Steel is commonly used in the manufacturing of hydraulic turbines. However, this material imposes geometrical limitations that restrain shape optimization of turbine blades. Therefore, GE is considering the replacement of a portion of these steel blades with more versatile materials; composite materials. This study, done in collaboration with GE, involves the design and manufacturing of a hydraulic turbine blade with a thin composite trailing edge.
The blade designed in this study is based on the replacement of about 20% of its chord at the trailing edge end by a composite material. The outside of the composite part is formed of a 2 mm thick skin made of carbon fibre and epoxy. The skin is joined to the blades main stainless steel part by a bonded joint similar to a double lap joint but characterized by an absence of any shape discontinuity on the surface of the blade. A charged epoxy core fills the inside of the skin adding stiffness to the composite part. Some mechanical and physical properties of materials were characterized. The impact of long term water immersion of the blades materials was studied. The epoxy core showed a 0.65% hygroscopic expansion and a tensile strength and elastic modulus reduction of 43% and 46% respectively after an eight month immersion. Mechanical tests performed on bonded joints representing the blades joint showed a mechanical strength reduction of 43% after a six month immersion. The composite materials’ transverse strength decreased by 23% after a three month immersion. Abrasive blasting of stainless steel proved to be the most uccessful surface treatment to promote adhesion of bonded joints among the treatments tested.
Using the finite element method, mechanical analyses were performed to evaluate the factor of safety of the parts of the blade. With numerical models, factors of safety were evaluated as 5.3 for the skin, 0.79 for the epoxy core and 2.0 for the adhesive layer in normal operation conditions. Destructive bending tests were performed on prototypes representing a small portion of the blade. These tests showed a factor of safety of 12.8 for a prototype without any conditioning and 10.1 for a second prototype that was immersed in water for 90 days which represents a 21% reduction. These results were in agreement with predictions made using numerical models. The manufacturing of a complete full scale blade prototype with a thin composite trailing edge validated the manufacturability of the design. Water influence on the mechanical resistance of the design could be avoided by adding a moisture barrier to the skin such as an aluminum film or a hydrophobic coating.
Bilton, K. (Author),
Dubé (Supervisor) &
Demarquette (Co-supervisor),
13 Jul 2016Student thesis: Master's thesis › Master in Engineering: Mechanical Engineering