Parametric inverse analysis of Johnson-Champoux-Allard acoustic model for weft knitted fabrics based on particle swarm algorithm
Objective Textiles are widely used in the field of acoustic absorption due to their porous texture,lightweight and formability.Due to the viscous inertia and thermal dissipation mechanisms in acoustic absorption of textile materials,the Johnson-Champoux-Allard(JCA)acoustic model is believed suitable for characterization their acoustic property.However,few studies were conducted on the acquisition of acoustic parameters in JCA model and the relations between acoustic parameters and fabric structure remains vague.This paper proposes a method to quickly acquire JCA acoustic model parameters and predict the acoustic absorption of weft-knitted knitted fabric.The particle swarm algorithm was chosen to obtain the JCA acoustic model parameters by inverse analysis.The relations between fabric structure and acoustic parameters were explored,and the sound absorption coefficients of knitted fabrics with different structures were predicted.Method Particle swarm algorithm was chosen to inversely analyze the acoustic absorption coefficient of weft-knitted fabrics,to obtain parameters of JCA acoustic model,including porosity,flow resistance,tortuosity,viscous characteristic length and thermal characteristic length.By adding inertia weights and learning factors,the inverse analysis process was restricted,thus reducing the number of iterations,avoiding the local optimal solutions,and improving the accuracy of the parameters obtained from the inverse analysis.Based on the results of the inverse analysis,the relations between the structural parameters of the weft-knitted fabric(unfilled coefficient)and the structural parameters of the JCA acoustic model(porosity,flow resistance)were established,and the JCA acoustic model parameters of knitted fabrics with different structures were obtained quickly for acoustic absorption coefficient calculation.The accuracy of the obtained JCA acoustic parameters of fabrics with different structures was verified by the finite element method.Results The acoustic parameters such as porosity,flow resistance,tortuosity,viscous characteristic length and thermal characteristic length were inversely analyzed by the particle swarm algorithm.After 100 iterations,the iteration speed slowed down and gradually stabilized,reaching the globally optimal solution.The final iteration number was less than 200,with a minimum value of 0.19.Comparison of the numerically calculated sound absorption coefficient with the experimentally measured curves showed that the particle swarm algorithm was able to accurately inverse-analyze the JCA acoustic parameters in the range of 500-5 000 Hz.When the structure was changed,the material parameters,including tortuosity,viscous characteristic length and thermal characteristic length,were empirically obtained from the inverse analysis.Porosity was determined by the unfilled factor according to the global optimal solution.Flow resistance was obtained by fitting the porosity and the flow resistance using the exponential function in the least squares method with known inverse analytical parameters.The coefficient of determination R2 was 0.994 6,indicating the effective fitting.The accuracy of JCA acoustic parameters obtained by above method was verified by finite element method.The sound absorption coefficient curves obtained from the finite element calculations for the weft flat-needle tissues fitted well with the inverse analysis and experiments.The coefficient of determination R2 was 0.809.The sound absorption coefficient curves obtained from finite element calculations of the double inverse organization fitted the inverse analysis and experiments well.The coefficient of determination R2 was 0.852.The work proves the accuracy and reliability of the fabric structure parameters deduced from the JCA acoustic model.Conclusion The particle swarm method was optimized to inversely analyzing the sound absorption coefficient of weft knitted acoustic-absorption material,and the number of iterations is less than 200,achieving the rapid acquisition of the parameters in the JCA acoustic model.For different textile structures,by directly obtaining the porosity and flow resistance coefficients and combining them with known material parameters,the sound absorption coefficients at different frequencies can be calculated quickly and with less error.This method provides new ideas for the acquisition of acoustic parameters and the prediction of sound absorption performance of acoustic-absorbing material.
万方数据
维普
