Search Results

The dynamic behavior of an OHC valve train system of a spark ignition engine is investigated to characterize the source and transmission of the valve train (VT) vibration and noise and to improve the VT design for better sound quality. The spectral properties of vibration caused by highly transient dynamics of VT system are characterized for the high frequency ranges over 3 kHz, although the overall sound pressure level due to the VT is negligible [1, 2]. For the analysis of valve train a lumped parameter model with 4 d.o.f.'s is developed and validated with the experimental results from a test rig. Experiments are performed on the test rig to measure the valve acceleration, the surface vibration of cylinder head during the operation, and the transfer functions. Also a measurement of cylinder head vibration in a real vehicle has been performed to correlate with the rig test results.

Experiments were done to quantify the dynamic motion of the piston and oil-film during piston impact on the cylinder bore, commonly known as “piston slap.” Parameters measured include engine block vibration, piston-skirt to liner separation, oil-film thickness between the piston and liner, and other engine operating conditions. Experimental parametric studies were performed covering the following: engine operating parameters - spark timing, liner temperature, oil-film thickness, oil type, and engine speed; and engine design parameters - piston-skirt surface waviness, piston-skirt/cylinder-liner clearance, and wrist-pin offset. Two dynamic modes of piston-motion-induced vibration were observed, and effects of changes in engine operating and design parameters were investigated for both types of slap. It was evident that engine design parameters have stronger effects on piston slap intensity, with piston-skirt/liner clearance and wrist-pin offset being the dominant parameters.

Cylinder combustion and fuel injector pump pressures and forces due to impacts of valve and seat and piston slap are signals of interest for diesel engine diagnostics. The vibration signals that these forces generate are easier to measure than the forces themselves, but unfortunately the signals are greatly changed as they travel through the engine structure. In addition, the vibration transmission of engine structures varies from one copy of the engine to another. Signal processing procedures for the recovery of fault signatures from vibration signals must take both the structural variability and the complexity of the transfer function into account. A robust/adaptive process that uses cepstrum windowing, singular value decomposition, and Kalman filtering as elements of its procedures has been applied to recovery of the combustion pressure pulse in a diesel engine.