Calculix Review

 As I was passing by on the web I noticed that no Calculix review had been received.  I couldn't pass through without correcting the situation.  I consider the program to be a real gem. I will be so bold as to say that it is the very best piece of English language GPL-ware on the web for finite element analysis.  There is a graphics pre/post-processor program Graphix, that supports the solver program, Calculix.  It handles a wide variety of mechanical, thermal, coupled thermomechanical, and contact problems,as well as field problems. It is well documented, regularly updated, and well supported. 

The format for the solver is a subset of Abaqus .inp, and true enough to the abaqus conventions that commercial file generators, like FEMAP, run with little or no doctoring.  All the files are ascii, so they are amenable to viewing, alteration, templating and modification with any text editor.  There are a number of solver options both linear and non-linear.  The solver program's creator uses long proven math libraries which gives the program excellent speed.  This has allowed him to concentrate on the material, physics, and engineering aspects.

The pre/post-processor Graphix has a unique interface that gives the user the usual display capabilities in the graphics window, and reserves a wide number of other commands for text input.  This makes it easy to learn the very basic operations.  The experienced user can do the more powerful operations without having to go through layers of menu, sub-menu choices.  The coding is such that the viewer can rapidly pan, zoom, and rotate very large grids, both meshed, and filled. There are  number of viewing options  that act upon the visual display of both geometry and data.The, Graphix preprocessor also generates input data for a number of other FEA, and CFD programs.  Very few people know that it can generate Code_Aster .mail files.

My general comment on the two programs is that they are designed to be used by people to solve actual problems.  On first view their simple user interfaces understate, and over-deliver. Like all powerful tools they have a significant learning curve to achieve proficient use.  Unlike most programs the time is not spent on learning what submenu has what funtion that you want.  The learning curve comes from a need to understand the structure and logic of the data to be passed to the solver.  This certainly bucks the current trend in programming.  With everything laid bare to the source code under GPL, there is little to be gained from hiding the internal workings of the program from the user (gasp!). 

Another item that seems to escaped casual reviews is that both programs are readily available as compiled binaries on both the Linux and Windows operating systems.  The windows program ports, by Jeff Baylor, run without modification with Windows 95 to Windows Vista.  No X server, no Cygwin, no Python installation required.

If you are looking for a tool to analyze real problems, then I highly recommend the Calculix programs, Calculix, and Graphix.  If you are looking for gloss on the outside and nothing under the hood, then these are not the programs for you.  Why Calculix has not become much more widely recognized is a mystery.  I could go on, but don't take my word for it, go look for yourself.