An effective methodology for optimal design of axial turbine blades is presented. It has been used for achieving stage maximal efficiency meeting both stress-strain and vibration reliability requirements and taking into account technological limitations.

INTRODUCTION

Problem formulations of the turbine flow path optimization reflect main phases of axial turbines design practice [1] and use increasingly accurate models of elaborated designs. As the industry moves forward, integration of 3D modeling of aerodynamic and strength of material characteristics into optimization process generates continuously increasing interest. Usually, such optimization refers to analysis of separate blade row or isolated stage [2 - 7, 9] and involves a large number of optimization parameters. This leads to enormous computational time and requires significant computational resources. Nevertheless, 3D modeling is an important part of numerical modeling along with conventional 1D and 2D approaches.

This paper describes a process of optimal flow path design that is achieved through the following steps:

• rapid flow path design and optimization using reduced order models and axi-symmetrical solver;
• blade cross-sections profiling according to aerodynamic criteria, blade stacking (3D profiling) with optimized twist/lean;
• generation of parameterized mesh for buckets and parameterized grid for inter-blade passages;
• detailed 3D CFD computations and finite element structural and modal analyses with commercial CFD and FEA tools;
• design optimization using design of experiment (DoE) methods and reduced order models.

Process begins from preliminary flow path design. Tools such as AxSTREAM™ allow significantly reduce the search range for bucket optimal configuration. AxSTREAM™ uses stage and airfoil optimizations that are based on DoE methods in combination with 2D aerodynamic and 1D structural calculations. Computed data can be exported to external tool for mesh and grid generation.