Getting Started: The First Ten Years

My active interest in stirling engines began in 1971 when I discovered the Philips literature and was inspired to design an air-charged stirling of several horsepower for use on a bicycle. The project was expected to take about a year.

My initial design was a complex mess, featuring twin double throw cranks, scotch yokes with rollers, twin cylinders, and a novel speed control system with several valves and passages upon which I had recently obtained a patent. After nine months of work, I wisely decided to scrap it entirely, and build a simpler engine based on Philips’ rhombic drive. After many more months of effort, this new engine was finally ready for its initial test run. Unfortunately, even with the heater tubes glowing bright red from the heat of a propane torch, the engine would do nothing more than turn six or seven feeble revo­lutions at a time, and then stop.

Putting this failure aside, I promptly undertook to design and build a V-2 alpha stirling based on an automobile freon compressor. This engine was to be a quick project that would be completed in one month and would raise my sinking morale. It actually took five months of spare time to complete, and it, too, showed no serious inclination to run. So it was that I spent two years of quite considerable effort in reading the stirling literature and designing, machining, and building engines, and yet was still unable to get an engine to run, let alone produce any useful power.

By this time, the "Iet’s put one on a bike and have some fun" idea was long forgot­ten. I was now obsessed with the idea of the modern stirling engine, per se, and I was determined to have the satisfaction of seeing one of my engines run. To that end, I wrote to Ted Finkelstein, to obtain the benefit of his wisdom.

In his kind response of August 1, 1973, Dr. Finkelstein wisely suggested that I go back to the basically sound rhombic design, eliminate the complex valved speed con­trol mechanism, and modify the heater to minimize thermal losses.

In accordance with this advice, the 65cc (cc = cm3 swept volume) rhombic was rebuilt with a simple stainless steel annular heater, which surrounded the engine’s hot space and thereby greatly reduced conduction losses from that space. The new heater also eliminated the excessive internal dead volume associated with the original tubu­lar heater. In addition, a new, thin-walled stainless insulation dome on the displacer replaced the original aluminum dome, further reducing the conduction of heat out ot

Getting Started: The First Ten Years

Getting Started: The First Ten Years

The hot space. In all other respects the engine was left as it was, since mechanically it was already quite good. Its friction was low and its seals were excellent. Both the pis­ton and the displacer used two-cycle engine racing-type piston rings, which are thin chrome plated steel rings with very low outspring. One unusual feature of the design is a displacer bore (2.280 inches, or 58 mm) larger than the piston bore (2.135 inches, or 54 mm), which eliminates the need for the conventional piston side clearance cutouts, turning what would have been dead space into displacer swept volume.

With these changes, the engine ran on its first attempt, in late January, 1974. At that time, it had no proper burner, but was merely heated by a hand held propane torch. Nor did it have any regenerator matrix. Despite these handicaps, the engine qui­etly turned 200 rpm, and I was quite happy.

It took alittle over a month to build an annular ring burner with a single row of jets and a rather crude prony brake. When these accessories were at last operational, I was shocked to learn that peak power was a mere 1.5 watts, @ 750 rpm, atmospheric, without any regenerator. With a 0.018 inch diameter (0.5 mm) wire regenerator power doubled to 3 watts.

WaIker’s book (see bibliography) suggested a good starting place for the gap between the heater wall and the displacer (or inner sleeve) was 0.015 to 0.030 inch (0.38 to 0.76 mm). The gap inthe rhombic was 0.060 inch (1.5 mm), so I added an in­ner sleeve to bring this gap down to 0.020 inch (0.5 mm). Performance improved to 3.9 watts, atmospheric, and 8 watts at 2 bar charge pressure. Next, I tried fine steel wool as a regenerator matrix, and power jumped to 7 watts, atmospheric, and 9 watts at 2 bar. Moreover, for the first time, the engine’s free speed was higher at 2 bar (1800 rpm) than atmospheric (1500 rpm). By the end of April, 1974, the engine had produced 16.3 watts at 1265 rpm at 2.3 bar, with a propane torch assisting the burner. Even then, only a small portion of the heater head was even at a dull red heat. It was obvious that a great improvement in both speed and power would be available when a burner could be devised to keep the entire heater glowing bright red.

Of course, these power levels are absurdly low for a 65cc engine, but the fact that minor changes sometimes doubled the engine’s previous performance was extremely encouraging, and served to further heighten my enthusiasm for experimental stirling work.

At this point it seemed reasonable to concentrate on improving the burner. I was aware that the annular type burner (similar to those Philips had used) would need 100% aeration for proper combustion, and that the ratios of the propane orifice cross section­al area to the mixer tube area, and of the mixer tube area to the burner jet area, were critical. What I did not realize was the magnitude of improvement available when the ratios were right. This knowledge came as I experimented with a set of interchangeable orifices I had made, each with a slightly different size of hole. I was used to rather mushy dark blue flames issuing out ofthe burner jets, which would barely turn the adjacent heater head a dim red in low light when the engine was not running. Most of the ori-

Getting Started: The First Ten Years

Fices produced just such flames.

But upon trying one of the smallest orifices, the burner immediately changed its entire personality. In the first place, it was cranky and hard to keep lit when the engine was cold (a variable mixture control later solved this problem). The jets would ignite and go out in a circular pattern around the burner. As the heater head became warm, however, the burner stabilized nicely. Instead of the mushy dark blue flames I was ac­customed to, the jets were now producing hard, bright blue miniature torch points.

And they were no longer silent; they produced a sort of sizzling sound. Best of all, the narrow strip of heater adjacent to them glowed bright red. At once, the answer to the burner problem seemed obvious. Simply stack four or more rows of jets into a burner, and find the right size orifice and mixer tube. I now knew how to put significant heat onto the engine’s heater. Several new burners were made along these principles, and each one boosted performance. With the last burner, made in 1980, the engine pro­duced 32 watts at 1350 rpm, atmospheric; 66 watts @1050 rpm at 2 bar; and 81.7 watts @1250 rpm at 2.7 bar inlet air pressure. The engine under the burners was relatively unchanged.

A great deal of what I learned on this engine was qualitative, rather than quantita­tive. For example, on various occasions the engine was running on air, and then helium was introduced into it. The speed and power would immediately increase by 50%, and the red glow of the heater head would rapidly disappear. In the darkening heater one could actually see the improved transfer of heat into the engine. Interestingly, helium has little apparent effect on later prototypes with extended surface area heaters, since they already have good heat transfer with air.

One early improvement was afinned aluminum alloy cooler, which achieved a substantially increased surface area over the original drilled cast iron cooler without any increase in dead volume. I was surprised when performance remained unaltered, but later realized that the engine was heater-limited, not coolerlimited, so the superior cooler could make no difference until the heater was improved.

Tests were conducted on various regenerator materials, including steel and stain­less steel wool, woven and wrapped stainless wire, ceramic wool, and dimpled stain­less foil. The stainless steel wool had numerous small particles that broke away and got into the heat exchangers. The ceramic wool broke down completely and was blown throughout the engine. The woven stainless wire and foil were the most promising, with the foil moving the peak power up to a slightly higher speed. ln these and other early tests my aim was to explore as many ideas as I could in as little time as possible. I was seeking breadth of knowledge rather than depth, simply because there was so much interdependent territory to cover.

This 65cc rhombic truly was the workhorse of the first 10 years of my stirling work. It has run numerous hours on minimal lubrication without giving any trouble; and, indeed, is still doing so as a student test engine at Ohio University.

During these early years l was actively corresponding with, and meeting, just

Getting Started: The First Ten Years

About everyone I could in the field, both professional and amateur. The value of such personal contact cannot be overstated. Not only is information shared, but also morale is sustained. Stirling engine development, like any other creative work, can be lonely, frustrating, and difficult. There is no guarantee, and often little evidence, that one’s ef­forts are not a complete waste of time. But if the small triumphs that occur and provide such satisfaction are shared with others active in the field, then all can continue their work with a better spirit.

Добавить комментарий

Ваш e-mail не будет опубликован. Обязательные поля помечены *