DARPA orders miracle motor for its flying car Intriguing news today on the flying-car front - and indeed many other fronts. A Californian space-rocket company says it has received US government funding to develop a miraculous engine that would offer as much power-for-weight as a gas turbine, but would be much cheaper and simpler to make and maintain. SR/C concept for …
Perhaps the most damning thing about the grant is that it is only for 1 million. A couple of years ago the combined F1 circus would burn though that in less than a morning's work on engine research. Rather than pissing away trivial amounts of money on such hopeless odds, perhaps DARPA would be better served by opening up the technology and convincing the FIA to change the rules to allow it to be used.
Unfortunately F1 engine research is pretty limited these days. The specs were frozen a few years ago at 2.4l V8s, no forced induction and all sorts of other boring stuff. And they're not allowed to improve the power output of what they've got. And any changes for reliability have to be approved by all the other teams. There's a new spec on the cards for 2013 but it looks like 1.6l four cylinder engines with turbochargers and hybrid drivetrains.
Sad to say that established motorsport is not the place to look for radical new engine technologies. There are a few electric formulae around, but the biggest change in the "mainstream" in recent years has been diesel engines in sports cars.
I think the trouble is that engine technologies are hugely expensive to develop, while things like drivetrains (4x4 rally cars) and aerodynamics (F1) are rather cheaper. So things like wankel engines and turbines just get banned outright, if not explicitly then by specifying piston engines.
Check your history there... A Deltic is (was?) a reciprocating engine. Sure, it was a weird one, but reciprocating nonetheless. (triangular groups of cylinders each with two pistons and no head driving three geared-together crankshafts. that sort of weird. Wikipedia has a good description.)
"it's a fairly trivial technical exercise to make light aircraft that can take off, fly about, land themselves and even interface with air-traffic control systems without human input."
Really.
It's that trivial that we haven't yet managed it with a vehicle that basically only travels in 2 dimensions on specially prepared strips of land where all said vehicles are going the same way.
But doing it with 3 dimensions, no specially prepared substrate, harsher environment (ie turbulence), etc... is trivial.
Keep drinking the coolaid.
No pedestrians, cyclists, lamp-posts, traffic lights or other stationary objects to contend with in the air, as long as you maintain a reasonable altitude.
Routes from A-B are simpler (even straight, as the crow flies) without a road to adhere to.
Interaction with the ground is usually within a clearly defined area and under fairly strict control. Also there's a third dimension to escape into when collision looms.
The hardest parts of flying are working out where you are (easy with GPS) and landing safely.
All very valid points, however, you're forgetting tiny & looming issue: Idiots. The reason flying is so much more safer than driving is that most pilots are extensively trained, whereas any fool with a room temp IQ can get a license to drive 2 tons of 2D transportation. While I would be the to stand in line for a flying car, the notion of people texting, drinking or flicking boogers on to the windscreen of the aircraft behind, for tail-gating them suddenly snaps me back to reality.
What's really sobering is the notion of an older person (such as myself) getting behind the controls and flying the absolute minimum airspeed to keep the vehicle airborn, just to piss the people off behind me.
In case you didn't realise many airliners are probably flying themselves most of the time, and an awful lot of planes can land themselves without human input.
Just because the hectic conditions of road traffic are tough for robots doesn't mean the same if true for flying.
"It's that trivial that we haven't yet managed it with a vehicle that basically only travels in 2 dimensions on specially prepared strips of land where all said vehicles are going the same way."
Actually, Google has managed it, in California of all places -- which speaks to the next point.
With cars, you are restricted essentially to not two dimensions, but one half-dimension: the road, with very limited options for changing or reversing direction. Then you're further constrained by the most difficult part for autonomous cars: traffic. Your tiny half-dimension is populated by other vehicles going different speeds which often move irrationally (sometimes to the point of NOT going the same direction.) Add to that the inconsistencies of road traffic management signs and symbols (especially across different countries), and you have a much more complex problem than that of autonomous flight.
With aircraft, you have three complete dimensions, global mapping and GPS technology, not to mention a relatively consistent system of identifying and communicating with traffic which has been in constant use and improved for over a half century.
The US military has had autonomous and semi-autonomous drones for YEARS now. It's not just a trivial exercise, it's a fact.
Oh, and obligatory pedantry: Kool-Aid is a brand name, spelled as shown and is a registered trademark of Kraft Foods, Inc.
Yes, it is in fact much easier. You see, in your 2-dimensional word, you only have to fail to control things properly for maybe 100ms and you have crashed. You are also maybe 2.5M wide, but the lane you are travelling in is only 3M wide, with objects just outside this space.
Now in 3-D (apart from landing), you have several hundreds (thousands) of meters in each direction, precision is not really required, fast response neither (unless your aircraft is unstable), and the closest objects are a very very long way away.
Once everyone and their dog has a flying car, and there are hundreds all travelling in different directions between lots of high buildings; then it will be harder. (Although probably before then there will be an agreement that all of them have a common type of control system with co-operating control)
Coolaid is not required (unless you want to see practical flying cars now)
I love these comments. It is always a pleasure when somebody reads something they were unaware of and do not understand, feel that it can't be right, disregards the fact that what they feel doesn't really have any impact on reality, and finally, without hesitation, doubt and any humiliation whatsoever, explain to the rest of us why the world isn't what it is.
It is trivial, and it is even more trivial with ground vehicles. However, the common limitation is that it requires that they navigate in an environment with nearly no obstacles, and absolutely no moving obstacles that isn't clearly broadcasting it's position. It is of course very rare that cars can operate in areas which are completely flat, where all other traffic is broadcasting it's position and the safety distance to said traffic has to be at all times several kilometers (or something like that). For air traffic this is fundamental.
Anyway disregard all this, the fact that most passenger planes today do not really need a pilot and we have drones capable of flying over today's battlefield without human input should at least make you wonder that maybe, just maybe, there is something you have missed.
IIRC there's fully automated shipping container carriers carting around in the Rotterdam port guided by below-the-surface electrickery and some steering electronics, been there for years. Works fine with the caveat that humans are not allowed to walk around because it might confuse the carts and you don't want a fully automated lorry with a full shipping container to run you over. But that's perhaps 20+ years old technology now.
And of course, various virtual road train and self-steering research projects have been going on for so long I was actually a bit surprised it's taken them so long to get it down to family saloon car size. In fact I get the feeling the only reason google did it in cars was because it's now well-known that flying drones can do it and shoot rockets at beardy turbans at the same time. Besides, plenty of aircraft are flying under "full authority" electronic systems that will intervene right so when the puny human fscks up.
In that sense the main thing holding it all back is public perception, ie. human psychology. Of course, you need a lot of testing and showing and so on. And even so I still would not be comfortable hopping in a fully automated taxi and have it cart me around New York. Or anywhere else for that matter. But technically it's probably entirely possible both on the road and in the air. If not now then within a very few years. Whether it actually will go mainstream is another thing entirely.
From the description, that seems most likely.
Mazda never took the production Wankel much beyond the dual-rotor gasoline-powered version but, methinks that there is more potential in that line of design. Pressure seals were always difficult and would be no less so in diesel at higher RPMs, but who knows. if one has gobs of money to throw at materials engineering, almost anything is possible, if not practical for mere mortals.
The ROTAPOWER ENGINE TECHNOLOGY from MOLLER. They have been trying to build a flying car years (decades). They advertise it as...
It improves the Wankel-type engine performance
Maximum power with minimum weight and size
Ability to operate on any fuel
Sounds line the same technology for the same purpose.
So, no patents then?
Combined with the derisory amount of the grant, I'd suspect their design is on the back of a napkin, which is apt since they'll blow through $1 million just on "working lunches" to give status updates on their fantasy engine.
It's described in WIPO patents:
WO/2010/047960A2 http://www.freepatentsonline.com/WO2010047960A2.html
WO/2010/047962A2 http://www.freepatentsonline.com/WO2010047962A2.html
WO/2010/047961A2 http://www.freepatentsonline.com/WO2010047961A2.html
WO/2010/042693A2 http://www.freepatentsonline.com/WO2010042693A2.html
WO/2010/042692A2 http://www.freepatentsonline.com/WO2010042692A2.html
Haven't actually read all that legalese, but from a cursory scan of the images it pretty much looks like a Wankel engine. Could be that I'm wrong, of course.
Not being a mechanical engineer I might be entirely deluded in wondering if a more oval instead of "cilindrical" rotor would be attainable and if so whether that would bring any advantages, like smaller end seals. At least with modern computer modeling an optimal shape migth be modeled.
Tangentially, I don't oppose at all the notion of alternative motor principles. In fact, the early experiments, like some of the advances in rotating radial piston, "real" rotary engines, often floundered on inadequate materials and poor machining. And that was back around the turn of the previous century. We've come a long way. Why stick with boring old piston engines anyway?
Chrysler had a multi-chamber rotary like a lumpy Wankel in the 70s. Earlier commenter is right about RR diesel. Cessna also studied Wankels for light aircraft last century.
Tip seals able to take higher pressures and temperatures seemed to be the big problem for all of them. Several projects are privately funded to build very light high performance diesels for aircraft.
One claims to have the power/weight ratio of a Merlin.
Turbines were meant to be become "cheaper" using preformed ceramic blades or complete fans in 1980s. Not heard anything further.
Now add this donk to the Triplex, cross it with a Reaper and hello Skynet killer drone.
Are you denying that, at the time, a world spanning damage-resistant computer network allowing nigh-instant transfer of small amounts of data and rapid transfer of very large amounts of data was considered an utterly ridiculous idea? Possibly even an impossible idea, there's no way you could get that many computers on a single network all cooperating with each other to get data around. Right?
Okay, so DARPA projects typically sound pretty insane and impossible at the time. And sometimes they work out.
In regards to autonomous ground vehicles, the biggest obstacle to that remains us humans. Pedestrians, cyclists, human controlled cars are all hugely unpredictable. In an autonomous aircraft future, there are no pedestrians (Humans haven't learned to levitate themselves yet), cyclists or even non-automated vehicles in the airspace these automated vehicles will be using. That simplifies things a lot.
So far, it seems that everyone who's tried to build a 'flying car' has started with a regular car and attempted to make it fly, usually without success. This makes me wonder if the entire approach to the problem is flawed. It might be more fruitful to start with a small plane and improve its driving capability. Already flies well, and can taxi around on the ground, so it wouldn't be adding entirely new abilities.
First FI meant June *2009* so this has been in the works for a bit.
It also mentions the actual *work* is subcontracted to Andrews Aerospace. Previously known for their championing of the *seriously* bonkers liquid Hydrogen driven air liquification and separation technology they seem to have transformed themselves into a jobbing R&D outfit.
So P&W R will probably slice a chunk of cash off the top for "Contract management" and hand the rest to Andrews to get on with things.
I agree it did sound *remarkably* like the Voller Wankel concept, which sort of vindicates Voller's view all along.
Wankel's seem to have material issues. IIRC the key enabler for the Mazda engines was the ability to flame spray a layer on the corners of the rotor which was both very hard and very temperature resistant.
State of the art for high speed bearings would be the the foil air bearings used for aircraft aid conditioning (so called air cycle machines). These run to 100k RPM for years with no wear.
No current model flying use ball bearings any more.
The gearbox issue *might* be more of a problem. USAF has sponsored improved tooth design work which might work and roll forming puts the teeth under compressive stress (always a good idea). For *very* high ratio gear down some kind of epicyclic design might be better.
Just a thought.
DARPA is an interesting department. It's DoD (Department of Defense) official risk taking group. They specialize in high risk / high reward ventures that require a little money to see results. For instance, DARPA runs its annual competition for the best automated full size vehicle. Teams bring prototypes to run an obstacle course designed by DARPA and a prize is awarded to the team able to complete the course in the least amount of time. For only a couple million dollars US (chump change compared to what our Dept. of Human Services spends on an hourly basis) they've managed push major advances in automated navigation and obstacle avoidance.
Projects like Land Warrior may not have panned out as expected but components developed in the program are either deployed already or about to be. Stealth started life as a DARPA project though later the US Air Force picked up on the idea.
Sadly (or ironically depending on your point of view) I suspect DARPA is one of our government's better investments.
And if anyone can do making this engine work, I'd bet on P&W. They have a little bit of experience in making high performance aircraft engines. They've only been doing it since 1925.
Let's see.... the SR71 ran J58's, C141's and the E3A used TF33's, the Navy used J52s in the A4's and A6s for almost 30 years (only recently retiring the A6 fleet) and that doesn't count all the radial engines used by the military and civilian airliners over the years. I'd say they've done this a time or two.
Can't wait to see what this little monster looks like. A light weight rotary diesel? Wonder how long it'll be before you see them at surplus auctions.
Was the power to weight ratio is 2HP/Lb.
In perspective this is about 2x the best of a high performance motorbike engine. Presumably with a target of similar maintenance and life expectancy.
Challenging but *should* be within the knowledge base of the companies involved.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
