Mechanical Engineering - Theses
Permanent URI for this collection
Search Results
1 - 2 of 2
-
ItemTurbulence Model Development and Implementation for Low Pressure Turbines using a Machine Learning Approach( 2019)The design of the gas turbine, which is the work horse of the aviation industry, has reached a high degree of maturity; given that the first gas turbine flew in the late 1930s. Despite this, the industrial sector is looking towards harnessing even incremental points of efficiency with novel methods, which can translate to millions of dollars of savings and large reductions in carbon emissions. Current gas turbine design is primarily carried out using low-fidelity simulations due to their low cost and user-friendliness. However, these simulations lack the accuracy of high-fidelity simulations, largely due to the use of a linear stress-strain relation – the Boussinesq approximation. With the increase in the power of computing, high-fidelity simulations are becoming increasingly commonplace but are still not feasible as an iterative industrial design tool. In order to bridge the gap between high and low-fidelity simulations, certain high-fidelity data sets can be harvested to extract meaningful physics-based insights with machine learning processes to improve the accuracy of iterative low-fidelity calculations. This thesis focuses on improving low-fidelity modelling strategies (Reynolds–Averaged Navier–Stokes (RANS)) for low pressure turbine (LPT) flows, by harnessing meaningful physics-based information from high-fidelity data using a machine learning approach – gene expression programming (GEP). Improvement in the accuracy of the existing linear stress-strain closure relations is sought by developing machine-learnt explicit algebraic Reynolds stress models (EARSM). Of the many physical phenomena that occur in an LPT, designers are very interested in being able to accurately model the wake mixing using RANS, as this phenomenon governs the stagnation pressure loss in a turbine and also because existing RANS-based turbulence models fail to accurately predict this phenomenon. Therefore, the goal of this thesis is to develop and implement non-linear EARSMs to enhance the wake mixing in LPTs using GEP and high-fidelity data sets at realistic engine operating conditions. Firstly, an extensive analysis of the existing RANS-based turbulence models for LPTs with steady inflow conditions was conducted. None of these RANS models were able to accurately reproduce wake loss profiles based on high-fidelity data. However, the recently proposed k-v2-omega transition model was found to produce the best agreement with high-fidelity data in terms of blade loading and boundary layer behaviour and was thus selected as the baseline model for turbulence closure development. Using different training regions for model development, the resulting closures were extensively analysed in an a priori sense (without running any CFD) and also while running CFD calculations. Importantly, to assess their robustness, the trained models were tested both on the cases they were trained for and on testing, i.e. previously not seen, cases with different flow features. The developed models improved prediction of the Reynolds stress, TKE production, wake loss profiles and wake maturity across all cases. The existing GEP framework was extended to include RANS feedback during the model development process. It was found that the models generated via this method allow greater flexibility to the user in terms of selecting metrics of direct interest. The models returned offer a higher degree of numerical stability and robustness across different flow conditions and even geometries. Models developed on the LPT were tested on a high pressure turbine case and vice-versa and some of the models were able to reduce the peak wake loss error by up to 90% over the Boussinesq approximation in this cross-validation study. A zonal based model development approach was proposed with an aim to enhance the wake mixing prediction of unsteady RANS calculations for LPTs with unsteady inflow conditions. High-fidelity time-averaged and phase-lock averaged data at a realistic isentropic Reynolds number and two reduced frequencies, i.e. with discrete incoming wakes and with wake ‘fogging’, were used as reference data. This is the first known study to develop machine learning based turbulence models for unsteady flows, and also the first study to use phase-lock averaged data for the same. Models developed via phase-lock averaged data were able to capture the effect of certain prominent physical phenomena in LPTs such as wake-wake interactions, whereas models based on the time-averaged data could not. Correlations with the flow physics led to a set of models that can effectively enhance the wake mixing prediction across the entire LPT domain for both cases. Based on a newly developed error metric, the developed models have reduced the a priori error over the Boussinesq approximation on average by 45%. Based on the analysis conducted in this work, a few best practice guidelines have been proposed which can offer future designers an insight into the GEP-based model development process. Overall, this study showcases that GEP is a promising avenue for future RANS-based turbulence model development.
-
ItemNumerical methods and turbulence modelling for large-eddy simulations( 2015)Turbulence is of interest in many engineering applications, ranging from aerospace design to naval architecture. The inherent complexities of turbulence make it difficult to measure experimentally, and, to simulate numerically. The focus of this dissertation is the simulation of turbulent flow with the computational methodology known as large-eddy simulation (LES). LES uses a filter to partition a flow-field into large- (or resolved) and small- (or subgrid) scales and solves only for the large-scales. This method provides more accuracy when compared to other computational methods, such as those based on the Reynolds-averaged Navier--Stokes equations. The increased accuracy, however, comes with an associated increase in computational cost. Indeed, the computational cost of LES can often be prohibitive, especially for cases involving wall-bounded flow over complex geometries at high Reynolds numbers. This high computational expense is one of the primary limitations of LES. Methods to reduce the cost of LES form the focus of this dissertation. The high cost of LES is in great part due to the near-wall resolution requirement. To accurately represent a flow-field with LES, it is necessary to sufficiently resolve all of the dynamically important scales of motion. This is relatively inexpensive in free-shear flows, where the large-scales are the most energetic, but it is more difficult in wall-bounded flows, where the energy-containing scales get increasingly small near a wall. These near-wall small-scales make it impractical to resolve all of the energy containing scales. Therefore, models that mimic the effect of the near-wall turbulent structures on the wall and on the core of the flow are often used. These models are known as wall-models, and, if accurate, they are able to significantly reduce the computational cost of a large-eddy simulation. At present there is no wall-modelling approach that has been shown to be apposite in a broad range of applications. In particular, current wall-models are often inaccurate when applied to separating wall-bounded flow and are limited by their inability to predict fluctuations of wall-shear-stress and the near-wall velocity. Because of this, a key focus of this dissertation is the proposal and investigation of a new wall-model that aims to overcome these two limitations. In addition, the new model aims to reduce the computational cost of LES by significantly reducing the near-wall resolution requirement. Before introducing this new wall-model, flow over a circular cylinder is investigated in order to gain familiarity with the large-eddy simulation methodology and to assess the effect of some key computational parameters in LES. In this investigation, the effects of mesh resolution, discretisation schemes, SGS-models, and wall-models on prediction of the flow-field are assessed. One of the primary outcomes of this study is the finding that `standard' wall-models are inadequate for turbulent separating flows. This motivated the investigation of the new wall-model. The new wall-model is able to predict the fluctuating wall-shear-stress from a large-scale velocity input. The model is based on the spectral structure of the turbulent boundary layer and the interaction between large-scale events in the logarithmic layer and small-scale events near the wall. Importantly, the model includes many important parameters that are able to preserve the structure of the boundary layer while remaining relatively straightforward to implement in a solver. Further, the model does not increase the computational cost of a simulation compared to current wall-modelling approaches. The model is implemented in large-eddy simulations of channel flow to assess its efficacy compared to a standard wall-model. The influence of two subgrid-scale models on the large-scale velocity input is also investigated. Results show that the new wall-model is able to resolve more of the wall-shear-stress variance when compared to a standard wall-model, and it has a small effect in the outer-regions of the boundary layer.