MACHINE LEARNING FOR THE DEVELOPMENT OF DATA DRIVEN TURBULENCE CLOSURES IN COOLANT SYSTEMS

This work shows the application of Gene Expression Programming to augment RANS turbulence closure modelling, for flows through complex geometry designed for additive manufacturing. Specifically, for the design of optimised internal cooling channels in turbine blades. One of the challenges in internal cooling design is the heat transfer accuracy of the RANS formulation in comparison to higher fidelity methods, which are still not used in design on account of their computational cost. However, high fidelity data can be extremely valuable for improving current lower fidelity models and this work shows the application of data driven approaches to develop turbulence closures for an internally ribbed duct. Different approaches are compared and the results of the improved model are illustrated; first on the same geometry, and then for an unseen predictive case. The work shows the potential of using data driven models for accurate heat transfer predictions even in non-conventional configurations and indicates the ability of closures learnt from complex flow cases to adapt successfully to unseen test cases.

Bayesian Mass Averaging in Rigs and Engines

This paper introduces the Bayesian mass average and details its computation. Owing to the complexity of flow in an engine and the limited instrumentation and the precision of the sensor apparatus used, it is difficult to rigorously calculate mass averages. Building upon related work, this paper views any thermodynamic quantity's spatial variation at an axial plane in an engine (or a rig) as a Gaussian random field. In cases where the mass flow rate is constant in the circumferential direction but can be expressed via a polynomial or spline radially, this paper presents an analytical calculation of the Bayesian mass average. In cases where the mass flow rate itself can be expressed as a Gaussian random field, a sampling procedure is presented to calculate the Bayesian mass average. Examples of the calculation of the Bayesian mass average for temperature are presented, including with a real engine case study where velocity profiles are inferred from stagnation pressure measurements.

OVERVIEW OF THE BEST 2020 AXIAL-FLOW FAN DATA AND INCLUSION IN SIMILARITY CHARTS FOR THE SEARCH OF THE BEST DESIGN

The paper deals with the aerodynamic performance of ducted axial-flow fans available in the 2020 market and aims to create a general picture of the best designs and design trends, as a tool for fan designers. To this end, the paper first presents the general formulation of the similarity approach to the fan performance analysis, including the effects of rotational speed (which affects the validity of the Reynolds similarity) and turbomachine size (which can hinder the perfect geometrical similarity of some shape details). The second part reports a statistical survey of the axial-flow fan performance based on data from catalogues of major manufacturers, and compares the resulting Cordier-lines with optimum fan designs from empirical or CFD-based models available in the literature. In addition to the global performance at maximum aeraulic and total-to-static efficiencies, this survey uses the form of dimensionless Balje-Cordier charts to identify the trends and values of other design parameters, such as hub-to-tip ratio, blade count, and blade positioning angle. As a result, a summary of the aerodynamic performance of year 2020 best designs, the improvements achieved during the last forty years, and the present design trends in contra-rotating, vane-axial, and tube-axial fan types are made available to fan designers.

Effects of Rotation on the Flow Structure in a Compressor Cascade

In turbomachines, rotors and stators differ by the rotation of the former. Hence, half of each stage is directly influenced by rotation effects. The influence of rotation on the flow structure and its impact on the performance is studied through Wall-Resolving Large Eddy Simulations of a rotor with large relative tip gap size. The simulations are performed in a rotating frame with rotation accounted for through a Coriolis force term. In a first step experimental results are used to provide validation. The main part of the study is the comparison of the results from two simulations, one representing the rotating configuration, one with the Coriolis force removed, without any other change. This setup allows very clean assessment of the influence of rotation. The turbulence-resolving approach ensures that the turbulent flow features are well represented. The results show a significant impact of rotation on the secondary flow. In the tip region the Tip Leakage Vortex is enlarged and destabilised. Inside the tip gap the flow is altered as well, with uniformization in the rotating case. At the blade midspan, no significant effects are observed on the suction side, while an earlier transition to turbulence is found on the pressure side. Near the hub, rotation effects are shown to reduce the corner separation significantly.

IN-STALL COMPRESSOR PERFORMANCE AND THE EFFECTS OF REYNOLDS NUMBER

Previous research into axial compressor stall has mainly focused on stall inception and methods to extend the stable operating range. This paper considers the performance of an axial compressor beyond stall and investigates how the characteristics of stall cells depend on Reynolds number. An experimental study has been conducted using a single-stage axial compressor capable of operating across the Reynolds number range of 10,000 – 100,000. Detailed unsteady measurements have been used to measure the behaviour across a range of in-stall flow coefficients. These measurements have been used to extract the stall hysteresis and to determine the size, speed, number, and spanwise extent of the stall cells. The results show that for the stalled compressor, as Reynolds number increases, the size of the minimum stable stall cell decreases. This means that a larger change in throttle area is needed to reduce the stall cell down to a size where the compressor can recover from stall. At the design Reynolds number, the number of stall cells that form transitions from one, to two, and then to four stall cells as the flow coefficient is reduced. At lower Reynolds numbers, the two-stall-cell state becomes unstable; instead, a single stall cell transitions directly into five stall cells. As the number of stall cells increases, so do the speed of the stall cells and the total size. Further reductions in the flow coefficient cause an increase in the total size and a decrease in the stall cell speed.

CALIBRATION OF A CFD METHODOLOGY FOR THE SIMULATION OF ROUGHNESS EFFECTS ON FRICTION AND HEAT TRANSFER IN ADDITIVE MANUFACTURED COMPONENTS

It is well-known from the literature that surface roughness significantly affects friction and heat transfer. This is even more evident for additive manufactured (AM) components, which are taking an increasingly important role in the gas turbine field. However, the exploitation of numerical approaches to improve their design is hindered by the lack of dedicated correlations and CFD models developed for such high roughness conditions. Usually the additive manufactured components are simulated considering the surfaces as smooth or applying an equivalent sand-grain roughness (ks) that results in a velocity shift in the boundary layer. However, determining a priori the most appropriate value of ks is challenging, as dozens of correlations are available, returning scattered and uncertain results. A previous work proved how the CFD prediction of friction and heat transfer returns significant deviations, even exploiting the ks values obtained from experimental tests on the very same test case. That work also allowed identification of a promising CFD methodology based on friction and thermal corrections proposed by Aupoix from ONERA. The aim of this work is to further the assessment and calibration activity of the model, by analyzing additional experimental data of friction factor and Nusselt number from new test cases considering different geometries and flow conditions. The new coupons consisted of straight circular channels and wavy channels. This work represents a further step in the generation of a more validated and general methodology for the high-fidelity CFD analysis of additive-manufactured components.

Analysis of Swirl number effects on effusion flow behaviour using time resolved PIV

The analysis of the interaction between the swirling and liner film-cooling flows is a fundamental task for the design of turbine combustion chambers since it influences different aspects such as emissions and cooling capability. Particularly, high turbulence values, flow instabilities, and tangential velocity components induced by the swirlers deeply affect the behavior of effusion cooling jets, demanding for dedicated time-resolved near-wall analysis. The experimental setup of this work consists of a non-reactive single-sector linear combustor test rig scaled up with respect to engine dimensions; the test section was equipped with an effusion plate with standard inclined cylindrical holes to simulate the liner cooling system. The rig was instrumented with a 2D Time-Resolved Particle Image Velocimetry system, focused on different field of views. The degree of swirl is usually characterized by the swirl number, Sn, defined as the ratio of the tangential momentum to axial momentum flux. To assess the impact of such parameter on the near-wall effusion behavior, a set of three axial swirlers with swirl number equal to Sn = 0.6 − 0.8 − 1.0 were designed and tested in the experimental apparatus. An analysis of the main flow by varying the Sn was first performed in terms of average velocity, RMS, and Tu values, providing kinetic energy spectra and turbulence length scale information. Following, the analysis was focused on the near-wall regions: the effects of Sn on the coolant jets was quantified in terms of vorticity analysis and jet oscillation.

HEAT TRANSFER IN ROTATING, TRAILING EDGE, CONVERGING CHANNELS WITH FULL AND PARTIAL HEIGHT STRIP-FINS

A wide variety of pin-fins have been used to enhance heat transfer in internal cooling channels. However, due to their large blockage in the flow direction, they result in an undesirable high pressure drop. This experimental study aims to reduce pressure drop while increasing the heat transfer surface area by utilizing strip-fins in converging internal cooling channels. The channel is designed with a trapezoidal cross-section, converges in both transverse and longitudinal directions, and is also skewed β=120° with respect to the direction of rotation in order to model a trailing edge cooling channel. Only the leading and trailing surfaces of the channel are instrumented, and each surface is divided into eighteen isolated copper plates to measure the regionally averaged heat transfer coefficient. Utilizing pressure taps at the inlet and outlet of the channel, the pressure drop is obtained. Three staggered arrays of strip-fins are investigated: one full height configuration and two partial fin height arrangements (Sz=2mm and 1mm). In all cases, the strip fins are 2mm wide (W) and 10mm long (Lf ) in the flow direction. The fins are spaced such that Sy/Lf = 1 in the streamwise direction. However, due to the convergence, the spanwise spacing, Sx/W, was varied from 8 to 6.2 along the channel. The rotation number of the channel varied up to 0.21 by ranging the inlet Reynolds number from 10,000 to 40,000 and rotation speed from 0 to 300rpm. It is found that

Experimental and Numerical Study of the Turbulent Flow and Heat Transfer in a Wedge-shaped Channel with Guiding Pin Fin Arrays under Rotating Conditions

Experiments and numerical simulations under stationary and rotating conditions have been conducted to investigate turbulent flow and heat transfer characteristics of innovative guiding pin fin arrays in a wedge-shaped channel, which models the internal cooling passages for gas turbine blade trailing edge. The Reynolds number range is 10,000-80,000, and the inlet rotation number range is 0-0.46. With the increase of Reynolds numbers, the enhancement of heat transfer performance with guiding pin fin arrays is significantly higher than that with conventional circular pin fin arrays. At the highest Reynolds number of Re=80,000, the overall Nusselt number of the channel with guiding pin fin arrays is about 33.7% higher than that of the channel with circular pin fin arrays under the stationary condition, and is about 23.0% higher than the latter under the rotating conditions. At the highest inlet rotation number of Ro=0.46, the heat transfer difference between the trailing side and leading side of the channel is significantly lower with the guiding pin fin arrays. Both the experiments and numerical simulations indicate that the heat transfer uniformity and enhancement of the channel endwall is significantly improved by the guiding pin fin arrays under stationary and rotating conditions, which provide more reasonable flow distribution in the wedge-shaped channel, and can further produce obviously improved heat transfer in the tip region for the trailing edge internal cooling channel.

Experimental and Numerical Study of Heat Transfer and Pressure Loss in a Swirl Multi-pass Channel with Convergent Jet Slots

This paper presents a comparative experimental and numerical study of the heat transfer and pressure loss in a swirl multi-pass channel with tangential jet slots, and another baseline multi-pass channel with 180-deg U-bends as comparison baseline has also been investigated. Transient liquid crystal thermography is used to obtain the detailed heat transfer distribution on the internal surfaces of the multi-pass serpentine channels. The heat transfer patterns in the swirl multi-pass channel are quite different from that of the baseline multi-pass channel. Compared with the baseline multi-pass channel, the experimental globally averaged Nusselt number ratios of the last two passes in the swirl multi-pass channel can be increased by up to 82.9%, 104.8% and 124.6% for the Reynolds numbers 20,000, 40,000 and 60,000, respectively. The high and circumferentially uniform heat transfer is mainly due to the large-scale swirling flow induced by the tangential slots. More detailly, the large-scale swirling flow impinges onto the surface and further induces high tangential velocity near the wall, which destroys the boundary layer flow and thus improves the heat transfer rates at the wall. However, the notable pressure loss of the swirl multi-pass channel should be further controlled reasonably, which is about 5.4 times that of the baseline multi-pass channel. As supplements to the experiments, three-dimensional numerical computations provide more insights into the turbulent flow structure in the two kinds of multi-pass serpentine channels.