Sunday, November 13, 2011

131> Climbing a Chain


Suspended PRT has the advantage (over bottom supported systems) of being well suited for very steep slopes.  As I have pointed out in previous posts, I believe this attribute is essential to achieving flexibility in routing and station placement because it enables a completely three-dimensional transportation solution.  Such flexibility could be important for keeping costs down and increasing ridership.  I do not think that starting with the supposition that each station must handle a minimum of many dozens of passengers per hour gives the kind of flexibility that cities need.  Such pedestrian-rich station locations may not be close enough to each other to make a walkable grid of coverage and getting to such strategically placed stations might well involve more car travel than the PRT trip is likely to save.  This makes dirt-cheap stations very important.  Bare-bones ground level boarding, say, at bus stops, is an option that is out of the question with most supported systems because the ramps would be in the way.

The picture above shows a minimal station in a park. While it clearly is a low throughput design, it also would not require a lot of passengers to pay off.  Small neighborhood parks are very common in the US, often created by the real estate developers to help attract home buyers to a subdivision.  Better still, they are usually located at or near a subdivision’s entrance.  Such small neighborhood stations could exist along a routes that are important but aren’t part of a grid, say from an airport to a downtown area, or between two metropolitan areas that have grown together.  Once residents realize the benefit of such an access point and station’s capacity becomes a problem the station could be upgraded, this time based on real-world ridership numbers.   The foot-in-the-door approach to eco-friendly commuting!

This (more multi-axis) approach should also allow options that will help in the “not-in-my-front-yard!” confrontations that any new transportation infrastructure is sure to create.  For example, it enables the track to be run at a much higher altitude than systems that rely on elevators or gradual ramps.  Instead of a “take it or leave it” approach, there is the option of running the track above the tree tops if necessary.  

None of this, however, is possible without a straight-forward and inexpensive way to engineer this capability into the system.  


Here is such a method.  I have illustrated it with only the relevant wheels and engageable surfaces of the track shown.  The last post shows the track and vehicle in more detail.  (To any newbies to the site out there, I advocate direct-drive “hub motors” which rotate around electrically fed axles, so actually the whole drive system is shown!)  This design is fashioned after the rack and pinion system used by cog railways.  The trick is to make it engageable and disengageable without complicated mechanisms that would drive up costs and reduce reliability.  It uses two stretched lengths of modified roller chain as a rack, and a sprocket that is of much reduced diameter (compared to the wheels) to eliminate the need for a low gear for slow but powerful climbing.  This method adds next to nothing to the costs or complexity of the system.  Such custom links as I use here (and, indeed, whole specialty chains) are widely available because they are commonly used for material handling tasks in industrial production lines.  Here is how it engages, step-by-step.  


The forward wheels move off of the flat running surface, so they are supported by the flanges alone, which rotate separately and freely from the motorized, rubber-clad drive wheels.  For a moment the front wheels can then spin freely as the bogie is being pushed forward by the rear wheels alone.  The motor controller then increases the rotational speed of the front wheels to compensate for the smaller diameter sprocket.  (The front and rear wheels must now be synchronized at two different speeds.)  As the sprocket engages the chain there is near certainty that the teeth will not be in the exact position to mesh exactly.  To adjust for this, there is a compressible rubber cushion (shown in green) which allows the chain to “stretch.”  Additionally, it is possible to design some limited rotational play into the sprocket hub.  Since we are talking about roller chain here, any occasional slipping will not “grind the gears” causing excessive wear as would be the case in a typical rack and pinion.  Keep in mind that this transition is not expected to be taken at anything approaching normal operating speed.  Once the chains mesh with the front sprockets, engagement is kept captive by an upper bar which presses on a freely rotatable ring (blue) mounted on the wheel’s axle.  This pushes the sprocket teeth into the chain and holds them from slipping out.  Although this may seem redundant because the middle, hold-down wheels do the same thing, (push down) I added this component because I was worried that the compression of the rubber tire would allow the sprocket teeth to slip, and I prefer that rubber because of noise and vibration dampening at high speeds.  Anyway, at this point none of the track’s normal wheel support surfaces are necessary for the front wheels to pull or stay centered.  With the front wheels engaged, the rear wheels will follow a similar sequence but without any question about meshing, since the distance between wheels can be precisely set to match the chain.  Of course the whole process can happen in reverse as well, to go from chain back to ordinary track .

Some readers might be concerned that the chain and sprocket are not strong enough to be used in this fashion, especially since normally at least a half a dozen teeth of each sprocket engage a chain. Each sprocket tooth, however, will have a test strength of at least 3500 lbs. The test strength of chains of the size pictured is at least 12,500 lbs. each.

Footnotes… The design as shown is based on four drive wheels, with the hold-down wheels being free-turning. This is a somewhat arbitrary decision. (The hold-down wheels make the system “half-track capable,” meaning the vehicle can travel on the either its left or right wheels alone without possibly twisting inside the track from lack of support – an essential attribute for switching tracks without the tracks themselves having moveable parts.) These hold-down wheels could just as well share propulsion and braking duties, and so could also be configured with additional sprockets and chains themselves.

Finally, A personal opinion:  I think PRT designers should take a page from the late Steve Jobs. A lot of what he did was to take existing ideas and products and redefine them by engineering them so they were a joy to the senses. He had zero tolerance for the slightly funky design choices that help rush products into the marketplace, only to limit their acceptance in the long run.  When I look at the current PRT offerings out there I see such design compromises by the boatload. We, at least in the developed world, will never fully embrace any system that is clunky, bumpy, noisy or slow! We must recognize that the bar has been set very high by a hundred years of automobile engineering and try our best not to disappoint. But then again even a Ferrari can’t rescue you from the congested city streets by going straight up!  


8 comments:

Juho Laatu said...

I agree that ability to build cheap tracks and cheap stations in all possible evironments is a key requirement. Otherwise we would unnecessarily limit the scope of use of the PRT system. (Private homes, forest roads, already built areas etc. to be covered.)

One alternative to this design is to build the "elevator" mechanism as part of the track (instead of part of the bogie). One could connect the bogie to a local moving element (bogie style or something simpler) that woud first connect to the bogie and then lead it up or down the steep section. That would keep the bogie simpler an could allow different implementations of the actual "elevator" mechanism. Only the connectig element would be standardized (like between railway cars, or simpler).

(In the picture the bend from the level section to the steep section seems to be too tight fro the bogie.)

You may need some fence around the "landing area" if the system is fully automated. Or maybe there is just an emergency brake for the passengers if the track is not clear.

Andrew F said...

So Dan, I take it that you're saying that having two chains with two wheels each is enough redundancy in the event that there is a defect or failure in one of the sprockets or chains. I suspect that you might need some additional independent system to brake the vehicle in the event of failure. What happens, for instance, if the vehicle loses power? The motors would just spin and the vehicle would more or less freefall, no? So, maybe you would need to add some kind of ratchet mechanism as well. Or would the emergency brakes that clamp the track be sufficient to arrest the vehicle and support its weight?

What kind of speeds are you envisioning for vertical ascent/descent? Honestly, the vehicle could well be capable of higher speeds than most users would adopt--most elevators don't have windows for a reason. You might be able to address this by frosting the windows (this technology is used on the 787 Dreamliner rather than shades). I think in the past you envisaged having guideways climb buildings on the exterior of the structure in order to reach stations on, say the 100th floor. I can see some users getting a bit nervous if the vehicle is more than 40 or 50 feet above the ground.

Dan said...

Dan the Blogger responds - in typical, long-winded fashion...

I see little benefit to putting machinery into the track at this point. (Although I was thinking about it way back in post 16) Adding four sprockets, four bearings and four rings to each vehicle is next to nothing – probably cheaper than adding a coupler. Not that it wouldn’t be a simple matter mechanically, and I would not want to “design in” something that would preclude it. It’s just that the need for such a system strikes me as being application-specific, and I’m sort of adding a bare-bones generic capability at this point. I am especially interested in keeping station costs down, which weighs against that option. Of course “not designing something out” actually means that you have to “design it in” just to check if you’re designing it out!

As far as fencing goes, that is something I need to figure out. Obviously sensors are just getting better and better, and many are so cheap that there could be a lot of redundancy… camera range finders, infrared and broken beam sensors are examples. There is also this emergence of QR codes. Paint one of those on the landing spot, and if there is a code error, switch to plan B, whatever that is. One idea uses the security cameras. Could an anomaly trigger a real-time visual for some guy looking at screens in a control room? Does “plan B” need human judgment? (If there are multiple, alternative docking bays, maybe not, unless there are multiple aborted dockings) Also the tilt sensors (for self-leveling the cabin) would pick up physical contact before anyone would get squashed. And there are warning sounds… I have even thought of landing on a collapsible platform or simply using weight sensors on the landing platform to detect obstacles. I am most worried about pranksters and false positives…and the “Plan B.”

As far as the bogie getting squeezed, it is true that the inner dimensions of the track need to change for curves – especially tight ones. This is true for both side and vertical turns. The track ribs will have slight variations and the steel fabricators will need to use some jigs and spacers to make sure they don’t goof up. Part of why I have eliminated angle iron in favor of round stock has to do with the fabrication procedures resulting from this issue.

Dan said...

Andrew, you are certainly right about the need for a braking mechanism. My working assumption is that each vehicle would have a truck-style brake where it automatically engages if there is no power. I am still working out the details. Power, though, would still come from the battery in an outage, and there should be plenty. Also, do not assume that the motors would be free turning without power. Low RPM, permanent magnet motors generally exhibit extremely strong “cogging” between the magnets and the steel in the coils, and circuits can be added that mean that any rotation feeds the electromagnets to further freeze the motor. This might not brake a falling bogey completely, but my guess is that it would slow it a whole lot.

As for speed of ascent, let me point out that I am designing sort of a super-PRT vehicle to see what designs won’t limit future capabilities. I really doubt a system would ever start out with the full functionality and power I have described and therefore I would want the system to work decently with a lower power vehicle. I guess I would start at something like a fast walking speed.(but going straight up) I wouldn’t be trying to climb buildings, at least not very high. I just can’t see needing a PRT vehicle on the upper floors anytime soon. At least I wouldn’t design a vehicle for it. Really fast and powerful climbing capability would be best accomplished with a special track or a PRT elevator with counter weights, I would think. I’m thinking pretty tame stuff here. Keep in mind that with a dry track and rubber wheels, the inclines can be fairly steep even without the chains, so ascent probably would be a bit faster that way. I don’t think 20 seconds or so to get up to the treetops is prohibitively slow, do you?

BTW the test strength I gave for the sprockets is very conservative. If the sprocket teeth are about pinky-finger sized, and they are hardened alloy steel, the working load limit should be over 10,000 lbs. each. A sprocket wouldn’t slip on a single snapped tooth anyway, since there are always a couple of others that are semi-engaged.

Unknown said...

Great Work and nice sharing really. Thanks for this and keep posting ike this useful information. It is really helpful for me.

Gear Hobs

Anonymous said...

What to say i praise of this blog, which contains a lot of amazing information as well as the thoughtful writes.

Chain Sprocket Hobs

Anonymous said...

Great post! The fact that you means someone is reading and liking it! Congrats!That’s great advice.

Chain Sprocket Hobs

tjejojyj said...

Very nice. I was thinking about a rack-and-pinion system myself since I don't see anyway elevated PRT will work without ground level stations.

The only other possibility was an LIM brake on the descent and LIM assist on the ascent.

There are three other suspended PRT designs which have vehicles descending down off the main guideway: SkyTran, Metrino, VubaPRT. The first two have segments of test track and the third say they are on the verge of building one but I've never seen any practical demonstrations.

Here's a video about Vuba as they are the latest to produce anything. (It's always easy in CGI!) https://www.youtube.com/watch?v=630YdXigPws

FWIW this blog recently got a post in the PRT Facebook page: https://www.facebook.com/groups/Personal.Rapid.Transit/permalink/3926553580799723/