Getting started

To get OOFEM installed on your system, please follow the Installation instructions first.

OOFEM is a console application; it should be executed from the command line with arguments specifying the path to the input file. When no arguments are provided, the help is printed on the output.

$ ./oofem

____________________________________________________
  ____  ____  __________  ___
 / __ \/ __ \/ __/ __/  |/  /___  _______
/ /_/ / /_/ / _// _// /|_/ // _ \/ __/ _ \
\____/\____/_/ /___/_/  /_(_)___/_/  \_, /
Copyright (C) 1994-2025 Borek Patzak /__/
____________________________________________________

Options:

  -v  prints oofem version
  -f  (string) input file name
  -r  (int) restarts analysis from given step
  -ar (int) restarts adaptive analysis from given step
  -l  (int) sets treshold for log messages (Errors=0, Warnings=1,
            Relevant=2, Info=3, Debug=4)
  -rn turns on renumbering
  -qo (string) redirects the standard output stream to given file
  -qe (string) redirects the standard error stream to given file
  -c  creates context file for each solution step
  -p  runs in parallel mode (MPI)
  -t  (int) sets number of threads for OpenMP parallelization
  -m  shows solution status monitor output,
      redirecting standard output to files,
      oofem.stdout, oofem.stderr by default, use -qo -qe to override
  -Werror  promote warnings to errors

Copyright (C) 1994-2025 Borek Patzak
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

The most useful options in daily work are

-f <file>

The input file to solve. This is the only required option.

-l <level>

How verbose the console log should be. -l 2 keeps only the relevant messages, -l 4 adds debugging detail.

-qo <file>, -qe <file>

Redirect the console log (standard output) and the error stream to files. Note that these control the log; the result output file is a separate thing, named in the input file itself (see Understanding output file).

-c and -r <step>

-c saves a context (restart) file after each solution step; -r restarts a previously interrupted analysis from the given step. Both are needed for long running analyses.

-t <n>

Number of OpenMP threads. Requires a build configured with USE_OPENMP_PARALLEL.

-p

Run as an MPI process of a distributed memory parallel analysis. Requires a build configured with USE_MPI_PARALLEL and a partitioned input.

-Werror

Abort on the first warning instead of continuing. Useful when preparing a new model, because warnings about ignored or misspelled input fields then cannot pass unnoticed.

Running oofem -v prints the version together with the list of compiled-in modules and the exact commit the binary was built from, which is what you should quote when reporting a problem on the forum:

$ ./oofem -v

OOFEM version 3.0 (x86_64-Linux, fm tm sm sm_elements sm_materials mpm IML++)
Git RepoURL: https://github.com/oofem/oofem.git
    Branch: devel, Hash: v3.0

Running the first analysis

The problem to be solved is fully described in the input file. The structure of the input is explained in the Understanding input file section and fully documented in the OOFEM Input manual.

To run oofem with a specific input, use the -f option followed by the path to the input file. For illustration, we demonstrate the execution using the beam2d_1.in test, which is part of the OOFEM test suite and is located in the tests/regression/sm directory.

$ ./oofem -f /home/user/oofem.src/tests/regression/sm/beam2d_1.in

____________________________________________________
  ____  ____  __________  ___
 / __ \/ __ \/ __/ __/  |/  /___  _______
/ /_/ / /_/ / _// _// /|_/ // _ \/ __/ _ \
\____/\____/_/ /___/_/  /_(_)___/_/  \_, /
Copyright (C) 1994-2025 Borek Patzak /__/
____________________________________________________
Job: /home/user/oofem.src/tests/regression/sm/beam2d_1.in
Computing initial guess

StaticStructural :: solveYourselfAt - Solving Metastep 1,  Step 1, Starting Time 0.000000e+00, Time Increment 1.000000e+00, Final Time 1.000000e+00,  (neq = 15)
NRSolver: Iteration ForceError
----------------------------------------------------------------------------
NRSolver: 0      D_u:  1.000e+00  D_w:  1.000e+00  R_v:  1.000e+00
NRSolver: 1      D_u:  4.005e-06  D_w:  7.393e-06  R_v:  3.385e-16
EngngModel info: user time consumed by solution step 1: 0.00s

...

ANALYSIS FINISHED

Real time consumed: 000h:00m:00s
User time consumed: 000h:00m:00s
Total 0 error(s) and 0 warning(s) reported

The lines starting with NRSolver: report the convergence of the equilibrium iterations, one line per iteration, with the relative error of each unknown group. The last line of the run is a summary you should always check: an analysis that finishes with reported errors, or with warnings you did not expect, has probably not solved the problem you intended.

By default, a text output file with the results is created (as specified in the input file). In this case, the beam2d_1.out file is created in the current working directory; its content is described in Understanding output file. Note that the paths in the input file, the output file name included, are interpreted relative to the current working directory, not to the location of the input file.

Where to go next

  • Understanding input file — how the input file is structured

  • Understanding output file — how to read the results

  • Postprocessing — how to get the results into Paraview

  • Using OOFEM from Python — how to drive the same analysis from Python

  • The tests/regression directory of the source distribution contains several hundred small, working input files covering nearly every element, material and analysis type. Finding a test close to your problem and adapting it is usually the fastest way to build a new model.