Numa Manson on velocity deficits and detonation stability

This memorial paper pays tribute to Professor Numa Manson’s contributions to the understanding of detonation velocity deficits and wave stability. Manson and his colleague Guenoche postulated that a velocity deficit exists in a tube because the chemical reactions are inhibited in a thin layer adjacent to the tube walls. The hydrodynamic theory of detonation was modified to account for this, and it was shown that the deficit varies inversely with the tube diameter. Manson and his students measured detonation velocities in tubes of various diameter. An estimate of the detonation velocity for an infinite tube diameter was obtained by plotting the velocity against the reciprocal of the tube diameter, \({\phi^{-1}}\) , and extrapolating the line through the data to \({\phi^{-1}=0}\) . The relative contributions of tube geometry and surface roughness to the deficits were systematically studied. Manson was also one of the early investigators to shed light on the cellular structure of detonation by reporting “vibratory phenomena” seen as striations in streak schlieren photographs. An attempt was made to relate this phenomenon to “dispersions” in the propagation velocity and hence the wave stability. The author has extended Manson’s work by investigating detonations in tubes with yielding walls. Whereas boundary layers were responsible for the gasdynamic expansion and deficits in Manson’s rigid tubes, it was the moving boundaries that caused similar effects in the author’s investigations. The author has repeated the “nozzle” analysis of Fay and Dabora using the detonation cell length as the relevant chemical kinetic length scale, and found reasonable agreement between his experimental results and the model. When the Poitiers data are reinterpreted in light of the modified model, the trends are described quite well. More recent studies have shown that the measured deficits for mixtures characterized by irregular cellular structures do not agree with the Fay–Dabora model. Possible reasons for the discrepancy are discussed.