Gianfranco Guidati

Self-organizing Maps for Pareto Optimization of Airfoils

This work introduces a new recombination and a new mutation operator for an accelerated evolutionary algorithm in the context of Pareto optimization. Both operators are based on a self-organizing map, which is actively learning from the evolution in order to adapt the mutation step size and improve convergence speed. Standard selection operators can be used in conjunction with these operators.The new operators are applied to the Pareto optimization of an airfoil for minimizing the aerodynamic profile losses at the design operating point and maximizing the operating range. The profile performance is analyzed with a quasi 3D computational fluid dynamics (Q3D CFD) solver for the design condition and two off-design conditions (one positive and one negative incidence).The new concept is to define a free scaling factor, which is multiplied to the off-design incidences. The scaling factor is considered as an additional design variable and at the same time as objective function for indexing the operating range, which has to be maximized. We show that 2 off-design incidences are sufficient for the Pareto optimization and that the computation of a complete loss polar is not necessary. In addition, this approach answers the question of how to set the incidence values by defining them as design variables of the optimization.

Active Combustion Control System for Reduction of NOx and Pulsation Levels in Gas Turbines

This paper presents two different active combustion control systems (ACCS) for the reduction of NOx levels, and suppression of thermo-acoustic instabilities within stationary gas turbines. Dependent on the actual measurement the ACCSs steer the fuel split between different burner groups or within the burner itself in order to find an optimum operation point. In a first step, an active control system has been developed for the ALSTOM GT13E2 gas turbine where the fuel ratio between two burner groups has to be handled to optimise NOx and pulsation levels. Since perturbations in combustion operating conditions have a direct effect on combustion pulsation behaviour, it is possible to control combustor equivalence ratio by using combustion pulsation measurements as an input for the control system. By doing that, lower operating NOx emissions are achieved as operating safety margin to lean blow out (LBO) may be reduced by more than 50% due to a more accurate and controlled handling of combustor equivalence ratio. Although combustion instabilities due to lean blow out are of minor concern within the GT26, the success of this combustion control approach has led to the development of a more advanced method where both, NOx levels and pulsation amplitudes are feedback-controlled simultaneously in order to track the optimum operating point. By using two premixed stages in the burners fuel supply, the equivalence ratio within the combustor is adapted. Engine and single burner test results confirmed control model dynamics predictions. This paper illustrates the applied closed-loop controls concepts and the successful controller verification on single burner and on engine level.Copyright

Detailed Compressor Degradation Effect Modeling for Single Blade Rows While Assessing Its Local and Overall Consequences

The detailed local and overall aerodynamic impact due to the basic degradation mechanisms of changed surface roughness (fouling), caused by deposits on the blading surface and/or corrosion, leading edge deformation, caused by erosion, and clearance increases, caused by rubbing due to structural ovalization and/or rotor sagging are assessed. The generic aerodynamic impact of these effects is assessed in a first step by detailed local issue focused CFD analyses. The resulting trend changes in the aerodynamic behavior for every individual blade row are in a second step transferred to an empirical model for loss and deviation prediction of a 1-dim. compressor mid-streamline code. This tool is used to predict row specific and compressor overall aerodynamics for a wide range of operating points. Applying the generic trend changes due to degradation therefore provides information on the local and overall compressor mismatching. Details on the process of assessing and transferring these physical consequences are shown and discussed.The different degradation contributors are assessed individually and in combination. Beneath detailed aerodynamic changes, e.g. stage pressure ratios and feed pressures, resulting effects on the compressor map, reflecting mass flow, efficiency and operating range changes are considered. Some typical overall engine degradation measurements are used to validate the numerical assessments and to draw some final conclusions.For end-users the conclusions of this paper can be used to optimize site-specific operation cycles, based on corresponding field findings and by application of the theoretical conclusions derived in this paper. This can finally trigger a re-assessment of washing intervals and washing types to be executed and/or even to give special focus on certain blade rows during major overhauls/inspections in respect to hand cleaning or even blading exchange.Copyright

Compressor Design From Specification to Validation: Application of a Fast and Reliable Process

To support customer requirements for existing power plants inside the Service business Alstom did develop over the years a fast and reliable design process for compressors. This has meanwhile been proven with several engines. Besides high-end technologies, main focus is given on the interdisciplinary alignment. The application for an upgraded product for the mature fleet will be briefly shown. The blading re-design is based on Controlled Diffusion Airfoil (CDA) technology, individually optimized for every single blade row. Together with mechanically optimized thin blades this results in high compressor efficiency and surge margin. The CDA blading technology has been introduced in Alstom during 1990s. Therefore extended experience on this technology is available and has been manifested in design rules & criteria. The interdisciplinary blading concept and optimization is done with in-house 2D & 3D CFD codes for aerodynamics and FEM codes for mechanical integrity. To speed-up the design process, task-optimized tools are used for specific topics. All tools are linked into a design system and have an interface to a CAM system. Aerodynamic and FEM codes have been tuned/well-calibrated according to the past experience. A review process with experts is applied at certain milestones for technical and commercial issues. During blading design usually several disciplines are involved. This causes a challenge to organize the corresponding resources. Alstom uses a different approach: tasks are partly executed independent from the disciplines with standard tools. Nevertheless the final release stays inside the ownership of the expert discipline. This means e.g. that simplified tools are developed for pre-checks to be done by any discipline and the final check with high-end tools is done by the experts. This ensures a highly flexible and fast interdisciplinary approach. To validate the design and operation range of the upgraded compressors special attention is paid on testing. This reveals usually in a pre-test for the baseline engine and a post-test for the upgraded engine. For validation purposes a full compressor mapping under load and idle is done for both test series. Additional measurements for pressure rise inside the engine and validation of natural frequencies and stresses to certain blades is applied. The evaluation of the measurement data is used to confirm the improved compressor design and to check achievement of guaranteed performance figures. In addition the surge margin is measured to confirm that the upgraded compressor can be safely operated under all condition.Copyright

Automated Blade Optimization and 3D CFD Analysis for an Axial Multistage GT Compressor Redesign

This paper focuses on (1) the basic compressor layout based on meridional through flow analysis and (2) the re-design of blades and vanes using sophisticated automated design optimization methods. All tools and processes are integrated into a consistent Compressor Design System, which runs on a powerful Linux cluster. This design system allows designing, analyzing and documenting blade design in mostly automated way. This frees the engineer from repetitive tasks and allows him to concentrate on a physical understanding and improvement of the compressor. The tools and methods are illustrated on the basis of an actual ALSTOM compressor. The main objectives of this upgrade are a modest increase in mass flow and an efficiency improvement. The latter is to be achieved through the replacement of NACA blades by modern Controlled Diffusion Airfoils (CDA). Results are presented including a CFD analysis of the front stages of the baseline and upgrade compressor.Copyright