节点文献

燃气涡轮冷却结构设计与气热弹多场耦合的数值研究

Numerical Research of Gas Turbine Cooling Structure Design with Aerodynamic Thermal Mechanical Conjugate Methods

【作者】 陈凯

【导师】 韩万金; 黄洪雁;

【作者基本信息】 哈尔滨工业大学 , 动力机械及工程, 2010, 博士

【摘要】 提高涡轮前燃气温度是提升涡轮效率、增加推重比的重要手段。涡轮前进口温度的不断提高,远远超过了叶栅材料所能承受的温度范围。只有采用有效的涡轮热防护措施才能够保证叶片安全地正常工作。常用的的涡轮叶片冷却方式包括气膜冷却、冲击冷却、扰流肋片、尾缘劈缝等。现代燃气涡轮必须采用多种冷却方式的组合才能达到强化冷却的目的,以满足涡轮前温度升高叶片材料不变对冷却系统提出的新要求。随着CFD技术的不断进步,数值仿真越来越成为涡轮冷却设计中的有力工具,准确预测涡轮叶片温度以及应力分布,是提高涡轮寿命和可靠性的基础。本文的主要工作就是采用气热、热弹耦合计算方法对某燃气轮机叶栅进行数值模拟,通过对计算结果的理论分析,研究影响涡轮冷却效率的因素,并且将CFD计算结果加载到有限元节点上,分析叶片的等效应力状态以及变形量。最后对某两级四列叶栅进行了冷却结构设计,并且根据实际冷却效果,改进初始设计以满足冷却设计要求。几何建模和网格离散是数值仿真研究的重要步骤,对于现代燃气涡轮叶片来说,为了保证叶片正常工作通常采用多种复杂冷却手段,使得叶片总体的冷却结构非常繁杂。本文中对比了传统手工方法建模与参数化叶片建模两种方法,详细介绍了两种方法的建模过程。非参数化的燃气涡轮冷却叶片建模方法灵活、丰富的建模手段能够精细刻画局部冷却结构,多种形式的冷却孔如复合角度孔、后倾扩张孔都能够按照当地的冷却特点布置。但是,该方法存在建模周期长、操作繁琐而且易于出错等弊端,而且离散点的方式生成叶片实体结构精度有限。相对于非参数化的几何模型,涡轮叶片型线经过了参数化,即成为了一种由函数表达曲线的形式。相对于通常采用的一组离散点表达曲线的方法,函数表达的形式避免了由于离散点数量少、间距不合理、位置偏差等问题造成的插值误差,能够在造型过程中保证较高的精度,而且经过了参数化过程,在需要修改叶片几何外形的时候,通过改变相应的控制参数就能方便快捷的生成新的几何模型,大大缩短了设计周期。影响涡轮叶片冷却效率的因素有很多,本文中研究了不同主流湍流度和不同冷却气体配比条件下对于冷却效率的影响。结果发现,主流湍流度对于叶片型面上的压力系数和速度比的影响较小,而于对壁面上的换热作用影响较大。在叶片前缘处,低湍流度方案下的的冷却效率略高于高湍流度方案,低湍流度的情况减少了喷射冷气与主流的作用,使得冷气覆盖情况比高湍流度下的情况要好。而在叶片中后部区域,加速流动的主流在高湍流度下使得冷气的贴壁性较好,因此,叶片表面的冷却效率在高湍流度下要略高于低湍流度的情况,在叶片尾缘处,由于冷却气膜的散失,失去了对叶片表面的保护作用,各湍流度下的表面冷却效率趋于一致。在不同冷却气体流量配比条件下,高的喷射动量对压力面冷却是有益的,平均冷却效率整体均略高于低冷却气体流量配比的情况。在吸力面,从总体上看冷却射流动量的增加降低了冷却效率,高喷射动量的方案仅仅在叶片前缘迎着主流的小范围内提高了冷却效率,而后低喷射动量的方案冷却效率一直高于高喷射动量的方案。对某燃气轮机四列冷却叶栅进行了冷却结构的初始设计与改进,对初始设计的冷却结构进行了冷却效果的校核,并且以计算结果为基础改进了冷却效果不达标的区域。在冷却结构设计过程中,以典型的导叶和动叶冷却结构为基础,如导叶的全气膜覆盖、冲击冷却、尾缘扰流肋柱结构、动叶的带肋蛇形通道、光滑蛇形通道结构。对初始设计的冷却结构进行气热耦合计算,然后修改局部没有达到冷却要求的区域,通过参数化的几何模型快速生成新的模型,从而大大节省工作时间。

【Abstract】 Increase the turbine inlet temperature was the important way which could improv the efficiency and thrust to weight ratio. At present, the inlet temperature has exceeded the turbine material sustainable limit. It was necessary to adopt effective cooling techniques which made the turbine run well. The most effective turbine cooling methods were film cooling , impingement cooling, turbulent ribs, trailing edge slot and so on. Compound cooling methods should be employed to fulfill the demand of modern gas turbine. With the rapid development of CFD technique, numerical simulation became powerful tools of turbine cooling structure design. Predict the blade temperature and stress status exactly was the foundation to improve the turbine cycle life and reliability. The main work of the dissertation investigated several rows of turbine blades using conjugate and thermal mechanical methods. The cooling efficiency which influenced by corresponding factors was studied through numerical analysis. After that, the CFD results was transferd to FEA mesh nodes and blade stress status and total deformation were presented. Finally, cooling structure was designed for four rows of gas turbine blades. The initial cooling configuration was improved according to the practical cooling effect.Geometry modeling and mesh generation was the key step of numerical simulation. In order to protect the blade, muiltple cooling mehthods were employed which made the whole cooling configuration comlicated. The dissertation compared two types of modeling methods which one was non parameterized methods and another was parameterized methods. The detailed modeling process was introduced. The non parameterized modeling methods was flexible to set cooling configuration at any area and many types of cooling holes, such as compound holes, laid back diffuse holes could be arranged based on the blade local flow character. Therefore, long time consumption, fussy operation and easy to make mistakes were the disadvantages of non parameterized modeling methods. The accuracy of the non parameterized modeling method was also limited. Compara to the non parameterized modeling methods, the blade profile became function format when parameterized methods was employed. This format avoid the drawbacks of interplotion error which result form less control points, non reasonable distance, location deviation and maintain high precision in the modeling process. When the blade was parameterized, it could esay to generate new geometry through control parameters and short the design period.The cooling efficiency was influenced by many factors,this dissertation studied the influence of different main flow turbulent intensity and cooling air mass flow rate. It was concluded that the main flow turbulent intensity has much more influence on blade surface temperature rather than Cp and Cu. The higher cooling efficiency was appeared in low turbulent intensity case at the leading edge. Low turbulent intensity reduce the mix of cooling air and main flow which result to high cooling efficiency. At the middle and rear of the blade , the accelerated mainflow made the cooling air close to the surface. Therefore, the cooling efficiency was little higher in high turbulent intensity case. Because of the film cooling air dissipatation, the cooling efficiency became identical at the rear of the blade in both two cases. The pressure side film cooling benefited from the high cooling flow injection. On the suction side, the average cooling efficiency was lower in that case. The high injection momtum just improved the situation at leading edge aera and the cooling efficiency in low turbulent intensity case was higher at the rest of areas.Initial cooling configuration design was implemented for four rows of gas turbine blade and the cooling effect was testified. Low cooling efficiency area was improved base on the calculation. In the cooling configuration design process, typical turbine blade cooling structure was adopted such as, full film cooling, impingement cooling, trailing edge turbulent ribs, rib roughed serpentine passage, smooth serpentine passage. The initial cooling configuration was testified using conjugate method and local cooling structure was modified. The new model was generated through parameterized blade which save much more time.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络