Search Results

Superior NVH performance is a key focus in the development of new powertrains. In recent years, computer simulations have gained an increasing role in the design, development, and optimization of powertrain NVH at component and system levels. This paper presents the results of a study carried out on a 4-cylinder in-line spark-ignition engine with focus on growl noise. Growl is a low frequency noise (300-700 Hz) which is primarily perceived at moderate engine speeds (2000-3000 rpm) and light to moderate throttle tip-ins. For this purpose, a coupled and fully flexible multi-body dynamics model of the powertrain was developed. Structural components were reduced using component mode synthesis and used to determine dynamics loads at various engine speeds and loading conditions. A comparative NVH assessment of various crankshaft designs, engine configurations, and in- cylinder gas pressures was carried out.

Current modeling techniques of the powertrain noise, vibration and harshness (NVH) involve fully meshed structural components and rely, in general, on predefined excitation loads to evaluate linear transfer or structural attenuation functions. While effective for comparative assessment of various designs, these methods neglect the complex dynamic interactions between the powertrain structure and crankshaft, piston, valve train, timing drive, and accessory drive systems. This paper presents an overview of modeling methods of low and high frequency powertrain NVH with focus on dynamic interaction among structural components. A coupled and fully flexible multi-body dynamics model using AVL/Excite is presented. The model includes the cranktrain, crankcase, cylinder head, covers, oil pan, mounts, and transmission housing represented as finite element meshes.

Vibration signal measurement and analysis is an efficient non-intrusive method for engine diagnosis. In this paper, we propose a method based on wavelet transform for vibration signature extraction to detect, locate, and diagnose a range of common faults in valve train mechanism and combustion process. For this purpose, a flexible multibody dynamics model has been developed using ADAMS. The model includes the valve train system, cylinder head, and gas pressure effect. The vibration response of the cylinder head surface under simulated faults such as misfire, cylinder pressure variations, and excessive valve clearance is investigated. The spectral properties of vibrations caused by highly transient dynamics are analyzed using continuous wavelet transform. Wavelet decomposition is performed to identify and correlate dynamical response of the system with corresponding sources of excitation. Fault detection procedure is presented and discussed.