Hmmm. Another piece of technology which increases the cost when buying the car, increases the cost when you need to replace it, for a marginal increase in efficiency (up to 10% -- more likely to be ~2%, and on urban routes, I'm not betting on those gains on the highway).
Just like xenon headlights, huge starter engines for fast start-stop times, complex injection systems, etc, these increase the costs of building and maintenance, for little benefit, and huge profits for the auto industry.
When do we get mass-produced electric engines already?! They're not as complex as Otto engines (thus lower maintenance costs), so when do we start producing them in large scale to get lower prices?
I don't believe the 10% number. I would think energy lost to vibration and lateral motion would be negligible compared to air resistance, tire friction, etc.
Only problem is hydraulics aren't very efficient and waste much heat. They claimed a recovery of 1 kilowatt, which is just over one horsepower (746 watts by definition.)
Now if that vehicle has a 200 horsepower gas engine and is boosted by a 1 horsepower, a 10% claim might be stretching it. An older generation Toyota Prius has a 10 horsepower electric motor, which can propel it over 30 mph. One horsepower might be good for about 5 mph. I suppose there might be unusual driving conditions that might realize a 10% increase in mpg, but not in our world.
The average car needs ~25HP to maintain highway speeds. So 1hp could be ~4% of the car's energy needs. It takes a lot of energy and HP to accelerate quickly but you don't spend a lot of time accelerating like that. At lower speeds over bumpy roads I guess 10% might be possible.
You'd think there would be some back of the envelope physics we could do. Anyone want to jump in?
Here's a start .. maybe
A bumpy road only uses extra energy because the bumps cause the vehicle to move upwards at the expense of its forward momentum. This invention would in the best case be able to recapture that energy and thus simulate being on a perfectly smooth road. So the most power it could produce would be the potential energy gained by raising a vehicle N cm off the ground every M seconds. N, and M would be affected by speed, and road bumpiness.
Continuing on this line of though, say there's a 0 incline fairly bumpy road where on average there's 3 3cm bumps every meter and the vehicle is traveling 60km/h (~17m/s). So the car has to provide energy to raise the car 9cm for every meter traveled. That's 1.5m/s vertically. So for an average car of 2000kg, we have to provide 2000x9.8x1.5 J per second, which is ~29KW. (That's gravitational potential energy mass x grav. x h)
Wow, so maybe it's right. Or my math is wrong somewhere, or my hypothetical road is way to bumpy?
Shock absorbers prevent you from lifting the full weight of the car for each bump. So replace your 2000kg with the un sprung weight of the car say ~200kg. (Your tires also act as mini shock absorbers so you might need to subtract 1mm from the size of each of those bumps.)
This is the same Zack Anderson that built the warcart (http://web.mit.edu/zacka/www/warcart.html) and created the fiasco with the charlie tickets. He is quite an impressive hacker.
I remember tossing this idea at a friend, and being told that someone was probably already working on it. I didn't have the drive and/or to follow up on it anyway, but it's nice seeing a good idea being executed.
The article neglected to mention the similar Bose [Amar G.; MIT 1951] system, which also generates electricity: http://www.bose.com/learning/project_sound/bose_suspension.j...