The collapse of spark-induced cavitation bubbles in a flowing system was studied by means of high speed photography. The migration of cavitation bubbles toward a nearby solid boundary during collapse and rebound was observed. Near its minimum volume the bubble typically formed a high speed microjet, which struck the nearby surface causing individual damage craters on soft aluminum.

The dynamics of high-speed impact between a compressible water droplet and a rigid solid surface is investigated analytically. The purpose of the study is to examine the mechanism leading to the erosion of a material due to liquid impingement. A Compressible-Cell-and-Marker (ComCAM) numerical method is developed to solve the differential equations governing the unsteady, two-dimensional liquid-solid impact phenomena. The method is designed to solve this unsteady portion up until the flow reasonably approaches the steady-state solution. The validity of the method is confirmed by comparing its numerical results with the idealized exact solution for the classical one-dimensional liquid impact problem. The accuracy of the numerical reresults is found to be very good in that only slight numerical oscillations occur. Viscosity and surface tension are neglected as seems resaonable with the relatively large drops and high velocities considered. Pressure and velocity distributions are solved as a function of time. The deformation of a drop is also recorded for three different shapes: cylindrical, spherical, and a combination of the two, which may more closely model the actual droplet shapes to be encountered in such impacts. Typical liquid impact Mach numbers of 0.2 and 0.5 (sonic velocity referred to water) were studied. Thus impact velocities of about 980 and 2450 fps are considered. Compression predominates during the early stages of the impact, while rarefaction governs later, during which time the radial lateral flow velocity exceeds the initial impact velocity. The reflection of compression waves and the lateral flow leads to the possibility of cavitation within the drop, due to the consequent generation of negative pressures, exists. The maximum pressure calculated in this two-dimensional liquid impact problem is found to be less than the one-dimensional maximum pressure for all three different droplets in various degrees. It is found that droplet shape impact angle and liquid impact Mach number are the only important parameters of the problem for a flat fully-rigid target surface. As more time elapses, i.e., up to 2–3 μsec for a 2.0 mm-dia drop, the maximum pressure shifts from the center of the contact area radially outward, while the pressure at the center attenuates rapidly toward conventional stagnation pressure.

Cavitation damage to specimens of stainless steel, carbon steel, aluminum, and plexiglas, placed in a cavitating venturi using water and mercury as test fluids is mostly in the form of irregularly shaped pits which do not change with additional exposure to the cavitating field within the limited durations utilized. The rate of damage is very high initially, decreases for a relatively short period of time, then increases again up to the maximum test durations of 150 hours with water and 270 hours with mercury. Observation of damage effects by several independent techniques, using a variety of specimen materials, with two different fluids under various fluid dynamic conditions, leads to a suggested correlating model in terms of the cavitation bubble density and energy and specimen material strength.

The impact process of a cylindrical, a spherical and a conical water drop upon a rigid plane is studied with numerical methods. Time variation of drop cross-section, pressure and velocity distribution are presented. The maximum pressure remained at the center of the contact area for the cylindrical drop, but expanded outward with the edge of the contact area for the spherical and conical drops. The peak pressure for both cylindrical and conical drop reached approximately the value predicted by a one-dimensionless analysis. The pattern of pressure distribution on the liquid-solid boundary compared favorably with the existing experimental data.

A comprehensive set of cavitation damage data has been obtained in a vibratory facility using water, mercury, lithium, and lead-bismuth alloy as test fluids, and covering temperatures ranging from room temperature to 1500 deg F. Materials tested include a wide variety of metals and alloys. From this data a simple, reasonably precise, damage predicting equation has been derived, including only ultimate resilience as a material property, but also corrections for cavitation “thermodynamic effects” and NPSH. It has been found that of the conventional mechanical properties, ultimate resilience is the most successful in this regard. A direct comparison between venturi and vibratory cavitation damage shows that the relative rankings of materials remain about the same for mercury, and a good correlation is obtained between the mercury data from the venturi and ultimate resilience. Neither statement applies for the water venturi data, possibly because of the greater effects of corrosion in the low intensity cavitation field.