108 Garfield Avenue Madison, New Jersey 07940
How many times have we all watched the man on the engine operate the injector to fill the boiler? There would be some steam coming from the overflow followed by some water then that unmistakable sound of an injector working. Steam from the boiler was being used to put water into the same boiler. Isn't that perpetual motion? Not quite, as we shall see.
The principles involved and their discovery go back many years to perhaps as far as Bernulli and Venturi. If we are not careful, however, we'll be having Adam and Eve as the inventors of the atom bomb. So let us begin with outlining just what it is that the injector does. That is, let's dissect the steps.
First, it must draw water from the supply tank and bring it to the injector. We can call that priming although many will recognize it as lift. Then the pressure is increased to something above that of the boiler onto which the water is to be injected. For the first part, the prime or lift, we can thank the Marquis Mannuary d'Ectot, who in 1818, was granted a French patent for a steam jet apparatus for pumping water from one height to another. Or, said another way, from one pressure to another.
His device contained all of the components that we now associate with an injector. That is, a combining tube and a delivery tube. But he was not thinking of a boiler feed device. It was Henri J. Giffard that put it all together and was the one to receive a patent from the French government in May, 1858, for a boiler feed device. At first his development was looked upon with skepticism for it was thought that he was trying to invent a perpetual motion machine. Truth did triumph and he received an award from the Academy of Science in Paris; their mechanical grand prize for 1859.
But just what had prompted M. Giffard to develop such a device? But of course it was for his airship! True. Giffard is the inventor of the dirigible. The steer able balloon or dirigere from the Latin. On September 24, 1852, he flew his 143 foot long by 39 foot diameter cigar shaped balloon powered by a 3 horsepower steam engine weighing 350 pounds some seventeen miles. As we all know, it really was necessary for internal combustion engines having much greater horsepower to weight ratios to be developed before flight was to become completely practical. But, as Giffard was a man of great technical capability it is fitting that he should be the one credited with devising a boiler feed device of such physical simplicity but of such thermodynamic complexity.
The need must have been acute for the device took the industrial world by storm. Patent rights were taken up by Messrs. Sharp, Stewart & Company in England in 1859. By 1860, William Sellers & Company of Philadelphia were offering them for sale from their factory. Although the basic working parts of the injector have remained the same as in the original device, there have been many improvements added over the years. Injectors for specific services such as railway locomotives were built to quite different designs than those for, say, a traction engine. The ones used even today on portable boilers and our favorite traction engine are probably the simplest, trouble free, device of anything used around steam.
Although they appear simple in construction, don't let that be deceiving. The design of the nozzles; the taper, their diameter and the relative diameters between combining and delivery tubes is critical. The diameter of the steam tube outlet has to be larger than the delivery tube, or the injector will not work. There are at least two popular explanations of the mechanism of operation. But these are simplistic explanations. There is only one true thremo dynamic and hydraulic explanation which is all tied up in Bernulli's Theorem and which at this stage of my career I am not certain that I could go to the blackboard and go through again.
We can, however, look at the at E letting atmospheric pressure lift the water from the supply thus filling from the supply pipe A through valve B enters the steam nozzle where it expands thus reaching a high velocity at C. When velocity increases in a closed system then pressure must decrease and vice versa. So there is a vaccum formed at E letting atmospheric pressure life the water from the supply thus filling the space with water. The steam from the jet is condensed and joins with the supply. The ratio here is about one pound of steam to 13 pounds of water. The impact of the steam increases the velocity of the water. As the mixture passes along the enlarging passage of H its velocity decreases and therefore its pressure increases. By selecting the various dimensions correctly it is possible to achieve a pressure at the end of the delivery tube that is greater than boiler pressure and therefore a flow will take place into the boiler.
Perhaps this is the time to add a few numbers to the discussion to help with the understanding of the operation. We can take the case of a Case (J. I., that is, with pun intended) where 120 pounds pressure is common. At this pressure the velocity of the steam in the jet at C would be around 1,400 feet per second at the time valve B is first opened. But, as steam flow is established and a vacuum is formed of perhaps 20' then the velocity could reach 3,500 feet per second. By the time the steam and water have combined and have passed through the expanding tube H the combined velocity will be around 170 fps. Let us suppose that we ran boiler water through a nozzle and did so at the 120 psi pressure. We would find the water velocity was only about 133 fps. We have, therefore, a margin between the velocity of the feed water and the velocity equivalent of the static water in the boiler. QED, the injector works. And it is not a perpetual motion machine. The differential velocity is, shall we say, wasted. Or, put another way, there is no loss of heat units, they all go back into the boiler. But we have changed the conditions and we have not used the total potential. Should you wish to look it up in your old Funk & Wagnall's, it's spelled entrophy.
If instead of an injector we had installed a simplex reciprocating steam pump for our boiler feed then the seeming paradox would not exist. We could have the boiler feed itself by simply letting the steam cylinder have a larger diameter than the water cylinder. The water side pressure would be higher than the steam side by the ratio of the areas of the two cylinders. Now, then, think of the diameter of the steam jet in the injector and that of the inlet to section H of the diagram. When the injector begins to work there is always some water in the overflow just before the water flows to the boiler. When steam is first condensed by the lifted water there is a water piston formed in the condensing tube being pushed by steam issuing from a nozzle of greater area. Think of these as the simplex pump. It is just another way of our explaining the operation.
The next time that you watch an Iron Man start his injector, just remember, it is a technological spin-off from the aviation field. We have come a long way in the past hundred or so years on the basic underlying theories developed by the men of early science.