Terrain-Induce Turbulence Prediction for Aviation Safety

Motivation Methodology Real application: Forecasting Special Analysis: Alta Airport

Terrain Induced Turbulence Prediction for Aviation Safety

Adil RASHEED, Karstein SØrli

Applied MathematicsSINTEF ICT

[email protected]

June 29, 2010

Adil RASHEED, Karstein SØrli SINTEF ICT Applied Mathematics



Outline

1 Motivation

2 MethodologyTheory

3 Real application: ForecastingNestingValidation

4 Special Analysis: Alta AirportHaugesund AirportKristiansund AirportAlta Airport




Incidents and accidents

On 11 March 1982, Widerøe Flight 933, operated by the Twin Otter LN-BNKcrashed into the Barents Sea near Gamvik, on route from Berlevag Airport toMehamn Airport. All investigations have concluded that the crash resulted fromstructural failure of the aircraft’s tail caused by severe clear-air turbulence.

On 12 April 1990, Widerøe Flight 839, operated by a Twin Otter, crashed into theseas one minute after take-off from Værøy Airport, killing all five on board. Thecause of the crash had been strong and unpredictable wind gusts during take-off,which had exceeded the plane’s limits and created a break-up in the plane’s tailrudder, so the plane became uncontrollable.




Hammerfest Airport: 1 May 2005

On 1 May 2005, the Widerøe Dash-8-100 airplane LN-WIK was buffeted by strongwinds upon landing, and the landing gear on the right wing collapsed. While severalpassengers sustained injuries, there were no fatalities in the incident, but theNorwegian Civil Aviation Authority imposed the strictest wind regulations in Norwayupon the airport.




Amsterdam, Geneva, Paris, Frankfurt

Flat terrain




Norwegian airports

Characterized by fjords, hills, mountains, valleys resulting in rotors, mountain wavesetc. ...




Theory

Governing Equations

∇ · (ρsu) = 0 (1)

DuDt

= −∇(

pd

ρs

)+ g

θd

θs+

1ρs∇ · R + f (2)

DθDt

= ∇ · (γT∇θ) + q (3)

DKDt

= ∇ · (νT∇K ) + Pk + Gθ − ε (4)

DεDt

= ∇ ·(νT

σe∇ε)

+ (C1Pk + C3Gθ)ε

k− C2

ε2

k(5)

SIMRA: Semi IMplicit Reynolds AveragedStructured version for forecastingUnstructured version for special analysis




Theory

Boundary Conditions (for special analysis)

u0(z) =u∗

κ

(ln

zz0

+ W( z

D

))(6)

K (z) = C−1/2µ u2

∗

(1−

zD

)(7)

270 275 280 285 290potential temperature (K)

0

1000

2000

3000

4000

z(m

)

Potential temperature profile for StableStratification

For forecasting: Nesting with Bigger scale model (UM1)




Theory

Terrain induced shear

du

dx

du

dz

L = 12 CLρinf(|U − Vp|)2




Theory

Effect of horizontal and vertical shear

Horizontal shear Vertical shear




Theory

Safety Analysis

F = −cg∂u∂x

+wc

`f

= −c

g`f[u(x + `f /2− u(x − `f /2)t] +

w`f

c(8)

ε1/3 ≈((C1/2µ K )3/2

`t

)1/3

≈ 0.67K 1/2`−1/3t (9)

c is the fly path, g gravitational acceleration u is the wind component along the fly path,w vertical wind component, ε turbulent dissipation, K turbulent kinetic energy `tturbulent length scale `f minimum response distance for landing configuration and is ofthe order of ∼ 500 m, Prevalence of the two conditions F < −0.1 andε1/3 > 0.5 m2/3s−1 correnspond to severe turbulence for commercial aircrafts andrepresent potential danger. These conditions are easily met when

√K > 3 ms−1.




Nesting

Nesting: HIRLAM12− > HIRLAM8− > UM4− > UM1− > SIMRA




Validation

Brønnøysund




Validation

Evenes




Validation

Fordes




Validation

Hammerfast




Validation

Honnigsvag




Validation

Mosjøen




Validation

Ørsta-Volda




Validation

Narvik




Validation

Sandane




Validation

Værnes




Validation

Right tool for the right job

So ... Downscaling does not always work. Use an appropriate model to resolveinteresting scales




Haugesund Airport

Haugesund Airport




Haugesund Airport

Special Analysis: Haugesund airport

Located on the west side of the islandand municipality of Karmøy, southwestof Haugesund.

Operated by Avinor since 1975

Scandinavian Airlines, Norwegian AirShuttle.

Runway dimension: 2120 m × 45 m.

Elevation of 86 m above mean sealevel

The airport had 514, 947 passengersin 2007




Haugesund Airport

Haugesund: Windrose

Spring

Summer

Autumn

WinterThe airport is not bothered by intense turbulence. Uneven side wind from the SSW andsometimes from the north, with strength 13m/s or more, can provide turbulentconditions on the final




Haugesund Airport

Haugesund: Terrain and Mesh

Terrain in the vicinity of the Haugesundairport

300 × 300 times 40 mesh




Haugesund Airport

√K = 3 contours for free stream velocity of 30 m/s

α = 60o

α = 120o

α = 90o

α = 150o

No potentialdanger.Turbulent zonesare located faraway from theairport.




Haugesund Airport

√K contour along the fly path

α = 60o

α = 120o

α = 90o

α = 150o




Haugesund Airport

Haugesund:√

K contours and velocity field projected on the conecontaining the gliding path for a free stream velocity of 20 m/s

α = 60o

α = 90o

α = 120o

α = 150oAdil RASHEED, Karstein SØrli SINTEF ICT Applied Mathematics



Kristiansund Airport

Kristiansand Airport

Braathens SAFE Boeing after landing accident 1977-10-31Adil RASHEED, Karstein SØrli SINTEF ICT Applied Mathematics




Sample Application: Kristiansand Airport (Kjevik)

Situated 4.3 NM (8.0 km) north-east ofthe city Kristiansand, Vest-Agder insouthern Norway

The airport serves the Agder districtwith domestic and international flights.

The airport is operated by Avinor.

Surrounded by water on three sidesand hills on the fourth

Vigorous turbulence is experienced forNorthwesterly wind





Kjevik: Mesh and Terrain

300× 300× 40 meshResolution: 50− 200 mDomain size:30 km × 30 km × 3 km

β

β is the gliding angle





Turbulent Kinetic Energy along the surface of the cone





3-D Velocity Field for α = 240o

Flow channelizes through the valleys





3-D Velocity Field for α = 320o

Flow ascends up the mountain





3-D√

K contours for α = 320o





3-D√

K contours for α = 320o

These plots are for free stream velocity of 20 m/s√K scales with this velocity




Alta Airport

Alta




Alta Airport

Special Analysis: Alta airport

Situated 4km northeast of the towncenter of Alta on a plain where AltaRiver flows in the fjord

2, 088 m long runway

Operated by Avinor.

334, 132 passengers served in 2009(busiest in Finnmark)

Provision for wind and turbulencemeasurement on Komsa hill (200m)

SAS, Norwegian




Alta Airport

Alta: Terrain

For 20− 25m/s wind speed strong turbulence is experienced. Wind from SSW to NWis most problematic.




Alta Airport

Alta: Windrose

Spring

Summer

Autumn

Winter




Alta Airport

Alta: Terrain and Mesh

Terrain in the vicinity of Alta airport andgliding plane

300 × 300 times 40 mesh




Alta Airport

Turbulent Kinetic Energy along the surface of the cone

α = 60o

α = 120o

α = 90o

α = 300o

Severeturbulence forα = 60o andα = 90o




Alta Airport

Turbulent Kinetic Energy along the fly path

α = 60o

α = 120o

α = 90o

α = 300o

Severeturbulence forα = 60o andα = 90o




Alta Airport

Alta:√

K = 3 contours for free stream wind speed of 30m/s

α = 60o

α = 90o

α = 120o

α = 300o




Alta Airport

Alta: Effects of Stratification

Ua = 10m/s,Fr = 1.0

Ua = 15m/s,Fr = 1.5

Ua = 18m/s,Fr = 1.8

Ua = 20m/s,Fr = 2.0Adil RASHEED, Karstein SØrli SINTEF ICT Applied Mathematics



Alta Airport

Alta: Effects of Stratification and Komsa hill

Ua = 10m/s,Fr = 1.0 Ua = 15m/s,Fr = 1.5




Alta Airport

Thank you

Do not be afraid: WECOME to Norway :)



Terrain-Induce Turbulence Prediction for Aviation Safety

Education

alta airport terrain

turbulence prediction

special analysis adil

mathematics terrain

karstein srlisintef

alta airporttheory terrain

alta airport amsterdam

alta airport incidents