[Full disclosure: I work for The 3D Printing Store's branch in Monte Vista, CO. The main store in Denver can be visited at: http://the3dprintingstore.com/ I am not an engineer, so the design, manufacture, strength, etc., I leave to the experts! :-) ]

Apologies to those who are already familiar with 3D printing...

A previous thread "Machine shop looking for work" (viewtopic.php?f=1&t=36375) mentioned 3D printers as an option for making parts.

3D printing is 'additive manufacturing', where the base material (plastics, resin-based materials, ceramics, metal, etc.) is added via a computer process to build an item/items.

For the plastic-build machines, it's similar to a hot glue gun that takes a solid material, heats it up, then uses a computer program to apply the melted material in a specific pattern in layers of 100 microns or less.

Other printers use different processes to build - including lasers - and the print 'envelope' size varies with the printer/manufacturer.


This video link shows the test-firing of a 3D-printed rocket engine injector:

http://www.youtube.com/watch?v=kFFeqmMprV4


...and this link is a 3D-printed rocket engine from students at UC-San Diego:

http://www.youtube.com/watch?v=VFXIN3ZvJYw


The upshot is that 3D processes are available to build and/or reproduce parts for trains, planes, tractors, automobiles, and lots, lots more. Existing machine shops can add to their manufacturing and/or design opportunities; and depending on the design, 3D-printed items just might of less cost to the organization...or the individual.

Thanks for reading!

-Jim

I think many of us are interested in the possibilities but know too little about the materials science in the technique. I can easily see using it to make a plastic loss pattern for casting an iron part, but can it make an iron part as string as the casting itself would be? So, that headlight bezel yes, but the superheater header probably not as far as I know........but I'd be very interested in learning otherwise.

dave

We used this process to make a pattern for a soft iron break shoe.
The nice feature here is to use 3D CAD to allow for casting shrinkage by increasing the design size before you go to the printer.
If you or someone in your shop can work 3D CAD, and or has a printer, this is a lot quicker than making wood patterns.
If not warm up your checkbook.

I expect Matt Austin to have some choice caustic comments about using contemporary printing technology for a very significant percentage of the structures making up large steam locomotives.

The printed rocket engine was part of a NASA challenge last year, which involved the Shapeways process (which is not selective laser sintering; it involves using an organic adherent, and the hard-material particles that are assembled into a 'green' form are then brazed together. See the Shapeways design guide (this is for stainless, but the operational requirements for cobalt and Hastelloys, etc. are not radically different):

http://www.shapeways.com/materials/steel

Selective laser sintering properly applied can give high strength to a PM structure -- but it is slow. Consider the bounding volume represented by the pistons of a SLS printer, and consider the "4 to 5 days" process time as a component of machine financing. If you have a part that could be fabricated by HIP, SLS may be an attractive idea. Yes, you could scale this to fairly large part areas by synchronizing the pistons (and the control logic to do this in realtime exists). Just don't go calling it cheap enough to replace conventional fabrication techniques for most railroad-restoration activities!

The rocket-engine projects (the full printed engine, and NASA's printed injector plate assembly) are carefully chosen to look sexy and high-tech, but realize that the actual stresses are not comparable to what locomotives are going to produce, particularly with respect to long-term creep or stress-fracture resistance.

Now, as a method for pattern-making, many of the 3D printing methods make better sense; for example, some of the complex cores involved in casting complex valve passages can be made directly from 3D simulation (there are ways to do this with lost foam, but they're harder to shape with high precision). But here too, expect very high per-unit cost; you'll still want to make a 'second cast' of the printed original, perhaps with one of the silicone/epoxy casting systems of known expansion, and do production cores (or castings) using that approach.

Nobody will be happier than I will when 3D printing becomes so pervasive that I can set up to make large pieces in my basement or back yard. The laser technologies to make this 'practical' aer already established, and even largely costed-down in other applications. But I would not hold my breath waiting for this to become 'the' method of choice for other than special boutique items, at boutique prices.

Wasatch has an established agreement with our local Community Collage to do 3D printing. To date, we have made cylinders, pistons, piston rods and pistons using a 3D printer. In addition we have made three injector bodies that have in turn been used as molds (patterns) for castings. We have also made name plates. If you want a real trick....whenever you need a name plate or a badge plate made, do it in 3D then have that pattern cast as the actual......cheeaaappppp.....way of getting it done!

I would LOVVVEEEE to post pictures of stuff we have done, except, a poster, mentioned in another post on this thread is very critical of anything we put up.....so, if you want to see this technology used in restoration, send us a note, we will send you back some photos.

It is great technology and most certainly has a place in restoration!

Kindly,

Sounds like a great subject for a new YouTube video, John! Just post the link.

By all means, John, send me what you got. I've got a lot of little bits and pieces needing to be found or made in the next few years.

The cylinders, piston rod and piston you mention having "made" with 3D - do you refer to patternmaking or to the actual metal parts?

dave

There have been some long sweeping learning curves for us in this area. Here are a few examples:

A.) We have made 3D parts as "actual parts" for trial basis on actual units. For example, the piston rod and assembly that we made was for the New Cagney Locomotives that we are building. We wanted to ensure that our piston rod length and our cross-head would be good mates, so, instead of spending money to have a "real" one made (either in house or otherwise), we simply did it in 3D. The actual, 1/1 piece was then installed as a "real part" and we tested it. It worked great. We then made the actual steel pieces.

B.) When using 3D for castings (bronze, brass, cast iron or otherwise) you have to consider a number of factors. Shrinkage is a real factor. We ended up making some lost wax molds, which, were made from the 3D patterns that we made. We then had to up-size the 3D piece to compensate for the lost wax mold and the shrinkage we would see in the final item. This is a trial by error deal. Same goes for name badges and or name plates.

C.) The biggest issue with 3D (in my book) is resolution. When you draft an item in CAD, for example, if you are drafting in lower resolution....what you see on the screen may not be what you get in real life. You learn this when you make a type of round object and you get it out of the printer and it has flat sides. Though it looks like a circle on the screen, when you print it, it ends up with lines in it. This is a deep function of the software and once you learn how to set resolution (without crashing your system) you come out with almost perfect printed parts. However, the higher the resolution, the longer the print time.

D.) Dave, if you have parts to be done, send me some pictures and examples. I will then give you the pro's and con's to each and what we would do if we do them for you.

A video, eh? Hmmmmm.....we could consider that and see if we could do a fast one.

Kindly,

Thanks John - so you are using it as a prototyping tool as well as for making patterns. Sounds very useful and as costs come down, possibly also economical.

Shrinkage and draft need to be considered in any kind of patternmaking, and it seems that there may be an algorithm developed along with a resolution standard that is scalable for this use.

Thanks for the fascinating information.

dave

It's very encouraging to see the replies to this thread...and their diversity!

I, too, have wondered about the strength, reliability, and cost of a 3D-printed item, especially as they pertain to railroading and other applications.

'Copyrights' is also a question for which I do not have an answer, e.g., if you print a copyrighted part from XYZ Corporation (personal use? for retail sale?), are you liable for royalties, or might you be opening yourself to a lawsuit?

A slightly-related topic: area students have asked if we can print air, glass...and money!

:-)

At the risk of diverging from strict on-topic (preservation) discussion:

A slightly-related topic: area students have asked if we can print air, glass...and money!

1) Theoretically you can print gas molecules -- CVD is essentially this. But 3D printing a gas is obviously silly... you can't even print normal-viscosity liquids and have then stay; diffusion effects in gas are much more pronounced. About the closest thing you can do is attempt to control the diffusion characteristics, similar (in a more sophisitcated way) to what those 'environmental harmony disc players' tried to do a couple of years ago. That new research on 128 possible distinct smells clearly indicates that some means of modulating chemical stimuli in 3D + time has applications...

2) You can print glass, but not in just one pass. And the surface finish is going to suck just as badly as the SLS or brazed rocket nozzles, unless you use some sort of aftertreatment. Pate de verre, of course, has been doable since at least 2009, whether 'faked' with permanent binder or used to produce greens for firing. For optical pieces, just 'braze' with one of the materials with equivalent refractive index to your spheres -- the stuff used for windshield cracks being an obvious 'feedstock' starting place. Assume a technology will be developed to implement the science of optically polishing this composite material (it may already substantially exist, but is likely largely proprietary at present).

3) Yes, but it reminds me very strongly of the old 1940s joke about the Air Force tactical weather computer, 'predicts the weather perfectly 24 hours in advance, but takes 48 hours to finish the run'. Making hundred-dollar bills that conform to all the chemical and physical tests can be done... but is still illegal if you 'miss a stitch', and it's relatively easy for Treasury to implement and roll out more tests if any appreciable volume of this stuff gets out there.

On the other hand, I recall a story about BART, where dollar bills made via some form of copy process came to represent some very large proportion of the bills processed through their TVMs. So there is a market for 'printed money' -- just not fancy 3D-printed money with all the little systems provided at 'good enough' resolution in the appropriate materials.

George O. Smith, the science-fiction writer, had some interesting stories about what happens when 3D duplication becomes a mainstream technology. Money was the explicit focus of one rather good (if technology-dated) story...

Now, before we all get big ideas about 3D printing, we need to all take a small step back. The 3D printing we do is made by a printer that prints in one media....a plastic type of media.

Printing money is a different media. Money is made primarily of cotton, right. So, until we have the technology to do 3D printing that includes, cotton, paper, ink, and other medias all in one, 3D printing of money is out. For that matter, doing 3D printing of glass and other items is out until they come up with the technology to print in that media.

We use the technology to sample a design before we actually make something. Included is a photo of a 3/8 inch injector body that we are working on. This first casting is rough, but it shows the 3D printed mold fairly well. If you look closely enough at the photo, you will see lines in the casting. These lines came from the 3D printed part that was used as the pattern. The resolution on our 3D part was not the best. The name plate, which was also a 3D printed piece turned out well.

The purpose of this project was that of determining if 3D lost wax molding would be good enough to produce multiple injector bodies in one pour. Indeed, our trial worked and we are now in the further stages of finishing this injector project and selling these injectors.

A lot of areas of manufacturing are using 3D to print items for trial and or testing or as mock-ups to what a actual piece would be. We at Wasatch are doing it and I believe that 3D will be a substantial part of future restoration work.

(Edit: Took me a bit to get the right file uploaded.)