The tendency in the past has been to assume that when wakes or non-uniform total head profiles are fed into an axial compressor then substantially constant static pressure prevails at the entry, the variations in total head appearing as variations in velocity. This variation in velocity causes variation in incidence on the early stage blade rows and thus can give rise to excitation of blade vibration. This assumption is implicit, for instance, in References 1 and 2, but we think has been a common assumption by most of the people working in this field.

Where the compressor is fed by a duct of substantially parallel walls for a reasonable length ahead, such an assumption appeared justifiable. Such a duct when given an air flow test with its outlet discharging, for instance, to atmosphere instead of to the compressor, then the distribution assumed would normally be obtained and in fact many surveys of such ducts have been represented in this fashion. The object of this note is to show that, in fact, this distribution will not normally occur when the compressor is present and we may normally expect much more nearly a constant velocity into the compressor with attendant static pressure distributions to match with the total head variations ahead of the intake, with of course, the attendant curved flow to support the static pressure gradients.