【作者】 王威；

【导师】 王军；

【作者基本信息】 华中科技大学 ， 流体机械及工程， 2022， 博士

【摘要】 叶轮机械的运行效率主要取决于叶道中分离、泄漏、二次流等复杂流动引起损失大小,其形成机理及其流动控制研究是本学科近年来的热点问题。得益于现代测试与数值计算技术的发展,叶道中精细化流动结构的获取并不困难,如何合理、准确的评价各处流动损失及其分布规律成为叶轮设计和优化中的关键。本文依托国家重点研发计划项目课题,针对叶轮内部流动机理及相应损失机制认识不足的问题,建立一种基于熵产理论的损失计算和分析方法,设计构建叶轮机械多参数、多目标的优化,并将两者结合开展风机叶型气动数值优化与实验研究。论文主要工作和研究结论如下:(1)依据湍动能输运与涡粘假设,构建一种湍流条件下的熵产直接计算模型与分析方法,并用三种典型流动结果对该模型进行验证。基于熵产数值计算,发展了一种利用熵产积分的翼型阻力计算方法,对比理论分析与数值结果验证此方法的可行性和准确性。结果表明:与近场积分法和尾迹远场积分法相比,熵产中场积分不仅对物面网格要求低,且数值耗散阻力易分离,有利于得到较精确结果。在高雷诺数状态下,湍流耗散造成的流动损失占主导,当翼型流动失速时,湍流耗散损失占比高达90%左右。(2)完善基于熵产的叶轮机械流动损失分析方法,并将其应用于低压轴流风机和压气机不同工况下的损失诊断。区别于全压损失系数等过程量,该方法不仅能基于当地熵产率准确定位流场中的高损失区域,更可通过熵产率积分对区域损失大小进行量化。结果表明,低压风机近失速工况下的流动损失主要由叶顶泄漏流主导,叶顶区流动损失在叶轮总损失中的比重高达56.5%。跨音速压气机近壅塞工况下的高损失则主要由角区分离导致,叶根处的流动损失占叶轮域总损失的29.5%;但随流量进一步减小,压气机叶顶泄漏明显恶化,叶顶区域的流动损失也提高至40.7%。(3)构建耦合叶型参数化建模、网格自动生成以及并行遗传算法选优的风机叶型气动优化平台,并以最小熵产率为目标,对翼型、叶栅以及三维风机叶片的气动外形展开优化。探索外形参数对翼型升阻力、激波位置以及叶栅性能的影响规律。熵产理论的优化策略成功应用于低压轴流风机三维叶片的多变量、大空间优化,解析出影响流动损失的叶型参数。优化后的两种低熵产风机全压分别提升了42.6%和61.7%,对应的全压效率分别提高3.9%和3.0%。论文基于熵产理论,深入开展风机叶型的气动优化与试验研究,精准给出叶道流动损失分布特征及主要影响因素,为叶轮机械内流减损与性能提升提供内在依据。更多还原

【Abstract】 The operating efficiency of turbomachinery mainly depends on the flow loss caused by complicated flows in the blade passages,such as separation,leakage,and secondary flow.In recent years,the formation mechanism of loss and the corresponding flow control has been a hot issue in this field.Benefit from the development of modern testing and numerical technology,it is not difficult to obtain the refined flow structure in the impeller,but how to evaluate the flow loss and its distribution law accurately has become the key to the design and optimization of turbomachinery.Supported by the national key research and development program,a flow loss calculation method based on entropy generation theory was established in this study,and a multi-parameter and multi-objective optimizer was also constructed.Then the numerical optimizations and experiments are carried out on the shape of airfoils and fan blades.The main work is as follows:1.According to the assumptions of turbulent kinetic energy transport and eddy-viscosity,a direct calculation model and analysis method for entropy generation under turbulent flow are constructed and then verified by three typical flows.Based on the numerical calculation of the entropy,a calculation method of airfoil drag using entropy generation integral was developed,and the feasibility and accuracy are verified through theoretical analysis and numerical examples.The results show that the entropy generation mid-field integration method depends less on the mesh quality than the near-field and the wake far-field integration method.The spurious drag is also decomposed,which is beneficial to the calculation accuracy of the airfoil drag force.The results show that turbulent dissipation is primarily responsible for flow losses of the flow state in a high Reynolds number,and the turbulent dissipation loss reaches 90% when the airfoil stalls.2.The entropy generation loss model for turbomachinery was improved and successfully applied to flow loss analysis of low-pressure fan impeller and high-pressure compressor rotor under different working conditions.Different from process metrics such as total pressure loss coefficient,the entropy generation loss model can not only accurately quantify the loss in any region by integrating the entropy generation,but also effectively locate the high loss area in the three-dimensional flow field by calculating the local entropy generation ratio.It is shown that the flow loss in a low-pressure fan is dominated by the corner separation and the leakage flow;the loss at the blade tip region reaches 56.5%under the near-stall condition.However,the high loss in a transonic compressor under near-choke conditions is mainly caused by corner separations,and the flow loss at the blade root accounts for 29.5% of the total.As the flow rate decreases,the flow loss at the blade tip increases significantly to 40.7% under near-stall conditions.3.An aerodynamic optimization platform for the fan blade was built by coupling parameterized modeling,rapid mesh generation,and a parallel genetic algorithm.Taking the minimum entropy generation as the optimization object,the effects of shape parameters on airfoil lift-drag,shock wave position,and cascade performance was explored.The entropy generation is successfully applied in the multi-variable,large-space,three-dimensional optimization of a low-pressure axial fan,and the factors dominant the flow loss of the fan was also concluded.The total pressure of the two optimized fans with low-entropy generation increased by 42.6% and 61.7%,respectively.The corresponding improvement of their total pressure efficiency is increased by 3.9% and 3.0%.Based on entropy generation theory,in-depth aerodynamic optimization and experimental research on blade profiles were carried out in the present dissertation.The flow loss distribution characteristics and the dominant factor were accurately obtained,which provides internal references for turbomachinery in flow loss reduction and performance improvement.更多还原