Void thickness

Discussion in 'Technologies and Hardware' started by Magic, Mar 24, 2009.

  1. Magic
    Magic Well-Known Member
    Hi all,

    I wanted to try some simple object with a moving part, and so I designed a cube with an embedded sphere.
    The difference of radius between the sphere and the spherical hole of the cube is 0.5mm (the diameter of the sphere is 2cm).
    I ordered two prints of the same model with different materials.
    I received my first prints, and here are the results:
    cubospheres.jpg
    [list type=disc]
    [*] the one in Tranparent Detail material is perfect: the sphere is freely moving inside the cube (and I can spend a lot of time playing with that :D )
    [*] The one made of White Strong & Flexible material instead is just one solid piece, with no moving part :cry: .
    [/list]
    I have seen topics about wall thickness, but what about "void" thickness? I mean what is the minimum amount of space you should have between 2 objects to be sure they will be disconnected ?
     
  2. Magic
    Magic Well-Known Member
    Oops :blush: , I've just seen the message from Whystler ($50 Experiment - ball & socket ) that confirms that:
    - 0.5mm is enough for Transparent Detail but not for White Strong & Flexible.
    Is it possible to update the material table with this information?
     
  3. frankbuss
    frankbuss New Member
    There is a blog entry which says 0.25 is sufficient:

    http://www.shapeways.com/blog/archives/141-Creating-hinges-a nd-moving-parts.html

    and a thread about clearance:

    http://www.shapeways.com/forum/index.php?t=msg&goto=2194

    The thread from Whystler you mentioned:

    http://www.shapeways.com/forum/index.php?t=msg&goto=2070

    Looks like the key point from this thread is, that you need to provide extra air gaps for the powder to be removed, then even smaller distances than 0.5 mm are possible.

    Maybe this depends on how big the contact area is, too. If it is too big, 0.5 mm is not sufficient for WSL. Can you remove the powder with a needle? Can you cut it to see if it is really fused?
     
  4. jdavidbush
    jdavidbush New Member
    From my experience, minimum gap thickness (for WS&F) really depends on two things: surface area (between contact surfaces) and print direction.

    Here are some very small test models:
    IMG_4298.jpg

    In this picture, the gaps are 2.0, 1.8, 1.6, 1.4, 1.2, 1.0, 0.8, 0.6 and 0.4mm.
    In WS&F, only the 0.4mm gap was (at least partially) fused. All the other gaps (0.6mm and up) printed just fine. I didn't have a 0.5mm gap, so I don't know if would have been fused or not.

    Also, the uploaded model was 3mm (H) x 11mm x 21.8mm. The measured dimensions of the printed model are 3.45mm (H) x 11.11mm x 21.90mm. So, the width and the depth are only about 0.1mm larger but the height (in the vertical print direction) is 0.45mm larger.

    And, that's why print direction plays a part in minimum gap thickness. The "resolution" of the printer in a horizontal direction is better than its resolution in the vertical direction.
     
  5. jdavidbush
    jdavidbush New Member
    Here is a little more experimental data. I created these models to test minimum gap thickness around an "axle".

    View attachment 380

    The axles are 2.0 mm in diameter and the gaps are 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 and 1.0mm (on both sides of the axle).

    For WS&F, all the gaps 0.5mm and smaller were fused. The gaps 0.6mm and larger printed (and turn) just fine.

    The cylindrical holes that the axles pass through are all 10mm long (that'll give you an idea how how much surface area is shared between the two surfaces).

    The size of the largest "handle" was modeled at 8mm (H) x 8mm but is measured to be 8.15mm x 7.99 mm. It is about 0.15mm too large in the vertical direction. (That's just more evidence that the printer has better resolution in the horizontal direction.)
     
  6. jdavidbush
    jdavidbush New Member
    Hi Magic.

    I apologize for hijacking your thread here, but I thought it would be a good opportunity to share one more test print for moving parts.

    I have one more print that is similar to the prints in the last post with a few changes.

    IMG_4310.jpg

    In this model, the axles have diameters of 5mm and the gaps are 0.75, 0.7, 0.65, 0.6, 0.55, 0.5, 0.45, 0.4, 0.35, 0.3, and 0.25mm.

    I built this model in the shape of an "L" so that it would be shorter than it was wide or deep. (I was hoping Shapeways would stage it so the shortest dimension was along the vertical axis and the axels would be "standing up" when it was printed)

    Unfortunately, the model was still printed with the axels "laying down" and all but the 0.75mm gaps were fused.

    The cylindrical holes that the axles pass through were all 20mm long.

    Anyways, I hope this shows that the minimum "void thickness" really depends on a number of factors. So far, I haven't run into any problems where 0.75mm was too small of a gap (although it makes a pretty wobbly axle.)

    -David
     
  7. frankbuss
    frankbuss New Member
    David,

    thanks for your idea for the printing direction. I assume Shapeways rotates it always that the height is minimum, because then the printing process is faster (at least looks like my Case model has slightly visible stripes perpendicular to the minimum height, in spite of the preview in my model page, where the minimum height was horizontal oriented). So chances are good that my Balance Weighing Scale works :cool:
     
  8. Magic
    Magic Well-Known Member
    Thank you very much for your answers.
    From what I undestand, with White Strong and Flexible material, 0.5mm clearance can work most of the time but can be fused in the worst case (depending on printing direction, friction surface, no extra air gap etc.).
    In my example, as the moving object is a sphere, I guess I am always in the worst case :( .
    Instead, 0.75mm should be always fine.

    PS: I double-checked and the WSF sphere seems really fused into the cube but I am sorry, I am not cruel enough to cut my very first creation, even if it is not perfect :)
     
    Last edited: Mar 25, 2009
  9. Whystler
    Whystler New Member
    Hey David,

    I would love to be able to buy your first axel test.

    -Whystler
     
  10. jdavidbush
    jdavidbush New Member
  11. Whystler
    Whystler New Member
    Ordered. Thank you!

    -Whystler