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).
The Bethlehem mechanical torque amplifier, a device for precise control of mechanical motions.
| Identifier | ExFiles\Box 154\3\ scan0061 | |
| Date | 26th November 1923 guessed | |
| Bethlehem Torque Amplifier A device for controlling accurately with a small expenditure of energy various types of mechanical motions THE Bethlehem mechanical torque amplifier (Nieman patents) recently developed by the Bethlehem Steel Company is a mechanism provided with a control shaft at one end and a work shaft at the other, the control shaft requiring only a feeble torque to operate it in either direction, while the work shaft yields a torque of sufficient amount to accomplish any purpose desired, at the same time accurately following the control shaft in all its angular movements. Its broad purpose is to perform the functions of an electrical or hydraulic servo-motor, but so different are its operating characteristics that it should be considered as an entirely new type of control apparatus. There are three elements to the torque amplifier: "work shaft," "control shaft" and "drive shaft." The drive shaft is driven by any outside source of power, such as an electric motor, and revolves continuously in one direction, a single motor being capable of operating a number of different amplifiers whose operation is entirely independent. The control shaft is actuated by any mechanical or manual control means, or by such weak forces as can be delivered through electrical recording instruments or telemetric transmission systems. The work shaft is directly coupled to the work to be done, as for instance the elevating or training gears of a gun, a ship's rudder or the steering wheel of an automobile. The control shaft can be freely revolved in either direction, with only a small amount of effort; the work shaft maintains at all times its angular synchronism with the control shaft, and in addition exerts a heavy torque to overcome outside resistance. The lag between the work shaft and the control shaft is so small as to be inappreciable in any practical application. Measured in angle it never amounts to more than a degree or two of arc, and this is true even if the control shaft is actuated in the most erratic manner; that is, with sudden changes in speed or quick reversals of direction. It is, in fact, utterly impossible by any means, even if the outside source of power is shut off, to get the control and work shafts out of step. The work shaft follows slow or rapid motions of the control shaft with absolute smoothness, and the internal parts of the apparatus operate noiselessly. The device has a definite maximum speed depending on the speed of the drive shaft. If attempt is made to revolve the control shaft faster than this maximum speed, it will meet with a positive resistance which will hold its speed down to the maximum regardless of the torque applied. Unlike other forms of servo-motor, there is always a definite ratio between the torque applied at the control shaft and that delivered at the work shaft, this ratio being known as the "amplification." This ratio may, however, be made as high as desired. In devices already constructed it ranges, according to the design, from 1:10 up to 1:50,000. Fundamentally, the torque amplifier consists of two oppositely rotating drums provided with friction bands which may be brought into contact with the drums through actuation of the control shaft, this frictional contact causing the friction bands to exert pressure on the work shaft. The force applied to the control shaft is thus enhanced, or amplified, when it reaches the work shaft by the extent to which the bands are urged forward because of their frictional contact with the rotating drums. In its simplest form the amplifier is shown in Fig. 1. Two oppositely rotating drums A, B, are driven by a motor through gears integral with the respective drums; C is the control shaft and D the work shaft. Fig. 2 is a section through Fig. 1 at the line M-M and viewed in the direction of the arrows; while Fig. 3 is a longitudinal section through both drums. The control shaft C is concentric with work shaft D and is supported by a socket bearing in D.{John DeLooze - Company Secretary} Integral with C is the forked control arm E which passes through a hole in the side of hollow shaft D, the size of this hole being sufficient for clearance and to allow a considerable angular movement of C in relation to D.{John DeLooze - Company Secretary} Keyed to D is work arm F, also forked. The friction bands G and H fit inside the drums A and B, respectively. These bands have sockets at each end, one socket of each band engaging with a stud on the work arm F, the socket at the other end of each band engaging with the respective ends of the control arm E.{Mr Elliott - Chief Engineer} The bands are disposed in opposite directions in the two drums so that in each drum the band extends from the control arm around to the work arm in the direction of rotation of that drum. It follows that if the control arm is revolved in a given direction it will expand and tighten the band which lies in the drum rotating in the direction of the movement given to the control shaft, at the same time contracting and loosening the band in the other drum. [2] etc., provided that the gearing is doubled, one set for driving in each direction. Application of the lashlock to a screw-feed is illustrated in Fig. 8. Here the same scheme is followed as with the gearing. Two nuts, A and B, run on a screw-shaft. The wedge, C, is mounted on a stem, D, which is supported by the housing of the driven member, while rollers E and F are attached to nuts B and A respectively. The wedge as it is pressed forward by the spring, G, urges the nuts apart. When the screw-shaft is turned in one direction one of the nuts forces its roller against one side of the wedge, a frictional binding is created, and through the stem the drive is transmitted to the driven member. Reversal of the screw-shaft produces a drive through the other nut, in the opposite direction. There is no lost motion during the reversal, nor can any binding due to unevenness of the threads on either screw-shaft or nuts strain the members. Screws usually become worn irregularly because one portion is used more than the rest. This fact makes it impossible to set a fixed adjustment for the sake of taking up backlash in the worn portion, since this would cause binding and strain when the nut traveled from the worn to the newer portion. The lashlock, however, will take up the backlash in the loose portion and nevertheless work freely in the tight portion, adjustment occurring automatically. It must be noted, however, that if the nut travels from the newer portion to the worn portion, without any reversal being made during this travel, and if reversal is then made while the nut is entirely on the worn portion, there will be detectable a certain amount of play during the moment of reversal while the wedge is adjusting itself, since the spring cannot push the wedge downward until pressure is relieved. At first sight this would seem to limit the accuracy of the device in screw-feeds, but this is true to only a limited extent, for the lashlock always takes a central position and therefore averages irregularities in threads and gear teeth; so that in moving from a new to a worn portion the over-all error will only be half as great as it would be with a fixed adjustment perfectly set for the newer, unworn portion. In the extremest cases the action of the lashlock is superior to the ordinary method of hand setting, which is to take up all the backlash on one side. In the ordinary operation of the machine the above described effect is no drawback at all. Applied to a piece of machinery—as the cross-feed of a lathe—the lashlock screw-feed makes it entirely unnecessary to turn the feed handle backward a fraction of a revolution in order to take up backlash. If a machinist overfeeds his carriage half-a-thousandth of an inch, he can correct the error by moving the carriage back a half-thousandth and stopping there—he will be set with absolute accuracy, and his carriage will be held at that point positively, with no tendency for the tool to jump forward into soft spots in the work. In the vertical feed of a planer the same precision of feed results and, further, the tool will not drop down when riding over a hole in the work or at the end of each back-stroke both of which are sources of chatter. The same beneficial effect occurs with the feed of a grinder. In actual construction the lashlock is combined with a feed-screw in a compact design of great strength. For thrust bearings—to eliminate end-play in a shaft —the lashlock installation is similar to that in a feed-screw, with the threads and nuts replaced by collars. Two collars are keyed to the shaft a slight distance apart; two other collars are mounted inside the first two, kept from revolving with the shaft but permitted a small endwise movement. Each of these two inside collars carries a roller so mounted as to press against a side of the wedge, the stem of which is attached to the fixed support which bears the thrust. The spring will then urge the inside collars outward against the collars fixed on the shaft; the shaft will not be able, of course, to move endwise in either direction because to do so would mean moving the wedge by pressure on only one roller. Bearing metal is set between the collars and lubrication effected as usual. As the bearings wear, the wedge works inward, taking up this wear and holding the shaft always in the same position, endwise, as at the start. Great flexibility is possible in designing the lashlock to meet situations, for the principle is capable of an astonishing variety of forms, some of which appear to [7] Fig. 1—Simple form of torque amplifier. Fig. 2—Section M-M through Fig. 1. Fig. 3—Longitudinal section through Fig. 1 Fig. 8—Pinion, with lashlock of double cam type. The pinion is split, one half driving in each direction. Lashlock mechanism may be positioned at a distance from pinion, or in case of a large gear wholly within periphery of teeth Fig. 9—Model of gun drive gearing. Backlash in the large gear during a reversal of direction of the knob is less than six seconds of arc | ||