But I've already PRINTED IT BEFORE!

Discussion in 'Bug Reporting' started by MaxSMoke777, Aug 13, 2010.

  1. MaxSMoke777
    MaxSMoke777 New Member
    A quick show of hands:

    How many people have had models they've already successfully printed in the past, suddenly being rejected and marked as "Unprintable" by some random KNOB working at Shapeways, because that twit decided some tiny part of it was unfit for printing?

    I don't think I'm out of line being just a little bit upset to see an order I've been waiting for, suddenly rejected, because some idiot decided that some tiny, well supported part of the model was suddenly unprintable, EVEN IF that same model has been printed before, several times, without incidence!

    Not only am I personally offended by having my model condemned without recourse, but I am completely put out to find my work delayed. It's bad enough I have to wait a week to a month, it's a hundred times worse to find out, after waiting, that it's NEVER coming.



    Nevermind. Turns out the problem was that when I moved up the holes that hold the magnets, I inadvertently moved up the bottom of the tile, reducing it's depth from 3mm to 2.5mm. So it was 1/2 a millimeter thinner. Unlikely this would make it unprintable, but it was thinner.

    I wouldn't normally have jumped to the conclusion that it was Shapeways Staff that was mistaken, but I did have issues before with other staff members marking items I already received as unprintable after the fact. That was not the case this time. In this case, it's entirely my fault. My apologies for the error.
    Last edited: Aug 20, 2010
  2. dadrummond
    dadrummond Member
    MaxSMoke777 -- I understand your frustration. I suspect that the Shapeways folks do as well. Your insults and attacks are most unwelcome, regardless of your frustration. What you dislike is the outcome, so please remain constructive.

    Perhaps you could show us what the model looks like, so that we could sympathize?

    This is an experimental time for Shapeways. Imagine someone submits a model that prints correctly 50% of the time, but the other 50% fails and disrupts the quality of nearby prints in the bed, causing costly reprints. Or suppose it's 5% of the time. At some point, SW has to refuse to print it, even if (say) 95% of the time there's no problem, simply so that they can profitably provide the service many of us love and some of us rely upon.

    Hope you take this in the best way. You're a strong contributor here and a talented modeler, and your message sounds like it was sent in anger. Hope we can provide some backstop for you.
  3. MaxSMoke777
    MaxSMoke777 New Member
    {blah blah blah Retracted]

    This is the old version I've coated with Wooden Floor Sealant (the toughest clear coating material I know of). The new version has the exact same dimensions, but the magnet holes on the sides are up a bit more to give them more strength. Also the brown areas are now white.


    This is what the full set should look like.

    Last edited: Aug 20, 2010
  4. stannum
    stannum Well-Known Member
    Hey, calm down a bit, you aren't alone. The "rules" about what will be printed seem random, there're some docs but even following them gives problems. You think it was right... BAM, now it isn't. If you had kept the same mesh and only changed the texturing from one version to another... you'd have a heart attack. ^_^

    Shapeways needs to build a serious knowledge base of what prints and what not, and of what was printed, what was risky and will probably be rejected next and mark it while shipping instead of waiting for the next order to take place, so everyone knows (new or old designers and staff). Even show photos of the misprints or ship the trashed parts too, so designers see what's really going on. Otherwise people will keep on having problems and complaining that prototyping with a "black box" isn't viable.

    [Ed: typo]
    Last edited: Aug 15, 2010
  5. bartv
    bartv New Member
    Hi MaxSMoke777,

    I don't think you're doing our people justice by calling them things like that.

    Having said that, I fully understand your frustration. The situation is like dadrummond described: we're always learning about printing tolerances and success rates, and we have to make sure we can print close to 100% of our models successfully. If only one model in a tray fails, it will destroy all the other models in it. A reprint will take over 24 hours and is quite costly, so we're trying to operate on the safe side (if we didn't we'd have to increase our prices..).

    From time to time we adjust our internal checks so yes, sometimes a model that was printed successfully in the past may be rejected now - as annoying and puzzling that may be. It has nothing to do with censorship or with us trying to squeeze more money out of you. I'll make sure that the team becomes aware of this problem, and ask them to find an acceptable solution for it.

    I'll also point out your personal situation to our customer service team, and I'm sure they'll be able to help you out. Feel free to contact me personally on bart@shapeways.com if you need to discuss this further.


    PS: we never had to create or enforce any 'forum guidelines' before, but from now insulting messages will not be tolerated anymore - both to other members or to the Shapeways team.
  6. Luukw
    Luukw New Member
    Hi MaxSMoke777

    I am very sorry to hear that you have some problems getting your models printed. Shapeways supposed to be fun. And I can image that when your models are canceled while they have been printed before is very frustrating.

    Shapeways tries to print as much as possible and if models have been printed before our people do not check it again. If you upload the same model again the model number changes, then we handle it as a new model. The same applies for new versions. Because it is all manual labor, sometimes a problem in a model does not get detected the first time, but the second time. That is why sometimes models which were printed before get canceled the second time.

    We are working on something to get a database with models which have been printed before. This database also flags models which have problems but we got them successfully and models we can not print.

    In the meanwhile when you have problems like this, the best thing you can do is send a mail to service with you order ID so we can provide you with a suitable solutions to the problem. There are several things which might be the cause of this, so with your order ID we can try to fix it.
  7. virtox
    virtox Active Member Moderator
    To answer your question :

    None :) I find that if you stick to the design rules, the chances of this are minimal..

    That said, I have noticed that the design rules for color sandstone are very strict. All models where I tried to add parts <3mm were rejected in sandstone.
    Not surprising, as it's apparently very brittle prior to the coating they add.



    PS. Cool project though ! I've been putting of building my own Catan-3d board, because of the costs involved ;)
    Last edited: Aug 23, 2010
  8. photosfromrob
    photosfromrob New Member
    I have heard this before but can't understand how this could happen. Since there is no movement and the models are build, layer by layer, everything is embedded in powder until the cleaning....how can one model affect others?
  9. KevinN
    KevinN New Member
    As the models are stacked in the tray in such a way that we optimize the tray space the models are all aligned at varied heights.
    When printing fails due to a thin wall the printing process crashes so many models are not completely printed and all other models which where planned to printed in the tray are postponed to a following print run. All semi-completed models are scrapped.
    On some occasions there may be some models on lower positions within the tray which can be salvaged.
    **This is only applicable for SLS printing. For detail, metal and glass we only print one layer of models at a time and therefore all models in the tray would have to be scrapped in the event of the print run crashing**
  10. MaxSMoke777
    MaxSMoke777 New Member
    [I'm not retracting this, because I think I still have a very valid point]


    I think the problem here is that you're thinking of the minimum thickness of a Rabbit Ear. Shapeways has a demo model of a Rabbit, and it's long, poorly supported ears require a certain thickness in order to avoid snapping off during printing.

    This isn't a Rabbit. This is a Tank-like tile. It's well supported on ALL SIDES. In order to make any form of detail, at some point all models are going to have very thin parts. This is the nature of a model, after all, we aren't making Sphere's here. But to have thin parts, you need to have good support, or the model will fall apart. And I've CLEARLY shown that my tiles are very well supported.

    2mm thickness is plenty for a well supported model with other areas 5mm or greater in thickness. The thickest part of my model is 6mm, adding to it's strength.

    If I was printing a ball of porcupine quills, I can see the problem, little bits snapping off and forming a mess in the tray. But this isn't a set of needles, this is a Hexagonal Tile. By it's very nature, it's strong and resilient.
    Last edited: Aug 20, 2010
  11. photosfromrob
    photosfromrob New Member
    thank you. NOW, I understand!
  12. photosfromrob
    photosfromrob New Member

    I don't know anything about your software, but if it were me, there are two things I would explore:
    1). Run a "virtual" batch for each setup. because there is on mechanics involved, it will only take a few minutes. If it's going to "crash", nothing will be destroyed and you'll know what not to print. This will give you more flexibility.

    2). rewrite the software so that it throws an exception and skips the offending piece rather than crashing.

    anything that gives you more control and options makes things better.
  13. TomZ
    TomZ New Member

    It doesn't work that way. It isn't a software problem. It's a mechanical problem.
    Thin features may break off and may be picked up by the printhead. The printhead will drag them all over the build area, disturbing the other parts. After a little while, the printer will notice things have gone horribly wrong and stop printing.
    Basically, at the point the printer notices it's gone wrong the printrun is already unsalvageable.
    There's no way to accurately predict whether this will happen but Shapeways has to draw the line for thin features somewhere.
  14. fx2
    fx2 New Member
    Question to KevinN:
    I always thought PolyJet printing was made without contact (ink jet type of printing). I know some of these processes have a sharp blade to egalize the layer height, though. Is it because of the blade that the whole tray is for the bin ?
    Last edited: Aug 17, 2010
  15. KevinN
    KevinN New Member
    Sorry for not having replied to this yesterday but I had asked our supply chain manager to gather some more info at our production facility first.

    For all materials other than detail (Acrylic-based photopolymer) the thin walls issues during printing are as described previously, printer can crash (mechanically) due to the thin parts being broken of by blade/wiper.

    For detail material the printer does not have a blade/wiper put injects the material from above (layer by layer - drop by drop). Here the only mechanical failure that can occur is that a model is so tall to support itself and collapses onto adjacent models. This is relatively rare.
    The reason why we do not attempt to print models with thin walls is that the removal of the support material is quite difficult and thin models are prone to breaking in post-printing processing.
  16. fx2
    fx2 New Member
    Thank you for the detailed answer ! :)
  17. bartv
    bartv New Member
    For your information: MaxSMoke777 has edited his post - he *did* make the object thinner after all.