The new version of the library is issued 2015/04/19:
Correction of a bug in the conversion of Nastran MAT8 material to ClaMat object. (The “STRN” parameter was not taken into account.)
the “LaminateAnalysis.xls” example using COM component has been slightly improved.
The major modification in this issue of FeResPost is that it is now distributed under the Lesser General Public License. Practically, this means that it is now allowed to distribute proprietary applications linked with the library. (See Appendix X.F.4 for more information.)
The new version of the library is issued 2015/08/26:
Addition of an example with ruby extension illustrating the access to optimization results (RUBY/EX21). The manual is modified accordingly (section IV.2.7).
In “NastranDb” class, the “writeNastranCard”, “writeNastranCards” and “writeNastranCardToVectStr” have been modified. The “cardName” String argument has become optional. The User Manual and corresponding examples have been adapted.
Addition of several “raw XDB access” methods to the “NastranDb” class. The methods are described in section III.1.1.11 and illustrated by examples in section IV.2.8.
The new version of the library is issued 2015/11/22:
Correction in the reading of shell element stresses at mid-thickness. The stresses are associated to “NONE” layer, and no longer to “Z0”. (This has been done to ensure coherence with other shell results reading. User manual is updated accordingly.
The reading of Nastran XDB result files produced with option “DBCFACT=4” has been nearly completely re-programmed. It seems that all the results that can be read with “DBCFACT=0” can also be read with “DBCFACT=4”.
Correction of a major bug in the “modifyRefCoordSys” method of “Result” class. (The modifications from spherical, or to spherical coordinate systems were completely erroneous before version 4.3.2.)
The new version of the library is issued 2016/01/01:
Correction of a few minor bugs:
FIXED: in GMSH output, no output was produced when one or several Result keys failed.
FIXED: program crashed when operands were of wrong type.
FIXED: wrong error message in Result extract operation.
Small optimization of XDB files indexed with hash keys.
Addition of a “BulkOnly” Logical parameter to the “readBdf” method in “NastranDb” class.
Addition of method “writeNastranCardsToVectStr” to the “NastranDb” class.
Addition of methods “insertCard” and “insertCards” to the “NastranDb” class. (These methods allow to modify the model defined in the “NastranDb” object.)
Addition of “atan2” dyadic function in the “Post” Module.
The “Post.pow” method now also allows a Float as first argument.
The “Post.vectProduct” method allows vectors (Arrays of three Float) as first or second argument. Note also that the “vectProduct” operation in ”deriveDyadic” method of “Result” class has also been modified and also allows an Array of three real values as second argument.
Addition of the “setComponent” method to the ‘Result” class.
Some of the “xdb” methods in NastranDb class have been transformed in ruby singleton method, or in .NET static methods. (It does not change the bahaviour of COM component.)
Addition of the special node IDs “CbarGrdA” et “CbarGrdB” to allow the reading of moments and stresses in CBAR elements that are not defined in the dataBase. (See Table III.1.6.)
A few corrections in the examples, and in the user manual.
The new version of the library is issued 2016/03/06:
Correction of a bug in “getNodeCoords” in “NastranDb” class. Now the coordinates wrt coordinate system 0 can be obtained.
Correction of a bug in “Post.cmp” method.
Correction of a bug in the writing of GMSH “mesh” files.
Addition of support for the unisgned int and unsigned long int in NastranDb raw accessor to XDB tables (“u” and “U” conversions).
Some modifications on the manual and the distributed binaries:
Modifications that correspond to the modifications of FeResPost libraries.
The procedure for building the binaries from sources have been cleaned somewhat, and should be easier to adapt and use for anyone.
The number of available binaries has been increased. (More versions of Ruby are now supported, in 32bits or 64bits. The .NET CLR versions 2 and 4 are supported.)
The new version of the library is issued 2016/06/05:
Correction of bugs in the XDB raw iterator. (The unsigned int conversion was missing in COM component and .NET assembly.)
Correction of a bug in the “writeGroupsToPatranSessionFile” in the COM Application object. (Sometimes, the “Groups” argument was not well converted.)
Correction of a bug in the calculation of CLA finite element load response. (The thermo-elastic and hygro-elastic contributions were not correct.)
Relaxation for the syntactic rules in BDF files read with “readBdf” method of NastranDb class. Integers are now accepted as real values when no ambiguity is possible.
A few corrections in the manuals.
Binaries for the versions 2.3.* of ruby are also distributed.
Addition of examples showing how the temperature distributions applied for thermo-elastic loading of Nastran models can be retrieved and used for the calculation of laminate load response. (See sections IV.2.8.5 and X.E.4.)
The new version of the library is issued 2016/08/16:
Correction of several methods with variable number of arguments:
Method “setComponent” in “Result” class.
Methods “getAttachmentNbrSubCases” and “getNodeCoords” in “NastranDb” class.
Methods “getAttachmentNbrSubCases”, “generateShellOffsetsResult” and “getNodeCoords” in “SamcefDb” class.
Correction of a bug in the “insert” method of “Result” class. (The “NONE” element ID of key was not accepted.)
Correction of arguments in methods “removeLayers” and “removeSubLayers” of “Result” class in .NET assembly.
Correction of arguments in methods “removeLayers”, “removeSubLayers” and “extractResultOnEntities” of “Result” class in COM component.
In COM component: correction of a bug in the “writeBdfLines” singleton method of “NastranDb” class.
Correction of a bug in the extraction of corner nodes in Nastran scalar elements “CMASS1”, “CMASS2”, “CDAMP1”, “CDAMP2”, “CELAS1” and “CELAS2”. (This bug affected several other methods as “writeGmshMesh”.
The limit on lines length when Nastran BDF or Samcef DAT files are read has been elmininated.
Correction of a bug in the “Result” class modification of reference coordinate systems. (Previously, transformation was erroneous when curvilinear coordinate systems were used in the orientation of material properties in shell elements.)
Introduction of the “Ilss_b” version of the composite inter-laminar shear stress failure criterion in CLA classes. (See section II.1.10.27.)
Addition of method “containsEntity” to the “Group” class.
In the description of generic “DataBase” class in User Manual, methods “getNbrCoordSys”, “getNbrElements”, “getNbrNodes”, “getNbrAbbreviations” and‘getNbrGroups” are now presented as they are: as attributes “NbrCoordSys”, “NbrElements”, “NbrNodes”, “NbrAbbreviations” and “NbrGroups”.
A few “cosmetic improvement of the manuals (addition of internal links).
COM component installation program “registrySetup.exe” has been renamed “modifyRegister.exe”. (This has been done to “dodge” Windows 10 User Access Control that does not like executables containing the word “setup”.)
Elimination of dirty spurious “QueryInterface” error messages from the COM component.
The new version of the library is issued 2016/11/01:
Correction of a bug in “setComponent” methods of “Result” class when two parameters are passed to the method. (One bug for all extensions, and another bug for the COM component only.)
Correction of a bug in the XDB reading/attachment, when anonymous load cases are present in the XDB file (SUBCASES with no SUBTITLE).
In Ruby extension, correction of a bug in “modifyRefCoordSys” method of “Result” class. (The method returned “Self” instead of nil.)
Correction of a bug in the XDB access with BBBT index (option DBCFACT=4). (It seems that no index is produced when the table is stored in a single page.)
Improvement of the test for big/little endianness of XDB files. A warning message is issued in case of “doubt”. The manual is modified accordingly (section III.1.1.10.10).
Correction of a bug in the extraction of nodes in Nastran scalar elements “CMASS1”, “CMASS2”, “CDAMP1”, “CDAMP2”, “CELAS1” and “CELAS2”. (This bug affected several other methods as “writeGmsh”.
Addition of method “calcRandomResponse” in the “Post” module.
Addition of method “deriveScalPerComponent” in the “Result” class.
Addition of “VonMises2D” derivation method for the “deriveTensorToOneScal” method.
Addition of method “deriveVectorToTensor” to the “Result” class.
A sixth optional argument has been added to the “calcResultingFM” class method in “Result” class. This argument allows to provide the list of coordinates used for the calculation of global moment. This can be handy when the forces are not associated to nodes. (Element forces extracted from CBUSH elements, for example.)
The new version of the library is issued 2017/01/02:
Correction of a bug in “deriveVectorToTensor” method of “Result” class.
Correction of a small error in User Manual (description of the “calcRandomResponse” method in “Post” module.)
Addition of information in Appendix X devoted on installation instructions. (Discussion of execution environment, of installations of both 32 bits and 64 bits versions of the COM component, on GNU C++ re-distributable libraries.)
Correction of a bug in the Result modification of reference coordinate system when one of the system is curvilinear and the result values are associated to element center. (Error was related to the fact that the location of element center is obtained by averaging the nodal coordinates in a curvilinear CS instead of a cartesian one.) Errors with previous versions were very small in most cases.
Correction of a bug in method “calcResultingFM” of the Result class. In some cases, results were erroneous when curvilinear coordinate systems were used in the calculations.
Improvement of the “modifyRefCoordSys” in Result class. New version should be faster when several modifications are done on results related to the same FEM entities. In some cases, the time saving can be very significant.
Improvement of the different operators in the “Result” class. Among other things, a better support for calculations with complex numbers is provided.
The “Post.vectProduct” method also supports calculations with complex numbers.
The new version of the library is issued 2017/03/19:
Correction of a bug in the reading of Nastran “PBEAML” property cards. (Method “readBdf”in “NastranDb” class.)
Correction of a bug in the reading of Nastran BDF files. Comment lines inside Nastran Cards were causing problems. (Method “readBdf” in “NastranDb” class.)
Minimum support for the “CORD3G” Nastran card.
Correction of a bug in a few dyadic operators in Result class. (Bad initialization of cId in some cases.)
Modification of the “operator” methods in ruby “Post” module (COM component “Application” class). Now, the new methods cover the corresponding operators and dyadic functions in “Result”, “Group” and “ResKeyList” classes. (See section I.6.4 for more information)
Improvement of “generateCoordResults” method in DataBase class. Now, error messages are printed when missing nodes are detected, but no exception is thrown.
A few optimizations of the code. In particular, the composite Result calculations have been significantly accelerated. (See section II.1.9 for a detailed explanation.)
Method “reInitAllPliesAccelMatrices” added to the “ClaLam” and “ClaDb” classes. (This modification is related to the acceleration of composite calculations.)
Correction of a few bugs in the conversion of array arguments of COM component.
Addition of a ninth parameter to “calcFiniteElementResponse” and “calcFiniteElementResponse” methods.
The new version of the library is issued 2017/05/28:
Modification of “calcFiniteElementResponse” method. ILSS criteria are always calculated at bottom ply if requested, even if no output request at bottom sub-layers.
Introduction of the concept of “Strength Ratio” in classical laminate analysis (sections II.1.10 and II.4.7.5).
A few corrections in “calcRandomResponse” method of “Post” Module. The User Manual has been corrected too.
Addition of method “calcPredefinedCriterion” in “Post” Module in order to accelerate computations (optimization of post-processing operations).