On Wed, 31 Mar 2004 10:20:59 +0100, "JimB"
wrote:
.....
I fixed a card to a bit of wire which I could bend at various
angles to the card (rudder angle). I then hung the card upstream
of the fan so that it could pivot only along the fore and aft
axis (above the fan) and again so it could pilot only along the
lateral axis, and again so that the rudder could rotate around
the vertical axis of its front post. Smoking fag ends came later,
with a rigid mount. Fan was turned on.

Forces were observed by noting the degree of card deflection
around the relevant hinge. There was a net force on the rudder,
primarily exerted towards the fan. It has little lateral
component, but lots of fore and aft component.

Smoking fag ends showed strong non-linear flow when the rudder
was deflected, but the net flow is an s bend zig-zagging around
the rudder towards the fan and two carpet burns. Smoke speed
dropped markedly with distance away from the fan.

The rudder kicked hard over (either way) when allowed to pivot
around its forward vertical axis. Within the limitations of my
crude experiment, rudder kick is probably caused by the net
effect of the fore and aft component of force, not a lateral
component.

I think this explains the rudder kick I've witnessed in astern in
some boats engaging astern gear, and probably explains Derek
Rowell's observation that the rudder rotates when allowed to (If
I understood his experiment design right).

However, the zig-zagging airflow proves to my satisfaction that
the rudder may not create a net lateral force, so I'm stuck with
the revelation that the yaw effect that my old skipper
demonstrated to me was wind, mirrors, inertia, prop walk and my
hero worship.

Ah well.

But now I've got to get rid of those carpet burns before the wife
comes home.

JimB

Asking for explanations from experimental rigs is the royal road to
progress. Congratulations!
Couple of your observations bear talking about.

If a hinge surface is hinged more than about 1/4 aft of its present
leading edge it is unstable in the fluid flow. ('rudder kick')

If a surface *is* hinged about 1/4 from the leading edge, it can
still break into oscillations which are quickly destructive, unless
the mass is balanced closer to the hinge line.

If a FLAT surface is inclined slightly ( 20 degrees) to the fluid
flow, the flow over the 'upper' surface is faster and provides lower
pressure than the flow over the lower surface. The streamlines do not
follow the (flat) surface of the test article (of course!), they kick
up in a smooth curve over the top. This applies to an airfoil flown
upside down too. The streamlines look similar to the streamlines
over a right way up foil, but less efficient and with lower pressure
difference from top/bottom.

It is not necessary for a lump of fluid dividing past the foil to
join up again after it has passed..
When providing lift, the lump of fluid does not join up again, in
fact.

Brian Whatcott