Abstract
The so-called ultrasonic shot-peening is a new process using calibrated peening hard balls motioned by walls vibrating at 20000 Hz. As in conventional shot-peening, a fully peened surface coverage is required tor the process to be efficient. However, it is intricate to determine the coverage ratio experimentally. A statistical approach, issued from the Random Sets Theory, was used to establish relationships between coverage, impact superficial density and impact crater area. The random modelisation was tested with different materials treated by ultrasonic shot-peening. Individual crater diameter measurements and local repartition analysis were made using quantitative image analysis to test the model. Calculations provide the mean number of impacts everywhere on the surface for any coverage ratio. This statistical approach can also be applied to any impact treatment involving an impact flow.