| dc.description.abstract | The uncertainty quantification in the turbine components' aerodynamic and heat transfer performances is widely considered to be the most challenging topic due to its intricate and nonlinear characteristics. This paper first proposes an efficient uncertainty quantification method based on an original parallel framework combining Polynomial Chaos Expansions (PCE) with two forms (stochastic response surface-based and Galerkin projection-based) and the Universal Kriging method. The rigorous mathematical tests are performed to verify the reliability and computational efficiency of the proposed method, and the results support that this method can dramatically reduce computational samples compared to the conventional PCE method while maintaining computational accuracy. Then, the genetic algorithm was introduced to establish an efficient uncertainty quantification framework, and it is applied to the aerothermal performance robustness investigation of the GE-E3 rotor blade tip with and without film cooling. Based on the findings of uncertainty quantification, the injection of cooling air drastically enhances the unstable tendency of the flow and thermal fields, resulting in the actual aerothermal performance of the squealer tip being much lower than that predicted by deterministic calculations. The setting of the film cooling, although effective in reducing the heat flux around the cooling holes, also induces more chaotic flow and thermal fields, leading to sharp heat flux fluctuations around the cooling holes. Finally, our novel reliability analysis algorithm, rooted in the quantification of uncertainty, corroborates the assertion that the introduction of coolant gas, while extending the operational longevity of turbine blades, confers only marginal improvements in the mitigation of lifespan variability. The comprehensive lifespan assessment elucidates that the mean operational longevity of the conventional squealer tip design stands at an estimated 16,169.44 h, accompanied by a standard deviation of 2,750.31 h. In stark contrast, the mean operational longevity of the squealer tip integrated with film cooling measures a significantly enhanced 17,035.17 h, exhibiting a standard deviation of 2,492.73 h. Consequently, the operational lifespan of the conventional squealer tip experiences a decrement of 10.17% in comparison to the anticipated mean lifespan, whereas the reduction for the film-cooled squealer tip registers at 5.36%. | |
