This paper presents a numerical study to investigate the feasibility of transporting mist through the internal cooling channel in high-pressure turbine vanes for film cooling over the vane’s surface. The idea of using mist film cooling to enhance conventional air cooling has been proven to be a feasible technique in the laboratory conditions and by computational simulations. However, there is a challenge to this technique to prove that the water mist can survive in the very hot environment inside the gas turbine casings and internal air passages and be delivered to the film injection holes. Both a zero-dimensional mist evaporation analytical model and 3-D computational fluid dynamic (CFD) scheme are employed for analysis. In the CFD simulation, the Lagrangian /Eulerian method is used along with the discrete phase model (DPM) to track the evaporation process of water droplets. The high-pressure water mist is injected into the stream of cooling air extracted from the compressor through the outer gas turbine casing near the vane before it reaches the vane internal cooling cavity. Using the mist equivalent of 10% of the cooling air mass flow rate, the results show that, when the liquid droplets are atomized to 30 μm in diameter initially, the droplets can survive inside the internal cooling passages and be delivered to the film cooling injection hole location with droplets of 20 μm in diameter; and alternatively, an initially 20 μm droplet can be delivered at 12μm in diameter, which is sufficiently large for completing the required external film cooling task.

Motivated by the need to further improve film cooling in terms of both cooling effectiveness and coolant coverage area, the mist/air film cooling scheme is investigated through experiments using fan-shaped holes over an extended downstream length in this study. Both an existing wind tunnel and test facility, used in previous work, have been retrofitted. The first modification was extending the length of the flat plate test section to cover longer distances downstream of the injection holes, up to X/D = 100, in order to investigate whether mist cooling can be harnessed farther downstream where single-phase film cooling is not effective. X represents the axial distance downstream of the cooling hole of diameter D. The second modification was to incorporate fan-shaped (diffusion) holes which are proven to have a higher film cooling efficiency, than cylindrical holes. The objective is to investigate whether mist can further enhance the film cooling performance of the already highly effective fan-shaped holes. A phase Doppler particle analyzer (PDPA) system is employed to measure the droplet size, velocity, and turbulence information. An infrared camera and thermocouples are both used for temperature measurements. Part I is focused on the heat transfer result on the wall. The results show that, at low blowing ratios when the film is attached to the surface, the enhancement of the mist film cooling effectiveness, compared to the air-only case, on the centerline of the hole ranges from 40% in the near hole region to over 170% at X/D = 100. Due to the diffusive nature of the fan-shaped hole, the laterally averaged enhancement is on par with that on the centerline. The significant enhancement over the extended downstream distance from X/D = 40–100 is attributed to the evaporation time needed to evaporate all of the droplets. Each droplet acts as a cooling sink and flies over a distance before it completely vaporizes. This “distributed cooling” characteristic allows the water droplets to extend the cooling effects farther downstream from the injection location. At higher blowing ratios, when the cooling film is lifted off from the surface, the cooling enhancement drops below 40%. Although the enhancement in the near hole region X/D < 40 is about 20% lower than that achieved by using the cylindrical holes, the magnitudes of the mist adiabatic film cooling effectiveness using fan-shaped holes are still much higher than those of the cylindrical holes. Part II of this study is focused on analyzing the two-phase droplet multiphase flow behavior to explain the fundamental physics involved in the mist film cooling.

Modeling liquid droplet evaporation in a flow stream is very important in many engineering applications. It was discovered that the result of predicted droplet and main flow temperatures from using commercial codes sometimes presents unexplainable phenomena; for example, the droplet temperature drops too low. The objective of this study is to investigate the issues involved in the built-in droplet evaporation model by using three different approaches: (a) use the existing built-in correlations model in a commercial code, (b) use the lumped analytical analysis, and (c) actually solve the heat and mass transfer by directly using CFD without employing the built-in correlation model. In the third approach, the evaporation process is simulated by imposing water evaporation in a very thin layer at the surface of a stagnant water droplet; in the meantime, the evaporation energy is subtracted from the same place. This is performed by imposing a positive mass source term and a negative energy source term in a thin layer of cells wrapping around the droplet surface. The transport equations are then solved using the commercial CFD solver Ansys/Fluent to track the mass and energy transfer across the shell sides into the liquid droplet and out to the ambient. Unlike the built-in evaporation model in commercial codes, which assumes that all the evaporation energy (latent heat) is supplied by the droplet, in the direct CFD calculation, the evaporation energy is absorbed partly from the droplet and partly from the surrounding air according to the natural process based on the property values and the heat and mass transfer resistance inside and outside the droplet. The direct CFD result (without using evaporation correlation) is consistent with that of the lumped analytical analysis (2nd approach). During the development of the direct CFD calculation, several technical difficulties are overcome and discussed in detail in this paper. A revised equation is proposed to improve the existing built-in model in the current commercial code. Both the direct CFD method and the zero-dimensional lumped method show the droplet temperature always increases.

This paper presents a numerical study to investigate the feasibility of transporting water mist to the rotating blades of a high pressure turbine. The idea of using mist film cooling to enhance conventional air cooling has been proven to be a feasible technique under laboratory conditions. However, there are challenges in implementing this scheme for real gas turbine systems. The first challenge is how to transport the mist to the rotating blades and the second challenge is delivering the mist to the injection holes and getting the particles to survive within the harsh gas turbine environment. Both a zero-dimensional mist evaporation analytical model and a 3D computational fluid dynamics (CFD) scheme are employed for analysis. In the CFD simulation, the Lagrangian-Eulerian method is used along with the discrete phase model (DPM) to track the evaporation process of each individual water droplet. For transporting the mist to the blades, the high-pressure water mist is injected into the stream of cooling air extracted from the compressor through two different passages. The first passage passes through the rotor cover-plate cavity before entering the blade base. The second passage passes through a diaphragm box on the base of the second vane, then tangentially through a cooling passage in the rotating shaft, and eventually to the blade base. The results show that it is feasible to transport the mist from the turbine casing to the blade through both passages, provided that droplets with sufficient particle diameter and mist loading are used. The shorter passage, through the nozzle diaphragm, alleviates a lot of challenges facing the passage through the blade cavity, and seems to be more practical. A side benefit of transporting mist through the internal passages is the additional cooling of the pre-swirler and rotor cover plates. The results are encouraging for implementing the mist cooling technique under real gas turbine conditions.

A phase Doppler particle analyzer (PDPA) system is employed to measure the two-phase mist flow behavior including flow velocity field, droplet size distribution, droplet dynamics, and turbulence characteristics. Based on the droplet measurements made through PDPA, a projected profile describing how the air–mist coolant jet flow spreads and eventually blends into the hot main flow is prescribed for both cylindrical and fan-shaped holes. The mist film layer consists of two layers: a typical coolant film layer (cooling air containing the majority of the droplets) and a wider droplet layer containing droplets outside the film layer. Thanks to the higher inertia possessed by larger droplets (>20 μm in diameter) at the injection hole, the larger droplets tend to shoot across the coolant film layer, resulting in a wider droplet layer than the coolant film layer. The wider droplet layer boundaries are detected by measuring the droplet data rate (droplet number per second) distribution, and it is identified by a wedge-shaped enclosure prescribed by the data rate distribution curve. The coolant film layer is prescribed by its core and its upper boundary. The apex of the data rate curve, depicted by the maximum data rate, roughly indicates the core region of the coolant film layer. The upper boundary of the coolant film layer, characterized by active mixing with the main flow, is found to be close to relatively high values of local Reynolds shear stresses. With the results of PDPA measurements and the prescribed coolant film and droplet layer profiles, the heat transfer results on the wall presented in Part I are re-examined, and the fundamental mist-flow physics are analyzed. The three-dimensional (3D) droplet measurements show that the droplets injected from the fan-shaped holes tend to spread wider in lateral direction than cylinder holes and accumulate at the location where the neighboring coolant film layers meet. This flow and droplet behavior explain the higher cooling performance as well as mist-enhancement occurs between the fan-shaped cooling holes, rather than along the hole's centerline as demonstrated in the case using the cylindrical holes.