This work sets out the creation, verification and demonstration of an automated end-to-end tool for the multi-point fluid-dynamic optimization of turbomachinery. Taking into account multiple operating conditions during optimization aims to benefit the overall performance of turbomachines which are characterized by off-design operation and to produce more robust designs with respect to deviations in operating conditions. The feasibility and effectiveness of the tool is demonstrated by a two-point fluid-dynamic optimization of a two-dimensional Aachen turbine stator cascade, performed on the High-Performance Computing cluster of the TU Delft. By utilizing the stochastic NSGA2 optimization algorithm, the numerically noisy problem is successfully optimized. The entropy generation of the blade is reduced at both nominal and off-design operating condition by an average of 7.57%, while satisfying a constraint on the flow turning when passing through the stator cascade. Additionally, a numerical noise analysis shows that disregarding the most noisy design variables increases the probability of obtaining an improved design.

Turbomachines are prevalent within the aerospace industry and vital for both power production and flight systems. To achieve better performance and higher efficiency, a numerical optimization technique is used to optimize designs. Previous work has primarily focused on performing steady-state optimization of turboma- chinery using Reynolds Averaged Navier Stokes simulations. In efforts for further performance increases, the next step of turbomachinery optimization is to incorporate time-accurate solvers. These time-accurate solvers are computationally more expensive than steady solvers, which hinders usage in optimization. The harmonic balance method was developed to reduce computational requirements compared to a time-accurate solver while maintaining unsteady flow characteristics in the simulation results. Optimization using the harmonic balance method has shown better performance in 2D cases, but there has not been work reporting on a 3D case. Fur- ther analysis of how optimization changes flow irreversibility in a design has not been shown between harmonic balance and mixing plane based optimization. This work’s novel contribution is to perform a comparative study of harmonic balance and mixing plane based optimization of a 3D multistage turbine and perform a loss quantification comparison of the optimized designs. In this work, the flow around a 3D 1.5 stage axial turbine in the subsonic to transonic flow regime was simulated and optimized using the mixing plane and harmonic balance methods. The turbine blade shapes are parameterized using a CAD-based NURBS parameterization and optimized using a gradient-based adjoint algorithm. Flow features, entropy trends, and computational costs between harmonic balance and mixing plane simulation and optimization were compared. This study shows that most entropy creation differences between harmonic balance and mixing plane methods occur in the rotating domain. A turbine design optimized using the harmonic balance method has better fluid-dynamic performance than a design optimized using the mixing plane method. Turbine designs combining blades from both harmonic balance and mixing plane based optimization show better performance than a design using a single flow simulation method for optimization. Results also indicate that a grid converged mesh may not be necessary for entropy generation minimization of turbomachinery designs for transonic flow.

The scientific community is consistently focused on identifying new sources of energy, which can reduce the consequences of climate change and depleting natural resources. Organic Rankine Cycle (ORC) based power systems have been touted as one of the promising technologies of extracting thermal energy from waste-heat and renewable sources such as geothermal, solar radiation and biomass, to name a few. Use of an organic fluid makes it theoretically possible to design a compact, high volumetric flow turbine that can extract energy from sources with low temperature head. However, realization of this turbine becomes challenging because of their unsteady nature which stems from high expansion ratio across the cascade and low speed of sound in organic fluids. The expansion of fluid in these turbines occur in the dense gas supersonic regime where ideal gas assumptions do not hold and the flow is characterized by shock waves and expansion fans. The success of ORC systems thus depends on how well these effects are modeled into the design stage. Stator vanes have a major contribution to the stage losses in these high-expansion low reaction turbines. In current literature, there exists no co-relation that takes into account dense gas effects for the preliminary design of these stator vanes. Thus the objective of this thesis is to study the trend of stator losses with the variation of geometric de- sign variables and inlet conditions, subsequently to qualitatively and quantitatively assess the performance of the blade and finally to propose new design correlations for the organic fluid Toluene. To achieve this objec- tive, the stator design variables affecting 2D loss mechanisms are identified and a design space is constrained. It was hypothesized that the total losses are minimum for an optimum vane design corresponding to a unique value of post-expansion ratio. A semi-automated analysis framework was made, that assessed the fluid dy- namic performance of vanes with varying post-expansion ratio, which upon analysis backed the hypothesis successfully. Further, the trend behavior of this optimum post-expansion ratio with design variables such as stator blade angle, total-static expansion ratio, nozzle solidity and fluid non-ideality is investigated and val- ues of post-expansion ratio corresponding to optimum design vanes are obtained. These discrete optimum points are interpolated to obtain a continuous second order function which is the proposed co-relation that can be used for the preliminary design of a stator vane in an ORC turbine.

In order to reduce the carbon foot-print of turbo-generators to global warming, a step-change in the design process is needed. Extensive CFD simulations coupled with optimization algorithms are required, resorting to novel techniques to capture the complex flow phenomena occurring in the passage. However, the computation of the optimization gradients and the imposition of geometrical constrains is computationally expensive and non-trivial due to the large number of design variables. In the 1980s, the adjoint method arose as an alternative technique to compute the sensitivities at a cost independent of the number of design variables. This technique is extremely popular in external aerodynamic applications. However, its full potential is yet to be exploited to internal flows. The use of an adjoint solver, together with a surface parametrization technique based on traditional blade design parameters, presents a unique opportunity towards a fully-automated design methodology based on high-fidelity models for turbomachinery applications. Stemming from the above considerations, the aim of this thesis is to develop a common open-source numerical framework for the optimization of both axial and radial turbomachinery. In order to achieve so, the open-source CFD suite SU2, that includes an adjoint solver, is coupled with ParaBlade, an open-source blade parametrizer based on traditional turbomachinery design variables. The proposed methodology is successfully applied in an axial turbine stator, where a reduction of 7.59% in the entropy generation across the passage is achieved. The study performed in a mixed-flow turbine rotor highlights the need to parametrize the hub and shroud surfaces. However, its optimization revealed a reduction in the objective function value, showing the robustness of the proposed methodology in two types of turbomachinery.