Version 4 beta 2

Victor

Thanks everyone who gave feedback on the first beta. I've incorporated many of your suggestions in beta 2 which is available from http://mecway.com/download/app/mecway40beta2.msi

Changes:

CCX support:
*STEP uses the INC parameter to allow more than 100 increments to be solved.
Custom model definition and custom step section input boxes show the units that must be used there.
Strain output is turned on with nonlinear analysis.
Face and element selections are carried through to the solution.
Recognizes DISPR and STRESSR in .frd files for compatibility with CCX 2.7
Bug fix: custom model definition and custom step section buttons were hidden on some resolution settings.
Bug fix: File->Import imports the items under the CCX branch of the outline tree.

Other:
Factor of safety is reported according to either von Mises (for solids) or Tsai-Wu (for shells) failure criteria, including first ply failure for laminates.
6 view orientation buttons added to the tool bar.
If it can't find a linked STEP/IGES file, an option to change the path appears in the context menu.

[Deleted User]

Very nice

VMH

Victor,
General: can we also add the Isometric View in the Tool Bar. I actually use that and the zoom fit the most?

For the quick tests:

Referencing to filename = beamcontact_try1

1. The element named sets are not carried to the solutions when using CCX: they grouped all of the mesh sets together in the default set in the solutions (see Plot1).
2. The vector total strains (TOSTRAIN) are automatically included in the solution of CCX but have "SIunits" (See Plot2).

Referencing to filename = beamcontact_try2

3. When adding material nonlinearity to beamcontact_try1 (saved as beamcontact_try2), the contact boundary conditions that were active as shown on Items 1 and 2 are no longer active (see Plot3)? Did I do something wrong.
4. For the material nonlinearity analysis, are equivalent plastic strain (PEEQ) automatically include in the solutions. If not, could we have them for material nonlinear analysis.
5. For the material nonlinearity analysis, is it possible to run static analysis with no geometric nonlinearity but with material nonlinearity that have equivalent plastic strain (PEEQ) outputs. Sometime with don't have large deflections but still have material stresses in the plastic ranges.

The files are also attached for references.
Thanks!

VMH

Victor, please disregard comment on Item 3. I was comparing "apples" with "oranges" (using entirely different young modulus). Also disregard comment on Item 5. In CCX manual, "If any of the materials defined in the input deck is not a linear elastic material, geometric nonlinearities are automatically taken into account (i.e. NLGEOM is activated)."

I made the corrections and changed the applied load and deleted one of the top supports for the comparison (attached).

Victor

Thanks for the detailed comments VMH

OK to reinstating the isometric view button.

1. It only does named selections for now, not components. This should be enough to address the problem of filtering the results in the table. I realize having components too will be better but maybe for the next version unless it's important.

2. I agree "SIunits" is a bit ugly. Low priority to change though. It means use the unit "1" which is the SI unit for strain.

4. No PEEQ. The reason is that to use plastic strain, you have to define a material manually in the custom model definition. So for typical situations it will be useless and shouldn't appear. If you're already writing a custom model definition, it's only a small step to also add *EL FILE to the custom step definition.

5. I'm not sure if CCX allows this or how you would specify it. What value would it have? Speed up convergence?

VMH

Victor, thanks for the reply.

1. Having components will be great for turning on and offto see the results of each component individually and also interface results of the connecting components. I agree that it's not a priority.
2. Can we include an option for a blank dropdown for user to manually select to hide it? I agree that it's not a priority.
4. I was interested in Equivalent Strain (vonMises) in Ansys or Strain Magnitude (PEMAG) in Abaqus to see the the strain as a whole instead of each direction (xx,yy,zz,etc.). Maybe Calculix cant do that yet.
5. Yes, that's just to reduce the runtime if we don't have to run geometric nonlinearity. But CCX automaticaly turn on geometric non-linearity if using material nonlinearity.

Thanks for your reply.

DaveStupple

Good effort Victor, graphics speed is a a big improvement: I I will upgrade as soon as I get the nod here. I would like to see each component in the tree with an off/on button controlling visibility (like for layers in Gimp GNU) - right clicking is a pain (though the added visibility options are good). Is there a way to reduce the transparency of small elements so that hidden edges are less visible (see attached)?

Victor

Thanks for the suggestion about buttons. This would need to be configurable because most people won't need them. I'll keep it in mind for a possible future UI redesign.

Seeing the lines through the faces is certainly a problem with small elements. I'll try to add a labs tool to manually adjust this but it won't be properly fixed for v4.

protontim

Hi Victor,

Great work as usual. I was wondering if you might consider adding support for a 3D mouse like the 3D Connexion products. This allows for simple examination of the part in 3D space.

Thanks,
Tim

Victor

Last time I looked into 3D Connexion, their SDK didn't work well with .Net, which Mecway uses. Things might have changed, so I'll try again sometime in the future. Quite a few people have asked for this from time to time.