Stuck? 
support@mecway.com
support@mecway.com
Mecway
FEA
Victor
About
- Username
- Victor
- Joined
- Visits
- 4,659
- Last Active
- Roles
- Administrator
Comments
-
Yes. AlT+V, S now toggles element surfaces
-
Oh. I understand now. Thanks! I'll fix that.
-
There's now an option for the Gmsh scaling change in the Options window.
-
Yes, shell edges have to be the slave. The should also be not much longer than the master faces to ensure all master nodes are connected to a slave node.
-
Probably, yes. A slave node can only belong to one bonded contact.
-
Here are the reasons I think. I only tested the stress at the X=0 edge which did get smoother with refinement.
-
Yes. The 6 zero-Hertz modes correspond to the 6 rigid body degrees of freedom that a completely unconstrained object has. Displacements and stresses are only relative to each other so the scale doesn't matter.
-
Error -9 from modal analysis can mean that density is 0. Make sure there are no red items in the outline tree. A zero density should be shown as the material name being red.
-
Yes, only normal.
-
The initial temperatures don't cover the entire model (all nodes) so some parts, especially the interior, are still at the default 0 K and the extreme temperature change between 20°F on the surface and a uniform -460°F on the interior causes error. …
-
You may be only looking at the first time step which shows the initial state. Dragging the slider on the timeline at the bottom of the screen shows higher temperatures. You may also want to set an initial temperature on all elements instead of the …
-
It's because one of the bonded contacts is connecting the Trace part to the opposite side of the pipe. If you click the bonded contact in the outline tree, it shows lines crossing the diameter. You should remove the opposite part of the pipe from th…
-
That's strange. I don't understand either. He's my one that solves.
-
No stress usually means it didn't converge. The stiffener comes away because the stiffness of the bonded contact joint is too low. I changed it to 1000 GPa/m and it stays pretty close. I also changed the pressure load from constant to ramped by re…
-
A common way to get things to mesh is to set the max. element size. Here I used 0.5m and it worked.
-
Don't use the internal solver for nonlinear. It's inferior to CCX in just about every way. *INFO and *WARNING messages are often normal and harmless. These ones look OK. *ERROR is what makes it fail. This one looks like it might be the same proble…
-
That's correct. Both the elastic and *TIE options close up the gap in a strain-free way. EDIT: By leaving a gap, you're specifying something physically impossible so the software has some liberty in deciding what to do with it. The default option f…
-
It's the dot product of displacement with the unit radial vector.
-
Signed radial displacement is (x*u.x + z*u.z) / sqrt(x^2+z^2)
-
It turns out the Y displacement constraint on the adjacent nodes is what's causing it. That error message was really misleading, sorry. I replaced it with a 10 000 GPa/m elastic support on the whole bottom surface of the 50rsa2 component. I've also…
-
Thanks for reporting this bug. It's a known one that will be fixed in version 10. It happens with the Gradient contour plot option (not the Banded one) and a formula that has a divide by zero or similar problem.
-
You can track it down using Edit -> Select nodes by number and enter 3498. Then see if either: * There are conflicting constraints which aren't physically possible, such as displacement by two different distances in the same direction or * T…
-
Here are some tests. Connectivity of the mesh has a huge effect on memory requirements so I used two models at opposite extremes - a cube and a 1-element-thick plate. Real models would usually be somewhere between the two. Your 250000 nodes sounds …
-
Oh. That does sound wrong, unless you have ~30 GB or more of RAM, in which case it's reasonable to have 10 GB unused by both CCX and the internal solver. I'm running some benchmarks and will post graphs tomorrow showing my observations of memory us…
-
This number of nodes looks like about the limit for default CCX which I think is restricted to 16 GB so you might be over the limit with only 10 GB. I can't reproduce your model because the STEP file isn't included in the .liml file but I tried it …
-
There's no direct CoG readout, so you might have to do it by supporting it at a couple of points and using the reaction forces to calculate the location of the line of action for the weight load.
-
1. That's an artifact of the display. It displays the offset shell with the same shape as a non-offset shell but different position. Since the surface becomes curved with Open cracks, that causes gaps between them. 2. Yes except except for stress a…
-
Yes, you can use imperfections to cause buckling in FEA, but with nonlinear analysis, not with linear. Perhaps you're looking for wrinkling like this? - No, this isn't buckling because linear static analysis can't do that. Displacements are prop…
-
The number one rule in FEA is that it's wrong until you estimate the size of the error. The basic reliable way to that is with a mesh convergence study. Here, you only have a single element in the thickness direction which is probably not enough for…
-
Yes. 3-node beam elements are only for the CCX solver. There are also line3 thermal fin elements for the internal solver.
