首页|Multi-objective optimization to minimize pumping power and flow non-uniformity at the outlets of a distributor manifold using CFD simulations and ANN rapid predictions

Multi-objective optimization to minimize pumping power and flow non-uniformity at the outlets of a distributor manifold using CFD simulations and ANN rapid predictions

扫码查看
原文链接
  • NSTL
  • Elsevier
Minimizing the pumping power and the flow non-uniformity at the outlets are two important goals in the design of liquid metal distributor manifolds. In the present study, a multi-objective optimization for a distributor manifold under the influence of an external magnetic field has been performed for the first time. In the first phase, the pumping power and the non-uniformity coefficient in terms of the different input variables are obtained using CFD simulations. An efficient artificial neural network (ANN) has been developed in the second phase to measure the expected outputs according to the design variables. Finally in the third phase, the proposed ANN and a multi-objective genetic algorithm are used for a Pareto-based optimization study. LINMAP and TOPSIS techniques identify the best points of Pareto front data. The former reduce pumping power by about 12% and the latter lessen the non-uniformity coefficient by approximately 25% relative to each other.

CFDDistributor manifoldANNMulti-objective optimizationDecision-making3D MAGNETOHYDRODYNAMIC FLOWSARTIFICIAL NEURAL-NETWORKSUNIFORM MAGNETIC-FIELDMHD MIXED CONVECTIONHEAT-TRANSFER2-IMMISCIBLE FLUIDSSQUARE CYLINDERMICRO-MIXERSCAPABILITIESNANOFLUID

Shahri, Mohammad Farahi、Nezhad, Alireza Hossein

展开 >

Univ Sistan & Baluchestan

2022

10.1016/j.measurement.2021.110566

Measurement

Measurement

SCI
ISSN：0263-2241
年,卷(期)：2022.188
  • 6
  • 95