1. This flow unsteadiness could also impact on the techniques presented in this paper in the following manner. Results for each time (or azimuthal) step are calculated from the average of 1024 voltage samples from each sensor. To the extent that the calibration equations are nonlinear, the more correct method would involve calculation of the flow parameters for each of the 1024 samples followed by averaging. (i.e., a function of the mean is not equal to the mean of the function when the function is nonlinear.) The use of the computer would allow this sethod to be used. For the results presented in this paper, this potential error was negligible because of tne relatively steady profiles noted during data acquisition. Aeronautics Series, Volume 54, (1977). 2. Weyer, H., and R. Schodl: Development and Tcsting
2. ments Downstream of a Moviq Blade Row. ASblX Paper 74-GT-3, (19973). 5. takshminarayana, B. and A. Poncet: A Nethod of