It is a very good and a very sound idea (and not only because I imagined it myself some time ago, only to find others well ahead of me).
As others point out there are fundamental challenges including heat, rpm, incompatible torque/rpm ranges and so on. None of these are insurmountable however and I believe the trade space will result in a worthwhile difference in fuel economy mainly by allowing a smaller engine to propel a given car without loss in power delivered to the wheels.
The KEY thing with this idea is to look at the energy stealers on a car: steering and brake pumps, water pump, oil pump, a/c.
All of these can be electrically powered. In the case of the water pump and oil pump it may be a good idea to leave as is but the brakes and steering waste energy all of the time when they are not even doing anything useful.
It is only in getting rid of these 'drags' (as well as the alternator on the belt) that the turbo alt will payoff.
Alternators running up to 30,000 rpm have been designed. Such is likely possible for an automobile alternator. Getting the turbo output down below 30,000 rpm is needed (and will increase torque which is a good thing). The turbo can be made larger in diameter (lower gas volume rate over the circumference) too (which also increases torque).
Keep at it. This is a good idea. Just wish I was more of a tinkerer.
Taking the high pressure are from the windshield area and venting it through to a low pressure rear vent area will reduce air drag not increase it but you will need a big hole from front to back to pass the air through.
The reason I was thinking wastegate was; with a clutch, you'd still be spinning the turbine, causing some backpressure. Another good thing about this whole idea is that you'd cut some exhaust sound, so you could use a less restrictive exhaust.
Hello, sorry to resurrect this zombie thread but I was googling and had to give my 2 cents :P.
I don't believe the turbo-lag concept is valid in this situation, as you won't be running a heavy air pump. You can make the turbine very light (easy spinning at say 500rps) and spin very fast at high RPMs and accomplish the same goal.
If you worry about heat damage you could always use a Stirling engine attached to the exhaust pipe very close to the engine. Have a very thin piece of aluminum/copper with thermal paste (for efficient energy transfer) and place the cold side of the engine somewhere where it would get fresh air flow. The bit getting fresh air wouldn't have to be very big just a small heat sink; all it has to do is stay cooler than the hot side to work, which isn't hard when you look at how hot exhaust gas is. The beauty of a stirling engine is that it may continue to work log after the engine has been turned off.
Or even mount it on top of the engine and put the cool side in the air intake lol.