An unusual cooling system idea

Sorry, no drawings or photos in this post.  Brain work only.

In warm seasons like this one, my thoughts sometimes turn to DIY air conditioning (AC).  In our climate, AC is one of those rarely-needed things so buying some off-the-shelf solution doesn't appear to be a very cost effective solution.  I'd much rather spend lots more time and less money on something that works well enough to take the edge off the heat when it occasionally occurs.  I've seen approaches using evaporative cooling, but our climate also has an additional challenge -- relatively high humidity.  So-called "swamp coolers" work best in regions where the relative humidity is low.  They're more efficient, and the added moisture to the air is welcome.  On the other hand, when you start with 60% relative humidity, the cooling action isn't as good:  and the increased humidity due to the swamp cooler can actually make you LESS comfortable, even if the temperature is slightly lower.

There are indirect cooling approaches where a heat exchanger comes into play, cooling inside air without adding moisture.  The heat exchanger is an added complexity, and also has its own impact on efficiency.  What if we could start with much-cooler water, cooler than can be achieved using evaporative cooling?  If we can get that, perhaps we can live with less-than-perfect heat exchanger technology.

There may be a way to do this.

Long ago I performed an experiment to see if it would be possible to make my own turbomolecular vacuum pump (I refer you to Google to get educated on this type of vacuum pump). I mounted two thin mylar plastic disks on an axle. They were spaced a few tenths of an inch apart. Then I spun them up, using either a Dremel tool or an electric drill (this WAS a long time ago so I don’t remember that part all that clearly). Centrifugal force would pull the disks into flat planes, but, if some pumping action were taking place, the space between the disks would be at a lower pressure so the disks would be pushed together. I observed that the disks were indeed pushed toward each other so some pumping action was going on. I could not increase the speed enough to get the disks to touch, due to instability and vibration problems, but this problem probably could be solved with some refinement.

If the axle was made from a tube, and we drilled some holes in it between the two disks, it is likely that some sort of pumping action would occur that would pull gas down the tube toward the disks: but now we’re faced with making a good rotary seal. The angular velocity of the tube would be less than the outer edges of the disks, but it still is a nontrivial problem if you want a reliable vacuum pump.

So that approach languished for a long time. But I recently had an idea where the scheme could still be useful. As a chiller. In point of fact, it would function as a single-pass refrigeration system. In this approach, the spinning tube is dipped into a water reservoir. The tube also has a restrictor to reduce the water flow so the water doesn’t completely fill the pump. If the water does fill the pump, I believe the water probably will boil in the disk portion of the pump and screw up the cooling cycle (see below). If sufficient vacuum is developed the water will boil, extracting heat from its surroundings. The water vapor is ejected by the spinning disk pump, where it immediately re-condenses (and liberates the heat it absorbed when it boiled).

In its simplest form, the rotating tube would be submerged in the water so the expansion of liquid water into water vapor would cool the water on the immediate exterior of the tube, which could then be circulated into a secondary heat exchanger. Efficiency could be improved by increasing the surface area of the submerged tube, perhaps with aluminum or copper disks (but they can’t be too large or frictional forces would limit the maximum RPM, and also heat the water).

The advantage of this method is that the minimum-achievable temperature is not determined by the relative humidity of air, unlike a standard evaporative air conditioner. The minimum temperature would be the freezing point of water (0C/32F).

NOTE: this scheme could be foiled by the buildup of minerals from the water feedstock, since the minerals would be concentrated by the liquid-vapor conversion.  Some type of purge or ballast-water approach would be needed for a commercially viable system.

Also NOTE: while one might think the water vapor exiting the pump would be cold & therefore useful for cooling purposes, in fact it will re-condense as soon as its pressure returns to room pressure. When this happens, it will release the heat it absorbed.  Now, of course, the water will be very finely dispersed and some to all of it will evaporate in the ambient air – again cooling down in the process. But there won’t be any “gain” offered by the pump, and the ultimate minimum-low temperature will be determined by the dew point of the ambient air – much higher than the freezing point of water.

So it appears the energy input of the spinning-disk scheme is best used as a way to implement a single-pass refrigeration system.  I use the term "single-pass" because the input fluid, water, is available without any special condensor.  Unlike a classic closed-system refrigeration system.

 

Rockin' out, part 1. An observation and pendulum theory.

Some time back I was playing with an old-fashioned bottle opener, commonly known as a "church key".  At one point I placed the opener down so the rounded and pointed ends were facing down, touching the counter top (polished granite).  Here's a photo of the opener:

  When I put the opener down, I noticed it rocked back and forth surprisingly slowly.  Intrigued, I looked more closely at what was going on.  I saw that the opener rocked back & forth across the curved end, and in that configuration the opener was pretty stable.  It would eventually tip over if pushed over too far.

And then I tried an experiment to see if I could further increase the period of this simple mechanical oscillator.  We have some small (~1/8" square) super-magnets that are used to hold photos, coupons etc. on our refrigerator.  I stacked several of them together to raise the overall center of mass of the system, and put the stack on the opener.  Sure enough, the opener rocked even more slowly.  See below (sorry, no videos yet):

I was able to increase the period to about 1/2 second/cycle, pretty amazing considering the relatively small size and mass of the system.  Due to the relatively poor finish on the rounded end of the opener, it rocked in an irregular fashion.

I started thinking about making an "improved" version of this,  for a fun little machining project.  I did finally make one, with one false start.   But at this point, rather than just showing what I did I want to start by explaining the physics behind the mechanical oscillator, and what determines its frequency.   I will start with the pendulum, as shown below (two different positions of the sphere are shown).

The sphere is hanging on a cord of "R" length.  So what causes the sphere to swing back & forth?  Take a look at the right-hand drawing of the pendulum.  The sphere has moved over, and, due to the fact that the cord is a constant length, the sphere rises slightly.  Since the force of gravity always points down but the cord is at an angle, a restoring force appears which opposes the deflection of the sphere (this assumes that the forces due to gravity and acceleration are transmitted along the cord at angle "w").  In a dynamic situation the system exhibits a periodic transfer of energy between potential energy (due to the lift "H") and kinetic energy.  What determines the oscillation frequency?  If the cord is lengthened, for a given angle "W" "H" becomes smaller, and the restoring force becomes less.  The effect is to slow the pendulum oscillations down.  If we increase the mass, the acceleration decreases due to the relationship F = Ma where M is the mass of the sphere and a is the acceleration.  Solving for acceleration:  a = F/M.  Therefore the mass accelerates more slowly under the influence of the restoring force.  So frequency also decreases as mass increases.

Another way to look at the pendulum is as a system with a center of mass that is constrained to move around a given radius of curvature.  In these terms, the oscillating bottle opener is a similar type of mechanical oscillator.  Increasing the mass (by putting magnets on top of the the opener) decreased the oscillation frequency, just as it does for a pendulum.  We could continue to add mass until the center of mass is above the center of radius.  At that point the system would become unstable and flop over.  Unlike a pendulum.

Next time:  some implementation considerations with my rocking toy.