Kinetic energy recovery system (speculation)

[bazzawill]

After watching GCN's coverage of the Taipei bike show tech from a mad ebike got me speculating if part of it could be used on a commuter to address my biggest pet peeve when doing so. TRAFFIC LIGHTS
So my idea involves an electric motor hooked up to a capacitor to (possibly) replace the brake (offset it's weight).
The motor would be used in reverse to stop and capture kinetic energy in the capacitor only to be used to assist and quickly accelerate (thanks to the high torque) back up to speed. Now I am probably dreaming as thermodynamics is a bitch but electric motors are getting much better as are high capacity capacitors so could it be possible?
Otherwise we need to build cycle bridges over every intersection when cyclists finally become the master road user

 

[bladderhead]

Everyone knows the advantage hub gears have over Disraeli. Having a motor means you can get away from a standing start when you are stuck in a high gear. Also, imagine hurtling downhill into a steep valley at 40mph and turning it on in dynamo mode and still pedalling. Then you can switch it to motor mode as you go up the other side.


You do not need a capacitor. You could use a flywheel driven through a CVT.

 

[McWheels]

I remember discussing regenerative braking on another forum (Candlepower). I think we're still at least 10 years (prob more) from the benefits trains and cars can achieve. The general point was the weight and efficiency of capturing the KE massively tilts the scales against doing it.

I don't have the perfects maths to hand, but consider this. A 3W hub dyno can produce up to 6W at speed with a well matched load. Power to accelerate a bike is in the 150 to 300W region (2k if you're Chris Hoy!). Best weight so far for the dyno-hub is in the region of ½kilo. Motors aren't as efficient the other way round, but they will typically add 5+ kilos plus battery and accouterments.

KE of a bike plus rider at 20mph, assuming 90kg combined. 20mph = ~9m/s. = 0.5*90*9^2=3645J

1 Joule is 1W for 1 second, so if you can decelerate that in 5 seconds, it's the average power dissipation of 729W/s, and that's not trying too hard!

That's a chunky motor, plus a very impressive capacitor to store that power, since a battery won't eat it that quickly. Best power density I could find was around 5 grams for 1.5 Farads at 5.5v. So..... that makes C=Q/V, and Energy (idealised) = QV = 45Joules. So that needs 81 caps, call it 100 with charging efficiencies = 0.5kg. So actually almost ok. But capturing that energy and keeping it below 5.5v will be tough. If you spike up to even 20v, then suddenly the capacitance density changes dramatically. As in I can't find anything above mF and they're getting big, bigger, BIGGER!!! I've not even considered the charging rates, but let's just accept that for now.

Still, I'm enjoying writing this, so let's keep going. The retardation of a motor in 'Collect' mode isn't at all similar to the propulsive effort when it's in Drive. But let's be daring and suggest anything above 5 mph can be collected. So that's 93% of the KE and 75% of the momentum.

What I think I'm coming to is that steady state collection is a reasonable aspiration, see the igaro 2 for example. But it's not been commercialised yet beyond running gadgets.

A final thought on the subject; many buses use a mechanical flywheel to store the energy of start/stop, so I used an online calculator. I reckoned on a few variables, but a 2.5kg flywheel, 10cm dia (got to package it somehow), needs to spin at 10,000 rpm to hold 3427J of energy if it's all in the rim. For an actual disk, the rpm is now 14,300 for roughly the same KE. Racing cars keep their KERS flywheels lying flat, since any other orientation has massive gyroscopic implications - which sounds fun but also slightly ludicrous on a bike!

 

[bladderhead]

You could put it under the seat. The gyro effect would stabilize the bike. And it is a recumbent bike. It could not look much weirder.

 

[tiltmaniac]

If you want to look at efficiency, you need to consider the whole system, which includes the human (motor).

Our motor is pretty efficient at low power, and much less so at sustained high power.
Which leaves the middle...

So the question of efficiency begs the question of how much "smoothing the spikes" helps.

I don't have an answer, but I can say that starting from a light (as separate from stopping or waiting) is one of the more annoying things I have to regularly do on my commute.

 

[bladderhead]

starting from a light

Bloody YES! There is a light at the top of a steep hill. Hardly any driver in my area is a speed-freak. At this point I wish they were. Green, and the driver slams on the gas, squeezes past, and then their foot moves from accelerator to brake. So I have to brake. One of the few places where I can outrun them, if they would only bloody let me. Sometimes I actually manage to overtake a car. I wonder what they think as I edge past.

 

[McWheels]

I think there might be a capacitor answer to acceleration only, since they absorb and deliver power very quickly, but their energy density isn't as good for long-term assistance. It'd save the weight of a battery but still cost in a motor/dyno weight way.

Or just be lighter. Easier to accelerate something that doesn't weigh much. More carbon anyone?

 

[tiltmaniac]

If the frame itself was the outer container for the other energy storage/capacitance stuff, perhaps we could decrease the additional weight some fair bit?

Call it 'structural capacitors'

 

[bladderhead]

Rubber handlebar grips. And definitely a layer of rubber under the seat. I am talking about a shock. Not an air-shock.