How long before we order rolling stock this way . . . ?

[jdcolombo]

The now-12-pages-and-counting thread on Atlas' acquisition of MTL and its associated tooling got me thinking about how the world of model railroading might change - drastically - in the next few years.

So here's a question for speculation.

How long will it be before we order a model by going to a web site, clicking on drop-down boxes to specify "NKP 40' PS-1 Boxcar, as-delivered, #6251" and then the order is fulfilled by having the boxcar 3D printed, painted and shipped from some large warehouse-looking building in Middletown, USA that has rows of automated 3D printers, a robotic painting system, etc.  No more injection molding; no more "tooling"; just a computer network that holds drawings and printing/painting info for every piece of rolling stock ever produced for a U.S. railroad.  The car comes to you as a simple kit: you snap the printed and painted shell onto the printed black floor; screw on the trucks and couplers, do some weathering (or maybe you can specify a weathered model in your order), and plop it on the layout.

5 years?  10 years? 

John C.

[robert3985]

@jdcolombo - John, the fly in the ointment right now to this scenario is 3D UV resin compatibility.  By "compatibility" I am meaning "robustness".  Most "robust" UV resins now compensate for brittle fragility by being more flexible...sometimes waaaaay more flexibility.  So, how does this flexibility get along with painting??  If flexible paints are used that will stick to flexible resins, then great(!)...if not...

Printer resolution is better than the resolution of the resins, meaning most resins today don't resolve what the printer is capable of at the voxel level...which is larger than the size of the pixel produced by the LCD.

Sooo...3D UV Resin Printing technology is making great leaps presently, with the mechanics of printing better than the medium with which the printing is done, and I'm pretty sure that more robust...maybe even less flexible, finer resolution resins won't take long to be readily available. 

But, that day is not quite here yet.

I'm pretty sure the automation/robotics of printing/painting are presently available, so that's not a problem.

I'm gonna guess that five years is overly optimistic...so I'll split the difference and say around 7 to 8 years down the road.

The whole idea of a "kit" needs to be modified because a complete car, with interior details, fine standoff details (grabs, running boards, brake wheels, underbody piping, chains, tanks, valves..etc.) can easily be printed monolithically...but, at least in my experience, some need to be printed separately to facilitate painting...which would result in the customer having to assemble a kit.  Also, metal weights would need to be provided. 

Also, I have serious doubts that 3D printed fine parts will ever be as robust as etched metal parts...so, maybe some etched parts might need to be thrown in the package too...or (thinking positively) maybe not.

I've been thinking about this scenario since I got into 3D printing about six years ago.

Cheerio!
Bob Gilmore

[MK]

I would say 10 years, maybe even more.  I'm not sure if there's enough demand in the US for this type of ordering.  I can see periods of 3D printers sitting idle, e.g. summer hot months when everyone wants to be outside.

Also, the quality has to be there.  Sure, small parts and what not great but entire shell or parts of shells?  Take a look at the injection and paint quality of Kato or Tomix Japanese trains.  I don't think we are there yet.

Lastly, the "some assembly required" feature may turn people off.  Yes, the majority of  here like to build and tinker, but many don't especially those outside of TRW.

Japan has a much bigger N-scale market than the US and they are not even close to getting to the point of order on demand.

[Maletrain]

I think that type of completed car to be "order on demand" is a step too far to predict at this point in time.

But, what about developing a set of stl files for specific cars that can be ordered on a one-off basis, similar to the old Shapeways mode of operation?  That seems doable today.  There could even be 2 options: one with the details molded printed on, and the other "bare" of such details with an added packet of etchings, bent wires,  etc.  Painting (pad printing) could be an optional added cost to the buyer, available for some, but not every prototype scheme.  That would be similar to the product availability today, but without the "limited run" issues of now-it's-available-tomorrow-it's-not.

That seems to me to address much of the head-end costs of mold tooling and craft assembly.  The basic model is relatively cheap, and making it better can be either DIY or cost more - buyer's choice.

Who knows, maybe "some day" there will be a process that can actually print the colors from some extrapolation of an an stl file.  But, I think that is too far over the horizon for this geezer to wait for.  Still, more people are likely to be willing to buy airbrushes and paint there own, compared to buying 3-D printers and printing their own.

And, if it ever gets to "Siri, find me some drawings of a B&O D-14ab and create an stl file for "Ben" (my Wifi enabled 3-D printer) to make 3 copies with the 1953 paint scheme, with track powered light circuit, please," there may not still be anybody left that knows what a D-14ab was used for in the "old days".

[Point353]

I would say 10 years, maybe even more.  I'm not sure if there's enough demand in the US for this type of ordering.  I can see periods of 3D printers sitting idle, e.g. summer hot months when everyone wants to be outside.

Adjust the pricing based upon demand as the airlines and hotels do.

[signalmaintainer]

The now-12-pages-and-counting thread on Atlas' acquisition of MTL and its associated tooling got me thinking about how the world of model railroading might change - drastically - in the next few years.

So here's a question for speculation.

How long will it be before we order a model by going to a web site, clicking on drop-down boxes to specify "NKP 40' PS-1 Boxcar, as-delivered, #6251" and then the order is fulfilled by having the boxcar 3D printed, painted and shipped from some large warehouse-looking building in Middletown, USA that has rows of automated 3D printers, a robotic painting system, etc.  No more injection molding; no more "tooling"; just a computer network that holds drawings and printing/painting info for every piece of rolling stock ever produced for a U.S. railroad.  The car comes to you as a simple kit: you snap the printed and painted shell onto the printed black floor; screw on the trucks and couplers, do some weathering (or maybe you can specify a weathered model in your order), and plop it on the layout.

5 years?  10 years? 

John C.

Ten to 15 years. It's inevitable that 3D printing will supplant traditional mass injection molding and no longer be an alternative. Print quality in smaller scales can only improve. Single-car orders that are reasonably priced (and profitable, too) still might be stretch, but perhaps a run of 10 or more would be.

It's an exciting prospect, John C.

[robertjohndavis]

If flexible paints are used that will stick to flexible resins, then great(!)...if not...

Bob - I think the idea of paint will seem quaint sooner than we think. Noch is already making outstanding figures in HO, N and Z that are printed in color.

I don't expect an incremental change in how models are made; I expect an entirely different model model... so to speak.

Rob

[rail_nut]

How long does it take to print a single colour box car shell with two doors and then clean it up ready to decorate? For mass production how many can/could you print at a time? I want to compare with mass production injection molding.

[robert3985]

Bob - I think the idea of paint will seem quaint sooner than we think. Noch is already making outstanding figures in HO, N and Z that are printed in color.

I don't expect an incremental change in how models are made; I expect an entirely different model model... so to speak.

Rob

Yup.  I'm aware of the figures printed in color.  However there are other factors than mere color to take into consideration such as surface texture (gloss, semi-gloss, flat), opacity, and resolution (for lettering, striping, emblems, etc.).  Although I'm pretty impressed with the Noch figures, figures are lumpy, whereas freight cars and engine bodies have large expanses of untextured flatness where any print artifacts are extremely easy to see.

I agree that the future probably holds the probability of color 3D printing, but, for realism in model railroading, using paint (at the present time) is by far the best option since the prototypes were also painted. It's fairly easy to replicate a surface color and texture using the same techniques and materials but with finer pigments.

Also, model train bodies need to be OPAQUE, and right now, only a few very specialized resins are truly opaque, most being plainly translucent in thin cross-sections...which N-scale has plenty of. I notice that the Noch figures, while not really translucent, are not fully opaque either.

But, maybe things will happen so that color 3D printing really takes off...but right now, it is sloooow, relatively imprecise and expensive as Hell.

Cheerio!
Bob Gilmore

[peteski]

Yup.  I'm aware of the figures printed in color.  However there are other factors than mere color to take into consideration such as surface texture (gloss, semi-gloss, flat), opacity, and resolution (for lettering, striping, emblems, etc.).  Although I'm pretty impressed with the Noch figures, figures are lumpy, whereas freight cars and engine bodies have large expanses of untextured flatness where any print artifacts are extremely easy to see.

Exactly.  I reviewed those figures here and the material is translucent. The figures look to me like they are coated with a thin layer of ice. But we still are in the infancy of this technology.

[NtheBasement]

Lots of products are built as specified by a customer after they order it; look how the marketing buzzword "bespoke" has cropped up in the last few years.  The pattern is that cheap stuff is mass produced before someone orders it, while the more expensive stuff that take a long time to build like ships, buildings and airplanes are built to order.  You used to be able to order a car but they are built "on spec" these days.  Unless someone invents Star Trek replicators for cheap stuff I wouldn't expect model railroad supplies to ever be built to order.  But notice all the pre-ordering that has become common in model railroading - takes the risk off the manufacturer, but only so they get an idea of how many to mass produce.

[thomasjmdavis]

How long will it be before we order a model by going to a web site, clicking on drop-down boxes to specify "NKP 40' PS-1 Boxcar, as-delivered, #6251" and then the order is fulfilled by having the boxcar 3D printed, painted and shipped from some large warehouse-looking building in Middletown, USA that has rows of automated 3D printers, a robotic painting system, etc.  No more injection molding; no more "tooling"; just a computer network that holds drawings and printing/painting info for every piece of rolling stock ever produced for a U.S. railroad.  The car comes to you as a simple kit: you snap the printed and painted shell onto the printed black floor; screw on the trucks and couplers, do some weathering (or maybe you can specify a weathered model in your order), and plop it on the layout.

5 years?  10 years? 

John C.

It won't happen.  The way it will work is that the AI that runs the factory will consult with the psychology and medical AI's, and determine what hobby best suits your temperament and physical abilities,  and produce the appropriate hobby materials for you.   With production scheduled during lulls in production of new AI chips and other things AI's will need to thrive and produce more AI's. 

If you object, the AI in charge of labor allocation will allocate you to the lithium mines.

[NSEGeorge]

The printer Noch is using for their figures is the Mimaki 3DUJ-553, which sells for $135,000. Thats a LOT of product to output to justify the cost, and you are right, it is in its infancy. It was originally designed as a prototyping tool, but enterprising folk can use a tool for anything, right? They also make a smaller one, which is $52,000 base price (the add ons take it up to 64K) We use that one to create props for filming.

It's a very difficult process to get right. Hewlett Packard tried and failed. Their version of a full color printer ($85K base) is discontinued. Statysys has one that is way more expensive per print than the Mimaki. There were the powder printers from 3D Systems that were based around gypsum powers and superglue that were used to make miniature versions of people (6-8" tall) and were only good for modelers if you were making a brick wall or an asphalt road (texture).

I expect that in the next generation we will see the democratization of this technology, just as we see things like the Euphymake E1 UV printer bring that tech down. However, there is nothing as crisp as Tampo printing for cars .......yet.

[jdcolombo]

It won't happen.  The way it will work is that the AI that runs the factory will consult with the psychology and medical AI's, and determine what hobby best suits your temperament and physical abilities,  and produce the appropriate hobby materials for you.   With production scheduled during lulls in production of new AI chips and other things AI's will need to thrive and produce more AI's. 

If you object, the AI in charge of labor allocation will allocate you to the lithium mines.

Heh.  You might be right.

But I do think "objects in the rear view mirror are closer than they appear."  I really can't believe the quality I'm seeing from a project I'm working on with John LeMerise to do a 1000-series NKP caboose for N scale.  This is something that will never be produced commercially (they were unique to the NKP, but were as much a part of the railroad's identity as the Berkshires), so a 3D print project was the only way to go.  And yes, we're going to use etched detail parts.   But when I look at the "test" body printing John has done that is sitting on my workbench, now painted in Caboose Red, I see something that looks every bit as good as the body of my Atlas NE-style caboose.  So I suspect the progression here will be that more people will turn to 3D printing for "special" models over the next few years, and then we'll have another Shapeways-like company whose business model will be doing these sort of custom projects.  And once that happens, it won't take long for that business model to expand as the Pennsy guys demand more specialized Pennsy rolling stock, the B&O guys clamor for wagontops; I get a F-M H10/H12 shell to pop on the Atlas VO1000 mechanism, etc.  I'm going to go with Bob's initial 7-8 years.  Hope I'm around that long!

John C.

[rail_nut]

Short or low volume and one offs might make sense to 3D print.  With CAD/CAM mold making, injection molding is quicker and cleaner for longer runs. I think it will still be less expensive seeing the price of the 3D printers mentioned and having no idea of the service life of these printers, injection molding machines last a very long time. Once set up for a run they don't require very skilled operators to produce high quality parts. If part is a single colour they can be molded without needing painting, or if multiple colour can be molded in one and additional colour painted. I don't know if it still the case but injected molded black parts were way less expensive than white. Exceptional quality clear parts are easy.