Shapeways parts arrived (analysis)

[bbussey]

I figured I'd start another thread so this analysis didn't get lost in the other thread.

I received my first set of test parts today, and I have to say I'm encouraged with what I received.  While design adjustments must be made in order to realize the full potential of the process, I can see with what I received that this may be a viable option for rolling stock body shells as well as more complex road vehicles such as buses.

All of the pics here are full size, so just click or right-click (depending upon your browser) to see the big pic with all of the detail.

Let's start with the items that can go to market as is.  The price quote I received from China to include the Heavyweight A/C Ice Bunkers as a part piggy-backed with the X58 detail parts was more than I expected and prompted me to explore this process.  I have to say, I'm very pleased with the results.  The photos don't do the product justice because of the translucent nature of FUD, but the detail is evident and clean.  Not as crisp as injection molded plastic obviously, but clean enough for this purpose.  Shapeways built the bunkers with the detailed front faces facing up, so there are no steps or cross-hashing of any kind on the visible face of the parts.





I will make these available to the public through Shapeways later this week.  While more pricey than originally expected due to their revised cost structure (the purchase price will be $10 for six pieces), it's better than not selling enough at $4 per set to cover the tooling expense.  And by being sold through Shapeways, there is no investment to recoup.  I designed the parts for some Micro-Trains heavyweight kitbashes I'm working on, so the digital models would have been designed regardless of whether they were sold to the public.

The second test item I submitted was the Chesapeake & Ohio F-9 well car.  While it technically remains as the next ESM car to tool after the X58, I remain apprehensive about releasing a lesser-desired style of car that happens to be based on a prototype of which only six were built (regardless of the number of roads that owned them at one time or another over a 50 year span).  Couple that with the realization that it's unlikely modelers will need more than one for their layouts, even if painted in their home roads.  I've only seen one photograph of two of the cars together.  The MSRP would have been deep into the $30s even back at the time the G26 gondola was released.  So it was worth exploring the Shapeways process for this model, and it yielded some interesting results.



The first thing I noticed is that these parts are delicate.  The ice bunkers were thick enough not to be an issue, but there is a lot of delicate relief on the well car.  One of the stirrups, which are similar in design to those on the BLMA gondolas, was broken upon arrival.  The "wood" floor is super-thin due to the car weight and low profile, and it was warped.  The bolts on the small plates on the end decks may or may not be visible, tough to tell with the FUD translucency.  Some of the deck horizontal edge detail didn't render, yet all of the vertical braces which were the same thickness did, as did the brake line.  You can see the stepping on the top deck slopes toward the well, yet it is comparable to the PerFacatory level of stepping.  But the big thing is that there is a notable shrinkage factor, as the recessed locks of the body were at least .030" closer together than those on the stainless steel underframe.  The body also was too narrow for the underframe.  There might be a shrinkage factor posted on the Shapeways site.  But if not, that will have to be determined before creating STLs so that it can be compensated for.



That said, I was very pleased with the stainless steel underframe.  It was not totally smooth on all surfaces, but it is smooth enough for a locomotive chassis and definitely smooth enough for rolling stock underframes.  It also has enough weight to keep the car on the track regardless of where it is positioned in a freight consist.  The part had a slight bow in the middle section, but it was easy enough to bend back to straight.  It has a tint of bronze in color, as opposed to the normal blue-gray hue.





I was able to force the body onto the underframe.  The slight bow on the left end is due to the fact that the underframe was flush into the recessed area of the body on the left side but not on the right side.



The underframe sits perfectly on the MTL Buckeye trucks.





Here you can see how the model is designed, and by how much the FUD shrank.

So the bottom line is that the stainless steel parts would meet all expectations, as standard Atlas/Kato/InterMountain etcetera engine components should work flawlessly in a properly designed stainless steel frame.  The FUD design will take some tweaking, both to compensate for the shrinkage factor as well as to be aware of any delicate detail parts on a model.  Unlike some of the war-game miniatures, or static vehicle models, rolling stock gets handled and is constantly in motion, so those issues are more of a factor.

[DKS]

Bryan, thank you for posting this analysis. It's extremely helpful for those of us who are exploring these media.

The stainless looks surprisingly "granular," with the edges being quite rounded, and the level of detail looks limited. Satisfactory, as you say, for an underframe, but have you rendered anything else in stainless that would reflect the level of detail possible? I'd be concerned about overall dimensional accuracy (as opposed to detail) for mechanism frame applications. Thoughts?

[bbussey]

David,

The edges are rounded in certain areas, but I believe this is a viable option for rendering mechanism frames.  I haven't taken a micrometer to the part yet, but it appears to be accurate.  I take measurements tonight and juxtapose them with those of the digital solid model and verify the accuracy of the build.

There is an ESM logo on the bottom of this part that did not render well, yet the smaller ©2006 copyright date CHINA did render and the emboss depth on that was the same.  You can see it in the next to last photo on the right left side, along with the square "bump" that was supposed to be the ESM logo.  I think we are limited to basic shapes for the stainless and limited shallow relief for consistent results.  But I still think it will be accurate enough for mechanism frames.  I might move forward on the EP-4 later this summer, depending upon how much spare cash I have.  I'm hoping that the shrinkage factor of the FUD is consistent, so that we can determine was the factor is and stick by that guide.

[sirenwerks]

Great news. The stainless steel sounds promising, especially for those wishing to explore kitbashing new locos requiring different frame lengths than what is offered. I've been thinking about Baldwin DS4-4-1000 and VO660 units for a while using the Atlas VO1000 as a source of parts. Maybe even a SW1 using LL and Kato parts and a Bachmann 44 tonner motor.

Maybe there's even hope for a BLW DR-6-2-1000 some day? Probably a stretch...

[DKS]

Thank you for your further observations, Bryan. I was able to see the copyright embossing you referenced. Comparisons to drawings will be useful, because assuming the accuracy is good, then this is indeed a boon for custom frame applications.

Somewhat disappointed about the shrinkage factor of the FUD. Seems like quite a lot. However, as you say, if it's consistent, then we can compensate.

Great stuff all around. Thanks once again.

[sirenwerks]

Bryan, I noticed there are a few Euro model producers using Shapeways. Do you think they're already figured the shrinkage thing out?

[Sokramiketes]

Did the FUD shrink or did Shapeway's overcompensate for shrinkage in the stainless steel process?

Don't know what FUD is, but usually the rapid prototype stuff is correct dimensionally (check against your original solid model)

The stainless would have a shrinkage factor in the molding and subsequent molten metal pour...  When I was going Perfactory to Brass it was 2%.

I'm just wondering if Shapeways took your underframe model, bumped it up 2-4%, and it ended up larger than what you wanted when it didn't shrink enough.

[DKS]

The stainless would have a shrinkage factor in the molding and subsequent molten metal pour... 

There is no "pour." The stainless powder is mixed with brass powder and a binder, and then built up in the usual RP process, just like the acrylics. At this point it's "green" and very delicate, prone to crumbling (dependent on the binder mixed in). The green parts are then fired to melt the brass, which fuses the stainless powder together.

Shapeways also offers a service to make silver parts, and these are made by creating an RP master with a wax-like material, and after that they follow a traditional lost-wax process to make the final parts. These are supposed to be very high resolution.

[Chris333]

Cool stuff. If you get the kinks worked out I could go for a FUD well car.  Even in person it is hard to see the detail. I sprayed mine flat black soon as I opened the box. Would be interesting to see how well the coupler hole threads.

[sirenwerks]

Apparently, the issues not just yours Bryan - http://www.shapeways.com/forum/index.php?t=msg&goto=19061&

[bbussey]

That could be a problem if the material is not consistent in its shrinkage factor.  The website says that FUD shrinks 0.1%, but the well flat definitely shrank more than a tenth of a percent - unless they compensated incorrectly.

I'm pretty sure the stainless part is correct, but I'll measure everything later tonight and report the findings.

I'll zip an email to Shapeways customer service after I measure, and see what kind of response I get back.

[wm3798]

What about creating a multi-part shell to go over the stainless weight?  That would allow for a reasonable margin of error, and the ability to "cheat" the parts together during the assembly process.

This would be useful if you're talking about marketing a kit.  It would obviously have its limitations if you plan to offer a factory built up model.

Lee

[bbussey]

My preference is to use the FUD as a master for making polyurethane castings, which would be stronger and more durable.  And if the shrinkage is constant, we can compensate in the STL.  But the problem appears that the FUD shrinkage isn't constant.

I measured both the stainless steel underframe and the FUD bodyshell for the well car.  The stainless steel part is dead on to the solid model - never off by more than .02mm in the few areas where there was a difference.  So the mechanism chassis possibilities are limited only by the imagination at this point. 

The FUD, however, is approximately 2% too small.  I sent an email to the customer service rep that I was corresponding with earlier in regard to the million triangle limitation, so we'll see what the response is.  I told him I can live with a 2% shrinkage factor as long as it's consistent, since we can scale the STLs at 102% to compensate.

As an aside, I noticed some minor rendering errors on the F9 body, and the bolt detail in the deck plates along with other subtle relief isn't visible.  It appears that the best option still is to tool the model.  So how soon this would be tooled depends on how popular the X58 is in the next year or two.

[Dave V]

So I was thinking about this...

I wonder if Shapeways would be a means to a conversion kit to turn Bachmann Spectrum 2-8-0s into Pennsy H-class consolidations?

[bbussey]

Well the mechanism would remain the same I'm guessing, so you're talking about a new engine shell?  It could be done, but it would be difficult to hide the layer stepping in the FUD.  And I don't see how a piece of rolling stock with an FUD body would not be damaged in day-to-day handling, let alone a motive power model.  I definitely would make castings of an FUD bodyshell master if I were doing it.