for Windows operating system (WIN32) -- per William Chan. Changed extension from .f90 to .F90 to accommodate ifdef 2007/01/11 * added component names to zone titles when tecformat=ascii in WritePatches.F90 2007/01/09 * corrections to --plot-velocity logic for overflow and laminar or iwfn=0 OVER-REl cases * added --prefix option to use RMS-* or SAV-* files rather than RST-* files for the OVER-REL solution data * remove parentheses from (time2-time1) when it appears in an I/O list, as this is evidently an extension to Fortran 95 2007/01/08 * incremented version to 2.38 * added --plot-velocity option to write velocity vector to usurp-surfaces data 2007/01/02 * modified CutOutStrings to work on all types of patches instead of only jmin surfaces 2006/12/08 * added component handling to PreReadUNCLEMM.f90 and allow_fringe command line option for OVER-REL and UNCLE-M 2006/11/28 * added interp_clip to OVER-REL namelist variables * minor revision to DPLR welcome message in DPLRM.f90 2006/11/16 * modified ReadDCI in StorePatchesUNCLE to accommodate changes in the DCI file format 2006/11/14 * minor correction to nalloc counter in StorePatchesOVERFLOW * add cgt directory with scripts to reformat USURP for CGT 2006/11/13 * incremented version to 2.37 * redefined qv array in StorePatchesOVERFLOW such that the triq file will be consistent with changes in overint (extensive rearrangement of StorePatchesOVERFLOW) 2006/11/02 * incremented version to 2.36 * added DPLRM module to handle reading the DPLR input file and various possible DPLR grid formats * edit makemake.pl and Make.sys to account for environment variable FXDR_HOME, which gives the location of the FXDR library * remove extra "nalloc = nalloc + 1" from the initialization section of StorePatchesUNCLEM left over from a deallocate of "done" array removed in version 2.25 or so * add documentation for DPLR users to UpdatesToManual.pdf 2006/10/12 * incremented version to 2.35 (copy 2.33 and merge CGT 2.34) * add UpdatesToManual.pdf to doc/ * remove GuideUpdates.txt and README.Compile from doc/ * modify PreReadUNCLEMM.f90 and StorePatchesUNCLEM.f90 so that grids are assumed to be in GRIDS/ for UNCLE-M 2006/08/24 * update to read_over_rel_nml namelist variables 2006/07/18 * allocate niq,njq,nkq to size nbq in StorePatchesOVERFLOW 2006/07/17 * Change error in StorePatchesOVERFLOW from nb/=nbq to nbq < nb 2006/06/29 * Correct unitialized logical variable "one_edge" in GarbageCollector * do not initialize "first" during declaration in CheckTriangles * move "myfmt" to header of my_cpu_timeM * remove initializations from declarations in CreateTriangles, StorePatchesOVERFLOW, PreReadOVERFLOW, and usurp (pvalue) * add implicit none to sort2 * move initialization of nalloc, etc. to NetAllocM.f90 2006/06/28 * Convert all Fortran procedures to module procedures and combine closely related procedures into common modules 2006/06/27 * Replace InterfaceDefinitions module with module-contained procedures * Nullify result->hole and result->contour in gpc.c upon error 2006/06/26 * Ensure that every USE statement has an accompanying ONLY 2006/06/22 * Complete MapTriQ to generate grid.i.triq w/o performing clipping using use.map 2006/06/21 * Echo command line input in CommandLine * Ignore viscous effects if only one grid plane is provided 2006/06/19 * incremented version to 2.33 2006/05/23 * BUG fix: for CFD-SHIP, UNCLE-M, and N-PHASE, allocate tag_list and ncomp_surfs in PreReadBCS, PreReadUNCLEM, and PreReadNPHASE * BUG fix: change gaminf to qv(iv,6) in calculation of pressure in StorePatchesOVERFLOW.F90 2006/05/22 * incremented version to 2.32 * Add subroutine SplitContours for self-intersecting contours * Switch --verbosity=3 to --trap[=clip,tri] 2006/05/19 * Rewrite ShrinkVertexList using a list of vertex "aliases" * Differentiate between never_skip_tri and never_skip_clip * Add wake cut resolution for i and j surfaces to ShrinkVertexList * Restored original timestamps to files older than Oct 21 2005 2006/05/17 * Added logic for j-surfaces to ShrinkVertexList * For k-surfaces, check wake cut prior to singularity * Change ProcessPairProcedures so that new vertices are not introduced if they are "too close" to existing vertices of the clipping polygon (--watertight option only) 2006/05/05 * incremented version to 2.31 2006/05/05 * incremented version to 2.30 * Added logic in ProcessPairProceduresM.f90 to randomly perturb the points of the clipping polygon if GPC fails and the --never-skip command line option is used. * Added logic to ShrinkVertexList to reduce repeated nodes due to axis, periodic, and wake-cut boundaries. Only works within one given subset (or "patch"). Currently only k-surfaces and part of i-surfaces complete. 2006/05/02 * Added gpc_free_polygon statements to gpc_c_translator_clip.c if gpc_polygon_clip returns with result.num_contours < 0. * Check for e0->next == NULL in gpc.c 2006/05/01 * incremented version to 2.29 * Made tag_list,ncomp_surfs,refc arrays allocatable in PreReadOVERFLOW and PreReadBCin (in the latter, the arrays are simply allocated to a fixed size) * Modified format statement for area decreases in CalculateProjectedAreas 2006/04/28 * incremented version to 2.28 * Fixed TriQuad to account for orientation of quad. * Added check for reversed normals in WriteTriQ. * Modified "original area" calculation in CalculateProjectedAreas to ignore any "original" area that was already blanked out by the user. 2006/04/19 * Added subroutine sort2, invoked from ResolveEdgeIntersections 2006/04/18 * Fixed component names in PreReadBCin to match OVER-REL logic * Bug fix in ShrinkVertexList for J-surfaces (effects any full-surface behavior related to split domains) 2006/04/17 * Rewrote ResolveEdgeIntersections -- removing the linked list and replacing it with a simple allocatable array for the edge intersections is much faster. * Introduced array ltri into CreateTriangles/SelfIntersect to identify self-intersecting contours rather than self-intersecting polygons 2006/04/14 * incremented version to 2.27 * Modify parallel criterion in SelfIntersection.f90. Previously, if abs(denom) was less than 1d-10 and rr=0 or ss=0, then triangulation was skipped. Change to abs(denom) less than 1d-10 and abs(rr) or abs(ss) less than 1d-10. * Significant revisions to WriteTriQ, including addition of capabilities for full surface triangulation with GPC and tecplot binary format. * Add subroutine TriQuad to triangulate unclipped quadrilaterals for full-surface and watertight operations. 2006/03/30 * If a problem occurs in gpc_polygon_clip, causing an early exit out of the subroutine gpc_f_translator, the polygon "opoly" must be deallocated prior to returning in order to avoid a memory leak. * Allocate priority pair array (ipri) and allow them to over-ride USURP's ranking in ProcessPair. 2006/03/29 * incremented version to 2.26 * Distinguish between "full surface" and "watertight". Modify so that "--full-surface" will triangulate the full surface and output grid.i.tri(q), where no attempt is made to resolve hanging nodes, whereas "--watertight" will do the same thing but also try to resolve hanging nodes. Using neither option will skip the full surface triangulation and not write out the grid.i.triq file. 2006/03/08 * Added subroutine CheckTriangles to check for unused vertices and edges belonging to more than one triangle. 2006/02/16 * Corrected a bug in WriteGridIBI.F90 where ibdir(ip) should have been referred to as abs(ibdir(ip)). The bug affects the output to the grid.ibi file when "max" surfaces are used by OVERFLOW users but more likely manifests itself as an allocation error for the arrays xd or yd. 2006/02/15 * Corrected a bug in StorePatchesOVERFLOW.F90 in which the "second layer" of velocities was being accessed as j+istep when it should have been j+jstep. The result was that data could be accessed from out of bounds and was manifesting itself as a zero density in a case where density was 1.0 everywhere. The bug would effect OVERFLOW users with q.save files who were using j-solid surfaces. 2006/02/10 * Add the capability for incompressible flow to StorePatchesOVERFLOW.F90, which is activated by FSMACH =0.0. To FOMOCO users, this is essentially a bug, which could have manifested itself as a "ttinf < 0" fatal error in the calculation of the viscosity using Sutherland's law. 2006/01/27 * Added a generic input capability consisting of a list of the solid patches (generic.bc), a formatted plot3d grid file (generic.grd), and an unformatted file containing the cell-centered iblank info (generic.ib) 2005/12/23 * incremented version to 2.25 * Rewrote most of the coding associated with full-surface triangulation, which is now activated using the --full-surface command line option. Decreased tolerance in GarbageCollection routine, and added test for self-intersecting polygons being passed to Triangle. 2005/09/27 * Added more detail to the "could not find CFD environment" error message in GetUserInput.f90. * Added UNCLE-M capabilities to GetUserInput and UserInputM * Added read_unclem 2005/09/19 * Refactored CalculateVolumes to reduce the metric arrays to one-dimensional temporary arrays. Moved all grid and flow arrays associated with StorePatchesOVERFLOW, ReadGrid, and ReadFlow into VolumeArrays module in order to eliminate grid-block-sized temporary arrays and subroutine arguments. 2005/09/15 * started Revisions file * current version number is 2.24 * Created module VolumeArraysM.f90 to transport xb,yb,zb,volv from StorePatchesUNCLE, StorePatchesOVERFLOW, and BodyForces to subroutine CalculateVolumes. Intent was to avoid stack problems possibly relating to passing large blocks through the argument list. New Makefile to include VolumeArraysM.