The question here isn't whether or not this would 'work': clearly, the physics are sound. However, will it work efficiently enough to be of any use to anyone? I

The question here isn't whether or not this would 'work': clearly, the physics are sound. However, will it work efficiently enough to be of any use to anyone? I'm a little rusty, but let's try and run some numbers (feel free to jump in and correct me if I go wrong). Estimating some spring dimensions (0.7mm wire diameter, 13mm outer diameter, 30mm free length) and looking up comparable springs from an online supplier (http://www.leespring.com/browse_catalog.asp?rbunitOfMeasure=Metric&springType=C&partnum=&UnitOfMeasure=Metric&specsCriteria=%28fl_dimension+between+0.984+and+1.378%29+and+%28size_dimension+between+0.02+and+0.035%29+and+%28od_front_dimension+between+0.461+and+0.563%29+and+%28%28int%28cCUR%28iif%28isnull%28solid_ht_dimesion%29%2C0%2Csolid_ht_dimesion%29%29*+1000%29%2F1000%29%3C%3D0.787%29+and+%28%28int%28cCUR%28iif%28isnull%28to_work_over_rod%29%2C0%2Cto_work_over_rod%29%29*+1000%29%2F1000%29%3E%3D0.433%29+and+%28InStr%281%2C%27MW+++%2C+SS+++%2C+OTCS+%2C+SS316%2C+17-7+PH%2C+ELGILOY%2C+ULTEM%27%2Cmid%28%5Bmaterial_alloy%5D%2C1%2C5%29%2C1%29%3E0%29&subType=&isMIL=&searchHistory=2 http://www.leespring.com/browse_catalog.asp?rbunitOfMeasure=...) seems to indicate a spring constant (k) of around 0.2 N/mm (200 N/m). Hooke's law states that the potential energy stored in a spring is equal to (k(x^2))/2, where k is the spring constant and x is the displacement (distance squeezed/stretched). If we take a wild, optimistic guess that each 360 degree rotation of the battery head produces 1cm of compression in the spring, that means each rotation stores a nicely round (200(.01^2))/2 = 0.01J of potential energy in the spring. Let's continue our optimism and say that the mechanism connected to the spring is able to quickly convert that compression into rotation, spinning a tiny generator which converts the kinetic energy into electric potential energy in the battery at, let's say, 80% efficiency. Each twist therefore stores 0.008J, or 0.002 mWh, in the battery. A decent NiMH AA battery stores around 2500 mAh. If we assume the cell in the design maintains the same energy density as a normal NiMH battery, at ~30% of the size, we can estimate a fully charged capacity of ~750 mAh. Since AA batteries operate at 1.5V, this gives us a total energy capacity of (1.5 x 750) = 1125 mWh. This means that, to fully charge the battery, you'd need to twist the top all the way around (1125/.002) = 562,500 times. Based on my home experiments fondling the top of a normal AA battery, I estimate that it's possible to rotate the top approximately twice per second, which means that fully charging the battery would take (((562 500 / 2) / 60) / 60) / 24 = 3.25 days of continuous twisting. (note: this ignores wrist and elbow fatigue, which I started to experience after only a few hundred twists) So more than 3 straight days of twisting one of these suckers, for a third of the power of a standard AA battery. I'd call that unfeasible, unless you had some serious time on your hands and were absolutely desperate for a few milliwatts.