Application of Multipoint Optimization to the Design of Turbomachinery Blades

Demeulenaere, Alain; Ligout, Alban; Hirsch, Charles

doi:10.1115/gt2004-53110

Cited by 36 publications

(18 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The algorithm of designing turbine flow parts can be automated. To his, methods for solving optimization tasks should be used to automate the process of ysing the results and generate new values of variable parameters (geometric acteristics of the flow part) [20][21][22][23][24]. An automated approach usually requires several dred to several thousand iterations [18,25].…”

Section: Introductionmentioning

confidence: 99%

Designing and updating the flow part of axial and radial-axial turbines through mathematical modeling

Rusanov

Lampart

2015

Open Engineering

View full text Add to dashboard Cite

The paper describes an algorithm for the design of axial and radial-axial type turbines. The algorithm is based on using mathematical models of various levels of complexity -from 1D to 3D. Flow path geometry is described by means of analytical methods of profiling using a limited number of parameters. 3D turbulent flow model is realised in the program complex IPMFlow, developed based on the earlier codes FlowER and FlowER-U. Examples of developed or modernized turbines for differentpurpose power machines are presented. They are: an expansion turbine, ORC turbine and cogeneration mediumpressure turbine.

show abstract

Section: Introductionmentioning

confidence: 99%

Designing and updating the flow part of axial and radial-axial turbines through mathematical modeling

Rusanov

Lampart

2015

Open Engineering

View full text Add to dashboard Cite

show abstract

“…The software Numeca/Design3D, in which genetic algorithm and artificial neural network are jointly adopted, is used for multistage local optimization. The major principle is that the successive design evaluation is performed by using an artificial neural network instead of a flow solver, and the genetic algorithms may be used in an efficient way [4].…”

Section: Introductionmentioning

confidence: 99%

Aerodynamic design by jointly applying S2 flow surface calculation and modern optimization methods on multistage axial turbine

Wang

Zhao

Wan-jin

et al. 2008

Front. Energy Power Eng. China

View full text Add to dashboard Cite

A three-stage axial turbine was redesigned by jointly applying S2 flow surface direct problem calculation methods and multistage local optimization methods. A genetic algorithm and artificial neural network were jointly adopted during optimization. A three-dimensional viscosity Navier-Stokes equation solver was applied for flow computation. H-O-H-topology grid was adopted as computation grid, that is, an H-topology grid was adopted for inlet and outlet segment, whereas an O-topology grid was adopted for stator zone and rotor zone. Through the optimization design, the total efficiency increases 1.1%, thus indicating that the total performance is improved and the design objective is achieved.

show abstract

“…Owing to their simplicity and flexibility, ANNs represent a class of RMSs that have been receiving increasing attention in the context of blade optimization [7]. Gradients have been used by ANNs as a regularization mechanism [8], which ensures a degree of smoothness for the output.…”

Section: Introductionmentioning

confidence: 99%

Multi‐objective turbomachinery optimization using a gradient‐enhanced multi‐layer perceptron

Duta¹,

Duta²

2008

Numerical Methods in Fluids

View full text Add to dashboard Cite

SUMMARYResponse surface models (RSMs) have found widespread use to reduce the overall computational cost of turbomachinery blading design optimization. Recent developments have seen the successful use of gradient information alongside sampled response values in building accurate response surfaces. This paper describes the use of gradients to enhance the performance of the RSM provided by a multi-layer perceptron. Gradient information is included in the perceptron by modifying the error function such that the perceptron is trained to fit the gradients as well as the response values. As a consequence, the back-propagation scheme that assists the training is also changed. The paper formulates the gradient-enhanced multi-layer perceptron using algebraic notation, with an emphasis on the ease of use and efficiency of computer code implementation. To illustrate the benefit of using gradient information, the enhanced neural network model is used in a multi-objective transonic fan blade optimization exercise of engineering relevance.

show abstract

Application of Multipoint Optimization to the Design of Turbomachinery Blades

Cited by 36 publications

References 0 publications

Designing and updating the flow part of axial and radial-axial turbines through mathematical modeling

Designing and updating the flow part of axial and radial-axial turbines through mathematical modeling

Aerodynamic design by jointly applying S2 flow surface calculation and modern optimization methods on multistage axial turbine

Multi‐objective turbomachinery optimization using a gradient‐enhanced multi‐layer perceptron

Contact Info

Product

Resources

About