This idea can be used for compression of gases, pumping
of liquids, or turning a fluid pressure differential into
electrical power.
First, gas compression / liquid pumping:
Take two pipes, and connect them in an acute V shape.
At the point where the axes of the pipes intersect,
place a
high efficiency piezoelectric speaker.
At the other end of each pipe, place a valve which can
be efficiently opened and closed at the frequencies that
the PZ speaker can most efficiently produce sound
(Somewhere around 1MHz, I think), and which can pass
a good amount of fluid through when opened. These
valves would be opened and closed by linear PZ
actuators.
I'm thinking the valves would be similar to the valve in
an auxetophone, but there might be something more
modern and efficient.
Have the speaker make sound at high volume and at it's
optimal efficiency, and the valves open and close at
that same frequency.
The phase of the speaker relative to the intake valve is
continuously adjusted so that the low pressure portion
of each sound wave reaches the valve as the valve
opens, and the high pressure portion of each sound
wave reaching the valve as the valve closes.
The phase of the output valve relative to the speaker is
also continuously adjusted, but so that the high pressure
portion of the wave reaches the valve as the valve
opens, and the low pressure portion of the wave
reaches the valve as the valve closes.
The maximum possible pressure difference between
input and output will hopefully come close to the
amplitude of the sound waves.
The reason why dynamic adjustment is needed is
because the speed of sound in a fluid changes as that
fluid's temperature changes.
For converting a fluid pressure differential into
electrical energy:
Opening and closing the input valve (particularly with a
pressure differential across it) will of course produce
sound waves.
A portion of each sound wave will bounce off of the
piezoelectric mic and continue to the output valve; a
portion of each sound wave striking the mic gets
converted to electricity. Some of the sound, of course,
will bounce off, but that's not particularly harmful here,
as some of the bounced sound will go back to the input
valve, and then return here, and some of the sound will
go to the output valve, then return here.
The output valve for the expander will do what the
input valve did in compressor/pump; namely bounce
back the high pressure part of each sound wave, while
being open for the low pressure part of each sound
wave. Since energy is being removed (not added) by
the PZ speaker, the pressure of the low pressure part of
the sound wave will be higher than the fluid pressure
beyond the valve.
For converting liquid pressure into electricity (or for
pumping liquids), it should be possible to get pretty
good efficiency with a single stage.
For converting gas pressure into electricity (or for
compressing a gas), multiple stages will be necessary for
efficiency.
What's the big deal? The extremely small number of
moving parts, and the direct-to-electricity aspect.
Let's suppose we want to build a toy steam engine using
this tech: there are two valves and a speaker for the
feedwater pump, then the boiler, and two valves and a
piezoelectric mic for the expander, then a condenser.
If we want a compound steam engine, each additional
stage of expansion requires one additional mic and one
additional valve.
The PZ speaker and mics need no seals, and thus should
last longer than the rest of the system. The valves
could be given permanent hydrophobic coatings, which
would allow the water / steam to act as a lubricant,
which would allow the valves to last for a very long
time too.
There is a downside: our toy steam engine has no visibly
moving parts, makes no audible sound (humans can't
hear the 1MHz vibrations), and can't spin a wheel unless
we hook up the output to an electric motor. Thus, it's
not an exciting toy to look at.