Development of a new combustor liners thermal design procedure through low order codes and uncertainty quantification tools

Turbomachinery plays an important role in the propulsion and heavy-duty industry. Improving the efficiency and reliability of gas turbines continue to be an important driver in the development of modern engines and power generation. There are two straightforward and effective methods for improving the performance of a gas turbine engine. One is increasing the engine pressure ratio to raise the thermal efficiency, the other is increasing the outlet temperature of a combustor to raise the specific thrust. Therefore, the gas turbine combustors working conditions are moving towards higher temperature rise and higher heat capacity. Therefore, the design requirements for a combustor become stricter, such as a wider working range, shorter length, and smaller distribution of outlet temperature. Simultaneously, the combustor is required to have a longer life and lower pollutant emission. In this scenario, a fundamental role is played by the cooling system. The definition of the most appropriate scheme represents one of the most challenging tasks in the combustor since it directly determines the components life. During the design and subsequent optimization phases of a combustor cooling system, the designer must consider several uncertainties related to manufacturing, geometry and operating conditions. These gaps can be very impacting on the system performance, so it is obvious that the design becomes a matter of optimization of the whole system. This requires an accurate assessment of trade-offs to meet all requirements. The design choices made in the first phases influence the following developments and it is essential to have a tool as efficient and flexible as possible to rely on. During the initial stages, 1-D codes are still widely used in industrial practice, and a low-order approach is preferred over high-fidelity simulations. These tools are important for designers because they allow having a good understanding of the problem, in relatively short times and with low general costs. Although these analyses have a good predictive level, they are often used when input quantities that characterize the problem are roughly known. These gaps lead to the inclusion of uncertainties within the code, which propagate and eventually influence the solution. The final common objective is to optimize the various components to find out the configuration in which the machine is independent from the uncertainties that may afflict it, thus arriving at a robust design. The aim of this thesis is the development of a numerical procedure for the preliminary thermal design of combustor liners (Therm-1D/Dakota). This procedure is based on the coupling of a one-dimensional tool (Therm-1D), developed by DIEF of the University of Florence, and a software that allows uncertainty quantifications analyses (Dakota). This has allowed the development of an innovative, faster, and more reliable procedure for the preliminary design and optimization of combustor cooling systems that is able to estimate the uncertainties affecting the results of this numerical simulations. In this way, the output quantities are as independent as possible from input uncertainties.