Gas Turbine Cooling System Design Procedures

Introduction

State-of-the-art gas turbine engines usually work under extremely high temperatures. This is directly related to efficiency of the gas turbines – in order to receive the maximum thermodynamics value, it is necessary to increase the gas temperature after the combustion chamber. Engine temperature can be higher than blades’ metal temp up to 500-600 K. Blades, nozzles, and the GT details are manufactured with special heat-resistant steels and in some cases, they require a special coating. That allows them to resist turning into liquid metal under these working temperatures like the T-1000 did in the “Terminator 2: Judgment Day” movie even under high temperatures :).

Picture 1 – T-1000 from Terminator 2
Picture 1 – T-1000 from Terminator 2. Source

However, metal has the property of “creep” – this is the tendency of hard metal to move slowly or permanently deform under stress. This occurs as a result of prolonged exposure to high stresses above the yield point, especially when exposed to high temperatures. Obviously, the solution to this problem is a cooling system for heat-stressed parts, which has allowed the gas temperature to increase by 600 K compared with uncooled machines. Since the gas turbines usually work with air, the simplest way to cool the system is by using this. Typically, the air exhausts to different parts of the compressors and is supplied to the cooling paths and blades which influence the thermodynamics efficiency of the gas turbine engine. Thus, it is crucial to ensure enough cooling to remove the heat on the one hand and on the other hand – to receive the lowest amount of air which requires cooling.

Picture 2 – Heavy Duty Gas Turbine, GT 26
Picture 2 – Heavy Duty Gas Turbine, GT 26. Source

Have you ever thought about how complex a gas turbine cooling design is?

The secondary flow cooling system must ensure high reliability throughout the entire service life, minimum possible airflow rates and pressures, absence of leakages, additional stresses arising from uneven cooling of individual parts should be absent or compensate for stresses from the action of gas and centrifugal forces, maximum temperatures and temperature differences in all modes must be within the limits permissible for a particular GTE. Moreover, the cooling depth is increasing with gas temperature growing before the turbine.

The requirements are quite complex and contradictory. No matter how complex, the design of any cooling system could be divided into design estimation as the first stage and verification of the design as the second stage. Not so long ago, the design task also required experimental expansive performances. Now it is expedient to perform the design of the cooling system as a series of verification calculations and consider both fluid-dynamic and thermal conduction in one approach.

To address these issues the Conjugate Heat Transfer (CHT) analysis can be used as an approach to improving the estimation of the requirements to the gas turbine engine.

Design of Gas Turbine Blades and Nozzles Approaches

The most thermal stressed parts in any gas turbine engine are undoubtedly the stator nozzles and rotor blades. If the maximum total temperature less than 1270 K before the nozzles and less than 1170 K before the blades, then the cooling system is not required. A total temperature increasing of every 100 K requires a cooling method be applied such as convective (internal cooling), film or transpiration cooling. (More info about the approaches is described in our previous blog).

Traditional approaches to estimating design tasks and obtaining temperatures and loads have evolved with computation methods. Mainly the procedures can be classified into methods call a Coupled Approach and a Decoupled Approach. The Coupled Approach allows coupling interfaces to become part of the solution domain (fluid and solid) and solving simultaneous solutions of the equation set which characterizes different fields. In contrast, the Decoupled Approach provides the opportunity to solve each field separately and bind the boundary conditions to each other even using the different solving tools. In a practical way, these analysis methods of the blades and nozzles can be evaluated in three ways. Thus, the gaining of approximated 1D correlations for internal and external turbine blade flow-field provides the possibility to apply them and estimate necessary data, which could be used as inputs to 3D finite element analysis. Such an approach requires a number of iterations because of the preliminary blade temperature assumption.

Picture 3 – 1D Correlations-based and 3D FEA Analysis Approach
Picture 3 – 1D Correlations-based and 3D FEA Analysis Approach

The 3D CFD-based approach was realized by separate 3D CFD simulations of the internal and external blade flow path. Results obtained from these simulations were used to 3D the FEA estimate metal temperature gradients. The CFD-based approach also required the initial value of the temperature to assume the iteration process. The third approach is CHT/CFD analysis which assumes performing the entire multiphysics design and calculates internal and external flow paths together while obtaining the blade temperature.

Picture 4 – CHT/CFD - Based Design System
Picture 4 – CHT/CFD – Based Design System

Entire Secondary Flow Path Design

The design of the entire cooling path requires estimation of flow and leakage through a cooling path which can include rotor-rotor and rotor-stator cavities, seals, orifices, deflectors, and the other components and channels of the system. Each element influences the flow creating resistance and, as a consequence, pressure losses. It should be noted that airflow inside the system is swirled, because of the influence of rotating walls on the flow, the absolute velocity contains the tangential component. Thus, accurate estimation of the pressure distribution in the secondary flow path provides the value of load influenced on the bearings and can be used in the structural calculations.

The organization of the air supply to the blades is also of great importance. For example, the direct air supply to the blades leads to an increase of cooler total temperature compare with the temperature of air at the exhaust parts because of turbine rotation. In schemes with pre-swirled air, conversely, the temperature is decreased which is a suitable solution.

Aero-Engine High Pressure Turbine and Combustor
Picture 5 – Aero-Engine High Pressure Turbine and Combustor (Source: R. S. Bunker – “Cooling design analysis”)

Usage of 3D CFD based simulation or CHT/CFD approach for the entire cooling path design is a pretty complex task. These approaches are time-consuming and resource-intensive for the full cooling flow path including all system components and blades. To address this challenge, it is reasonable to consider a collaborated CHT/CFD approach and 1D accurate method for estimation the secondary cooling flow path. The internal cooling system is advisable to model using a 1D thermal-fluid network, which contains heat transfer coefficient and the pressure loss correlations, which were evaluated, validated during years, and provides the possibility to accurately calculate flow through elements of the system. The blades and nozzles estimation can be provided by CHT/CFD approach. That integrated approach can be successfully performed via using 1D steady-state and unsteady AxSTREAM NET™ software for internal cooling path design  and three-dimensional STAR-CCM+ software for CHT/CFD calculation of the blade via AxSTREAM ION™, which provides the possibility to link tasks and solve the residuals between models. This approach will save time while obtaining an accurate design of the cooling path and optimizing the system as well.

Picture 6 – Internal cooling flow path modeling in AxSTREAM NET
Picture 6 – Internal cooling flow path modeling in AxSTREAM NET™

The design of the gas turbine cooling path requires the solving of complex and contradictory issues. Nowadays, computation methods can address most of them, but for the entire system design, it is necessary to take into account the time of the project and resources. Decreasing the time while obtaining accurate results can be achieved using a multidisciplinary toolset, which covers everything from secondary cooling flow path prediction to thermal and structural solutions. On our upcoming webinar “Cooled Gas Turbine Simulation Coupling AxSTREAM NET™ & STAR-CCM+” we will discuss the prediction of secondary flow path using a 1D approach alongside CHT/CFD calculation. Register here to see this process in action

If you need more information about integrated modeling, please contact the SoftInWay team at info@softinway.com.

References:

  • Shvetcs I. T., Diban E. P. “Air cooling of gas turbine details”, 1974.
  • Oleynik A. V., Sharkov S. Y. “Calculation of thermal strain state of cooled blades and turbines”, 1995.
  • Andrei, A. Andreini, B. Facchini, L. Winchler “A decoupled CHT procedure: application and validation on a gas turbine vane with different cooling configurations”, 2014.
  • S. Bunker – “Cooling design analysis”.
  • Siemens Digital Software “Siemens energy uses 3D multiphysics-based virtual prototyping to improve has turbine efficiency and reliability”, 2017;
  • S. Akerman. L. A. Zarubin, V. P. Reshitko, A. V. Rosinskaya – “Gas turbine GTE-115M”, 2009.