Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: Singing pipes



At 09:30 12/26/97 -0800, Bill Beaty wrote:

If anyone here can quickly set up a Rijke Tube demo, I have something for
you to try out. The Singing Pipe normally goes silent when turned
horizontally....
To further investigate this, I'd like to find out if the Rijke Tube still
makes noise if positioned horizontally while being provided with a gentle
air flow from a nearby fan.
....
William J. Beaty

Starling and Woodall call this setup a 'howling tube' from the
overtones present. They contrast it with a singing flame produced
by inserting into the same tube not a gauze but a small flame from a
bunsen (fed by a rubber tube of some critical length.)

This arrangement is said to produce a rather pure tone
( if you hold your mouth just right ) accompanied by an oscillation
of the flame size.

Here they assert it has been shown the flame burns most strongly
and hence supplies heat to the air at the greatest rate when the air
surrounding it is in a state of maximum compression and therefore
just on the point of expanding.

The flame is most effective, they say, when placed at the position in
the vertical tube of greatest pressure changes - that is when it is
at a node.

But returning to the gauze in tube: the transfer of heat from the gauze
which accounts for the energy of the oscillations, takes place at its
greatest rate just before before the compression is greatest, "for the
air is then streaming upwards with its greatest velocity towards the node
at the center of the tube. This supply of heat assists the expansion of
the air which takes place in the next phase of the cycle and the heat
transfer ... drops as hot air is again being driven again over the gauze."

It occurs to me one might try an electrically heated gauze in a tube
closed at one end, where a lump of solid CO2 is stationed. This might
be pursuaded to sound off even when horizontal for the reasons you adduce.


brian whatcott <inet@intellisys.net>
Altus OK