CREDOS Project. CREDOS Scenario Definitions D3-3

CREDOS Project CREDOS Scenario Definitions D3-3 Contract Number: AST5-CT-2006-030837 Proposal Number: 30837 Project Acronym: CREDOS Deliverable Title: CREDOS Scenario Definitions Delivery Date: T 15 (AUG 2007) Responsible: DLR Institute of Flight Systems Nature of Deliverable: R (Report) Dissemination level: PU (Public) File Id N : CREDOS_313_DLR_DLV_D3-3_CREDOS_Scenario_Definitions_v1b.doc Status: Approved Version: v1b Date: 19 September 2007 Approval Status Document Manager Verification Authority Project Approval DLR Airbus Project Management Committee Carsten Schwarz Gordon Höhne PMC Members WP313 task manager WP3 Leader

Institutsbericht IB 111-2007/ 38 CREDOS Scenario Definitions D3-3 C. Schwarz (DLR-FT) F. Holzäpfel (DLR-PA) Institut für Flugsystemtechnik Braunschweig 36 Seiten 14 Bilder 24 Literaturstellen Deutsches Zentrum für Luft- und Raumfahrt e.v. Institut für Flugsystemtechnik Abteilung Flugzeuge Stufe der Zugänglichkeit: I, öffentlich zugänglich Braunschweig, August 2007 Unterschriften: Institutsdirektor: Prof. Dr.-Ing. S. Levedag Abteilungsleiter: Dr.-Ing. K.-U. Hahn Verfasser: Dipl.-Ing. C. Schwarz Dr.-Ing. habil. F. Holzäpfel

Edition history Edition Nº Date Authors Section Comment 1a 15/08/07 C. Schwarz, F. Holzäpfel 1b 19/09/07 C. Schwarz, F. Holzäpfel page 29, table 4.1 Version 1a correction of listing of representative (generator and follower) aircraft according to sections 2.3.1 and 2.3.2 III

GLOSSARY A/C AGL AIP CREDOS DLR FRA IAS IATA ICAO IMC MTOW PLA RWY PDG PIO SID TAS VMC WV WVE WVL aircraft above ground level Aeronautical Information Publication Crosswind Reduced Separation for Departure Operation German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt) Frankfurt International Airport (IATA airport code) indicated airspeed International Air Transport Association International Civil Aviation Organization instrumental meteorological conditions maximum takeoff weight power lever angle (thrust setting) runway procedure design gradient Pilot-induced oscillation Standard Instrument Departure true airspeed visual meteorological conditions wake vortex wake vortex encounter wake vortex line(s) α angle of attack β angle of sideslip b wing span b 0 (initial) vortex separation γ flight path angle, vertical inclination Δ difference Γ circulation (vortex strength) H altitude ξ*, η*, ζ* normalised control commands (roll, pitch, yaw) L lift n z vertical load factor Φ, Θ, Ψ Euler angles (roll, pitch, heading (also horizontal alignment)) m mass ρ air density t time coordinate V velocity V 2 takeoff climb speed x, y, z Cartesian coordinates IV

Executive Summary CREDOS subtask 3.1.3 and this report provide a definition of wake vortex encounter scenarios during the takeoff phase as a basis for the simulations carried out in CREDOS WP 3.1. The parameters of influence for wake vortex encounters during takeoff are discussed and offline wake vortex encounter simulations are conducted to derive tendencies for the impact of relevant parameters. V

Table of Contents EXECUTIVE SUMMARY 1 INTRODUCTION 1 2 SURVEY OF WAKE VORTEX ENCOUNTER SCENARIO PARAMETERS 2 2.1 TRAFFIC SCENARIO 2 2.1.1 Air traffic distribution 2 2.1.2 A/C combination 3 2.1.3 A/C trajectory 4 2.1.4 Takeoff segment/ phase 7 2.2 ENCOUNTER SCENARIO/ CONDITIONS 10 2.2.1 A/C state 10 2.2.2 Orientation 10 2.2.3 Relative position 12 2.2.4 Encounter altitude 12 2.3 AIRCRAFT PAIRING 12 2.3.1 Vortex generator 14 2.3.2 Follower 15 2.4 METEOROLOGICAL CONDITIONS 16 2.4.1 Wind 16 2.4.2 Atmospheric turbulence 16 2.4.3 Temperature stratification 16 2.4.4 Visual conditions 16 2.5 AIRCRAFT CONTROL 17 2.5.1 Individual pilot behaviour 17 2.5.2 Performance of automatic controllers 17 2.6 SIMULATOR COCKPIT MOTION 17 3 OFFLINE SIMULATIONS 18 3.1 WAKE VORTEX ENCOUNTER OFFLINE SIMULATION 18 3.2 AUTOMATIC CONTROL FOR CLIMB 19 3.3 WAKE VORTEX ENCOUNTER PARAMETER STUDY 19 3.3.1 Simulation reference scenario 19 3.3.2 Parameter variation 21 3.3.3 Offline simulation summary 27 4 CONCLUSION 28 5 REFERENCES 30 V VI

1 Introduction Embedded in CREDOS work package 3 Risk Modelling & Risk Assessment according to the CREDOS Description of Work the objective of subtask 3.1.3 and this report is the Definition of scenarios of interest from a severity point of view considering air traffic scenario, a/c combination, meteorological conditions, etc. As a basis for the simulations carried out in WP 3.1 a framework for the simulator test matrix is to be established. This includes the provision of scenarios of interest and a spectrum of possible encounter conditions/ parameter combinations for the simulator tests. Comparable work has been performed within the scope of the EU project S-Wake dealing with the approach flight phase. A simulator test plan has been developed for piloted simulator tests with the intention to assess the hazard of wake vortex encounters during approach [1]. For wake vortex encounters during takeoff the parameters of influence are discussed and offline wake vortex encounter simulations are performed to derive tendencies for the impact of relevant parameters. However no absolute wake vortex encounter severity assessment will be conducted as this is out of the scope of this task. 1

2 Survey of wake vortex encounter scenario parameters This survey will discuss the parameters of influence for wake vortex encounters during takeoff. 2.1 Traffic Scenario The traffic scenario comprises the air traffic distribution according to aircraft weight classes, the actual A/C combination, the A/C trajectories, and the takeoff segment/ phase. 2.1.1 Air traffic distribution As aircraft type and size have a major influence on the assessment of wake vortex encounter situations, the distribution of the air traffic with respect to aircraft size/ weight classes is of interest. The class-wise air traffic distribution (ICAO weight classes) for selected airports (Table 2.1) is shown in Figure 2.1 according to the EuroBen study [4]. The fraction of heavy aircraft for major airports ranges from 10% to 30%, and light aircraft represent not more than 1% of the movements, which can be neglected with respect to capacity considerations. However from a safety point of view light follower aircraft demand attention. The hourly traffic distribution can vary significantly from the yearly/ monthly average. Typically early in the morning heavies from overseas arrive at major European airports, e.g. more than 90% heavies at that time at London Heathrow [5]. The air traffic mix does not indicate the distribution of A/C combinations, in other words there is no indication on how often parings with A/C from different weight classes occur. However, it can be concluded that obviously in most cases A/C pairings with heavy, medium or both occur at major airports, with heavy heavy predominating early in the morning at least for arrivals [5]. The traffic fraction for the different A/C types for Europe is documented in [18]. Recommendations for specific representative A/C types are made in section 2.3. London Heathrow (EGLL) Frankfurt International (EDDF) Paris Charles de Gaulle (LFPG) Amsterdam Schiphol (EHAM) London Gatwick (EGKK) Manchester Ringway (EGCC) Table 2.1 Airport ICAO codes for selected airports Madrid Barajas (LEMD) Zurich Kloten (LSZH) Munich Franz Josef Strauss (EDDM) Düsseldorf International Airport (EDDL) Barcelona Le Prat (LEBL) 2

Figure 2.1 Traffic mix at selected airports during February and July 2004 (ICAO weight classes and airport codes) [4] 2.1.2 A/C combination Since medium and heavy aircraft are dominating at major airports (see section 2.1.1) and following aircraft smaller than the leader are the most relevant combination from a safety point of view, medium behind heavy is an important scenario for investigations. Light aircraft are relevant as a follower from a safety point of view as the impact of a wake vortex encounter increases with lower weight and size of the encountering aircraft. As far as the air traffic distribution is concerned, the combinations light behind medium or heavy are not of high relevance for major airports; consequently the potential capacity gain is not very high. However for these combinations generally a 2 minutes non-radar departure separation time is applied [7]. In order to investigate possible separation reductions these combinations can become relevant. For the combinations medium behind medium and heavy behind heavy no specific non-radar departure separation minima exist. Hence there is no potential capacity gain as far as rulemaking is concerned. However from a safety point of view aircraft combinations within the same aircraft class can not generally be assumed to be unproblematic. First of all each of the present ICAO weight classes covers a large range of differently sized aircraft, and secondly even an aircraft following the same aircraft type does not necessarily have to be uncritical. Accordingly these combinations shall not necessarily be omitted; however they do not have a high priority when looking at worst case combinations. For the CREDOS concept medium or heavy aircraft departing behind heavy aircraft are taken into account [11]. 3

D3-3 CREDOS SCENARIO DEFINITIONS 2.1.3 A/C trajectory For IFR departures Standard Instrument Departure routes (SID), which are published in the Aeronautical Information Publication (AIP), are applied, frequently also for VFR departures. Figure 2.2 depicts the general route layout of FRA as an example. The ground tracks are fix according to the SIDs. However the altitude profile is only defined as a (minimum) climb gradient or a (minimum) altitude (e.g. cross a certain navigation point at 3500ft or above or maintain a climb gradient of 12%/ 750ft/nm ) and is therefore variable depending on aircraft performance and the applied takeoff and departure procedures (see section 2.1.4). Thus the complete 3D-trajectories are not as well defined as for the approach case. There is a greater unpredictability for the actual departure trajectories corresponding to a given SID which makes it more difficult to define a general corridor for the flight tracks. With respect to wake vortex generation and encounters the different possible flight path angles have to be taken into account (see section 2.2.1). According to [23] straight departures have an initial departure track within 15 of the alignment of the runway centre line, opposed to turning departures with larger turns. The turning departures assume straight flight until reaching an altitude/height of at least 120 m (394 ft). Figure 2.2 Frankfurt arrival and departure routes for jets [6] The different departure SID ground tracks (reference tracks, not only actual tracks) have a greater variability compared to the approach case where there is only one nominal flight path 4

aligned with the RWY direction for the final approach (typically approximately the last 11 nm before touch-down but in many cases even more than that). As an example the RWY 25L/25R departure routes at Frankfurt airport are depicted in Figure 2.3 (topographical) and Figure 2.4 (procedural). The first turn can take place after less than 2 nm already. And even within the first 5 nm the SID tracks can be crossing others. This means that wake vortices from aircraft using different SIDs could possibly be encountered. Therefore the corresponding trajectories have to be taken into account as far as the encounter scenario/ conditions are concerned (compare section 2.2), specifically the relative encounter angle and the altitude of vortex generation (which can be well above or below (rebound effect) the follower A/C trajectory). For the CREDOS project specific emphasis is placed on RWY 25R at FRA. Figure 2.3 Frankfurt RWY 25L/25R departure routes for jets [6] 5

Figure 2.4 Frankfurt RWY 25L/25R north departure routes [10] For the CREDOS concept validation it has to be proven that different tracks (SIDs) are safe when used in crosswind conditions and the following aircraft departs either on the same or on 6

an upwind track compared to the leading aircraft [11]. The same SID is normally not used for consecutive A/C, but this case can also occur. In CREDOS consecutive departures from the closely-spaced parallel runway system are not considered as the CREDOS concept aims at single independent runways. 2.1.4 Takeoff segment/ phase From a flight mechanical point of view the takeoff is segmented according to the changes in aircraft configuration (flap setting and gear), thrust setting, and flight state (flight path angle and acceleration). This is depicted for a typical procedure in Figure 2.5. The gear is usually retracted a few seconds after lift-off. So the time frame of a wake vortex encounter with gear down is relatively small (as well as the period of time for wake generation with gear down). On the other hand, the encounter probability for approaching aircraft is highest at altitudes below 100 m because wake vortices can not further descend below the flight path which has been specifically shown for the landing approach [21]. The thrust setting is either takeoff thrust or maximum continuous thrust, the more safety critical case for the encountering A/C being takeoff thrust because there is no further thrust margin to recover from disturbances. Flaps are extended in the beginning and then retracted, the more safety critical case for the encountering A/C being flaps extended because due to the higher drag aircraft performance (acceleration, climb) is restricted which is disadvantageous for recovery from disturbances. Figure 2.5 Noise abatement takeoff procedures (typically applied for takeoff) [9] Typical examples for takeoff and departures procedures are shown in Figure 2.6 schematically (altitude vs. distance). The modified ATA procedure (MODATA) in combination with reduced takeoff thrust (FLX) is typically used by a major German airline whenever feasible (i.e. for adequate runway length) [8] (which is also corresponding to the noise abatement takeoff, Figure 2.5). 7

TAKEOFF AND DEPARTURE PROCEDURES Modified-ATA Reduced Take-off Thrust (MODATA-FLX) ICAOA Reduced Take-off Thrust (ICAOA-FLX) Modified-ATA Max Take-off Thrust (MODATA-TOGA) ICAOA Max Take-off Thrust (ICAOA-TOGA) Height above Ground Level MODATA-FLX ICAOA-FLX MODATA-TOGA ICAOA-TOGA FLAP RETRACTION 3000 ft (914 m) 1500 ft (457 m) 5 km 10 km 15 km Distance Figure 2.6 Takeoff and departure procedures altitude profiles [8] For the CREDOS initial concept of operation the following three phases for takeoff and departure were defined early in the project with respect to wind dependant wake vortex separations [2]. Phase 1 The Phase 1 covers the whole runway and includes thus the rolling phase and the rotation but also the early stage of the climb phase as long as the aircraft is heading in the runway axis. The aircraft are in this phase at quite low altitude [ ]. Phase 2 The Phase 2 is a transition phase between the Phase 1 and 3 covering the short period of time while the aircraft has already turned but is still at proximity of other SID that could be affected by wake vortices or at proximity of SID taken by other preceding aircraft generating wake vortices. Phase 3 The Phase 3 covers the rest of the Standard Instrumental Departure from the end of the first turn up to the incorporation of the aircraft in the en route traffic. 8

Figure 2.7 CREDOS departure phases early definition [2] According to this definition phase 2 begins with the first turn. As far as the flight mechanical segments are concerned the first turn may take place before the thrust reduction but will generally not happen with gear down. Phase 3 is supposed to not represent any particular hazard by definition [2] however the sufficient amount of separation between departing aircraft or departure routes has to be investigated. 9

2.2 Encounter Scenario/ Conditions 2.2.1 A/C state The relevant parameters of the A/C state (generator and follower) comprise the flight situation (turn/ climb/ acceleration), flap, gear, and thrust setting, flight path angle, Euler angles (Φ, Θ, Ψ), aerodynamic angles (α, β). For gear, flap, and thrust setting refer to section 2.1.4. The flight situations during takeoff and departure comprise climb, acceleration, and turn. The reference case would be climb. However acceleration and turn should also be investigated to achieve a complete picture of possible encounters. The actual aircraft climb angle (flight path angle respectively) is strongly depending on A/C performance and will usually be around 10, occasionally up to approximately 15 (for some A/C types in some cases can also reach approximately 20 ) [12]. Generally the aircraft will not be descending during the departure phase however there could be segments with low climb angles (for acceleration). According to CS-25 [17] (FAR Part 25 respectively) the required minimum climb gradient below 400 ft even with engine failure (gear retracted) is 2.4% for 2 engines, 2.7% for 3 engines, 3.0% for 4 engines, and above 400 ft 1.2% for 2 engines, 1.5% for 3 engines, 1.7% for 4 engines. And no configuration changes are allowed below 400 ft except gear retraction. The standard procedure design gradient (PDG) is 3.3 % if not specified differently [23]. For the symmetrical (standard) case and without a vertical wind component the pitch/ inclination angle is directly related to the flight path angle and the angle of attack. Θ = γ + α (2.1) Typical vales for the pitch angle are at least 15 and for the angle of attack around 8 to 10. The angle of attack is no independent parameter but is depending mainly on A/C weight, speed, and configuration. The angle of sideslip β is usually zero due to minimisation of drag. Aircraft heading (Ψ, orientation of the aircraft longitudinal axis relative to North) and also the flight path azimuth angle (χ, direction of the flight track) do not directly influence the encounter situation, only indirectly with respect to the encounter horizontal angle (see section 2.2.2). For climb and acceleration the bank angle Φ can be assumed to be zero. During a turn the bank angle can reach up to 30. 2.2.2 Orientation The relative orientation of flight path and WV lines (Ψ WVL and γ WVL ) define the encounter angles (Ψ WVE and γ WVE ) which are crucial for the encounter duration and the type of the encounter (e.g. roll dominant etc.). 10

ΨWVE = ΨWVL Ψ (2.2) γ WVE = γ WVL γ (2.3) The horizontal wake vortex encounter angle Ψ WVE can be between 0 and 90 (all other cases are not relevant due to symmetry) for takeoff and departure due to possibly crossing flight tracks (see section 2.1.3). The vertical encounter angle γ WVE is depending on the vertical inclination of the vortex lines and the flight path inclination of the encountering A/C. Vortex evolution is treated in detail in WP2 of the CREDOS project (WV behaviour modelling). For the flight path angle see section 2.2.1. Because of the sinking of the vortices the vertical inclination of the vortex lines γ VWL will be slightly higher than the flight path inclination of the generating A/C γ (Figure 2.8). Due to the variation in flight path angles the follower can have a higher or smaller climb angle than the generator. Hence the vertical wake vortex encounter angle will be relatively small, i.e. between around -10 and 15. flight path vortex line γ WVL ( γ ( x g z g Figure 2.8 Side view of generator A/C during climb with flight path inclination and vortex inclination angles (not to scale) Under certain circumstances the two vortices can have different altitudes, i.e. have a vortex bank angle. This also contributes to the relative encounter orientation combined with the follower flight path and follower bank angle. Since the follower flight path for takeoff will generally have a positive inclination not only the vortex bank angle is relevant for the relative encounter orientation. Hence the vortex bank angle is assumed to have no significant direct impact on the follower. For sufficiently long times within the area of influence of a wake vortex the reaction of the encountering aircraft can build up to possibly severe A/C states. If the duration of the encounter is relatively short, no noteworthy changes of the A/C state occur. However accelerations within a short period of time may occur, possibly being harmful to aircraft structure and impacting 11

passenger comfort. The encounter duration is not an independent parameter but is depending on the encounter angles and the follower airspeed. 2.2.3 Relative position The relative position of the WV lines and the encountering aircraft is decisive for the encounter distance. The closest distance of the encountering aircraft to the vortex pair is generally influencing the severity of the wake vortex encounter. The closer the wake vortex the more severe is the encounter in general. A flight through one vortex usually constitutes the worst case scenario. Which encounter distances pose no noteworthy effect on the encountering aircraft depends mainly on vortex strength and follower aircraft. An absolute wake vortex encounter severity assessment is out of the scope of this task. 2.2.4 Encounter altitude The altitude of the WVE is an important parameter for the severity assessment affecting the time to recover and avoid ground impact. The parameter is strongly connected to the A/C trajectory (section 2.1.3) and the takeoff/ departure segment/ phase (section 2.1.4). Within the CREDOS project is has been decided to be limit considerations to altitudes below 3000 ft above ground level (AGL). However as far as certification is concerned the takeoff ends at 1500 ft AGL [17]. The assessment of the effect of encounter altitude has mainly to rely on subjective pilot judgement and therefore has to be subject to investigation by piloted simulations. In any case altitude has to be considered for vortex habitation (and possibly by vortex velocity distribution) models taking into account the ground effect. In ground proximity vortex descent is restricted by the surface. As a consequence transport by weak crosswind away from the runway centreline can be compensated by the self-induced velocity of the luff vortex and the vortices may rebound which both increases the encounter probability [22]. 2.3 Aircraft Pairing This section deals with the relevant parameters directly belonging to each of the involved A/C and representative A/C types for weight categories. In any case the most important aircraft characteristics are wing span, speed and mass, and for the generator vortex strength respectively. The specific A/C which are subject to investigation should cover the possible ranges of mass, wing span, and vortex strength for the respective weight category. Based on the Eurocontrol BADA database [24] covering over 80% of the European air traffic [18] the wing span of 87 A/C is shown related to the maximum takeoff weight in Figure 2.9 (with indication of ICAO weigh classes (dashed lines)). The wing spans in the medium category range from 10 m to 40 m and in the heavy category from 40 m to 65 m. The takeoff speeds in the medium category range approximately from 50 m/s to 80 m/s and in the heavy category approximately from 65 m/s to 85 m/s (Figure 2.10). 12

80 70 60 wing span (b) [m] 50 40 30 20 10 0 0 50 100 150 200 250 300 350 400 maximum takeoff weight (MTOW) [t] Figure 2.9 Wing span vs. maximum takeoff weight with indication of ICAO weigh classes (dashed lines), representative heavy generator A/C o, and representative medium follower A/C O [24] 90 80 70 takeoff speed (V) [m/s] 60 50 40 30 20 10 0 0 50 100 150 200 250 300 350 400 maximum takeoff weight (MTOW) [t] Figure 2.10 Takeoff speed vs. maximum takeoff weight with indication of ICAO weigh classes (dashed lines), representative heavy generator A/C o, and representative medium follower A/C O [24] 13

2.3.1 Vortex generator For the vortex generating A/C the relevant parameters are (initial) vortex strength (depending on the actual A/C mass, speed and vortex spacing), the vortex spacing, depending on the wing span and wing loading (respectively aircraft configuration), and the resulting wake vortex flow field in the plane perpendicular to the flight path, which is depending on the vortex strength, core radius and vortex spacing. It has been shown that the direct influence of vortex spacing on the encountering A/C is low for moderate encounters based on a hazard area definition which is validated for approach and landing [14]. initial vortex strength (actual A/C mass, speed, vortex spacing) vortex spacing (wing span, wing loading/ aircraft configuration) wake vortex flow field (vortex strength, core radius, vortex spacing) Γ 0 b 0 V WV = f(y,z) Table 2.2 Vortex generator parameters The initial vortex strength (circulation Γ 0 ) is depending on the lift, airspeed, air density and the wing span respectively the vortex separation according to the KUTTA-JOUKOWSKY theorem. Γ 0 L = ρvb 0 (2.4) Figure 2.11 shows the initial circulation for the same A/C database as above based on the maximum takeoff weight, takeoff speed and elliptical lift distribution. 800 700 initial vortex strength (Γ0) [m²/s] 600 500 400 300 200 100 0 0 50 100 150 200 250 300 350 400 maximum takeoff weight (MTOW) [t] Figure 2.11 Initial vortex strength (circulation) vs. maximum takeoff weight with indication of ICAO weigh classes (dashed lines), representative heavy generator A/C o, and representative medium follower A/C O [24] 14

Generally speaking larger and heavier A/C generate stronger vortices which tends to lead to more severe wake encounters. However stronger vortices will sink faster which impacts the encounter probability. These A/C types represent the ICAO heavy category: A/C type designator [19] A/C type European traffic fraction [18] FRA traffic fraction B744 Boeing B747-400 1.66% 7.59% B772 Boeing B777-200 1.11% 1.89% A343 Airbus A340-300 0.87% 1.91% A333 Airbus A330-300 0.33% 1.93% A306 Airbus A300-600 0.43% 3.78% A310 Airbus A310 0.71% 0.50% Table 2.3 Representative heavy generator A/C 2.3.2 Follower The relevant parameters for the following A/C are the actual mass (inertia respectively), geometrical dimensions, true airspeed V TAS (typically takeoff climb speed = stall speed + 20%), aerodynamics (e.g. lift coefficient), centre of gravity, and the available control power. The landing gear is supposed to have minor influence on the direct A/C reaction posed by the disturbance, however gear extracted would be the worst case, because due to the higher drag aircraft performance (acceleration, climb) is restricted which is disadvantageous for recovery from disturbances. But the gear is usually retracted a few seconds after lift-off (see section 2.1.4) and therefore not of noteworthy relevance. A lower actual A/C mass will present a lower inertia to disturbances and hence result in more severe A/C reactions. However A/C performance is higher with lower actual mass which is beneficial for recovery from disturbances. These A/C types represent the ICAO medium category: A/C type designator [19] A/C type European traffic fraction [18] FRA traffic fraction A320 Airbus A320 8.55% 14.25% B733 Boeing B737-300 5.50% 9.37% CRJ7 Canadair Regional Jet CRJ-700 0.74% 2.30% E145 Embraer 145 2.99% 4.22% AT45 ATR 42-500 0.75% 1.07% SF34 Saab 340 - - Table 2.4 Representative medium follower A/C 15

2.4 Meteorological Conditions During final approach and landing aircraft follow quite precisely the ILS. Due to the relatively small height deviations of subsequent aircraft the meteorological situations are of outmost importance which prevent the usually anticipated descent of wake vortices below the glide slope. Candidate meteorological conditions are updrafts in convective situations (generally speaking turbulence), stable temperature stratification, shear layers, and ground proximity. All the mentioned scenarios tend also to accelerate vortex decay. Because in departure situations follower aircraft are not expected to fly typically at the same altitude as the generator aircraft, meteorological conditions with potential vortex rebounds do not deserve the same special attention as for approaching and landing aircraft. For CREDOS, in particular, lateral transport by crosswind and, secondly, vortex decay due to various atmospheric parameters are of interest. 2.4.1 Wind The main factor for reduction of departures separations in this project by definition is crosswind. The crosswind speed is affecting the probability of a wake vortex encounter due to vortex transport and also the vortex orientation. The limit for the maximum crosswind speed for takeoff for a typical short-to-medium range commercial passenger aircraft according to [9] is 30 kt. Other wind effects like windshear may also play a role as the situation could be more demanding for the pilot and the wake vortices may tilt, reduce descent speed, and even rebound. On the other hand realistic values for wind speed & direction (constant or varying with time) should not significantly affect the encounter severity. Hence the reference case could be the case without wind or with moderate crosswind. 2.4.2 Atmospheric turbulence The degree of the atmospheric turbulence has a major influence on the vortex decay. In this respect no turbulence represents the worst case. In convective turbulence updrafts may compensate the descent speed of the wake vortices. Also the accuracy for the task to maintain the flight path and state is affected by turbulence as well as the workload of the pilot. High turbulence levels can be considered not relevant for wake vortex encounters because the quickly decaying vortices do not affect the encountering A/C more than a certain degree of turbulence. Hence the case without atmospheric turbulence can be considered as the worst case. 2.4.3 Temperature stratification Temperature stratification reduces wake vortex descent and may even cause rebounding wake vortices. On the other hand, temperature stratification accelerates vortex decay. Because follower aircraft are not expected to fly typically at the same altitude as the generator aircraft neutral stratification should represent the most critical case. 2.4.4 Visual conditions The visibility is categorised into two types, visual meteorological conditions (VMC) and instrumental meteorological conditions (IMC). Another parameter is the altitude of the ceiling 16

below which no clouds are present. With instrumental meteorological conditions present the pilot workload is significantly higher as one has to rely fully on instruments. In this case the same wake vortex encounter will have a higher severity compared to under VMC. Hence the IMC case represents the worst case. The ICAO Procedures for Air Navigation Services [7] allow maintaining own visual separations in VMC if authorised by ATC, also for departure. 2.5 Aircraft Control There can be no a-priori judgement whether manually or automatically piloted wake vortex encounters represent the worst case. This has to be subject to investigation depending on the specific problem. 2.5.1 Individual pilot behaviour For manually piloted WVE the individual skill of the pilot to cope with the situation does play a major role for the severity of the resulting flight situation. This comprises the reaction time as well as the amplitude and frequency of the pilot commands. In the worst case the WVE may trigger pilot-aircraft-coupling (also known as PIO), which could result in a more severe aircraft reaction than the wake vortex induced disturbance in the first place. In any case the manually piloted results will not be reproducible. 2.5.2 Performance of automatic controllers Opposed to human pilots automatic controllers (e.g. autopilot) do not have a reaction time like humans, however there will be a time delay due to signal processing and computation time. Like human pilots automatic controllers could overshoot due to the designed gain factors. In any case automatic controllers will generate reproducible results which depend on the performance of the control system. 2.6 Simulator cockpit motion The mode of cockpit motion (on/ off) has no strong influence on pilot perception and handling [3], but using cockpit motion can be recommended (if applicable) to provide the most realistic acceleration impressions to the pilots. it must be assumed that pilots rate simulators equipped with motion systems as being more realistic overall [3] 17

3 Offline simulations The identification of scenarios of interest for wake vortex encounters during takeoff is supported by offline wake vortex encounter simulations using parameter variation for the relevant parameters of influence. This way the influence and sensitivity of the selected parameters can be determined. 3.1 Wake vortex encounter offline simulation The wake vortex encounter simulation environment is a 6 degree of freedom aircraft simulation comprising models for vortex generation and aging, the representation of the wake vortex induced velocity distribution and the modelling of the encounter aircraft reaction with an aerodynamic interaction model. The simulation system is described including validation in detail in [13], [14] and [15]. The wake vortex encounter simulation environment is capable of simulating the flight through a wake vortex flow field of a pair of straight vortices of a certain age generated by a given generator aircraft. The simulated (encounter) aircraft is a VFW 614 twin engine jet aircraft (ICAO class 'medium', MTOW = 21 t) which is operated by DLR as a test aircraft (ATTAS Advanced Technologies Testing Aircraft System), specifications and 3-view see Table 3.1 [16], [20]. General Characteristics Crew 2 Capacity Length Wingspan Height Empty Weight Wing area Max Takeoff Weight Powerplant Performance Max Cruise Speed Range (with 40 passengers) Initial climb 40-44 passenger in 4 abreast seating 20.60m (67ft 7in) 21.50m (70ft 6.5in) 7.82m (25ft 8in) 12 179kg (26 850lb) 64.0m² (689sq ft) 19 958kg (44 000lb) Two 7473lb (33.2kN) Rolls-Royce/Snecma M45H Mk. 501 turbofans 380kt (704 km/h) 645nm (1195km) 3100ft (945m)/min 3-view graph Table 3.1 VFW-614 specifications [16] and 3-view [20] 18

3.2 Automatic control for climb Uncontrolled wake vortex encounters (without pilot/ autopilot commands) do not allow a severity assessment because the aircraft flight path is chaotically influenced by the initial conditions/ disturbance. For the task to maintain vehicle reference state and flight path a specifically designed automatic controller for the climb phase is used. The applied control concept is feedback control. Reference parameters are compared to the actual values and the error is fed back with a gain to the control surfaces. The reference flight path parameter is the flight path azimuth angle χ. For the vehicle reference state the velocity VTAS and the attitude Φ (bank angle) and Θ (inclination angle/ pitch) are considered. For damping/ attenuation (stability augmentation) the rotational rates p, q, r (roll, pitch and yaw) are controlled to be zero. The controller is tested with regard to reference value response behaviour and control input disturbance response behaviour, i.e. step, impulse and doublet excitations for the reference values and control commands. Furthermore the disturbance response behaviour is tested with wind step excitations, gusts (vertical, lateral and longitudinal) and turbulence. In any case no instabilities occur and the reference values are maintained. Also for some test wake vortex encounters with strong vortices flight path/ state deviations can be compensated. 3.3 Wake vortex encounter parameter study The aim of the parameter study is to come up with relative statements regarding tendencies for the worst case for different parameters, possibly also the sensitivity. No absolute severity assessment will be conducted (in terms of what is hazardous/ acceptable). Also only the parameter impact directly on the A/C reaction can be evaluated as the automatic controller will only maintain the given reference state. In a critical situation real pilots would consider changing the A/C state like increase thrust (if applicable) or change the A/C configuration or for example the desired flight path angle. 3.3.1 Simulation reference scenario The reference flight state for the simulations is the stationary takeoff climb with maximum takeoff thrust, flaps extended and the following parameters (Table 3.2). The reference wake vortex is a 50 s B747-400 vortex (based on the MTOW). For the discussion of generator parameters see section 2.3.1. 19

parameter value mass 44 000 lb (19 958 kg) thrust maximum takeoff thrust gear gear up flap setting 5 indicated airspeed (VIAS) 128 kt (takeoff climb speed) flight path (climb) angle 10.6 pitch/ inclination angle 19 Table 3.2 Reference flight state and path for encounter A/C The encounter reference scenario comprises a pair of wake vortex lines with a vertical vortex inclination angle of 20. The nominal lateral distance of the reference flight track to the closer vortex is 17.7 m (Figure 3.1). Significant control activity is required to maintain reference flight state and path (Figure 3.2). This reference scenario is chosen as a compromise between very weak wake encounters (with no significant A/C reaction) and extremely strong encounters (with chaotic A/C reaction) which both leave no room for interpretation of the results/ parameter influence. Generally speaking lower encounter distances will result in more severe wake encounters. Figure 3.1 Wake vortex encounter reference scenario with flight path, vortex lines and reference flight track, side view and top view 20

Figure 3.2 Wake vortex encounter reference scenario parameter time histories: airspeed, bank angle, vertical load factor (normalised by the gravitational acceleration) and normalised control commands (roll, pitch, yaw normalised by maximum control power) 3.3.2 Parameter variation In order to isolate the influence of a specific parameter an independent variation is required. This it not always possible due to coupling between the different parameters especially with regard to realistic parameter combinations. Angle of attack and pitch are in any case dependant parameters, see section 2.2.1, so no direct parameter variation is carried out. The flight situation (turn/ climb/ acceleration) is linked to the flight path (climb) angle γ (for climb/ acceleration) or the bank angle Φ (turn) and is covered here by the variation of these parameters. Independent parameters are the vertical and horizontal encounter angles (Ψ WVE and γ WVE ). The requirement for constant airspeed (in order to leave speed constant like other parameters) is contradicting to the definition of airspeed for the takeoff climb (typically stall speed plus 20%, with stall speed depending on weight and A/C configuration). In these cases the airspeed is maintained at stall speed plus 20% which means the absolute airspeed is not constant. However this reflects the more realistic case for takeoff. The following simulation results relate to the maximum absolute values for the deviations from the reference flight state for bank angle ΔΦ, lateral flight path deviation Δy and vertical load factor Δn z. Follower mass (speed/ gamma) A given follower mass results in a corresponding airspeed for takeoff climb (based on the stall speed for that mass) and flight path (climb) angle (with constant thrust setting). The vertical encounter angle is kept constant at γ WVE = 9.4 and the thrust at maximum takeoff thrust, which 21

results in maximum roll commands of 53% in any tested case for the encounter scenario. On the bank angle deviation there is no significant influence and according to the maximum lateral flight path deviation the worst case is low follower mass (with low speed/ high climb angle respectively), however the influence on the flight path deviation is not strong (Table 3.3). m [lb] V [kt] γ γ WVL ΔΦ Δy [m] 44000 128 10.6 20.0 7.14 6.8 42000 125 11.5 20.9 7.10 6.9 40000 122 12.5 21.9 7.08 7.1 38000 119 13.6 23.0 7.08 7.2 36000 116 14.8 24.2 7.10 7.4 Table 3.3 Follower mass (speed/ gamma) influence Follower flight path angle A given thrust setting (PLA power lever angle) results in a corresponding flight path (climb) angle for constant airspeed. The vertical encounter angle is kept constant at γ WVE = 9.4. Both bank angle and lateral flight path deviation exhibit a low sensitivity but a lower flight path angle constitutes the worst case (Table 3.4). It is assumed that the thrust setting does not directly significantly influence the A/C reaction. PLA [%] γ ΔΦ Δy [m] 100 10.6 7.1 6.8 89 7.8 7.3 7.3 78 4.5 7.5 7.9 Table 3.4 Follower flight path angle influence Follower speed Increased follower airspeed with constant thrust and mass results in very low variation of the flight path angle. Hence for a constant vertical vortex inclination of γ WVL = 20.0 the vertical encounter angle is approximately constant. Both bank angle and horizontal flight path deviation exhibit a worst case tendency for lower airspeeds (Table 3.5). 22

V [kt] γ ΔΦ Δy [m] 128 10.58 7.1 6.8 135 10.65 6.6 5.7 140 10.67 6.3 5.1 145 10.66 6.0 4.6 150 10.62 5.8 4.1 Table 3.5 Follower speed influence Follower flaps (speed) Extended flaps result in a lower stall speed and consequently in a lower takeoff climb speed. Angle of attack and flight path angle remain approximately constant. Airspeed is always set to stall speed + 20%, mass is constant and the vertical vortex inclination is γ WVL = 20.0 (vertical encounter angle approximately constant). Flaps more extended (with lower airspeed) constitutes the worst case according to the horizontal flight path deviation and to a lesser extent to the bank angle (Table 3.6). flaps α V [kt] γ ΔΦ Δy [m] 0 8.62 134 10.98 7.07 6.4 5 8.55 128 10.58 7.14 6.8 14 7.80 121 10.51 7.74 8.7 Table 3.6 Follower flaps (speed) influence Bank angle Flight with a bank angle results in a loss of flight performance and hence lower flight path angles. Airspeed is kept constant at V = 128 kt and the vertical encounter angle at γ WVE = 9.4. The simulated wake vortex encounter scenario with a follower A/C bank angle is the same as the reference scenario (Figure 3.1), i.e. no turn but straight climb with a bank angle. The straight climb with bank angle is chosen (opposed to a turn) for the parameter variation for sake of simplicity as this study constitutes a preliminary stage within the CREDOS project and final more detailed investigations are yet to be conducted. The worst case according to the vertical load factor n z and the horizontal flight path deviation (the bank angle deviations being less significant) is a high bank angle (Table 3.7). 23

Φ γ WVL γ ΔΦ Δy [m] Δn z [-] 0.0 20,0 10.6 7.2 6.8 0.13 2.5 19,8 10.4 6.9 6.9 0.12 5.0 19,2 9.8 6.6 6.8 0.13 7.5 18,3 8.9 6.3 6.4 0.16 10.0 17,2 7.8 5.9 8.6 0.19 Table 3.7 Bank angle influence Vertical encounter angle The vertical encounter angle γ WVE is not coupled to any other A/C or encounter parameter as it can be adjusted by the vertical vortex inclination. Hence follower mass, flight path angle and speed can be kept constant. The bank angle does not exhibit a clear tendency but negative encounter angles with a low absolute value result in high bank angles. And according to the horizontal flight path deviation the worst case tendency is clearly also towards a low absolute value for γ WVE (Table 3.8). Note that (the absolute value of) vertical encounter angles cannot be chosen too small in order to prevent chaotic A/C reaction. γ WVL γ WVE ΔΦ Δy [m] 20 9.4 7.14 6.8 17 6.4 7.05 13.2 16 5.4 6.81 16.4 14 3.4 5.37 20.1 7-3.6 53.0 62.6 6-4.6 48.1 71.3 5-5.6 7.2 19.8 4-6.6 6.6 11.1 3-7.6 7.9 8.2 2-8.6 7.6 6.6 Table 3.8 Vertical encounter angle influence Horizontal encounter angle The horizontal encounter angle Ψ WVE is not coupled to any other A/C or encounter parameter as it can be adjusted by the horizontal vortex orientation. Hence follower mass, flight path angle and speed can be kept constant. In this case the aircraft is not passing the wake vortex laterally but flying above the vortex pair. Both vertical vortex inclination and follower flight path inclination are equal γ WVL = γ = 10.6. The nominal vertical distance of the reference flight track to the vortices is chosen to be 20 m (above the vortex pair) again as a compromise between weak and 24

strong encounter reaction. The flight path of the encountering A/C for a horizontal wake vortex encounter angle Ψ WVE = 10 is shown in Figure 3.3. (Note that the horizontal and vertical angles are not to scale due to axis scaling.) Figure 3.3 Wake vortex encounter scenario with 10 horizontal wake vortex encounter angle (flight path, vortex lines and reference flight track, side view and top view, note: horizontal and vertical angles are not to scale due to axis scaling) The worst case is a low Ψ WVE exhibited both by the bank angle and the vertical load factor (Table 3.9). 25

Ψ WVE ΔΦ Δn z [-] 5 30.6 0.38 10 8.7 0.29 15 4.2 0.16 20 4.4 0.17 25 4.1 0.17 30 3.6 0.18 35 3.2 0.18 40 3.1 0.18 45 3.1 0.17 50 3.1 0.16 55 3.1 0.15 65 3.1 0.12 75 3.2 0.09 90 3.3 0.05 Table 3.9 Horizontal encounter angle influence 26

3.3.3 Offline simulation summary Table 3.10 summarises the offline simulation parameter variation results für the investigated parameters. The possible range is listed which is not necessarily the range covered by the simulations carried out. The worst case (tendency) is highlighted and the sensitivity and coupling with other parameters is noted, if applicable. parameter range worst case for direct impact follower flight path angle remarks 0-15 (20 ) lower not very sensitive parameter, coupled with thrust setting for constant airspeed, mass and configuration follower bank angle 0 (turn up to 30 ) high (30 ) coupled with flight path angle follower flap setting vertical encounter angle horizontal encounter angle follower mass follower airspeed in/ maximum extended extended -10 to 15 low absolute value 0-90 (due to symmetry) MTOW and smaller low low not very sensitive parameter, coupled with speed (stall speed) no coupling no coupling not very sensitive parameter, coupled with speed (stall speed) and flight path angle at least takeoff low very low coupling with flight path climb speed V 2 angle Table 3.10 Parameter overview encounter offline simulation study 27

4 Conclusion Both the results of the survey of wake vortex encounter scenario parameters and the offline simulation parameter variation are summarised in the following overview (Table 4.1). All parameters dealt with are listed with the corresponding range and a recommendation/ worst case tendency is given respectively. As stated in section 3.3.2 the parameter variation is only evaluating the isolated influence of a parameter (where possible). Hence no conclusion can be drawn for combinations of parameters. 28

parameter range recommendation/ worst case reference section A/C category/ combination heavy and medium medium behind heavy 2.1.1, 2.1.2 takeoff phase (including A/C configuration and dedicated altitude range) flight situation (A/C flight state) lift-off, climb, acceleration, climb to en route level turn/ climb/ acceleration climb, acceleration, climb to en route level climb (turn, acceleration) 2.1.4 2.2.1 follower flight path angle 0-15 (20 ) lower 2.2.1, 3.3 follower bank angle mostly around 0 (turn up to 30 ) most relevant case 0 worst case 30 2.2.1, 3.3 follower flap setting in/ maximum extended extended 2.1.4, 3.3 follower landing gear retracted/ extended relevant case: retracted (extended worst case) 2.3.2 follower pitch angle 15 and higher N/A 2.2.1 follower angle of attack around 8 to 10 N/A 2.2.1 follower angle of sideslip small 0 2.2.1 vertical encounter angle -10 to 15 low absolute value 2.2.2, 3.3 horizontal encounter angle encounter distance 0-90 (due to symmetry) several hundred metres and less low 2.2.2, 3.3 low 2.2.3 encounter altitude up to 3000 ft AGL low 2.2.4 generator A/C type follower A/C type heavy (possibly medium) medium (possibly heavy) B744, B772, A343, A333, A306, A310 A320, B733, CRJ7, E145, AT45, SF34 2.3.1 2.3.2 follower mass MTOW and smaller low 2.3.2, 3.3 follower airspeed wind at least takeoff climb low 2.3.2, 3.3 speed V 2 no wind moderate crosswind turbulence no turbulence medium turbulence temperature stratification no stratification medium stratification no wind moderate crosswind 2.4.1 no turbulence 2.4.2 no stratification 2.4.3 visual conditions VMC/ IMC IMC 2.4.4 simulator cockpit motion on/ off on 2.6 Table 4.1 Parameter overview for takeoff wake vortex encounter scenarios 29

5 References [1] Fuhrmann, M., Luckner, R. (Airbus), Verbeek, M.J. (NLR), S-Wake - Flight Simulations Test Plan, S-Wake TN-321, 2001 [2] Bonnier, M., Dowdall, R., Harvey, A., Lane, R., Treve, V., Vidal, A., Initial Concept of Operation (CREDOS D4-1 B.7), Eurocontrol, 2007 [3] Höhne, G. (Airbus), Reinke, A. (TU Berlin), Verbeek, M.J. (NLR), S-Wake - Wake Vortex Encounter Flight Simulation: Metrics, Hazard Criteria, and Influence of Cockpit Motions, S- Wake TN-320, 2002 [4] Galpin, D., Pugh, C. (NATS - National Air Traffic Services Ltd), Turp, D. (UK Met Office), EuroBen - European Wake Vortex Mitigation Benefits Study, Work Package 1 Deliverable: Identification of WV Constrained Airports, TRSD03/2005/WP1, National Air Traffic Services Ltd, Corporate Technical Centre, June 2005 [5] Galpin, D., Pugh, C. (NATS - National Air Traffic Services Ltd), Turp, D. (UK Met Office), EuroBen - European Wake Vortex Mitigation Benefits Study, Work Package 2 Deliverable: Categorisation of WV Constrained Airports, Met Office and National Air Traffic Services Ltd, August 2005 [6] DFS internet site http://www.dfs.de [7] International Civil Aviation Organization (ICAO), Doc 4444-ATM/501, Procedures for Air Navigation Services - Air Traffic Management, 14th Edition 2001 [8] König, R., Macke, O., Kreth, S., Design and evaluation of short-term realizable new noise abatement flight procedures, Internoise 2006, 3-6 December 2006, Honululu, Hawaii, USA [9] A320 AOM Airplane Operations Manual [10] AIP Germany (Aeronautical Information Publication) [11] Treve, V., Wennerberg, A., Vidal, A., Harvey, A., Morvant, J., Initial Concept of Operation (CREDOS D4-1) Draft v.0n, Eurocontrol, 2007 [12] N.N. DFS, EDDF-2 Data Collection Campaign Report (CREDOS D2-4) preliminary information [13] Schwarz, C., Hahn, K.-U., Full-flight simulator study for wake vortex hazard area investigation, AST Aerospace Science and Technology, 10 (2), Elsevier, p. 136 143, DOI 10.1016/j.ast.2005.09.005, March 2006 [14] Schwarz, C., Hahn, K.-U., Gefährdung beim Einfliegen in Wirbelschleppen (Hazard of Wake Vortex Encounters), German Aerospace Congress, Munich, 17.-20. November 2003, DGLR-JT2003-242, DGLR yearbook 2003. Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt, 2003 [15] Hahn, K.-U., Coping with Wake Vortex, 23rd International Congress of Aeronautical Sciences, Toronto (Canada), 8-13 September 2002, ICAS, 23rd International Congress of Aeronautical Sciences Proceedings, p. 732.1-732.14, 2002 [16] Jategaonkar, R., Identification of the Aerodynamic Model of the DLR Research Aircraft ATTAS from Flight Test Data, DLR Forschungsbericht (research report) DLR-FB 90-40, 1990 [17] European Aviation Safety Agency, Certification Specifications for Large Aeroplanes CS-25, Amendment 2, 2 October 2006 30