FrontISTR v5.8 Release Notes¶

Release Date¶

November 7, 2025

What's New¶

New Features¶

E-1. Implementation of the Element Activation/Deactivation Function¶

Implemented the "Element Activation/Deactivation" function that suppresses the computation of a specified element group. The enabled/disabled state of element groups can be controlled by the !ELEMENT_ACTIVATION keyword in the Analysis Control Data file. Visualization output in the deactivated state is also supported.

(Issue #657, solver !473, document !9)

Specification Changes¶

M-1. Restriction on CLOAD Specification by Node Groups When the Mesh Is Refined¶

When the mesh has been refined, the CLOAD specification by node groups has been changed to stop with an error when used, because it is difficult to distinguish them from the node groups for boundary conditions.

(Issue #272, solver !512)

M-2. Refactoring of the External Load Processing¶

The external load processing (fstr_ass_load.f90) was split into dedicated subroutines per load type, improving readability and maintainability. The processing of concentrated nodal forces, distributed loads, user-defined loads, thermal loads, etc. was organized into individual functions, and the common processing was extracted as helper functions.

(Issue #683, solver !484)

M-3. Review of the Default Values of the Automatic Increment Parameters and Contact Search Parameters¶

The default values of the automatic increment control parameters and contact search parameters of nonlinear analysis were reviewed.

(Issue #710, solver !511)

M-4. Change of the Principal Stress/Principal Strain Component Names in VTK Output¶

When reading pvtu files in ParaView, the display of the principal stress and principal strain components was changed from the default "X, Y, Z" to "1st, 2nd, 3rd".

(solver !505)

M-5. Fix of the Dependencies in Makefile.am¶

The dependency descriptions (.dependf90) between Fortran source files in Makefile.am were fixed.

(solver !507)

M-6. Fix of Compiler Warnings at Build Time and Dependency Library Linking¶

Compiler warnings such as buffer overflows that had been occurring in GitLab CI were resolved. Also, the duplication in the dependency library specification at CMake link time was organized.

(Issue #713, solver !510)

Bug Fixes¶

B-1. Fix for the Plate Thickness Not Being Applied in the Stress Update for 2D Elements¶

Fixed a defect in the subroutine UPDATE_C2 of static_LIB_2d.f90 where the plate thickness (THICK) was not multiplied into the computation of the internal force vector and the units did not match.

(Issue #276, solver !513)

B-2. Fix for NaN Errors Due to Missing Variable Initialization in contact and embed¶

The state variables of contact and embed (distance, lpos, gpos, direction, multiplier, tangentForce, etc.) are now initialized comprehensively, resolving NaN errors caused by uninitialized variables. Also, the element center coordinate retrieval function (getElementCenter) was extended to 3D elements such as tetrahedra, prisms, and hexahedra. It does not manifest in environments where the compiler zero-clears uninitialized variables, but it had caused errors in other environments.

(Issue #714, solver !511)

B-3. Fix for Insufficient Load Assembly Arguments During Contact Analysis in Implicit Dynamic Analysis¶

Fixed a defect where, when performing contact analysis in implicit dynamic analysis (dynamic_nlimplicit), the iteration count (iter) argument was not passed to the dynamic_mat_ass_load call.

(solver !515)

B-4. Fix for the Defect Where Contact Information Is Not Written to the Restart File of Explicit Dynamic Analysis¶

Fixed a defect where, in explicit dynamic analysis (dynamic_nlexplicit), the contact information was not included when writing the restart file.

(solver !508)

Acknowledgments¶

The following contributors contributed to this release. We thank them for their contributions.

• inagaki: E-1, M-2, M-3, M-6, B-1, B-2
• Kazuya Goto: M-1, M-2
• Kanato Kono: B-3
• toru ando: B-4, M-5
• Otasam-ihcada: M-4