New Experimental Apparatus. By W. ALEXANDER, A.M.I.M.E., A.M.I.C.E.

Progress, Volume III, Issue 4, 1 February 1908, Page 123

New Experimental Apparatus. By W. ALEXANDER, A.M.1.M.E., A.M.I.C.E.

more to employ the methods that are in vogue in the practical scientific world. The principles and laws to be learned by the student are deduced, under the guidance of his teacher, from quantitative observation of phenomena shown by the apparatus arranged for the purpose. Bven mathematics are now studied in this way, due chiefly to the initiative of Professor Perry, one of I y ord Kelvin's most distinguished students. Usually the student makes experiments in the subjects that apply more directly in the work he will be engaged in in after life, but whether the subjects themselves relate to his later work or not does not matter so much as his comprehending the power of, and his ability to apply, the deductive method in connection with any branch of scientific or technical investigation he may be concerned with. The above kind of instruction requires a considerable variety of apparatus which must not be crude and give rough results, but should be well designed to satisfy the following general requirements :—: — (a). The phenomena to be observed, and the measurements to be made, should be vitiated as little as possible by disturbing phenomena or errors. (b). Adjustments and observations should be arranged to be made quickly and easily. (c). The chances of the experiment breaking down, due to mishaps to the setting or to the apparatus itself, should be reduced to a minimum. When these conditions are fulfilled several advantages accrue : — The student's attention concentrates more on the phenomena or measurements under consideration ; his time is saved, so that he does not tire of the experiment ; and he gets over more of the subject ; and altogether he takes more interest in the work and in writing up his reports and descriptions.

In connection with the above method of teaching, the various pieces of apparatus described below have been designed for the mechanics classes of the Wellington Technical School by the Director and myself. They have been made in the school workshops partly by the students and my assistant, Mr. Dolby. They form only a very small part of the total equipment necessary for a mechanics laboratory, but it is hoped that, due to possible outside aid, the mechanics and other classes will be soon fully provided for. Most of the apparatus is arranged to fit on one pattern and size of stand. In this way, comparatively few stands are required. The utility of the apparatus is not lessened thereby, but is rather improved ; and there are the advantages that, as regards the making of them, the material, labour, and cost are much reduced, as is the storage room required ; while the convenience of handling due to lessened weight is increased.

of its f aU by means of a thread. The vibrator is then started, and the plate caused to fall at the same time by burning the thread. The bright wave-trace marked on the plate by the vibrating bristle forms a permanent, continuous, and complete space-time record of the motion, and from this trace all the laws of motion of bodies falling freely, and also the gravity constant, can be deduced. A Kater's pendulum is shown in Fig. 2. It is an instrument for the determination in absolute units of the force, or the acceleration, of gravity. The large masses, M, are made in two parts and are symmetrical about their axes, so that a rotation would not affect the period of vibration. In order that the rate of swinging can be altered they are movable along the rod, and can be fixed in any position by simply screwing one portion of the mass into the other, which causes a firm grip on the rod. N is a smaller mass in the form of a milled nut that can be adjusted in position more sensitively than the larger masses, for the purpose of bringing the periods of vibration for both knife edges to exactly the same value. When this is accomplished, the common period is the same as that of a perfect simple pendulum of the same length as the distance between the knife edges, and a simple calculation will determine the required result. It is intended to make as soon as possible a nearly perfect simple pendulum, and a

bifilar suspension pendulum, for comparing one type with another. The comparison is made by causing any pair of the three different kinds to vibrate together on the same stand. For studying the laws of coil friction the apparatus shown in Fig. 3 has been devised. It is well known that when a belt, string, or rope, is lapped on a plain or grooved cylindrical piece the larger pull applied at one end, necessary to overcome a constant smaller force at the other end, increases with the angle of lap, more rapidly than in simple proportion. The precise manner of increase has been determined and represented in a formula, the constant of which varies with the material and form of band, and on the material and form of the surface on which the band laps. How the constant varies with the diameter of the surface and the thickness of the band I don't yet know, but this can be determined from the apparatus. Ihis knowledge is applied every day in the correct design of belts and their corresponding plain or grooved pulleys. It also explains why a few turns of rope wound round a post on a wharf with a small pull at the free end will quickly bring to rest a large ship alongside ; and explains why it is so easy to lower a heavy barrel down an incline into a cellar by means of a rope fixed at the top, passing

round the barrel, and held in the hand ; and so on. A is a movable arm that can be fixed in any position by the milled hand-nut, H. P is an idle cylindrical pulley with ball-bear-ings. Cis a changeable piece incapable of rotation, of variable diameter, with plain or grooved surface, and of any desired material. sis a circular piece divided in degrees. It turns with, but can be shifted relatively to, the arm when setting to zero. To set the arm for zero angle of lap, the string or band with two equal weights attached is passed over the idle pulley, and the arm is adjusted near its top position until the string just touches the friction drum, when the arm is fixed. The scale is then set to zero, which corresponds with zero angle of lap. When the arm is shifted through any angle the string laps the drum to tlae same extent, and tke scale gives the angle exactly. The efforts necessary to pull up the constant resistence, and the corresponding angles of lap, are recorded. A comprehensive set of experiments can be done with the apparatus, for all the variables that influence the result can be easily altered. The experimental crane illustrated in Fig. 4 is arranged for tests of mechanical efficiency (which are typical of the tests for other machines), and for testing the truth of the stress diagrams that give the forces acting in the members The crane with the load suspended as in the illustration gives two speed ratios— 3 and 9

to 1. By using a pulley block another two can be got — 6 and 18 to 1. Thus, four different efficiency tests can be made wfth varying load. In addition to the two ways just mentioned of suspending the load, it can be hung from the junction pin of jib and tie, so that foi any load three different diagrams can be got for the forces exerted by the jib,.« tie, and string, at the pin. In order to make the test for the stresses more general, the length of the tie can be altered by means of the screw N. B is a balance weight, which has the effect of rendering the jib and tie weightless in as much as their weight affects the stresses in them ; Sis the stress indicator, graduated in pounds. The effort, E, is applied to the lower drum while the load is transmitted to the upper drum. The pawl shown prevents the load running back when stresses are being observed. To render the crane sensitive in respect of stresses the pulleys have been made light, and large in diameter, and the pins small. The figure explains an interesting case as regards the stress in the tie. The length of the tie has been adjusted to be the same as th?t of the post, so that the total stress in the string and tie bar should equal the load, 4 lbs. Now, the string itself has a force of 4 lbs., since it supports the load over a nearly frictionless pulley. Hence, the force in the tie bar should be zero, as the indicatoi shows. A statics table is illustrated in Figs. 5 and 6. The centre part is mounted on ball bearings and has a number of brass sockets in it for taking the pins at the ends of the load-support

ing strings, or the pins which, support the beam in Fig. 5. The beam has a number of equidistant holes of the same diameter, so that the some strings and pins can be used in loading the beam as are employed foi applying the forces to the revolving part of the table when in the horizontal position. The pulleys can be placed in any position on the circumference, and can act when the table is either in the horizontal or in the vertical position. The principles of beams can be derived by experiments such as Fig. 5 shows. The principles of moments can be studied by using the table horizontally and causing the forces to act on the revolving disc, which will take up a position of rest. The student first fixes a sheet of paper to the disc by inserting two pins in the hole's The paper will rotate with the disc until the position of equilibrium is reached. Then lines are drawn on the paper to sh >w the position of the strings. Arrows and figures are put on the lines to give the directions and magnitudes of the forces respectively. The centre point is marked, when the paper is removed, in order that the calculations can be made. Fig. 6 shows a funicular (or string) polygon with the applied forces. The record is taken on a sheet of paper in the same way as before, and the structure and force diagrams drawn

out on another sheet of paper. The results obtained experimentally and graphically are then compared. For studying the general conditions of equilibrium of any body at rest under the action of forces in one plane, a light disc of wood of any shape, and provided with socket holes, is placed on the centre part of the table, and the forces are applied to the disc. The principles of beams can also be studied fiom the table in the horizontal position by using a short and light wooden beam. In all the experiments friction can be eliminated entirely by shaking the table a little on its stand. Also, due to strings being close down to the table, errors from parallax are small. Hence it is that the apparatus gives results correct to one per cent. In Fig. 7 is shown an optical lever for measuring small thicknesses. Vis a telescope provided with cross wires. 5 is a scale divided into tenths of an inch ; C is a movable collar carrying screwed post, P, with milled head for fixing collar in any position along the horizontal rod. The telescops with scale can be clamped in any position on the post. Tis a tripod, with its feet constrained by a point, line, and plane ; it is provided with a graduated screw, N, and mirror, M . The scale is reflected from the mirror to the telescope and can be clearly seen at the same time as the cross wires when the focussing has been properly accomplished. The position of the cross wires on the scale is read off. The piece being measured is then placed under the screw, N, and the new reading taken. When the collar is in the position shown, the thickness is equal to the one-hundredth part of the difference in scale readings. The scale can be read to the onefortieth part of an inch, so that the thickness is given to the one four-thousandth part. The head, A 7",A 7 ", is divided into 25 parts, and one-fourth of these divisions can be estimated. The screw has 40 threads per inch. Consequently, by turning the screw so that the scale reading comes back to the original value, the thickness is given by the number of divisions turned through by the graduated head, N, also to the one-four-thousandth part of an inch. Thus the screw forms a check on the optical lever. In having the collar movable the instrument is better from an educational point of view, for the student has to make out his own calculations in transferring from scale readings to thicknesses for an}' other position of the collar.