Feasibility of 3D printing Erie Stillwell coaches

[jmlaboda]

Does priming and painting seal the surface enough to ******** shrinkage?  Or does this have no impact on the chemical process?

Its part of the printing process, not the completed print, so priming and painting will have no effect since it would be already present.  From what I have seen prints aren't that hard to get right providing the artwork submitted is done right.  Hopefully this can all be dealt with appropriately.

[Lemosteam]

Steady progress...

[Sokramiketes]

I bet Bryan Bussey has got all of this figured out.

By reading the low-side got thread, it all sounds like a tolerance nightmare!

[thomasjmdavis]

Thanks for the replies everyone (to the question on shrinkage).  On costs  for resin bodies, I will take your word for it.  I've done a couple small parts myself with an "everything you need to be casting in minutes is included" kit, and I know it takes some time and bother, and materials are not cheap.  You all are the experts, I am just happy you have turned this from a pipedream into a project in the course of a couple weeks.

I do find it interesting that rapid prototyping is turning availability of models and pricing on its ear.  A lot of stuff is beginning to appear for N and Z scales that is quire rare in HO.  And because it takes so much less material, and can be produced on smaller machines, N scale pricing on RP is much more competitive with injection molding.  We are getting to the best of both worlds- and if I ever do actually build that Dearborn station model (another pipedream, but I officially retired yesterday, so wondering what I will be able to get done when I can put in a 40 hour week on model building), from looks of things, the "Scott and John with some help from their friends" C&WI Stillwells could be pulling out on the short tracks on the east side of the station just before a Kato produced Super Chief pulls in on the longer west tracks.

[Upstate Gator]

I want to second @thomasjmdavis with my appreciation to you all for working towards moving this project from a question about possibilities towards reality!

[chessie system fan]

How many castings do you think we're talking about?  I've got all the casting equipment needed and wouldn't mind having a Stillwell or two.  Maybe we can work something out.   

[Erik aka Ngineer]

I was wondering about shrinkage- I have no real knowledge of the physical properties of the various shapeways materials, but I have read several complaints or observations on the phenomena.  Question for those more knowledgeable than I am...

How much shrinkage is there in percentage terms?

Does priming and painting seal the surface enough to ******** shrinkage?  Or does this have no impact on the chemical process?

Might it be better to make a master via rapid prototype and then cast resin copies?  I have several resin cars and have not noted shrinkage with resin, although I have noticed some that become brittle if not finished.

In Shapeways' production process the models are printed slightly larger to accomodate shrinkage. They shrink a tiny bit during production, but all the parts I have printed with them seem to be the size I wanted them to be. Shrinkage only occurs during production, not afterwards. I would recommend sealing/painting all SW parts just to make sure they are not affected by UV light etc. Nobody actually seems able to tell if this matters, but I want to be on the safe side. Painting the inside of parts in Frosted Ultra Detail also helps bring out the outside paint better since the material is translucent.

[randgust]

I did a very intensive attempt last year to take an RP master and do resin copies.   Did not go well at all, had shrinkage and warping of the RP shell printed in FUD. 

Priming and sealing is just critical if you're going to attempt using one as a resin master.   Darn things are porous.   When I made my first attempt, the rubber molds were just peppered with bubbles coming out from the master themselves.   Attempted sealing with Badger (a very thin acrylic) and they warped like crazy.

Did a second attempt where the designer primed and sanded it themselves, corrugated passenger car, went a whole lot better - no bubbling, no outgassing, but the dimensions of the RP print were measured SHORTER after the entire process, it most certainly shrank over the length of an 85' car.   Quality was good, but for me, the shrinkage and warpage issues are very real, and I'll add the outgassing as a bonus.

[Chris333]

I wonder if FXD prints slower? Maybe that would cause less shrinkage over FUD?  I haven't bought bigger stuff, but just about everything I have in FXD looks great.

I know a guy who designed a rod driven chassis in 3D. Using the same drawings he laser cut the siderods. Because of the shrinkage he had to re-draw to make the rods fit. And this was a short 2" long chassis:
http://s276.photobucket.com/user/jnj1097/media/IMG_3835_zpssgf8lbax.jpg.html?sort=3&o=1

[bbussey]

Been following this from the periphery through Scott.  A couple of observations:

PPD LTD actually has the ability to send the brass through the etcher twice.  You can get three levels of detail (plus cut through) without having to stack layers of brass!

This is good to know.  I have some heavyweight etching projects I'd like to try this on.

To go along with Scott's etchings...

John, it's less of a headache if you eliminate the window detail and just provide a long rectangular cutout for the insertion of "window glass" cut from clear styrene.  This will allow the glass to rest against the back of the etching overlay, and aligning the parts becomes less of a factor.  Also, design for the insertion of a fully-assembled MTL 1015 coupler box, rather than building in the coupler box in the FUD core.  I do that only when it's impossible to allow for mounting the MTL coupler (such as on gondolas), and even then I embed only the box and use MTL lids.  Drilling (or cleaning) and threading FUD/FXD pivot posts for #00-80 screws is not for the timid, as being too aggressive will crack the post.  It's far more delicate and brittle in that aspect than polyurethane resin.  If you have to include the pivot post due to model design, add the hole for #00-80.  It will render tight and still need to be drilled out, but it's much easier to enlarge a hole to #00-80 in FXD than drill one from scratch.

Priming and sealing is just critical if you're going to attempt using one as a resin master.   Darn things are porous.   When I made my first attempt, the rubber molds were just peppered with bubbles coming out from the master themselves.   Attempted sealing with Badger (a very thin acrylic) and they warped like crazy.

It is very porous.  The KV caboose masters were practically completely covered with etched brass insets and overlays, and it still needed multiple primer coats to seal the FXD. And, it does warp under pressure casting. The one best plus to the casting method is that the etching drawings can be matched to the one FXD core after the fact to insure the tolerance is tight.  But in this case with the Stillwells, the insets in the sides should be fine.  The Shapeways FXD machines (at least the ones in Queens NY) are calibrated perfectly.  I haven't noticed any deviation in the measurements regardless of orientation, which is why the Low Side Gons and the final batch of G32C gons will have FXD bodies.

[Iain]

In Shapeways' production process the models are printed slightly larger to accomodate shrinkage. They shrink a tiny bit during production, but all the parts I have printed with them seem to be the size I wanted them to be. Shrinkage only occurs during production, not afterwards. I would recommend sealing/painting all SW parts just to make sure they are not affected by UV light etc. Nobody actually seems able to tell if this matters, but I want to be on the safe side. Painting the inside of parts in Frosted Ultra Detail also helps bring out the outside paint better since the material is translucent.

If you clean it properly, it is not transparent.  You are leaving wax on your parts, which is going to cause issues with paint down the road.  You need to soak your parts in bestine for a while to remove the wax. 

[randgust]

Another observation to make here..... if you are doing an RP master for resin molding, you absolutely must make every attempt possible to remove any undercuts, notches, or anything else on the INTERIOR surfaces, where the mold 'plug' has to be removed after molding from the car interior.   There's a temptation when designing in RP to be too efficient and step surfaces up, down, and around to minimize material use and cost.  One of the problems I had on using an RP-designed shell as a master was having to fill in, with putty or styrene, every last bit of those shelves, undercuts, and hollows to allow the interior mold plug to come out cleanly.

And while the first attempt I did with a designer on a diesel shell was pretty much of a failure (in order to get the shell sealed it had to be sanded, and there went all the surface detail like louvers) the second attempt on a passenger car actually worked - the designer got the surface of the car pretty well sealed with primer inside and out.   Not sure how much was luck on the original.   But like the first time, while the cast part measured back to the master, the cast part ended up being about a foot and a half short on an 85' car, and when you measured the RP, it had shrunk over original dimensions as well.  That part was 2 for 2 on both attempts.   After it happened the second time, I'm pretty well convinced that between the printing, priming, molding... something happened to the RP master itself.

Also on design - you've got to be aware of undercuts.   The only thing that didn't mold cleanly was the area around the car end diaphragms.  On design, that would have been better as a separate part added after molding.  What I'm still wondering about are things like rivits and louvers, if those are better left off and added before molding (like Archer details), because sealing and sanding 'scale' printed detail like that has so far been a failure.

Don't think I'm against RP because I'm absolutely not.   I'm now doing a second custom build for a customer on a GE 25-tonner using the RP printed body.  It's really pretty darn good.  Sealed it with autobody primer, light sand and the end surface was very, very good.   And I just got a couple forklifts for myself.    On small stuff like that I'm not seeing many issues except for the brittleness factor of FUD.

[chicken45]

If you clean it properly, it is not transparent.  You are leaving wax on your parts, which is going to cause issues with paint down the road.  You need to soak your parts in bestine for a while to remove the wax.

I don't think that's what he's saying. If you take a thin piece of styrene and draw and X on one side with a Sharpie and hold it up to the light, you're going to see the X from the other side. It isn't thick enough to block out all light. That's why we spray the insides of buildings black if we light them.

In other news, I've seen FUD not frost even after soaking for a week. What's up with that? It's patchy.

[randgust]

I've seen the same thing.   The thicker the cross-section, the more likely it is not to 'white out'.   

[peteski]

The resin used in FUD and FXD printting is neither transparent nor white.  It is translucent.  The white appearance comes from the rough surface finish (printing artifact).  This is usually caused by the contact with the wax supports needed to support the resin during printing.

It is sort of like frosted glass.  When it is dry then the glass has a whitish appearance. This is because light is scattered in all directions due to the rough (frosted) surface. But wet that piece of glass and suddenly it becomes almost transparent. That is because water fills the microscopic voids in the frosted surface, smoothness them out.

Wax remains in the 3D printed parts are basically wetting their frosted surface showing the natural appearance of the resin.  Rinsing the wax off and drying the part gets rid of the smoothing so it becomes much rougher and it scatters the lights making it look white (like piece of dry frosted glass).

If you were to spray that part now with a clear lacquer, ti will become translucent again (since the clear lacquer will fill the microscopic voids in the resin.

There are sometimes areas of the 3D printer parts which were not in contact with the wax like tops of the object. Those are much smoother right after printing and will remain translucent even after rinsing the wax off. That shows the natural color and translucency of those resins.