Thèse de doctorat
Résumé : This thesis addresses the experimental investigation of turbomachinery flows by means of particle image velocimetry (PIV). While PIV is an attractive investigation approach due to its non-intrusiveness and capability to offer whole-field velocity information, the geometrical complexity and unique operating conditions pose various challenges to implementing this technique in turbomachinery test rigs. The main objective of this thesis is to verify the feasibility and capability of PIV in turbomachinery-related applications and provide guidelines for the successful implementation of this measurement technique. The experimental test campaigns took place in three test facilities at the von Karman Institute for Fluid Dynamics: The L-12 low-speed wind tunnel equipped with an active turbulence grid, the S-1/C high-speed wind tunnel hosting a linear low-pressure turbine cascade, and the CT-3 short-duration high-speed rotating test rig fitted with a single high-pressure turbine stage.The measurement on the active turbulence grid presented the core capability of PIV combined with the fluorescent particle. The synthetic jet injected into the main flow from the grid was distinguished by seeding the fluorescent particle to the jet and a typical non-fluorescent oil particle to the rest.The PIV measurements on the transonic linear cascade in the S-1/C wind tunnel were carried out at Re_out,is= 70k – 120k, M_out,is = 0.70 - 0.95, matching the typical operating condition of low-pressure turbine stages employed in modern geared turbo-fan engines. The measurement on the upstream blade-to-blade (B2B) plane involved a parametric study on the data acquisition and processing settings, showing a high sensitivity of measured TI to the particle displacement. The measurement on the blade passage B2B plane demonstrated the performance of PIV in a transonic flow with high-velocity gradients and cross-validated RANS prediction and 5HP measurement results. The stereoscopic PIV measurement on the cascade outlet plane (COP) investigated secondary flows near the cascade endwall. The measured three-component velocity information characterized the turbulence field at the COP in terms of TI, degree of anisotropy, and anisotropy invariants.Lastly, the PIV measurement on the high-pressure turbine stage in CT-3 was conducted as a feasibility study of PIV in a short-duration rotating test facility. The region of interest was on a tangential plane at 38%, 50%, and 58% of the channel span, covering a full rotor pitch and 5 - 15% rotor axial chord downstream of the rotor trailing edge. An evaluation of convergence history concluded a minimum of 200 pairs of snapshots were required to achieve a converged phase-lock-average flow field, manifesting the necessity of a high-speed PIV system for the blow-down test rig whose testing time lasts only 250 ms. The measurement results reasonably agreed with the RANS prediction, suggesting the prospect of PIV in such challenging turbomachinery applications.