FreeCAD 1.0 brings a true rev­o­lu­tion. But apart from the TNP mit­i­ga­tion, built-in Assem­bly work­bench, mate­r­i­al han­dling rework, and major Sketch­er improve­ments, there’s a less com­mon mod­ule that received many updates – FEM

In fact, it’s one of the work­bench­es that changed the most. Over 80 issues from the FEM Roadmap were fixed. This includes many bugs — some of them seri­ous and long-stand­ing — but also com­plete­ly new fea­tures and var­i­ous small usabil­i­ty improve­ments. I’ve been main­tain­ing the FEM work­bench for almost a year. In this post, I’d like to out­line some of the most impor­tant new fea­tures com­ing in FreeCAD 1.0 soon.

The biggest new FEM fea­ture is Con­strain­t­Rigid­Body – a tool pro­vid­ing sup­port for CalculiX’s rigid body con­straint. The con­straint is com­mon­ly used to apply moment/rotation or force/displacement to any select­ed face via a so-called ref­er­ence point which can be locat­ed at a dis­tance from the face. This makes it pos­si­ble to apply a force act­ing on a lever (remote force). The con­straint is par­tic­u­lar­ly use­ful since it final­ly pro­vides a way to sim­u­late tor­sion of arbi­trar­i­ly shaped parts in FreeCAD FEM.

The sec­ond major change is the intro­duc­tion of 2D analy­ses with Cal­culiX. That includes plane stress (for very thin parts with in-plane loads), plane strain (for very thick parts with in-plane loads) ‚and axisym­me­try (for revolved parts with loads uni­form around the cir­cum­fer­ence). The analy­ses can be per­formed by just fol­low­ing the approach for analy­ses with shell ele­ments (but using only the XY plane) and chang­ing the Mod­el Space prop­er­ty of the Cal­culiX solver.

Anoth­er new form of sym­me­try sup­port­ed with the Cal­culiX solver is cyclic sym­me­try. It can be defined with the tie con­straint and used for parts con­sist­ing of cir­cu­lar pat­terns of repet­i­tive seg­ments (sec­tors) around an axis and sub­ject­ed to loads exhibit­ing the same type of sym­me­try. Thus, many rotat­ing com­po­nents such as shafts, fly­wheels, tur­bines, or fans can be ana­lyzed this way.

Small­er but also impor­tant changes include more set­tings for tie and con­tact con­straints, beam ele­ments with reduced inte­gra­tion (nec­es­sary for pipe cross-sec­tions and accu­rate results with plas­tic­i­ty), addi­tion­al incre­men­ta­tion, and out­put fre­quen­cy set­tings for Cal­culiX, adjustable accel­er­a­tion mag­ni­tude for grav­i­ty load and sup­port for plas­tic­i­ty with kine­mat­ic hard­en­ing in addi­tion to isotrop­ic one, among others.

Sev­er­al improve­ments were intro­duced for ther­mo­me­chan­i­cal analy­ses. You can change their type to uncou­pled or pure­ly ther­mal, radi­a­tion heat flux can be mod­eled, vol­u­met­ric heat source is usable with Cal­culiX and heat flux was added to this solver’s outputs.

Some oth­er changes include capa­bil­i­ties of sup­press­ing analy­sis fea­tures, hid­ing the select­ed finite ele­ments and can­cel­ing ongo­ing meshing. 

The imple­men­ta­tion of Net­gen was also replaced to work on Lin­ux. Sym­bols of analy­sis fea­tures are now prop­er­ly scaled and their size can be adjust­ed. There are sev­er­al mesh appear­ance improve­ments as well. Old FEM exam­ples on the Start page were updat­ed and replaced with one file includ­ing three dif­fer­ent approach­es (1D, 2D and 3D) to the clas­sic can­tilever beam analysis.

Final­ly, the menu/toolbar names of sev­er­al analy­sis fea­tures were changed to elim­i­nate incor­rect usage of the word “con­straint” and make their mean­ing more clear.

A full list of changes, exclud­ing minor bug fix­es, can be found in the release notes

Such enor­mous improve­ments wouldn’t be pos­si­ble with­out our new devel­op­er – Mario Pas­saglia (marioalexis84), work­ing with whom is a plea­sure. His con­tri­bu­tions were spon­sored by you, the com­mu­ni­ty, via a grant from the FreeCAD Project Asso­ci­a­tion. There’s still a lot to do in order to make FEM as pow­er­ful as it could be, and more con­trib­u­tors are always welcome.

Cur­rent plans for fur­ther devel­op­ment include fix­ing any remain­ing bugs, stan­dard­iz­ing the core of the work­bench to make it eas­i­er to imple­ment addi­tion­al solvers (there is a new ini­tia­tive to add sup­port for Code_Aster), as well as fur­ther extend­ing the sup­port for Cal­culiX and Elmer and intro­duc­ing advanced mate­r­i­al mod­els based on the reworked mate­r­i­al sys­tem in FreeCAD. 

Hope­ful­ly, many of these goals will be achieved and avail­able in the next major release. All users are wel­come to test as much as pos­si­ble, report any issues and pro­pose new fea­tures if they aren’t on the roadmap yet.


Discover more from FreeCAD News

Subscribe to get the latest posts sent to your email.

7 responses to “Major FEM workbench improvements for FreeCAD 1.0”

  1. Applied Vehicle Technologies LLC Avatar
    Applied Vehicle Technologies LLC

    Amaz­ing work, team! Count­ing down the days until 1.0!

  2. renatozor Avatar
    renatozor

    great job , thanks to all the devlop­pers Team !!

  3. SNK Avatar
    SNK

    Will there be CFD sup­port or can we only rely on the cfd­OF module?

    1. Aleksandr Prokudin Avatar
      Aleksandr Prokudin

      Built-in CFD sup­port isn’t in plans at the moment

  4. joe Avatar
    joe

    +.1 for mul­ti mesh

  5. L3mming Avatar
    L3mming

    It’s so cool what you have changed and added in FreeCAD and the FEM work­bench! Real­ly look­ing for­ward to the release of ver­sion 1.0.

    I have to agree to a poster before though: CFD would be incred­i­ble! If CFD would be inte­grat­ed into FreeCAD, then we’d final­ly have a full pack­age of rea­son­ably easy to use but still pow­er­ful open source engi­neer­ing tools avail­able for hobbyists:

    - KiCAD for elec­tri­cal design/PCBs (I absolute­ly love it, it’s amaz­ing how far it has come in the past cou­ple of years)
    — FreeCAD for mechan­i­cal design
    — FreeCAD FEM work­bench for struc­tur­al analysis
    — FreeCAD CFD for flow analysis

    As a hob­by­ist, one could then build near­ly any­thing with­out hav­ing to know every detail of cur­rent open source soft­ware or spend­ing a for­tune on com­mer­cial soft­ware. That would be real­ly great.

    1. NewJoker Avatar
      NewJoker

      We already have add-on Cfd­OF work­bench which serves as a pre­proces­sor for Open­FOAM and is real­ly great. I would appre­ci­ate it being built-in but its dev does­n’t have such plans for now. Adding prop­er CFD to the FEM work­bench itself would require a lot of work and imple­ment­ing the same solver as the one used by Cfd­OF so it’s also not fea­si­ble for now. I’d like to extend the sup­port for Cal­culiX instead since many fea­tures for struc­tur­al analy­ses are still missing.

Discover more from FreeCAD News

Subscribe now to keep reading and get access to the full archive.

Continue reading