Theoretical Modelling of Ingress in Gas Turbine Rim Seals.

工程师们的目标是在不牺牲材料寿命的情况下提高燃气轮机的效率。入口是来自主流的热气体被吸入静止机壳和相邻旋转圆盘之间的轮空的现象。为了防止或减少进入，在阀瓣的外围安装了轮缘密封件，密封空气直接从压缩机中排出，对内腔施加压力。然而，使用太多的密封空气会降低效率，而使用太少的空气会因为高温而损坏材料。西门子目前使用巴斯大学孔口模型为其发动机设计轮缘密封件。在这里，只考虑了惯性效应，而流量系数被用来考虑粘性效应和其他相关损失。这些系数不能从理论上确定，它们被视为每个密封的经验常数。孔口模型被用来预测入口，尽管它与基于浓度的实验结果很好地吻合，但对于不同的轮缘密封几何形状(即，轴向和径向轮缘密封)，压力差预测出现了问题。最近，Svov和Atkins提出了另一种方法，用湍流扩散将入口模拟为粘性问题，而忽略了惯性影响。本项目的目的是将目前被Bath方法忽略的小孔模型和湍流扩散模型中的粘性项结合起来，开发一种新的用于预测入口的控制体模型。新模型有望解决压差预测问题，同时与基于浓度的有效性测量保持良好的一致性。学生将把实验和理论部分结合起来，以便更好地理解和解释结果。此外，将CFD用于验证目的是可取的，然而，在这个阶段，在多大程度上还没有决定。在项目开始时，学生将介绍有关入口的现有文献，同时在有经验的操作员的监督下在1.5级涡轮试验台上进行实验；稍后，学生将成为试验台的主要操作员。第二阶段包括收集压力分布和集中效果数据，这些数据将为建模提供信息。然后将对西门子特定的轮缘密封设计进行参数研究。最后，在与第二阶段重叠的第三阶段，学生和导师将把实验测试结果合并到新的理论模型中进行验证。通过该项目，希望透平机械研究中心(TRC)的同事将通过比较他们的结果并改进他们的模型和代码来参与模型的验证和验证。这将构成一个3.5年的博士课程，到2022年将建立一个新的控制量模式。

Engineers aim to improve the efficiency of gas turbines without sacrificing the life of materials. Ingress is the phenomenon where hot gas from the mainstream flow is ingested into the wheel-space between the stationary casing and the adjacent rotating disc. To prevent or reduce ingress, rim seals are fitted at the periphery of the discs and sealing air, which is directly bled from the compressor, pressurises the internal cavity. However, using too much sealing air reduces the efficiency, while too little can damage the materials due to high temperatures. Siemens currently use the University of Bath Orifice Model to design rim seals for their engines. Here, only the inertial effects are taken into account, while discharge coefficients are used to account for the viscous effects and other related losses. These coefficients cannot be determined theoretically and they are treated as empirical constants for each seal. The Orifice Model is used to predict ingress, and although it is in good agreement with concentration based experimental results, problems with pressure difference predictions arise for different rim seals geometries (i.e., axial and radial rim seals). Recently, Savov and Atkins proposed an alternative approach, where turbulent diffusion is used to model ingress as a viscous problem while ignoring the inertial effects. This project aims to develop a new Control Volume Model for the prediction of ingress by uniting the Orifice Model and the viscous terms of turbulent diffusion model which are currently ignored by the Bath approach. The new model is expected to solve the problem of the pressure difference prediction while maintaining good agreement with concentration based effectiveness measurements. The student will combine the experimental and theoretical parts for a better understanding and interpretation of the results. Furthermore, it is desirable to involve CFD for validation purposes, however, at this stage, it is undecided to what extent. At the start of the project the student will cover the available literature around ingress, while conducting experiments in the 1.5-stage turbine rig with the supervision of an experienced operator; later the student will be the main operator of the rig. The second phase includes the collection of pressure distribution and concentration effectiveness data which will inform the modelling. A parametric study of Siemens specific rim seal design will then be conducted. Finally, during the third phase, which overlaps with the second, the student and the supervisor will incorporate the experimental test results into the new theoretical model for validation.Throughout the project, it is expected that colleagues from the Turbomachinery Research Centre (TRC) will be involved for the validation and verification of the model by comparing their results and improving their models and codes. This will constitute a 3.5 - year PhD programme and by 2022 a new Control Volume Model will be established.