ObjectiveIn order to investigate the influence of process factors on the acoustic vibration performance of composites, the process parameters of composite preparation were optimized to improve the acoustic vibration performance of composites.

MethodThe test was designed according to the structure principle of laminated veneer lumber to prepare birch veneer/glass fiber composites. FFT was used to detect the acoustic vibration properties of composite materials. The comprehensive score after normalization of the specific dynamic elastic modulus (E/ρ), the ratio of elastic modulus and shear modulus (E/G), acoustic radiation damping (R), loss tangent (tanσ), and sound velocity (v) was used as the response indicators to analyze the influence of hot-press time, hot-press pressure and resin sizing amount on the acoustic vibration performance of composite materials. Based on the single factor experiment, the response surface methodlogy was used to establish the quadratic regression model of process factor and response value to optimize the preparation conditions of composite materials.

ResultWithin the scope of the single factor experiment, when the hot-press time was 10−25 min, the pressure was 0.6−1.3 MPa, and the resin sizing amount was 140−180 g/m², the acoustic vibration performance of the composite materials was significantly improved. Those experiments used Design-Expert to perform quadratic polynomial regression fitting on the acoustic vibration performance test results of composite materials, eliminating the factors that have no significant influence on the model, and the response surface model of composite scores was established. The optimal process conditions optimized by the response surface model were hot-press time 24.5 min, hot-press pressure 1.3 MPa, resin sizing amount 180 g/cm². Under the conditions, the E/ρ of the composite reached 25.27 GPa, E/G was 15.99, R was 6.48 m³/(Pa·s³), tanσ was 0.001 25, v was 5 026.55 m/s, and the comprehensive score reached 98.19.

ConclusionThe P of the comprehensive score model was less than 0.000 1, the deviation between the measured value and the predicted value was less than 5%, indicating that the response value has a highly significant relationship with the regression model. It also shows that the regression model is accurate and reliable.