Sunday, July 18, 2010

95> Call the Cable Guy!



Whenever I see an artist’s conception of something futuristic I always have to muse at the attributes they give the materials of the future. While up at the cabin, I stumbled on this old magazine. Note the total lack of support for the track of the main vehicle. (I won’t even get into the giant ULTra vehicles)

While this is common practice, it’s pretty tough to compete with designs made out of super alloys straight out of science fiction. In the case of PRT advocacy, the main job is to sell the concept of a whole new infrastructure, overlaid on the existing one. This is obviously much easier to do if it is minimalized. On the other hand, there is also a credibility issue here. If the system you are sold on bears no resemblance to what you are really going to get, how can any further claims by a PRT advocate (or vendor) be trusted?

I have given a lot of thought to PRT trusses, and I just want to make an observation. I have never seen any system meant to carry people that is more than about 35 times as long as it is high. Box beams, I beams, trusses, whatever. They rarely approach this mark and are usually much less. There is some pretty good reading on the subject in the Wikipedia entries on beams, bridges and trusses.

The bottom line is that PRT needs to have as skinny a support structure as possible, much more so than any other application I can think of. In other situations, generally, something needs to be built and the buyer has little choice but to take the recommendations of the architect or engineer. Here they can just decide not to involve themselves with PRT in the first place. 

Next time you pull down your wooden attic stairs or have a wooden stepladder handy, note how they reinforce the steps. You will see that they have a thin metal rod beneath each step that can be tightened, squeezing the step end-to-end. The step then becomes a compression member resting on a tension member, resulting in a very strong “beam”. Tension members (cable) are also employed to great effect on pre and post-tensioned concrete in a very similar way. Below are some drawings of cables that are integrated into a traditional truss. I am quite confident that this type of addition would increase span substantially, beating that 35 to 1 ratio.  

 
Fig. 1 shows geometry similar to a suspension bridge. Support points can be had by periodic attachment along the length of a stretched cable. In a Suspension bridge secondary cables can hang straight down to support the bridge decking. Suspension bridges must have the ends of the cables be anchored to the ground, however, although multiple spans can attach together instead. Pulling the cable tighter makes the decking arch upward. Figs. 2 and 3 show how cable could be stretched within a truss and terminated in a spool, which could be tightened. Tightening both spools equally (with a BIG torque wrench) would be essential, or the support would be pulled over. Straight runs would need to terminate by ground anchoring, just like a suspension bridge. The cable could also be continuous. (Between ground anchors) Fig. 4 shows how cable-stayed bridges differ from suspension bridges. Instead stretching a cable from two ground anchors and hanging a bridge off of the cable, the cable-stayed design balances cantilevered loads on a support column. At these slight angles there would obviously be a lot of compression put on the truss itself, pushing it toward the support posts. What is interesting to me though, can be seen in FIG. 5. Note that the truss also becomes a tension member, because the cables work to pull the structure in opposite directions in the middle of the span. This gives it characteristics similar to the continuous cable in Fig 3.

Finally, to really confuse the reader, figure 6 is a full crossbreed. Depending on how the cables are tensioned, this can be either a suspension or a cable-stayed design. Confused? I know I am. But I have used similar techniques to make impossibly long and thin unsupported shelves with great success, and even took the sag out of a roofline once. I know it would work to some degree. Also, never underestimate the wisdom and practicality of the farmer. They use cable to keep stuff from sagging all the time, like this irrigation system. 
 One final photo. It is nearly impossible to down a single telephone pole, since they are all cabled together at the top. I have seen them broken in half by trucks, and the snapped pole just hangs there. Such a quality would seem to be ideal for PRT from a safety point of view. By the way, cable is, relatively speaking, dirt-cheap. Each half-inch steel cable has a tensile strength of over 20,000 lbs. It would sure make ME feel better in an earthquake!

16 comments:

Andrew F said...

Confusing is right! I think I might need to read it again later to really understand it.

My question is what your goal is here. You mention that this could let you have spans more than 35 times longer than the truss is high. IIRC, you proposed your guideway to be perhaps a meter in height, so the truss supporting it would likely be at least that, so we're talking about pushing spans past 35 m. Do we really need that? It'd be handy for bridging highways, rail ROWs, and small rivers. Would this allow you to lighten up the truss for the regular portions of the guideway that are more regularly supported? Or be able to carry more weight for a given truss weight/support frequency? Both would be useful, I guess.

I'll read it again later and see if anything else jumps out at me.

qt said...

Andrew:
Part of his goal, from what I'm reading, is to cut down on the number of support columns cluttering up the landscape--one of the objections that is often raised to PRT in any area where aesthetics become important (which is a whole lot of places).

Another advantage, as seen by this over-the-road trucker, is that it makes the use of medians and other major right-of-way real estate more doable. I don't drive a day without seeing signs of multiple "the car/bus/semi crossed the median" incidents. Recent scars on concrete dividers, tire tracks plowing almost the whole width of an Interstate median before (apparently) the out-of-control vehicle stopped. One thing that has often bothered me about people blithely discussing the use of major-highway rights-of-way is the need to protect those inexpensive, lightweight columns PRT people are so proud of from the next runaway 18-wheeler. The medians are usually adequate to protect oncoming traffic, but look how the bridge abutments are reinforced. It's not all to support the weight...

qt said...

I didn't finish before I hit the key--sorry. I meant to conclude as follows:

I've always feared infrastructure planners would insist on much more heavily reinforced (therefore more expensive and more obtrusive) columns at least on major traffic corridors. This way, you'd have to put up fewer of them.

Now.
Dan:
I wondered if you'd considered cable support. I'm not engineer enough to have known if it would work. Good to see the overview.

If you want to take it to a logical wretched-excess conclusion, take a look at http://stringtransport.com/.

I suppose you could hang your truss underneath such a rail. Cute idea, though I don't know if I'd try it. I keep imagining the failure modes...

cmfseattle said...

qt, although there are a lot of things that could only be delivered by multi-ton trucks, i'd like to see PRT utilized for pallet deliveries. http://openprtspecs.blogspot.com/2009/10/chasing-trucks.html

qt said...

cmfseattle:
No argument here. Frankly, pallet PRT appeals to me, too. So do things like MegaRail and other alternative ways of moving truckloads between cities. OTR trucking is just a way of getting a paycheck for me, not a really satisfactory way of moving stuff.

My point in this post was purely from a construction standpoint. Dan's talked about using freeway medians for PRT corridors before, and lots of other people seem to go for that too. I was just pointing out that building a PRT corridor on a major right-of-way won't be as cheap and easy as many advocates think.

Putting something where out-of-control vehicles can slam into it at 60 mph, and then saving money by supporting it with skinny columns--NOT a good idea. And making the columns safe in that environment will add to the cost.

Andrew F was wondering why Dan was interested in increasing spans in any but special environments. I was pointing out one reason to think about that--besides the aesthetics, which Dan did mention in the post.

Longer spans means fewer supports, which decreases cost in any environment that requires extra spending on those supports. And there are quite a few places where you just can't go with minimum-cost supports. If he can get his cable-braced trusses to work at a low enough cost, it might be cheaper than the extra collision protection for the extra columns in a freeway routing.

Andrew F said...

To protect supports in an situation like a highway median, it's not necessary to improve the supports per se. They can be protected by bollards or deflectors that are not part of the actual PRT structure. Of course, these things still do cost, so of course it would be beneficial to reduce the number of supports. Also, most highways have extra room in their right of way, so PRT could be put to one side or the other, rather than down the centre. This way it could also avoid the worst of the overpasses.

About the appeal of 'megarail' style transport for long distance freight, I think it would have to be smaller capacity (than existing 53' trailer) vehicles that operate on a very sparse network of heavier duty guideway. So, each city can have intermodal facilities, and distribution centres on this network to facilitate last mile (either using trucks or breaking the load into smaller vehicles). This actually could have a lot of appeal in the developed world, because Europe has a horribly fragmented freight rail network (it's impossible to move a container using the same locomotive across the continent) and the US has ripped out much of its rail network. But I don't think we could put full-sized sea containers (40') on PRT-style guideway without making it very, very heavy. But such a continental, low-energy, automated, continuously moving freight system would create hundreds of billions of dollars in value and slaughter the intra-continental air freight business. All that is to say, it's very worthwhile to think hard about accommodating these uses in designing a system. And given the heavy-duty needs of a more bulk-freight system, would it necessarily make sense to put both these freight vehicles and inter-city PRT vehicles on the same guideway, especially anywhere the traffic volume warrants separate guideways.

Sorry for the tangent.

Dan said...

Dan the Blogger Responds...

Gentlemen, thanks for the posts. Even if I don’t have time to reply right away, I enjoy hearing your thoughts, as do the other readers, I’m sure. About my intention in bringing up the cables… I look at it like this: It’s sort of like doing a mural and talking about it, and taking in ideas, and you have this big can of a color you never showed anybody. Who knows how it could be used? Sometimes it’s worth bringing stuff up just to be on the same page. I guess it could be useful in making the spans longer, or the track skinnier, or cheaper or safer, or some combination.

Cmfseattle, you know I’m in agreement on the freight thing. I would particularly love to see some open-source PRT software prove itself in some sort of private sector warehousing or distribution, or even airline baggage handling. I have a 3D SketchUp model of a warehouse/industrial track intersection with a hundred different possible routes. Ten in and ten out, if I recall correctly. I was saving it for someday when I have writer’s block.

I’ve been doing some thinking about the size of the SMART specification and it does worry me that it is as thick as it is. I think it’s a few inches bigger than Skyweb’s and I wish it were a few inches smaller. Of course it would be a full eight feet higher, so it would be that much further away. The whole freight discussion also plays right into the subject. The advantages of longer wearing, slower rotating wheels (that have largely created the track size recommendation) also mean that the track could then take more weight. Actually the internal dimensions would be fine for heavy freight, right up to half-sized shipping containers. All that would be needed are frequent supports.

Qt, one thing about the string transport as well as the stuff I have been showing is the extreme tension on the cables. Unfortunately I really don’t have much experience with very long, tight runs of cable. I can’t help but wonder about expansion/contraction, and stretching. You are right, though. If we were coming down a mountain, supports would be mostly just for controlling wind or other forces that would get the thing rocking.

BTW, Andrew, tangents are always welcome here!

cmfseattle said...

fewer supports would seem to save cost, but maybe not.

Einar Svensson, who did engineering work on the Seattle Monorail guideway, has moved toward longer-span designs
http://urbanaut.com/Guildway%20and%20Construction%204.htm

"The Seattle-ALWEG monorail guide-way beams are supported on piers that feature a design based on mass rather than high-strength materials. The construction drawings for these piers specified the use of relatively low strength 3,750psi concrete and 10,000ksi steel reinforcing bar. These materials are only slightly stronger than what is commonly used in house foundations. Yet, the wisdom of this design strategy has proven itself through 41 years service. The piers have survived several major earthquakes, collisions from vehicles including heavy busses and trucks, and the ravages of time and the environment without failures and with minimal or no maintenance."
http://www.historylink.org/index.cfm?DisplayPage=output.cfm&File_Id=4282

http://www.seattlemonorail.com/history.php


from the small scraps of info i've found, it sounds like the SkyTran/Unimodal system design goes the other way: smaller posts, spaced closer together (30 feet, according to one of the commenters below)
http://www.mv-voice.com/news/show_story.php?id=3083


i think that public entities would be much more interested in a system that generates revenue, from taxes on freight pallets, moving around on their elevated infrastructure. also, freight vehicles should be limited to ~2000 lbs. (including vehicle), so as to be nearly the same weight as a loaded passenger vehicle.

qt said...

Dan:
I wasn't really advocating the StringRail approach (note the "wretched-excess" crack). If it works, fine. But when I imagine failure modes I see explosive breakdown of rails if a cable snaps under that God-awful tension. I don't have to be an engineer to flinch.

Just thought you'd be amused and intrigued. I admit it will be neat if it turns out to be workable...

Andrew F said...

cfm, I think it might be desirable to have the guideway capable of supporting more than 1 tonne vehicles. Even increasing the limit to 2 tonnes would make the system much more useful. And I'm talking about the base network here.

For a heavier-duty freight network, especially one capable of carrying sea containers, we're looking at supporting up to 4 tonnes for the container and another 30-35 tonnes for the cargo. I was looking at the spec for sea containers, and apparently 20' containers are designed to carry the same weight as 40' containers. I suppose this makes sense in terms of sizing cranes, etc. What's probably called for is some statistical analysis of how much of that capacity is used. If only 5% of containers would weigh more than 20 tonnes, then it would probably be justified to limit the weight there, and require anything above that to be broken down (or shipped in 2 20' containers at origin).

An elevated freight network is also appealing in the way it reduces the creation of 'islands' of wilderness that are difficult for wildlife to move between, causing challenges for migratory species or large ranging mammals. If the infrastructure is elevated, there's no need for fences, concrete barriers, etc. I'd imagine there would still be access roads to allow for maintenance, etc. but without the constant stream of traffic at-grade, this should be much less of a barrier.

The other interesting thing that arises from a heavier duty subnetwork would be the potential for roll-on, roll-off car carriers. I don't think it would be wise to allow the passengers to remain in the vehicle, but the ability to bring your car with you on a trip could also be valuable.

Dan said...

I have to agree with cmfseattle on the weight issue. If I recall correctly, Ed Anderson blames the fatally high cost (50 million per mile?) of the Raytheon PRT system on them not keeping the weight down. I believe the going wisdom is that doubling weight capacity tends to cube the cost. I wish I had some links handy to back me up… Also note that the average vehicle occupancy is only 1.2 passengers, so even going with 4 passenger vehicles represents a lot of wasted structural strength and energy moving the oversized vehicles around, made necessary by ADA compliance.

If I am not mistaken, the Skyweb Express track is sized so that it can support fully loaded vehicles bumper to bumper, something that would never happen, but something regulators would surely demand. I also recall something to the effect that such a weight falls below the what pedestrian bridges must carry, thus placing itself outside of the much heavier weight classification normally required for motor vehicles. Anyway, it is my impression that the argument for a heavier, more versatile track is a dangerous trap, at least until PRT gets to compete equally with roads for tax funds. None-the-less, I have endeavored to allow a hierarchy where lighter, smaller vehicles can enter heavier track. I just think the advantages of elevation and automation become less as the weight goes up and the payload is bigger. In a bus, for example, the driving duties require only a single driver for all passengers, a more efficient arrangement than, say, a taxi. The same goes for cargo. Therefore I suggest the most “bang for the buck” is on the lighter end of the spectrum, where tons of metal, plastic and glass replace a 15 lb. bicycle. I agree it would be nice, though.

Somewhat heavier vehicles might be possible by spreading their weight over more track, however. (post 46) But as many alert readers pointed out in that post, there is also the problem of the whole system slowing down for the worst performer. This brings up the question of restricted access, either physically or by designating special times for more cumbersome vehicles.

By the way, “habitat fragmentation” happens to more than wildlife. Freeways can destroy neighborhood “ecosystems” as well. Elevating transportation solves all kinds of problems, as will be evident if we ever get a system built.

akauppi said...

Saw this TED video today and would like to share it with you:

http://www.ted.com/talks/ellen_dunham_jones_retrofitting_suburbia.html

It's about how US suburban way of life can be turned into more local, more beautiful islands of habitat, sharing both work, living and shopping instead of needing to drive miles and miles between such. Gave me hope that the US infra can actually be turned around.

AFTER such a retrofit, PRTs make sense imho. Not before.

Andrew F said...

I saw that video when it was first posted.

I think PRT can help facilitate the process, for that matter. But then, in my neck of the woods, dying malls tend to get built over with condos. I'm beginning to think perhaps Canada and the US are more different than people give them credit.

Dan said...

Thanks Akauppi, for the clip. I'm going to devote an upcoming post to the sprawl spreading issue, but I would make these points in answer to you and Andrew...
First, not all US cities have this problem. Age, growth and layout seem to be the controlling factors. Older, slower growing cities hemmed in by water or mountains fare better, as to cities with tourism.
Secondly, the longer ranging, faster PRT model has much greater environmental benefits per passenger, so I don't want to preclude it by design. The route causes of the sprawl need to be addressed, though.

Bengt Gustafsson said...

Going back to the technical issue I am interested in how a steel cable can be used to reinforce a steel guideway design. Or can it? As far as I know steel does not get stronger by being spun into a wire, so reinforcing a steel girder with a steel wire, does it really make sense? Isn't it just the geometrical positioning of the steel (and the steel quality of course) that makes the difference?


One thing that may differ between a steel wire and a steel beam could be if the wire is made out of austernitic steel (spring steel) and has a higher modulus of elasticity (less elongation at a certain pulling force). I have been told that this is not the case, but it seems logical that the springing back function of spring steel seems to indicate that it is, and if so the reasoning is probably valid, which is really interesting.

Dan said...

Bengt, That’s a great question. Here are some thoughts you and other readers may or may not have considered. First of all, it is not straight, but wound into multiple spirals. As it tightens they squeeze against themselves and try to rotate so as to unwind. In the process it tends to distribute tension between the strands any one of which could be weak. Ever experiment with winding your own string from a cotton ball? It’s pretty impressive how weak cotton fibers get stronger with a few twists. Also, ever note how traditional rope is almost always wound from 3 windings? There is a science to this stuff. Some steel cable winding designs are fairly sophisticated, by the way. I know nothing about how stretchy steel cable is but it must stretch under tension, as the windings compress against themselves trying to unwind.

I am not sure, but I don’t believe steel is very uniform in strength as it comes from the mill. (impurities, crystallization) Roll forming tends to compress and strengthen it, but drawing rod or wire? I don’t know. I have wondered the same thing, particularly in stressed concrete applications, where you’re not going to ever re-torque or wind it. You would think threaded rod would be a lot easier. They used to use turnbuckled rod quite a bit in bridges in the old days. Now you never see it. Anyway, one thing is for sure about cable. It is resilient, even after being bent or unbent, something that cannot be said of solid steel. Its failure mode is completely different from the rest of the truss, something that attracts my attention from a safety point of view.