Search Results

This paper presents the principles of robustness design of axle system dynamics to reduce vehicle system related axle gear whine. Through examining the physics of the axle gear noise, the influence of the system dynamics are identified as two parts, i.e., the dynamics mesh force generated at the gear meshing per unit gear mesh motion variation; and the force transmissibility from mesh to the axle housing, and then further to the bracket attachments. The noise sensitive design parameters are identified and discussed. Component design requirements are proposed to minimize the system resonances in the typical gear mesh frequency range. The use of FEA models for system understanding and further design tuning is illustrated.

It is generally recognized that a composite dynamic impedance method using component frequency response functions obtained experimentally is an effective procedure for analyzing complex dynamic systems extending over a wide frequency range. However, previous attempts to apply composite dynamic impedance methods to analyses of dynamic systems involving multiple connection degrees of freedom for several components have been largely unsuccessful because of numerical error that occurred in the matrix inversion process. In an effort to avoid this problem, researchers at Structural Dynamics Research Corp. have recently proposed an improved composite dynamic impedance method known as SMART, an acronym for System modelling and analysis using the response technique. SMART incorporates the singular value decomposition (SVD) theorem as a numerical technique for reducing error in matrix inversion.

The design of automotive components for low structure-borne interior noise and vibration is facilitated by the ability to reliably simulate total vehicle system response over a wide operating frequency range. This requires that the car body, its interior acoustic cavity, and critical chassis components must be included in the overall dynamic model. Unfortunately, most noise and vibration problems occur in the 200-1000 Hz frequency range where finite element and experimental modal methods have limited applicability. This is due to the high modal density, high damping levels, and sensitivity to fine geometric detail. A simulation method has been proposed earlier which uses component finite element models and component experimental transfer functions to predict combined system response [1]. This method has allowed for a practical approach to automotive system noise and vibration simulation.

The standard approach often used to reduce gear noise in automotive system is to minimize the transmission error. This is done by using stringent quality control measures in the gear manufacture, selecting desirable gear parameters, and applying profile modifications. This approach may be effective in many instances. However, there are numerous examples where the gear quality is the best that can be achieved within the manufacturing constraints, and the noise levels still exceed acceptable limits. In many cases, the system dynamics cause the gear train design to be highly sensitive to manufacturing induced transmission error. Therefore, it is advantageous to perform dynamic analysis to examine the influence of gear train dynamics and design parameters on gear noise. Proper design modifications may then be identified and applied to reduce gear noise levels.

The design of automotive components for low structure-borne interior noise and vibration requires the ability to reliably simulate total vehicle system response over a wide operating frequency range. This implies that the car body, its interior acoustic cavity, and critical structural components must be included in this overall dynamic model. Unfortunately, most noise and vibration problems occur in the 200-1000 Hz frequency range where existing finite element and experimental modal methods have limited applicability. This is due to the high modal density, high damping levels, and sensitivity to fine geometric detail. Moreover, it is highly doubtful that these methods will ever be practical tools for the study of the total body dynamics over the frequency range of 200-1000Hz. In this paper, a practical hybrid experimental-analytical approach is proposed in response to the need to simulate high frequencies structure-borne noise and vibration in automotive systems.