Quasi-Static Transmission Error Behavior Under the Composite Effect of Temperature and Load

Abstract

Current demands for enhanced rotational speed in electric vehicle transmissions, aeronautical gearboxes and industrial machinery is known to affect the thermal behavior of mechanical parts by increasing their steady-state temperature. In geared transmissions, such condition is detrimental as the lubricating film is reduced thus increasing failure probability, furthermore, if temperature levels are sufficiently high thermal distortion can affect mesh behavior. Scientific literature review has shown that no experimental evidence on the composite effect of temperature and torque on transmission error exists up to date. Although some authors already pointed out that temperature influences positioning accuracy, no previous reference to peak to peak behavior has been found and comparisons to torque effects have not been performed. In this work, quasi-static transmission error behavior is experimentally analyzed under increased thermo-mechanical conditions. First, the development of a custom back-to-back test rig is described, and test specimen geometries, operating conditions and measurement procedure are presented. Next, loaded transmission error tests are carried out in order to validate the expected mechanical behavior and then the influence of temperature is analyzed by heating up the system in a controlled manner. Composite effect of temperature and load are studied in terms of backlash, mean level of transmission error and its peak-to-peak value. Finally, experimental measurements are compared to analytical predictions, results are discussed, and conclusions are withdrawn. It is shown that the effect of temperature and torque coexist in transmission error diagrams. Both parameters have a significant role in the mean level of transmission error while the influence of torque on peak to peak is prominent relative to that of temperature. Although the correlation between the change of mean level and that of backlash for increasing temperatures is clear, peak-to-peak variation due to temperature is not obvious.