SINCE air-cleaners of the radiator-fan type cannot be tested satisfactorily by the older method, in which a known weight of dust is fed directly into the airstream entering the air-cleaner, a special method was found necessary in which the air-cleaner under test is mounted in its normal position behind a radiator fan that is located inside of an elliptical wind-tunnel within which the fan circulates air. A tractor engine running at constant speed and load drives the fan in the wind-tunnel and draws the air for its carbureter from the wind-tunnel through the air-cleaner under test and an absolute air-cleaner connected in series.
A 100-gram charge of a standardized dust is introduced into the wind-tunnel. By averaging the results obtained from repeated tests, using three different collecting-type dry centrifugal air-cleaners, it is found that under normal conditions 15 per cent of the total dust-charge actually reaches the air-cleaner under test in the described apparatus. Therefore, 15 grams is the basis on which the efficiencies of dust separation are calculated. With one exception, the air-cleaners tested showed rather low efficiency. The vacuum or restriction effects, measured by the usual open U-tube manometer, were found low and were unaffected by accumulation of dust except in the case of two air-cleaners. The test method and the apparatus described can be used to test air-cleaners of practically any type, but the accuracy is less than that of the older method. Descriptions of the air-cleaners, tables of summarized data, and numerous observations are presented.