An experimental study of ventilation of turbo alternators

WHEN designing a large steam turbine-driven alternator, the volume of air per minute needed to cool the machine may easily be calculated after having estimated the losses and after having decided upon the value of the temperature increase of the air while passing through the machine. The pressure delivered by the generator thereof, (the fan) can be predicted with reasonable accuracy from data to appear in an early paper². On the other hand the distribution of the air, and the pressure needed to drive it through the various paths are too difficult to predict with our present knowledge of the laws governing the flow of air. In any of the recognized systems of ventilation, there are combinations of various series and multiple paths, and in practically each path there are changes in direction or section, and at each change there is a loss of head. In order to design large generators intelligently, it is important to predict what the quantitative relations are within a reasonable percentage of error. The only feasible way to determine them at present is by means of experiment, and that, too, becomes involved.