Search Results

There are two distinct classes of finite element models that can be used to support vehicle body design and development. The most familiar of these is the detailed body model, which achieves computational accuracy by precisely simulating component geometries and assembly interfaces. This model type is quite useful for conducting trade-off studies after detail drawings become available. The second class is an architecture concept model that simulates the basic layout and general structural behavior of major load-carrying members (e.g., pillars, rails, rockers, etc.) and joints in the body. Such modes are valuable for design direction studies in the earliest phases of the vehicle development process. This paper presents a generic process for building architecture concept models that include a mathematical representation of the major body joints derived from existing CAE models.

Gear whine is one of the NVH attributes that have become noticeable as performance of other NVH attributes is improved. The predominant root cause of gear whine is transmission error, defined as the deviation of angular position of driven gear/s from the ideal conjugate position/s. Presence of Transmission error can be attributed to three major sources: 1) dimensional variability of individual gears during manufacturing, 2) misalignment during assembly and 3) dynamic tooth deflections during operation. This paper describes a method to estimate statistical distribution of Composite Manufactured Transmission Error (cMTE) for a planetary gear system based on measured surface variation of the gear tooth profiles. The statistical distribution of the surface variation is derived from measured left and right profiles for four equally spaced teeth per gear in both lead and involute directions.

Distinguishing physical modes from mathematical modes in the modal analysis of complex systems, such as full vehicle structures, is a difficult and time-consuming process. The major tools frequently used are stabilization diagrams, mode indicator functions, or modal participation factors. When closely spaced modes are to be identified, the stabilization diagrams and mode indicator functions are no longer effective. Even the reciprocities of mode shapes and modal participation factors cannot be well satisfied to indicate whether a mode is a physical one, when measurement errors are large. To overcome these difficulties, an optimization procedure is developed, whereby physical modes can be sorted out in a given frequency range while the error between measured and synthesized frequency responses is minimized. An optimal subset selection algorithm is used in this procedure.