Dear OOF users: Preliminary versions of OOF 1.0.8 and PPM2OOF 1.1 are now available for download at http://www.ctcms.nist.gov/oof/BETA/ This version fixes a number of bugs in OOF. Some of the bugs were serious. We apologize for that. In penance, OOF and PPM2OOF have been ported to Sun Solaris2. ------- The following BUGS have been fixed in OOF: * The shear stress, sigma_xy was computed incorrectly, due to a misplaced factor of two. * The force on enslaved node groups was computed incorrectly. * The strains reported by version 1.0.6 and earlier were the two dimensional in-plane elastic strains. In order to do the equilibration calculation for plane strain, the out-of-plane components of thermal strain have to be included as a "correction" to the in-plane strains. This "correction" wasn't removed before reporting the strain. So strains were wrong for elements with thermal expansion in plane strain. * The energy density was computed incorrectly when an element had thermal stresses in plane strain. This was a long-standing known bug. * When mutating, the damage element applied knockdown2 to C_45 twice. * The zimmer elements didn't revert to their original orientation when reset. * The OOF manual states incorrectly that for damage, griffith, griffith2, and zimmer elements that ALL components of C other than C11 are reduced by knockdown2. That was untrue, but the fault was with the manual, not the program. The on-line html and postscript manuals are now correct. Refer to the section on damage elements in the chapter on Element Types. http://www.ctcms.nist.gov/oof/download/Manual/Manual.html * The griffith2 element didn't appear in the menus, but the griffith menu commands created griffith2 elements. When initialized from goof files written by PPM2OOF, the correct elements were created. Binary goof files written by OOF, however, created only griffith elements, while ascii goof files created only griffith2 elements. In all cases, the Element Info dashboard would report that all elements were "griffith". Please note that none of these errors affected the equilibration calculation. The confusion between griffith and griffith2 elements affected the mutation calculation. ------- CHANGES in OOF since version 1.0.6 include: * The bugs listed above have been fixed. * The strains reported by /output/strain and the graphics windows are now the total (geometrical) three dimensional strains, not the two dimensional elastic strains. The next version will let you choose between total, elastic, and stress-free strains. * Goof files containing mutated elements cannot be read by the current version of OOF. This is a consequence of fixing the zimmer element bug. If this is a problem, send mail to oof_manager@ctcms.nist.gov, and we'll help convert your goof files. * Actually, goof files containing damage, griffith, griffith2, and zimmer elements can't be read, because a new "only_once" parameter has been added to these elements. If only_once=true, then an element will mutate only once, even if its yield condition is exceeded a second time. If only_once=false, the element can mutate repeatedly. To make old goof files readable by the new OOF, save the files in ASCII format with the old OOF's /output/grid/ascii command. Then, at the Unix shell prompt, type sed -e 's/zimmer/zimmer only_once=true/' < old.goof > new.goof (replacing "zimmer" by "griffith", et al, and "true" by "false", as appropriate). The new goof file will be readable by the new OOF. * The mutate and equilibrate commands can be interrupted with a Stop button. (The Stop button runs in a separate process, so the Unix ps command will show two OOFs running. The second process uses almost no CPU time and little memory.) * The "applied force" button in the Node Info dashboard has been changed to "force". The sign of the force reported by the dashboard and by the "output force" command has been changed. It's now the force that the boundaries exert on the system, rather than the other way around. * The Element Info dashboard has been slightly restructured, so that you can query the full three dimensional stress and strain on individual elements. * The "all_comps" commands in the /output/individual/stress and /output/individual/strain menus have been replaced with "tensor" commands. These print the full three dimensional tensor, as do all of the other stress and strain output functions. * A few minor interface bugs have been fixed. ------- Changes in PPM2OOF since version 1.0.8 include: * The adaptive mesh can be made to follow the internal boundaries of the material more accurately, without jagged corners. The user interface is still a bit jagged, though. For advice, see http://www.ctcms.nist.gov/oof/download/ppm2oofhints.html The improved adaptive meshing scheme allows the positions of the nodes to change through an algorithm which tries to make the elements homogeneous as well as equilateral. It can be used in concert with the previous adaptive meshing scheme which subdivides triangles with fixed vertices to increase homogeneity. * The mutating elements have been updated to include the "only_once" parameter (see above). * Canceling a "save" window in the image gallery now works. * Minor bug fixes. ------- The OOF web pages are being moved around a bit. The main OOF page is now at http://www.ctcms.nist.gov/oof/ and the download page is at http://www.ctcms.nist.gov/oof/download Questions and comments can be directed to the OOF developers at oof_manager@ctcms.nist.gov or to the OOF mailing list at oof@nist.gov The OOF Team, Steve Langer Craig Carter Ed Fuller