While making some phone calls today I was thinking about a solution to a problem that plagues so many places... making some of the junk look better. I work part time during the summer on the 195, which has been sitting outside for decades "on display" in Skagway. The engine is missing just about everything. Rather than find inappropriate replacements to slap on, how hard would it be to have some false water glasses, safety valves, whistle, etc. printed up? Most of these parts were universally purchased by the 1940's, and once the files were made up it seems like it would be pretty cheap to have 200 fake safety valves printed that could be used on display engines all over the place. No worries about scrap thieves or expensive replacements and plastic lasts forever. Thoughts?

195 is sitting way off the beaten path, behind the museum in Skagway, but plenty of folks off the ships 'discover' her every time they dock there. Plastic stuff wouldn't last very long in that climate, especially as cruise ship people swarm around it if they stumble across it.
I doubt plastic stuff would last long on that locomotive where it sits...

There are some changes happening around the engine, including a greenway for people to walk from the junction up to the museum along Pullen Creek. The "abandon porn" status the engine had in years past is fading.

A fair number of people do use the cab as a place to get stoned. I'd rather have some cheap replacement parts to use which will make the engine look better as a complete display rather than worry about acquiring hard-to-get parts that could be damaged. If the engine was going to operate, my interest would be different.

Print up some patterns and have a foundry cast some fakes out of iron.

3d printed fakes could work well if the locomotive was in a climate-controlled area, which 195 is apparently not.

3d printing as many know is simply puting layers of melted plastic on top of one another, slowly building up the layers into a 3 dimensional part. How well these layers bond depends on the type and quality of the plastic, and how hot is is when it is "printed". The quality of the printer could also have an effect as well.

Exposure to the heat and cold, and the sun (as well as moisture with some plastics) will cause the layers to separate and break down, effectively turning it into junk. It would definitely work, but the 3d printed parts would likely have to be replaced from time to time. I'd expect the parts to last a couple years, maybe a little bit more, depending on the quality and severity of the elements. 3d printed fakes definitely wouldn't be permanent stand-ins.

Edit: Speaking from experience with a 3d printed flower pot, any water that gets inside and freezes will also rip the layers apart and ruin the item.

Try it out and see how they really hold up, this could be something that could become more widespread if successful, and if not, castings like what Mr. Rowlands suggested will always be long term fixes.

I think the concept is a good one, but you could whip up a wooden fake as well. Probably you want to cover the fake in a water tight paint to help prevent the damage mentioned above. I like the idea for a cheapie display fix to fill out the details.

Aside from that you could sell copies for fund-raising. IE..this fake whistle was mounted on the engine for display, has officially been mounted on the engine, take a pic of it for the buyer, sign it somehow with the buyer's name on the whistle.

765 regularly makes copies of their builder's plate, mounts it for a run, and fund-raises sells them, putting a new one on.

I've often thought someone could have a business manufacturing replica display locomotive parts out of the same kind of heavy duty plastic they make outside garbage cans out of. There could be generic molded bells, headlights, number plates, and so on that would dress up any park locomotive and not be too expensive to replace if someone walked off with them.

3D printing is getting far more sophisticated.

Carbon fiber printed parts are just round the corner. Those will outlast just about anything.

For smaller parts, a desktop FDM printer can do the job. Anything over the average 6x6x6 or 10x10x10 though would require some creative splitting of the model into smaller pieces and then adhering them back together. While services like Shapeways exist, expect to pay some ridiculous prices--I heard one story of over $400 for a pattern for a builder's plate.

I have a little experience using one of these myself (printing some stuff right now, actually), and a lot of the quality from consumer-level printers is going to be based on how much you tweak the settings, the type of plastic, and your finish process. They're not at the point where you can just 'push button, get part/pattern/prototype' yet.

Personally, I print primarily in Acrylonitrile butadiene styrene (ABS, about $25 for a 2.2 pound spool). Since it's hygroscopic, it will absorb latent moisture from the surrounding air, which will affect its rate of degradation if left fully exposed. Given that FDM is just laminating paper-thin layers of plastic (in my case, at 210C), any moisture that gets in there and freezes will absolutely wreck the printed part. The contact surface is only 0.4mm wide (depending on the extruder tip size), and most prints aren't 100% solid--I use an infill as low as 7% to decrease weight, material usage, and (importantly) time.

You can counter the degradation if you're willing to sit down and mess with it a bit. First thing is to get rid of the visible layering. Even if you're sending it to a foundry for casting, you probably don't want it looking like a stack of paper that's been cut out. If you had to split the part, that includes hiding seam lines (get out the bondo!). ABS has an amusing property that will help smooth/seal the layers, though: it dissolves in acetone, a fact that I use to 'weld' multi-part prints together with at home as well as finish parts prior to priming and painting. Though I note that with multi-part prints, you need to adjust the fit tolerances slightly as the feature is only the center 0.1mm of that 0.4mm extrusion width. 0.15mm is on the inside, the 0.15mm on the outside will mess that up.

If using the printed part, it will definitely need to be coated in a waterproof coating inside and out. In 195's case, I'm not how well even that would hold up.

Long and short--using 3d printed parts could be viable, but your mileage may vary. Considerations should be taken into whether a wooden copy would be cheaper/easier/faster to produce and the longevity of all possible options. As Mr. Rowlands implied, nothing will last like cast iron. Heck, even making a urethane resin cast could work. Silcone molds aren't too difficult to play with (though the lower-end of 2 part 1:1 silicones run ~$180/gallon) Decently strong, durable, and you're left with a pattern already done (and the mold) if the casting goes missing.

Hot metal that's a great idea.

Start collecting data, blueprints, photos, etc.. on the accurate replacement parts that the loco will need to be made operational some day, and get started on the CAD files for those parts. Get your patterns made and instead of casting the parts as brass/bronze, etc.. cast then in plain old iron for now, and you will have usable cheap parts soon that can be replaced later. The patterns will be stored just waiting for the day when the real bronze or brass castings are needed.

That is a most efficient and Fantastic idea.

Fabricated non-functioning replica whistle and safety valves we built and installed on ARR 556 on display in an Anchorage park to replace items that disappeared 40-50 years ago.

They have 3D printers that can print in metal now.

Are we talking about the same process? There are three 'common' processes that we should probably be differentiating between here: Stereolithography (SLA), Direct Metal Laser Sintering (DMLS), and the more common Fixed Deposition Modelling (FDM).

DMLS an produce parts that are up to 80-85% the strength of their cast counterparts, yes, but FDM? Not exactly. DMLS works by essentially welding powdered metal together with a Ytterbium fiber laser, 20 micrometers at a time. The typical FDM printer's extruder doesn't cross the 250C (~482F) mark, barely enough to melt pewter, much less bronze, brass, or iron.

There are 'metal' filaments for FDM printers out there, but they're really just powdered metal suspended in a plastic binder. The plastics will suffer from the same problems with the metal powder suspension as they would without them.

The thing is that in terms of availability, DMLS machines are less common and far less accessible to the average home user (or preservation group with a shoestring budget). From the perspective of a group looking to make display equipment parts using AM techniques, FDM machines are more widely available and have a large community of support to ask questions online, plus are relatively inexpensive (~$500 for a 6x6x6 assembled) and have a lot of common, widely-available components. When a comparable SLA machine runs $3000, and a DLMS unit being even higher, cost becomes a factor. For what use it would get, FDM would do the job (be it patterns or actual dummy parts).

3D metal printing is not as good as it sounds, resolution is not good and requires machining afterwards to get a decent looking part. At work we have been looking at this process for several years waiting for it to get good enough to make prototype molds. At this time it is still more expensive to make metal 3d parts then to just machine it out of solid blocks of metal.

That's been my take on the 3D metal printing as well. The technology is still new, and the results aren't that good yet. Also, while it is metal, it's probably not as strong as you'd like.

I'm wondering about an alternative. What if you were to 3D print the part you wanted, and then used that to do a casting in the material of your choice? I realize there may be issues with shrinkage, but if it's just for a replica, it shouldn't matter much.