{"title":"压气机叶栅湍流的大涡模拟","authors":"Syed Anjum Haider Rizvi, J. Mathew","doi":"10.1115/gt2022-83081","DOIUrl":null,"url":null,"abstract":"\n A study of large eddy simulations (LES) for compressor blade flows has been undertaken and assessed by comparisons with the measurements of Hobson et al. (AIAA J. Prop. Power, 17, 154 (2001)) on a linear cascade of controlled-diffusion airfoils at Reynolds numbers based on blade chord and inflow velocity of 210 000, 380 000 and 640 000. LES of the latter two conditions are reported here. LES is by an explicit filtering method. Compressible flow equations in curvilinear coordinates were solved with 8th-order, central, finite differences and 2nd order Runge-Kutta integration. At 380 000 there is close quantitative agreement of the blade surface pressures and blade normal profiles of velocity magnitudes and fluctuations. Features of the initial development of the suction surface boundary layer is typical of transition induced by freestream turbulence, but end-stage turbulent spots are missing because of premature completion of transition over a mid-chord separation bubble. At 640 000 this separation is absent and turbulent spots appear before transition is completed. Blade surface pressure distribution from the LES agrees closely with that from experiment and is consistent with absence of separation, as are velocity profiles. Velocity profiles from experiment however had shown substantial open separation from the suction surface. This discrepancy had been noted before and was supposed to be an endwall effect, but our LES, including one with endwalls, does not support that view.","PeriodicalId":191970,"journal":{"name":"Volume 10C: Turbomachinery — Design Methods and CFD Modeling for Turbomachinery; Ducts, Noise, and Component Interactions","volume":"60 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Large Eddy Simulation of Turbulent Flow Through a Compressor Cascade\",\"authors\":\"Syed Anjum Haider Rizvi, J. Mathew\",\"doi\":\"10.1115/gt2022-83081\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"\\n A study of large eddy simulations (LES) for compressor blade flows has been undertaken and assessed by comparisons with the measurements of Hobson et al. (AIAA J. Prop. Power, 17, 154 (2001)) on a linear cascade of controlled-diffusion airfoils at Reynolds numbers based on blade chord and inflow velocity of 210 000, 380 000 and 640 000. LES of the latter two conditions are reported here. LES is by an explicit filtering method. Compressible flow equations in curvilinear coordinates were solved with 8th-order, central, finite differences and 2nd order Runge-Kutta integration. At 380 000 there is close quantitative agreement of the blade surface pressures and blade normal profiles of velocity magnitudes and fluctuations. Features of the initial development of the suction surface boundary layer is typical of transition induced by freestream turbulence, but end-stage turbulent spots are missing because of premature completion of transition over a mid-chord separation bubble. At 640 000 this separation is absent and turbulent spots appear before transition is completed. Blade surface pressure distribution from the LES agrees closely with that from experiment and is consistent with absence of separation, as are velocity profiles. Velocity profiles from experiment however had shown substantial open separation from the suction surface. This discrepancy had been noted before and was supposed to be an endwall effect, but our LES, including one with endwalls, does not support that view.\",\"PeriodicalId\":191970,\"journal\":{\"name\":\"Volume 10C: Turbomachinery — Design Methods and CFD Modeling for Turbomachinery; Ducts, Noise, and Component Interactions\",\"volume\":\"60 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-06-13\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Volume 10C: Turbomachinery — Design Methods and CFD Modeling for Turbomachinery; Ducts, Noise, and Component Interactions\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1115/gt2022-83081\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Volume 10C: Turbomachinery — Design Methods and CFD Modeling for Turbomachinery; Ducts, Noise, and Component Interactions","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1115/gt2022-83081","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
对压气机叶片流动的大涡模拟(LES)进行了研究,并与Hobson等人(AIAA J. Prop)的测量结果进行了比较。动力,17,154(2001))在基于叶弦数和流入速度分别为21万、38万和64万的雷诺数下控制扩散翼型的线性级联。后两种情况的LES在这里报告。LES是通过显式过滤方法实现的。采用八阶、中心差分、有限差分和二阶龙格-库塔积分法求解曲线坐标系下的可压缩流动方程。在38万时,叶片表面压力和叶片的速度大小和波动的法向曲线在数量上非常一致。吸力面边界层的初始发展特征是典型的由自由流湍流引起的过渡,但由于在中弦分离泡上过早完成过渡而缺少末端湍流点。在64万时,这种分离是不存在的,湍流斑在过渡完成之前就出现了。从LES得到的叶片表面压力分布与实验结果非常接近,并且与无分离的情况一致,速度分布也是如此。然而,从实验中得到的速度曲线显示出与吸力面有明显的开放分离。这种差异之前已经被注意到,并被认为是端壁效应,但我们的LES,包括一个端壁,不支持这种观点。
Large Eddy Simulation of Turbulent Flow Through a Compressor Cascade
A study of large eddy simulations (LES) for compressor blade flows has been undertaken and assessed by comparisons with the measurements of Hobson et al. (AIAA J. Prop. Power, 17, 154 (2001)) on a linear cascade of controlled-diffusion airfoils at Reynolds numbers based on blade chord and inflow velocity of 210 000, 380 000 and 640 000. LES of the latter two conditions are reported here. LES is by an explicit filtering method. Compressible flow equations in curvilinear coordinates were solved with 8th-order, central, finite differences and 2nd order Runge-Kutta integration. At 380 000 there is close quantitative agreement of the blade surface pressures and blade normal profiles of velocity magnitudes and fluctuations. Features of the initial development of the suction surface boundary layer is typical of transition induced by freestream turbulence, but end-stage turbulent spots are missing because of premature completion of transition over a mid-chord separation bubble. At 640 000 this separation is absent and turbulent spots appear before transition is completed. Blade surface pressure distribution from the LES agrees closely with that from experiment and is consistent with absence of separation, as are velocity profiles. Velocity profiles from experiment however had shown substantial open separation from the suction surface. This discrepancy had been noted before and was supposed to be an endwall effect, but our LES, including one with endwalls, does not support that view.