TY - JOUR
T1 - Radial turbine optimization under unsteady flow using nature-inspired algorithms
AU - Mehrnia, Seyedmajid
AU - Miyagawa, Kazuyoshi
AU - Kusaka, Jin
AU - Nakamura, Yohei
N1 - Funding Information:
The authors gratefully acknowledge Mitsubishi Co. for supporting a one-year research program at Waseda University in Japan.
Publisher Copyright:
© 2020 Elsevier Masson SAS
PY - 2020/8
Y1 - 2020/8
N2 - This paper investigated the performance of a radial flow turbine by coupling metaheuristic algorithms with Computational Fluid Dynamics. We performed the optimization of the casing and wheel of the turbine simultaneously. The computer codes of four metaheuristic algorithms, namely the Genetic Algorithm, Flower Pollination Algorithm, Grey Wolf Optimizer, and the Grasshopper Optimization Algorithm were developed using MATLAB. The optimization results indicated that Grey Wolf Optimizer is the most powerful algorithm to achieve a higher temperature drop in comparison with other algorithms. Revealed by the study, choosing the best angle at the blade inlet is the most influential factor for efficiency improvement. Besides, casing optimization has a positive effect on the pressure recovery of the turbine by eliminating swirling flow. A comparison of the physics of fluid in the optimized and base wheel showed that the flow is more attached to the optimized blade since the backward-facing step produces a favorable pressure gradient mainly in the recirculating bubble. Employing pulsating flow confirmed that the efficiency of the optimized turbine has a significant increase in comparison to the base turbine by more than 2% in high mass flow rates.
AB - This paper investigated the performance of a radial flow turbine by coupling metaheuristic algorithms with Computational Fluid Dynamics. We performed the optimization of the casing and wheel of the turbine simultaneously. The computer codes of four metaheuristic algorithms, namely the Genetic Algorithm, Flower Pollination Algorithm, Grey Wolf Optimizer, and the Grasshopper Optimization Algorithm were developed using MATLAB. The optimization results indicated that Grey Wolf Optimizer is the most powerful algorithm to achieve a higher temperature drop in comparison with other algorithms. Revealed by the study, choosing the best angle at the blade inlet is the most influential factor for efficiency improvement. Besides, casing optimization has a positive effect on the pressure recovery of the turbine by eliminating swirling flow. A comparison of the physics of fluid in the optimized and base wheel showed that the flow is more attached to the optimized blade since the backward-facing step produces a favorable pressure gradient mainly in the recirculating bubble. Employing pulsating flow confirmed that the efficiency of the optimized turbine has a significant increase in comparison to the base turbine by more than 2% in high mass flow rates.
KW - Computational fluid dynamics
KW - Metaheuristic algorithms
KW - Sensitivity analysis
KW - Turbine efficiency
KW - Turbocharger
KW - Turbomachinery aerodynamics
UR - http://www.scopus.com/inward/record.url?scp=85086407644&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85086407644&partnerID=8YFLogxK
U2 - 10.1016/j.ast.2020.105903
DO - 10.1016/j.ast.2020.105903
M3 - Article
AN - SCOPUS:85086407644
SN - 1270-9638
VL - 103
JO - Aerospace Science and Technology
JF - Aerospace Science and Technology
M1 - 105903
ER -