coaster

Happy December everyone! After a few prototypes and redesigns, I believe I have my final solution for the overall track design as to how it will assemble as a piece-part and within the track system. Here's a pic! I had to do away with the rail clips, as I found they interfere with the power connector. I also had to widen the rails from the classic "I" shape to more of an inverted "T" to give room for the mounting screws. Note that the rails themselves will be square cut, with exception being for the adapter tracks for standard Lego 9V tracks. The electrical connection will be maintained through the small metal plate. One of the advantage of doing it this way is you will be able to remove this plate to create isolated track segments, e.g., with proper controls, a balloon turn will be possible with 9V tracks. Most importantly though, it simplifies tooling, a necessity for the short runs we'll be dealing with. For the adapter tracks, the plate will be removed and the adapter-side rails will be cut to match the Lego "handshake". The spring of the standard Lego rails will be sufficient to maintain proper contact. The other side will remain square cut. I have prototype R40, R104, and R200 curves in process. Switches are being modeled still, but I should have those to show soon (trying to improve that mechanism as well). Once I do, I will layout a dimensional cut sheet of the tracks with their lengths, angles, etc, for final comments before I initiate the kickstarter. Again, I apologize for the delays. I spend every spare moment I get on this, but I just don't get that many of those.

Thanks for the support everyone! Just a quick update, I am working on changing everything over to the 2x8 sleepers, but it requires a few changes to the electrical connections. My goal is to get that designed such that I can use the same ABS base piece for adapting to LEGO 9V tracks. I've had a few small delays as I updated my computer. Naturally, once I did, nothing worked. But I'm back up and running now, so hopefully can have something new to show in the next couple weeks.

I would agree, but that's what the OP said worked to take it off. I think if you're careful, it'll be fine, but otherwise, yes, rubbing alcohol or floor/countertop cleaner that doesn't destroy ABS should do the trick.

This may sound a little odd, but are you sure it was actually corrosion and not simply a build-up of crud? I know a number of my tracks look like that, and it's just gook from God-knows-where that's gotten stuck on to it. What makes me question the corrosion diagnosis is that the rails are nickel-plated commercial bronze. Nickel is actually very resistant to corrosion; in fact it's one of the main corrosion resistant elements to stainless steels. Nickel doesn't rust, so it wouldn't care less about humidity or proximity to rusty nails. The chemicals that will corrode nickel you probably don't want to be in contact with anyway. But the thickness of the coating is extremely thin, typically around .0002-.0003" (5-8µm) thick, so if the nickel were truly corroded, that light sanding to clean it up would have revealed the base bronze metal. More likely, it's trace amounts of oil and dirt that have been electrostatically attracted to the rail surface. Could be from storage, or could even be from use at shows or around your house. A wooden crate would indeed have a high concentration of such matter. The good news is, it should be easy enough to remove.

That 1/8" KS tubing seems like it would be a little sloppy to me. The ID is Ø.097", in which a Ø2mm shaft will really rattle around, more than the standard wheel assemblies do. I checked with a few of my barstock sources, I can get Ø.085" ID tubing for pretty cheap. Shoot me a PM with how many you need and I'll just cut them for you. Shortest tube I can get is 1 ft, which will probably yield about 8 pieces, but if you need a handful of them, should only be a buck apiece, or less even.

Just to be clear, you mean for the tubing to press into the clip light piece: ...such that it sucks up the slop from the wheel axle? EDIT: ...and provides a stop so the wheels stay centered.

Hmm, interesting idea Aaron, but it still doesn't resolve the issue of having different left and right curves, and as such, different left switches and different right switches. It kills me to say it, but as much as I like the look of the 1x8 ties, there doesn't seem to be an elegant way to implement it to the stud grid. I think I need to scrap that idea and go back to the 2x8 sleepers. It just simplifies everything.

At 16 studs long, the gap between has to be 1 stud long, otherwise the spacing overall doesn't work out. Unless you're showing something else that I'm missing.

I started looking at that last night a bit. It's a little awkward, but not terrible. This is more a sketch than an actual track, but it would look roughly like this: Bottom of each crosstie looks like a 2x8 plate with 2mm sawed off of each side. The studs would have to be recessed from the top instead of bottom, or simply not present. I think there's a lot of value to keeping the studs though, even if it means some LUGs need to tile each one.

That won't be necessary; the transition tracks would still be 16 long. The tracks themselves are normal to the grid, only the crossties are offset 1/2 stud. And this is necessary to facilitate even tie spacing while still keeping the track segments symetrical, i.e., we'd have to have different curve pieces for going left or right otherwise. One thing I could consider is designing the bottom of the crossties to naturally assembly between studs. That'd take some work to change, but could also look kinda weird. We could always go back to the 2x8 crossties too, I just think the 1x8s overall look more realistic. As for the color, that's easy enough to have options. I figured to make them in dark grey, dark blue-grey, dark brown, and black, regardless of what design we settle on.

Don't worry; way ahead of you on that. Curves have been done for a while now, and I have a sandbox layout to make sure everything fits like it should (here's your crossover, zephyr): But you are absolutely correct, with the 1x8 crossties, we do have to have them in between studs in order for this to work, unless you'd rather have a 2x8 crosstie at every connection, which, frankly, looks terrible. I'd start investing heavily in the 1x2 plates with 1 stud.

A 16-long track adapter is in the works, yes. Yes, the switch will allow for a single crossover to a 16-stud centerline, while keeping the studs horizontally in line as well. Some notes on the curve segments and switches though: In order to do the crossover (or a parallel siding for that matter) properly, the switch needs to terminate at an angle of 22.62° (standard R40 curves are 22.5°). While normally 0.12° is pretty trivial, at this radius, it translates to a .435 stud length gap (a little more than the thickness of a tile, or 3.5mm / 9/64"). As such, the switch will indeed be 22.62°, but will also need to include 2 special angle curve segments, 22.62° and 7.38°. The normal curves will be 15°, so the special curves will be easily identifiable. As for the kickstarter, it is the plan, but I don't want to go there just yet. I want to make sure I know what my costs are going to be and have all the design issues buttoned up first. Are issues going to crop up? Of course, but that doesn't mean I shouldn't do my best to mitigate them on the front end. Based on my current rate of work (which, sorry, I have 2 little kids at home, so I only get to work on this after they go to bed), my very rough timetable looks like this: Finalize designs and models - Nov. 1 Get quotes and prototype units built - Dec.1 Initiate kickstarter (assuming minimal changes to prototypes required) - Dec. 12 Second round of prototypes (if necessary) - mid-January Complete kickstarter campaign (1 month) - Jan. 12 Tooling & approval units- 10 weeks Initial Production & Assembly - 4 weeks Shipping of kickstarter units - End of April Shipping of all other order - Mid May There's naturally a lot that can slip between here and there. I've spent the last 10 years in product development, so I have no delusions of it going that smoothly, but I think May 2016 is a realistic timeframe of when I could be in full-tilt production.

That's the reality of 3D printing. For what it's worth, 5 years ago it would have been double that. And unfortunately, the studs being on the sides means you'd need action on the mold tooling, so you're probably looking at about $15-20k for tooling, get you down to about $.60 apiece, so break even is somewhere around 400-500 pieces if you injection molded it. Doesn't really become economical though till about 5000 pieces.

Thanks guys! I'll move forward with the new mechanism then. As for the rails, I got a quote on custom rails that was pretty reasonable, and at production quantities will end up being cheaper than just buying stock rails. The best part about new rails is I can make them exactly the same height and width as the standard Lego rails, so no playing games trying to hold the gauge and allow for the power connector or other brick connections. Of course, I have a buy a mill run as a minimum, which is about 1000 lbs, but as small as the rails are, that translates to about 3.7km, however, that's realistically not that much if we go production. I will probably have to buy more than that even. I'll figure it out. Let's make some tracks!

Need some input on switches! After many up-till-2-in-the-morning nights, I have the switch design mostly finished, and boy was it time consuming. There's still some detail work to do, but it's more or less there. I need some input though as I've redesigned the switch mechanism. I don't care for the Lego version, as it has a tendency to jam on me, and isn't particularly conducive to automating. Anyway, here's the new one: The new mechanism is a quarter-turn of this 2x2 barrel section, and it turns in the direction of the train, i.e., turn it left to take the left branch, right to go straight: Some additional notes: Mechanism is spring-loaded to lock left or right; there is no in between settings and no clicky-sections. Points are still spring-loaded, so trains running in the opposite direction can still pass through. Note the two moving points instead of one. Track is wired the same way as the standard Lego track. Power is to whichever track is active. Switch mechanism extends out the same length as the standard switch. Height is 1 brick. Barrel section has axle slot 1L deep. Thoughts on the mechanism change? Keep it, change it, or go back to Lego's version? Next step is to finish detail work, and then prototype switches, straights, curves, and to-Lego adapters. Hopefully will have something up and running by mid-October.

Got straights and curves modeled, getting quotes. Have had to make a few small changes along the way, but we should be in good shape. For the rails though, they're apparently too small for most extruders to do in brass or similar non-ferrous, non-aluminum metals, so I'm seeing if I can get it cold drawn to shape. Failing in that, we'll have to go Aluminum, but there are things we can do to improve conductivity.

Here's you go:

It's not a big deal, but it sounded like it was from your earlier comment about the extruded design not being compatible. As for the ME metal rails, no idea. I think they had issues with their supplier and with the connections. Don't know what the price is/was. As a separate issue, there is a tiny problem with the R120 curve: it doesn't fit nicely for a switch. I think it's generally agreed that the centerline spacing between tracks is 16 studs, which is why ME adopted the 40, 56, 72, etc. sequence for radii. However, maintaining that for switches, such that the siding is 16 studs offset while also ending on the full length of track grid (i.e., we can't have a siding that can't then reconnect to the main) doesn't work. That falls on a different pattern of 40, 104, 200. Even if we allow half length straights on the siding, our R120 still doesn't match. But, as you may have noticed, the 104 & 200 do, so do we adopt 104, 200, or both? Open for discussion. So, the short list of things that should be available: Track adapter Standard 16-stud length straight (for those that don't want to bother with the adapter and want everything the same) Half length (8-stud) straight Wide radius curve Wide radius left switch Wide radius right switch Long-term wish list: Flexible track segments 2x and 4x straights 3-way switch wye switch 90° cross 45° cross double cross-over switch What about standard radius curves and switches? My thought is to have the switches arranged a little differently, maintaining the curve to 37° (there's some funny math behind this number, but it makes sense), and then have an 8° and 37° curve segment to go with it. Would these be useful for sidings and yards, or skip it and just stick with new tracks? I'd like as many opinions as possible, so please comment if we're going in the right or wrong direction.

No, they'd be a custom extrusion, similar to O-Gauge, but made specifically for this. Extrusion tooling is dirt cheap. I can screw it up a bunch and still be ok. For me though, it's 9V or nothing. If we were just making plastic tracks I'd do it all in one solid piece and be done. That'd be easy. Well, that's where I was going before: losing compatibility is the compromise for getting 9V track in different configurations without spending a fortune. I don't think ME has shipped any metal tracks yet, but there's a pretty good reason why nobody else (LGB, etc) is doing the handshake connections like that: they're hard to do and expensive to machine. The stampings are flexible and give you spring-loaded positive contact, but with a solid extrusion, you're relying on as small of an interference fit as possible to make the connection. Too much interference, and it's too tough to put tracks together. Too little, and you could easily lose contact, and the tolerance for that middle ground is very small, which means machining becomes critical. I haven't personally used the ME metal track, but from what I've seen at shows, the few guys that had it were using either copper tape or solder at the joints, so obviously it wasn't making good contact. I'm still trying a few things on the original design, but failing in that, I'd like to have a Plan B going. And there may be some way to make it compatible, might just take some work.

That's the real objective. I need a ton myself, and I'm obviously not alone in that. Short of this turning into a truly insurmountable task, I plan to fully manufacture straights, curves, turnouts, crosses, etc. However, there's quite a bit of work to do between here and there. I agree; I don't think it's that bad. The problem is, because of the complexity of duplicating the Lego rails (thin-gauge stamping), most of the stamping places that are capable of working with material that thin want progressive tooling, the cost of which is, frankly, soul-crushing. At Lego's volumes, I doubt they'd bat an eye at it, but in our terms, at $180 for a half-loop, we'd have to produce over 8000 curve pieces to break even. Possibly doable for the curves, but I doubt that's realistic for the turnouts. I don't think I'd have much luck pitching $2k for a left-hand switch. In other news, I drew this: Base is a single solid piece, end rail connectors lock into the base and are crimped to the rails, holding the assembly together (much like your typical O-gauge track). Tracks are 16 studs long like standard, however, the rails overhang by a 1/2 stud on each end, that way we get the correct crosstie spacing on assembly, like this: I'll work on curves next, but they take a bit longer to model, so wanted to get these out there first. Thoughts?

Well, that's pretty close to what I'm trying to put together right now. Give me a few more days to work out some of the finer details. I don't mind printing some prototype pieces, but production units would be traditionally injection molded. Honestly, I've never actually seen flex O-gauge or G-gauge track. Smallest flex tracks I was aware of was HO, I just figured the rails were too thick on the bigger ones to be practically flexible enough. But even so, and this may just be a difference of opinion, I've never cared for the flex track designs. I don't like that you have to cut the rails once in position. If you're building a permanent layout, then sure, no problem there, but a lot of us build a layout, tear it down, and build something else. Of course, no reason we couldn't then do rigid and flex options, but like I said earlier, let's walk before running. Give me 2 or 3 days, and I'll post what I've got in mind, and we can discuss further from there.

I wasn't going to use aluminum; I was commenting on the ME design. Extrusion tooling, whether it's Al, Brass, or something else, is generally pretty cheap. I'd probably use something like a grade 260 Bronze and then nickel plate it. Not entirely sure, I'd have to look into it a bit. As for the radii, since we're locked into a defined 8mm grid system, it makes sense to stay within those parameters. The standard curves are R40 (40 studs/320mm), and it's pretty well decided that any additional curves should be R=40+16*N. Also, there's not really a consensus as to what scale Lego actually is, so it's hard to say what a #4 turnout would even be. The problem isn't so much the physical connection, rather also getting the electrical connection. Probably could have an adapter of sorts punched out, but I'd have to look at it. Sorry, not sure what you mean. I have an o-gauge track here too, but it's much the same way: pretty rigid.

Well, typically, the G scale track isn't flexible, at least, not the ones I'm familiar with (I do a bunch with LGB, and that's all rigid). However, from a manufacturability standpoint, it's a piece of cake to make different lengths and radii, so in that sense, it's super flexible. Switches and crosses, I can figure out, shouldn't be a problem. And I hear you about being heavily invested in 9V. I've got a number of Lionel and LGB tracks lying around at home. I'll pull some stuff out and model something up this weekend if I get a chance.

That's the idea, but really, anything that is compatible with the existing 9V track. However, I've seen the ME metal track, and I can't say I'm a huge fan of it. The extruded aluminum design is nice since extrusion molds are cheap, but all the machining that has to be done afterwards...ugh. I sadly find it unlikely they'll be able to get it to work well and do it cheaply. I'm quite familiar with the G scale design, and that would work, and in fact I don't want to discount it altogether. For those not familiar, it looks like this: The issue though is transitioning from existing tracks to whatever new ones we make. Although, that does beg the question: What if we simply do-away with everything we currently have and implemented a new G-Scale type track design? It could still be made compatible within the Lego stud system (i.e., crossties would still be 2x8 with studs and receptacles on top and bottom) and stand-off height would still be the same. Track segments would still assemble in a similar manner, It just wouldn't connect to existing track pieces, though we could still make it compatible with the 9V contacts. If it meant wider radius curves and switches are readily available, would you guys pull up all your existing straights as well? Thoughts?

Why 312? It certainly fits the R = X*16+8 stud requirement, but I don't see it fitting nicely into any particular pattern (though admittedly, I haven't looked that closely at it). I think if I were to do something that wide, it'd probably be 360, which fits the pattern and is an even 9x standard. I'm not discounting it, just trying to understand it. At worst, I could have some printed for you; you'd just have to sand and finish them yourself.