Gauging material loss in WSF Polished

I'm interested in finding out what others' experiences have been with material loss in resulting prints when using the Polished version of White Strong & Flexible.

I received an order back in June with two prints of a simple one-piece design in WSF Polished. As far as I can tell, they were printed in a true vertical orientation. The measurements in X & Y are undersized to various degrees, but corresponding X and Y dimensions are fairly consistent with each other per print, and the two prints were fairly consistent with each other. Z-dimensions are the most accurate, if (slightly) large.



The big kicker, though was the thickness of the fins. STL model fins were 0.05"; final print fins were in the range of 0.038" - 0.040", constituting a 20-24% loss of material relative to the digital file and yielding a part that was more flexible than anticipated. I sent photos of caliper measurements and discussed the issue with customer service, who looked into the problem and concluded:

As a designer and Shop Owner, I wanted to know more about the rationale behind the problem for future reference:

So at this point I'm wondering why the problem can't be pinned down more precisely, given that I'm surely not the first person to order Polished WSF. I was really hoping for more feedback from the production team instead of speculation from customer service (no offense to the guy who helped me, I believe he did his best but may not have been in a position to look further into the issue).

More information from the people who understand how the printers work would be more helpful, even if merely speculation or a simple "if you want accuracy, get the UNpolished material." We designers are working with trial-and-error to see what works; I need more information than "we give up, here's a refund" especially when the problem exceeds the stated tolerance for the material.

That was the last of the conversation. Anyone who can offer something more based on their experiences with WSF and WSF Polished, please feel free to share.

That's interesting. I've had WSF cylinders printed with uniform wall thickness (as designed). But the printed parts clearly had an offset in wall thickness which made them look oblong. I still have them so I could measure them but my feeling was that this had something to do with settling or movements of parts in the powder during fabrication, and it seems to be worse for thinner walls. I've never seen this effect so severely in alumide (aka metalized plastic or whatever they call it now). These kinds of nonuniformities are extremely noticeable when you put a part on a lathe for trimming or polishing. You can definitely feel any wobble with your finger while it is spinning or notice cutting on the bigger side when using a cutting tool.

I've also seen this happen with sandstone but that's more of a sagging or warped central axis issue during handling and hardening after the printing, which is something that is not completely unexpected due to the sandcastle warnings on the materials page.

Depending on what you mean by "offset in wall thickness which made them look oblong," you might be talking about a different issue altogether, such as print orientation, which can screw up symmetrical objects due to minor discrepancies in the calibrations of the printer axes (probably unavoidable and therefore a good case for allowing us to specifying orientation). I have an FUD prototype part with that very problem in a cylindrical component (layers indicate it was printed at a 45 deg. angle relative to the Z-axis and it has a definite curve).

Print orientation may be the same thing that caused your odd shape, and is likely an issue with other materials, as well. I plan to bring my concerns up in a separate thread at some point (because there are multiple areas affected by orientation, not just curvature), but one issue at a time. The fin units I received actually were printed vertically, so at least in this case, odd shapes and print orientation are not the problem, but material loss is.

I'll try to dig them up this weekend. It's possible they were printed at an angle so I'll check on that. I've got a few boxes of experimental parts filled with wrong turns.

It may be worth your while reviewing the information SW gives about the printing characteristics of their printer w/SWF. Makes for interesting reading, in any event, and very nice of SW to publish it. https://www.shapeways.com/tutorials/design_rules_for_3d_prin ting

That said, however, the referenced page shows they measured a deviation from desired wall thickness of only 0.06 mm at a nominal thickness of 1.0mm.
Naturally, that's not the polished version - it's what comes out of the machine.

If it were me, I would probably just bump the thickness up a bit. Or, perhaps take advantage of the fact that you're not confined to using traditional flat stock for the wings...you could perhaps get equal or better buckling strength even with the thinner cross section if you used a corrugated shaped sheet???

Regards,
...Shel