Use of Statistical Energy Analysis to Predict Articulation Index (AI) in a Vehicle Interior for Real-World Operating Scenarios

<div class="section abstract"><div class="htmlview paragraph">Vehicle interior noise is a crucial assessment criterion for automotive NVH. It has a significant effect on customer opinions about the quality of a vehicle. Articulation Index (AI) is one of the key sound metrics used to describe speech intelligibility and quantifies the middle and high frequency spectra associated to the internal noise of vehicle. In reality, Vehicle operating under dynamic condition experiences various air-borne noise sources such as tire rolling noise, powertrain noise, intake-exhaust noise & wind noise along with structure borne excitations such as powertrain vibrations, suspension vibrations. It is very challenging to predict cumulative effect of all these excitations to interior noise level and Articulation Index (AI) of vehicle over complete frequency range. The statistical energy analysis (SEA) is a well-known methodology being used to simulate & predict mid & high frequency noise. Objective of this paper is to present the process of development of a SEA simulation model designed to investigate vehicle interior noise & Articulation index and associated correlation against test measurements for various real- world operating scenarios. The SEA simulation model was meticulously developed with close attention given to structural representation which allowed to consider the structure borne excitations along with air borne noise sources during the analysis. The interior trims & sound insulation pack were also in detailed in the model. Both static & dynamic real-world operating scenarios of vehicle or load cases are demonstrated to validate the model against test measurements. The contribution study was performed to determine dominant noise sources and weaker transfer paths for improvement of Articulation index & interior noise quality of vehicle.</div></div>