The problem of measuring the mean rate of mass flow of a fluctuating stream is one which has attracted the notice of many investigators but, although a limited degree of success has been achieved by more than one method, it may be said that a simple, inexpensive, and compact device which is capable of dealing with flows subject to severe pulsations without loss of accuracy and without ancillary apparatus has yet to be described. Meters of the orifice type, used in conjunction with large smoothing capacities, have been widely employed, in particular for the measurement of the air consumption of reciprocating internal-combustion engines, but the smoothing capacity often requires to be inconveniently great and too small a volume will lead to surprisingly large errors (Watson and Schofield 1912; Kastner 1947)‡ whilst the range of flows covered by a single orifice is rather restricted since the rate of flow is proportional to the square root of the head. Electrical methods, depending on the transfer of heat from an electrically heated wire to the stream, have been tried (Moss 1924) but are also not free from errors when velocity fluctuations are present. Other methods adapted to particular circumstances have had some success (Bailey 1939; Cornelius and Caplan 1952) but do not represent a satisfactory general solution of the problem. So-called ‘viscous meters’ in which the pulsating stream is directed through a number of narrow passages across which the pressure difference is measured and in which the Reynolds number is kept so low that the flow is laminar (Alcock and Ricardo 1936) may be shown theoretically to have certain advantages, notably that of a straight-line law between head and mass flow even when the instantaneous velocity is subject to fluctuations. Within limits these advantages have been confirmed in practice but under conditions of severe pulsation such meters have shown large and puzzling errors. The investigation described below was undertaken with the primary aims of studying the performance of viscous meters under the most severe conditions of pulsating flow, with special reference to the measurement of the air consumption of reciprocating internal-combustion engines, and of developing an accurate meter suitable for use under such conditions. It is believed that these aims have been at least partly achieved and, whilst many of the experiments relate to engine air-flow measurements, it is considered that the results are capable of wider application and that they will be of value in a variety of problems concerned with fluctuating flows.
Read full abstract