【作者】 单晓明；

【导师】 黄金泉；

【作者基本信息】 南京航空航天大学 ， 航空宇航推进理论与工程， 2012， 博士

【摘要】 以旋转失速和喘振为代表的流动失稳对航空涡轮发动机安全运行构成极大威胁。流动失稳一旦发生，不但使发动机性能（推力、经济性）大为恶化，而且更严重的是它会引起发动机的突然熄火，或引起压气机叶片剧烈振动以致叶片断裂而造成整台发动机的损坏。因此，有必要在以下两方面开展进一步研究：一是失稳控制技术，探求最有效、快捷的工程办法，能准确地捕获压气机失稳信号，从而快速控制发动机退出非稳定工作状态；二是扩稳技术，探求有效的控制失稳的方法，在准确地捕获到压气机失稳先兆信号后，不仅能控制发动机退出非稳定工作状态，使发动机不发生失稳，而且还能够使发动机喘振裕度增大，将发动机非稳定工作边界外推。鉴于此，本文在分析目前国内外该领域的研究思路和方法的基础上，在涡轴发动机流动失稳在线预报和扩稳技术方面进行部分探索工作。主要工作和研究成果如下：流动失稳信号特征研究：通过对壁面压力信号进行时域信号分析、小波分析以及自相关分析，获取压气机流动失稳信号特征及其发生发展过程，找出表征失速先兆的参数及其判据，进而提出一种基于自相关分析的失稳在线预报方法。分析表明，“Spike”先兆仅可在失速前1～2转通过时域分析发现；小波“斑团”在失速前几十转甚至几百转时就可发现；压气机转子壁面动态压力的自相关系数在靠近失速点降低幅度较大（较大流量点下降约7%～9%），与小波分析观察到的16倍转子频率带处出现的斑团相对应，为压气机流动失稳时壁面压力锯齿波周期性遭到破坏所致。流动失稳在线预报技术研究：基于流动失稳时转子顶部壁面压力信号特征，提出了一种以压力信号自相关系数为判据的流动失稳先兆预测方法，即当自相关系数降低到某一阈值（文中实例为0.87），且概率达到某设定值（30%）时，则认为流动失稳即将发生。在准确捕捉失速先兆信号的基础上，基于现代DSP控制技术，研制了一套针对轴流级压气机流动失稳的在线预报系统，并在某高亚音速轴流压气机试验台上进行了试验验证，在失速前175转（约1.28s）时发出失速报警信号，使流动失稳的预警时间较常规预测方法大幅提前。流动扩稳技术研究：针对轴流压气机低转速下前面级易于出现流动失稳的特点，开展了针对叶顶的一系列喷气扩稳实验。实验结果表明：喷气（主流流量的0.256%）可使轴流压气机稳定工作范围扩大7.1%，效率提高1%，并提出了该方法的工程应用建议，即采用0°即沿轴向喷气，喷气位置安装在动叶前缘附近。通过对叶顶压力信号的小波分析和自相关分析，首次解释了喷气流动扩稳的机理，即喷气能够有效改善壁面压力的周期性，消除导致流动失稳的微小扰动，拓宽流动稳定工作范围。针对某组合压气机离心级的流动特点，专门设计了针对径向扩压器前缘的喷气扩稳系统，试验结果表明，利用喷气量（主流流量的1.5%、2.0%、2.5%），可以有效地改善离心压气机径向扩压器进口的流动状况，提高组合压气机的喘振裕度（0.6%～2.3%），提高效率（0.4%～2.5%）。退喘监控系统研制：基于不同折合转速下涡轴发动机压气机试验结果的统计分析，提炼出了在不同工作状态下发生喘振时压气机进出口压力脉动幅值和频率的变化特征，脉动幅度一般为（40～85）%，喘振频率一般都在（0.5～10）Hz。在此基础上，具体凝练出某型涡轴发动机压气机喘振发生时压力脉动幅值的变化阈值，设计了一种独特、新颖的喘振判别系统。当喘振发生时，喘振判别系统输出阶跃信号，并进行声光报警，启动退喘阀门，从而达到准确测量涡轴发动机喘振边界和快速退出喘振状态的目的。该套喘振在线监控系统在某型涡轴发动机压气机部件台及整机试验台上进行了应用验证，结果表明，涡轴发动机喘振在线监控系统响应时间快（小于100ms），报警准确度高（优于95％）。本文的研究成果可为先进涡轴发动机流动失稳在线监测和扩稳系统的研制提供技术支持和工程参考。更多还原

【Abstract】 The flow instabilities represented by rotating stall and surge are the biggest menaces for aeroturbo engine running safety. When flow instabilities occur, it will not only degrade the engineperformance (power, s.f.c, etc.). More seriously, it might cause an abrupt flameout, or intensevibration of the compressor blades possibly inducing blades rupture and destroy the whole engine.Therefore, it is necessary to study more on the following two aspects. First, the instabilities controltechnology, i.e. exploring the most effective and shortcut engineering way to capture the instablesignal well and truly, sequentially, to control the engine getting rid of the instable operation conditionrapidly. Secondly, the stabilities extending technology, i.e. searching for effective method to controlinstabilities, not only to control the engine running away from the instable condition, but also toenlarge the surge margin, to have more clearance of the engine instable operation line.Hence, by analysis and absorbing the thought and methods in this field domestic and abroad,based on the relevant researches on turboshaft engine, some investigations have been carried on flowinstabilities online predicting and stabilities enhancement technology for turboshaft engine. In thispaper, the following aspects are mainly studied:Research on the flow unsteady characteristics. Based on the analysis of pressure signals obtainedfrom the sensors located on the casing of compressors using time domain, wavelet, andauto-correlation methods, the characteristic and its developing process of the flow instable signal hasbeen abtained. With the found parameter criterion representing the stall foreboding, an on line stallpredicting method based on the auto-correlation has been promoted. Several exciting rerults havebeen revealed by detail analysis in this paper. Stall incepted by spikes can only be detected by timedomain analysis before1～2rotor revolutions, while the spots in wavelet can be found before tens tohundreds revolutions. The amplitude of auto-correlation coefficient of the wall pressure drops rapidlywhen the compressor is close to the stall point (about7%～9%compared to the lager flow rateworking condition), which is consistent with the wavelet spots appearing at the16times frequency ofthe rotor revolution. Both are all caused by the breakage of the compressor casing pressure sawtoothperiodicity when the flow encounters instability.Flow instabilities prediction technology. Grounded on the flow unsteady characteristics of thepressure signal from the rotor tip casing, an instable signal estimation criterion by monitoring thethreshold of auto-correlation declining (exampled as0.87) while it reaches at certain probability (setas30%) has been established. Based on the truly stall precursor detection and modern DSP technology,a set of online instabilities prediction system has been developed and demonstrated on a highsub-sound velocity axial compressor test rig, whose results showed that the stall precursor alarming before stall175rotor revolution (about1.28s) earlier, which brings the pre-warning time muchforward compared with the conventional prediction method.Stabilities enhancement technology: Blade tip injection experiment research on a specific axialcompressor whose front stages tending to unsteady at low speed has been implemented. Resultsshowed that the compressor stability has been improved (7.1%with0.256%mass flow rate injection)and the efficiency has been increased (about1%) correspondingly. Application suggestions have beenproposed, i.e, blow injecting along0°axial line, with the nozzle installed close to the rotor front tip.Both the wavelet and auto-correlation analysis results disclose, for the first time, that the reason ofenhancing stabilities in axial compressor is that the end-wall pressure micro turbulence can besuppressed by jet injection, hence the pressure periodicity can be improved efficiently and the flowstability can be broadened.Demonstration of flow control stabilities enhancement has been applied on the centrifugaldiffuser of a combined compressor, in which the jet holes are drilled on the casing where the instableflow arises more likely. With certain amount of mass flow rate injection (1.5%,2.0%,2.5%of theprime flow mass), the inlet flow field has been improved effectively, featured by the surge abundantenhanced by (0.6～2.3)%, while the efficiency raised by (0.4～2.5)%of the combined compressor.Surge monitoring and control device development. The fluctuant magnitude (40～85)%andfrequency (0.5～10)Hz of the turboshaft engine compressors outlet pressure have been found bystatistically analysis of the tests results under different rotation speeds. Taking the threshold value ofcompressor surge fluctuating pressure, a specific novel surge judging device is designed. When surgeoccurs, a step signal feeds out by the surge judging circuit, to startup the digital acquisition and recordsystem, on the mean time, sound and light alarm, compressor retreats from the surge synchronously.In this way, the purpose of measure the surge line correctly and eliminating surge rapidly are allachieved. This online surge monitoring and control device has been applied on the compressor andturoshaft engine test rigs, where fast response (less than100ms) and high alarm accuracy (better than59%) have been demonstrated.All of the above research results in this dissertation lay a solid foundation and a technical supportfor the design and development of flow stability online management for China next generationturboshaft engine.更多还原