> Very impressive stuff. I am wondering whether it would be possible to save
> the blade etc breaking by just retracting the blade very quickly instead.
> Like Michael points out, it may not work well with wet wood, humid
> environments etc and you wouldn't want to be replacing blades every five
> minutes due to false alarms.

A good trick if you can manage it.
Somewhat akin to airbag action.
Blade needs to retract from immediate area at a rate well in advance
of feed rate.
I posited an upper feed rate of about 1 m/s or 1 mm/ms.
To bring the blade up to maximum feed rate  of 1 m/s in 1 ms requires
1000g acceleration.
Quicker or faster takes proportionally more g.
Very doable, but the forces and energy involved are significant and,
while you don't need to lock the teeth, you are still imparting
significant extra energy into the blade which then needs to be removed
again to bring the blade to rest.

The system may JUST kick the blade away from the user or away and then down.
There is a risk that the backwards motion may cause a hazard unless
adequate protection is provided.

Spring or explosive charge or compressed gas or a gas generator may all work.
Metal azide probably viable.

Blade kicked back and guard flicked over such that guard falls into
area where blade was "before finger can get there" should work.
(Grossly exceed maximum feed speed and you get whacked by the
descending (or rising) guard

You present a finger to the blade and suddenly there is a guard in
place a few mm back from the original contact point. Blade is behind
guard and decelerating fast under very heavy  traditional brakes.

Energy input rate is interesting [tm].
I keep getting answers re energy input rate that I appear to be
missing something.
Actual energy needed is not large as time to accelerate is small.


        Russell
.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist