I’m a mechanical engineer in the field of heavy machinery. I’m considering my options with the open source community, to implement more FEA work into the design work. I’ve been tinkering with a few different software. I’ve grown fond of Salome-Meca for meshing, and Paraview seems powerful (although not too user friendly), but find code_aster to be too complicated for me to actually be productive with my time.
I’m very impressed with PrePoMax so far - what a wonderful GUI, and wonderful implementation of Calculix. I’m wondering if anyone is using it for commercial work, or if its mainly being used for academic/learning purposes.
It all depends on what you want to analyze but PrePoMax and CalculiX can be suitable for many scenarios, including the analyses performed in industry. Since PrePoMax is just a pre- and postprocessor, its capabilities depend on what CalculiX has to offer (ccx features not yet implemented in ppm can always be added using keywords). So I would suggest you to ask on the CalculiX forum - some users may share their experience with industrial applications: https://calculix.discourse.group/
Code_Aster with Salome_Meca is powerful but for me and many others users it’s no go due to extremely low user-friendliness. CalculiX is pretty much the best choice for general FEA cases (unless you need some specific multiphysics capabilities - then Elmer might be needed). Check this thread on the CalculiX forum: Where to start with open source FEA - CalculiX
Very good point - of course the limitations lie in CalculiX. Im asking more to get a sense of the “community” around PrePoMax, and to understand the vision for the project.
Overall im very impressed with it so far. Im going through your tutorials on youtube currently. I wonder a few things:
How comfortable is it to work on larger assemblies? Say, in the spectrum of 20 to 100 parts?
What are good learning materials for Prepomax/Calculix?
Any recommendations as to advanced workflows?
Like for example, meshing in Salome, solving in Prepomax, postprocessing in Paraview? Something else? Calculix plugins for IDEs? Any recommendations are appreciated.
It seems beams are not available in PrePoMax, despite CalculiX allowing it, although i believe its limited to solid circular and rectangular sections - what will the future hold for PrePoMax?
Lastly, as its open source, whats the best way to support the development of PrePoMax to become even better ?
Of course, there can be a significant performance decrease when working with very large models but this aspect was improved recently and it shouldn’t be very problematic. But I suggest you test it with your typical assemblies - import them and try to perform some operations. However, CalculiX’s default solver may have issues with solving large models (it was reported several times here and on the CalculiX forum). In such cases, you can try using the Pardiso solver.
Documentation is your best friend, especially for CalculiX since PrePoMax is very user-friendly. Apart from that, YouTube tutorials and active forum communities make it easier to start working with ccx and ppm.
Advanced workflows usually involve meshing in external software (especially if you want to create a hex mesh). One of my tutorials shows meshing in Gmsh and the use of this mesh in PrePoMax. Salome_Meca might be even better for hex meshing if you are patient and can get used to inconveniences related to GUI.
PrePoMax is usually sufficient for postprocessing but if you need some advanced capabilities (like hiding individual elements and creating custom field output) you can convert the CalculiX’s .frd file to .vtk using for example FreeCAD or small converters available online.
For advanced cases it might be necessary to do the initial setup in PrePoMax, export the .inp file and edit it manually. Then you could also try automating runs and using GraphiX (CalculiX’s native pre- and postprocessor with some scripting capabilities).
Beams will likely be added to PrePoMax in the future. For now, you can use them in FreeCAD’s FEM module which also utilizes CalculiX as a solver.
I think that the most important thing is testing and reporting bugs. Currently, there’s a 1.3.5 dev version available for testing (it includes a large new tool for pressure loads imported from OpenFOAM). If you find any bugs, please report them on the forum with all the necessary details so that they can be fixed. Participating in forum discussions is also helpful - for example when ideas for new features are discussed or when other users ask for help.
user can manage by grouping trough compound part, an option to merge or separated mesh is available.
in case of large models in element and node, high performance computer is required and investing RAM capacity is highly recommended. speed of processor, number of cores/thread is a plus.
PrePoMax does not have feature in modeling a part, only by importing common 3D CAD file format. so, an external software to modeling is a must.
also, automatic tetrahedral element is supported only from imported part to be meshing. structured mesh is not, user need an external software with the capability to produced an UNV or INP formats. in this case, someone could chosen Salome CAD/CAE due to it’s completeness of feature…
right, it’s not yet supported. hopefully in the future PrePoMax can be capable to import of line based element and provide basic interface to defining cross section dimensions.
As a main/only PrePoMax developer I am happy to hear that you like the GUI. Thank you.
The PrePoMax project started some 6 years ago with the intention to provide a user friendly graphical interface for one of the open-source FEM codes. The CalculiX was selected for this purpose due to its maturity and similarity of the input files with Abaqus. The aim of the development is to support FEM engineers in their professional work, not only to solve test cases or academic examples with students.
It definitely has drawbacks and a lot of features are still missing but overall I think that a lot of engineering work can be done using it and some users are already using it for their professional work.
The best contribution to an open-source project is to be an active user. Test the software, post questions, ideas, answers. The PrePoMax community is not huge but when you add the CalculiX community to it it is large enough to get answers.
The development is mainly based on user feedback. The other thing affecting the development is the financial support. After PrePoMax was released, we were contacted by a Korean Institute KIMM with a proposition that they would like to participate in the development of the PrePoMax. Since they have no developers to help they are helping the project by founding it.
If PrePoMax is good enough for you only you can decide. So test it throughly!
Thanks for all the great responses. I’m keen to get more invested into the software, and will make sure to make use of the community, and help it grow. Thank you Matej for making the effort. It’s a beautiful thing when developers takes the (spare)time to make powerful AND user-friendly FOSS for the masses. Especially within the open source CAE, where the potential seems enormous.
I am introducing it to our company. its an uphill battle unless you can show that the results are useful and equivalent/real compared to the other software packages that people have used in the past. We use specialized software for some esoteric simulations. Otherwise the only needs we have have been stress/strain, thermal and vibration mode search. So PrePoMax/Calculix have been great at covering this.
Early on while learning how Calculix worked or better said, how to use Calculix, I learned that if you want results, you need to make the model as simple as possible. avoid assemblies where parts “almost” touch each other with a very tiny and flat gap. Same for circular or curved shapes, don’t have a hole really close to an edge. What that would do is, it would need more refinement, and if its not necessary to your subject, just delete the hole and get results faster.
That said, I’ve loaded up 1000 parts from solid works. Nothing fancy, lots of imported parts from Mcmaster. Its really of no use to do that. So for example, split lock washers come with a helical shape and also a round OD. That’s just going to make your case hang for a long time and you’ll get nothing in the end. I would look into simulating one screw connection, then using some sort of average to load the assembly without the screws. Simplification at first, then you can add complexity as you learn your particular model’s behavior.
Thanks for your input.
I dont expect it to be an uphill battle in my work, as no one questions what software people use, only that you can somehow validate your results through either hand calcs or experimental data. I suppose this would be the case with any FEA software, though i would suspect you could have a tendency to “lean” more on the software if it was a fancy/expensive one. What’s your experience regarding the pushback, and how have you dealt with it?
I really appreciate your not-to-do’s as these are some of the things i’m currently learning by doing - painfully. I’ve pondered over the best ways to learn to use calculix, but i find the ressources somewhat limited. Instead, what i’ve ended up doing is following all the examples from a book i have lying around, “Engineering Analysis, with solidworks simulation 2020”, and try to replicate the cases with prepomax. I find it’s equally informative, and satisfying, to do replicate even the simplest cases, and figure out stuff on my own. Ofcourse, FEAnalyst has great youtube videos, and Matej too, but beyond that it seems im left with the Calculix documentation/prepomax manual (still limited).
One thing i currently really miss is beam elements in prepomax, and i know its high on the list for TODO’s in prepomax. Looking forward to that.
I can also recommend the book titled “Analysis of Machine Elements Using SolidWorks Simulation” since it’s the only book I’ve found that includes verification with analytical calculations (something that I always include in my tutorials as an important but often ignored aspect of FEA studies).
Apart from that, there’s a great practical book “Finite Element Analysis for Mechanical Engineers”. It also features some examples and you can learn a lot of good practices from it.
Ah, thanks for the heads up. I agree with you, and i REALLY think your videos are great for showing first the analytical solution, followed by the FE analysis! I will look into the SolidWorks book you mention, although i suspect it must cover a lot of the stuff that the one i mentioned does? Perhaps it’s a more advanced book? May i ask your opinion on using PrePoMax vs. SW simulation?
The book you’re referring to in the end of your post - is it possibly this one you mean?
It’s from the same publisher but different authors and different kinds of examples. The level is rather similar though.
SW Simulation is also very intuitive since it’s meant for designers, not for analysts but the workflow is different (after all, PrePoMax and CalculiX are based on Abaqus and its specific approach to analyses with sets, steps and so on). For a CAD-embedded FEA module, SW Simulation offers quite a lot of features. Contrary to e.g. Fusion 360 which has very limited FEA module.
Yes, that’s the one. There’s also another, older practical book titled “Building Better Products with Finite Element Analysis” but with no example problems.
Dear Matej!
Thank you for your work! You have made a great program. I would like to show the program to the students in my former technical high school.