Blank Rejection?

I got an e-mail saying there is a problem with my model (nothing new there). This time though, it didn't give me any really issues! Has this ever happened to anyone before?
Maybe I'm just mistaken on what the real error is...

The automated rejections can be a bit understating in the issue. If you can share the file, we can take a look to see what the issue could be. If you don't want to make it public, my email is below.

I have to change a small part in it that someone found after I tried the blank fail, but then I'll re-upload it and see what it responds. If it's blank again I'll probably send it to you.
Is there any idea why it sends a blank fail?

The only automated emails you will get besides that one I think are if your item is too big/small physically, too big in file size, unsupported file type, or non-manifold.

The manual checking that goes on after you order gets a full response as a human looks at your file. We're training our computers to get better at it though.

I made the corrections I mentioned, and it came out with the same error.
So I'll send you the file.

I most likely won't be able to look at it till tomorrow.

That's just fine. Thanks for your time looking at it.

Check out to see if this helps you. I just downloaded your file, I'll update if when I see something.

Did you happen to find anything? I keep trying slightly different versions, they all keep coming back with a blank error. I don't think my pieces are too small, at least not by my calculations, but I can't be sure.

I've tried NetFabb a little, and Netfabb cloud services with no luck so far. Seeing if I can get someone that know's better than I about using Netfabb can help.

Yeah, I started messing with that, and got it to repair 100% and save the stl. But it still sent me a blank rejection, so now I'm at a loss. Unless it could just be it's got a piece that's too small to make, I I wish it would just list that as the problem though.

I got Stony to take a look and here's his response.

Shapeways has a limit on uploaded models of 1 million triangles. Yours is at 1.5m and you need to try to reduce that some.



I believe that the upload process still has an issue that if you upload a model with more than 1m triangles, the background processing software aborts without producing a usable message.. hence the blank email.



Personally, I'd remove a LOT of the holes around the edge and such, but another option would be that you seem to be using 20 facets per circle, if you dropped that to 8 or 6 facets per circle, it is likely that you'd not be able to see the difference (these are VERY small holes), and it would drop a significant number of triangles out of the file.



A few notes ...
1) your "C" beam supports under the model are too thin. They measure out at 0.19mm thick and that will be rejected upon printing.
2) the "X" cross-supports are not 'connected', they are floating independently. that makes for 56 very small parts that the operator will have to pick out of the printer with tweezers, and a number of them are likely to be lost. Did you intend for those to be separate parts? If you do weld them to the main body, I'd consider removing the holes in the end of each beam.. again, those holes are contributing to the overall triangle count. Otherwise, you need to try to do something to collect them together in one area and sprue them together such that they don't get lost.
3) in the center of your model, you've got a bit of bad geometry that will fill in as solid and cost you more. The Boolean join you did for the hole under that center "knob" failed.. it needs to be redone.
4) the main 'skin' of the model (the part like the fabric on a umbrella) is also too thin. You've got it at 0.1mm and it must be at least 0.3mm thick.



Also.. printing this is going to be rather expensive:
http://stonysmith.com/wired/VolumeEstimator.asp?L=481&W= 481&H=30&T=.3

Thanks for that extensive list! I'll see if I can fix all of it in different ways and probably try again. I'm surprised it didn't at least give me the too thin error in the e-mail.

We figured it would be about that cost though, there is just not another way to make a piece that small and that detailed.

Thank you for your time and looking into this so much. It's going to be extremely helpful for trying to fix all of the problems... hopefully that's actually all of them.

Thin walls are not something that is automatically detected yet.

Ah, that would make sense then... Well thank you again for the help.

Is there a way to check the number of parts in a print to see if I properly attached the 56 small parts you mentioned?

Download Netfabb Studio Basic from www.netfabb.com
Open your STL file with it and then click "Repair" then "Update"
It lists the number of shells, as well as the number of holes.

Any holes in the model will be hilighted by yellow dashed lines.
Any inverted triangles will be hilighted as red.

I decided to make the skin of the old model removable, and it actually can be printed now! So success there, but I'm struggling with finding the error that is causing the structure to print. I went through and measured the wall thicknesses of everything I can find. Made sure I'm under 1m triangles and removed a large number of the holes. I also can't tell if the cross bracing parts between the C beams are actually connected. I believe they are, but don't know netfabb well enough to be certain. If anyone could take a look at the new model I'd be thankful.