The concept of attentuation is abandoned and instead the problem is formulated as a sequence of two linear couplings: the turbulent boundary-layer fluctuations excite the fuselage skin in lateral vibrations and the skin vibrations induce sound inside the fuselage. The techniques used are those required to determine the response of linear systems to random forcing functions of several variables. A certain degree of edalization has been resorted to. Thus the boundary layer is assumed locally homogeneous; the fuselage skin is assumed flat, unlined, and free from axial loads; and the "cabin" air is bounded only by the vibrating plate so that only outgoing waves are considered.
An Adobe Acrobat (PDF) file of the entire report: