Sunday, June 13, 2010

90> Let's Take It From the Top

I want to go back to the beginning in my discussion of PRT track, specifically track for hanging style PRT, for the benefit of the majority of readers who have not followed the ongoing development of the corresponding SMART (Standardized Modular Automatic Rail Transport) specification.

The most fundamental concept that needs to be understood is that of switchless steering. Unlike a railroad, which requires a track segment to be mechanically moved to provide switching, PRT designers have generally agreed that this can be more easily done by the cars themselves by means of engagable/disengagable guide wheels. These require a vertical surface to engage upon. Since the main wheels (or even a Maglev system) require a horizontal running surface, the two, in combination, look something like this.

 It should be noted that only one half of this book-matched arrangement is theoretically needed, except for places where tracks are merging and diverging, though leaving the redundant half in has advantages (load distribution, simplicity of symmetry) which I believe out-weigh the disadvantages. This picture,
from post 54, illustrates the concept.

The second concept is the combining the functions of multiple wheels into one. A look at a roller coaster wheel setup  shows how multiple wheels can hold a vehicle tightly to a track. A simpler approach, however, can be achieved by the use of flanges, as on a train wheel, or dished in or out wheel or track profiles. The angle of contact, though, in such systems, should be as close to a right angle as possible or the flange will tend to wear out, since it introduces a point of friction to the system. A quick look at my older posts will show various examples of radii on both wheels and track.

A third concept is that plate that keeps appearing in my track designs, pictured below.

Whenever a tube-like structure is bent, it tends to flatten out in the process. If a section of track is spanning a great distance and is bearing a great deal of weight, the downward forces will tend to want to flatten, then fold the track under that weight. Stopping this flattening arrests this tendency and strengthens the track. In the case of an encased track with a slot in the bottom, the initial flattening would cause the sides to fold inward or outward, thus narrowing or widening the slot. These plates, placed periodically, prevent this. In the second picture you can see how such bracing can be connected by long plate steel to make a complete box beam. Personally I find this design ugly and worry about single wall designs for PRT because of noise, condensation, and uneven expansion.

A fourth concept is modularity. Almost all discussions on PRT touch on the possibility of movement of goods as well as people, particularly at night. Additionally, many envision a variety of vehicles, and the current diversity of motor vehicles that populate our roadways would tend to bear out the desirability of this approach. Such ideas are not put forth by actual PRT vendors, however, since it currently would be counterproductive for them to develop and offer a confusing array of alternatives. Then there is the matter of routing. Long commuter type routes would call for faster vehicles or possibly vehicles for groups, if there were sufficient numbers of people with common origins/destinations to support this scenario. Therefore there would seem to be a very strong case for creating a track that is compatible for all or any of these vehicle types or speed ranges. This logic leads to a track profile where smaller vehicles are compatible with the track of the largest anticipated vehicle type. I submit that this “largest vehicle” would be GRT. (Group Rapid Transit) In the U.S. many cities have “Park and Ride” systems already in place, which aggregate passengers in outlying suburban locations. I would not want to rule out shuttling in these passengers in groups of, say, 10 or 12. Although GRT is a contentious issue in PRT circles, it should be noted that from a track design point of view, there is very little difference dimensionally between the two. Heavier gauge steel, more frequent supports, or simply spacing heavier vehicles more sparsely are ways to accommodate such a possibility. Whereas PRT track could be made smaller (in profile) than track for GRT, the size of the truss that is required to span reasonable distances can fit either. I would suggest that any such heavier track be used for arterial routes only, because cost is everything when it comes to expanding limited routes into meaningful networks, which, after all, is what PRT is all about. This modular approach recognizes that no one set of engineers is liable have all of the answers as to the very best vehicle design.(s) any time soon. Therefore I suggest removable running surfaces. This has the added advantage of reduced noise, a smoother ride, and essentially no chance of condensation. It also looks to a future with possibilities like Maglev. In order to accommodate the largest number of vehicle types I suggest something like this general shape.

The design allows for several centering means through the use of either convex or concave surfaces. My only concern is the fabrication of this profile, especially radiused sections for turns, because the required roll-forming equipment for that is not universally available on a local basis. Although I have tried to keep my designs easy to produce anywhere, this may be something that would need to be shipped in. I suspect, however, that many local job shops would be happy to fabricate the shape by splitting stock pipe and making curves by a tack-and-bend-as-you-go method.

Although I promised some further thoughts on use of tension cables to extend the span of track sections, I would like to put that off for the moment and return to standard truss design. As far as box trusses go, the round tubular type seems to be acknowledged as the strongest, and was the choice for the Skyweb Express demo track. I believe 90 ft. was the maximum span.

It has the disadvantage, however, of being relatively complicated to fabricate, because the pipe ends must be cut to match a curved surface. (I have seen the tube ends are flattened in some applications) I have wondered about this for some time, as it seems that square tubing would be so much easier, since it just needs to be mitered. Nonetheless, in most demanding situations, such as cranes, these complex joints are almost universally used. Here is a picture showing he complex cut. These have just been tack-welded.

In the illustration below I have taken a similar box truss but substituted those plates for the verticals. Because the plates are only rigid in one direction, I have stiffened them by joining the upper and lower gussets into a single piece. I have also used simple miter cuts on the diagonal pipes, again, using gussets to reinforce the joint. This is not exactly to scale and the diagonal and horizontal pipes don’t even touch in this picture, (much weaker) but this is just a conceptual drawing. The general design should, however, be very strong, cheap and easy to build anywhere where CNC flame-cut plate and pipe is available.

In the next picture the red surfaces indicate the running surfaces for a PRT bogey, taken from Post 83.

If it looks like I haven’t decided exactly what becomes part of the truss and what becomes removable, you’re right! I have not. There’s also a radius missing. It’s a work in progress. Also not shown is the skin, and provisions for electrification, communications, carrying utility wires. There’s plenty of work to do, and now you all know were I stand at the moment. For shorter spans or hanging sections, BTW, the design would be completely different, but, as they say, “One step at a time.”


cmfseattle said...

bravo, Dan. I agree with all of the above.

re: evacuation & suspended systems: again, my only concern w/suspended is the cost of higher-strength materials. hanging track from a post will be more expensive than supporting said track, no?
(also, note how the artists' renderings compare with the real thing; if you count the 2 curbside parking lanes, this street was 6 lanes wide. arches would tend to favor a suspended design, i guess.)

Dan said...

Dan the Blogger Responds -

Hi Cmfseattle – Interesting! One mile and no stops? Those better be two pretty compelling places to be! Almost seems like it should be expanded, not demolished. It’s pretty hard to get the network effect when it’s a network of two.

About cost of hanging vs. supported:

There is no doubt that you can’t do any better, cost wise, than to have the track resting on a cylinder. Offsetting it by, say, 5ft. would probably add $750.00 (USD) per support. (Materials only - steel with concrete foundation) If the posts are designed to be particularly graceful it could be a lot more. T’s (dual direction) would seem to be about the same either way. Arches or an “A” shape would be pretty cheap, except most sidewalks don’t have room.

Post-stressed concrete, (my favorite) would be one way to make the hanging design almost as cheap as supported, but the molds for a good looking shape would be expensive initially. Then there is the matter of some kind of color/finish treatment, which tends to be expensive with concrete products because the migration of alkali means that coatings don’t generally adhere very well. One possibility is to spray the inside of the mold with a colored bead, sand or gravel mixture prior to pouring the concrete. I have used aquarium gravel and crushed glass experimentally for some interesting effects, and once simulated the sparkle of natural rock by dusting wet cement with mica flakes. But I digress…

Andrew F said...

At 30 m spacings, an increase in cost of $750 per post works out to $40k per mile. Sounds like a rounding error to me.

I think it's important that the guideway and support be unobtrusive. It's certainly work slightly higher cost to achieve that, at least in areas where people will spend a lot of time up close and personal with the supports.

cmfseattle said...

cmfseattle said...

the failed expansion of the seattle center monorail in november, 2005 was what led me to discovering PRT. for 40 years, the monorail was the only U.S. public transit system that turned a profit.

so I was curious about what went wrong.

(also, it was on Thanksgiving weekend that year that the 2 trains collided) (doesn't appear to mention that after the 2006 stalls, the trains were limited to 200 psgrs; not sure if that has changed, now that both trains have been overhauled and the guideway power rails completely replaced).

I learned a lot about what people thought of the old monorail, and the new plan
I also have a PDF written by Jon Magnusson to the seattle city council, which I think should be required reading for anyone seriously considering elevated transportation (a.k.a. "looming" structures).

and then I found the "get on board PRT" site.

I should also mention that Dr. Anderson did a PRT presentation for SMP; what they learned: PRT would require 4 times as much guideway as monorail. yes, really.

for me, probably the most important lesson is that these things take time. lots and lots. too much time. and since everyone uses transportation, nearly everyone considers themselves experts on transportation. :|

Dan said...

Andrew, I have to admit it has been a while since I have been to the steel yard. I was last paying 75 cents per pound for cold-rolled. Last time I bought concrete it was $85.00 per cubic yard delivered. As I said, I was just talking about cantilevering by 5 ft. That's not much. For example, that much "I" beam @ 18" high, 11" wide, 1/2" web thickness and 3/4" flange thickness weighs about 500 lbs. I think I guesstimated an additional 850 lbs of steel and a yard of concrete. I probably should check current prices and figure a little closer before shooting off my mouth, but I don't think I'm WAY off...unless steel has doubled or something... Don't forget, you can do a lot a material saving with the right design.

CMF, you certainly have a lot of neat stuff, transportation-wise, in Seattle! BTW You guys make me realize that I better get to designing a beautiful support column one of these days.

Andrew F said...

Sorry Dan, I re-read my comment and perhaps I was unclear. I meant to say that if the cost of cantilevering is $40k per mile, then I would consider it to not be materially different in terms of system cost, ie, go for it.

I definitely agree, though, that the supports have to be attractive. I'm not sure that's something you have to worry about at this stage, except where attempts to make the supports more attractive impact the design of the rest of the system.

I've been thinking about how to integrate the track/supports with the existing street clutter we have. In particular, I'm thinking about electricity and telephone poles/lines (which can be spectacularly ugly--it's just that they are so ubiquitous no one notices) and streetlights. I know you posted on the streetlighting issue before.

I have to say, I'm not entirely crazy about your idea of including utilities within the track structure itself. Seems to me that this adds complication where it's not necessary (and I'd think it'd be worth it to avoid running high voltage wire in the same structure as our steel track). As you noted earlier, streetlights are a challenge because we'd want to avoid passing vehicles casting large shadows on the ground. Not sure of the best way to address this: your innovation of a more linear streetlight along the track or to further cantilever the light over the track so the vehicle would pass outside of the light cone from the light.

I imagine that supports will need to be beefier than the standard issue suburban concrete light/utility pole that is used in my suburban neighbourhood. But if we can combine these applications, we can minimize the increase in clutter at ground level and help to reduce visual clutter overhead.

In terms of aesthetic impact, I think of the streetcars that are common in the central part of Toronto. The overhead catenary is lower to the ground and quite possibly more visually intrusive than a PRT track would be. It's heartening that the visual impact isn't a deal breaker for the large expansion in the streetcar network here. I only wonder what we could do if we instead spent the upwards of $10 billion in planned streetcar infrastructure on PRT in terms of system coverage, average speed, and ongoing operating subsidy. The funny thing is that the streetcar network is a cost compromise here. Toronto is a subway-or-nuttin' kind of town, and our transit operator insists on designing subway to a very high standard, resulting in costs of upwards of $300 million/km ($500 million/mile). So, our streetcar expansion is a 'bargain' at only $150 million per mile (for no average speed increase over existing bus service). Ouch!

Anonymous said...

You mentioned half-sided track designs:

"It should be noted that only one half of this book-matched arrangement is theoretically needed, except for places where tracks are merging and diverging, though leaving the redundant half in has advantages (load distribution, simplicity of symmetry) which I believe out-weigh the disadvantages."

I believe our (BM Design) design is the only one so far actually presenting such a semi-track design. We intend to default on the semi-side track and use both sides only for non-junction faster speed sections (if such are required in a network).

I made renderings public at SlideShare so you can have a look:

Note that unlike our patent application, these pictures are not expected to be physically flawless. In particular, the grabber arms must extend longer, reaching below the support track. Just wanted to mention that.

While we (and myself personally) embrace the open source approach, we don't intend to open source the whole design. We may open source parts and especially on the software side. We intend to be pragmatic as to using the business approach that brings PRT fastest to everyday use.

qt said...

Good post, Dan. You need to repeat yourself in some detail from time to time (a serious problem with the blog format, when you're discussing a more-or-less unified topic).

With that in mind, you might want to find a place to put some kind of summary document. Update it from time to time if you like, but your readers need it--and not just the brand new ones. The rest of us need the reminder sometimes...

Andrew F said...


Thank you for the visualizations! I can't say I'm crazy about your system design, though. The vehicles don't look like they'd past muster in terms of accessibility, and I imagine the track would still face issues with ice buildup in severe winter weather and noise from wheels running on exposed track.

Anonymous said...

Hi, Andrew.

Our vehicles will fit 2-3 persons and you are right, no room for wheel chair. We cope with that by having a separate kind, higher, on-call vehicle for wheel chair passengers. Whether a certain local track will have those is an option. Some customer cases will require it, others don't.

In my understanding, it's not enough to simply have room for a wheelchair. The chair also needs to be firmly attached to the cabin, otherwise it may topple. There are elaborate safety belts in buses for this reason (I live in Finland, we're very safety concerned here). The trouble is, these require someone else to operate them. Our goal is to allow independence also to wheelchair passengers.

Winter conditions are rough enough here for testing. I don't expect severe problems in that front, though. There are absolutely no protrusions in the track, which allows rain, snow and sand to pass through without accumulating. The most friction is needed near station stops (acceleration and deceleration) where we can heat up the track if required. The trickies part is sub-zero rain which happens only a few times a year. Desert conditions will naturally not require that kind of precaution.

( We should probably open up some discussion area of our own in the near future. Dan, thanks for letting me use this space for now. )

Dan said...

Pretty interesting stuff, Akauppi. I like the switching scheme especially as it might apply to dual-mode. Systems like RUF and TriTrack (I believe) need to get off the rail to switch. Also, your very light track reminds me to address the issue of how I could get mine smaller where many smaller supports can be used. I don’t think it is always going to be desirable to use track that is sized to make 90m plus spans unsupported, as I just mentioned in my post. There’s nothing like skinny track to make PRT look like it will fit in anywhere. That being said, you are lucky than an “artist’s conception” of anything futuristic is generally held at least partially exempt from the laws of physics. You’re pushing the limits of material science pretty hard if these things have any speed, power or weight at all. Nice renderings!

Resume Writing Service said...

It is glad to see this blog, it is good and detailed fun to read this, nice informative

blog, I really like this blog, It's always nice when you can not only be informed, but also

get knowledge, from these type of blog, nice entry. Thanks