Optimization of jacket design for large wind turbines Mathias Stolpe & Kasper Sandal Wind Turbine Structures and Component Design
We present a software which can bring advanced mathematics to practical use in wind energy JADOP Jacket Design Optimization Can do conceptual design of jackets in a few minutes through sizing of members with continuous design variables Can also design structures based on predefined catalogue values 2
This is how numerical optimization can be a valuable tool for structural engineers minimize subject to f(x) Ax b 0 σ σ x σ ω ω x ω Design considerations Optimal design problem 3 Design trends
This is how numerical optimization can be a valuable tool for structural engineers minimize subject to f(x) Ax b 0 σ σ x σ ω ω x ω Design considerations Optimal design problem 4 Design trends
Our target is efficient computer aided conceptual optimal design of jackets Conceptual structural optimization of the primary steel structure of jackets Design variables are the cross-sectional properties of the members in the jacket (diameter and thickness) JADOP is a software for conceptual optimal design of jacket-type support structures Developed at DTU Wind Energy as part of the strategic research project ABYSS Advancing BeYond Shallow waters funded by Innovation Fund Denmark 5
JADOP is a design software with analysis capabilities for jacket structures Beam finite elements Decoupled rotor RNA is modelled by lumped masses and inertias DNVGL standards & guidelines Constraints on fatigue, frequency, etc. Validation with aero-elastic simulations (collaboration with AAU & NTNU) Soil modelling & integrated foundation design (collaboration with DTU Civil Engineering) 6
JADOP can do very fast design optimization which is suitable for conceptual design studies Parameter studies on Loads (using damage equivalent loads) Soil properties (including CPT-data) Geometry Height Leg distances Number of sections Number of legs Tower Transition piece etc. 7
Parameter study on leg distance, where 100 jackets are optimized (2 minutes per design) 8
JADOP also has capabilities for design from catalogue dimensions Each variable chosen from a catalogue of values 9
The catalogue design problems become large and challenging to solve DTU 10 MW reference wind turbine, mounted on a jacket in 50 m water depth The jacket is inspired by the INNWIND.EU reference jacket, which is a four-legged, 67 m high jacket with a bottom leg distance of 34 m and top leg distance of 14 m > 40000 nonlinear constraints (yield stress, buckling limits, equivalent fatigue, etc.) Catalogue diameter ([mm]): 600:Δd:1400 Catalogue thickness ([mm]): 15:Δt:80 10
The catalogue design is both different & heavier than the continuous design Δt = 10 mm Δd = 100 mm Whole lines Catalogue design variables 930 tons Dashed lines Continuous design variables 790 tons 11
Δt = 10 mm Δd = 100 mm With larger catalogues, the catalogue design approaches the continuous design Δt = 5 mm Δd = 50 mm Whole lines Catalogue design 930, 852, 817 tons Dashed lines Continuous design 790 tons Δt = 2 mm Δd = 20 mm 12
We continue to develop JADOP to provide additional functionality Integrated design of tower Larger modules in the modelling Transient analysis as part of the optimization Direct coupling to aero-elastic software Other suggestions??? 13
We present a software which can bring advanced mathematics to practical use in wind energy JADOP Jacket Design Optimization Can do conceptual design of jackets in a few minutes through sizing of members with continuous design variables Can also design structures based on predefined catalogue values 14
Design optimization of jacket structures for mass production Kasper Sandal PhD defence Friday 6 October 2017 at 10.00 in H.H. Koch DTU Wind Energy Risø Campus 4000 Roskilde Online webcast: See www.abyss.dk for information 15