Reject It Anyway

How many times I had to fix the same rejected model for different issues?

Two days ago this model

https://www.shapeways.com/model/1781724/1-50-sep-parts-g-wag en-for-bp-42.html?li=shop-results&materialId=4

is rejected with that explanation:

"We cannot print your model due to: Thin Wires. Here are some details to help you resolve this issue:

We attempted to print this, and the wire pictured in the image broke on two separate parts of this product.

Our nylon materials are printed with a process known as "selective laser sintering." During this process, a layer of powder is pushed out by the printer, and only the areas specified as solid per the 3d file are lasered. The product, therefore, is composed of hundreds of microscopic layers joined together.

The areas of powder not melted by a laser are left as loose powder. At the end of printing, we have a vertically built print job with solid models encased by loose powder. We remove the solid models from the loose powder by hand, then we use sandblasting and compressed air to achieve a fully cleaned surface.

When the product is encased in loose powder, it is similar to as if it were encased in a snow bank. The powder, especially that closest to the model, can be tough or difficult to remove. To avoid a breakage in this stage, it is important that the structure of the part be sturdy.

In this particular design, the structural integrity broke at the above mentioned stage of handling. The wires were too thin to withstand the pressure of the loose powder around them and the attempts at freeing it from the powder cake. The design may be more successful if the broken areas are thickened in diameter."


I thought, "well, no more, five minutes of easy work, and a kind appology to my customer (good customer, fortunately).

But today, they have found another new "issue":

"We cannot print your model due to: Minimal Clearance. Here are some details to help you resolve this issue:

Parts of your model are located too close to each other, printing it will result in them being fused together. Make sure the clearance is > 0.5mm."


This "big issues" detected by the "terminators" were printed without problem in these another models (these parts were a clon copy of the late model!).

https://www.shapeways.com/model/1622781/1-50-sep-parts-a-wag en-for-bp-42.html?li=shop-results&materialId=6

https://www.shapeways.com/model/1509217/1-50-sep-parts-k-wag en-for-bp-42.html?li=shop-results&materialId=6

How can we trust on PIA if an usual order checking isn't made correctly and have to be corrected again, and again and again?

What is worrying about your rejection is the little note about clearance.

Apart from the weird notation (a point or comma is the usual notation for a decimal, not a single quote) the model checker seems to think that the clearance for wsf is 0.25, not 0.5, even though they note 0.5 on their annotation. To give a correct clearance for a cylinder of 2.0mm the surrounding hole should have a diameter of 3.0mm - or am I wrong?

If the checkers don't know the rules...

Annotation aside, the checkers math adds up. The larger diameter minus the smaller =.6mm. 6mm divided by 2 is .3mm of clearance on each side of the cylinder which is less than the required .5mm.

The clearance was not caught in the manual checks previously, completely on us. Can models not be fused less than .5mm apart? yes. But if it fuses, and we send it out fused, we sending a product that doesn't function as intended. As we get better at checking, a down side is we'll catch more and more of these things that were missed previously. So it will take time to see the benefits of better screening. Which will mean more reliable production, faster turn around on products that you can be sure won't have issues in production when your customer orders them.

This means that checking method evolved from monday march 31 to wednesday april 2? In two days? I'm a lucky man!

And... why do you think this under 0,5 mm clearance isn't by design? It takes months and many test to find the correct clearance FOR MY DESIGNS to not have loose parts or melted parts. All this test are going to trash in one second.

I'm sure next time the rejection message will be:

"you are making an ugly model. Please go to art school and upload a nice model for once"

There's a second reading behind this consecutive rejection?

Part of the Shapeways answer:

"Subject: Re: Help us resolve issues with your customer's order
APR 02, 2014 | 05:42PM CEST
Hi Luis,
We highly recommend that you place an order for the piece
before making it available for sale, in order to avoid these situations for your customer.
By not printing beforehand, it essentially puts the customer in the position of testing
your model. ..."

Sorry, I'm not millionaire to test all the models people ask me.

I have today 190 models for sale in shapeways. They are recommending I have to retire 150 models from my catalog because I don't print it before put it to sale.

Think about shapeways catalog with a 75 % less of models because designers are too avaricious to print models. Without variety, It will be a crash.

0.2mm clearance for WSF is totally doable
Dragonscale earrings - the majority of the rings on these earring are seperated by 0.2mm. There is a lot of powder in the closest regions upon delivery and the rings do not move fluidly, however none are not fused. This is by design - WSF is printed with the powder temperature at near melting temperature under a vacuum. The powder that is hit with the laser to fuse the grains retains heat, it takes a while to get back down to the chamber temperature. The thicker the part is that gets fused by the laser the longer it takes for the part to cool = more thickness, less clearance is more likely to fuse.

If Luis says this has worked ok before, why not let it ride?

Luis, Shapeways don't check a model for all aspects of a material's rules, they find the first and order cancelled let you know - just one of those things to get used to. If you haven't got it already, grab NetFabb Studio Basic (free), asides from model integrety & error checking, the measurement tools are worth their weight in (free) gold.

Paul
[hr]

stop4stuff Modeller for hireShapeways Shop - Controller PendantTwitterYouTube​

[hr]

p.s. unless the model is a commission, it is good practice to buy in the model, not only to make sure it works, but also to get some photographs of the printed itm.

Paul

Thanks, Paul.

Well, I don't really love netfabb, but that's a personal choice (I think it's a stone age axe compared with my autodesk inventor-solid work working tools).

My real upset is about the infallibility of shappy "terminators" checking team.

In a year of designing -with its funny rejections included- I learn to take air, deep breath and fix the "awful issues" without send a complaint.

But, why the "terminator" don't check the entire model to avoid consecutive rejections? What's the game when you have two rejections for the same model in two days for two different "issues"?

And we have also the printing test model. I have also the bitter experiences (sadly, more than two) that a printed before model, then now (oh, surprise!) can't be printed "because your grandmother smoke too much" when a customer order it.

Then, who assures me that a before printed model will be printed back in a customer order?

Do i have ordered it again with the "crucial changes"? and again, and again, and again,......

Regards

Luis

That's just what I said!

I've received JUST NOW the following rejection message:

"While we've been able to print this model before, as we continue to produce the model we've noticed that it does not print successfully due to: Weak Geometry

Despite our best efforts we've only been able to successfully print this model 50% of the time. To continue printing this model we need your help to resolve these issues:

Plastic material might be better for this part. If not the part need to be thicker in the circle area. at least 20% more"

Best efforts? My b.......................... It's a color model for ....... sake!

Model:

https://www.shapeways.com/model/1746585/2xbuckle-v5-burgundy .html?li=aeTabs

Seems they have printed this last model several times and half of the times it would fail at some point so they had to do a reprint.
Not that strange for them to suggest you make it a bit stronger so it won't break as easily.

About the clearance in your earlier model. As you say you've tested the clearances extensively and this 0.3mm has always worked fine, I think shapeways should flag it to be printed even though it doesn't meet their guidelines.
Perhaps try to contact some of the active shapeways crew on the forum to see if they can do something.

But why did we think that the problem is from the "fat fingers" designer?

If my design fullfil all the rules and specifications AND A "Terminator" CHECK IT AND GAVE ITS OK, the problem could also be wrong working of a machine, cleaning method, crew, etc...

But it's easiest to say: "could you change your design to a brick 80x80x80 mm., but without sharp edges and completely solid?"

I think a lot of these rejections and bad feelings could be solved through better education by Shapeways. The problem is that we as designers don't see much of the work that goes into making these models, and thus don't appreciate the rejection messages. The guidelines are a great start but are unable to cover every circumstance. If we could see a video of HOW a model is breaking when cleaning, or examples of other common rejection problems we can empathize with the hard-working factory team and make it easier for them.

I own a desktop printer so I understand how many ways a print can fail. Probably 50% of all my prints fail for some reason or another. Sure, Shapeways uses industrial machines that work much better than my little $700 toy, but you'd be amazed at the creative ways the machine conspires to ruin a perfectly good model.

So Shapeways, can we continue the "Ask an Engineer" series with more examples of common rejections? Better info for designers will save a lot of time and frustrations for both parties.

Of course you're right, if the people who operate the machine were more careful, it may have not broken half the times.
If the operators are that much more careful all the time, they could change the rules such that you can make thinner structures.
But then there would be a new model which is just at the limit of what they can produce, and breaks half the time. Then when the operators were even more careful, it may not break as much, etcetera.

From a pratical point of view, it doesn't really matter who's fault it is that it repeatedly breaks during production. SW has a certain level of carefulness when handling the models, and they design guidelines to match that. If you follow all the guidelines, SW should normally be able to print it. But the guidelines are quite generic, so there will always be cases where even though you adhere to the guidelines, the part may not be printable.

Likewise, there will also be cases where the model will print fine, even though you did not follow the guidelines (like it probably should for the other model).

But I can imagine how frustrating it must be when you do everything right and they still can't print it properly.

And herein lies the problem IMHO, as that "level of carefulness" seems to have shifted towards requiring sturdier (and clunkier)
models since shapeways started to outgrow its Dutch roots.

^^

True dat!

I had the same feeling about my OO scale (original) phone boxes in FUD with 0.3x0.5mm window bars - seemed to me like, they got popular, and the rules changed to favour more sturdy items... I'd love to own a ProJet 3500 HD, just to see what really can be done with careful handling.

Paul

Hi all,

I am the checker that rejected this model after the 2nd attempt (I rejected because of the clearance). I would like to express my concern to avoid a bad rejection experience; our clearance guideline (I would like to remark "guideline" but not rule) states that we can guarantee that 2 parts won't be melted as long as the distance among them is greater than 0.5. A direct consequence, then, is that the minimal difference between a rod and its allocation should be at least 1.0.

Regarding the notation I am aware that some parts of the world don't use this notation, but I should accept that I am not that sensitive in these aspects. We learned this notation to avoid confusions with vector & matrix notations.

Best regards to all,

Albert

Then, thanks for reject it for a "guideline".

Perhaps somebody in shapeways could add a list of "guidelines" to material portfolio, and all we could go working in gardening.

What does this means?

Do the tools of Shapeways check team are wrong or they try to fool me?

Shapeways picture have sent me:



My own measures (netfab and inventor):





Why do they have to lie?

Talking with CS, having them look into the discrepancy. Regardless, the production facility doesn't feel they can produce that part without that area being thicker. Even at 3mm, that area is supporting too much weight to survive handling and will probably break before being dipped in the sealer.

They printed it before.

One in this model

https://www.shapeways.com/model/1662745/buckle-v1.html?li=sh op-results&materialId=76

and this:

https://www.shapeways.com/model/1746585/2xbuckle-v5-burgundy .html?li=shop-results&materialId=26

Total: 3 printed buckles sent to customer.