Pointwise Reliable People. Reliable Tools. Reliable CFD Meshing. The Connector newsletter Pointwise Facebook Pointwise GitHub Pointwise Instagram Pointwise LinkedIn Another Fine Mesh blog Pointwise Pinterest Pointwise Twitter Pointwise YouTube Y+ Calculator



Gridgen Version 13.2.1 - 22 September 1998

Gridgen Version 13.2.1 is primarly a maintenance release. However, a new feature has been implemented through the re-work and repair of the hyperbolic grid methods for domains and blocks. These solvers allow you to quickly create a structured mesh by marching outward from an initial grid.

SPR 13.0063

Improved the stability of manner in which the elliptic grid method's foreground control functions are calculated when applied to an edge adjacent to an edge with a floating bc.

SPR 13.0067

Added more improvements to the IGES reader in order to be able to import trimmed surfaces from IGES files exported from CATIA that contain severe syntax and/or geometry problems.

SPR 13.0070

Gridgen now properly exports GCL files containing the "smalltext" variable.

SPR 13.0073

Fixed a problem with the VCR button's hot keys. Hot keys are now consistent with documentation.

SPR 13.0074

There was a problem with the database entity count in the Diagnostic window which has been fixed.

SPR 13.0075

Added a tolerance check to the handle and anchor points used for entity scaling so that scaling is disabled if the two points are within the same point tolerance value. Prior to this fix, very minor differences in the handle and anchor points would cause very large scale factors.

SPR 13.0076

When movement restrictions are set to the screen-y coordinate during creation of a database constrained connector segment, movement of the 3D cursor is now controlled by horizontal mouse movement in order to be consistent with all other modes (it used to require vertical mouse movement).

SPR 13.0077

Changed structured domain grid initialization so that pole domains on databases are initialized using the Pole method and NOT Parametric TFI. The option to manually apply Parametric TFI to pole domains on a database still exists.

SPR 13.0078

Fixed a problem in the WIND interface that caused Arbitrary Inflow BCs to be written as Diffuser BCs.

SPR 13.0079

Made two changes to the manner in which Gridgen checks for database swapping during Gridgen file import. First, an error was fixed that caused the swap to be skipped if the node tolerance in the file was bad. Second, added checks to be sure that parametric coordinates in the Gridgen file were trimmed to the parametric limits of the new database prior to evaluation.

SPR 13.0080

In the Examine command, changed the logic used by the page-up and page-down buttons used to control the scan plane location. Since these buttons increment/decrement the scan plane by 10 indices, problems had occurred on grids with less than 10 planes in any direction.

SPR 13.0082

Added the letter "f" to the hiliter movement buttons in the Examine command to emphasize that these buttons move the hiliter to the location of the maximum and minimum diagnostic function values. Conversely, if no function is currently calculated, these buttons are greyed-out. Also fixed a problem that prevented the hiliter from moving to an entity other than the focus entity when these buttons were used.

SPR 13.0083

Decreased the memory requirements of the unstructured tet solver.

SPR 13.0084

Made many improvements to the IGES reader to be able to import trimmed surfaces from IGES files from SolidWorks that contain severe syntax and geometry errors.

SPR 13.0086

Found and fixed an error in the hiliters used during domain and block splitting that had resulted in the hiliter being restricted to a portion of the domain/block.

SPR 13.0088

Image manipulation in the Examine command with a crinkle surface cut through a tet mesh with a diagnostic function displayed used to cause a crash. This has been fixed.

SPR 13.0089

Made sure that the diagnostic window tally of entities only counted active entities and not those with temporary status, such as those currently being created.

SPR 13.0091

Fixed a problem in the application of the hyperbolic tangent distribution function for connectors that sometimes caused crashes on DEC and Win32.

SPR 13.0092

Fixed a memory management bug in the control function initialization routine for structured blocks that sometimes resulted in a crash when two blocks were joined and then run in the elliptic solver.

SPR 13.0093

Fixed a bug in Ortho TFI for structured domains.

SPR 13.0094

A data type error for Polar TFI axis points was fixed. This bug had caused grids with Polar TFI'd domains to crash when an automatic regeneration was triggered.

SPR 13.0095

The Apply to All button was having no effect during the specification of Ortho TFI in the structured solver. This has been fixed.

SPR 13.0096

A structured block's volume grid points are now updated consistently with the domains used on its faces when the algebraic solver is used to reinitialized any of those domains. Prior to this fix it was possible for the volume grid to become inconsistent with its faces.

SPR 13.0098

Changed a consistency problem with hot keys during connector segment editing. The "original location" menu button now has the "l" hot key.

SPR 13.0099

Diagnostic functions and the hiliter have been fixed for 2D structured blocks in the Examine command. 2D unstructured blocks may NOT be examined.

SPR 13.0100

Added a missing "zone velocity" card to the FDNS bc file. Also, export of FDNS data has been split into separate commands for grid and bcs. Changed the default extension for the grid file to .grd.

SPR 13.0101

Fixed a picking bug for sorted lists that sometimes resulted in entities other than the ones picked being used by the command.

SPR 13.0102

Deactivated the Auto Next Edge feature when creating the initial edge of a hyperbolic domain.

SPR 13.0103

Changed the convergence criteria on the secant method used to relate a connector's arclength and the artificial "s" parameter so that splitting db-constrained domain would not result in a complete mess due to failure of the secant method to converge.

SPR 13.0104

Fixed several errors associated with IGES subfigure import when the subfigure contained trimmed surfaces. Also, fixed a bug in the import of bounded surface domain curves and the scale factor in subfigures.

SPR 13.0105

Improved the manner in which consecutive duplicate points are eliminated from segment files.

SPR 13.0107

Added logic so that Gridgen could detect periodic, db-constrained, unstructured domains in order to increase the odds of automatic triangulation on the db.

SPR 13.0108

A logic problem in the determination of a suitable database surface for automatic domain initialization has been fixed that previously caused crashes.

SPR 13.0110

Added further improvements to Gridgen's ability to detect an unstructured grid constrained to a database surface that closes onto itself (eg, surface of a circular cylinder). Also, made sure that the surface deviation is calculated each time prior to application of the unstructured solver for domains to make sure the values are consistent.

SPR 13.0113

Fixed the domain and block hyperbolic solvers.

SPR 13.0115

Changed the format of the field used to display the total number of block grid points in the Examine command to handle 9 digits (ie, 999,999,999 points).

SPR 13.0116

Toggle buttons were added to the "Assemble Edges" method of unstructured domain creation to allow meshes to be created without any interior points. This allows the user to bypass the sometimes time consuming task of domain initialization, in cases when it is known that the domain will have to be reinitialized again with different parameters.

SPR 13.0117

A minor typo in the control function calculation on structured block faces sometimes caused the structured block solver to run forever without completing a single iteration. This has been fixed.

SPR 13.0119

While Unix allowed Gridgen to compute the square root of -1, NT was less forgiving. This has been fixed.

SPR 13.0123

When .rgb image files were saved from Gridgen on an SGI O2 the colors were bogus. Fixing the offset and shift values for the color map eradicated this bug.

SPR 13.0125

Fixed another math error that Unix had been letting Gridgen get away with: taking the acos of a number greater than 1. NT caught this so that we could fix it.