Excellent! Would be fun to try out on some future 3d-prints, as well as CNC-stuff. I need to migrate to FreeCAD fully so that I can build an automated pipeline from online-orders of custom things (such as a parametric wooden box) to g-code.
Great video, but as stated before the post processing pipline in FreeCAD lets you make post processing as nice as solid works! All off you guys are very welcome to share your FreeCAD FEM experience on the FreeCAD forum. cheers bernd
The small difference is most probably due to meshing strategies. It's a simple analysis. My guess would be that with more elaborate designs, the meshing techniques, and the differences they produce, are probalby more visible. This does not mean to say that the difference would all imply a freecad error.
I don't know how to do it in Free CAD but in SolidWorks, you can make a sketch on the surface and do a split line command. Then in the simulation, you can pick the area you did the split line on instead of the whole face
I keep encountering an issue with FEM, saying FEM mesh has no volume elements, either define a shell thickness or provide a FEM mesh with volume elements.
I have not tried a dynamic analysis like that, it would be an interesting question on the freecadweb.org forum. I would probably approach it by placing the equivalent force of the bullet on the surface area it would impact and see if it would go beyond the yield strength. Tracking the deformation of the bullet would be some real complex stuff though.
Both softwares are equivalent...since there is no exact math formula to predict stress distribution with accuracy...only lab experience do...like aerodynamic profile for planes: Only miniature models in wind tunnels tell if the prototype is valid...no algorithm work
: You are so right. People tend to think that a FEA solution is an exact solution or a good representation of reality. It is not. Even if the model mimic 100% the real system, nothing warranties that the numeric solution is correct. All numeric methods can fail or yield wrong results when converging to a wrong solution.
There are several formulae to predict stress distribution with KNOWN accuracy. Know your physics and know your software prior to analysis. Being able to turn a guitar on doesn't make you a rockstar. That said, I am not familiar with any accurate method ( as you're ALL about accuracy) to measure actual stress distribution on physical model.
The equations of statics and the elastic compatibility for a true elastic material do have theoretically exact answers. Differences in meshing will produce different numerical results. The FEA are not developed from experimentation, they are developed from statics and compatibility of elastic materials. If you are dealing with non-linear materials, then you need to use experimentation for those properties. You would also not be using a linear static analysis.
you converted the object to mesh, but what if the object was joined using bolts? bolts have much smaller contact area than a monolithic object, therefore this simulation doesn't worth a dime, because it assumes the table is made out of an entire piece of wood (or whatever)
Excellent! Would be fun to try out on some future 3d-prints, as well as CNC-stuff. I need to migrate to FreeCAD fully so that I can build an automated pipeline from online-orders of custom things (such as a parametric wooden box) to g-code.
Sounds like some great projects!
Good job. People like me will be less afraid of using open souce FEA now :)
Great video, but as stated before the post processing pipline in FreeCAD lets you make post processing as nice as solid works! All off you guys are very welcome to share your FreeCAD FEM experience on the FreeCAD forum. cheers bernd
Very informative. Thank you!
Excellent comparison! Thank you.
This is a good 4 years old video. Can we get an update to the freecad fem.
Great side-by-side. Thanks!
Great video, thank you! I want to model my planned house additions now!
Thanks good luck on the house
The small difference is most probably due to meshing strategies. It's a simple analysis. My guess would be that with more elaborate designs, the meshing techniques, and the differences they produce, are probalby more visible. This does not mean to say that the difference would all imply a freecad error.
Cool Gracias from Santiago of Chile.....👏👏👏👏👏👏.....😷👷
How would you proceed, if the force would be applied unevenly, say, close to only one side or close to only one leg of the table?
I don't know how to do it in Free CAD but in SolidWorks, you can make a sketch on the surface and do a split line command. Then in the simulation, you can pick the area you did the split line on instead of the whole face
Looks like fun to play with!
I keep encountering an issue with FEM, saying FEM mesh has no volume elements, either define a shell thickness or provide a FEM mesh with volume elements.
Can you import an STL and do stress test on it?
Maybe freecad fem developer can work on the color temperature resolution if we all ask them :D
Absolutely! Also see this on visualization if you have not already:
th-cam.com/video/zCi632QRcAE/w-d-xo.html
Very cool video. Thank you.
Thank you. Very helpful
What do you say regarding SelfCAD?
Thank you.
Is it possible to do impact simulation in FreeCAD? For example a bullet hitting a armor plate
I have not tried a dynamic analysis like that, it would be an interesting question on the freecadweb.org forum. I would probably approach it by placing the equivalent force of the bullet on the surface area it would impact and see if it would go beyond the yield strength. Tracking the deformation of the bullet would be some real complex stuff though.
great works
Great video.
Apparently the simulation is about a table of solid metal. The elements are not hollow. Am' I right?
That's correct, if you look at the description you can download the models to see for yourself.
@@JokoEngineeringhelp I did it. I was surprised because incredibly strong, compared with a real metal table, thank you.
thanks
So cool.
I mean to say is there any open source cam software for 5 or 6 axis
Not that I know of!
Hey, thanks for the video. I want to know if there is any 5 or 6 axis free cam software
Hummmmmmmmm, very interesting a stress peak on one node point: the table leg that is. A bit of local remeshing would be nice.
Bad colors in FreeCad if you're a bit color blind, the greens and yellows almost look the same.
All FEA uses this sort of color scale, but the FreeCAD pipeline can put the color on a greyscale
Both softwares are equivalent...since there is no exact math formula to predict stress distribution with accuracy...only lab experience do...like aerodynamic profile for planes: Only miniature models in wind tunnels tell if the prototype is valid...no algorithm work
:
You are so right.
People tend to think that a FEA solution is an exact solution or a good representation of reality. It is not.
Even if the model mimic 100% the real system, nothing warranties that the numeric solution is correct. All numeric methods can fail or yield wrong results when converging to a wrong solution.
There are several formulae to predict stress distribution with KNOWN accuracy. Know your physics and know your software prior to analysis. Being able to turn a guitar on doesn't make you a rockstar. That said, I am not familiar with any accurate method ( as you're ALL about accuracy) to measure actual stress distribution on physical model.
@@Eng_Simoes
Rafael
Knowing how to use a FEA software does not make you an expert in stress analysis
The equations of statics and the elastic compatibility for a true elastic material do have theoretically exact answers. Differences in meshing will produce different numerical results. The FEA are not developed from experimentation, they are developed from statics and compatibility of elastic materials. If you are dealing with non-linear materials, then you need to use experimentation for those properties. You would also not be using a linear static analysis.
Is FreeCAD doing Delaunay triangulation?
you converted the object to mesh, but what if the object was joined using bolts? bolts have much smaller contact area than a monolithic object, therefore this simulation doesn't worth a dime, because it assumes the table is made out of an entire piece of wood (or whatever)