The gyroscopic forces on the mill when it yaws, which can make blades bend toward the tower, strike it, and shatter, go up with the product of the rotation rate and the yaw rate. So letting the prop speed up in a high wind (which can also yaw it more quickly when it shifts direction) increases the chances of such a smashup.
More significantly, the centrifugal force (the real reaction force exerted by the mass of the blades against the centripetal force pulling their motion into a circle) goes up with the square of the speed. If your rotor speeds up by a factor of 10, the forces trying to tear the blades off and let them fly away at (wind speed * TSR)/2 or tear the tips off and let them fly away at (wind speed * TSR) go up by a factor of 100.
Additionally, a lift-type mill loaded for maximum power may be turning about half the speed it would turn if it were unloaded at the same wind speed (and angle to it). If it's not furled, removing the load will double its speed and quadruple the tear-apart forces in it. But the actual speedup may be much more, for several reasons:
- The load of a simple alternator/rectifier/battery system rises more with higher wind that what is necessary to track the maximum-power point. It will be designed to run at max-power-point for a wind speed on the low side of the typical wind distribution. This means the speedup may be much more than a factor of two in typical-to-high winds if the load is lost or the wind overcomes the generator drag.
- The furling force in an offset-rotor, pivoted-tail mill is a result of wind drag on the rotor, and most of that is the result of lift on the blades from the alternator load. If the mill is furled and the load is lost, the furling force is lost. The mill will turn full-on to the wind as if there were no furling system.
- Current goes up with the speed above cutin and heating from it goes up with the square of current. (Available energy goes up with the cube of the wind speed so there's no lack of it to drive this failure mode.) So if your mill didn't furl adequately in a high wind and the genny doesn't load it enough to drive the rotor into aerodynamic stall, the genny heating will rise greatly and the genny wiring will melt (and perhaps the genny will catch fire and/or drop molten metal on the weeds below) or the diodes will fail open, after which it provides no load. Then the prop runs away as above.
If you lose all or part of a blade due to centrifugal tear-apart or impact with the mast due to gyroscopic forces on yaw (or impact with the tail), the remaining rotor will be massively imbalanced. So it will shake the mill and tower mightily until enough stuff fails (probably including other flying pieces of blades) that the rotation stops. This may cause shaft bending, rotor-stator rubbing, bearing failures, guy or tower failure, or even make the tail jump off its pivot or the whole mill jump off the mast.