Under the Supervision of Professor Edwin J. Crosby

     An apparatus was developed to produce controlled collisions between pairs of drops moving in free flight through a gas. With the aid of high-speed motion pictures and photographs, collisions between water drops ranging from 150 to 900 microns in diameter were studied. Pairs of drops with diameter ratios of 1:1, 2:1 and 3:1 were investigated for which the relative velocities ranged from 30 to 730 cm/sec and the directness of collision ranged from head-on to barely grazing. Electric charges on the drops were minimized and the humidity in the collision chamber was maintained near saturation.

     The conditions governing the various outcomes of collisions for six different categories of size were determined and mapped on plots of relative velocity versus impact angle with the impact angle being a measure of the directness of collision. Coalescence, rebound and disruption were found with each combination of drop sizes. Drops with diameter ratios of 1:1 and 2:1 also underwent reflex disjection while drops with a diameter ratio of 3:1 experienced partial coalescence. A sixth outcome, spatter, was recorded in collisions between drops of more than 1 mm in diameter. Composite photographs illustrating each of the observed outcomes were prepared.

     Coalescence was found to be favored by a decrease in drop size and, at relative velocities above 100 cm/sec, by an increase in diameter ratio. For the smaller pairs of drops, all collisions at relative velocities below 100 cm/sec resulted in coalescence. The delay time between apparent contact and the development of a detectable liquid bridge between the colliding drops was found to increase with an increase in either impact angle or drop diameter. The delay time was observed to decrease rapidly as the relative velocity was increased.

     Qualitative explanations were suggested for some of the collision results observed. A theory was developed to predict the conditions for transition from coalescence to disruption. The implications of the experimental results for collisions occurring in a random manner were considered briefly.

Research equipment