Experimental identification of the parameters of a multilayer coating architecture

Composite components with specifically designed and optimized vibration damping behaviour can be obtained by means of multilayered coating technology solutions [1-3], and these authors presented a multilayer beam model that takes into account of the dissipative actions at the interface between different layers [3], modeled by means of a complex interlaminar impedance. This model can be used as a virtual prototyping tool in order to find and optimize innovative multilayer coating solutions. Since the model needs laminar parameters as input, i.e. the density and modulus of the layer materials and the interface parameters values that define the complex impedance between different coating materials, dynamic mechanical measurements and an optimization based numerical technique are used to find these laminar and interlaminar parameters. An identification procedure is applied to measurement data from three-layer and five-layer beam specimens in order to identify the unknown interface parameters modeling the interlaminar impedance. Normalized estimator z [3] is used to evaluate the damping behaviour. Error vector e, function of the unknown parameters, is defined as the difference between z calculated from the model and z measured at the same frequency values. The objective function Ψ=0.5·e·eT is obtained. A multi-step optimization algorithm based on a quadratic, constrained non-linear optimisation technique is employed to identify the interface parameter values that minimize Ψ. Fig. (1a) shows examples of identification, i.e. measurements and the resulting model fit for five-layer beams, adopting Al and E1, E2 epoxy based layers. Identified parameter values are then used as input to the multilayer beam virtual prototype model in order to simulate the damping behaviour of different multilayer composite architectures. Fig. (1b) shows the z model estimate from the C1 (E2-E1-Al-E1-E2) five-layer beam virtual prototype model, and the z experimental validation. Some different coating solutions are then applied to gear pump Al casings to be tested and compared, by means of sound pressure measurements, in a real industrial application with real operating conditions, i.e. revolution speed n [1000-2000] rpm, fluid (oil) pressure p [50-230] bar. Results concerning the application of the E1-E2 coating architecture are reported in Fig. (2). A discussion and a critical analysis of the results is presented.