CCX looks like as one binary file (in Windows it is ccx.exe). If you will try to run it as usual (by double-click) nothing will happen. CCX is console application with text file interface. At first you should create text file (File.INP) with mesh, named groups and input (boundary conditions) applied. After that you can run solver from terminal emulator "ccx file". CCX will read text file and make calculation. The result will be many text files stored at the same folder. For example .FRD file will contain all results (as you will order in File.INP)
What is CGX? CGX is pre-postprcessor. What you can do with CGX? You can run input file for preprocessing (to see and check groups of nodes, element faces and elements). You can apply some boundary conditions (for example pressure or film coefficient for convection) with using "send" card in CGX.
When calculation is finished, you can run .FRD file for the postprocessing (to check stress, displacements and etc).
What is Salome? It is powerful open-source CAD/Mesher, developed by EDF (Electricity of France). There is a simple reason why it is open, because there are no reasons to make it proprietary. The goal of development is to design and build industrial things, which is much more expensive and important than possible income that can be obtained from proprietary software. Also the normal proportion for salespersons/developers in software industry is 9 to 1 (not any software, even perfect can be sold easily). You can take part in usage and testing and submit your issues to developer. Salome is mainly for Linux but there is experimental Windows version.
What is Salome-MECA? It is Salome-Platfrom + Code_Aster FEA package. Salome-MECA is for Linux only. There is a special Linux distribution (64 bit), called Caelinux that combines Salome-MECA and other open-source software for engineers. You may get .ISO file from official website, from SF.NET or using this link: https://yadi.sk/d/nfKAyCsdmWME7(md5sum fa8f1d551f0ad9ccfdb295367fc9e9cf for caelinux2013.zip) See all files at this shared folder: https://yadi.sk/d/Uwq053wjmWMKk Installation:
1) Salome For Windows: download archive (from here or from official website) Unzip it onto C:\ and double-click salome_run.bat file. Link for old 32-bit version.
The reason why older versions can be required is that many Code_Aster books and examples has been created for older version of Code_Aster (v10). Some commands can be changed and the old sample will require some changes in the newer versions.
Installation of Salome-MECA under Linux is very simple (because it is generally portable software). Untar archive into any folder, and ./...run it from terminal.
Under windows you can try VirtualBox or programs like that. Google for "how to install 64 bit under virtualbox" if you're going to use 64 bit. Some extra actions can be required. In this case Linux will work as normal windows program at separate window.
If you want to install it as normal fully functional OS, google for "install linux from ISO". If you have separate flash stick, use UNetBootin from SF.NET (or similar programs) to make and run it in live mode (without installation). Also you can install linux onto usb flash drive or SSD and use it from there.
Many other lessons for Salome can be found on this channel.
What you can do with Salome in general? You can build any geometry (bodies and shells) and export it into STEP/IGES standard formats. Salome is little bit less efficient in terms of productivity than Solidworks or other feature-based software but you can do the same things just slower.
What else you can do with Salome? You can prepare mesh with named groups (to apply boundary conditions) and export it into many mesh formats (like .UNV, .MED, etc) or you can use Code-Aster to make calculation inside Salome-MECA.
Open meshers are less productive than commercial ones, but it do the same thing (just little bit more sensitive to your skills and experience).
There are many programs and services (like SimScale) use opensource libraries and can be used along with Salome.
1. In this lesson we will use geometry in .STEP file. There is a steel cantilever (E=29e6 psi) with 1500 lb applied to the tip as pressure directed along Y axis. See this first: http://youtu.be/FIKBtefm-es
2. Elmer uses Opencascade kernel to deal with STEP and when you will open the model, all dimensions will be scaled to millimeters although you use something different when saving. In this problem we will use Imperial units and scaling factor=0.03937 (1/25.4) to scale initial mesh coordinates to "inches". I am adding this line to the "Free text" at the "Simulation" section of .sif file
3. Default type of elements in Elmer is linear one, what is giving weird results even for good meshing.
It is why we need to add "element=p:2" to the "solver" section of .sif file. I am doing it manually in this sample. Also I am correcting Young's modulus to be counted in "psi" to be compatible with "inches". Other values and constants doesn't necessarily to be changed (for static problem).
In this example I use "Direct Solver" for linear solution and 1 iteration at "Nonlinear" section (steady state, linear). Specially for Elmer you need to follow log file when making solution to be sure that it is converged. Otherwise the results can be weird.
4. Elmer GUI is still in development and it saves all data in the xml file. When you corrected .sif file manually and save it, it will come back to previous state when you will load your project again from xml file. Sometimes it can be hard to figure out what is happen. Check .sif file every time before running solver, because solver uses just .sif , not xml. Learn Elmer Model Manual of actual version for command language and use it.
Elmer is good and correct scientific software but it requires some knowledge about console applications principles, to be successful with it.
1. Create the model in Freecad and export as step
2. Run Elmer-FEM, open the .STEP file and mesh it
3. Apply equation, material and boundary conditions
4. Edit .sif file. Coordinate scaling =0.001 is because Opencascade kernel is always transfer .step into "mm", but we use SI units and need "meters". "Element=p:2" is for using second order p-elements and avoid weird results
5. Run Solution
6. Run Post-processor and check results