A way to guarantee print orientation

Discussion in 'Technologies and Hardware' started by danzen, Oct 31, 2011.

  1. danzen
    danzen New Member
    Hi all,

    I'm new here and like many, would like my designs to be printed in a particular orientation.

    I'm mostly going to be printing mechanical parts; gears, rollers, wheels, etc. So the print orientation is rather important.

    I figured that since our designs get reorientated to optimize the space available in a tray, I could sprue or cage my parts together such that the overall bounding box is a perfect cube (equal length on all sides). This way, the packing algorithm (or human operator) would not reorient my design as it will not serve any purpose.

    I suppose the limitation is this will only work for smaller parts and when you need enough copies that you can arrange them more or less into a cube.
    The sprue or cage can then be design to ensure a perfect cubical bounding box.
    It's a lot more work, but I think its worth the trouble.

    What does everyone else think? Will this work?
    Last edited: Oct 31, 2011
  2. Bunrattypark
    Bunrattypark New Member
    I am interested in this too. I find that the orientation of printing is critical if my models are to be polished without getting damaged, as the wrong orientation leaves weak spots which are susceptible to breaking under polishing. The correct orientation means a successful polishing.

    I am beginning to realise that my ability to continue using the Shapeways service is dependent upon printing orientation.

    As the OP asks, can we organise a cube, and pay for that?
  3. danzen
    danzen New Member
    Hi Bunrattypark,

    I'm not sure you understood what I was trying to convey.

    I wasn't asking Shapeways to do anything, but rather proposing a method for us users to circumvent the inability to specify or fix the print orientation. This is a temporary solution until the feature is available (if it will ever be).

    Since designs will be rotated in 90° steps, no matter how a cube is rotated, it will not optimize the layout or packing at all. Therefore, I would imagine that a cube shaped object would be printed in the orientation that it was uploaded.

    So, an array of small parts arranged more or less in the shape of a cube and then enclosed in a cubical cage with equal sides will be less likely (?) be reorientated.

    There might be a problem with this idea if you don't pack the parts closely enough. I'm not sure, but it is possible that if there's a large hole or gap in the "cube", Shapeways might decide to reorient the "cube" such that the hole fits over a long, narrow part of some other design to optimize the space available.

    Anybody else thinks this will work?

    P.S. Btw, being new here, what does OP stand for?
  4. It's a general forum term, not a Shapeways-specific term. It stands for "Original Poster".

    Regarding the idea, since I believe the models are manually optimized for build, I suspect the cube would only make it more likely that orientation is in a given rotation. I seem to recall a post somewhere that materials optimize differently, rather than all based on height. If I find that post, I'll link it here.
  5. danzen
    danzen New Member
    Hmm...What else is optimized? I had the impression that the goal was to print as many models in one go as possible given a limited volume. So I my thoughts are that it is essentially a packing problem and the orientation of a cubical object is insignificant in solving it.

    Also, thinking about it, solving a packing problem manually doesn't seem very efficient. A decent packing algorithm can easily perform better and faster than most humans (no offense to the folks doing it) given the complexity and variety of sizes of the models.
  6. I looked for a while, but I couldn't find the comment I remember. Since memory can be disturbingly unreliable, I could be remembering something that never happened.

    I did find the thread below, and lots and lots of other examples of the exact same question. I don't know whether Shapeways uses a manual process to organize the prints. If Shapeways does do that, I don't know why. I do know that Shapeways seems to be continually seeking improvement, so I think with some patience, we'll see some improvement in this matter.

    http://www.shapeways.com/forum/index.php?t=msg&&th=6 634&goto=36065#msg_36159
  7. danzen
    danzen New Member
    I've seen that thread before. Actually, I had previously looked around for quite a bit and found that many have requested to fix the print orientation.

    Since it seems that's not going to happen anytime soon, I started to think about how I could "beat the system" and came up with that idea.

    I guess doing it as I suggest probably wouldn't guarantee 100% that your model will be printed in the desired orientation, but it should increase the likelihood significantly.

    Perhaps as a final bid, one could incorporate text labels onto the empty areas of the cage or sprue. The texts would read something like "â–²Pleeeease print this way up â–²" in hopes that the person optimizing the prints would notice (perhaps subconciously) and sympathize. ;)
  8. Bunrattypark
    Bunrattypark New Member
    I understand the point in the OP. However, even arranging models in a cube would appear to me to be no guarantee of a particular orientation.

    I also understand Shapeways very valid reason for not offering a choice of orientation.

    However, it is looking like it will mean the difference between using the service, or not using it at all. Shapeways will save no money if I remove my models altogether, so that negates the argument of greater efficiency.

    I realise now I need a guarantee of orientation before I upload my next model, which promises to sell a hundred or more. I have shown my work so far to prospective customers at a recent model exhibition, and have enormous interest in my next proposed model, a popular bus model. But the unanimous verdict is that wrongly orientated models are simply unacceptable to my customers.

    A hundred of my next model is about €5,000 - €6,000 worth of turnover to Shapeways, if I upload it and make it available. I cannot do so if half of them are orientated the wrong way, and I get a ream of complaints from my customers. I am the one who will be blamed.
  9. danzen
    danzen New Member
    I completely agree with you. I'm also holding back my very first order as I'm concerned it will not turn out right simply due to unfavorable print orientation.

  10. BillBedford
    BillBedford New Member
    So, what you are saying is that you are willing to forgo all these orders rather than redesign you models so that the orientation is less important?

    PS. from long experience of model railway shows, a hundred peple enthusing over a model will result in ten paying orders -- if you are lucky.
  11. danzen
    danzen New Member
    Perhaps I'm inexperienced, but I don't see how it can be done.

    Selling defective products results in unsatisfied customers demanding refunds. This not only means one does not gain any profit, but potentially having to fork out more than what was put in to begin with. I think Bunrattypark's logic is sound. Rather than lose money, it would be better to not try to make a profit to begin with.

    I have experience rapid prototyping with FDM and the print orientation matters a lot, not just aesthetically, but mechanically as well. I've yet to tried SLS and perhaps the machines that I had used before was inferior, but this is definitely a limitation inherent in all forms of 3D printing.

    A wheel printed with its cross section in the x-y plane will always be rounder and smoother. It might be less significant if the model was large, but not everyone wants/needs to print at that scale. Conversely, a slender, long rod printed upright would be weaker and more prone to breaking. Again, printing at a larger scale may solve the problem but is not always an option. There is simply no way to redesign such objects so that the orientation is less important.
  12. Bunrattypark
    Bunrattypark New Member
    These are customers who have been buying resin kits over a period of the last ten years or more, a very reliable market. I know most of these people personally. The advantage of producing different variants of one model is what gives Shapeways an advantage over resin, even at a higher price.

    However, it looks like I have to forgo the advantages of Shapeways, and look at resin production, because these models cannot be engineered to take any 3D orientation. They can go in one orientation only. Wrongly orientated model buses suffer a lot of breakage under polishing, due to the inherent weakness between layers. Vertical seat frames and stanchions shear off, and cannot be printed thicker without going too far overscale. They didn't break with the correct orientation. In addition, wrongly orientated models suffer from unsightly layer lines all along the length of the roof area, which are absent if the model is orientated correctly.

    Already I am left with a handful of models that were orientated wrongly, and returned to me by the customers as sub standard. That's all right with a few models. It's no good if I start releasing models with a potential of a hundred each or more.

    I have already produced a few models for the small Irish market. Those models have been seen by the much larger UK market, and there was much excitement initially about the possibilities of using Shapeways for model buses. However, if I cannot guarantee print orientation, this market will be lost to Shapeways, and it will go to cheaper but more restrictive resin production instead.

    I have been in discussion with a resin manufacturer, and a run of a hundred model buses is planned before Christmas. That is a hundred sales Shapeways could have had. It's a pity, but at present, I see only a small requirement for Shapeways models, due to the unreliability. I cannot afford to soak up the cost of returned models, or answer for sub standard quality.
  13. aeron203
    aeron203 New Member
    The cube concept is a good idea, but in reality there is a difference in efficiency with different orientations at the same volume. This would be different on different machines, but it would make the build longer if they are not optimized. If you were to print a bunch of vertical cylinders on most machines the build time would be far longer than printing them laying down.

    The difference would be measurable, so it seems possible that one could simply charge more per cc to compensate and get their model printed however they'd like. I am sure as they improve the software this will get addressed, but the more voices are calling for it the quicker that will happen.

    I also would like to be able to orient my prints in some specific cases, and would pay more for it. Though this issue has been raised before, if anyone else would benefit from being able to orient their prints, now is a good time to say so.

  14. danzen
    danzen New Member
    Thanks. Now that you mention it, I can imagine that the amount of support material required would also be significantly different for different model orientations, though this doesn't apply to SLS.

    I could be wrong but I believe that the laser in SLS goes through every column of every row anyways, so there's no time saving here. You could say that it takes time to deposit each layer of powder and hence it is desirable to minimize the height, but my guess is that Shapeways stacks models anyways to fully use the entire printing volume available. So again, no time saving. Of course, these are just my educated guesses, perhaps someone from Shapeways can shed some light on this?

    I agree. As of now, requests for fixing printing orientation is scattered throughout the forums. But even if enough people were to bug Shapeways for this feature, it will take time for it to actually happen, perhaps months.

    I hate waiting, but I hate paying a dear price for defective items even more. It would be great if someone from Shapeways can confirm that the cube idea could work or at the very least has a high chance (>80%?) of working, or even if it could work for some of the materials. I also honestly wouldn't mind any other temporary solution even if it requires more effort or additional but reasonable costs.
    Last edited: Nov 2, 2011
  15. robert
    robert New Member
    Hey guys,

    Setting the print direction is definitely something we want to offer but we need to be able to support this in our production process. It is more of an operational / administration issue than a printing / technology problem.

    At this moment the production planner makes the decision based on the best possible outcome. We do not optimize print direction for efficiency of building our trays or build platforms.
    Of course this is a manual process and depends on the judgement of the planner.


  16. GWMT
    GWMT New Member
    Bunrattypark said:

    "I have already produced a few models for the small Irish market. Those models have been seen by the much larger UK market, and there was much excitement initially about the possibilities of using Shapeways for model buses. However, if I cannot guarantee print orientation, this market will be lost to Shapeways, and it will go to cheaper but more restrictive resin production instead."

    I'm in the same situation. I've got dealers interested in what I can design but they are hesitant to set up accounts and order if the product they receive may be unsellable. If something is printed wrong and if Shapeways agrees to reprint it in the correct orientation the buyer's looking at about 2 months from when they paid before they get a piece they can sell; not many are willing to tolerate that.

    For now we'll have to wait.

  17. edi
    edi New Member
    I got the same idea, "The cube concept" and tried.
    But it didn't work. :confused
  18. underitall
    underitall New Member
    I've not had problems with Print orientation, as have only used WSF and it's always seemed fine. (My models are very small scale, 2mm/ft, so expect to see some stepping)
    One idea is, for example in FUD the maximum model size is:
    12 x 18 x 15cm.
    It doesn't state what axis is which though, but just say the Z axis was 15cm, then if you were to make a model with one side longer than 15cm, but shorter than 18cm, then it would be have to be orientated flat on the X/Y axis...
    Just a thought,
  19. bossmoss
    bossmoss New Member
    Sorry to introduce a newbie question here, but how do you know the best orientation for a given model? I ordered two copies of the same model in Silver and while one was perfect, the other was too crude for use. Reading this thread, I now realize this was probably due to print orientation.

    Is there any way of knowing which direction my good piece was aligned, versus the bad one?

    I know we still can't choose a print direction yet, but when (and if) this becomes available I would like to know what direction works best.
  20. GWMT
    GWMT New Member
    Last edited: Dec 11, 2011