Revolutionising High-Performance Chain Drive Testing: A Precision Dynamometer with Empirical Bearing Friction Compensation

In elite competitive cycling, where races are won and lost by the smallest of margins, bicycle hardware is under continuous optimisation. However, such is the maturity of today’s bicycles that step changes are implausible and efforts are instead focussed on achieving marginal gains. This presents a distinct challenge for the engineering of chain transmissions, requiring discrimination of fractional performance differences between design or setup changes. To meet this need, we present a bespoke chain dynamometer system together with a novel empirical method of compensation for the losses in support bearings, and an associated experimental methodology for high-fidelity emulation of real-world conditions. The mechanisms of power losses in chains are explored, including the influence of sprocket-sizing, polygonal action and span dynamics. The challenges of measuring losses in highly efficient transmission systems are discussed and the factors influencing measurement uncertainty are described. The dynamometer support bearings’ frictional moments are characterised using a secondary experiment, from which interpolated losses are eliminated from the dynamometer data. The design and operation of the measurement systems are described, and the measurement uncertainty of the bearing data is propagated to evaluate the uncertainty of the chain dynamometer losses by empirical and analytical means. The presented system provides the accuracy and precision necessary to optimise transmission design and lubricant selection for Olympic track cycling competitions. Independent repeats of a baseline chain and commercially available lubricant combinations across multiple test campaigns have yielded a bounds uncertainty of < 1 . 75 % of the measured power loss at the 99% confidence interval. • A high-performance dynamometer for chain drive loss measurement is presented. • Loading and boundary conditions specific to Olympic track cycling are recreated. • Support bearings are empirically characterised for accurate parasitic compensation. • Efficiency shown to depend on cyclic torque input, sprocket sizing, chain dynamics. • Tests are demonstrated to have good repeatability.