Damage prediction of rotating blades using displacement residuals

This paper presents an optimization methodology for prediction of damage in rotating blades. Blade damage is often characterized by relatively high amplitudes of response due to a local reduction in stiffness factors. An explicit relation between the amplitudes of response and stiffness reduction coefficients can therefore be conveniently developed. Based on the available response amplitudes along various points in a harmonically excited blade with modal data, a displacement residue is defined in terms of unknown stiffness reduction coefficients. These coefficients are predicted by minimizing the norm of the residual vector using genetic algorithms. The approach is illustrated with a finite element model of a viscously damped rotating blade of aerofoil cross-section.