Model orientation

Discussion in 'Software and Applications' started by 4477_deleted, Aug 11, 2008.

  1. 4477_deleted
    4477_deleted Member
    Hi,

    I made this statue in blender:
    [​IMG]
    the up direction is along the (positive) Z axis, which how it's mostly done I presume

    After converting to STL and uploading it to shapeways, the generated preview however displays the statue flat on its nose:
    https://www.shapeways.com/model/2700/druplicon_statue.html

    Is this expected (because I got my XYZ-axis wrong)? Is this because of the server side postprocessing? And will it be printed that way?

    I added the foot (or should I call it pedestal?) at the bottom especially so that the drop shape could stand fixed (and could be printed without rolling around).

    Thanks
     
  2. 3194_deleted
    3194_deleted Member
    Too bad you do not allow seeing your model, I don't know how to help you.
     
  3. 4477_deleted
    4477_deleted Member
    oh, sorry
    I forgot to enable it as a public model :rolleyes:
    should be fixed now

    thanks for your time
     
  4. bartv
    bartv Member
    Hi saxofaan,

    yeah, that's a result of our server processing - we'll be adding better controls for that in the near future. In the mean time, don't worry about the print result, our operators always manually re-orient each model for the best printing quality.

    Cheers,

    Bart
     
  5. 4184_deleted
    4184_deleted Member
    I was worried about that, so it is not the most economical orientation, then? For example, if I print a gear, it is printed flat like you draw it on paper?
     
  6. 4808_deleted
    4808_deleted Member
    Usually, yes. One of the things the printers need to do is keep the support structures to minimum. So a gear would be printed flat because that needs no support structure during the printing process.

    As for the axis directions - the conventions of pretty much every 3D file model format is that X and Z form the horizontal plane with Y pointing upwards (unless it comes from the geospatial community, which then has Z up). So we don't do any processing on the coordinate axis because knowing what is "up" is impossible to determine automatically.
     
  7. 4477_deleted
    4477_deleted Member
  8. 4184_deleted
    4184_deleted Member
    Thanks, I didn't think about that. Makes sense.
     
  9. 4808_deleted
    4808_deleted Member
    So, funny coincidence. Here I am sitting in a siggrah paper presentation titled: Upright orientation of Man-made Objects. Basically automating this problem of determing which way is up.
     
    Last edited: Aug 13, 2008
  10. 4184_deleted
    4184_deleted Member
    Well, apparently not.

    I've printed a mutilated gear and it is printed in the least preferable orientation.

    I can not make a picture of the gear because my camera just can't focus and can't see the details, but I've made an image to explain.

    To get good results in gear printing I think the orientation is critical. I've built a router a few years ago (when I had little money and a lot of time) for about 1000 euro. The result of the milling is better than the printed gear. And that is not what I want (now I have more money and less time, I do not want to make them myself).

    This single gear I've printed will probably work, I'll assemble it tomorrow. But then again, the final design I have in mind uses a lot of gears. And then I'm worried.

    I'll probably order again to try out other things, but still...

     

    Attached Files:

    Last edited: Apr 13, 2009