Detecting and predicting the onset of cavitation in automotive torque converters

This investigation will detail a study of the effects of multiple torque converter design and operating point parameters on the onset of cavitation during vehicle launch. The most critical operating condition for the occurrence of cavitation is stall, or zero turbine speed. The noise produced by the collapse of cavitation bubbles occurring in the automotive torque converter during stall operation could become objectionable to vehicle occupants as manufacturers continue to improve noise levels in new vehicles. It is then important to identify the general torque converter geometries and operating conditions that increase speed and torque thresholds at the onset of cavitation.; An automated torque converter dynamometer test cell was developed to characterize the effect of torque converter design parameters on the onset of cavitation and noise. A nearfield acoustical technique was utilized to acquire cavitation noise occurring within the torque converter during a standardized stall speed sweep test. After post processing the noise data, the onset of cavitation was readily identified by a sudden increase in filtered sound pressure level. This technique was applied to a wide range of design and stall operating conditions. The effects of the design parameters of diameter, torus dimensions, and pump and stator blade designs were determined.; Dimensional analysis and response surface methodology were used to develop dimensionless empirical models for predicting the onset of cavitation based upon design and operating point parameters. The corresponding dimensionless empirical models can be used in the torque converter design process to produce designs in which the onset of cavitation occurs at the highest possible combination of speed and torque at stall. At higher speeds and torques cavitation noise will be masked by other sources of powertrain noise. A dimensionless empirical model encompassing nearly all torque converters tested can be used to predict the onset of cavitation for general torque converter design to within +/- 7.00% root mean square error.