System Engineering - NREL¸ DTU, Technical University of Denmark System Engineering 2 Title of the presentation 21-aug-2008 Figure 1. The Systems Engineering Process from A. T. Bahill

System Engineering

Flemming Rasmussen, Peter Hauge Madsen

Wind Energy Division

Risø DTU

Risø DTU, Technical University of DenmarkRisø DTU, Technical University of Denmark

System Engineering

21-aug-2008Title of the presentation2

Figure 1. The Systems Engineering Process from A. T. Bahill and B. Gissing, Re-evaluating systems engineering

concepts using systems thinking, IEEE Transaction on Systems, Man and Cybernetics, Part C: Applications and

Reviews, 28 (4), 516-527, 1998.

Risø DTU mission:Quality – Relevance – Impact

Wind Energy Division - Risø DTUTechnical University of Denmark

Research Themes:• WindpowerMeteorology

• Aeroelastic design methods

• Wind Turbine Structures

• Offshore Wind Energy

• Remote sensing & Measurements

• Windpower control& integration

WIND ENERGY

DIVISION

EDUCATION

Wind

Turbines

Wind Energy

Systems

Aeroelastic

Design

Test &

Measurement

Meteorology

INTERNATIONAL

CONSULTING


Wind turbine aerodynamics and aeroelasticity- optimisation

turbulent inflow

Atmospheric boundary layer

2D aerodynamics

terrain roughness

wave forcescode development

terrain induced wind

tip aerodynamics

rotor aerodynamics

3D airfoil data

rotor-tower interaction

root aerodynamics

nacelle aerodynamics

wakes

surface roughnesstransition

dynamic stall

airfoil design

3D aeroelasticity

rotor design

boundary layer turbulence

aero-acoustics

aero-servo-elasticity

large deflections

optimization

grid interaction and

integration


Aeroelastic codes and simulations

Engineering sub-models for simulation of:

yawed flow

dynamic stall

unsteady blade aerodynamics

unsteady inflow

tip loss

tower shadow

wakes from neighboring turbines

simulation of turbulent inflow

control

Aeroelastic codes for

time simulations:

• FLEX4

• BLADED

• FAST

• HAWC2

- simulations in real

time or faster


HAWC2 – Risø DTU’s code for wind turbine load and response

• A tool for simulation of wind turbine load & response in time domain.

• Normal onshore turbines; 3B, 2B, pitch control, (active) stall

• Offshore turbines (monopiles, tripods, jackets)

• Floating turbines (HYWIND concept for now, later… Sway, Poseidon).

• Based on a multibody formulation, which gives great flexibility

• It is a knowledge platform!

• New research/models are continuously implemented and updated.

• Core is closed source. E.g. Structure, aerodynamics, hydrodynamics, solver…

• Submodels are open-source. E.g. water kinematics, standard controllers, generator models.


HAWC2: Numerical platform for time simulation studies of complex wind turbine loading.

Meandering wake

Distributed control

Large deflectionsNear wake

Offshore structures

- and floating


Components in System Engineering at Risø

• HAWC2 aeroelastic optimisation

• HAWCStab2: combined aeroelastic and control design

• Airfoil optimisation (including noise)

• Blade design including structure, noise

• Topology optimisation of wind farms (Site specific design)

In progress:

• (Integrated design of offshore support structure and turbine)

• (Turbine, wind farm and grid-interaction)

21-aug-2008Title of the presentation8


Optimisation of blade pre-sweep

Athens 7-9 April, 2010

• 5 sweep curve exponents combined with

24 tip offsets = 120 + 1 (ref.) blade variants

• Steady wind speeds (4..26 m/s, 1m/s steps)

• Turbulent wind speeds (4..26 m/s, 2m/s steps, 10 min series) same

seed number, TI=0.18

• Total of 1573 simulations

Analysis of various pre-swept geometries and wind conditions


Equivalent Loads (fatique)

Flapwise Edgewise

Torsion


HAWCStab2 – Aeroservoelastic modal analysis Example: Swept blades

First flapwise mode shapes at 12 m/s First flapwise frequencies and damping


Aeroelastic stability

Local torsion

about e3

Prebend beam

Forward pre-bending

increases edgewise stability


Airfoil optimisation: AirfoilOpt - flow chart

• A direct multipoint and interdisciplinary design tool

• An optimization algorithm coupled with a general 2D flow solver: XFOIL (or EllipSys2D)

• B-spline formulation of the 2D airfoil shape

• A Simplex optimizer – or other type of optimization algorthm

• Multiple angles of attack are analyzed

• Model of moment of inertia with shell thickness

• 3D tool with Cubic B-spline formulation of the 3D blade shape and rotor flow calculations

INTERFACE

Optimum

airfoil shape

Initial

airfoil

shape

Objective function

Constraints

Design variables

FLOW

SOLVER

STRUCTURAL

CALCULATIONS

OPTIMIZATION

ALGORITHM

p u( )

x

y


Airfoil design

• Several lessons have been learned from four airfoil series designed 1998 to 2007 (Risø-A1, Risø-P, Risø-B1 and Risø-C2) :

1. Roughness insensitivity is very important

2. High aerodynamic efficiency is very important

3. Structural stiffness is important

4. High compatibility is important

5. Low noise is important

6. High lift is important for some concepts

Risø-A1-18

Risø-P-18

Risø-B1-18

Risø-C2-18

Designed for stall regulation

Designed for pitch regulation

Designed for pitch regulation variable speed

Designed for pitch regulation variable speed


New development: Noise (TNO model)


Rotor designHAWTOPT: Possibilities

• The analysis part can contain:

• Aerodynamic models (e.g. BEM)

• Aeroelastic calculations (e.g. HAWC2)

• Aeroelastic stability calculations (e.g. HAWCSTAB2)

• Noise calculations (e.g. BPM-model)

• Simple structural models (predicting mass, stiffness, tip deflection)

• The general part can contain:

• Object functions:

• CP, CT, power, thrust, Annual Energy Production (AEP), noise, fatigueloads etc.

• Design variables:

• Blade chord, twist, thickness distribution, tower height etc

• Constraints:

• Noise, CT, thrust, maximum chord, blade thickness, tip deflection, fatigue loads etc

Rotor designHAWTOPT: Program structure

Objective function

Design variables

Constraints

Initial guess

OPTIMIZATION

ALGORITHM New designANALYSIS

MODULES

Response

Operational

conditionsAssumptions

Evaluate

design

No

Yes

Convergence

General part Analysis part


Tower base flange, Mx

7.29 7.3 7.31 7.32

x 105

6.176

6.177

6.178

6.179

6.18

6.181

6.182

x 106

1.9

1.95

2

2.05

2.1

2.15

2.2x 10

4

Electrical power

7.29 7.3 7.31 7.32

x 105

6.176

6.177

6.178

6.179

6.18

6.181

6.182

x 106

1.655

1.66

1.665

1.67

1.675

1.68x 10

7

Topfarm wind farm optimization approach- loads and power

• Fast and robust optimization scheme for placing of wind turbines for optimum (lowest) cost of energy

• Wake modeling using DWM (Dynamic Wake Meandering)

• Quick lookup and summation for lifetime fatigue loads in a database based on HAWC2 aeroelastic simulations

• Cost function including: Annual energy production and costs of: Turbines, Grid, Foundation and O&M

0.005 0.01

0.015 0.02

0.025

30

210

60

240

90270

120

300

150

330

180

0

Tower base flange, Mx

7.29 7.3 7.31 7.32

x 105

6.176

6.177

6.178

6.179

6.18

6.181

6.182

x 106

1.9

1.95

2

2.05

2.1

2.15

2.2x 10

4

Electrical power

7.29 7.3 7.31 7.32

x 105

6.176

6.177

6.178

6.179

6.18

6.181

6.182

x 106

1.655

1.66

1.665

1.67

1.675

1.68x 10

7

Tower base lifetime fatigue loads in wind farm

Annual energy yield for each turbine

Wind rose

Example: A 20 WT wind farm

Turbines in wake have higher loads produce less energy!!

Risø DTU, Technical University of DenmarkRisø DTU, Technical University of Denmark 20 40 60 80

-2

-1.5

-1

-0.5

0

time [s]

pitch/r

oll

[deg]

platform motion; H0 = 0.48 m and Ws = 9.5 m/s

Combined wind and wave energy – PoseidonWorld’s first floating combined wind and wave energy plant

19

Mooring pointWave energy

conversion device

Three GAIA 11kW free yaw and teetering down wind turbines

Grid connection point

• Primarily a wave energy platform

• Large dimensions makes it stable

and suitable for wind turbines

• Has been operating for 4 months

• Comprehensive measurement

campaign

• Modeling and simulation work is

ongoing

Code to code validation of new modeling tool


Each feather is a ”blade”:

-curved blade axis (main spar)

-airfoil shape (ribs)

The Vision

An example of Integrated Design

Flap-torsion coupling

due to sweep


Cross section of feather

Flap-camber coupling due to

rotation of ribs from shear web

deformation


Leading edge

Flap-camber coupling


Flap-camber coupling

Leading edge


Integrated Design:Wing tip is a propeller

The tip feathers forms

a propeller to deliver

thrust from flapping

due to passive built –in

flap-torsion-camber

coupling

Increasing main

spar stiffness

Torsion turns lift

to thrust

Documents

System Engineering - NREL¸ DTU, Technical University of Denmark System Engineering 2 Title of the presentation 21-aug-2008 Figure 1. The Systems Engineering Process from A. T. Bahill