|
Title: |
1907 Article-Johnston Oil Engine Co.-Hoizontal Crude Oil Engine |
|
Source: |
Gas Engine, V9, Aug 1907, pg. 311 |
|
Insert Date: |
3/4/2016 12:31:49 PM |
Among the more important reasons for the low "commercial efficiency" or "power per dollar ratio" attained by the present well-known types of liquid fuel internal combustion motors, a few may be briefly stated as follows:
(1) A refined product is used for fuel.
(2) A large percentage of the heat developed by the combustion of the fuel is absorbed by the water jacket, and thus lost.
(3) The limit of allowable compression pressures is very low, due to the danger of pre-ignition of the mixed air and gas during compression.
(4) The lowering of compression pressures in throttle governing engines, when running under full load increases the fuel consumption.
Other serious defects in the ordinary systems are due, first, to the electric ignition devices being complicated and liable to disorder, and, second, to the carburetors or equivalent mechanisms, being uncertain and failing- to give correct mixtures under variable loads or when starting the engines
The problem, then, as it presented itself to me, was to discover what new features could be devised, and what old devices eliminated so that the resulting product would be an engine which would, as far as possible, be subject to none of the undesirable features mentioned above.
(1) As to the quality of fuel:
It was evident to me that if all grades of heavy, unrefined oils were to be used, a vaporizing system would be of no use, as a large percentage of the constituents of heavy oils will not vaporize, but merely form a thick deposit of tar and carbon in the vaporizing chamber. Hence some other system had to be devised.
Having had some experience with oil furnaces for heating steel in which heavy crude oil was broken up into a fine spray with compressed air, and in that state completely burned, with no sign of residue, I was firmly convinced that just as perfect combustion would take place in an engine cylinder under proper conditions.
(2) The only method of reducing the water jacket loss seemed to be in the use of a combustion chamber and cylinder that needed no water jacket, but it was plainly apparent that if the cooling system were eliminated the charge would reach a very high temperature during compression, and hence the fuel and the air should not be allowed to come into contact until the moment for ignition had arrived.
Mathematically the requirements would call for the walls of the combustion chamber to be of a non-conducting material, with no capacity to retain heat, but as far as the writer's present knowledge extends no heat insulating or non-conducting material has been found, which will serve for the interior lining of a cylinder though many experiments have been made for the purpose of finding one. Hence it has been necessary to fall back to the engineer's old standby, cast iron, and to use an asbestos packing to prevent the escape of heat.
This descent from the theoretical, idea chamber with non-conducting walls, to the practical cast-iron asbestos packed cylinder, is one of the many approximations that have to be adopted in working out our mathematical equations with commercial materials.
The iron absorbs a certain amount of heat from the burning charge, which it gives out to the succeeding fresh charge and thus, in a small degree, detracts from the theoretical efficiency; but, as will be shown later, this works out to the general benefit of the proposition.
(3) The limits noted as to the degree of compression allowable in the ordinary gas engine cycle would cease to apply the fuel and air were not in contact during compression.
If, then, the fuel and air were not to be in contact during compression and the combustion must take place before expansion, the fuel must be injected at the end of the compression stroke and be ignited immediately.
The question of ignition had now solved itself to a certain extent. After the engine was warm and running, there would be sufficient heat retained from the previous charge to ignite the fresh fuel immediately when injected.
As the fuel was now to be injected into the already compressed air, and burned, instead of being vaporized, mixed and exploded, the matter of vaporizers or carburetors needed no attention, and the only requisite was a pump which would automatically measure and deliver to the engine at the proper time each working stroke, the quantity of oil required by the load.
(4) As there is no explosive question to consider, the air supply does not need to be regulated with the oil, and hence the compression pressure will remain practically constant.
After working out these ideas, to demonstrate their value, it was necessary to construct an engine embodying them. I therefore made complete working drawings for an 8x8 inch engine and set out to find capital for building the machine. After a very great deal of trouble and the lapse of several months I secured what I estimated would be a sufficient amount of capital to construct the engine try out the principals involved. But as usual, in experimental work, the original amount raised proved quite insufficient and a further considerable sum of money was expended before the success of the invention was assured.
Fig. 1 is a sectional elevation of the engine as originally designed.
Though it was originally intended to operate the engine on a two-stroke cycle, compressing the air charge alone to its full compression pressure in the crank end of the cylinder, the mechanical difficulties in working out this cycle were so serious that it was decided to omit the feature of preliminary compression and to arrange the engine to work on an ordinary four-stroke cycle.
I had designed a compressed air spraying device to break up the fuel before entering the cylinder, and also a diverging conical nozzle, which I expected would attain a temperature high enough to ignite the oil instantaneously as it passed through.
This nozzle is shown in Fig. 2. It proved a complete failure and was the first of a long series of devices intended to be attached to the cylinder head in front of the injected fuel to cause ignition and distribution of the fuel.
The method of trying out these devices was as follows: A jump-spark plug was inserted in the cylinder head, and the necessary electric connections made, gasoline was fed into the air supply pipe and the engine started up by hand on gasoline.
After a few minutes' running the ignition became automatic and the spark was shut off. The point of ignition gradually advanced as the internal ignition devices became hotter, until it was much too early and caused pounding. The peculiar fact was discovered that if more gasoline was now turned on, the pounding ceased entirely, and by feeding a heavy excess of gasoline the engine could be run until the lighter parts of the interior of the cylinder showed red hot.
During the latter part of these runs on gasoline the exhaust was as black as the smoke from a fresh-fired locomotive. When the temperature became so high that we could no longer run on gasoline, the oil was turned on through the sprayer.
For two months the engine would not continue running five minutes after the gasoline was shut off.
Much time was spent at this period in experimenting with different substances for the interior lining of the cylinder, but no non-metallic substance could be found suitable for the work.
Sprayers were tried by the dozen; all the ignition devices that ingenuity could contrive. At last, by a process of elimination, certain forms were found to be more favorable than others, certain natural laws made themselves apparent and were gradually formulated, thus establishing guiding lines for future experiments.
The first indicator cards showed a mean effective pressure of only 10 pounds. This was gradually worked up until 44 pounds was reached. At this stage the engine would continue to run, but only if the load were varied to suit it. If the load went too light the engine stopped; if it were too heavy it stopped; a variation of the load 10 or 15 per cent either way resulted in a stoppage.
At this time it was decided to make considerable changes in the design of the cylinder and combustion chamber. A new air pump was also built. The M. E. P. was then turned up to 78 pounds. My theoretical card had shown 101 pounds, and further experiments were unnecessary. Another long period of trying out followed. The M. E. P. rose by small step until 95 pounds was reached. This is high as is now reached in average work though individual cards are sometimes obtained up to 105 pounds. The highest card ever obtained from the engine showed 120 pounds M. E. P.
During all this period of development the engine would never run on a very light load. If the load were thrown off while the engine was running it would very soon begin to pound, the pounding in a very few minutes becoming so severe as to make it advisable to shut off the oil. If it were not shut off the engine stopped after a few minutes in any case.
Over 18 months elapsed between the time when the engine first ran on oil and the time when it could be started and run on no load and yet will take care of full load if thrown on at any time.
While the no-load difficulty was bad enough, the worst trouble was due to explosion waves forming in the cylinder. This trouble did not show itself until after the M. E. P. had been worked up to about 50 pounds. The external evidence of these explosion waves was a pound in the cylinder varying from a mild thump, which could easily be mistaken for a loose bearing on the shaft rod, to a terrible metallic bang, which sounded as if the whole engine was being pounded to pieces with a steam hammer.
The violence of these waves was so great that in many cases the pencil lead in the indicator would be snapped off on each side of the holder immediately when the cock was opened.
Fig. 3 is a reproduction of a card showing the effect of the waves on an indicator card, while Fig. 4 shows the shape of normal cards.
Fig. 5 shows a sectional elevation of the Johnston oil engine cylinder as now constructed of diameters of 12" and under. If the cylinder is over 12” diameter the non-water jacketed part a better made separate to allow for unequal expansion.
Note that no water jacketed surface is presented during the early part of the expansion stroke.
The cast iron liner on the piston during long runs, full load, reaches a temperature sufficiently high to make it show a dull red, but no harm results, as no strain is taken by this liner.
It has taken two years of the most tedious and discouraging work to find out why these waves are formed and how to prevent them forming. It is now possible to design engines such that explosion waves will not form, but the chance of a designer striking the right combination accidentally is less than one in a thousand.
It has generally been the opinion among gas engine experts that explosion waves are largely due to pockets in the combustion chamber, but in all engines experimented with there have been no pockets of any kind. Neither, as may seem strange to those accustomed to gas engines, does the time of injection of the fuel make very much difference, the waves form apparently almost as easily when the fuel is injected after the crank is past the center, as when it is before it.
The principal cause of failure of the previous attempts to use crude oil or crude residues in an internal combustion engine lies in the tendency for tar and other carbonaceous material to accumulate in the vaporizers, cylinders, valve chambers and other parts of the mechanism. It was freely predicted by those in authority that the Johnston oil engine would be subject to the same difficulties, and among all the experts to whom the proposition was submitted in the beginning the number who at all appreciated the advantages of the new system could be counted on the fingers of one hand.
In over three years of constant experimental work there has never been the slightest trouble from this cause, though every kind of cheap, heavy oil that could be obtained has been tested in the engine.The whole of the fuel is absolutely burned and the exhaust is clear and colorless if the engine is properly adjusted and not overloaded. It is found that a considerable overload may be carried if an extra-large supply of oil is fed; the exhaust under these circumstances is quite brown, but there is never the least tendency for deposits to form anywhere in the system.
About 90 per cent of the oil used in our tests is a fuel oil supplied by a Canadian oil company. It is a residue from crude oil after the gasoline, benzene, distillates and kerosene have been extracted, and thus is the cheapest of any obtainable fuel. It is a dark greenish black in color. Specific gravity about .875 flash point, open test 170 degrees F.
The engine has been run on crude oil from Texas, crude oil from Petrolea, cheap fuel oil from anywhere bought in the open market, kerosene, benzene and gasoline. No adjustments in the engine are necessary in changing from one fuel to another, nor is there any appreciable difference in the amount of gallons of the different fuels required to produce a given amount of power.
The first tests of fuel economy showed a consumption of 1¼ gallons per 10 H. P. hours; about six months later 1 gallon.
The consumption now guaranteed by the Johnston Oil Engine Co., Ltd., of Toronto, Can., on their 9x12 inch engine is three-fourths gallon of any petroleum product per 10 H. P. hours. The best test on record to date is a consumption of two-thirds gallon per 10 H. P. hours.
The speed regulation is very simply worked out. As no particular proportions of oil and air are required, no explosive mixture being formed, but the oil being simply burned, it is necessary to control only the quantity of oil injected per working stroke, while the air supply may be left constant. The oil pump is controlled by the governor and measures out each charge of oil to suit the load. Any type of governor may be used, and as the movement of a small cam one-quarter of an inch is the total motion necessary for complete control, the speed variation can be reduced to a very low percentage.
As noted above, the original method of starting the engine was to use a gasoline explosive mixture and an electric spark, continuing to run on gasoline until the combustion chamber was hot. There were several serious objections to this method, the principal one being the extremely violent pre-ignition explosions, which occurred just after the gasoline was shut off As explained before, when running to heat up the engine, it was necessary to feed a great excess of gasoline to prevent pre ignition. When the gasoline was shut off, the proportion of gasoline to air fell rapidly with each succeeding stroke, and when the mixture became approximately correct the resulting explosions occurring at the latter part of the compression stroke were simply terrific. The original intention had been to heat the interior of the combustion chamber in starting with a plumber's torch, and after the experimental work had progressed sufficiently so that there was a reasonable possibility of the engine starting when turned over the torch was used to raise the parts to the temperature necessary for the ignition of the first charge. This method was fairly satisfactory, but it had the disadvantage that it was necessary to remove the torch from the cylinder and plug the opening before the engine could be started. If the least time were lost in putting in the plug or in getting the engine turned over the parts would be cooled off and the whole routine had to be gone through again.
After a considerable number of devices had been tried out the "thimble" starting igniter, such as is now used, was evolved. As the name indicates, this device is in the form of a thimble and is about one inch in diameter. It is secured to the engine in such a position that some of the oil, as it is injected into the cylinder, strikes the interior walls of the thimble. In starting the engine the flame of a torch is played upon the exterior of the thimble until it is red hot, the time required averaging three minutes. The engine is then started and the torch is removed afterwards. This method is perfectly reliable and is better than any other for inexperienced operators.
As it was necessary to equip every engine with a compressor to pump air for spraying the oil, the use of compressed air for starting naturally suggested itself. This starting mechanism was very easily worked out for the second engine, and all engines now built are supplied with, an extra tank for starting. The starting tank may be pumped up to any desired point while the engine is running, and will hold its pressure indefinitely. To start it is only necessary to open a valve and move a small lever. The absolute certainty with which an engine of this type will start is a feature which appeals most strongly to gas engineers.
There are only three conditions to be fulfilled to make the starting of one of these engines certain, and these conditions are such that it is possible to determine them beforehand. It is not necessary to try to start to see whether he conditions are correct.
The three conditions are: There must be at least 200 pounds air pressure in the tanks, there must be oil in the sprayer and the thimble must be red hot. There is a pressure gauge on the tank; there is a try cock on the oil line; when the thimble looks red it is hot enough.
The latest achievement in connection with the engine is the successful application of the jump spark for starting. This is the only heavy oil engine in the world which can be started cold with a sparker. The ordinary jump-spark plug will not fire the oil, but by combining a little theory with a whole lot of practice, one has been designed which will do the work. The current is usually left on for two or three minutes when starting, after which the engine is hot enough to run without it.
To sum up the matter in a few words, the engine is now commercially perfect.
It will operate on any liquid fuel. It is not subject to either ignition or carburetor troubles. It is self-starting and reliable.
No part of the engine is subject to deterioration, with the possible exception of the igniter, which may have to be renewed in from six to twelve months, at expected the cost of a few cents.
For marine work a compressed air reverse is fitted, which is absolutely certain and positive in action.
These engines are now being built commercially by the Johnston Oil Engine Co., Ltd., in Toronto, Can., and it is expected that arrangements will soon be their manufacture in the United States. |
|
This shows a sectional elevation of the Johnston oil engine cylinder as now constructed.
1907 Johnston Oil Engine Co., Engine Indicator Cards
|
|