engineering design centre blade design for axial compressors
TRANSCRIPT
Engineering
Design
Centre
Timos Kipouros – [email protected]
Department of Engineering, University of Cambridge
Propulsion Engineering Centre, Cranfield University
Using Parallel Coordinates to Guide
Optimisation Processes
Blade Design for Axial Compressors
Objectives:
• Minimise blockage, entropy generation rate, profile and endwall losses
Constraints
• Mass flow (equality), mass-averaged flow turning, leading edge radius and tip clearance (inequality)
Definition of the Design Space:
26 design parameters of Partial Differential Equations parameterisation – combination of these associated with actual 3D
geometrical characteristics
Introduction
Modern Engineering Design involves the deployment of many computational tools. Research on challenging real-world design
problems is focused on developing improvements for the engineering design process through the integration and application
of advanced computational search/optimisation and analysis tools. Successful application of these methods generates vast
quantities of data on potential optimum designs. To gain maximum value from the optimisation process, designers need to
visualise and interpret this information leading to better understanding of the complex relations between parameters, objectives
and decision-making. This work has identified that the Parallel Coordinates visualisation method has considerable potential in
this regard. This methodology involves significant levels of user-interaction, making the engineering designer central to the
process, rather than the passive recipient of a deluge of pre-formatted information, and building on the human-in-the-loop
design approaches.
The methodology is applied and demonstrated in different engineering design problems with six different aspects:
• To identify critical and detailed characteristics of the designed product that actually distinguish the impact to specific optimum
behaviour – deep insight and understanding of the complexities of the design problem.
• To distinguish the correlations between feasible and infeasible areas of the design space and identify the physical relationships
between the design parameters, objective functions, and constraints.
• To explore discontinuous areas of the design space that translate to topologically different areas of the parameter space –
potentially leading to decisions of deploying appropriate evaluation models to the appropriate areas of the design space.
• To identify patterns and specific trends and combinations of the design parameters that directly relate to optimum behaviour
of all, or particular set, of the objective functions of interest – leading to decision making mechanisms.
• To apply multiple criteria and filter the practicality of the produced optimum design configurations and express preference of
the human designer for decision making.
• To dynamically monitor and steer computational engineering design processes.
Analysis with ||-coords: Identification of Patterns
• Although the two identified patterns are in neighbourhood
areas in the objective functions space, they correspond to
topologically different areas in the design parameters space
• Informative decision making for further multi-disciplinary
analysis and design
Aerodynamic Design of Axial Wind Turbines
Objectives:
•4 Critical aerodynamic and mechanical metrics that express the Annual Energy Production
Hard Constraints
•Transportation, manufacturability, aerodynamic noise
Definition of the Design Space:
12 parameters that control the chord, thickness and twist distributions along the span
Workways Environment
Identifying Causes of Feasible and Infeasible Aerodynamic Behaviour
Analysis with ||-coords: Exploration of Discontinuities
Identified Patterns
Acknowledgements To Tiziano Ghisu, Cambridge Engineering Design Centre, for providing the optimisation data for the preliminary design of core compressor test case.
To Gunter R. Fischer, Nordex Energy GmbH, Hamburg, Germany, for providing the optimisation data for the aerodynamic design of wind turbines test case.
To David Abramson and Hoang Anh Nguyen, Research Computing Centre, University of Queensland, Australia, for building the Workways environment.
• Common geometrical characteristics between the two families of compressor geometries, as well as differences at the root
camper
Blue: feasible solutions
Cyan and Purple: Subsets of feasible solutions
Green: Subset of infeasible solutions
• Visual representations to investigate causes
of infeasible solutions
• Combination of Scatter Plots and ||-cords
• Utilising human pattern recognition skills
• Integrating intuition and experience into the
computational engineering design process
Preliminary Design for Core Compressor
Objectives:
• Maximise isentropic efficiency, maximise surge margin
Constraints
• De Haller number, Koch factor, Static pressure rise coefficient
Definition of the Design Space:
45 design parameters controlling stage pressure ratio, annulus area, flow angles, and number blades
Human-in-the-loop Computational Engineering Design Cycle