Einstein Refrigerator

35 views
Skip to first unread message

Marlin

unread,
Sep 18, 2012, 11:48:58 AM9/18/12
to Tx/Rx Labs
In the late 20's Leo Szillard and Albert Einstein invented a
refrigerator (actually an adaptation of an earlier similar design)
that had no moving parts. What's interesting about it is it consists
of a series of tubular "circuits", heat exchanges and three chemicals:
ammonia, water and butane. By applying heat to the left hand side of
the device (as shown in the patent diagram: http://en.wikipedia.org/wiki/Einstein_refrigerator)
you get cold output on the right hand side of the diagram. The
circuit is at a constant 60 P.S.I. and has no motors, no valves, no
moving parts and is completely sealed. The only possible requirement
for a pump is to dissipate the heat through a grid. A Georgia Tech
grad student, Andy Delano, did his PhD dissertation on it and
published it online: http://www-old.me.gatech.edu/energy/students/andy.htm

My first thought is it might be possible to make a version of this by
cutting channels in a material rather than fitting tubing together.
If the base material is non-heat conductive, then the places where
heat exchanges exist would need to be inserted. For simplicity,
simply make it 2D to an arbitrary thickness.

My Next thought is you could make an A/C system using this technology
by scaling by stacking.

And of course, this would be ideally be solar heat powered.

Forrest Flanagan

unread,
Sep 18, 2012, 12:14:42 PM9/18/12
to txrx...@googlegroups.com
Einstein didn't actually have a hand in the creation, he just pushed the patent. This technology is still used in kerosene refrigerators.

There's a simplified (batch) version of it here, based on late 1920s technology: http://crosleyautoclub.com/IcyBall/HomeBuilt/HallPlans/IB_Directions.html

It'd be great if something other than ammonia could be substituted, since it's so nasty. I am interested in how these would scale to a microfluidic level.



--
You received this message because you are subscribed to the Google Groups "Tx/Rx Labs" group.
To post to this group, send email to txrx...@googlegroups.com.
To unsubscribe from this group, send email to txrxlabs+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/txrxlabs?hl=en.


Marlin

unread,
Sep 18, 2012, 1:29:22 PM9/18/12
to Tx/Rx Labs
Yes, I was thinking about a microfluidic-scaled version as well.
There might be some problems where in the patent, a few parts of the
system rely on gravity to achieve directional flow. Also I'm not sure
how it would be possible to achieve the 4 bars that enable the butane
to maintain a liquid state in a microfluidic setting.

A mini system could be interesting, say, 1 foot^2? Maybe a mold
could be printed out on a printer. Hmm, pressure is still going to be
an issue, I'm afraid.

On Sep 18, 11:14 am, Forrest Flanagan <solenoidcl...@gmail.com> wrote:
> Einstein didn't actually have a hand in the creation, he just pushed the
> patent. This technology is still used in kerosene refrigerators.
>
> There's a simplified (batch) version of it here, based on late 1920s
> technology:http://crosleyautoclub.com/IcyBall/HomeBuilt/HallPlans/IB_Directions....

Forrest Flanagan

unread,
Sep 18, 2012, 2:54:54 PM9/18/12
to txrx...@googlegroups.com
It'd be much easier to maintain pressure in a microfluidic system.

P = 2st/(d-2t)

where,

P = max. working pressure (psig)
s = material strength (psi)
t = wall thickness (in)
d = outside diameter (in)

Marlin

unread,
Sep 18, 2012, 5:54:30 PM9/18/12
to Tx/Rx Labs
I found another kink. The system relies on a "bubble pump" where the
low point in a vertical channel is heated to gassify and produce
bubbles. The bubbles rise, causing the fluid to move vertically with
the bubbles. So for a microfluidic system to function you probably
need some form of externally powered pumping mechanism, unless you can
think of another way...

Forrest Flanagan

unread,
Sep 18, 2012, 8:02:06 PM9/18/12
to txrx...@googlegroups.com
microfluidics is an entire discipline of moving fluids through tubes in controlled manner. There's probably a playbook of ways to achieve equivalent effect without moving parts.

I'll admit, the thermoacoustic discussion you had going was a lot more intuitive to me, I just knew about the icy ball because I printed a bunch of stuff before power cut out for two weeks during Ike.

Mark Sullivan

unread,
Sep 18, 2012, 8:09:37 PM9/18/12
to txrx...@googlegroups.com
> It'd be great if something other than ammonia could be substituted,
> since it's so nasty. I am interested in how these would scale to a
> microfluidic level.

Modern absorption chillers use water as the refrigerant and lithium bromide as the absorbent.

My grandma had a Servel refrigerator powered by natural gas. They later turned into Arcla.
My favorite ammonia refrigerator was the "Icy Ball".

- Mark Sullivan -

Marlin

unread,
Sep 18, 2012, 11:39:34 PM9/18/12
to txrx...@googlegroups.com
I actually first read about the Crosley Icy ball a few days ago.  It's kind of scary because it's operating pressure is 250 psi.  The operating instructions are kind of involved, like you heat one ball while the cold ball is immersed in cool water.  The motivation was that if you had an Icy ball you didn't have to pay the ice man to come and deliver ice every day, so a somewhat involved daliy chore didn't seem too bad.   I've also studied the Arcla AC units a bit.  They are a little strange to me because I don't think they were very efficient and as I learned more I began to wonder if they weren't a sort of conspiracy by the gas company to get customers to use gas even in the summer!

I need to learn more about modern absorption chillers.  Do they have any moving parts?

Marlin

unread,
Sep 19, 2012, 11:41:48 AM9/19/12
to Tx/Rx Labs
I'm coming to the realization that it would be a pretty bad idea to
make an A/C unit that uses strong ammonia. It's one thing to make a
fridge but an A/C is an order of magnitude greater than that. Also,
it seems that the so-called Einstein design is limited to about 28 F
output temperature. Andy Delano's PhD thesis outlines alternative
chemistry to improve that, However, the only alternative to Ammonia
that Delano suggests and provides information on is Hydrogen Chloride.

Freon came into existence because it is non-toxic, didn't smell bad
and had an excellent boiling point. It's a bummer about the
environmental consequences.

Brian Dawson

unread,
Sep 19, 2012, 12:03:16 PM9/19/12
to txrx...@googlegroups.com
I've been interested in 'Einstein's Refrigerator' For some time.  Attaching a schematic from Delano's thesis and the description.

"Figure 1 shows a schematic of the Einstein refrigeration cycle. In the Einstein cycle, ammonia acts as an inert gas to
lower the partial pressure over the refrigerant, butane, and water later provides separation by absorbing the ammonia.

Starting in the evaporator, liquid butane arrives from the condenser/absorber at point 2. In the evaporator, saturated
ammonia vapor is bubbled into the liquid butane (point 4), which reduces the partial pressure of the butane. With its
partial pressure reduced, the butane evaporates near the saturation temperature of its partial pressure and cools itself
and the ammonia, and provides external refrigeration. The ammonia-butane vapor mixture leaves the evaporator at point
3 and enters the pre-cooler, where it cools the hot vapor ammonia counter-flowing from the generator (point 5). The
now superheated ammonia-butane mixture flows out of the pre-cooler at point 6 into the condenser/absorber, which is
being continuously cooled by an environmental heat sink. Meanwhile, liquid water from the generator is sprayed into the
condenser/absorber at point 9. With its affinity for ammonia vapor, this water absorbs the vapor ammonia from the
ammonia-butane mixture. The absorption of the ammonia vapor increases the partial pressure on the butane vapor to
nearly the total pressure, allowing it now to condense at butane’s saturation temperature for the total pressure. Note
that this is higher than butane’s saturation temperature at the partial pressure in the evaporator. The butane and the
ammonia water separate due to their respective density differences and the fact that ammonia-water is immiscible with
butane at the condenser/absorber’s temperature and pressure. Since liquid butane is less dense than liquid ammonia-water, it
is the top liquid and is siphoned back to the evaporator at point 1. Meanwhile, the ammonia-water mixture leaves from the
bottom of the condenser/absorber and enters the solution heat exchanger (point 7). Here, the mixture is pre-heated before
entering the generator.

Inside the generator, heat is applied to the strong ammonia-water solution driving off ammonia vapor where it
rises and is carried to the evaporator (point 5). The remaining weak ammonia-water solution is pumped up to a reservoir via
a bubble pump at point 8. In the reservoir, any residual ammonia vapor from the bubble pump is sent to the
condenser/absorber (point 8g). The weak ammonia water solution falls to the solution heat exchanger (point 8f) where it
gives up its heat to the strong ammonia-water solution leaving the condenser. Finally, the water is sprayed into the condenser/
absorber.

While the overall pressure of the cycle is constant, there are slight pressure variations within the cycle necessary for
fluid motion. These are due to height variations and are not large enough to significantly affect property evaluation."



--
You received this message because you are subscribed to the Google Groups "Tx/Rx Labs" group.
RefrigCycle.png

Mark Sullivan

unread,
Sep 19, 2012, 3:07:16 PM9/19/12
to txrx...@googlegroups.com
 
I think most LiBr chillers use a pump.  They run at high vacuum, so sealing is as important as it is with a pressurized system.  There's no reason you can't make a no-moving-parts-but-fluids chiller but I don't know how common it is.
 
When I was a kid, I spent hours at Grandma's house reading her encyclopedia.  There was an extensive article about refrigeration and that's where I learned about the Icy Ball.  It was probably contemporary with the encyclopedia.  It's always struck me as a very elegant solution.  Compare it to today's domestic refrigerator with half a dozen motors.  I think there's an important engineering lesson here.
 
Absorption chillers have a low COP compared to vapor compression, but if include the efficiency of making prime-mover heat into electricity or burning it in an engine and then using that to run the compressor, I bet the comparison gets a lot closer.
 
I've seen them in RVs where they needed dual-fuelled refrigeration.  The refrigerator used an absorption chiller which could either be heated with propane on the road or with an electric heater when plugged-in.
 
When I was in College, ca 1980, the Chemistry building was cooled by a big absorption unit that was powered by steam from the central plant that heated the whole campus.
 
I don't know about the COP of the Arcla but the flame in Grandma's was very small.  I've seen bigger pilot lights.  Maybe it just had good case insulation.  I remember it as massive, but then I was much smaller at the time.
 
 
- Mark -
 
Reply all
Reply to author
Forward
0 new messages