Mecway 31

edited January 15
Mecway 31 32 is now available at the link below.

https://mecway.com/download
https://mecway.com/download/mecway320.msi


Changes
Allows multiple custom model definition/step contents/engine input/bulk data for external solvers.
Revolve tool merges on-axis nodes and corrects the shapes of their elements.
Mirror tool with copy has an option to merge nodes on the mirror plane.
Automesh 2D/3D can generate linear elements. The option is also available in the API.
Solution data from external solvers is stored in single precision which reduces file size, memory use, and load/save time.
Solution data with multiple time steps or modes is stored more efficiently in the .liml file.
Strain measures are identified in the outline tree for nonlinear analysis types and in the material stress/strain curve input.

Bug fixes
Couldn't open assemblies with Alibre Atom3D. It showed "License key is not found or expired or invalid for attempted operation".
Bug 47. Accounts for a change in the OpenRadioss solution beam moment definition according to the release version number which you specify. It also shows warnings on beam moments which are incorrect with some versions of OpenRadioss.
Split step assembly failed with GEOMETRICALLY_BOUNDED_WIREFRAME_SHAPE_REPRESENTATION.
If the solution contained global values (internal energy, kinetic energy, or time step) but no node or element values, then none of it was saved to the .liml file.

Comments

  • OOooh, Great!
    I see that you did multiple things to reduce the file size and I am very grateful for that as the file sizes for the OpenRadioss crashbox simulations I did before christmas was huge and it gave me issues when opening the files.
    The file size improvement can be obtained for existing files by opening them and saving them again. In one case, the file size reduced from 12,7 Gb to 9,8 Gb, so while it is still a large file, a 23% size reduction is very nice! I have lots of other files from iteratively developing those crashboxes that are around 20 Gb and a few that are as large as 32 Gb, so now I just have to open them and save them again to release about 100 Gb of space. It might be ok for me to outright delete some of the iterations though...

    Also thanks for the multiple custom model definitions/engine input and so on. For those who doesn't follow the github page where we can request improvements to Mecway, the benefit of this is to be able to make custom contacts, timestep size control and so on and be able to suppress them easily it as they are separate features in the feature tree.
  • Good to hear it helps, and thanks for showing me what a problem file size was. There's even greater saving after solving in v31. Just doing open and save still keeps it in double precision to ensure fidelity but solving again goes to single.
  • I'm glad to hear that. I don't think I will rerun the files just to save some space but it is good to know that the space savings are higher from rerunning the analysis. Single precision is half the size of double precision, right?
    If it halves the file size, then that's a massive saving.
  • A 20 Gb file reduced 25% in size to 15 Gb by opening and saving. I will rerun it and report back what the resulting file size is.
  • The file that was 20 Gb in version 30 is now 7,8 Gb after being rerun in version 31. So it is true that the single precision results takes up about half as much space as the double precision results and that this is on top of the 20-25% reduction achieved by the improvements in how the data is stored, which is obtainable without rerunning. :)
    I am very happy with these smaller file sizes :D

    I haven't timed the time to open the files but I can say that they do open much quicker as well. :D
  • edited January 14
    I timed the time to open. The old 20 Gb file was 6 min 30 sec to open while the new 7,8 Gb file was 1 min 45 sec.
  • I've had a few days working with V31, up to now I have seen no issues. However this afternoon I had the program crash. I got this error
    Error. Please save your work in a different file, close Mecway and report this to support@mecway.com

    The given key was not present in the dictionary.

    Version 31
    System.Collections.Generic.KeyNotFoundException
    System.ThrowHelper.ThrowKeyNotFoundException
    System.Collections.Generic.Dictionary`2.get_Item
    SelectionPropertyCollection.ManagerTreeNodes
    ModelData.ManagerTreeNodes
    ProjectClass.ManagerTreeNodes
    ViewPort.RefreshAll
    frmMain.pnlRender_Layout
    Control.OnLayout
    Control.PerformLayout
    ScrollableControl.OnVisibleChanged
    Control.OnParentVisibleChanged
    Control.OnVisibleChanged
    Control.OnParentVisibleChanged
    Control.OnVisibleChanged
    Control.OnParentVisibleChanged
    Control.OnVisibleChanged
    Control.OnParentVisibleChanged
    Control.OnVisibleChanged
    Control.OnParentVisibleChanged
    Control.OnVisibleChanged

    It may have nothing to do with the new version but I was able to get back into the program with an old version and the new version does not appear to want to open it again. I was in the middle on making some loads a constraingts and was trying to make a new group - it was a big file with 300,000 or so elements so that may also have something to do with it
  • Oh! This is a new bug with Groups in Loads & Constraints. Thanks for reporting it. I'll make a quick update soon.
  • Because of this serious bug Fatmac found in loads & constraints groups, I've rolled back publishing version 31 on the download page.

    The bug is new fixed in version 32 which is available here https://mecway.com/download/mecway320.msi and that will eventually become the next published version if no other new bugs are found.

    Sorry for the trouble.
  • edited January 26
    Hi Victor, thnks for this update. I have installed and there are some visual glitch in the pretension bc. I have applied in several horizontal bolts, and they are represented in the center position (paralell to X), but for the vertical ones (Z direction), they are very far away from the selected faces.

    EDITED: I have meshed with coincidents nodes bolts and base, so some of the faces that define the pretension force was shared between two components. Once I remove those shared faces from the selection, the visual representation is ok.

    Best regards
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!