Probabilistic performance-based geometric tolerancing of compressor blades

The relationship between tolerances of statically measured geometric parameters and the aerodynamic performance of a sample of manufactured airfoils is investigated in this thesis. The objective is to determine which geometric parameters are the best discriminators of performance, and how these best discriminators are affected by changes in design methodology, manufacturing precision, and desired minimum performance levels. A probabilistic model of geometric variability for a three-dimensional blade is derived. Using this geometric variability model, probabilistic aerodynamic simulations are conducted to analyze the variability in aerodynamic performance. Tolerance optimization is then applied, in which allowable ranges for each geometric parameter are determined so as to maximize the quality of the accepted blades. Optimization is performed at several performance limits to observe how the effectiveness of tolerances and the best geometric discriminators of perfromance change with performance limit. A set of compressor blade data is used consisting of an original geometry, a deterministic re-design, and a probabilistic re-design. Using the geometric variability model, the variability is also artificially increased to investigate the impact of manufacturing precision on tolerance effectiveness. Results shows the best geometric indicators of performance are leading-edge thickness and measures pertaining to the curve quality of an airfoil. While design can affect the performance of tolerances, tolerances are no less effective on probabilistic designs than deterministic designs. Additionally, manufacturing precision affects the best geometric indicators of performance but tolerance effectiveness is not affected. Thesis Supervisor: David L. Darmofal Title: Associate Professor of Aeronautics and Astronautics

[1]  John G. Proakis,et al.  Probability, random variables and stochastic processes , 1985, IEEE Trans. Acoust. Speech Signal Process..

[2]  Mehmet Imregun,et al.  A Probabilistic Analysis of Single-Degree-of-Freedom Blade Vibration , 1993 .

[3]  Paul J. Drake,et al.  Dimensioning and Tolerancing Handbook , 1999 .

[4]  Jack L. Kerrebrock,et al.  Aircraft Engines and Gas Turbines , 1977 .

[5]  K. Nara.  Simulated Annealing Applications , 1999 .

[6]  David L. Darmofal,et al.  On the Aerodynamic Design of Compressor Airfoils for Robustness Under Geometric Uncertainty , 2004 .

[7]  R. Preisendorfer,et al.  Principal Component Analysis in Meteorology and Oceanography , 1988 .

[8]  Sharon L. Padula,et al.  Probabilistic approach to free-form airfoil shape optimization under uncertainty , 2002 .

[9]  Meherwan P. Boyce,et al.  Axial-Flow Compressors , 2006 .

[10]  Wei Chen,et al.  Methodology for Managing the Effect of Uncertainty in Simulation-Based Design , 2000 .

[11]  Víctor E. Garzón,et al.  Probabilistic aerothermal design of compressor airfoils , 2003 .

[12]  David L. Darmofal,et al.  Impact of Geometric Variability on Axial Compressor Performance , 2003 .

[13]  W. B. Roberts,et al.  Axial Compressor Performance Restoration by Blade Profile Control , 1984 .

[14]  C. Y. Wu,et al.  Arbitrary Surface Flank Milling of Fan, Compressor, and Impeller Blades , 1994 .

[15]  V. Rich.  Personal communication , 1989, Nature.

[16]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[17]  Patrick N. Koch,et al.  A MULTI-STAGE, PARALLEL IMPLEMENTATION OF PROBABILISTIC DESIGN OPTIMIZATION IN AN MDO FRAMEWORK , 2000 .