Larry Jaques wrote:
On Thu, 29 Sep 2011 22:33:35 -0700,
wrote:

(...) -- Indicates snippage, *not* three nostrils.

Me too, since the same blade cuts through steel
really nicely when the workpiece is clamped in
a bench vise and the saw is guided by hand.

Are you absolutely sure it's not caused by vibration of the workpiece?
If the piece is hopping, it could -induce- a waggle, oui?

I haven't seen any high speed video, but you are probably right.

Probably the first stroke of the blade goes well
then *any* contact of the blade to the workpiece on the return
stroke induces movement in the handheld workpiece, which causes
the largely undamped blade to 'violin', side to side.

Once this lateral oscillation of the blade starts, cutting just
about ceases as the workpiece gets trapped by the displacement
caused by the resonant standing wave in the blade. (It doesn't
really matter if it is fundamental or harmonic resonance.
A dynamically floppy blade is effectively wider than the kerf
at most points along it's length.)

http://www.youtube.com/watch?v=-n1d1...eature=related

At that point, the workpiece spends much of it's time riding
the blade.

So the guides dampen the oscillation of the blade at it's
resonant frequency. The blade remains 'straight enough'
laterally and therefore does not trap the workpiece.

The blade probably still oscillates laterally when it cuts a
clamped workpiece and the saw is hand held. It does not
approach resonance because the base pressure of the saw
against the clamping device (vise, workbench)
prevents the workpiece relative movement that causes blade
lateral resonance.

A first - order test would be to clamp a workpiece in a vise
and attempt to cut it with much less base force than usual.
I suspect that the lowered mass coupling would cause the
blade to enter resonance at some FPM and pressure settings,
resulting in a bouncy ride for the user. BTDT
(Queue SFX: "trombone gobble")

That is my story and I'm sticking to it.

My take-away:
What if you made a 'virtual' top guide, immune to swarf?
Several displacement sensors and electromagnets dynamically
phase - cancel lateral acceleration of the blade,
straightening it in real time.

--Winston