Abstract
With the advent of progressively quieter vehicle interiors, NVH testing of whole vehicles and or their components has grown in scope and complexity. While manufacturers constantly strive for ways to compress test time, often times it is necessary to run successive rather than concurrent tests as the requisite all-in-one transducers may not exist, or mutually exclusive boundary condition requirements may demand multiple tests.
In addition to defined seat acceptance noise and vibration criteria (ISO532B, non-stationary loudness, dB(A) functions, S.E.A.T. respectively) many specifications still demand that the sample must exhibit 'no objectionable noises during testing'. Clearly what may be objectionable is open to interpretation, and is often a function of where (spatially) the time domain data are obtained as well as overall boundary conditions. Much work has been done recently to assess these signals under controlled conditions (i.e. juried studies) and correlate to human impressions and preference. Further, it has been observed that seat / occupant interface often influences seat noise performance and quality during BSR testing, especially during in-vehicle tests on a four-post simulator or component tests on a shaker table. Ideally, a human surrogate that properly loads the seat (static and dynamic) while sensing noise like a human is desired. While binaural methodology exists and is widely utilized in vehicle NVH development, to our knowledge, none of the commercially available systems (head-alone, head and shoulders, head and torso etc) are mated to a mannequin that possesses the representative mass and compliance of a human. Canal-borne binaural microphones do exist, however, subjecting the same human to repeated tests may not be cost-effective or practical. As such the existing apparatus appear better suited to seat / interior vehicle noise tests where significant road energy is not imparted to the vehicle which could otherwise cause a BSR response from an improperly loaded seat. For the case of rough road excitations, (and thus BSR response) these deficiencies in seat loading may lead to misleading noise and or human vibration data. Thus in situ vehicle measurements may otherwise contain seat-related BSR properties that would never be observed were an occupant present in the seat.