In Part I, I summarized my efforts to build myself a testing platform for refrigeration projects. I cut that post short at about the point where I began doing some heavy brazing work. Also, the amount of media in that post was getting to be quite extensive, though this post will be no different.
I want to again point out that I have no specific goal with this project, just a general interest in refrigeration, the end to to which will be guided by wherever the winds may blow my short attention span.
In order to assess the performance of these systems, and to evaluate alterations, data collection along with model construction, will be critical. As of this date I am still in the qualitative, “build it and beat on it” stage. As I hone my construction skills with copper work, and develop more controllable (and better insulated) outcomes, I begin to ask questions like, “Well it seems to work well, but how well, and is it any better than before?”.
For now, I have a bit more media to bring the reader up to date on the state of the project.
We begin with a picture of the transformation of an old coil into something quite new. The larger diameter coil in my hand has been recovered from the test bench, identifiable by its apparent diameter as the dry evaporator in the pot. As I mentioned before, I had grown tired of the dry evaporator, and the constant necessity to adjust the throttle in order to maintain superheat. Dry evaporators, so common in modern HVACR equipment, often utilize thermostatic expansion valves to regulate superheat. They are quite effective little devices and their utilization make for lightweight, economical evaporators, when properly sized and installed. However, my interest in antique refrigeration systems has led me to revitalize a technology which has laid dormant in domestic systems for some time now.— I am constructing a gravity flooded evaporator.
Here you can see the coils I made; two or three from the dry evaporator, and the remainder from fresh copper. In the previous picture the reader can see the spring bender in my hand which prevents the copper tubing from collapsing when it is bent. There are other methods to keep the copper from kinking when bending, including filling the tubing with sand, salt, or even ice, but I found this method to be much less messy. The only drawback so far with this method, is that there is a lower limit to the radius which can be bent because the spring will bind onto the tubing and will not slide freely once a portion is bent. I of course did not bend these by hand, but used a 3″ ABS pipe as a form. These spring benders are quite handy for “free hand bending” as well. I purchased a large set of these in many sizes at Harbor Freight Tools. Don’t over pay for a simple tool like this. They are wonderful.
A preliminary fitting of the evaporator coils to the central column or “separator”. The coils do not form one continuous run, like in a dry evaporator, nor do they have a distributor. When operating, low pressure liquid refrigerant will “flood” the central column and the coils from the bottom to just below the top coil. The suction line will draw of saturated vapor from the top of the separator.
The end of the first night of work. A liquid line sight glass/moisture indicator is being used as a liquid level indicator in the low side of the system. The 3/8″ lines flared to the sight glass communicate with the central column in a similar fashion to the evaporator coils; the bottom portion is connected to liquid, and the top to vapor. I will from this point on refer to this apparatus as the “Ebullator”, a name I will explain shortly.
Preparing for brazing. There is a bit more to explain here. The brass flare fitting connected to the top of the separator will be the suction line. The small 1/4″ line behind it will be a low pressure gauge. The 1/4″ tubing pointing up and to the right, will be where high pressure liquid refrigerant from the condenser will come in, heading for the throttling device. That line enters the separator at an angle, but does not directly communicate with it. Passing through the 1″ separator, it passes through the bottom cap, takes a 90º turn to the right, a 180º degree back, and then it goes through the throttling device which in this case is a 1/4″ needle valve, finally entering into the side of the separator below the low pressure liquid surface. Got it?!
I made a clamp to hold the awkward piece securely, but still allow me to braze everything at once. It’s simply a piece of 1/2″ copper pipe flattened, shaped around a shovel handle in the vise, and bolted to a piece of steel secured by clamps and the bench vise.
The whole setup after brazing was completed. I purge with dry nitrogen to ensure oxidation does not occur inside the Ebullator. I braze with oxygen and either MAPP gas or propane. This brazing job was the vast majority of any copper brazing I’ve ever done. It wasn’t easy, and by the end of it I felt more confident than when I began. I was delighted to find it held pressure without any detectable soap bubbles!
Even though they are not the prettiest joints, I am happy with how it turned out, considering the complexity of the job, for a novice that is.
Here it is completed with all the ugly brazing joints hidden under the necessary insulation. Even though the central column will be filled with relatively cold liquid refrigerant (corresponding to suction pressure for the refrigerant chosen: propane), heat transfer is not desirable in the separator or the sight glass. This will be apparent in due course.
As high pressure liquid refrigerant passes through the throttling valve, a pressure drops occurs where a portion of the work instilled to the substance as pressure is converted to internal energy, as the process is essentially isenthalpic. A portion of the liquid “flashes” or otherwise evaporates suddenly, expanding in volume almost proportionally to the pressure dropping, going from a liquid to a gas. -Lost work.- Additional flash gas is produced due to the higher temperature of the subcooled, high pressure refrigerant (presumably a few degrees below the condenser temperature). What exits the throttling device and enters the separator is a combination of low temperature liquid and saturated vapor. I hope I explained this adequately.
The action doesn’t stop there. The previously described process is essentially true of all throttling systems in vapor compression circuits, but from here on, things are a little different. In a normal dry evaporator, the liquid gas combination would be carried through the entirety of the evaporator loop, gradually increasing in quality (A percentage of how much of the refrigerant is vapor), returning to the suction side of the compressor as a saturated vapor, or more likely a superheated vapor.
In the ebullator on the other hand, low pressure liquid is maintained to the height of the sight glass, and thus throughout the device at that level. When the liquid/vapor combination enters the separator, the vapor – bubbles to the top and is carried away to the compressor directly, leaving only 0% quality saturated liquid in the bottom of the separator and in the evaporator coils. This cold liquid keeps the inside surface of the coils wet, and absorbs heat much more effectively compared to the high velocity (relatively speaking) liquid/vapor “mist” of a dry type evaporator. As the temperature of the liquid inside the coils is presumably maintained lower than the temperature of the surrounding medium (air, water, brine, etc.), heat naturally conducts from the copper to the liquid, and of course from the environment to the copper. In other words, the environment is cooled.
Now since the liquid inside the coils is at saturation, any heat absorbed will cause the liquid to go through phase change, forming vapor bubbles which will, by their lower density, rise to the surface and be drawn back to the compressor as a saturated vapor. These vapor bubbles can form anywhere in the liquid volume that heat can be absorbed, since the entirety of the liquid volume is held at suction pressure, with the exception of a slightly higher pressure – lower in the ebullator due to static head. The formation of these vapor bubbles throughout is commonly known as “boiling”, also commonly called “ebullition”; thus the term “ebullator“.
By maintaining the liquid level no higher than the top of the evaporator coils, the bottom is always in communication with liquid in the separator, and the top with vapor. As the ebullition becomes increasingly violent, the intended consequence is for the vapor bubbles rising upward (due to lower density) to carry liquid refrigerant back into the separator, creating a “spill over” effect, leading to “makeup” liquid to enter the bottom of each coil from the separator. If the central column was not insulated, vapor bubbles would form and rise, going against the intended natural flow. This is the primary reason for insulation. Likewise, there is no reason for the lines communicating the sight glass with the separator column to absorb much heat, as this could lead to inaccurate readings.
Together, these processes could be described as thermosiphoning. However, much of these intended effects were theoretical, based on textbook reasoning and the study of many different gravity flooded evaporators over a half century or so. In Part III, I will share some of my experience putting this device into service and some observations made.
[…] work I have been doing with ejector and flooded evaporator is much too valuable (and so much fun), that I can’t let the dust collect on that […]