A

over the contact points Q1, the circuit is closed and this is shown on the paper by a break in the line traced by the pen L1. similar arrangement is used to record the number of revolutions of the turbine. The eccentric R, on the turbine shaft T closes the circuit at Q2, and the pen L2 records a break in the line for each revolution. The third pen L, is connected to the time circuit, every 12 second being recorded on the chart.

The method of procedure in making a turbine test is as follows: sufficient time is allowed for the discharge to become steady, after which the brake is adjusted. At a given signal the

3

diaphragm is then dropped into the water, the switches O,, O2, Og on the electric circuits are closed, and the recording drum is set in motion. The time, number of revolutions of the turbine, and the velocity of the diaphragm are then recorded automatically upon the chart, but the elevation of the water surface in the canal must be observed at gages H and I, Fig. 7. Vertical lines are then drawn on the chart at the beginning and ending of the gaging distance, which cut off the corresponding time and number of revolutions. Thus, for the diagram shown in Fig. 8, the time in traveling over the gaging distance of 32.8 feet was 13.37 seconds, and the number of revolutions of the turbine during that time was 21.51. Hence,

The recording chart moves at a sufficiently high rate, in this case 0.033 feet per second, so that an error in scaling the chart is inappreciable. A reproduction of the recording chart showing four runs is shown in Fig. 9, and a view of the recording apparatus in Fig. 10.

The discharges at this station were formerly measured by current meters, as the head was too small to permit using a weir. A current meter gaging at 8 points in the cross-section took from 2 to 3/4 of an hour, during which time it was difficult to keep operating conditions constant. The time for computing the result is also an important factor, and if one considers that sev eral measurements of the discharge must be made for a complete turbine test, it may readily be seen what a saving in time the diaphragm method accomplishes. A complete test of a turbine can be made in half a day with the new method, and the results are considered quite as accurate as with any of the usual methods employed in the testing of turbines.

DESCRIPTION OF THE DIAPHRAGM APPARATUS AT THE BERLIN TECHNISCHE HOCHSCHULE

In the testing of turbines at the hydraulic laboratory of the Technische Hochschule at Berlin, the diaphragm method of measuring discharge has now been used for a number of years. The facilities for the installation of this method were not very favorable at this place, as the tail-race canal (see Fig. 11) is built of brick and is only 33 feet long, but nevertheless an instrument was built which gives satisfactory service and accurate measurements.

Description of Car and Diaphragm.-The structural details of the diaphragm (see Figs. 12 and 13) and recording device differ somewhat from those designed at the Voith testing station. The frame work of the car is made of very thin steel tubing, brazed together and although weighing but 88 pounds is very rigid. The diaphragm is built of light angle-iron covered with oiled canvas and is hung from the horizontal axis A, Fig. 12, about which it is free to swing in a forward direction. It can also be raised or lowered by two small cables attached to the two hand wheels N, the guides K sliding along the two vertical tubes T. A brake B, Fig. 13, is attached to one of the wheels so that the speed with which the diaphragm is lowered can be easily regulated. Its descent is limited by the two rubber buffers P. When in a vertical position the diaphragm is held rigid with the vertical frame by means of the clutch R, and when the clutch is released the current swings it around the axis A to the dotted position shown in Fig. 13.

The bottom and sides of the canal were carefully plastered with cement mortar, so that the clearance between the diaphragm. and walls is only about 0.2 to 0.3 of an inch. A distance of

about 10 feet is necessary for immersing the diaphragm, so that the actual length of the gaging distance is only 23 feet. Experience has shown, however, that this distance is sufficient with the rapid vertical immersion of the diaphragm.

Method of Procedure.-The method of procedure in making a discharge measurement is as follows: the car is placed at the upstream end of the canal, with the diaphragm raised but locked in a vertical position with the sliding frame. At a given signal

the diaphragm is dropped, its descent being controlled by the hand brake, so that it falls gently against the rubber buffers. These buffers are set so that just a small clearance occurs between the diaphragm and the bottom of the canal. As soon as the diaphragm is partly immersed the car begins to move, but the diaphragm usually reaches its lowest position before the car has traversed 3 or 4 feet. After a distance of 10 feet the motion of the car is so uniform that the remaining distance can be used for determining its velocity. As soon as the car reaches the end

of the canal it is held by two bumpers which also release the clutch; the diaphragm then swings about the axis A, the frame is raised and the car is pulled back to the starting point. As the entire operation can be completed in but a few seconds, several

measurements can be made as a check while the turbine is held under constant conditions. The depth of water in the gaging channel is read several times during a run by means of a floatgage located in a recess of the canal. The back water occasioned