Experimental Validation of a Multi-Row Tapered Roller Bearing Analytical Model for Relating Preload to Frequency Response Characteristics

Roller bearings are used in many rotating power transmission systems in the automotive industry. During the assembly process of the power transmission system, some types of roller bearings (e.g., tapered roller bearings) require a compressive preload force. Those bearings' rolling resistance and lifespan strongly depend on the preload set during the installation process. Therefore, accurate setting of the preload can improve bearing efficiency, increase bearing lifespan and reduce maintenance costs over the life of the vehicle. A new method for bearing preload measurement has shown potential for both high accuracy and fast cycle time using the frequency response characteristics of the power transmission system. An open problem is experimental validation of the multi-row tapered roller bearing analytical model. After validation, the analytical model can be used to predict the assembled system damped natural frequency for a desired bearing preload. This work presents the experimental validation of the analytical model including the experimental test stand, test method, test results and comparison to analytical model. The developed test stand represents an automotive pinion shaft/gear as might be found in a rear axle and can be mounted either vertically or horizontally to simulate assembly and operational positions of the axle and to consider the effects of gravity. To measure bearing preload, the test stand is instrumented with a load cell, and each shaft has strain gages installed. Multiple accelerators are used to measure the system frequency response to an impulse provided by a modal hammer. For this work, three different bearing pairs in back-to-back configuration are tested at eight different preload values. The analytical model is evaluated using the same bearing designs and the preloads measured from the test stand. Results from the analytical model are compared to experimental results to validate the analytical model.