Evaluation of uncertainties in classical and component (blocked force) transfer path analysis (TPA) 2019-01-1544

In situ component TPA (originally named in situ source path contribution analysis [Elliott et al. J. Sound Vib., 332 (24) (2013)] has been attracting considerable attention recently. The measurements required with the in situ method are almost identical to conventional TPA but there are two major advantages. First, there is no need to dismantle the vehicle or assembly since all measurements are conducted in situ. Secondly, the forces obtained by matrix inversion are the blocked forces which are an independent property of the source component and are therefore transferrable to different assemblies. The current paper extends this idea of mathematically decoupling components from in situ measurements. The in situ measurement of blocked forces is first considered. A substructure decoupling approach is then described from which independent properties (accelerances) of coupling elements and other components can be extracted from in situ measurements. The independently described components can then in principle be (mathematically) recoupled with other components to build a virtual acoustic prototype. However, the inverse methods used for indirect measurement of blocked forces and for substructure decoupling can be sensitive to measurement errors as is well known. Therefore, a framework is presented for estimating the uncertainties associated both with the blocked forces and with the substructure decoupling and recoupling. The principles are illustrated through detailed in situ measurements made on a laboratory assembly. A practical example of field measurements is then presented.