New design rules for WSF Polished & FUD/FD -- Feedback Request

Hey all, over the past few weeks we've been working with our production partners to better define design rules for the Frosties and the Polished WSF (White, Strong & Flexible). Many products have needed re-printing, which has a bunch of negative downstream impact (unhappy customers, bad lead time, more cost for everybody, and yes--price increases). So we are attacking this from various perspectives. One perspective is to add some design rules that will lower the number of broken pieces if followed.

But, before we make these official, I'd like to throw it out to the community and see what you guys think. I'd also like to get your opinions based on the prints you've received so far. Do these numbers seem reasonable, do you have better ways of presenting these findings or tips of your own to help others create stronger structures?

Please see attached PDF for the design rule draft, and let the discussion begin!

Thanks,
Nancy

In reference to the wall thickness for WSFP, would it be possible to opt out of polishing? Some people may want to hand-polish prototypes if given the option of making the wall thicker or hand-polishing.

Also, this looks like a typo:

"3: Frosted Detail/Ultra Detail Issues to Note

...

Suggestions? We recommend to try to design things with a clear base. You can also look at Frosted Detail, which does not have this issue."

I believe this describes Frosted Detail as having this issue, but then also suggests Frosted Detail as the alternative.

Thanks for the notes, I'll make sure the text gets change to something clearer.

Re: In reference to the wall thickness for WSFP, would it be possible to opt out of polishing? Some people may want to hand-polish prototypes if given the option of making the wall thicker or hand-polishing.

Opt out of polishing--why not just order the parts in White Strong, & Flexible? That is unpolished. Am I understanding you correctly?

I'd like to know what is considered a wire? does it only apply to round structures and at what point does something protruding become a wire?

Further on you are talking about the downfacing side - but we don't have any control over which orientation you print them in so how are we to know which side is going to be downfacing?

The suggestion is "We recommend to try to design things with a clear base." What exactly do you mean by a clear base ? - a flat base?

Tom

Thanks

Re: Opt out of polishing--why not just order the parts in White Strong, & Flexible? That is unpolished. Am I understanding you correctly?

I only have parts ordered in FUD so I didn't realize that there was a difference between WSF and WSFP Thanks

Nancy, please could you find out if the FUD wax support material is printed at the same resolution as FD i.e. half FUD resolution... I'm thinking that the wax is printed at FD resolution and that the white rough surface is the first layer of plastic filling up the spaces in the wax, whilst also trapping the wax in that layer (my thinking could very well be wrong).

If FUD design rules are going to be changed, I hope there will be some kind of liason between the staff who check the models and designers to save carte-blanche rejections when a design has been previously printed without reported issues.

fwiw, I had a FD design printed that arrived broken, the broken off parts were missing compleely, to me this is a handling issue somewhere along the line as subsequent repeats of the same model arrived fine. At one point, I even had an email from Ralph (hi Ralph) notifying me that a model was broken, when in fact it turned out parts had just come off their sprues... perhaps models uploads could have a check box for sprued parts to save this issue.

Re: Wire, yes I need to add a section explaining what wires are. Here's some examples..

Wire "mesh" structure: https://www.shapeways.com/model/276378/heart_pendant___3d_and _stylized.html?gid=cg14

Wire structure: https://www.shapeways.com/model/209888/snap_bangle.html?gid=m g

Wire connectors (the things connecting the wings of the airplane): https://www.shapeways.com/model/271692/1_144_sopwith_triplane .html?gid=mg

And the obvious one.. wire poles: https://www.gearwest.com/images/W/exel-S5-jr-touring-xc-ski-p ole.jpg (ok, not a Shapeways image, but you can see how this is a wire if this was like 10 cm long model)

Clear base: yes I meant to say flat base. Or a slightly curved base.

I will find out the resolution question for you, stop4stuff.

As for the rejection, we will also work with customers to make their models stronger as well. Of course, we want to respect the models that have been printed before as well.

Thanks,
Nancy

Hi Nancy,

I would also like to know does this mean that we can control how you orient our models during printing, at least for FUD? If so, you should also provide us with a guide line.

Since I observed that the coordinate system of 3D software might be different. Something like this would be helpful: "If you see your model's front is facing forward in Shapeways' preview after uploading, then your model's bottom will be downfacing during printing."

CGD

I would like a bit more clarification about the wall size issue before I decide whether these rules would negate all the work I've done as nearly all my models have walls bounded on three sides at ~60mm length, by ~10mm high. I have not heard of any issues with warpage. Could these walls be strengthened with a few carefully chosen braces?
I checked my one model with a wire feature and I suppose I could live with the new rules even though it looks a bit chunky with the increased diameter, although I too would be interested to know when something stops being a detail and becomes a wire.
It would also be nice to be know or be able to set which axis becomes X, Y, and Z in the printer, because the resolution is dramatically different from X to Z in FUD.

The base suggestion sounds a bit "uh?". For example, you model a humanoid with base and arms pointing down, then the arms will still need support, wouldn't they? The machine page has a photo showing this "arms" problem, with a ring. The only fully safe shapes would be those that repeat or become smaller than the base footprint as they grow in height.

I've got a basic problem with this:
This means this one and this one aren't printable anymore with the new rules. This is very annoying as both were printed several times with no apparent problems. Both have "wires" with respectively 0.5 and 0.35mm thickness. The tractor has got "wires" on the box top and dumpster the two handles on the side.

I sincerely hope this material will not be cluttered with a lot of rules reducing it's usability. And I hope people will not be plagged with load of rejections two or three weeks after they ordered...

Maybe you should allow thinner wires, but at customers own risk (no reprint).

With wires in FUD why has the minimum thickness been set at 0.6mm? 0.5 Seems a reasonable compromise between robustness and visual fineness, but 0.6 starts to look a bit chunky.

From the pdf linked to in Nancy's post

You miss the point. 0.6mm appears to be an arbitrary measurement, What criteria was used in choosing 0.6? why is it not 0.55 or 0.65?

And what about links like these? http://shpws.me/1ZHj] They are from 0.5 mm wire, and have been tested with a 100gm load with out any sign of failure.

It looks as if we are heading for the same sort of debacle that we had with Detail where the minimum wall thickness was applied in such a way that pieces that had been made many, many times suddenly became unprintable.

As a designer I want to see my failures, because then I can design around the problem. For example. I have noticed that, in FUD, if you have a wire that is fixed to a heavier section, say where an axle meets a wheel or where the loops meet the sprue wire in the coupling fret above, the thinner wire will shear at the change of section. This can be overcome by blending the wire into support. This means that a thinner wire can be used where the end(s) are blended. How much thinner and how big a blend is needed is something that designers need to work out for themselves.

One final point, If I am making patterns for other processes, then I only need one good one. I would be quite happy to receive 5 or even 10 broken pieces to get the one good one I need.

Reading the whole of the (after trial period) revised rules about wires shows that 'wires' are "...(wire with no support or supported at one end)", something like a 1:76 scale walking cane would fall into that category... chain links do not.

I do not expect my K6 or K2 telephone boxes to be rejected, even though the window frame bars are just 0.5 x 0.3mm 'wires' as the 'wires' fall within the (after trial period) revised rules.

If models are suspect for failing and you're willing to carry the cost for developement purposes, I'm sure that Shapeways will be accomodating if you ask.

Hey all! I am traveling around the world this weekend so pardon if my responses are delayed. I have some time in Seoul International Airport (huzzah for airports with free wifi!) and will respond to some general concerns I gleamed from skimming this post.

Re: arbitrary rules, bad design rules, limitations, etc.
Please remember that this is a DRAFT of the design rules. The purpose of this discussion is to bring a starting point to the table, and have an open discussion w/ the community before laying down the law, so to speak. Understanding where the concerns are will help us make sure the rules are fair and helpful.

Re: prints that have been printed before
I understand that some of the prints you have already received before, but also some of these prints may have been the result of many reprints and in the end are more costly to the community as a whole (hence the price increases, etc). To be sure, we will try our best to grandfather in some of the prints that have been successful before, or at least work with you guys on thickening parts that need to be thicken, or shorten wires that need to be shortened.

Re: 0.6 minimum & arbitrary number
The most important issue is the structural integrity--answering the question "Will it break?" A lot of this is thickness, yes, but a lot of it is also ratio and the geometry of the way the models are designed. Rather than putting all of our creative designers through Engineering 101 Lectures, we start by creating some basic guidelines. It may be that, just as we have an under 3mm guideline for stainless steel, we also have an under 0.6mm guideline for wires in FUD. Again, this is all a DRAFT and this conversation is meant to improve upon the draft.

Thanks all for your discussion and please continue.

It would be very helpful to hear what you think the design rules should be based on how fragile the parts are that you've received. If you received parts that you can break very easily, or need to handle with great care, how would you design so that it is more structurally strong? If you've received parts that are under 0.6mm and know are structurally strong (like the Dumpster), how would you change or amend the design rules for including those types of designs?

Thanks!
Nancy

Well if you don't tell us about problems how are we to know and be able to adapt out designs so that they do print first time?

Maybe you need to start rating designs and giving a discount to those that are easy to print.

The other thing that would help us a lot is either show us pictures of what has gone wrong or indeed send us the faulty items. Some of us might even be prepared to pay a reduced fee for some of the faulty items as in part they might be usable and certainly having them will give us a much better idea of what has worked or not worked.

Tom

I agree. We need feedback from Shapeways about how well our designs print. I am quite prepared to let Shapeways work out the best way of doing this, but I will make a couple of suggestion that may help.

While you are running trials on new materials I think that you should return the pieces as they print and with any printing errors. We need to know what a new material will not do as much as what it will do.

I also think that it is essential that there is some way of marking pieces as tests that we don't expect to be printed perfectly. If a test piece is going to be useful it need to show were the line is between those component that work and those that don't.

How long will it be before the new rules kick in?