Stuck? 
support@mecway.com
support@mecway.com
Mecway
FEA
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
support@mecway.com
FEA
It looks like you're new here. If you want to get involved, click one of these buttons!
Comments
The three examples you have shown are a little bit different, you have chosen 20 color ranges in Mecway and in, picture 2 has got 16 color ranges and picture 3 has 21. Unify the amount of color ranges for all of them and compare then, I am interested how different it will be.
I remember when I had a bit of numerical methods at uni and we used turbo pascal to solve and "plot" values as an array of numbers, then having a contur display as in novadays software is like being in Haven.
Personally I think that coulours in Mecway are really nice and warm
You can use clip planes in FEA as well and some funky feauture displaying range of isofields dynamicaly and easily expand your field of e.g. stress fields from where it is highest untill it covers whole body or just a set range of delta20 MPa etc. This is why I was mantionning it to Victor, that it would be very useful in Mecway.
When you have a singularity and your colour bar shows really high red values but the model is all in blue, even if you spin it hundreds of times, then great option is to switch to dynamic iso fields and start scrolling scroll button in the mouse so the model turns transparent and you can start scrolling up so highest values show first, then those lower ones untill you reach the bottom range.
I think that constructive criticism is what Mecway needs from our comunity so Victor can be succesfull and we can have better and better product and a lot of satisfaction that we were/are part of the story. Right now I am trying to crack on with OpenFOAM and when I get there I will combine Mecway with OF as pre procesor and as Structural part of FSI simulation. It's a long way in front of me but it's worth it.
Take care Guys and thank you very much for Mecway Visctor!
I think that being able to rename positions in the tree could be helpfull as well.Generic information about certain possitions can be quite confusing if there is lot more branches then just few. E.G. I am creating now constrain equations and sets of them don't differ at all if it comes to names.
Thanks,
apadzak
I'll add some more flexibility for colors in v4 - possibly an option with more distinct colors or a customizable set. I agree there's too much similar looking green. It's really only 5 colors with everything else being interpolated between those. Using different darkness looks like a great idea to get more visible contour lines.
apadzak:
You can already turn component visibility on and off in the solution without re-solving. But you have to use the separate "Components" item under the Solution branch in the tree. I know it's weird having two of them. It's because otherwise you'd be able to do things like delete a component from the model then be unable to change its visibility in the solution anymore.
As for constraint equation, would it be of some use to have them consolidated into a list rather than tree nodes and you'd see the equations themselves instead of their names? The reason I suggest this is that there are uses where the number of equations might become huge - into the thousands or so. Though that's probably not going to be very helpful if you have several sets of similar looking equations and mainly want to distinguish the sets.
About scrolling through the colors, what you describe sounds quite similar to what the sliders on the color plot already do. Once you click one of the two thumbs, you can use the mouse wheel to change the range. I find myself using this for exactly the case you described - a small but high peak and everything else near zero. Sometimes the step size isn't fine enough though.
VMH:
Version 4 will have multiple selections for geometry surfaces but probably not the multi-select in the outline tree or quicker bonded contact setup. I know these features will be good but they're also disproportionately hard.
i would like to say sorry, may i has been chose inappropriate words to explains my opinion. please don't take seriously, english is not my mother languages.
back to topic,
for ParaView of course yes
hi Victor,
thank you for concern in colour contour ranges.
greetings,
about contact between parts, i seeing this problem from many others FE software document. many developer says this problem's harder to solve than plasticity, due to convergences of contact springs force especially in quadratic element. MECWAY already have contact features using frictionless support, but it's limited to perfect rigid or infinity spring stiffness.
hi Victor,
i have some minor suggestion, MECWAY already good in automatic planar 2d meshing. it'll be great if it has further calculation for arbitary section properties (Area, Inertia, Elastic&Plastic Section Modulus, Torsional & Warping constant, neutral axes, shear centre,etc). Calculation torsional properties only possible using FE software, hand calculation are just an approach and conservative because fillet radius or chamfer are ignored. another problem are many formula exist has limitation.
greetings,
Thanks for the suggestion about section properties. I'll add it to the list. You can already calculate areas using Tools -> Surface area.
ups, i forget. thank for correcting me. in MECWAY "frictionless support" works as rolled, node dof's translation in perpendicular face support are restrained for both compression and tension. it's different with "compression only support" i verified by simple case (attached).
thanks for concerning section properties calculation. yes i checked, it's already for area calculation from group of selected faces.
other suggestion from my old posting about concrete shell reinforcement, i found more general method to calculate using sandwich analogy were it's already available (?) but i didn't not found plate sandwich in MECWAY documentation. it's already available in previous program LISA.
in sandwich model, calculation of reinforcement bars are not too complicate. plate bending moment (M) only converted to equivalent imaginary membrane force (N) by dividing with thickness (d) of core layer sandwich. tensile strength (fy) of reinforcing bars need to be defined by user. Bar areas requirement (A) can be simple dividing membrane force with tensile strength.
Example for plate bending ignoring twisting moment and only for bottom reinforcement,
M = 1*10^7 N*mm
d = 100 mm (*user defined)
fy = 200 N/mm^2 (*user defined)
T = M/d = 1*10^5 N
A = T/fy = 500 mm^2
if you're interested and concern i have and can sending a documentation about the background theory of concrete shell reinforcement which including effect of twisting moment and membrane force. or you can ignored if it's out of the radar
greetings,
The sandwich material in LISA is/was just a special case of laminate, which Mecway has. However a soft core with stiff skin can be very inaccurate so this is often a bad way to model an actual sandwich panel. See this other discussion showing the problem http://mecway.com/forum/discussion/comment/261/#Comment_261
I don't understand what you're saying about using a sandwich model for rebars. If you can make use of the laminate material for that, all the best, but beware of the limitation shown in that other thread.
johan:
[edit: deleted my misunderstanding about units]
I haven't heard of a need for normal force or stress per unit length before. They don't sound too useful because the stress profile through the thickness is, in general, not uniform so there would be regions of high stress that don't appear in such summary results. Can you give a bit more detail on the use case for them?
Principal stress visualization is something that I realize is valuable but haven't got to yet. Maybe in a future version after 4.
Regards
johan: I see what you mean now, thanks for the suggestion. I might try to add it as a labs option in v4 but not sure. What are the useful ways to represent the stress in this case? von Mises, components, etc?
i have another minor suggestion:
MECWAY already has great feature in elastic spring support and compression only support, it will be useful if result reactions are also displayed in stress contour for contact representation. this can be done by dividing reaction force (already available) with tributary areas (already knows) of the spring itself.
greetings,
At the moment I am tinkering with MecWay demo version. It seems to be a very useful tool in an education environment. Even small 3D-analyses are possible, so homework can be given to students. The price tag for non commercial is not too high (even for students!).
I want to begin with a contribution to the wishlist, although this might look greedy. In this thread I read the wish for UNV export. For me, a UNV import might be even more useful. We all know that preprocessing is the weak point of all lowcost systems ... There is a piece of software called Salome (maybe you have heard of), developed mainly in France as a preprocessor for Code_Aster (which is powerful, but not very user friendly) in a Unix-world. Salome comes as Windows binary, too, and it has really nice preprocessing properties. The only generally usable output format of Salome is UNV. Exported items are meshes and grouped mesh entities (nodes, elements). No boundary conditions, no physical properties! So it seems to be not too difficult to implement a UNV import interface. I know about Netgen, and Salome contains Netgen. Anyway, Salome looks far more CAD-like, so I think it is better suited than Netgen.
So much for now, greetz,
best
hp
Best Regards