Welcome to the von Karman Institute for Fluid Dynamics Store

Multidisciplinary Turbomachinery Component Optimization Considering Performance, Stress, and Internal Heat Transfer

Be the first to review this product

Availability: In stock

€25.00
OR

Quick Overview

VKI PHDT 2008-06, Multidisciplinary Turbomachinery Component Optimization Considering Performance, Stress, and Internal Heat Transfer, T. Verstraete, 2008, ISBN 978-2-930389-33-8

Multidisciplinary turbomachinery component optimization consider

Double click on above image to view full picture

Zoom Out
Zoom In

More Views

  • Multidisciplinary turbomachinery component optimization consider
  • Multidisciplinary turbomachinery component optimization consider

Details

Multidisciplinary Turbomachinery Component Optimization Considering Performance, Stress, and Internal Heat Transfer
by Tom Verstraete
Published in 2008
ISBN 978-2-930389-33-8

PhD Thesis from von Karman Institute / Universiteit Gent, Belgium, June 3, 2008

Abstract

This work deals with the design of turbomachinery components and consists of two parts. The calculation of the heat transfer between blade and flow is discussed in the first part. These calculations are gaining importance as the turbine inlet temperature is increasing and the blades need to be cooled more effectively to guarantee their structural integrity.

The second part is devoted to the multidisciplinary optimization of different designs. An automated method for the design of turbomachine components is proposed, by which the flow around the blades, the mechanical stresses in the blades and the heat transfer between both is taken into consideration. It is demonstrated that this method has a high potential for industrial applications.

Part I The first part deals with the numerical modeling of the heat transfer between a fluid and a solid. This so-called conjugate heat transfer is of increasing interest for the design of turbine blades. Solving the flow without heat transfer to the blade is no longer acceptable in modern cooled turbine blades. The whole system of solid conduction in the blade and the external flow around the blade needs to be solved together for those applications. This can be done in two different ways. A first method uses the same discretization technique in both domains and solves the complete system at once. The second method solves both domains separately and tries to obtain the continuity of heat flux and temperature at the interface in an iterative manner.

Present work focuses on the latter method. The three different iterative techniques known from literature were thoroughly analyzed. A fourth new technique was proposed. A convergence criterion for the four methods was analytically determined and validated by different test cases.

The four iterative techniques were validated by computing the heat transfer for the flow over a flat plate. The results of the numerical computations and analytical models were compared for different boundary conditions. Special attention was given to the stability and the accuracy of the different methods and confirmed the earlier mentioned simple analytical convergence criteria.

The heat transfer in a micro gas turbine was calculated by means of the newly developed iterative method. The small distance between the hot turbine and the cold compressor results in a large heat transfer. The negative impact on the performance of both components was evaluated.

The same iterative method was used for calculating the heat transfer in a cooled axial turbine blade. Three different models were coupled. The flow in the cooling channels was modeled by a 1D flow prediction based on experimental correlations, the flow around the blade was calculated by a Navier-Stokes solver and the heat transfer inside the blade was computed by a finite element method. The thermal stresses inside the material were computed upon the determination of the temperature distribution. A simple model for calculating the lifetime of the blade was proposed. It is based on a damage model that takes into account the impact of temperature and stresses on the material.

Part II The second part of this work deals with the use of optimization methods for the design of turbomachine components. The optimization method developed at the von Karman Institute and in use since a decade has been used as the start point. One of the most critical components of the optimization algorithm is the metamodel. The Radial Basis Function (RBF) network has been implemented as an alternative to the existing Artificial Neural Network (ANN). Both metamodels were tested on their capabilities to minimize analytical test functions with the optimization algorithm. The influence of architecture and training model of both metamodels were compared.

A pure aerodynamic optimization is useless when the resulting optimal geometry fails due to the centrifugal forces. The existing aerodynamic optimization of radial compressors and turbines was extended to a multidisciplinary optimization system by integrating the mechanical and thermal aspects. The stress computation in the material was added to the aerodynamic optimization in the first step. Different radial compressors and turbines were optimized by using this multidisciplinary method. Results showed that respecting the mechanical constraints comes at the cost of a small drop in aerodynamic performance.

Finally, the conjugate heat transfer method from part I was used for the multidisciplinary optimization of the positioning of the cooling channels in an axial turbine blade. The objective was to maximize the life time of this blade and to minimize the coolant mass flow. Results of the two different metamodels (ANN and RBF) were compared. A considerable increase in lifetime was obtained with similar coolant mass flow.

The results of the multidisciplinary optimization show that not only the design cycle is shortened, but also new, surprising decisions are made by the optimizer that provide better insights to the designer. Moreover, the method is fully automated, reducing the interventions of the designer. Less than 100 designs need to be evaluated in detail, which shows that the method is cost-effective and of interest to industry.

 

 

 

 

Additional Information

Manufacturer von Karman Institute for Fluid Dynamics

Product Tags

Use spaces to separate tags. Use single quotes (') for phrases.