This exciting method of heat-driven cooling came to us much 
later than most of the others.  It is based on Stirling engine
principles, but (amazingly) has no pistons!  The piston-like
objects are "displacers" which do not have to operate across
a pressure difference, and thus don't have to have much of a seal.

(diagram from http://ieeexplore.ieee.org/iel5/852/2490/00074788.pdf )

Richard McMahon from Down Under introduced us to this idea as follows:

From:     richard mcmahon
Date:  Fri Jul 4, 2003  2:23 am

You may get greater efficiency by using your surplus solar electricity 
to turn a stirling-cycle engine. But even more effective would be the 
following, based on the fact that you get surplus solar electricity 
when the sun is shining: Set up two cylinders (1 and 2), each 
containing a displacer linked so that they can move together in phase. 
Your surplus solar electricity can move the displacers backwards and 
forwards and the only work required is to overcome friction (say in the 
range of 10 to 50 watts) Then rig up a flatplate or trough parabola 
solar thermal collector to heat the hot end of cylinder 1. U
se either fins or water to cool the other (warm) end of cylinder 1 
to ambient, and also to cool the (warm) end of cylinder 2) to ambient. 
The other end of cylinder 2 will get very cold indeed. This is heaps 
cheaper than photovoltaics to turn a motor on a compressor as you are 
mostly using low grade solar thermal energy.
Have fun
Regards
Richard McMahon

From: Marc Ringuette

The Vuilleumier cycle is like a "non-kinematic duplex Stirling omitting
the shared piston".

Hey Richard, let me try and explain the cycle. I'll use the orientation
of the following ascii-gram (hot side to the left). I'll refer to the 
four spaces as A, B, C, D as follows:

hot side | A | displacer | B ==ambient== C | displacer | D | cold side

The entire device is built inside a single pressure vessel, which cycles
up and down in pressure once per cycle of the displacers, with all
parts of the system at the same pressure at any given time.

The displacer cooler cycle
---------------------------
Phase 1 (solar heating). Both displacers are stationary to the right.
As the air in space A is heated by the sun, pressure gradually increases.
This increase is transmitted to all parts of the system. The air in
space C is heated by compression, but cooling is applied to keep
the temperature in space C near ambient. When the pressure hits our maximum...

Phase 2 (air movement). Both displacers are moved to the left, moving
the air from A to B and from C to D. Ideally, each displacer should
act as a heat exchanger, cooling the air further and storing the heat for phase 4.
Richard, I suggest using the displacers as regenerative heat exchangers
(make them out of perforated metal, so that they store and release heat
once per cycle).

Phase 3. (cooling). Both displacers are stationary to the left.
As the air in space B is cooled by heat exchange with ambient,
pressure gradually decreases. This decrease is transmitted to the
rest of the system, where the air in space D is cooled by expansion,
and is able to absorb heat from the cooling load. When the pressure
hits our minimum...

Phase 4 (air movement). Both displacers are moved to the right, forcing
the air from B to A and from D to C. Ideally, the displacers will have stored
some heat from phase 2 and will pre-heat both masses of air ("regeneration").
Return to phase 1.

-------------------

I really like the fact that this can happen as slowly or as quickly as the
heat exchange can occur. I'm imagining a large one of these built inside
a 1000 gallon propane tank, cycling every few minutes. One end of
the (long cylindrical) tank is solar heated, one end is used for cooling,
and the middle portion is water-cooled to near ambient.

Richard, congrats, this is very promising! It's a way to use solar
heat directly in a compression based cooling system ... without any
intermediate steps involving electricity or mechanical work. Nice!

Kudos to Randal Perisho (and his anonymous Stirling engine friend)
who clued us in that this is called the Vuilleumier cycle.


Online references:

There is exactly one good online reference so far.

"Development of a gas fired Vuilleumier heat pump for residential
heating", by H. Carlsen (Denmark)   
   http://ieeexplore.ieee.org/iel5/852/2490/00074788.pdf



(the above diagram, and the one earlier, are from Carlsen's paper.)

(Note: the diagram indicates that the "cool" displacer is moved
first, then the "hot" one.  There needn't be a time delay between
the two moves, but this seems like a good choice in order to keep
a large fraction of the useful expansion (and harmful compression) 
in the correct chamber of the "cool side".)
----

I should also mention these two nice little Stirling simulations,
in case you'd like to learn (or remind yourself) how those work.
They run in your web browser as Java applets.

   http://techni.chemie.uni-leipzig.de/stirling/
   http://www.suction.co.jp/stirling/seMove/yamaMove.html

--Marc Ringuette

----

Offline references:

Stirling and Vuilleumier Heat Pumps by T. R. Roose et al., 1990, 256 p.
   http://www.stirmach.com/Books4.htm

there is also this book:
   Regenerative Thermal Machines (Stirling and Vuilleumier Cycle Machines)
   for Heating and Cooling, by Noboru KAGAWA, Dr.
   http://www.bekkoame.ne.jp/~khirata/book_ref/index_e.html

Also, Brad Ross, the SMW editor, is a really good guy, and extremely
knowledgable about all things stirling. he would be the best guy to
point you to a good source of info.  --Chris, the maniacal engineer