Inconsistent acceptance/rejection

I'm totally confused about the criteria for acceptance and/or rejection of dimensions of features that seem to be treated as "walls".

I've had a T gauge tunnel portal accepted and printed successfully 18 times in white detail plastic. One of the details on this model is copings on top of brick columns. I've also had connecting walls accepted and printed in the same material. The walls have the very same coping details of exactly the same dimensions. All have printed exactly as expected them...with a good level of detail.

Now I've had another item..wing walls..rejected because the coping detail of EXACTLY the same dimensions have been deemed to be walls and consequently at 0.45mm thick are too thin! I've emailed support querying why there is this inconsistency and all I got back was "Unfortunately we cannot print models which have been rejected." and very little else! No explanation why one model had a detail accepted and another model with the same detail was rejected. Nothing!

Anyone else had a similar experience??

I'm now left with half-built items that have cost me a great deal of money that I may now have to throw away and redesign!!

Did they send you an image of the rejected area, that you can share with us? We're working diligently to get rejections standardized. Some things don't exactly follow the rules, but still make it through review and printing. Some things, even being printed before, then get rejected later. I personally know how frustrating that can be. With some insight into your design, hopefully we can help you get around the rejection.

Yes they did...image attached.

Just to throw another pebble in the pond....I did actually have one of the rejected models previously accepted and printed.

And for reference, I've also attached a photo of the tunnel portal itself that has the same detail...which was accepted.

Is the part that sticks out .48 mm at least .48 mm thick? If not, 1:1 is requested for details. It's hard to tell, but it looks close.

I've just looked at one of the models and yes, it is close....and might be very, very slighter deeper than the thickness on one side....but well under on the other three!

Is this the rule for such a detail?? If it is, I can slim the outside dimensions of the coping down so the coping sticks out around half its thickness.

For strength purposes, details should be at least as thick as they are tall. So if you make them stick out less, or increase the thickness, that should fix your rejection issues. One of the problems will manual checks is one person may catch such details, when previously it wasn't caught. They don't want to be sending out items that will likely break in shipping. So we're working on better automated software that will check your model the same way every time.

Many thanks for this information....

I read the tutorial on thin walls through three or four times and looked at the diagrams numerous times and was somewhat confused. The diagrams are hopelessly unclear. I had to conclude that the solution was to make sure the width of a projection should be greater than its height...and you've confirmed what I thought was the case.

I've already started sorting out the detail dimensions, so I'll report back on this thread how I get on.

And some advice for the guys at Shapeways.....when reporting back on a rejected item, why not specify what test the item has failed - and a suggestion or hint on how to cure it!