A Multi-Variable Study Of Diesel Geartrain NVH Performance, 03-R9861Printer Friendly Version
Inclusive Dates: 10/01/08 10/31/09
Background - Review of the state of the art in diesel geartrain design shows definitive areas for noise vibration and harshness (NVH) research and development. Engine companies battle geartrain noise and vibration problems regularly, and many use band-aid fixes such as sound deadening covers and foam panels to combat the problem. Each engine company uses its own geartrain noise reducing recipes, some of which are costly, like anti-backlash gears. Little has been published on geartrain noise reduction, and published standards for design and development are not available. This effort is addressing these problems by developing a mathematical model for designing diesel engine geartrains with improved NVH.
Approach - A lumped-parameter vibration model is being developed using an open source scientific software package. The model will be verified through experiments performed on a four-cylinder diesel engine with a geartrain consisting of five gears. Eleven degrees of freedom will be modeled including the rotations of the five gears, the radial and axial displacement of the crankshaft, and three axial displacements of the main idler gear. Recently published gear tooth mesh stiffness and damping models have been used as a starting point and will be modified as necessary to correlate with test results. Two additional sets of gears have been designed with different tooth geometries, which will provide different input geometries to the model and different rattle behavior in the engine. These gear sets have been prototyped and are being tested in the engine, along with the engine's original gear set. Bearing clearance and backlash, both well-known to influence gear NVH, will also be varied and tested. Test results will provide multiple points for verification of the analytical model.
Accomplishments - The lumped parameter vibration model has been programmed, and initial troubleshooting has been accomplished. The key challenge was integrating the numerically stiff equations, which arise due to the backlash in the gears and the nonlinear contact stiffness of the gear teeth. This is now being done successfully. Some of the inputs to the analytical model include the motion of the crankshaft. The rotation of the crankshaft has been measured with an optical encoder, and the translational motions have been measured with lasers, as shown in the illustration. The two new gear sets have been designed. The first has been fabricated and is currently being tested. The second gear set is being machined. A transient engine dynamometer profile has been developed to provide a repeatable test to load the geartrain at three different levels throughout the operating speed range of the engine. In addition to developing the analytical model to predict geartrain NVH, this project is also helping to demonstrate the relative sensitivity that the many gear design parameters have on diesel engine NVH response.