Latin hypercube sampling is widely used design-of-experiment technique to select design points for simulation which are then construct a surrogate model. The exploration/exploitation properties of models depend on the size and distribution in chosen space. present study aimed at evaluating performance characteristics various depending (LHS) procedure (sample spatial distribution) diverse set optimization problems. analysis was carried out two types problems: (1) thermal-fluid problems...

Optimization of a centrifugal pump with double volute was performed using surrogate modeling and multi-objective genetic algorithm to minimize the radial thrust at off-design conditions minimal loss in hydraulic efficiency. Both steady unsteady numerical analyses were conducted three-dimensional Reynolds-averaged Navier–Stokes equations shear stress transport turbulent model. The three objective functions are efficiency design flow rate, coefficient 70% 120% rate. In addition, geometric...

This paper presents a multi-objective optimization procedure for the design of sigma micromixer. The combines three-dimensional analyses fluid flow and mixing, polynomial approximation objective functions, genetic algorithm (MOGA). MATLAB Optimization Toolbox (version 7.7, Mathworks, Inc., MA, USA) was used to invoke MOGA optimization. A brief discussion on application is introduced in paper. Three geometric parameters concerning shape sidewalls were exploited Mixing index non-dimensional...

A multiobjective optimization of a micromixer with convergent–divergent sinusoidal walls has been conducted using flow and mixing analyses, surrogate modeling, genetic algorithm. The ratios amplitude to wavelength the walls, throat width depth sections, diameter inner circular wall were chosen as design variables for optimization. full-factorial method was used discretize space. index nondimensional pressure loss selected objective functions. Radial basis neural network functions train...

Low fidelity model assisted design optimization of turbomachines has reduced the total computational and experimental costs. These models are called surrogate which mimic actual experiments or simulations. The can generate thousands approximate results from a few samples, making it easy to locate optimal solution. Ample articles reported centrifugal pumps. In this article, authors try give brief overview based technique along with its historical applications trend recent use. various key...

Optimization of a three-dimensional split-and-recombine micromixer with serpentine structure was performed using Navier–Stokes analysis and optimization techniques. The study at fixed Reynolds number 15, four dimensionless design variables, viz. the ratio subchannel width to main channel width, length pitch length, depth depth, recombination width. space investigated by parametric study, points within were selected Latin hypercube sampling method. Two different objective functions, mixing...

A passive micromixer with wavy channel walls a sinusoidal variation is proposed, and numerical simulations based on Navier–Stokes equations are carried out for Reynolds numbers (Re) from 0.1≤Re≤30. Ethanol water used as the working fluids mixing. mixing index employed to evaluate performance of micromixer. different realization chaotic Dean vortices observed shift in center rotation toward inner wall lower-wavelength mixers. The found sensitively increase wavelength decreases over entire Re...

Optimization of rectangular cooling channels with intersecting ribs was performed to determine optimal rib configurations for thermal performance enhancement using Reynolds averaged Navier-stokes equations at number 10,000. The study aims achieve two-fold objectives: (1) maximize the in three different aspect ratios equal 1, 2, and 4, respectively a fixed hydraulic diameter (2) analyze trend optimum design variables change ratio. Angle attack, position relative channel center-line, height...

Performance analysis and optimization of a circular pin-fin with inside gaps in rectangular cooling channel were performed at Reynolds number, 10,000, using three-dimensional Reynolds-averaged Navier–Stokes equations multi-objective genetic algorithm. The low-Reynolds-number version the shear stress transport model was used as turbulence closure. A parametric study also to identify geometrical effects on heat transfer pressure drop. straight reference gapped pin-fins yielded better...