DEVELOPMENT OF A PHOTO-VELOCIMETER FOR MEASUR- BY FIRST LIEUTENANT B. W. DUNN, ORDNANCE DEPARTMENT. INTRODUCTION. Our knowledge of the effects of powder pressure in modern ordnance will advance directly with the accuracy of our simultaneous measurements of small intervals of time and space. When we can trace accurately the change in velocity of a known mass, we can state the law of development of the resultant force acting on it. To the French is due the credit of attacking, from the standpoint of dynamic rather than static measurement, the more difficult problems that confront the Ordnance Engineer. The free recoil carriage of Colonel Sebert and the work done with it by Colonel Sebert and Captain Hugoniot suggested the best method for investigating problems in interior ballistics. The published records of experiments made with this carriage, however, justify doubt as to the reliability of data obtained from it. That different investigators should obtain different results is readily understood when practical details are examined. In the case of the 10 centimeter cannon, the total recoil during development of maximum powder pressure is about 0.02-inch and the total time about 0.001 second. It is simply impossible to obtain, with this instrument, a satisfactory time space record for these limits. Printed by authority of the Chief of Ordnance, U. S. Army. It will not suffice to use a high power microscope to translate an infinitesimal record. As in photography, the details in the original cannot be improved upon in the enlarged reproduction. The original record must be produced on a magnified scale; the pen must move more rapidly than the hand and yet in complete unison with it. Perfect conditions require a massless pen and American ingenuity has found this in a ray of light. In photography the experimenter has a new tool for his set and he can now form many new combinations which secure for him increased accuracy in all measurements and enable him to accomplish, in some cases, the previously unattainable. This new application of an old tool was introduced in 1891 by the Ordnance Department of the Army, when the Photo Retardograph was constructed.* During the same year the Ordnance Department constructed a photo-chronograph, whose accuracy may be judged by the fact that it was used, with target intervals of only two feet, to obtain five consecutive velocities of a projectile moving 1345 feet per second, and the extreme variation was only two feet per second. Figure 1 shows a reproduction of one of the records from this chronograph. Fig. 1. The next step was the introduction, some years later, of the beautiful Polarizing Photo-Chronograph of the Artillery School, an instrument which, from the wide publication of its merits, has obtained a well-deserved international reputation. In theory, at least, this instrument improved upon its predecessor, the ordnance photo-chronograph, by supplying a massles shutter to take the place of the induction coil whose secondary sparks had been photographed to show how much of the photographic time scale corresponded to the required time interval. The following paper contains some useful data, but its principal value is to be found in its illustration of the fact that, with photography added to the tools available, a simple and a cheap combination can be made for the investigation of any special problem in view-a combination which will give the investigator a sensitive, accurate and valuable instrument. When smokeless powders were introduced for small arms, See Artillery Journal Nos. 1 and 2, vol. VII. exaggerated claims were made for their non-recoil properties. The spring dynamometer, whose measurements are given in foot pounds, was the only instrument available at Frankford Arsenal for measuring recoil. The capacity of this instrument was exhausted without throwing much light on the subject, and a very interesting report by Colonel Farley was published in report of Chief of Ordnance, U. S. Army, 1893. The suggestion of the writer, that by using photography, in connection with other appliances on hand, a reliable free-recoil carriage could be improvised, was readily approved by Colonel Farley, and the experiments were made during the winter of 1895-96. GENERAL DESIGN OF APPARATUS. The special problem to be investigated was: A small arm rifle of known mass is mounted on rollers, and, excepting a small frictional resistance to be determined, recoils. freely. It is desired to know the velocity of recoil as a function. of recoil, the velocity of projectile as a function of the range and the relative positions of piece and projectile at any instant. The idea in assembled apparatus was to fasten a photographic plate to the recoiling piece, to record thereon a time scale and to use this time scale as a chronograph for tracing the successive velocities of projectile by photographing on the same plate the secondary sparks from an induction coil, produced as the projectile passed through successive screens placed at ten feet intervals. If only the motion of gun had been desired the photographic time scale would have sufficed. This simple method of attaching a plate to a body to be moved. and recording thereon a photographic time scale will give, in the majority of cases, the data required to measure the resultant of moving forces. The exact influence of any recoil check, the amount of energy accumulated in gun and carriage at any instant, the rate at which this energy is being absorbed at any point in path of recoil, etc., are illustrations of cases to which the method is directly applicable. The details of apparatus for securing the time scale should depend upon the accuracy desired. In this case the most convenient one was selected. A slow fork giving about 250 breaks per second was used as a “contact breaker" for a Ruhmkorf induction coil, the secondary spark being photographed to form the main divisions of time scale. For greater accuracy, a high fork DESCRIPTION OF APPARATUS FOR THIS EXPERIMENT.-PLATE I. and sunlight or electric light should be used to record a continuous sinusoid similar to that given by the photo-retardograph. The rifle, with surplus stock and magazine removed to offset, approximately, the extra weight of mountings and to prevent center of gravity being too far removed from axis of bore, is mounted to roll freely on a metal track in a dark box (Figs. 1 and 7). A sensitive plate, 10 inches long and about 11⁄2 inches wide, is securely attached to stock of rifle so that the plate can have no motion other than that of the rifle (Figs. 4 and 7). An attachment to the right prong (Fig. 2) of tuning fork permits its use as a "contact breaker" for the Ruhmkorf induction coil No. 1, Fig. 1. The spring, s, Fig. 2, and the screw, a, are in the electric circuit passing through magnets which sustain vibration of fork; the spring, s', and adjusting screw a' are in the primary circuit. of coil No. 1. The simplest arrangement would seem to be to put the coil in the circuit through fork magnets but this did not prove satisfactory. The breaking of contact between s and a does not occur at a definite point and with the extreme certainty and regularity required. With the fork at rest, the spring, s', and screw a, are adjusted to a delicate contact. This insures that when the fork is vibrating and the right prong is moving inward, the screw, a', will strike the light watch. spring when the prong has its maximum velocity and thus produce a quick and decided break in primary circuit. The tone of spring, s', when vibrating independently, is made higher than that of the fork so that it will not lose contact with end of screw, a2, until, on return path of prong, the spring is arrested by again coming into contact with end of screw, a', re-establishing the primary current. This arrangement works with perfect satisfaction. The breaking and making of primary current is accomplished in perfect unison with the vibrations of fork. The primary current of coil No. 2 flows successively through all the screen targets, Fig. 1. This current is broken when the bullet passes through any screen and by the arrangement shown in Fig. 5 and target No. 4, Fig. 1, the current is re-established through the succeeding screen before the bullet reaches it. The device for doing this is quite simple and gave perfect satisfaction. As the current enters the target (No. 4, Fig. 1, for example) the only path available for it is that through the screen -the alternate circuit through lower part of target (Fig. 5) being kept open by the force exerted on small pistons and springs by the tension of wires in screen. As soon as this tension is relieved, by the breaking of one or more of these wires as the bullet passes through, the piston affected moves quickly, under action. of compressed spring, across the small space separating it from the insulated metal plate under it. When it reaches this plate. the current is re-established through the next target. By this arrangement, a break of the primary current of coil No. 2 is secured as often as desired and the velocity of bullet can be measured at any number of points in the same trajectory. The electrical resistances of the various circuits are kept as constant as passible. The only one that tends to change is the resistance in primary circuit of coil No. 2, and it will be noticed that the arrangement described throws in the resistance of only one target at a time. The secondary circuits of both coils, well insulated, are led to the terminals in lens mounting shown in Fig. 4. A photographic record is secured of each secondary spark produced during recoil over first ten inches. There is practically no limit to the length of record obtainable. record. Fig. 8 illustrates the The origin sparks are made by breaking each of the primary circuits once, by hand, just before firing the gun. The tuning fork is then set in motion and, to prevent a great number of time sparks being recorded at the origin of time scale, between the turning on of current through coil No. 1 and the beginning of recoil,-thereby injuring that portion of negative. by solarization-a simple shunt for coil No. 1, which is broken by first motion of gun in recoil is used. The advantage of this shunt is questionable. It served its purpose and improved the appearance of negatives; but more extended use of apparatus will probably show that, to secure best results, the time sparks should be turned on before recoil begins. After the stream of sparks is once established a condition very nearly approaching a vacuum exists in the small space (about o".03) between terminals. The resistance offered by this space to the first spark should therefore be comparatively large. The Leyden jars connected with secondary circuits (Fig. 1) are also experimental. Without them, the secondary spark is comparatively dim and of long duration, its photographic record on a plate in motion being a line (see Fig. 1 page 2.) With them, |