顾毅楠，吴永泽，俞建峰，钱陈豪，化春键，蒋 毅

江南大学 机械工程学院，江苏省食品先进制造装备技术重点实验室，江苏 无锡 214122

引用格式：

顾毅楠，吴永泽，俞建峰，等. 涡流空气分级机补气结构设计及流场仿真［J］.中国粉体技术，2024，30（5）：158-170.

GU Yinan， WU Yongze， YU Jianfeng， et al. Air supply structure design and flow field simulation of turbo air classifier［J］.China Powder Science and Technology，2024，30（5）：158−170.

收稿日期：2024-03-29，修回日期：2024-06-28，上线日期：2024-08-30。

基金项目：国家自然科学基金项目，编号：51905215；江苏省研究生科研与实践创新计划，编号： KYCX23_2553；江苏省食品先进制造装

备技术重点实验室自主研究课题资助项目，编号：FMZ202302。

第一作者简介：顾毅楠（1999—），男，硕士研究生，研究方向为粉体智能装备。E-mail：gyn19991007@126. com。

通信作者简介：俞建峰（1974—），男，博士，教授，博士生导师，研究方向为粉体智能装备。E-mail：robotmcu@126. com。

摘要：【目的】 针对涡流空气分级机补气效果不佳的问题，为寻求最优补气方案。【方法】 采用Fluent软件对涡流空气分级机进行流场仿真，探究不同补气结构对气流滞留能力、均匀性、分散性的影响，以及对应分级机的颗粒分级效果。【结果】 仿真结果表明， Model 2在导流罩上方增添挡风环结构，有效遏制了气流流失，显著提升了补气气流在锥壳内部的滞留时间，d₅₀较Model 1的减小0. 62 μm； Model 3在补气结构内添置导流叶片，减弱了气流冲击并增加气流分散效果，d₅₀较Model 2的减小2. 88 μm； Model 4、 Model 5将补气结构进出气区域在竖直方向上错开，保证了气流有充足的滞留空间，进而阻止气流快速扩散； Model 5采用补气结构出气区域下置方案，利用锥壳壁面为气流导向，其轴向速度分布更加合理，气流分散性及分布均匀性最好。【结论】 对比5种补气结构的粒径累计分布曲线，Model 5的d₅₀为28. 5 μm，略大于Model 3的27. 3 μm，但在0~20 μm粒度区间段，Model 5的曲线位于Model 3下方，说明Model 5对大米粉细颗粒的分级效果优于Model 3，综合来看，Model 5对大米粉的颗粒分级效果最佳。

Abstract

Objective To address the poor performance of turbo air classifier， five different air supply structures were designed to find the optimal scheme to improve its classification efficiency.

Method The accuracy of the calculation model was verified through grid independence analysis and eddy current classification experiments using rice flour. The flow field of the turbo air classifier was simulated using Fluent software to explore the influence of different air supply structures on air retention capacity， uniformity， and dispersion. A discrete phase model （DPM） was used to simulate the motion trajectory of rice flour particles. The cumulative distribution of particle size at the discharge of the classifier was analyzed to determine the influence of different air supply structures on rice flour classification.

Results and Discussion Simulation results revealed that adding a windshield ring structure above the cone fairing in Model 2，based on Model 1， effectively reduced direct air loss from the cone shell wall， significantly increased the residence time of the replenished air in the cone shell， and reduced d₅₀ by 0. 62 μm compared with Model 1. Model 3， which added a flow guide blade to the air outlet plane of the air supply structure on the basis of Model 2， weakened air impact and increased air dispersion. The fluctuation in axial velocity value in Model 3 became smaller， and the d₅₀ was 2. 88 μm lower than that of Model 2.However， insufficient space in the air supply structure reduced airflow retention time and affected air replenishment. A completely different air supply structure was adopted in Models 4 and 5 from Model 3， with staggered inlet and outlet areas in the vertical direction to ensure sufficient retention space for airflow， thereby improving the residence time. The d₅₀ of Model 4 was 30. 22 μm， which was 2. 92 μm higher than Model 3， indicating suboptimal air-replenishment performance. In contrast， Model 5 adopted a downsetting scheme for the air outlet area of the air supply structure， using the conical shell as an air flow guide. Model 5 showed less fluctuation in axial velocity， a more reasonable distribution， and the best air flow dispersion and uniformity. Although the d₅₀ of Model 5 was slightly increased by 0. 62 μm compared with Model 3， its cumulative particle size distribution curve was below that of Model 3 in 0 to 20 μm particle size range， indicating better particle classification efficiency for rice flour.

Conclusion Among the five air supply structures， the baffle structure in Model 1 had the worst air flow retention effect， with the largest d50 of 30. 8 μm. The windshield ring structure in Model 2 effectively reduced air loss， significantly improving air replenishment， with a d₅₀ approximately 0. 62 μm lower than Model 1. The guiding vane structure in Model 3 effectively weakened air impact and improved air dispersion， demonstrating the most noticeable air replenishment effect， with a d₅₀ 2. 88 μm lower than Model 2. Model 4， which adopted the upper scheme for the air supply structure outlet area， performed poorly， while Model 5，adopting the lower scheme， performed better. Although the d₅₀ of Model 5 was 28. 5 μm， slightly larger than 27. 3 μm in Model 3， the cumulative particle size distribution curve was below that of Model 3， showing a better classification efficiency for rice flour particles. Overall， Model 5 performed the best in the particle size range of 0 to 100 μm.

Keywords：turbo air classifier； air supply structure； air retention capacity； uniformity； dispersion

参考文献（References）

［1］卢道铭，范怡平，卢春喜. 颗粒空气分级技术研究进展［J］.中国粉体技术，2020，26（6）：11-24.

LU D M， FAN Y P， LU C X. Advances in research on granular air classification［J］.China Powder Science and Technology，2020，26（6）：11-24.

［2］PUKKELLA A， CILLIERS J， HADLER K. A comprehensive review and recent advances in dry mineral classification［J］.Minerals Engineering，2023，201：108208.

［3］付敏，曹众，陈效庆，等. 空气分级技术及设备研究进展［J］.化学工业与工程，2024，41（4）：117-130.

FU M， CAO Z， CHEN X Q， et al. Research progress of air classification technology and equipment［J］.Chemical Industryand Engineering，2024，41（4）：117-130.

［4］HUI C X， LI Q， PENG J X， et al. CFD simulation and performance study of a three-separation combined air classifier［J］.Thermal Science，2024，28（2）：1589-1603.

［5］ESMAEILPOUR M， MOHEBBI A， GHALANDARI V. CFD simulation and optimization of an industrial cement gas-solid air classifier［J］.Particuology，2024，89：172-184.

［6］CURTIS R， LI X D， GENG X， et al. Manufacture of defatted canola meal with enhanced nutritive composition by air classification on an industrial scale［J］.Science of Food and Agriculture，2019，100（2）：764-774.

［7］ALTUN O. Air classification performances of the components within the varied feed blends［J］.Powder Technology，2022，399：117092.

［8］POLITIEK R， HE S， WILMS P， et al. Effect of relative humidity on milling and air classification explained by particle dis persion and flowability［J］.Journal of Food Engineering，2023，358：111663.

［9］HUANG L， YUAN J， MIAO P， et al. CFD simulation and parameter optimization of the internal flow field of a disturbed air cyclone centrifugal classifier［J］.Separation and Purification Technology，2023，307：122760.

［10］ISMAIL F， Al-MUHSEN N， HASINI H， et al. Computational fluid dynamics （CFD） investigation on associated effect of classifier blades lengths and opening angles on coal classification efficiency in coal pulverizer［J］.Case Studies in Chemical and Environmental Engineering，2022，6：100266.