Wednesday, November 30, 2011

132> Forward Compatibility

I know I said that my next post would be about standards, but I think this topic should come first since it weighs into that discussion. There has been something in the back of my mind, in every design that I have posted on this blog. It is the recognition that progress often happens in baby steps, yet baby steps often don’t go in a straight line or a logical direction.  

I am talking about forward compatibility. The vexing thing about designing PRT (or anything else) to be forward compatible is that it requires a design that allows evolution toward an end that must first be defined itself. In the case of PRT, that eventual product would ideally be very fast, silent, comfortable, able to be deployed with minimal cost, and be adaptable for every contingency a city could throw at it. These adaptions  would include a variety of station types for different locations and passenger volumes, being practical within buildings, being able to be elevated to a level deemed acceptable by the effected parties, etc. The vehicles themselves can be assumed to be, in this future world, mass produced to point of enjoying economies of scale.

I am sure that many readers have dismissed my designs as being too ambitious. The vehicles, in particular, have very advanced electro-mechanical systems which are designed to remedy situations which could largely be avoided in the first place with little negative impact. Well now you know the reason. By designing the vehicles as though Toyota had been building them for decades, one can better consider the best design for the track and stations that may still be around when that day comes.

In the meantime, however, PRT will be subject to the restrictions that our current economic, political and technological realities place on it. The problem is that designs for today and designs for tomorrow are strikingly different. The main culprit is the tire wear and noise/vibration associated with speed. It is true that smooth running surfaces and track clamping emergency braking capability enable harder rubber, solid tires which don’t need to flatten out on pavement to achieve the high traction requirements associated with gripping slippery roads. Still, highly wear resistant plastics, such as are seen in rollercoaster wheels, are a recipe for a very noisy system at high speeds, especially on pipe, which is notoriously good at amplifying sound. (This is the basis for many musical instruments)

I should point out that the design I show in the Oct. 30 post enables both steering guide wheels to be raised for high speeds, eliminating contact (and therefore noise and wear) by the plastic flanges. Also, these flanges are to rotate independently of the drive wheels, so that they may make contact with the track (pipe) anywhere on their surface and create their own rotational speed based on the diameter established by that point, rather than the smaller diameter of the tire. This reduces wear on what is a tiny contact point.

The main point remains, however, that high speed systems should have larger wheels (OK, not Maglev) or risk lots of wear and/or lots of noise. A system that requires wheels or tires to be changed every few thousand miles would be a disaster. But longer wearing, larger tires means bigger track, something that is nearly as bad, in that it raises the cost and visual impact of a system which will, in the beginning, be under intense scrutiny from critics. Also initial systems will probably be slower anyway, because such trial systems will have to first prove themselves for short-distance downtown use. 

Could there be a two tier system? Would it be crazy to start with a system for downtown that would preclude high speed vehicles? I know it sounds like a terrible idea, but it would probably shave 20% off of the track costs, and vehicles would be discounted considerably more. And let’s face it. The other PRT systems out there aren’t exactly fast or flexible either.

In the illustration above, the system on the left, which is obviously simplistic and incomplete, would only need to raise and lower the small pairs of wheels to steer. I do not believe that the middle “hold-down” wheels (illustrated in previous posts) would be required, so that’s really all there is. There are inexpensive “off the shelf” hub motors available that would fit in the drive wheels, and the flange and hubs could be cast as one, (in urethane) so that solid rubber tires would slip on. Such a bogie would be extremely cheap to produce. The complex (expensive) articulation capabilities of the swing-arm and gondola could also be dialed back in such a “starter” system.  

The obvious problem comes from the fact that the fast vehicles wouldn’t fit in the smaller track, although the slow vehicles could run in the high speed track. So what is a city to do? Well, there are a couple of things to note here. First, the high speed track would be equally usable for GRT. (Group Rapid Transit.)  A track going out to an airport, for example, might well be a good stand-alone investment used in this way. Passengers coming from the airport to downtown would need to change vehicles to use the downtown PRT, but the upside is that they didn’t have to make the long trip at 30 mph.  Faster, express PRT could share the track at some point, and slow vehicles could use the track at certain times of day. Because PRT is a smart technology, if high-speed track is running through a grid of low speed track, the slow vehicles could still, in theory, get on and off without disrupting the high speed service. (Assuming sparse high speed traffic) If the track is modular, standardized and interchangeable, the slow track could be removed (during an upgrade) and be reused elsewhere. In the airport example, for instance, slow track taken from downtown could be used to build a network around that airport. In such a case changing vehicles would be a minor inconvenience for relatively few passengers. It is also noteworthy that, in a downtown environment with mixed track, fast vehicles can’t get up to speed anyway, because of sharp turns. Therefore slow vehicles sharing the (fast) track would be no problem. In such a case the system could simply send a slow vehicle if a trip would involve a stretch of slow track.

I have come to the conclusion, reluctantly, that there are theoretically reasonable migration paths from slower, inexpensive PRT to faster systems capable of tackling longer distance commuter traffic. The examples above are just a sample of the possibilities.  They also show that it takes some creativity to undo what many of us would say is a very shortsighted decision. (To put down track that can’t take fast vehicles)  But at least it’s better than having no forward compatibility at all!  It is unfortunate that such a complicated situation should ever exist in the first place, but I have my doubts that we can ever get to PRT 2.0 without first dabbling in PRT 1.0.


Juho Laatu said...

Flexible specs that allow tracks and bogies of different categories is a good idea. Maybe typical categories would be some weight and speed classes. In your proposal one key problem was that some bogies can not run at all on some tracks.

Was the cost or the size of the track more important criterion here? If it is the size, then large wheens simply can not run on those small tracks. If we talk only about the cost of the track, then maybe adding some height would not be a critical problem. One can save maybe somewhere else. A track with higher profile maybe does not need as many poles to support it. Extra poles could easily take as much metal as was saved in the low profile tracks.

One alternative approach would be to have a standard interface between the bogie and the gondola. Then one could change the bogie when one moves from one type of track to another. One could make the bogie change automatic if one really wants.

One approach for very cheap tracks would be to use half tracks. This is not a very elegant idea, but if cost is everything, then why not. Btw, could the simplest half tracks consist of only the bottom section (one quarter) of the track, i.e. no "roof" at all? (Also large wheels would fit in that open space.)

Dan said...

Specifications and/or definitions, based on ranges and/or absolute values… I think there’s a lot of room for good or harm…we just need to be very sure there is solid reasoning behind any decision.

About size or cost… It’s really cost. Not that it is expensive… Actually it’s ridiculously cheap compared to the alternatives in use today. What I worry about is competition from other forms of PRT that are neither multi-axis nor high speed. After all, the advantages of a system like I describe would mainly become apparent only after a number of years – after it is actually an industry and people start wanting tracks and stations everywhere they are likely to go. I worry that cheaper systems, that can never achieve that flexibility, will dominate the market, and PRT will fail to live up to its potential. Also, early stage PRT is barely viable without the kind of subsidies that other forms of surface transportation enjoy. There isn’t really any fluff in terms of track or stations, and I’m crossing my fingers that a white knight(s) will eventually come along to build the vehicles.

About your point about a taller track needing fewer posts- I agree and would add that that is a principle problem with the ¼ track idea you mention. Yes, you could probably run around on a single piece of angle steel, if you are trying to simplify to the extreme. But there is no structural support, not to mention weather issues.

I’m not going to think much about gondola to bogie interfaces at this time, although there are many theoretical possibilities for such an eventuality. That is an area where you want to entice manufacturers, not cities or PRT operators though. It would probably be a good idea to let them patent their designs - up to a point anyway.

Something that you, in particular, might be interested in, (that wouldn’t justify a whole post or announcement) is that after looking into structural steel bending, I am now convinced that round pipe isn’t the only game in town. Square and rectangular tubing and bar, in particular, may be easy to get radiused locally, even in small cities. I mention this because I am particularly dissatisfied with running the high speed steering guide wheels on the round pipe, and I am seeking the best way to flatten that (currently angled) surface while keeping the track fabrication process as simple as it is now. Also, I forgot about emergency braking when it comes to the small bogie going in the big track. The plan was to press brake shoes against the “ceiling” and the running surfaces, since both remain even in a switching area, although adding a flange to clamp to instead would be straight forward enough, I suppose. Maybe it is even better since standard disc brake calipers could be used. I just hate to add steel that doesn’t have a structural purpose, even something that light. I’ll have to mull it over.

Juho Laatu said...

One design in my mind is to have some triangular parts where you now have the two tubes below one half track. Maybe that would mean angle steel + welding two (around) 45 degree surface at the bottom. This approach would allow quite steady contact with only three surfaces to lean to. That includes avoiding tilting to eiter side in curves (in switches with half track support only).

You said that you want to save steel where possible. One has to make the bottom of the track very strong in any case (to carry the weight of the bogie+gondola). The current design is good since it relies on the stregth of the bottom of the track and leans to it (or the tube below it) also from the other side (below). My point is that this strongest part of the track could be used also for braking. In the tirangle approach above you could use the two 45 degree surfaces to brake. Maybe also the main running surface would be used as a (third) braking surface. As a result a minimal (half) track would have only three (flat) surfaces for the wheels to run and (same three surfaces) for the brakes to contact. And all of them would be close to each others, requiring hopefully only small amount of steel. The rest of the track would have to support this core part of the track (also against bending when bakes are used), but its structure would be quite free.

I guess you may still need those C-shaped "claws" that "grab" the whole track in your design to minimize the amount of steel and keep the structure solid. That means that also the triangles that I discussed above would have to leave some space for the claws (just like in your tube based design where the outer small tube is a bit lower than the inner tube). I however note that it may be good to build the two half tracks quite separate so that for example braking would not rely much on the claws to keep the two halves of the track or the ceiling of the track where it is. The two halves would thus be independent, and the main forces to clamp on the track (to drive or to brake) would be limited (separately) to the bottom party of each half track only.

If the width of the track is small enough, you could imagine also (half) track that consists of one piece of angle steel and two triangles only (i.e. no explicit claws below the track). I'm not sure if that would make sense from both cost and strength point of view, but that would be simple. (Is the need to maximize the distance between two poles btw a sufficient reason to make the track profile narrow but high?)

Btw, my approach to standardization at this point is to batter your proposals as much as I can to make sure that they ate solid and stable enough to resist any other attempts to prove them wrong or inefficient in meeting the targets (of cheap tracks everywhere etc.). :-)

I've said this before, but I think this approach is a quite good approach for suspended C-shape tracks. I think also other approaches may work (also supported and even ground style), but this approach has many enough benefits to have the potential to become a (or the) standard track. I discussed earlier on this blog also the inverted T-shape approach for suspended systems. I have one pending reply on this topic (since my browser failed and ate the reply). I plan to come back on this design too - at least as a theoretical construct. But let's at least make the C-shape design solid enough for the world market.

Dan said...

I hear you, Juho. I have to confess that there are enough options and reasons to choose different ones that it calls into question the wisdom of any standard. just keep in mind that my last design can have an extremely tight turning radius for very close parking and storage maneuvers. The flanges are not for normal travel except exiting the main track. (half track)

The outer angled running surface you describe can't engage a very large wheel, so it seems like it is just another half-track requirement that would be disengaged for high speeds, no?

Juho Laatu said...

> tight turning radius for very close parking and storage maneuvers

I'll need tight turns also to keep the cost of simplest/cheapest tracks low (my home yard, small shops, when tracks are built in whatever small space there is available). Storage could be built also as straight track segments. Parallel parking based stations could need tight turns.

In the tight turns we can assume very low speed. Maybe that will allow us some liberties like not keeping all the wheels and flanges in contact etc. I'm not sure if turning wheels / bending bogie would help. But anyway, the ride need not be as smooth as in high speeds.

> flanges

Flanges may be a good idea for high speeds. In a half track whose profile is simply one large triangle the small wheels and brakes could be disconnected but very close to the surface. That could be a safe approach that protects against tilting and other surprises. The small wheels could btw be of the same material as the flanges (to avoid friction).

> disengaged for high speeds

Yes, I guess it is a good basic assumption that all small and extra wheels would be disconnected (but maybe not much) in high speeds. In some designs one could even use slightly turning wheels instead of the flanges.

It is possible that a wider base (wider than one can get with contacts near the main running surface) is needed to resist tilting (high winds, high speed). In that case some (normally disconnected) surfaces high up insde the C-tube would do.

As you have discussed earlier, some of the small wheels (that touch the track only occasionally) could be replaced with "pads". An extreme (and cheap) solution would be to have only the main wheels as wheels. A suspended approach works nicely here since it is possible to build the system so that 99% of the time only the main wheels touch the track.

First place where small wheels could be needed again in addition to pads is switching. The temporarily missing support of the other half track wheels could be replaced with support from below the other half track (edge side, further to the edge than the main wheels, maybe using "larger than small" vertical wheels).

An alternative approach would be to move the weight of the gondola below the continuous half track while switching. Just two (maybe turning) main wheels needed. I wonder how complex this would be. But maybe cheap tracks are even more important. This approach would allow me to build my (lightweight, slow speed) home track all the way using one (very simple) half track only.

One more observation. If you use mainly pads instead of small wheels, then the number of pads is not critical. One could use also vertical and horizontal surfaces instead of the round and 45 degree ones. A basic T-shape half track could do. (main wheels above, multiple pads below the track) The brake pads could use either or both surfaces.

Dan said...

Sorry Juho, I have had zero time. I will say that my primary concern is in the mass-production more than general geometry. I just bought a white board to sketch stuff out, and I was going to do a post on it, but now I may talk about ULTra's new project instead.

Are you talking about welding the angled pieces for their whole length? PS - you are right to suggest under the track as a substitute for the hold-down wheels (and the wheel flanges)for slower vehicles and an option for braking. I haven't found anything magically compelling, though, to want to build a standard around...

Juho Laatu said...

Yes, welding might be a natural approach for the angled pieces. I'm not an expert enough to say which technique would be the most economical one in each part.

In my proposals I was also looking for some very cheap solutions, like ability to build a quarter track (I sent some picturess too), maybe without electricity, in my back yard. And still compatible with the more complete main (C-shaped) track.