In the course of NASA evaluations of the statistical discrete gust method for determining worstcase gust loads on aircraft, it was found that the concept of the matched lter from signal processing theory could be applied to the analysis of time-correlated, worst-case gust loads for mathematically linear aircraft models. Much of the work on the matched lter approach to gust loads analysis that followed had been geared toward a fundamental understanding of the implications of the matched lter theory interpretation of the gust loads analysis problem, as well as various attempts at extending the techniques to nonlinear aircraft models. A rigorous development of the matched lter interpretation of gust loads within the context of the theory of random processes is presented and includes discussion of its relationship to existing statistically based methods for the calculation of design gust loads for linear aircraft. It is shown, using the matched lter theory formalism, that the load peak-to-rms ratio can be related to the energy of the matched excitation waveform. Also, it is shown that a candidate design value of the peak-to-rms ratio that has been reported in the literature to be around 3 may have its basis in the discretization process that accompanies both data reduction and stochastic simulation on digital computers.
NomenclatureA[ ] = temporal average Ā = load-rms-to-gust-rms ratio, Eq.
(33) E[ ] = ensemble average G(v) = gust frequency response function H(v) = system frequency response function h(t)= unit impulse response K = arbitrary constant, Eq. (3) P( y) = cumulative probability distribution of random variable y p( y) = probability density function of random variable y R(t) = correlation function t, T = time variable, interval U x,T = energy of waveform x(t) over time interval T U s = design gust velocity= normalized load hd = load peak-to-rms factor r = correlation coef cient s = standard deviation t = time F g (v) = power spectral density of atmospheric turbulence c = normalized load v = circular frequency, rad/s