Wind tunnel measurements in ski jumpers and simulation of the jumps

Abstract

Ski jumping performance can be divided into different phases as follows: gliding, takeoff, transient flight, actual flight, flare-out and landing. It has been suggested that in the phase of transient flight the drag should be at its minimum, during actual flight the lift-to-drag ratio a tits maximum, and in the flare-out phase the lift at its maximum. The purpose of the tunnel measurements was to investigate ski jumpers attempting to reach their minimal drag of the gliding position in the up-hill and individual maximal lift-to-drag ratio and achieve the optimal flying position with a low pitching moment coefficient for the said phase according to the feeling from their sensor motor feedback system. The jumpers were attached to an overhead three-component platform-balance with a modified "seat" and belts on the abdominal side of the hip. The system was pivoted close to the center of gravity in order to allow the jumper to adjust the angle of attack. A 3 dimensional flight posture model with sixteen points was created for the measurements. In the flight phases when the lift-to-drag ratio was at its maximum, the individual averaged positions for selected sample times were calculated for the angle of attack of skis, lower legs and upper body. The sweep angle of the skis (V - angle ) and the distance of the ankle joints were also calculated. The deviation was large in all measured variables. The height and "a real mass ratio" of the jumper had a positive effect on the lift-to-drag ratio. Positive regression coefficients for the lift-to-drag ratio in respect to the average values were found in the angle of attack of the skis, trunk and V-angle. The low pitching moment indicated that it was relatively easy to find a stable flying position in the wind tunnel. Therefore, wind tunnel conditions could be useful in training basic flying technique for a longer period of time. A modified Aquila simulation program for the center of gravity of the jumper-equipment system was developed to work in Excel. The input variables were as follows: air density, initial velocity at starting gate, mass of the jumper-equipment system, friction coefficient, aerodynamic reference area, takeoff force, aerodynamic parameters during gliding phases on the track and in the air, sensitivity of lift-to-drag ratio for head and side winds, profile of the jumping hill starting gate position and head and side wind velocity. The output variables were as follows: the tangential and normal component of the velocity at the takeoff table, resultant velocity, release angle, instantaneous x-, y- and z- coordinates, v x-, v y- and v z- components and resultant velocity as a function of time. The length of a jump was defined as the distance travelled when the path of the jumper's center of gravity intersected the profile the of the hill. In a hill of K = 120 m, the relative difference between the simulated and measured jumps was on average 2.8 %. The largest difference between the measured jump and simulated one could be 10 - 20 m, if the wind velocity was between 1.0 m s-l following and 3.0 m s-1 head wind. It can be concluded that the simulation model of center of gravity was in agreement with real jumps.