The latter has been found to he diminished by having a considerable thickness of water inter vening between the outside of the revolving wheel and the in side of the sta tionary casing. In the p shown there is only a very nar now clearance space at the sides of the wheels ; hut here unusual care b as been taken in con struction, th e wheel turned and 11.18,410 perfectly true after being keyed on the spindle.
The resistance greatly increases if the wheels are not perfectly true ; but up to the present time the data respecting friction in such eases are meager. As the water enters through one side only of the wheel, it causes a thrust in that direction which is equivalent not to the force of the suction only, as is generally supposed, but to the area of the inlet multiplied by the maximum pressure of the discharge. l'he pump being inverted, with the suction inlet at the top, the entering water flows downward, and the reactive force is consequently upward. The upward thrust, which in most cases would be objectionable, is here turned to practical account for supporting the weight of the vertical driving shaft and the pump-wheel. The plan of inverting the pump so that the suction enters at the top was introduced in California by the writer in the latter part of 1883, and was then believed to be of great importance, because of the difficulty of supporting the vertical driving shafts by other means in the deeper pits. In the case of one pump, completed in 1886, the weight of the shaft and its attachments was nearly 1,000 lbs. The shaft was of steel, 21- in. diameter, and ran at 600 revolutions a minute. The upward thrust was sufficient to carry this shaft, together with some additional weight which was found necessary. The lift was 00 ft.; inlet of pump, 10 in. diameter ; throat of discharge, 5 in. diameter ; uptake pipe, 10 in. diameter. This problem of thrust upon enclosed wheels taking water at one side is an intricate one. If the rear side of the wheel is exposed, as is common, to a pressure equal to the discharge, the thrust, as already stated, is equal to the inlet area multiplied by the discharge pressure. If the wheel is shrouded on one side only, the thrust will be equal to the whole area of the wheel multi plied by the discharge pressure. At starting there is, of course, no upward thrust until the pump is charged. Provision is, therefore, made at G for carrying the shaft on collars, which are already required for steadying the revolving wheel laterally in the pump casing, and arc so arranged as to support the shaft vertically for a short time, unassisted by the water thrust.
The collars are screwed upon the shaft, and several thin washers of steel are inserted between them and the seat which carries them. They run in a, pool of oil, or rather oil and water, because there is generally a small pipe leading a little water back from the discharge pipe, D. to the thrust box, C. The joint thus formed seals the pump, taking the place of a pack ing gland. The suction pipes, .5S, are shown as they are commonly arranged. for branches leading in from right and left ; their large area is intended to be equal to that of a number of branch plies, and to keep the flow in all at a uniform rate as nearly as possible.
Compound Centrifugal of the main problems to be dealt with in applying centrifugal pumps to high lifts are how far the impact or mechanical push of the vanes may be disregarded as a factor in the pump's duty, and how the bearings and driving gearing may be maintained in proper order at the high speed required.
Practically the speed at which the pump should be driven increases as the square of the height of lift. For example, the circumferential speed of the revolving wheel for a lift of 60 ft. will be at least six times as fast as the discharge column should flow ; while for a head of 80 ft. the circumferential speed for the same flow would have to be more than ten times that of the discharge current. It is, therefore, seen in how rapidly increasing a degree the revolving wheel must overrun the flow as the lift increases ; and how rapidly the effect (Inc to impact or mechanical push of the vanes falls off, as the velocity of the wheel increases. For lower lifts the extent of overrunning diminishes in the same degree, and the gain by impact is increased accordingly. It is easy to attain high efficiency in centrifugal pumps working against a low head ; but it is a difficult matter to arrange such pumps suitable for working in the deep pits in California, against a pressure of 43 lbs. per sq. in., or 100 ft. total lift, and to secure results that are satisfactory. Thus far it has not been possible to make experiments for determining definitely the efficiency attained in these high lifts. From such observations as have been made it would seem that from 35 to 45 per cent, of the indi cated power has been realized in water raised.