Print It Anyway: Want to Experiment More with 3D Printing? Tell us to PIA!

Pete,

Why is it in the 3 years or so I've been asking both publically and privately that practically no reasonable feedback about print difficulties has been conveyed as a matter of course... the only times I've ever got any kind of 'reasonable and intelligent' feedback about my model rejections is after a drawn out discussion with Shapeways CS team, and even then the reasoning has been on occasions pretty non-sensical.

If a model is problematic, convey that in a clear and consise manner at the fist available oportunity.

Just a decent amount of feedback about model print problems might be good enough... e.g. I'm not talking about 'probability of breakage during cleaning', I'm talking about issues faced during cleaning.

Sorry PIA, imo is another wasted so many months until the next 'big thing' comes along. I won't be using the PIA service unless there is a guarantee of the best possible treatment and handling of models and feedback of model printability issues (including every single step along the way).

Paul
[hr]

stop4stuff Modeller for hireShapeways Shop - Controller PendantTwitterYouTube​

[hr]

So what you are saying Pete is, given that it would take an intelligent person with expertise in the printing process to first determine if it was a design error and then explain what was wrong to a person with a different skill set( the designer), and that such people are neither in plentiful supply or cheap, it's cheaper for you, on a cost benefit basis, to just absorb the extra printing cost for now?

So it's not just Shapeways leaving us designers out of the loop again, thanks for taking to time to explain it to us , I can appreciate the reason and logic now.

But at some time in the future, you are hoping to semi-automate this, bring in a check list so a semi-skilled tick off the boxes or something on those lines? I can also appreciate that setting something like this up is not a trivial task, especially in a factory where every item you make is unique.

Then we will have full feed back . I shall just bide my time waiting for this then. Keep up the good work and try to continue breaking fewer things than you fix!

Tom

I don't mean to say something that might make the universe crack into pieces, but...

Can you let customers print it anyway?

OMG!! RUN FOR YOUR LIVESSSS!!! HAHAHA!

Tom - I think you hit the nail on the head. We have a very clear vision of what type of information we need to provide you:
1) Automated information for as much as we can determine upfront, as quickly as possible. (You're starting to see this in the form of WT Viz / Fix and other checks like Bounding Box, etc)
2) A CONSISTENT and REPEATABLE manual review of each part that follows a given process each time, regardless of the person involved. (We're working on this but still have alot of room for improvement given some of the examples I've seen popping up lately)
3) A very clear, thorough, and actionable set of reasons and analysis for why your model is not able to be produced. In most cases you should be able to learn from this and upload a new part that absolutely passes our process. In other words - the right information at the right time to help you make the right decision.

The 'end' vision is crystal clear, however the best path for us to get there is still a work in progress. Cutting over 100% of our manufacturers and factories to a new process may actually not give us the results we want, because we still need to learn (as do you) exactly what's printable. In many cases, we honestly don't know until we try it. This is why Print it Anyway is important - it allows us to work together to push the limits of the manufacturing process and improve our process, machines, etc.

It should be noted that there in a few cases, we may find that there's no way to produce the type of part you hope to make given the design constraints. An example of this would be a Stainless Steel hairpin that needs to fit snugly into a 0.2mm space. Given the limitations of the material and process, we will most likely never be able to produce a part that fits your needs so we can either recommend a different process, or give you the information to determine that the part is not manufacturable.

Hey guys,

Just a quick update, with today's release, we are taking the lessons we learn from your PIA prints to improve the regular process.

As we run the Print It Anyway pilot, we want to make sure that this positive experience for our designers and what we learn from your PIA orders carries over to our regular Shapeways experience. To provide better context and information to our 3D Printing Engineers while they are checking models, we are showing them if an order was a previously PIA and it's success rate.

This is crucial part of having this PIA pilot, we ultimately want to improve the regular ordering process!

The problem with rejections is often nothing to do with the design but is to do with the interpretation of the design guidelines. Thin walls is often used as an excuse to reject a model. But on detailed inspection of the areas where this has occurred I have found that the the thin wall is actually surface detail. If it protrudes less than 1mm from the side wall of an N Gauge wagon, and it isn't more than 3 or 4mm long, then it isn't a thin wall. It is detail. And it doesn't matter if it isn't perfect when it is printed either, because the customer will probably sand it and prime it and then paint it. Automated checks are even worse.

I have also had parts rejected in one order that are passed as okay in another order, even if both orders are going through at the same time.

Consistancy is absent, and I have argued with CS people on both sides of the Atlantic in order to get the prints I want doing actually done. It shouldn't have to come to this.

I have also had parts broken during the packaging process. We know this because the bags still have the indentations in them and the broken off parts are in the bag. And we have had several longer parts that have been printed bent, resulting in many expensive reprints you have had to do until you eventually got them right.

So PIA is no good unless you can guarantee there won't be problems with your production or packing department first. No credit or reprint, no PIA. But the main thing is to address the consistancy in your application of the design guidelines. These goal posts change far too often.

The attached images show a comparison of a FUD wagon verses one printed elsewhere. It is a poor comparison and we have had much better results. The FUD version is very sharp, and yet although this has obviously been printed the model currently sits in My Models with a thin walls flag raised for that beautiful fine door detail and hinge mechanism. The second picture shows the same model in final painted form and mounted on a chassis. This isn't a Shapeways FUD print, but one done in SLS. Poorer quality maybe, but you would hardly notice.

For consistancy of prints I am having to look elsewhere. Which is a shame as the FUD results are beautiful - if you print them.

Hello everyone,

I just received my first part prototyped. I wanted to share with you my great satisfaction.

I had doubts about the feasibility of the details of my model. Some details not exceeding 0.2mm deep and 0.3mm wide, but I still try. I was impressed by the result when I opened my package. oO

The part is also very solid, although it is not more than 5mm thick.

I know the SLS process, but it was the first time I pushed the design of my pieces so far.

Thank you very much, you will very quickly to prototype new models.

Pierre

P.S: I'm attaching some pictures.

What about the polygon count? Are you ever going to offer a service with a higher polygon count?

Forget polygons - the Shapeways' upload limit is counted in triangles, not some random non-defined n-gon (unamed polygon))
So, yep. Get the terminology right Shapeways (please) and there should be less posts like these, and maybe more people would have more faith in what can work


Paul

[hr]

stop4stuff Modeller for hireShapeways Shop - Controller PendantTwitterYouTube​

[hr]

All triangles are polygons. Not all polygons are triangles. Geometry or logic failure somewhere.

That is correct. Typed in hurry I'm afraid, but when you think of it what is a sqaure but two triangles sharing two points. A square has a hypotenus and that divides the square into two or more triangles.

My question still remains unanswered. Are there plans to increase the poly count?

They are 4 things ...

Not necessary to separate this out. You answered my question. I like your current move to print it as it is. I may have to use it quite a bit.

Add 5 ... more computing needed on the servers to handle that polycount , my laptop can`t handle more than 1 million.

I was able to print a prototype using print it anyway! it had a few small edges on it that didn't meet the design requirements for shape ways but it printed perfectly fine! Im very grateful they added this feature.

So far it seems you accept triangles and quads with formats that support both (you are right), but count the triangles (1 quad counts as 2 triangles) needed to recreate the mesh (others are right).

Moving to other thread... just fix the docs or the code instead. Docs are probably the simplest approach.