Roland Frederick Koontz

Abstract: The experiments described in this thesis were initially prompted by interest .. , the radiation loss of relativistic electron rings passing through periodic structures. Later the same experiments became relevant to the theory of energy loss of electrons in large storage rings. In both of these cases energy loss to the higher order modes of the respective structures could seriously limit their effective operation. In these experiments, single bunches of electrons with intensities up to 7 x 10S electrons per bunch are accelerated through the SLAC three-kilometer accelerator, and their energy spectra are analyzed. Early experiments over a wide energy range (900 MeV to 19 GeV) demonstrated that the energy loss was proportional to the total charge m the bunch but was im$,ependent of beam energy. The average energy loss of a single bunch normalized to 10 electrons was initially measured to be 38 MeV. While this work was starting, E. Keil at CERN, Geneva, Switzerland, was developing a theory and a computer program based on cavity modal analysis to identify the higher-order modes which exist in a cavity array and to calculate the total energy delivered to these cavities by a passing relativistic electron bunch. The average energy loss predicted by Keil's theory was in reasonable agreement with our early experiments . Later, more refined experiments at SLAC shed significant additional light on the physical radiation lose process, showing how the position of the electron bunch with respect to the accelerating wave affects the results. This prompted G. Loew at SLAC to devise a semiempirical analysis of the problem for which a computer program was written by R. Early and B. Woo. This analysis not only yields the average loss for the entire electron bunch but can also give the energy loss as a function of time within the bunch and the resulting energy spectrum. The only additional element that was necessary to complete this theory was the function giving the response of the SLAC three-kilometer cavity array to a delta-function beam excitation. This function was supplied by P. Wilson and K. Bane at SLAC, who had obtained Keil's program in order to apply it to the design of storage ring cavities for the PEP project, and who simply performed a modal sum in the time domain of all the accelerator modes. G. Loew's theory, which now incorporates the Wilson-Bane function, gives very good agreement with our measured results. …