Abstract

Pneumatic multi-hole probes are an indispensable tool for the investigation of flows in many applications. Non-intrusive measurement techniques require optical access to the flow. This is often difficult to achieve, e.g. in turbomachines. Therefore, flow measurements with pneumatic probes are widespread in turbomachinery studies. The geometry of turbine stages mostly prohibits use of the probes in nulled mode of operation, i.e. aligned with the flow. Thus, probe calibration in a known flow field is necessary. As there are a variety of types of pneumatic probes, a universally applicable standardized calibration procedure does not exist.

The subject of this paper is therefore the development of calibration procedures for two specific probe designs used in turbomachinery measurements. The aim is quantification and reduction of measurement uncertainties. Uncertainties related to pressure measurements, probe alignment and the probe traversing system are taken into account. The calibrations take place in the wind tunnel SKG of German Aerospace Center (DLR). This closed loop wind tunnel allows independent variation of Mach and Reynolds number through control of mass flow and total pressure.

The calibration procedures for the two probe types, one with slanted, planar faces at the head and one with a smaller, cone shaped head, are presented. Both probe types are equipped with a backpressure hole to increase Mach number sensitivity in the transonic flow regime. The pronounced edges at the head of the larger probe type are intended to serve as defined lines of flow separation. However, a comparison of Reynolds number dependency showed no clear advantage of either design. Mach numbers are varied from 0.1 to 1.3. Reynolds numbers with respect to the probe head diameter range from 700 to 14,000. Flow angles up to 32° are considered. Pressure coefficients are monitored during all stages of the calibration. Standard procedures for waiting times between the end of probe movements and the recording of calibration data are derived for both probe types. The necessary waiting time is found to be mainly dependent on probe type and total pressure. This is attributed to the longer time needed for the wind tunnel to reach the desired flow conditions for larger probe heads and lower total pressure. In these cases, a longer waiting time must be chosen. Moreover, an evaluation procedure using values averaged over a certain time span is presented.

The effect of pressure measurement uncertainties on the accuracy of probe measurements is found to be heavily dependent on Mach number and total pressure. Monte Carlo simulations with probe pressures randomly drawn from probability distributions were carried out. The application of this approach on the actual evaluation algorithm is time consuming and computationally expensive. Therefore, a simplified mathematical model was developed and implemented in GUM Workbench, a commercial program for uncertainty analysis. The results of both models are in satisfactory agreement for large parts of the calibration domain. However, a considerable discrepancy was found for low Mach numbers where the simplified model predicts far larger uncertainties.