5 years of NVH applications make Transfer Path Analysis (TPA) appear a commodity tool. Despite the fact that TPA is used in a large variety of applications today, the method remains an expert methodology requiring insight in both the NVH problem and all the related signal and system analysis constraints. TPA is proven to be reliable, but the main bottleneck remains the huge measurement time to build the full data model. For this reason, industry is constantly seeking for simpler and faster methods. One such method is Operational Path Analysis (OPA), which was introduced about two years ago. OPA is a fully-operational method, requiring only operational measurements of the path references (body-side mount accelerations, pressures close to vibrating surfaces, nozzles and apertures, etc.) and target response(s). The OPA method is indeed very time-efficient, but suffers from several limitations leading to false path contributions and wrong engineering decisions. Its major limitations are: (i) cross-coupling effects between path references which may lead to faulty interpretations, (ii) potential errors due to missing paths and (iii) numerical illconditioning problems related to the estimation of transmissibilities from operational data. So, despite the fact that OPA is a very time-efficient approach, its benefit is limited in most application cases.
This paper introduces a novel path contribution method which combines the advantages of classical TPA and OPA. The method is based on simplifications that allow balancing path accuracy and speed of execution. The principles of the method are first outlined. Then, the method is compared with the existing TPA and OPA methods using an automotive example.
