Bill Gates Applies for Patent on Electromagnetic Engine
appft1.uspto.gov
appft1.uspto.gov
The patent system is supposed to be a bargain: an inventor gets a limted-term monopoly on an invention, in return for teaching that invention to others. But this invention doesn't teach others how to make the device; to do that you'd need exmpensive man-years of engineering.
Similarly, in 1930, Frank Whittle patented the turbojet. But his patent didn't teach others how to make a turbojet -- how could it when Whittle didn't know himself how to do it, and only had a working engine in 1937?
Probably the patent system is unreformable, but if it is reformable, then one necessary reform would be for any patent application to include a working example of the invention, and all technical drawings and documentation (in computer-readable formats).
It will be interesting to see if the implementation is actually viable (and how efficient it can be), given the power levels required by a typical vehicle.
At first I was pretty skeptical if the usefulness of such a device. The crankshaft in an engine transfers power to a load, but it also performs an elegant and complicated synchronization task for multiple cylinders and ancillary systems (like lubrication, cooling, fuel supply, ...).
But the crank also imposes a practical limit on stroke length because a long stroke in a small space gives you pretty high angles and less efficient power transfer. The short stroke puts an artificially-low limit on the fraction of energy from the combustion that can be recovered.
Attaching a kinetic-to-electric conversion device directly to the cylinder allows the stroke length to be longer (basically unbounded) so a larger fraction of the energy can be captured, and less wasted as heat.
Then a multi-cylinder engine basically becomes a computer program, where it's easy to implement fine timing, synchronization, load-balancing and load-prediction tasks using sensor networks.
I'm skeptical that it would be more useful than just dropping the mechanical conversion and it's relatively high losses, and using something like a fuel cell instead.
I think this would be a logical way of making a car engine more efficient. Fuel cells are cool but the infrastructure already set up to refuel and maintain petrol engines is vast and not ready to go away yet.
And that's precisely why ethanol engines are such a nice idea. I have been driving ethanol-running cars since my youth and can tell you they work just fine and have a pretty low carbon footprint, specially when compared to fossil-fuel based ones.
I sincerely doubt they got their math straight.
(The wikipedia article is kind of awful, but it touches on the main concerns: http://en.wikipedia.org/wiki/Corn_ethanol )
I think I have owned only one car that ran on gasoline, an import from Argentina.
However, I couldn't figure out whether the article was accurate so I decided to post the patent application instead.
http://www.tomshardware.com/gallery/engine,0101-194921-0----...
Since there is no flywheel (or connecting rod) for that matter, they have to power the piston through its intake phase and then induce power from the ignition phase.
I like it. You could make a very small very fast little motor out of it.
edit:
I still think this:
http://www.gizmag.com/go/6245/
and this
http://news.bbc.co.uk/2/hi/technology/5386004.stm
are more they way of the future than a magnet being blasted back and forth inside a tube.
As to http://news.bbc.co.uk/2/hi/technology/5386004.stm you are going to have heat issues because at best you are going to be ~60% efficient which means you are generating a lot more heat than a battery. (Think of the worst laptop battery and then think about it generating 3x as much heat.)
However, that was October 2007, and the search towards alternative energy sources & radically efficient engines, still keep growing. By leaps and bounds these days.
This PATENTED innovation still concerns itself with the "pistons" i.e. an Otto-cycle which is doomed. So it's not much of a real innovation.