Control of ball-racket interactions in rhythmic propulsion of elastic and non-elastic balls.

Ball-racket interactions were investigated in a task where participants propelled a ball rhythmically into the air. The study contrasted two ball-racket conditions: (1) an elastic impact where the ball was able to rebound due to the elasticity of the colliding objects and participants bounced the ball, and (2) a non-elastic impact where the coefficient of restitution was zero and the ball did not rebound such that the participants had to throw the ball. The goal of the study was to contrast the situations where haptic information about the ball-racket interactions is either secondary (elastic bouncing) or becomes a primary factor for control (non-elastic propulsion). In the elastic condition, the performers controlled the parameters for ball-racket contact prior to contact: In agreement with the criteria for dynamical stability defined by a model, racket accelerations immediately before the contacts were negative, racket positions and velocities at the instant of the initial contact correlated negatively, contact durations were short (30+/-9 ms), and during the collision interval racket velocity and acceleration decreased monotonically. In the non-elastic condition, the parameters of ball release were primarily controlled during the collision phase: Racket accelerations before contact were positive, racket positions and velocities at initial contact showed weak correlations, and the contact intervals were significantly longer (116+/-15 ms) with a clear segmentation into two segments. Negative correlations were observed between the integrals of the velocity and acceleration computed over the two consecutive segments, giving evidence that in the non-elastic condition the CNS is able to introduce corrections during the very short collision interval. The results are discussed with respect to physiological mechanisms of movement corrections available during such short time intervals.