An accurate knowledge of turbine blade surface temperature is desired in order to obtain maximum performance from turbine engines. A limited spectrum radiation pyrometer can be used for blade temperature measurement. A model is presented which predicts the output signal from the detector unit of a pyrometer in a turbine engine application. Six inputs are required for the model. The inputs are the turbine blade geometry, location of the pyrometer with respect to the blades being viewed, focusing parameters of the pyrometer, type of detector, transmission curve of the optical system, and an estimate of the blade surface temperature. The model uses Fortran 77 and IBM CADAM to create a three dimensional representation of the pyrometer path across the blades along with the intercepted target spots. Once the target spot areas are determined, the photocurrent output signal of the detector is predicted as a function of percent blade chord and time. Results are shown for different detectors and temperature distributions. Experimental data is also included, and a comparison is made between the data and the model. Any of the model input parameters can be varied so that different pyrometer schemes can be evaluated at either the initial design phase or after installation. / Master of Science

Past research at Virginia Tech (VPI) explored the aerodynamic loss of the transonic VPI turbine blade, which 1s based on the pitchline profile of a high pressure turbine blade for a large commercial aircraft gas turbine. The current experiment explores the loss of the VPI blade for different axial solidity ratios near the design point. Ten percent changes in the solidity ratio were accomplished by varying the blade pitch and changing the blade stagger to maintain a constant throat to spacing ratio. Reaction, exit angle and exit Mach number were kept constant with this method. Cascades with three different solidities were tested in VPI’s transonic blowdown wind tunnel. Downstream total pressure loss and static pressure measurements were obtained. In addition, inviscid calculations were made for each case. Static pressure contours and Mach number profiles from the calculations were compared with the experimental results. A ten percent decrease in solidity caused no cascade loss penalty as compared to the Baseline solidity for a wide range of Mach numbers. Calculated blade Mach number profiles agreed well with experimental profiles except on the suction side near the throat and downstream of the shock/boundary layer interaction. Predicted downstream static pressure values agreed well with experimental values, except that the inviscid code tended to over-predict the pressure rise across the suction side shocks. / Master of Science