Eksenel gaz türbini kanat ucu geometrisinin hesaplamalı akışkanlar dinamiği ile aeroternal tasarımı

Abstract

Türbomakina performansı kanat ucu boşluğundaki akış yapısına bağlı olup, üç boyutlu ve karmaşık bir yapıya sahiptir. Kanat ucu boşluğunda, basınç ve emme kenarı arasındaki yüksek basınç farkı nedeniyle sızıntı akışı oluşur. Sızıntı akışının türbin kanadında oluşan diğer ikincil akış yapılarıyla etkileşimi ve oluşan karmaşık akış yapıları hem rotordaki aerodinamik kaybın hem de kanat ucuna ısı geçişinin ciddi oranda artmasına neden olur. Sızıntı akışının olumsuz etkilerini azaltmak için literatürde çeşitli pasif kontrol yöntemleri geliştirilmiştir. Squealer kanat ucu, gaz türbinlerinde yaygın olarak kullanılan bir pasif kontrol metodur. Kanat ucu yüzeyinde boşluklu bir yapı oluşturularak elde edilmektedir. Kanat ucunu mekanik hasarlardan korumak amacıyla geliştirilmiş bir tasarımdır. Kanat ucunun gövde ile temasını engelleyerek olası bir mekanik problemde kanat yerine squealer yapısının zarar görmesi amaçlanır. Bir diğer pasif kontrol metodu ise emme kenarı kısmi squealer kanat ucudur. Bu yöntemde squealer yapısı yalnızca kanat emme kenarı boyunca bulunmaktadır. Tez kapsamında eksenel gaz türbinlerinde düz, squealer ve emme kenarı kısmi squealer kanat ucu modellerinin aero-termal karakteri ve akış fiziğine etkileri Hesaplamalı Akışkanlar Dinamiği (HAD) çözümleri ile sayısal olarak incelenmiştir. Sıkıştırılamaz, daimi RANS denklemleri parametrik HAD analizleri ile çözülmüştür. Türbülans modeli olarak türbomakina gibi karmaşık akış yapılarına sahip akışlarda yaygın olarak kullanılan ve pek çok çalışmada deneysel verilerle doğrulanan SST k-ω kullanılmıştır. Kanat ucu ve gövde arasındaki kanat ucu boşluğu t/h tüm modeller için sabit olup %1.0 alınmıştır. Sayısal modelleme lineer kaskatta ve sabit gövde ile yapılmıştır. Tüm modeller için toplam basınç kayıp katsayısı, sızıntı debisi, toplam ısı yükü ve ortalama ısı geçiş katsayısı hesaplanmıştır. Squealer ve emme kenarı kısmi squealer uygulamalarının aero-termal performansı incelenmiş ve referans geometri olan düz kanat ucu ile kıyaslanmıştır. Sonuçlar incelendiğinde squealer kanat ucunun hem düz hem de kısmi squealer kanat ucu modellerine göre daha düşük sızıntı debisi ve daha düşük aerodinamik kayıp yarattığı bulunmuştur. Bu nedenle, aerodinamik açıdan açıkça daha üstündür. Squealer kanat ucu aerodinamik kaybı düz kanat ucuna göre -%7.7 azaltırken, emme kenarı kısmi squealerına göre -%6.7 azaltmıştır. Sızıntı debisi en düşük değerini squealer kanat ucunda ve en yüksek değerini kısmi squealer kanat ucunda almıştır. Squealer kanat ucu sızıntı debisini düz kanat ucuna göre -%20.9 azaltırken, kısmi squealer %11.1 artırmıştır. Termal açıdan ise en üstün tasarım emme kenarı kısmi squealer kanat ucudur. Emme kenarı kısmi squealerı ortalama ısı geçiş katsayısını düz kanat ucuna göre -%40.6 azaltırken en yüksek performanslı squealer kanat ucuna göre -%33.9 azaltmıştır. Literatürde squealer yüksekliği ve squealer genişliğinin aerodinamik ve termal performansa etkisiyle ilgili geniş bir aralıkta incelenmiş sistematik bir çalışma bulunmamaktadır. Tezin ana amacı squealer yüksekliği ve squealer genişliğinin aerodinamik ve termal etkilerinin geniş bir aralıkta incelenmesidir. Hem squealer yüksekliğinin hem de squealer genişliğinin akışın fiziğine etkileri, sızıntı debisi, aerodinamik kayıp ve ısı geçişi üzerindeki etkisi detaylı olarak irdelenmiştir.

In order to allow the relative motion of blades and to prevent the blade tip surface from rubbing, clearance gaps between blades and casing are required in most turbomachinery systems. The overall performance of the turbomachines is strongly related to the flow within tip gap. The flow in tip gap is 3-dimensional and highly complex. The pressure difference across the pressure and suction side of the blade forms a leakage flow passed over the blade tip surface. The pressure driven flow throughout the gap results in approximately one-third of the aerodynamic loss in the rotor of an axial gas turbine. The flow structure in the tip gap is a significant source of inefficiency in terms of aerodynamic loss and heat transfer to the blade tip and casing. The leakage flow passes over the blade tip without being turned as the passage flow, thus a reduction in work extracted from the turbine is observed. The leakage flow is also a significant source of higher thermal loads on the blade tip platform which is exposed to the hot gas stream. Leakage flow and its interaction with other secondary flows result in complex flow field around the turbine blade. It is responsible the aero-thermal inefficiency. In order to reduce the negative effects of complex secondary flow structures, there are several passive control methods in the literature such as squealer, partial squealer, winglet and carved blade tip. In the thesis study, numerical investigation of aerodynamic and thermal characteristics of squealer and PS partial squealer blade tip in a low-speed high-pressure unshrouded turbine rotor. The squealer tip provides an aerodynamic seal effect in the tip gap and an effective reduction in aerodynamic loss. It also protects the blade tip against the impact of hot leakage gases. Even the PS partial squealer tip is ineffective in terms of aerodynamic aspect, however it provides an excellent thermal resistance to high temperature leakage gases. Both two passive control methods have been widely used in modern gas turbines. In this research, the effect of squealer and PS partial squealer tip section were examined in terms of the flow and heat transfer. Reference geometry for the comparison was the flat tip. The aero-thermal performance of the squealer and PS partial squealer were compared with the flat tip. Complex secondary flow structures near the turbine blade and detailed flow physics were investigated. A numerical study was carried out in this study. Although experimental investigations in the field of turbomachinery provide great insight into the flow physics, they may become expensive, difficult at times and time consuming. The Computational Fluid Dynamics (CFD) method becomes a significant tool to analyze the complex flow structure within the tip gap region of a turbomachine. A special emphasis was placed on obtaining three dimensional and complex grid systems in a parametric effort. The number of the cases for CFD computations, the solid model and the grid generation became significant productivity issues. The parametric study considerably reduced the production time for complex tip configurations and the grid generation process. The axial turbine blade profile and experimental datas belongs to the Pennsylvania State University Axial Flow Turbine Research Facility (AFTRF). The tip profile of the AFTRF rotor blade that was used to create an extruded solid model of the axial turbine. The computational domain was obtained as a linear turbine cascade arrangement for a single blade passage. The inlet domain has the length of 1.0Ca and the outlet domain 3.0Ca. Circumferential periodicity in the tangential direction was imposed. Tip clearance (t/h) of the blade was 1.0% for all cases. Numerical model was carried out with stationary casing. The computational domain was divided into multi-blocks as inlet, rotor and outlet domain. The number of blocks was 37 in order to provide a parametric definition and achieve a fully hexagonal grid in a simple way. Creating a multi-block flow domain enabled to use the multizone method in the ANSYS Meshing. The multizone method was used for the grid generation. Multizone, a type of blocking approach similar to ICEM CFD uses automated topology decomposition and generates a structured hexagonal mesh where blocking topology is available. Fully hexagonal elements were used in calculations to reduce the solution time and increase the accuracy as shown. The average number of elements in the grid was around 5 – 7 millions. An O-Grid topology for the boundary layer mesh was introduced to keep the y+ value at a reasonable level. Its averaged value was 0.8 around the blade profile at the 97% of the span and lower than the 1.5 in this study. Numerical calculations were obtained by solving the 3D, incompressible, steady and turbulent form of the Reynolds-Averaged Navier-Stokes (RANS) equations were solved with a finite volume discretization using commercial code ANSYS Fluent 16.0. Two equation turbulence model, SST k-ω was used. It yields close agreement with the experimental datas in the passive control literature. In order to use SST k-ω model, it is recommended to keep y+ values smaller than 2. For all cases, the y+ condition was satisfied. Mass flow inlet and static pressure outlet boundary conditions were imposed. At the turbine inlet and outlet section, turbulence intensity and hydrodynamic length were defined as 0.5% and 110.36 mm respectively. Inlet velocity triangle at tip section, inlet mass flow and outlet static pressure datas were taken from the AFTRF rotor. For thermal boundary conditions, inlet and wall temperature were at 50°C and 25°C respectively. Maximum velocity in the computational domain was less than 102 m/s, thus compressibility effects were not considered. No slip boundary condition was applied to all cascade walls. Convergence level of the continuity, momentum, k and ω equations was the order of 10-3 and for the energy equation 10-6. Difference between mass flow at the inlet and outlet was monitored. Static pressure point monitor for the convergence test was introduced at the tip gap midpoint. The convergence test satisfied a sufficient level for the convergence. Grid independence test was performed for coarse, medium and fine mesh of the flat tip. For the numerical validation ANSYS CFX and ANSYS Fluent codes were compared in terms of flow and heat transfer. In the literature, there was not any study about effect of squealer width and height on both aerodynamic and thermal performance in a wide range. For this reason, the main aim in the thesis study was understanding the effects of squealer width and height on the aero-thermal performance of squealer tip. In this study, also squealer side walls were take into account for the heat transfer calculation. Numerical analysis were performed for 28 different squealer tip geometries. 4 different squealer heights and 7 different squealer widths were computationally investigated with regards to leakage mass flow, aerodynamic loss and heat transfer coefficient. Significant effects were discovered and explained in detail. From the numerical calculations, squealer tip reduced the aerodynamic loss 7.7% and 6.7% and the leakage mass flow 20.9% and 32.0% compared to flat and PS partial squealer tip respectively. From the aspect of thermal performance, PS partial squealer tip much better than the other passive control methods. It reduced the averaged heat transfer coefficient 40.6% and 33.9% in comparison to flat and squealer tip respectively. In conclusion, while squealer tip had the highest aerodynamic performance, PS partial squealer tip had the thermal performance.