A non-intrusive polynomial chaos (NIPC) model was proposed to efficiently and accurately quantify turbine blade performance change caused by flow variations

and verified by function model. It was used to quantify the change of energy loss coefficient for transonic turbine cascade when the inlet incidence angle varied at different scales. Meanwhile

it was compared with the direct Monte Carlo simulation (MCS) and sensitivity-based evaluation method. The flow solutions were statistically analyzed to reveal the impact mechanism of inlet incidence angle variation. The results showed that the adaptive NIPC quantified the uncertainty of energy loss coefficient with high performance at different scales perturbation of inlet incidence angle. Besides

the energy loss coefficient variation was nonlinearly dependent on the inlet incidence perturbation. Moreover

the shock wave in blade passage was most sensitive to the inlet incidence angle variation

becoming the most contribution to energy loss coefficient uncertainty for the transonic turbine cascade.