Two applications of mesh morphing based on RBF Morph will be presented at ASME Turbo EXPO 2015 that will be take place from 15th to 19th of June at Palais De Congres in Montreal – Canada. Both papers are the result of a joint research between the University of Rome Tor Vergata (Italy), the University of Leeds (UK) and ANSYS UK Ltd.
GT2015-42249
Technical Publication
Multi-objective CHT CFD Optimisation of Film cooling Turbine Blade Geometry with Mesh Morphing
Authors
Gordon E. Andrews, University of Leeds
Derek B Ingham, University of Leeds
Mohamed Pourkashanian, University of Leeds
Daniele Proietti, University of Rome
Alessandro Pranzitelli, ANSYS UK Ltd.
Marco Evangelos Biancolini, University of Rome Tor Vergata
Abstract
A Computational Fluid Dynamics (CFD) optimisation of a single row of film cooling holes was performed.
The aim was to achieve the highest adiabatic cooling effectiveness while minimising the coolant mass flow rate.
The geometry investigated by Gritsch et al. [1] was the baseline model. It consisted of a row of cylindrical, 30o inclined holes, with a mainstream inlet Mach number of 0.6, a blowing ratio of 1 and a plenum for the upstream cooling air flow.
The predictions agreed with the experimental data with a maximum deviation of 6%.
The geometry was then optimised by varying three shape parameters: the injection angle, the lateral hole expansion angle and the downstream compound hole angle. A goal driven optimisation approach was based on a design of experiments table.
The minimisation of the coolant mass flow together with the maximisation of the minimum and average cooling effectiveness were the optimisation objectives. The shape modifications were performed directly in the ANSYS Fluent CFD solver by using the software RBF Morph in the commercial software platform ANSYS Workbench.
There was no need to generate a new geometry and a new computational mesh for each configuration investigated.
The dependency of the average effectiveness along the plane centreline on the three geometrical parameters was investigated based on the metamodel generated from the design of experiments results.
The goal driven optimisation led to the optimal combination of the three shape parameters to minimise the coolant flow without reducing the cooling effectiveness.
The best results were obtained for a geometry with 20o hole angle and 7.5o compound angle injection, leading to a reduction of 15% in the coolant mass flow rate for an enhanced adiabatic cooling effectiveness.
The results also showed the preponderance of the centreline angle over the other two parameters.
GT2015-42251
Technical Publication
Goal Driven Shape Optimisation for Conjugate Heat Transfer in an Effusion Cooled Wall
Authors
Gordon E. Andrews, University of Leeds
Walter Savastano, University of Rome
Derek B Ingham, University of Leeds
Alessandro Pranzitelli, ANSYS UK Ltd.
Marco Evangelos Biancolini, University of Rome Tor Vergata
Mohamed Pourkashanian, University of Leeds
Abstract
Full coverage effusion cooling was numerically investigated by means of conjugate heat transfer (CHT) computational fluid dynamics (CFD) for an array of effusion cooling holes in order to maximise the overall cooling effectiveness for a fixed coolant mass flow rate G, kg/sm2.
The baseline case study consisted of a 152x152x6.35mm perforated wall with a 300 K, 0.18 kg/sm2 coolant flow through a square array of 90o effusion holes. The plate was mounted in a 770 K duct crossflow.
The numerical model was validated against the experimental work of Andrews et al. (1990) and showed a maximum disagreement between the prediction and the experimental data of 3%.
The effects of three geometrical parameters, i.e. the inclination of the holes, the pitch in X direction and the pitch in Y direction, on the overall cooling effectiveness were investigated without varying the coolant flow rate.
The inclination of the effusion holes was varied between -33o and +33o from the plane of the plate, while the pitch in both directions was varied between 10.64mm and 19.76mm. The numerical investigation was performed using the commercial software ANSYS Workbench following a design of experiments approach.
The geometrical modifications were obtained automatically in the CFD solver ANSYS Fluent for each design point by means of the RBF Morph software.
This avoided the manual modification of the geometry and the subsequent mesh generation. The optimised configuration was obtained considering the maximisation of the average overall cooling effectiveness as a goal and a chosen minimum value for the local cooling effectiveness of 0.4 as the constraint.
The results showed that the inclination of the effusion holes and the pitch in the Y direction have a greater impact on the cooling effectiveness than the pitch in the X direction, up to 17%, 28% and 5% from the baseline, respectively, within the range of values considered. An optimal combination of the three parameters was determined.