MICROPARTICLE IMPACTS AT ULTRA-HIGH VELOCITIES - THEIR RELATION TO MACROPARTICLE IMPACTS

In recent years the Hypervelocity Microparticle Impact (HMI) project a t Los Alamos has utilized electrostatically accelerated iron spheres o f microscopic dimensions to generate ultra-high velocity impact experi ments to about 100 km/sec, about an order of magnitude beyond the data range for precisely controlled impact tests with ordinary macroscopic projectiles. But the extreme smallness of the micro impact events bri ngs into question whether the usual shock-hydrodynamic size scaling ca n be assumed. It is to this question of the validity of size scaling ( and its refinement) that the present study is directed. Impact experim ents are compared in which two comparable impact events at a given vel ocity, a microscopic impact and a macroscopic impact, are essentially identical except that the projectile masses and crater volumes differ by nearly 12 orders of magnitude---linear dimensions and times differi ng by 4 orders of magnitude. Strain rates at corresponding points in t he deforming crater increase 4 orders of magnitude with the size reduc tion. Departures from exact scaling, by a factor of 3.7 in crater volu me, are observed for copper targets--with the micro craters being smal ler than scaling would predict. This is attributed to a factor 4.7 hig her effective yield stress occurring in the micro cratering flow. This , in tum, is because the strain rate there is about 10(8)/sec as compa red to a strain rate of only 10(4)/sec in the macro impact. The measur ement of impact craters for very small impact events leads to the dete rmination of metal yield stresses at strain rates an order of magnitud e greater than have been obtained by other methods. The determination of material strengths at these exceedingly high strain rates is of obv ious fundamental importance. Results are compared to recent theoretica l models by Follansbee, Kochs and Rollett. Finally, the problem is add ressed of predicting crater sizes in a target material with strain rat e effects. First some basic results are recalled pertaining to the lat e stage equivalence of hypervelocity impacts. It is then seen, for a s train rate dependent material, that the curve of dimensionless crater volume versus impact velocity is replaced by a family of curves, each member of which is for one final crater size. The spacing of the curve s is determined by the stress versus strain properties of the material .