Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Developer Items to Consider
#1
This is a list of items for developers to consider. Developers are free take on any project they like. These are just some basic ideas:

MOST STRAIGHTFORWARD:

1. Define and verify an open source Fortran compiler that works with MYSTRAN. MYSTRAN has currently been compiled with the Lahey Fortran compiler. We are now working towards using gFortran to compile MYSTRAN (In-Work item).

2. Investigate a 64 bit option. According to Bill, this is likely to be straightforward.

INTERMEDIATE:

1. For small DOF problems, the option the card "PARAM, SOLLIB, ZZPACK" can be used and is reliable (banded solver). However, for larger DOF programs, a sparse solver may be required. MYSTRAN currently uses an older Intel solver for this case (this only exists for the complied EXE but ***NOT*** for the files on GitHub, which is a known issue). However, this older sparse solver is not very reliable. We are investigating various sparse solvers. We previously looked at BCSLIB-EXT, but the cost is relatively high and it appears there may be other solvers that are just as good (if not better). The SPOOLES solver is an option, but it seems to be relatively slow compared to other options. PARDSIO with Intel MKL is an option, but requires an Intel compiler and the user must also download the Intel MKL. PARDISO 6 is more improved than PARDISO MKL, but requires the user download a free license every year. PaStiX may be the best option since it is modern and free. We can take a page from CalculiX and see how their investigation is going since they looking at other solvers. CalculiX has historically used SPOOLES, but recently added an option for the PARDISO MKL.

2. For the Eigen solution, beam elements have a differential stiffness matrix implemented in MYSTRAN. However, the shell and solid elements do not have the differential stiffness martrix coded. This would need to occur for them to be available for the Eigen solution.

LONG TERM - PRE/POST

1a. pyNastran is capable of reading a BDF input file: ( https://pynastran-git.readthedocs.io/en/latest/ ). However, it can not read MYSTRAN ouptut files. It is capable of reading MSC Nastran OP2 files, but MYSTRAN would have to be coded to create an OP2 file to view the results in pyNastran.

1b. CalculiX GraphiX is a pre/post that works with open source CalculiX solver ( http://www.dhondt.de/ ). It has basic functionality and is a lightweight program. This make it rather convenient and I have used it with some routines to generate deflection/contour plots and pass them to Excel in real time ( https://www.espcomposites.com/software/advanced.html ). It does not support MYSTRAN, but it may be possible for MYSTRAN to create output files that are compatible with CGX (CalculiX GraphiX). One challenge is that CalculiX only uses solid elements (it expands beams and shells into 3D solids), so this could pose an issue that requires time to modify either the CGX or MYSTRAN programs.

1c. Another pre/post option is to work on Roshaz. There is a separate forum for that ( https://www.structuralfea.com/forums/for...php?fid=11 )

LONG TERM - NONLINEAR

1. A nonlinear solver could be implemented. I think geometric nonlinearity would be the first and easiest to implement. Material nonlinearity could be addressed after that. This would be similar to the MSC Nastran SOL 106, which has some basic nonlinear capabilities. A full nonlinear implementation (with contact, etc.) would probably be beyond the scope though.

2. As a solver option, it may be possible to use the SPOOLES solver. This is what CalculiX uses (another open source finite element program that has nonlinear functionality).

LONG TERM - HIGHER ORDER ELEMENTS

1. An 8-node quad and 6-node tri could be developed in MYSTRAN. This would take quite a bit of work though.

--

Brian Esp
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)