Iterative eigenvalue method coupled with RSM for structural acoustic analysis and optimization of the CLD-damped orthotropic steel deck

In the present work, an iterative eigenvalue method (IEM) is proposed and combined with the boundary element method (BEM) to perform the structural acoustical analysis of the orthotropic steel deck (OSD) damped with constrained layer damping (CLD), through which the frequency-dependent properties of the viscoelastic core can be accurately considered. Then, a series of hammer tests are carried out in a semi-anechoic chamber to verify the acoustic prediction method. It shows that the simulated structural noises basically agree with the test results, and the overall deviations at the field points M2 and M8 are merely 2.4 and 2.3 dB, respectively. Subsequently, taking the thicknesses of the damping layer and the constraining layer, as well as the material properties of the constraining layer as design variables, a central composite experiment is designed and combined with response surface methodology (RSM) to fit the relationship between analysis objectives and design variables. With the objective of minimizing the acoustic radiation power, a program for finding the optimal values of design variables is developed based on the interior penalty function method. Through iterative computations, the optimal points are numerically determined as t1 = 2.3 mm and t2 = 0.3 mm for the thicknesses of the damping layer and the constraining layer, respectively. Steel alloy is found to be the optimal material for the constraining layer. After optimizing CLD treatment, the overall acoustical power level can be reduced from 97.7 dB to 93.1 dB and the overall sound pressure levels (SPLs) at the field points M2 and M8 can be reduced from 89.2 dB to 83.5 and 76.9 dB to 70.3 dB, respectively, indicating that the optimization of CLD design parameters can significantly improve acoustic performance of the CLD-damped structures.