ISBN 978-3-8439-4529-5

978-3-8439-4529-5, Reihe Luftfahrt

Max Hufnagel
Nondimensional Erosion Model Considering Aerodynamic and Mechanical Particle Size Effects

124 Seiten, Dissertation Universität Stuttgart (2020), Softcover, A5

Zusammenfassung / Abstract

Aircraft jet-engines are exposed to environmental particle mixtures featuring broad particle size distributions. Impacting particles damage compressor blades. In order to forecast the local material removal, experimentally calibrated erosion models are required.

Solid particle erosion experiments were performed using a sand-blast type test rig operating at high-pressure compressor conditions. The investigated materials were quartz sand impacting on flat plate specimens of Ti6Al4V. The quartz sand featured particle size distributions corresponding to standardized Arizona Road Dust grades A2, A3 and A4. The jet impingement angle and the jet free stream velocity were varied in the ranges of 15° to 90° and 275m/s to 400m/s, respectively. The particle impact conditions were investigated using a combination of high-speed shadowgraphy measurements and CFD simulations. A nondimensional description of the material removal process was derived. Aerodynamic particle size effects were successfully identified and quantified. With decreasing size, particles are increasingly prone to be affected by the wall-near flow field, causing particle deceleration, deflection and entrainment. As a consequence, particle kinetic energy and impact angle deviate from the free-stream conditions. Thus, at nominally constant free-stream conditions, the sand-blast type test rig produces inhomogeneous particle impact conditions for varying jet impingement angles and particle size distributions. The conventionally measured erosion rates were corrected for these aerodynamic particle size effects.

It was shown that uncorrected wear data lead to a substantially biased erosion model, since the influence of particle size on erosive wear is blurred by the aerodynamic effects. Hence, it was concluded that test-rig dependent flow-effects have to be taken into account in order to ensure model transferability from the experiment to other flow conditions.