Stuck? 
support@mecway.com
support@mecway.com
Mecway
FEA
JohnM
About
- Username
- JohnM
- Joined
- Visits
- 2,606
- Last Active
- Roles
- Member
Comments
-
@Sergio, thanks for yet another great validation of the solver!
-
here is an old benchmark that looks similar to yours: https://mecway.com/forum/discussion/750/propeller-hub/p1 I ran the cube with PASTIX 1 proc - 21% CPU usage 1:45 6 proc - 70% CPU usage 0:58 similar with PARDISO
-
Ran a few tests: bolted joint with contact, 100k DOF run times ccx_PARDISO.exe from calculix_2.17.1 with 1 processor 11:55 ccx_PARDISO.exe from calculix_2.17.1 with 6 processor 6:58 ccx_PASTIX 2.17 by 3rav 31 aug 2021 with 6 processor …
-
Good option, thanks Victor! We looked at the *SECTION PRINT command, which would be handy for this application, but it is very dependent on mesh density to get an answer close to correct. We often have curved surfaces, so the current "trick" is to…
-
Here's another one for you consideration. This is a test problem that is very "fragile". In other words, it can crash pretty easily. Some things I know - - Use auto-time stepping with no lower bound but restrict the upper time step (it will pro…
-
Pin/ bushing work best with hexahedral geometries when possible. Elements with mid side nodes will also help preserve the curvature characterization. Tetrahedral meshes, especially if not very fine, can give some strange results. Here is anothe…
-
Your press fit is a deflection-controlled problem, meaning it's important for the contact stiffness to be high enough that the penetration is low compared to your interference. For instance, if a 10mm pin is pressed into a 9.98mm bushing, the 0.01mm…
-
If you are not doing modal, try using Pastix solver - there is a thread on this.
-
The strain direction has the same issue as when you look at stresses - they often don't align with a global axis. You can look at principal strain (I would look at max prin strain for proof/burst limits) but to get the direction you will have to in…
-
Linear gages measure linear strain, rosettes will provide biaxial results. You should be able to look at linear strain, but you want to extract that result across the actual gage length. You can literally glue small soft elements and look at the del…
-
I think if you put a new step file with the same name in the directory and use the reload feature, it updates the geometry. I have found though that there is no guarantee that all of the surfaces will line up, but most of them often do.
-
Modal results are usually Mass Normalized, I think that is the case with Calculix. Look up Mass normalization, you'll find plenty of references. I'm a little rusty on this stuff but I also think you can obtain participation factors that allow you to…
-
The spiral sweep for threads works but you will find you need to leave a pin hole down the center of the bolt. Another option is to leave a larger cylindrical dead volume and fill this void with tet/pyramid elements. Also, you will find the beginnin…
-
Transfer displacements from solution (one of Victor's examples) # This example script replicates the functionality of # # Mesh tools -> Transfer displacements from solution # # but is limited to solutions without time steps or modes an…
-
Solve then Save mw.solve() mw.save()
-
Modify selected nodes to the same constant X value (for y and z edit the "x" lines below def mod_node(nodes, value): for point in nodes: mw.set_node_x(point, value) active_nodes = mw.selected_nodes() # Get selected nodes if len(activ…
-
Python script to create center node of circle between 3 points. I just adapted this from old FORTRAN (no jokes), and it "seems" to work. I can't tell you what it is doing with more than 3 nodes selected - I tested a few examples and it wasn't clear.…
-
Calculix expands beams and shells into solids, so you will find a lot of new nodes. This discussion may help: https://mecway.com/forum/discussion/1014/exporting-deformed-mesh-of-shell-model
-
We use the "12.000000123" hack, just double check how the number actually looks inside the liml file before you start scripting.
-
Static 3D is going to use small deflection theory, switch to nonlinear static 3D.
-
How did I miss that one all these years?!
-
@3rav, what version of pthreads are you using?
-
THANKS!
-
From our company intranet: Basically, run PASTIX but keep PARDISO handy We have found that 4-6 processors is useful, over that is diminishing return. For model size, we try to keep things under 500k nodes.
-
Just looked at that paper, fascinating stuff. I think it's useful to run the modal analysis first, as this will give you some idea about your required time stepping. For example: 1/(frequency * 10) might be a good start.
-
Attached is a simple example using temperature dependent properties using CCX custom model definition. Run with different temperatures, and deflections will change. CCX custom commands open up a lot of potential to extend your capabilities, so it'…
-
@3rav, we'd like to try a modification to CCX that I discussed with Guido, but we are not proficient at compiling a Windows executable. Can you provide a recipe here?
-
I think your mesh did not make solids, Mecway thinks your model is shell elements. You have multiple parts, they need to be read in as individual STEP files and meshed separately. You likely got a "Gmsh may be bad" message at the end of your meshin…
-
@prop_design F77 here as well. But everyone around here is a JavaKid or PythonKid. They've missed out on the thrill of debugging a 5000 line spaghetti code
-
Python script to create center node of circle between 3 points. I just adapted this from old FORTRAN (no jokes), and it "seems" to work. I can't tell you what it is doing with more than 3 nodes selected - I tested a few examples and it wasn't clear…
