The Quest of the Fool-proof.
Page 8
Page 9
If you've noticed an error in this article please click here to report it so we can fix it.
The Continuation (from page 531) of the Story of the Evolution of an Oil Pump and of the Investigation which its Production Entailed.
By Henry McLaren.
Wouldn't Pump Hot Oil.
In due course the first batch of 25 pumps, made to the new specification, were sent in, and one pump was taken and tested with hot oil, and worked well. Orders were then given for all engines to be fitted with these pumps. Within the next few months we began to receive complaints from our outside men, that some of the pumps would not work, whilst others worked well. On inquiry in the works we found the same thing. if a pump did not work well during the testing of the engine, it was returned to stores, and another tried, until a satisfactory one was found. When I again had to start a personal investigation into this oil-pump question, I found that a very large percentage would not work except at light pressure. I, thereupon, had a special rig made for testing them. The pump, fixed in a shallow bath of water kept at boiling point, soon picked up sufficient heat to keep the oil at a constant temperature of about 200° Fahr. Then provision was made for working the pump at all speeds from 1 to 180 strokes of the pump plunger per minute. The results were surprising. I found that none of these rejected pumps would pump against a pressure of 10 lb. with oil at 200°, although some would give a good pressure with cold oil. Just then a new batch of 25 pumps had arrived from the makers, so we determined to test every one, and reject all those that would not pump against a pressure of 250 lb., with oil at 200° Fahr. Only one stood the test, and it would pump up to 300 lb. None of the others could cope with 200 lb., and some could not pump against a higher pressure than 5 lb. per sq. in., though all of them would pump with cold oil. So we sent 24 pumps out of the 25 back to the pump-man, and told him to make them right. We continued our experiments with the good pump, and with our wasters from the stores. it was easy to see from the short time that it took to fill the oil bath, that the whole of the oil pumped was leaking past the plunger ; the leakage increased as the pressure was raised, until it came exactly to balance the amount pumped. The well-fitted plunger maintained a higher pressure than the badly-fitted ones, but in both cases all the oil leaked away past the plunger. 3-10,000ths of an Inch Wear in Six Years.
Meantime we had calculated that each pump would require to make 500,000 strokes for a year's work of the engine, so we started our good pump on an endurance test, making 180 strokes per minute, and pumping against a pressure of 160 lb., and giving a slight side pressure to the plunger, in order to cause wear. After making two million strokes, equal to say four years of work, we tested it as before at three strokes per minute, and oil at 200°, and found that it would not maintain 80 lb. We put it back and ran for another million strokes (or a total of six years of work). We found, on testing again, that we could only pump against 45 lb. pressure. As this pressure was no use to us, we stopped the test, took out the plunger, and gauged it. Where it had not gone in the hole, it was, of course, of the original diameter, and to our surprise was only worn 3-10,000ths of an inch where it worked in the barrel. This 3-10,000ths of an inch, on a plunger 5-16 in. diameter, makes all the difference between a really good pump and a waster, if it has to pump hot oil at slow speeds of the plunger. Meantime we had been trying our hands at making plungers fit better than those of the pump-man, but here we failed utterly. Not one of ours would pump against any considerable pressure. After some weeks the pump-man sent the 24 pumps back, and told us he had spent no end of money trying to meet our requirements. On testing them we found some much improved, but many of them were still quite useless for our pressures, so our stock of wasters kept accumulating. By this time I was in despair of ever being able to make a reliable pump for hot countries ; I was thinking of going back to the screw-plunger type.
The White-Metal Dodge.
Just then I thought of trying the white-metal dodge, for packing the plunger. This consists of drilling a hole (V), at least twice the diameter of the plunger hole, cross-over through the centre line of the plunger, then putting the plunger in place, and running in white-metal (W) for about in. on each side of the plunger. This metal contracts round the plunger, making a perfect fit ; but at the same time it contracts away from the walls of the cross hole (V), and allows the oil to leak round the outside of the packing. To prevent this we compress the white metal, after it has cooled, causing it completely to fill the hole (V), and thus to prevent leakage. Fig. 3 shows the method of applying the pressure to the white metal, through two steel plugs (XX1). After these plugs are taken out, the holes are filled up with metal, to make a finished job. [We must here refer our readers to the diagram which had to be included with the first portion of the article.—En.] 200 lb. Pressure at 2000 Temperature.
We applied this packing to the worst of the waster pumps we had in the stores, and we found that we could then pump up to practically any pressure. We put this pump on for an endurance test, and ran four million strokes, or eight years of work. We tested it with hot oil at 200°, at three strokes per minute, and we found we could still pump an ample supply of oil against a pressure of 200 lb. per sq. in. On careful examination we found the plunger slightly worn at the top end, near the crosshead, where it worked in the cast-iron part above the packing. This was no doubt due to the crosshead tending to rock the plunger in a fore-and-aft direction. We then decided to open out the plunger hole (Y), down to the cross hole, and to run the packing like an inverted T, so that the top end of the plunger would be wholly in contact with white metal. This was done, and the pump run for 12 million strokes, equal to 24 years of work ; it was afterwards tested, and it maintained a pressure of 300 lb., pumping a full supply of oil at a temperature of 200° Fahr., at three strokes per minute.
The Last Trouble, an Air-lock.
Our troubles were now nearing the end, but one more brought me on the job again. It was found to be very difficult to get these packed pumps to start pumping, but when once they started they went along first rate. After a great deal of trouble, I found out that this reluctance to start was in a measure due to our good packing, and also to a fault in the pumpman's plungers, He left them short of coming down into the pump chamber at Z, even when at the bottom end of the stroke ; so when the pump was first filled with oil, a blob of air was always trapped under the plunger at Z. If it happened that the plunger was at the top of its stroke, twice the amount of air was imprisoned, and owing to the packing's fitting so well this air could not leak away, but simply compressed and expanded alternately, and so prevented the pump from starting. Nothing short of taking the plunger out, filling up the hole with oil, and getting the end of the plunger on to solid oil would clear out that air. Afterwards I got the pump-man to lengthen his plungers, so that they should come right down into the pump chamber, and thus displace any air that might be trapped under the plunger. This proved quite effective, and now I think our troubles with oilpumps are over.
Praise to the Pump-man.
I cannot close without a word of praise for the pump-man. He stuck to it till the end, but I do not suppose he made much profit out of that first hundred pumps. He came again to see how we had got. over the difficulty of making the plungers fit, and eventually we granted his firm the licence for making these packings for oil pumps. Fig. 4 shows our method of altering the stroke of the lever (R) without taking out pins, etc. The clamp bolt not only secures the clamp to the valve spindle, but also takes the bottom end of the connecting link. As this clamp is moved to or from the pump, the number of teeth is reduced or increased respectively from one to four teeth per revolution.
The pump-man made a very good double steamcheck valve, fitted with two ball valves. When I saw our men starting up a new pump, undoing the coupling to see if the oil was coming through, I had the small bleeder valve fitted under the check valve, so that by opening it any intelligent driver could see if his pump was pumping. This is the finish of a long tale about a little thing. but, when every small detail of our engines has had as much consideration as this oil pump, they will be a little more than "fool-proof." I venture to think.