ABSTRACT
The new construction of large-scale wave flumes throughout the world has stimulated interest in the simulation of laboratory-generated random waves. Either analog or digital methods may be employed to generate the signal required to activate the wavemaker in order to obtain the desired type of random waves. Fourier transform (FFT) algorithms are frequently used to represent periodic-random types of digital-to-analog simulations. The measured data signals were digitized through analog-to-digital converters simultaneously with the activation through digital-to-analog converters of the motion of the wavemaker piston by the FFT-synthesized wavemaker signal. The sonic wave profiler was calibrated by displacing the sonic transducer through a discrete sequence of vertical distances measured above the still water level in the absence of waves and by correlating the sonic transducer output with these measured displacements. Each measured spectrum was smoothed by a box-car filter using discrete spectral estimates which gives 26 degrees-of-freedom for discrete spectral estimates computed from FFT coefficients.
