Wednesday, June 27, 2012

143> Waiting for SkyTran…

Recently several alert readers have called some aspects of my latest designs into question. It is extremely easy, especially with CAD software, to get on a roll with a particular set of beginning assumptions and start working out details from there, never revisiting those initial assumptions. After all, once you go through the trouble of drawing a component, it is very hard to throw it away, so you tend to modify your design to make it work. I think this may be the case with the recent conversion to the “diamond” track.  So I am taking a break to reassess the basics before going back to component design. I need fresh eyes, and to sleep on the problem for a while.

In the meantime  I thought I might share a few observations about another highly ambitious suspended PRT system under development, SkyTran. Naturally I have given the system some thought, since perhaps there is something in the design that would benefit our little project.

Of course SkyTran’s most notable characteristic is the Inductrack technology, which uses magnetic levitation to eliminate the need for wheels, (and the inefficient friction and periodic replacement associated with them.) But before going much further, there is one very important point to keep in mind in regards to wheels… Inflatable tires that are designed to absorb shock (and grab the road to prevent skidding) are extremely inefficient. Hard wheels, such as the steel wheels on a train, offer MUCH less rolling resistance. Perhaps some alert reader out there can quantify this with some numbers, but my gut tells me that the difference between hard wheels and maglev is not much greater than between hard wheels and soft, road-worthy tires. Of course this point is somewhat moot, since our little design requires a certain amount of “give” to the main wheels, so they probably will be of solid rubber. My best guess is that this will put us about halfway between maglev and automotive tires in terms of frictional resistance, and still give us enough cushion for a quiet, smooth ride.  That said, let’s take an “under the hood” look at Inductrack.

With Inductrack, vehicle-mounted permanent magnets pass over magnetic coils in the track to induce current. That electricity, in turn, induces magnetism that will be polarized to repel those permanent magnets, creating lift. While this has been described as “passive” levitation,” this description is a bit misleading. Actually the system is behaving as a linear generator, where it uses the power generated by moving permanent magnets over coils to create lift, instead of harvesting that power for other uses.   

Generators and motors are structurally nearly identical, and the same can be said for linear generators and linear motors. Therefore such a system, with only minor modifications, can be made to propel and brake the vehicle as well.

It sounds too good to be true. All you need to do to make the vehicle “float” is to get it going a few miles per hour, and it takes almost nothing to push a floating object! But there’s a catch… do you see it? The reason this sounds so darn good is because it has the attributes of a perpetual motion machine. It is a generator that feeds a motor that propels a generator that feeds a motor. We all know that perpetual motion is impossible – that energy must be added- and Inductrack is no exception. Inductrack is unique, though, in that levitation can be had from any outside form of propulsion, from jet engines to sliding down a hill. But for PRT, where the applied energy will undoubtedly be electrical, it raises an important question. “How efficient is this Inductrack design compared to other forms of maglev?” After all, we have established that energy needs to be applied to Inductrack to create the propulsion that creates the lift. Anyone who owns a generator knows that the motor works harder under load. In the case of Inductrack, this effect would manifest itself by creating some amount of drag on any source of propulsion, so the vehicle is not really floating freely, but rather encountering resistance because it, too, is “under load”.  Other forms of maglev simply create the lift and propulsion directly. Why should Inductrack be more efficient? Another problem I see is with linear motors in general, where the object being propelled is a big, wind catching object possibly filled with hyperactive children. That is the fact that the greater the gap between the vehicle and the track, the less efficient the linear motor. The motors I have seen need a gap of less than a quarter inch to be reasonably efficient, and half of that to be comparable to a rotary motor. 

Let’s take a brief look at maglev in general, including  SkyTran’s approach. The benefits are obvious. No noise, friction, or moving parts to wear out. The downside is that it puts a bunch of electrical hardware into the track, and that has consequences in terms of how any given transit project is routed and financed. If all of the propulsion hardware is in the vehicle, that investment is put to immediate and nearly continuous use, where it can pay for itself. If the expense is shifted to the track, then that track must be equally in constant and continuous use to achieve a timely payback. This seems problematic to me, because it would tend to inhibit expansion of routes. Or, put another way, is this the most efficient use of copper? In the vehicle (in a rotary motor) the same copper coils are used again and again, with each rotation of the wheels. In the track, between each passing vehicle the copper just sits there, paid for, but unused. The shear mass of the coils is many times what it would be if they were in the vehicles, so it is more expensive, period. More expensive, but, all else being equal, well … better.

Speaking of “All else being equal,” I was going to point out that maglev vehicles, like their LIM powered cousins, would tend to be less maneuverable in close quarters, especially tight vertical turns. Actually, though, I think that problem is surmountable.  This is something I wish the SkyTran people would address. As it stands, The SkyTran’s stations need to be elevated, and that means elevators. True, good maneuverability would complicate their simple design and might make the curved track segments more expensive, but I think the versatility would be worth it.

Maglev, it seems to me, requires more than a single track profile, in terms of how it is wired. Nobody, in this fast-paced world, wants to accelerate at a snail’s pace, and feeder tracks need to be kept short anyway. That means big, beefy acceleration coils in the track around stations. At speeds just under where aerodynamic drag starts to really take its toll, I would expect the track’s coils to be minimal. For very high speeds, air resistance again creates the need for high energy input. I have just mentioned that very tight turns might require special provisions to keep the vehicle centered and moving. (Actually Inductrack comes in two flavors, a high speed and a low speed type, but this still does not address the propulsion issue, at least as far as I can tell.)

Since maglev vehicles are untethered, transfer of power from the track to the vehicle is more problematic. Any onboard battery must presumably be recharged at the station, or additional magnets and coils can be used as a linear generator. (creating more resistance to propulsion) I wonder though. If there was a break in the track, cutting off power, would a SkyTran vehicle be stranded? Or is there a way to hobble to a station or emergency evacuation area on battery power alone? It is not impossible that a maglev system could be entirely battery powered, or battery powered but supplemented where higher power is required, such as acceleration ramps. But again, you can’t just harvest the energy that you are creating. Battery powered maglev would require charging stations or a way to harvest energy from an electrified track through magnetic induction, which sort of defeats the purpose.  

To summarize, it has been said the maglev is the future for transportation, and that may well be the case. SkyTran’s Inductrack technology demonstrates that the amount of power that is required to break the bonds of gravity (and loose the wheels) is not that great. (even the best rare earth magnets moving over coils at only, say, 12 mph represent VERY little generated power, since speed is everything when it comes to the amount of electricity a generator can make) The underlying principals involved in this, and other maglev designs are simple and well understood. Anyone who has ever opened up a typical electric motor has seen the components – copper coils to induce or harvest magnetism, and steel laminations to coax those fields where they are needed most, perhaps some permanent magnets.  We have been using the techniques for the last century. Using it for linear, instead of circular propulsion, and tweaking it to create repelling magnetic fields to create lift, well, that is more recent. Nonetheless, we are basically talking about geometry here. It is the clever arrangement of the magnets, coils and laminations that makes everything possible. Making it application specific – say, for a payload of 2-4 people with expectations of performance similar to their family car – that is, indeed, an undertaking! The people at SkyTran say they have it essentially nailed down, but all complex engineering tasks involve compromise, and we don’t yet know what specific trade-offs they have had to accept.

In the meantime, for what it is worth, our humble little system is suitable for a retrofit, if and when it comes to that point. And a hybrid is not out of the question! After all, their weakness, (low speeds, maneuverability) are our strengths, and our challenges (going through tires because of high speeds, too much machinery packed into a small track, getting silky-smooth high-speed switching) are their strengths. Who knows?


Andrew F said...

Dan, SkyTran seems like vapourware. There has been very little movement from them over the past 5 years.

Relying on battery power seems like a bad design choice. Until batteries get much better, the charge time, weight, and cost dynamics don't really support fully battery powered PRT operation in high traffic applications. We'll see how ULtra does with Amritsar.

Dan said...

Dan the Blogger says...
Hi Andrew - I wouldn't exactly call it "vapourware..." I think we just have to recognize who we are dealing with here. After all, NASA has got researchers. Researchers live for funding to research another day, otherwise they are out of a job. Therefore I think we will continue to see improvements in Inductrack, maybe even a test track... But it will continue to be a science project until someone else wants to commercialize it. One problem with research is that they may make decisions based on the most dramatic results, rather than what is practical outside the laboratory.

I am not sure all battery power is off of the table...ULTra,is, after all, not the most efficient PRT vehicle ever conceived. There are also many ways to recharge "on the go." One idea is to use capacitors to help charge the battery between stations. If a vehicle was getting charged anywhere within, say, 50 m of a station, and it got enough charge to capacitors to get another couple of minutes of run/charge time, real downtime for charging might be kept pretty darn low. Of course I have opted rails with line power, so obviously I think that is best.

cmf-seattle said... looks to be from late 2005; curious, that they'd use a beam-straddling design, but I guess they plan to keep switches to a minimum.

In the Aerospace Corp. PRT design, permanent magnets were in the track and a "rotor" was onboard each vehicle; 600VDC lines would have been inside the guideway.

I wonder if track coils could serve both for vehicle propulsion (and possibly levitation) and guideway structure...basically, your PRT track would be a collection of LSMs.

Dan said...

Dan the Blogger responds -

Thanks, cmf! So that is how General Atomics fits in to all of this…I had heard that name as a company somehow involved, but not much else.

I still question the cost effectiveness of linear motor and/or maglev in general, especially designs with lots of track based hardware and especially vs. hub motors. I just don’t think the rolling resistance is that big of a proportion of total drag - especially at speeds where aerodynamics is a concern. One new factor going for all of these technologies, though, is the advancement of accurate proximity sensors combined with lighting fast servos to keep the air gap between track and vehicle to a minimum. Otherwise that air gap creates inefficiencies that favor rotary motors, which do not suffer this problem.

As for the possibility of making Inductrack’s (or LSM) coils have structural value, insofar as electrical coils can be replaced with laminated structures, it sounds like a remote possibility, although that is where the heat gets dissipated, so it brings uneven expansion and contraction into play, really complicating things.

Bengt said...

Here are some comments regarding rolling resistance and SkyTran:

I have seen rolling resistance numbers such as:

Car tire: 1.5%
Hard plastic: 0.5%
Steel on steel: 0.05%

This is the resistance as a percentage of the borne weight. As a rule of thumb this means that with car tires aerodynamic drag and rolling resistance are about as large at "PRT speeds" of 50 km/h but above this aerodynamic drag dominates.

This means that maglev has no energy advantage over steel wheels at any speed and diminishing advantage over car tires at higher speeds.

Dan said...

Hi Bengt, thanks for dropping in!
I’m not sure I understand how that last assertion can be made without having the numbers (that represent the magnetic equivalent of rolling resistance) for a given maglev configuration. Nonetheless, what you point out about diminishing returns is a really, really relevant point when you start talking about higher speed systems like Skytran. If the breakeven point for inflated rubber tires is only 50 km/h (31mph) then it would seem that the wind resistance would absolutely dwarf rolling resistance for vehicles going much faster, since the wind resistance cubes every time the speed doubles. This would indicate that the energy savings with maglev at say, 150km/h would be almost insignificant compared to a vehicle with semi-hard wheels, as I envision.

HaraldH said...

Hi Dan, is noise pollution a relevant factor in the question maglev VS. wheels? I would think maglev is more quiet?

Dan said...

Harald, I think that noise is certainly a concern. One problem with the diamond track design is that by moving wheels outside of the track we don't get the benefit of the sound shielding that an enclosed track provides. Luckily I don't think that High speed lines are liable to be in quiet neighborhoods, but rather following highways with much louder vehicles. It is a drawback compared to maglev, but it seems like a minor one, all in all. I might add that, but for noise, I would seriously be considering steel wheels, which have much less rolling resistance than what we have been considering. Thanks for your comment!

merlen hogg said...

I loved reading this piece! Well written!

Merlen Hogg