STATISTICAL GROUND EXCITATION MODELS FOR HIGH SPEED VEHICLE DYNAMIC ANALYSIS

The method used to statistically describe a digitally defined ground roughness profile, based on measured data, is outlined and illustrated by reduction of the survey data of the Tracked Levitated Research Vehicle's (TLRV) Initial guideway section. The power spectral density (PSD) is the statistic of prime interest. Its use as the ground roughness excitation function in linear vehicle dynamic response analyses provides, the impetus. The PSD is not unique; therefore, when calculated for use in such analyses, the vehicle speed and suspension natural frequencies become critical parameters. The statistical estimates of the PSD and autocorrelation (AC) functions based on this TLRV data indicate a reasonable fit with mathematical models which consist of Fourier transform pairs. These PSD models contain less energy at low frequencies than the model presently used. A first-order vehicle dynamic response analysis incorporating one of these models results in substantial reductions of suspension stroke response. The roughness coefficient (A) of this section of thw TLRV guideway is approximately 1.2 x 10 to the minus 6th power feet-radians.