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).
Page from 'The Autocar' magazine discussing the design, efficiency, and causes of failure of worm gears.
| Identifier | ExFiles\Box 2\6\ B002_X 140 141 157-page140 | |
| Date | 9th December 1911 | |
| THE AUTOCAR, December 9th, 1911. 1729 B. Three thread worm, 20° angle. Worm 2in. pitch, 3in. lead, 3.55in. pitch diameter. Wheel 43 teeth, 13.7in. pitch diameter. Maximum efficiency, 89% at 10 h.p. C. Four thread worm, 20° angle. Worm 2in. pitch, 4in. lead, 3.52in. pitch diameter. Wheel 44 teeth, 14in. pitch diameter. Maximum efficiency, 92% at 15 h.p. These results are not only of the highest interest by way of comparison, but they are very remarkable as demonstrating excellence of design and workmanship combined with perfection of lubrication, all resulting in the extraordinarily low coefficient of friction of about 0.02. At 31° thread angle with this coefficient of friction, a worm gear alone (excluding journal and thrust bearing losses) has been shown to have an efficiency of 97.5%. Now the stress on the wheel increases quite materially with increased thread angle, so, it is submitted, angles above about 35° are not worth while. The objection to 45° thread angle is mainly on account of its extravagant constructional cost, not in cutting the particular angle, but because the angle puts up so largely the diameter of the wheel. This can be shown by taking a concrete case. Take a 5 to 1 ratio (corresponding to 30 miles per hour on 880 mm. wheels with an engine speed of 1,500 revolutions per minute), and assume that a 1½in. pitch tooth is considered necessary. For 30° angle (actually 31°) the gear might be, say: Worm 6 threads, 2.5in. lead, 1.4in. pitch diameter. Wheel 30 teeth, 7.15in. pitch diameter. For 45° angle the gear might be: Worm 10 threads, 7½in. lead, 2½in. pitch diameter. Wheel 50 teeth, 11.9in. pitch diameter. Thus is, the wheel and casing are some 64% larger for a 45° angle than for a 31°. This comparison is not a rigorously accurate one, because, as may be gathered from the actual cases previously quoted, the output is increased very materially with increasing lead and angle, so that one would choose a smaller pitch diameter and a smaller pitch with the ten tooth than with the six tooth worm. But the conclusion remains good, and the contention is that at about 30° to 35° thread angle it pays better to concentrate on lubrication and a low coefficient of friction than on increased angle. The Essentials for Successful Working. A low coefficient of friction involves three things mainly—good and adequate surfaces and good oil. Good oil alone will not do it, nor will good surfaces alone; the two combined will not do it without the adequacy, for the laving contact is entire absence of metal-to-metal contact. The oil film must be maintained unbroken, it is the oil that must wear, not the metals. If surfaces be inadequate the intensity of pressure rises beyond that at which oil can remain between them; it is squeezed out and metal-to-metal contact ensues, with immediate loss of efficiency and rapid destruction. It is right here, on this question of adequacy, that the text books omit to give the designer any help, and where a designer, with a true knowledge of the underlying principles, has failed it is perfectly certain that he has erred on inadequacy of surfaces. Broadly, it may be stated that the pressure sustaining area of a worm gear—the capacity of transmitting effective thrust—is, other things being equal, proportional to the diameter of the worm and to the square root of the diameter of the wheel. The things that are not equal, and which modify these simple relations, are speed of rubbing contact, material and physical character of the surfaces, lubricant and permissible temperature rise—every one of which must be suitably allowed for in the design. For successful design without extravagance one must have had experience and lots of it. After that with the best of design a failure will still result from either bad material, bad workmanship, bad lubricant, or bad support. In the design of the differential cover case it is a good plan to incorporate radiating fins as is now done on cross-channel 'buses, and as to the internal arrangement, that should be as open as practicable, so as to contain the greatest possible quantity of oil and to permit of its free circulation. The Causes of Failure. The present writer has had through his hands for rectification, or has been given the data of, many worm gear failures. In every case the cause has been quite readily diagnosed on investigation. One constantly running set of 100 h.p. was found to be of good design but the material of the wheel was bad. Frequently the thread angle has been found too low; sometimes with a great enough angle the worm diameter has been too great, resulting in excessive rubbing speeds. Not rarely the surfaces have been totally inadequate, while occasionally the machining has been at fault—wrong distance wrong or alignment defective. Any maker with a reasonably silent engine and gear box who, after trial, has abandoned worm gear on the live axle has made a mistake. He has either not understood the problem, or he has been betrayed by his works—or he isn't British. He may positively like noise! A word may be said in conclusion on the hollowed type worm. In a paper presented in February, 1907, before the Institution of Engineers and Shipbuilders in Scotland, the writer said: "The so-called Hindley worm gear, practically unknown in this country, is largely used in the United States. The substitution of a hollowed worm for a parallel one, without any other alteration, is definitely stated in certain cases to have converted failures into successes. If the contention that the hollow worm affords increased pressure-sustaining contact area can be maintained, and appearances are certainly in its favor—then the departure is justified, for the highest cost of the worm will be more than balanced by the higher rating of the whole gear." To this day the Hindley worm has never caught on in Great Britain. That it is justified is sufficiently proved by its continued use on the Lanchester car. The ordinary parallel worm can be milled as a repetition process, and it can be turned in a good lathe, whereas the Hindley worm cannot be milled with any usual equipment, and it requires quite a special equipment for turning; moreover, it demands exceeding accuracy in setting, and it must not be allowed even the very least end play. It is not known that any trustworthy test results have been published showing either the efficiency of this worm or its load carrying capacity. | ||