From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Technical discussion on the merits of different engine lubrication systems.
| Identifier | ExFiles\Box 16\1\ Scan018 | |
| Date | 12th October 1910 guessed | |
| to avoid the expense of a pump. It could be used without sight feeds with very little extra trouble to the driver, and is a very safe system. Another good way of ascertaining the oil level is to have a large pipe—some 2in. in diameter—coming out from the side of the crank chamber near the fly-wheel, and so as to be visible and accessible through the space between chassis and platform, where the front wing joins the platform. The pipe should have a screw cap, and be at such a height that the right oil level is just below its top. An objection is that it gives a correct indication only when the car is level. Another ingenious system, adopted in an imperfect form on the Ballot engines of the Mass cars, is perhaps the safest yet devised. The oil pump draws from the crank chamber at the right lever and delivers into a tank on the dash. This has an open connection to the crank chamber, so that the oil runs back, but the pump can draw faster than the oil can run back, so the level is kept always right. If the pump fails the engine suffers from over, and not from under lubrication. A tap can close the pipe when the engine is stopped. The trough system is really one for keeping the level right. Oil is pumped into the troughs and overflows into the sump, so that the troughs are always full. I first heard of this system on the Siddeley cars, but it rapidly came into common use. It does away with the safety of the splash system, because the supply of oil in the troughs is not enough to continue lubrication for long if the pump fails. In the present Wolseley cars the oil is pumped into reservoirs above the main bearings, and also into the troughs. The arrangement is shown in Fig. 2. I think it would be better if the oil pipes were led into the main bearing so as to feed them under pressure, the pipes to the troughs being restricted sufficiently to give some pressure. Except that the lubrication depends on the pressure supply, this is, to my mind, the most perfect system. The pressure-feed is where it is most wanted, and where the throw-off can be kept out of the cylinders, so that any quantity of oil can be forced through the bearings, and the splash level is always right for each big end and under all conditions of running. When the splash level is in the crank chamber itself, this has to be divided up into two or four compartments, otherwise when going up or down a hill the cylinder at one end smokes, while that at the other runs out its bearings. This means providing two or four means of ascertaining the levels, unless, as is sometimes done, small holes are made connecting the compartments—a device which merely increases the length of hill a car can go up. Another objection to the splash system with the oil in the crank chamber is that the amount of splash varies so much with the speed of the engine. The splashed oil covers all the inside surface of the crank chamber and runs back into the pool at the bottom. When the engine is running continuously at a high speed, it is capable of splashing the oil up quicker than it can run back, so that the oil is kept at a level at which the bottom of the big end just touches it. When the engine runs slow, the level is considerably higher. The level, therefore, cannot always be right; in fact, I have been told by one manufacturer that he arranged for a higher level to be maintained in cars that were required to run in France, where high-speed runs can be kept up for a long time. The supply to the crank chamber to keep up the splash level is sometimes taken to the main bearings and sometimes straight to the crank chamber. In one case it is taken to the main bearings and passes through the crank chamber to the big ends. Such an arrangement is, I think, bad. The oil ways in the crank chamber are necessarily small and are easily blocked up by dirt or congealed oil. Such oil ways should be supplied only under pressure. The same applies, though to a lesser degree, to the supply to the main bearings. Where, as is common, a single drip feed supplies all the bearings, one branch may get blocked up and no indication of the blocking will appear so long as the others remain open. There seems little object in taking gravity feed pipes to the main bearings, as it is easy to arrange channels to catch the splash and lead it to the bearings. Next comes the question of supply in the splash system. The simplest arrangement, which is still adopted on certain small cars, is to have an oil tank on the dashboard, either in front or under the bonnet, the position of the tank giving the necessary head. More generally the supply tank is on the chassis level, or is a sump in the crank chamber. In the former case exhaust pressure or a pump is used to force the oil to the drips or the bearings; in the latter a pump is necessary. In one case, a hand air pump is used to keep pressure on the oil tank. This offers a good scope for forgetfulness and leakage. Exhaust pressure is so unreliable that it should never be employed unless the petrol also is raised by it. This at any rate gives an effective indication if the pressure falls. In any case the exhaust pressure is extremely variable, and gives a variable supply of oil where the supply depends on the pressure. For these reasons the use of oil pumps has become more and more general. These, when used for a gravity feed, generally supply sight-feed drips, and it is remarkable how the use of these still continues in connection with a pump when the supply from them depends on the viscosity of the oil, and varies in consequence so much with its temperature. As indicators, they are not nearly so easily seen as a gauge. In regard to the regulation of the supply of oil, a pump forcing direct to the bearings supplies approximately a given quantity per revolution. Splash systems with the oil in the crank chamber probably supply less per revolution as the engine speed increases. Those with the oil in troughs probably supply more per revolution as the speed increases. Panhards during the last two years have adopted a system of supply under which the supply increases as the work done by the engine increases. This is arranged by joining the chamber into which the sight drips fall to the induction pipe, so that as the work increases, whether the engine goes fast or slow, the suck in the induction pipe increases the oil supply. After the sight feed the oil passes to a second pump which forces it to the main bearings under pressure. The system is ingenious, but I greatly doubt whether the engine wants oil in proportion to the work it does. I am inclined to think it wants so much oil per revolution irrespective of the work done. The system with double oil pump, in which the oil is forced to sight feed drips and after passing these is forced to the bearings, has become very common in the last two years. It seems to me a very complex way of getting an imperfect result. Sight feed drips are required only to keep a splash level right, but they cannot keep it quite right, and in this respect the system is imperfect. A trough system with pressure supply does the same thing perfectly instead of imperfectly; there is only one pump and there are no drips to regulate. In conclusion, my own opinion is that the only two systems which are really good are :—(1) The force feed into main bearings and through the crank shaft to the big ends, arranged with a relief valve near the bearings and large pipes, so that there is always a good pressure in the crank shaft and no fear of the ways blocking up. (2) A splash system with automatic means of keeping the level constant (not a constant supply), and preferably with force feed to the main bearings. The trough system is the best I know of, but greater safety would be secured if the splash reservoir were larger. Of the two systems I prefer the latter. In the force feed system I think provision should always be made for using splash lubrication in the event of the pump failing. | ||