旋涡自吸泵水力设计及流动噪声仿生优化

发布时间：2018-06-13 22:37

本文选题：旋涡泵 + 数值计算　；参考：《江苏大学》2017年硕士论文

【摘要】：本研究以XKM80-1型旋涡自吸泵为研究对象,其扬程高,流量小,工程实践中发现旋涡泵流动诱导噪声较大;为对其流动噪声进行优化,本文基于数值计算对旋涡泵全流场进行流场、声场数值计算,得到噪声的关键影响因素,对其进行仿生设计,在兼顾效率、扬程的同时降低旋涡泵运行噪声。本文主要研究内容如下:1)全流场定常、非定常数值计算及声学边界元数值计算。根据性能要求完成水力设计,试验达到性能要求,定常数值计算与试验值吻合;采用大涡模拟计算,选择隔舌间隙、蜗壳流道关键点作为监测点,对流场内的静压脉动进行时域及频域分析;得到压力脉动主频为叶频、轴频以及其倍频,脉动极值主要集中在1-3倍叶频及1倍轴频处。在非定常计算的同时导出偶极子噪声源作为声学激励源对全流场进行声学边界元计算,以主频点为声学计算频率点;得到隔舌处在主频下声压级较高,声压值明显大于其他区域,同时声源主要分布在隔舌区,确定隔舌间隙为噪声优化研究对象。2)隔舌间隙仿生设计及流场分析。参考鲨鱼表皮非光滑表面特征,将隔舌间隙设计为非光滑表面,初步定性设计梯形截面和曲线截面两种方案;采用大涡模拟分别对原设计方案、梯形截面方案、曲线截面方案进行数值计算;分析隔舌壁面一个周期内的静压、涡通量、涡流以及涡核的非定常特性,得到隔舌间隙壁面静压周期性变换明显,相比于原设计和梯形截面方案,采用曲线截面形式可降低隔舌间隙监测点压力脉动幅值,其壁面涡核分布较离散,壁面涡通量极值分布均匀,可减弱噪声激励源强度。3)仿生试验设计及其声学仿真分析。流场初步分析,确定采用曲线型进一步仿生设计,对曲线参数变量进行控制,构建仿生设计曲线方程,确定试验单因素变量为波长与幅值的比值。建立数学模型,波长与幅值比值作为自变量构建单因素试验样本空间,自变量分别取5.0、7.5、10.0、12.5、15.0。采用数值仿真方法求解样本空间的流场、声场,同时提取扬程、效率、监测点最大声压值等性能参数作为试验目标参数。结果表明,随着自变量数值的增加其效率、扬程逐渐降低;相比于原设计方案,五组试验的扬程平均下降约5%-10%;效率值平均上升约3%-5%;监测点噪声声压级变化较大,波长与幅值比值为5.0,7.5,15.0时声压级提高,声学性能变差;波长与幅值为10.0,12.5时监测点声压最大值分别降低4.8dB、5.75dB。4)数据分析处理。采用最小二乘法对目标参数与自变量进行多项式拟合,拟合优度保证在0.9以上;得到扬程、效率与自变量成二次关系,声压级与自变量成三次关系。对各性能拟合方程分别赋予不同权重,构建综合性能单因素多目标优化数学模型,优化目标参数为旋涡泵额定工况下的扬程、效率、声压最大值;对试验值与原设计性能值的优化幅度加权求和得到其综合性能优化值。其计算步骤为:先确定各性能的权重,之后计算优化值与原设计值差值除以原设计值,得到无量纲优化值;最后加权求和得到综合性能优化值。本研究主要对XKM80-1型旋涡自吸泵进行声学优化同时兼顾其效率、扬程,所以设定扬程、效率、声压级的权重分别为0.2,0.3,0.5。计算可得到波长与幅值比值为10.0时为最优方案,综合性能提升约0.5%。最后基于数值计算数据以及相关的数学模型,采用最小二乘法多项式拟合算法确定优化目标的数学方程,设计基于MFC的数据处理可执行程序,可进行单因素多目标的试验设计的数据处理。
[Abstract]:This research takes the XKM80-1 type vortex self suction pump as the research object, with high lift and small flow. In engineering practice, it is found that the flow induced noise of the vortex pump is larger. In order to optimize the flow noise of the vortex pump, the flow field of the vortex pump is carried out by numerical calculation, the sound field is calculated, and the key influence factors of the noise are obtained. The main research contents of this paper are as follows: 1) the constant flow field, the unsteady numerical calculation and the acoustic boundary element numerical calculation. According to the performance requirements, the hydraulic design is completed, the test meets the performance requirements, the constant numerical calculation is consistent with the test value; the large eddy simulation calculation is used to select the partition. The key points of the tongue gap, the key point of the volute channel as the monitoring point, are analyzed in time domain and frequency domain in the convective field. The main frequency of the pressure fluctuation is the blade frequency, the axial frequency and its frequency doubling. The fluctuating extremum is mainly concentrated at the 1-3 times leaf frequency and the 1 times the axial frequency. The dipole noise source is also guided by the dipole noise source as the acoustic excitation source. The acoustic boundary element method is used to calculate the frequency point of the main frequency point. The sound pressure level of the tongue at the main frequency is higher and the sound pressure is obviously greater than that of other regions. At the same time, the sound source is mainly distributed in the tongue zone, and the gap between the tongue is determined as the noise optimization object.2) the bionic design and flow field analysis of the gap between the tongue gap and the non smooth surface of the shark epidermis are referred to. Surface feature is designed as a non smooth surface, and two schemes of trapezoid and curve sections are preliminarily designed, and large eddy simulation is used to calculate the original design scheme, trapezoid section scheme and curve section scheme respectively, and the unsteady characteristics of static pressure, eddy flux, eddy current and vortex core in one cycle of the tongue wall are analyzed. Compared with the original design and the trapezium section scheme, the pressure fluctuation amplitude of the gap monitoring point can be reduced, the distribution of the wall vortex core is discrete, the wall vortex flux is evenly distributed, and the intensity.3 of the noise excitation source can be weakened. The bionic test design and acoustic simulation of the noise excitation source are reduced. Analysis, preliminary flow field analysis, determine the use of curve type further bionic design, control the parameter variables of the curve, construct the bionic design curve equation, determine the ratio of the single factor variable to the wavelength and amplitude, establish the mathematical model, the ratio of the wavelength to amplitude as the independent variable to construct the single factor test sample space, the independent variable takes 5 respectively. 7.5,10.0,12.5,15.0. uses the numerical simulation method to solve the flow field, the sound field, and the performance parameters such as the lift, the efficiency, the maximum sound pressure value of the monitoring point as the test target parameters. The results show that the lift is gradually reduced with the increase of the value of the independent variable. Compared with the original design scheme, the lift of the five sets of experiments is decreasing. About 5%-10%, the average increase of the efficiency value is about 3%-5%, the sound pressure level of the monitoring point is changed greatly, the acoustic pressure level is increased when the ratio of the wavelength to amplitude is 5.0,7.5,15.0, the acoustic performance becomes worse; the maximum value of the sound pressure of the monitoring point is reduced by 4.8dB and 5.75dB.4 respectively when the wavelength and amplitude is 10.0,12.5. The target parameters and the independent variables are used by the least square method. With polynomial fitting, the goodness of fit is guaranteed to be above 0.9, the relationship between the lift, the efficiency and the independent variable is two, the sound pressure level and the independent variable have three relations. Different weights are given to each performance fitting equation, and the comprehensive performance single factor multi-objective optimization mathematical model is constructed, and the target parameters are optimized to be the lift under the rated working condition of the vortex pump. The maximum value of the sound pressure and the optimum value of the test value and the original design performance value are weighted. The calculation steps are as follows: first determine the weight of each performance, then calculate the difference value between the optimal value and the original design value by the original design value, get the dimensionless optimization value, and finally get the sum of the comprehensive performance optimization value. The XKM80-1 type vortex self-priming pump is optimized with both its efficiency and the lift, so setting the weight of the lift, efficiency and sound pressure is 0.2,0.3,0.5. calculation can get the optimal scheme when the ratio of wavelength to amplitude is 10, and the comprehensive performance is improved about 0.5%. finally based on numerical data and related mathematical models. The least squares polynomial fitting algorithm is used to determine the mathematical equation of the optimized target, and the data processing executable program based on MFC can be designed to deal with the data processing of the single factor and multi-objective test design.
【学位授予单位】：江苏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TH317

【参考文献】