| Particle gluing is one of the important sections of particleboard production,and its key equipment is ring-type glue blender.The functional parameters,which is performance parameters and production parameters,of the ring-type glue blender will directly affect the gluing effect and fragmentation of the particleboard,thus affecting the production quality and mechanical properties of the particleboard.Traditional methods often use experimental methods or workers’ practical experience to optimize the functional parameters of the ring-type glue blender.In the optimization process,there are problems such as repeated experiments,long cycle and low accuracy.In this paper,from the perspective of improving the gluing effect of the ring-type glue blender and reducing the crushing degree of the particles,through theoretical research,simulation analysis,simulation experiment verification and mathematical model prediction,the influence of the functional parameters of the ring-type glue blender on the gluing effect and crushing degree of the particles is explored,and the functional parameters of the ring-type glue blender are optimized.It provides theoretical methods and technical means for improving the gluing effect of the ring-type glue blender and reducing the crushing degree of the particles with high quality,high efficiency and intelligence.Based on the analysis of the particleboard gluing process and the operation mechanism of the ring-type glue blender,based on the meso-mechanical analysis of the particle gluing of the ring-type glue blender,this paper analyzes the mutual movement and force between the particle and the adhesive,the particle and the particle,the particle and the stirring claw,and the particle and the inner wall of the stirring cylinder at the meso-scale,and reveals the particle and the discrete glue in the process of particle gluing.Based on the velocity gradient variation characteristics of the particle flow,the internal relationship between the functional parameters of the ring-type glue blender and the particle gluing effect and the degree of fragmentation is studied,and the influence of the functional parameters on the particle gluing effect and the degree of fragmentation is analyzed,which provides a research basis for constructing the simulation model of the particle gluing of the ring-type glue blender.Based on the theory of computational fluid dynamics(CFD)and discrete element method(DEM),the FLUENT-EDEM particle gluing simulation model of the performance parameters of the ring-type glue blender was established.According to the meso-mechanical analysis results of the particle and adhesive of the ring-type glue blender,the velocity gradient characteristics and mutual movement of the particle and the glue in the flow field were analyzed and studied,and the particle velocity distribution law of the particle flow in the ringtype glue blender was obtained.By using the particle mass change and particle size ratio of each performance parameter in the FLUENT-EDEM particle gluing simulation model,the average particle weight gain,standard deviation of particle weight gain and particle crushing ratio under each performance parameter were counted.The influence trend of each performance parameter on the particle gluing effect and its crushing degree was obtained.The orthogonal experiment method was used to simulate the particle gluing effect and crushing degree under each performance parameter.Based on the analysis of variance(ANOVA)and grey relational analysis(GRA),the influence of each performance parameter and simulation experiment results was analyzed and studied.The comparison results show that the FLUENT-EDEM particle gluing simulation model has the same influence degree and trend as the simulation experiment results,which verifies that the simulation model establishment method is feasible and has high reliability.Based on the ANOVA-GRA method,the correlation between the performance parameters and the particle gluing effect and the degree of fragmentation is analyzed,and the performance parameters of the ring-type glue blender are optimized.The FLUENT-EDEM particle gluing simulation model is used to evaluate the optimization results.The results show that the ANOVA-GRA method is at a better level for the optimization of performance parameters.Based on the parameters of the multi-objective optimization support vector regression(SVR)prediction model of the elite strategy non-dominated sorting genetic algorithm(NSGA2),the NSGA2-SVR multi-objective prediction model of the production parameters of the ring-type glue blender and the mechanical properties of the particleboard was established,and a multi-objective prediction method of the production parameters of the ring-type glue blender and the mechanical properties of the particleboard based on the NSGA2-SVR multiobjective prediction model was proposed.Different mathematical models and NSGA2-SVR multi-objective prediction models were used to compare the prediction accuracy of the production parameters of the ring-type glue blender,and the multi-objective prediction effect of the NSGA2-SVR model on the production parameters of the ring-type glue blender was evaluated.The results show that the established multi-objective prediction model has good prediction performance,and the actual application of the NSGA2-SVR model is verified and analyzed according to the production requirements.The results meet the requirements of the actual working condition prediction standard,which is the optimization of the functional parameters of the ring-type glue blender and the modeling and prediction of the gluing effect.The overall improvement of gluing effect in actual production provides support for innovative theories and technical methods. |
