Date: Fri, 03 Sep 2004 09:36:19 -0400

Author: cablem

Subject: Rank-Hilsch Tube


A friend and I did a little undergraduate research about this tube in
college. (see
The 3 variables that we played with were the chamber design, the length
of the exit tubes and the inlet pressure.

We found that chambers of different wall topographies (straight, sloped,
curved) gave different results. The best one we made was where the cold
side hole was smaller than the warmer side with the walls rough and
curved to look kinda like a tornado. Unfortunately, is was made from
plexiglass and the walls melted and smoothed back out, so we only got
the results once. But, it did tell us that the hot air was on the
outside while the cold had to be on the inside of the vortex. We never
had the chance to make it out of metal.

We never got below 0C, but we did end up with about a 30C difference, if
memory serves. Not bad for two kids messing around with plexiglass and a
compressor. The thing was LOUD too. The louder the machine, the better
the results we got. The faculty made us move into the accelerator room
that was being revamped at the time.

Personally, I think the machine works much like a hurricane or tornado
with high and low pressure areas directed in specific directions. Then,
as the low and high pressure exist the chamber, we get an even more
pronounced cooling or heating effect. We also noted that we got better
effects if the pipes exiting from the chamber were of different lengths
(as indicated in the figure).


On 9/2/2004 4:35 PM, Zigmund J. Peacock wrote:

> Taplers, I was perusing the McMaster-Carr site mention by Paul Nord
> and at the bottom of page 2433 is a Rank-Hilsch tube.
> A marvelous curiosity you put 80 psi air in and get -40 C cold air
> out of one end and warm air from the other.
> At $300 not bad price for a department that has almost everything
> and needs something cool and amazing for Thermo.
> I have been told that the only reason it works so well is the
> "Maxwell Demon" that is busy sorting hot air molecules from the cold
> one and assigning the to their respective ends of the apparatus.
> Ziggy
> Paul Nord wrote:
>> There are also commercial devices. Check the McMaster-Carr catalog.
>> Pages 2432-2434 in the current catalog.
>> Paul

Machele Cable
Lab Manager
Physics Dept.
Wake Forest University
Winston-Salem, NC 27109

Phone: 336-758-5532
Fax: 336-758-6142

"Foxtrot", 7/1/2004:
"3...2...1...Lift off!"
"I think the payload was too top heavy."
"Maybe we shouldn't try launching Barbies."
From Fri Sep 3 10:07:15 2004