Aerodynamic Optimization of a Low Pressure Exhaust Hood of Steam Turbine Using Bayesian Optimization

The exhaust hood, an essential component of the turbine with large aerodynamic losses, significantly impacts the turbine performance. With the objective of improving the static pressure recovery coefficient of the exhaust hood, the critical components of the exhaust hood are parameterized and then coupled with the final stage for joint optimization. This simulation-based optimization is conducted through the efficient and effective Bayesian optimization using an adaptive confidence bound acquisition function. After optimization, the static pressure recovery coefficient of the exhaust hood increases up to 0.563, and the total pressure loss coefficient decreases to 0.197, resulting in a significantly improved aerodynamic performance. In addition, the flow fields of different sections and conditions of the exhaust hood before and after optimization are analyzed and compared to obtain the flow mechanism contributing to the performance improvement.