Abstract
The behavior of timing belts used in automotive applications have to be defined and predicted at the preliminary design phases. Numerical simulations replace progressively experimental determinations that are time and money consuming.
The object of the work was to qualify from experimental results a timing belt drive numerical model. The model simulates the dynamic behavior versus time of any kind of tooth belt power transmissions. The model architecture, originalities and capabilities have been already presented, and the purpose is now to compare in details numerical and experimental results.
The experimental qualification has been carried out on a laboratory test bench with a medium size engine valve controlled distribution made of 3 pulleys and a tensioner. Tensions, camshaft torque, pulleys speeds and angular acylisms, dynamic transmission error between camshaft and crankshaft pulleys have been measured. Numerous tests have been made for different running conditions by changing : speed, angular acyclism, camshaft torque, setting tension. Several phenomena and influence of parameters have been identified, as the pulley eccentricity effect on camshaft torque, span tensions, and transmission error. Part of the experimental results are used as entries of the model : camshaft torque, crankshaft instantaneous speed, transmission error due to pulley eccentricities. Further, comparisons with the numerical results were made. Experimental and numerical results of tension, angular acyclism, dynamic transmission error, versus operation time are compared for the different tests performed. The agreement is good and shows that the model developed allows to simulate dynamic behavior of timing belt with high degree of confidence.
