![Ilmas Bayati
PoliMi Wind Turbine Properties
Target mass [g] 550
Total mass [g] 1270
Components [g] 903
Custom parts [g] 367
Rotor mass](https://image.slidesharecdn.com/bayativer1-180630152248/85/6-DoF-Hydrodynamic-Modelling-For-Wind-Tunnel-Hybrid-HIL-tests-of-FOWT-the-real-time-challenge-42-320.jpg)
6-DoF Hydrodynamic Modelling For Wind Tunnel Hybrid/HIL tests of FOWT: the real-time challenge
- 1. 6-DoF Hydrodynamic Modelling For Wind Tunnel Hybrid/HIL Tests of FOWT: The Real-time Challenge I.Bayati, A.Facchinetti, A.Fontanella & M.Belloli Politecnico di Milano Department of Mechanical Engineering International Conference on Ocean, Offshore & Arctic Engineering June 17-22, 2018 Madrid, Spain OMAE 2018-77804
- 2. Ilmas Bayati • Experimental Setup (H2020/Lifes50+ project) • Hybrid/HIL methodology • Modelling approach and case study • Real-time challenges • Conclusions and ongoing activities Outline
- 3. Ilmas Bayati • Experimental Setup (H2020/Lifes50+ project) • Hybrid/HIL methodology • Modelling approach and case study • Real-time challenges • Conclusions and ongoing activities Outline
- 4. Ilmas Bayati Experimental Setup Hybrid/HIL 6DoF system λ 𝐿 = 75
- 5. Ilmas Bayati Experimental Setup Hybrid/HIL 6DoF system λ 𝐿 = 75 Why Hybrid Testing? • Scaling issues • Separate facilities Wind Tunnel/Ocean Basin
- 6. Ilmas Bayati Experimental Setup Hybrid/HIL 6DoF system λ 𝐿 = 75 National Instrument PXI/VeriStand Wind 6-DoF Real Time ModelDisplacements Aerodynamic Forces Controller Wind Turbine Controller Robot Motion Angular speed + pitch Hydro/Substructure Aerodynamic Forces Load Balance PMAC dSPACE
- 7. Ilmas Bayati End effector 6-components balance Links Custome-made joints Motor-reducer units Experimental Setup Hybrid/HIL 6DoF HexaFloat Robot In-house • Design • Realization • Calibration • Control
- 8. Ilmas Bayati • Experimental Setup (H2020/Lifes50+ project) • Hybrid/HIL methodology • Modelling approach and case study • Real-time challenges • Conclusions and ongoing activities Outline
- 9. Ilmas Bayati Experimental Setup Real-Time: Measurements + Computations + Actuation λ 𝐿 = 75 National Instrument PXI/VeriStand Wind 6-DoF Real Time ModelDisplacements Aerodynamic Forces Controller Wind Turbine Controller Robot Motion Angular speed + pitch Hydro/Substructure Aerodynamic Forces Load Balance PMAC dSPACE
- 10. Ilmas Bayati Hybrid/HIL methodology Real-Time: Measurements + Computations + Actuation λ 𝐿 = 75 Wind Displacements Controller Wind Turbine Controller Robot Angular speed + pitch Aerodynamic Forces Load Balance PMAC dSPACE Aerodynamic Forces Motion 6-DoF Real Time Model Hydro/Substructure
- 11. Ilmas Bayati Hybrid/HIL methodology Real-Time: Measurements + Computations + Actuation λ 𝐿 = 75 Wind Displacements Controller Wind Turbine Controller Robot Angular speed + pitch Aerodynamic Forces Load Balance PMAC dSPACE Aerodynamic Forces Motion 6-DoF Real Time Model Hydro/Substructure
- 12. Ilmas Bayati Hybrid/HIL methodology Real-Time: Measurements + Computations + Actuation λ 𝐿 = 75 Wind Displacements Controller Wind Turbine Controller Robot Angular speed + pitch Aerodynamic Forces Load Balance PMAC dSPACE Aerodynamic Forces Motion 6-DoF Real Time Model Hydro/Substructure
- 13. Ilmas Bayati • Experimental Setup (H2020/Lifes50+ project) • Hybrid/HIL methodology • Modelling approach and case study • Real-time challenges • Conclusions and ongoing activities Outline
- 14. Ilmas Bayati Modelling approach Matlab/Simulink for Real-Time Implementation Platform+Turbine Inertia Tensor Platfform Added Damping (e.g. Ocean Basin) Platform+Turbine • Gravitational • Restoring Substructure • Radiation • Mooring Forces • Viscous forces • Wave Forces Derived from measurements
- 15. Ilmas Bayati Modelling approach Matlab/Simulink for Real-Time Implementation Platform+Turbine Inertia Tensor Platfform Added Damping (e.g. Ocean Basin) Platform+Turbine • Gravitational • Restoring Substructure • Radiation • Mooring Forces • Viscous forces • Wave Forces Derived from measurements Challenging Modelling for Real-Time!!
- 16. Ilmas Bayati Modelling approach Matlab/Simulink for Real-Time Implementation Platform+Turbine Inertia Tensor Platfform Added Damping (e.g. Ocean Basin) Platform+Turbine • Gravitational • Restoring Substructure • Radiation • Mooring Forces • Viscous forces • Wave Forces Derived from measurements Challenging Modelling for Real-Time!! Offline verification against FAST/HydroDyn/MoorDyn
- 17. Ilmas Bayati Case study Triple Spar/DTU 10MW (Innwind.eu)
- 18. Ilmas Bayati • Experimental Setup (H2020/Lifes50+ project) • Hybrid/HIL methodology • Modelling approach and case study • Real-time challenges • Conclusions and ongoing activities Outline
- 19. Ilmas Bayati Experimental Setup Real-Time challenges λ 𝐿 = 75 Wind Displacements Aerodynamic Forces Controller Wind Turbine Controller Robot Angular speed + pitch Aerodynamic Forces Load Balance PMAC dSPACE 6-DoF Real Time Model Hydro/Substructure Motion
- 20. Ilmas Bayati Experimental Setup Real-Time challenges λ 𝐿 = 75 Wind Displacements Aerodynamic Forces Controller Wind Turbine Controller Robot Angular speed + pitch Aerodynamic Forces Load Balance PMAC dSPACE 6-DoF Real Time Model Hydro/Substructure Motion
- 21. Ilmas Bayati Experimental Setup Real-Time challenges λ 𝐿 = 75 Wind Displacements Aerodynamic Forces Controller Wind Turbine Controller Robot Angular speed + pitch Aerodynamic Forces Load Balance PMAC dSPACE Real-time contraints • Limited number of computations • Limited memory Modelling • Discretization Platform and Mooring nodes (Natural Freqs. and damping) • Contribution of mooring lines’ forces • Dimension of frequency-dependent files Spectrum width Physical consistency • Time step Convergency and Accuracy 6-DoF Real Time Model Hydro/Substructure Motion
- 22. Ilmas Bayati Modelling approach Discretization Platform Mooring lines • Effect on natural frequencies • Effect on damping
- 23. Ilmas Bayati Modelling approach Discretization Platform Mooring lines Wave kinematics • Effect on natural frequencies • Effect on damping
- 24. Ilmas Bayati Modelling approach Discretization Platform Mooring lines Wave kinematics • Effect on natural frequencies • Effect on damping
- 25. Ilmas Bayati Modelling approach Mooring lines force contributions MoorDyn Lumped-elements model
- 26. Ilmas Bayati Modelling approach Mooring lines force contributions MoorDyn Lumped-elements model
- 27. Ilmas Bayati Modelling approach Mooring lines force contributions MoorDyn Lumped-elements model T = Tensile Load C = Damping W =Weight B = Contact with seabed Dp = Transverse viscosity Dq = Tangential viscosity
- 28. Ilmas Bayati Modelling approach Mooring lines force contributions MoorDyn Lumped-elements model T = Tensile Load C = Damping W =Weight B = Contact with seabed Dp = Transverse viscosity Dq = Tangential viscosity
- 29. Ilmas Bayati Modelling approach Mooring force relevant contributions Combined free decay tests - Viscosity depends only on the nodes velocity - No wave-exciting forces
- 30. Ilmas Bayati Modelling approach Mooring force relevant contributions Combined free decay tests - Viscosity depends only on the nodes velocity - No wave-exciting forces
- 31. Ilmas Bayati Modelling approach Mooring force relevant contributions -Constant - Potentially constant Varying Neglected Anchor FairleadSeabed Off-seabedIntermediate T = Tensile Load C = Damping W =Weight B = Contact with seabed Dp = Transverse viscosity Dq = Tangential viscosity Summary
- 32. Ilmas Bayati Modelling approach File dimension and sampling handling Topic Issue Real-Time Contraint Solution/Final Parameter Integration time-step Mooring dynamics = faster FAST: 0.015/0.001s Unique 0.015s (FS) Irregular Sea Viscous forces Wave kinematics (η, u, v) • Loading u,v matrices: N platform elements x time length • Dependency on platform istantaneous position (wheeler) Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) • No Ptfm dependency Wave exciting forces 1° order • Based of WAMIT outputs • Loading force matrices: 6 x time length Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) Wave exciting forces 2° order (Newmann) Wave exciting forces 2° order (QTF) Simulation (measurement) time • Consistency with Ocean Basin meas. • Affecting freq. resolution • Memory available • Simulation time Vs measurement time 2000-2500 s (FS) FS = Full scale
- 33. Ilmas Bayati Modelling approach File dimension and sampling handling Topic Issue Real-Time Contraint Solution/Final Parameter Integration time-step Mooring dynamics = faster FAST: 0.015/0.001s Unique 0.015s (FS) Irregular Sea Viscous forces Wave kinematics (η, u, v) • Loading u,v matrices: N platform elements x time length • Dependency on platform istantaneous position (wheeler) Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) • No Ptfm dependency Wave exciting forces 1° order • Based of WAMIT outputs • Loading force matrices: 6 x time length Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) Wave exciting forces 2° order (Newmann) Wave exciting forces 2° order (QTF) Simulation (measurement) time • Consistency with Ocean Basin meas. • Affecting freq. resolution • Memory available • Simulation time Vs measurement time 2000-2500 s (FS) FS = Full scale
- 34. Ilmas Bayati Modelling approach File dimension and sampling handling Topic Issue Real-Time Contraint Solution/Final Parameter Integration time-step Mooring dynamics = faster FAST: 0.015/0.001s Unique 0.015s (FS) Irregular Sea Viscous forces Wave kinematics (η, u, v) • Loading u,v matrices: N platform elements x time length • Dependency on platform istantaneous position (wheeler) Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) • No Ptfm dependency Wave exciting forces 1° order • Based of WAMIT outputs • Loading force matrices: 6 x time length Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) Wave exciting forces 2° order (Newmann) Wave exciting forces 2° order (QTF) Simulation (measurement) time • Consistency with Ocean Basin meas. • Affecting freq. resolution • Memory available • Simulation time Vs measurement time 2000-2500 s (FS) FS = Full scale
- 35. Ilmas Bayati Modelling approach File dimension and sampling handling Topic Issue Real-Time Contraint Solution/Final Parameter Integration time-step Mooring dynamics = faster FAST: 0.015/0.001s Unique 0.015s (FS) Irregular Sea Viscous forces Wave kinematics (η, u, v) • Loading u,v matrices: N platform elements x time length • Dependency on platform istantaneous position (wheeler) Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) • No Ptfm dependency Wave exciting forces 1° order • Based of WAMIT outputs • Loading force matrices: 6 x time length Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) Wave exciting forces 2° order (Newmann) Wave exciting forces 2° order (QTF) Simulation (measurement) time • Consistency with Ocean Basin meas. • Affecting freq. resolution • Memory available • Simulation time Vs measurement time 2000-2500 s (FS) FS = Full scale
- 36. Ilmas Bayati Modelling approach File dimension and sampling handling Topic Issue Real-Time Contraint Solution/Final Parameter Integration time-step Mooring dynamics = faster FAST: 0.015/0.001s Unique 0.015s (FS) Irregular Sea Viscous forces Wave kinematics (η, u, v) • Loading u,v matrices: N platform elements x time length • Dependency on platform istantaneous position (wheeler) Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) • No Ptfm dependency Wave exciting forces 1° order • Based of WAMIT outputs • Loading force matrices: 6 x time length Memory available • Offline generation Lookup Tables • Downsampling (x10) = 0.15s (FS) Wave exciting forces 2° order (Newmann) Wave exciting forces 2° order (QTF) Simulation (measurement) time • Consistency with Ocean Basin meas. • Affecting freq. resolution • Memory available • Simulation time Vs measurement time 2000-2500 s (FS) FS = Full scale
- 37. Ilmas Bayati Modelling approach Results Natural frequencies and damping Irregular SeaDecays
- 38. Ilmas Bayati Conclusions and ongoing activities • Hybrid approach for wind tunnel tests • Real-time hydrodynamic modelling = heavy constraints • Sensitivity analysis for semplification • Finalizing LIFES50+ wind tunnel tests • Olav Olsen - Semisub • Tecnalia/Nautilus - Semi
- 39. Ilmas Bayati Experimental Setup Politecnico di Milano Wind Tunnel Boundary Layer Test Section • 36m long • 13.8x3.8m, 16m/s • Turbulence < 2% • Turbulence generators = 25% • 13m turntable Low Turbulence Test Section • 4x3.8m, 55m/s • turbulence <0.1% • open jet / closed test section Boundary Layer Section Low Turbulence Section
- 40. Ilmas Bayati Experimental Setup DTU 10 MW Wind Turbine Scale Model
- 42. Ilmas Bayati PoliMi Wind Turbine Properties Target mass [g] 550 Total mass [g] 1270 Components [g] 903 Custom parts [g] 367 Rotor mass
- 43. Ilmas Bayati PoliMi Robot Properties Parameter Limit value zWsd Center height 523 mm x Surge ± 150 mm y Sway ± 75 mm z Heave ± 75 mm α Roll ± 5 ° β Pitch ± 8 ° γ Yaw ± 3 ° Design displacements (3Hz)
- 44. Ilmas Bayati Analytics Aerodynamic forces from measurements
- 45. Ilmas Bayati Analytics Mooring Lines Forces
- 46. Ilmas Bayati Analytics Mooring Lines Added Mass
- 47. Ilmas Bayati Mooring Lines Forces
- 48. Ilmas Bayati Mooring Lines Forces
- 49. Ilmas Bayati Mooring Lines Forces