All Documents | PDF version

Chapter 2

2.1 Emergency descent (memory item)
2.2 Windshear (memory item)
2.3 Unreliable airspeed (memory item)
2.4 Evacuation
2.5 Forced Landing (inc. Ditching)
2.6 Overweight landing
2.7 Engine failure in cruise
2.8 Bomb on board
2.9 Single engine circling
2.10 Stall recovery (memory item)
2.11 Computer reset
2.12 Abnormal V Alpha Prot
2.13 Overspeed Recovery
2.14 Volcanic Ash Encounter
2.15 Incapacitation

2.1 Emergency descent (memory item)

If an emergency descent is required, the Captain should consider taking control if not already pf. pf initiates the memory items by announcing “Emergency Descent.”

Don oxygen masks and establish communication.

pf then flies the emergency descent. Descent with autopilot and autothrust engaged is preferred. The configuration is thrust idle, full speed brake and maximum appropriate speed, taking into account possible structural damage.1 Target altitude is fl100 or mora if this is higher. If speed is low, allow speed to increase before deploying full speedbrake to prevent activation of the angle of attack protection. Landing gear may be used, but speed must be below VLO when it is extended and remain below VLE. If on an airway, consider turning 90° to the left.2

pm’s only memory action is to turn the seatbelt signs on; their primary task is to ensure that pf has promptly and correctly initiated the descent.

Once the memory actions are complete and the aircraft is descending, complete the cab pr excess cab alt ecam if it is available, then the qrh emer descent checklist, which covers much of the same ground as the ecam but adds a couple of useful items. This will lead pf to finesse the speed and altitude targets and inform atc of the descent; pm to set continuous ignition on the engines and set 7700 on the transponder. Both pilots then set their oxygen flows to the n position3 and, if cabin altitude will exceed 14,000ft, pm deploys the cabin oxygen masks. On easyJet aircraft, the cids/pram will automatically play a suitable pa, so it is not necessary for the flight crew to carry out the emer descent (pa) action.

Once level, restore the aircraft to a normal configuration. When safe to do so, advise cabin crew and passengers that it is safe to remove their masks. To deactivate the mask microphone and switch off the oxygen flow, close the oxygen mask stowage compartment and press the “press to reset” oxygen control slide.

[ eomb 3.80.2, qrh aep.misc, fcom aep.misc, fctm aep.misc ]

2.2 Windshear (memory item)

2.2.1 Takeoff roll

Windshear encountered on the takeoff roll is only detectable by significant airspeed fluctuations. It is possible that these fluctuations may cause V1 to occur significantly later in the takeoff roll then it should. In this case it falls to the Captain to make an assessment of whether sufficient runway remains to reject the takeoff, or whether getting airborne would be the better option. If the takeoff is to be continued, call “Windshear, toga” and apply toga power. Rotate at Vr or with sufficient runway remaining4 and follow srs orders. srs will maintain a minimum rate of climb, even if airspeed must be sacrificed.

2.2.2 Reactive

The reactive windshear detection system is a function of the facs. It only operates when below 1300ft ra with at least conf 1 selected. In the takeoff phase, it is inhibited until 3 seconds after lift off and in the landing phase it is inhibited below 50ft ra.

A warning is indicated by a red windshear flag on the pfd and a “Windshear, Windshear, Windshear” aural warning. Call “Windshear, toga” and apply toga power.

The autopilot can fly the escape manoeuvre as long as the required aoa is less than αprot. If the autopilot is not engaged, follow the srs orders on the fds. If the fds are not available, initially pitch up to 17.5°, then increase as required.

Do not change configuration until out of the windshear. Once clear of the windshear, clean up the aircraft: leveraging the go-around procedure is useful for this.

In severe windshear, it is possible that Alpha Floor protection will activate. As toga will already be selected, this will have no immediate effect. Once clear of the windshear, however, toga lock will be active. This, combined with the unusual aircraft configuration, leads to a significant threat of overspeed. The most natural way to disengage toga lock is to disengage the autothrust using the instinctive disconnect pb on the thrust levers then use manual thrust until the situation has sufficiently stabilised to re-engage the autothrust.

[ fcom pro.aep.surv ]

2.2.3 Predictive

Below 2300ft agl, the weather radar scans a 5nm radius 60° arc ahead of the aircraft for returns indicating potential windshear.

Alerts are categorised as advisory, caution or warning, in increasing order of severity. Severity is determined by range, position and phase of flight. Alerts are only provided when between 50ft and 1500ft, or on the ground when below 100kt.

All types of alert produce an indication of windshear position on the nd, provided that the nd range is set to 10nm. A message on the nd instructs the crew to change range to 10nm if not already set. A caution additionally gives an amber w/s ahead message on both pfds and an aural “Monitor Radar Display” warning. A warning additionally gives a red w/s ahead message on the pfds and either a “Windshear Ahead, Windshear Ahead” or “Go Around, Windshear Ahead” aural message.

If a warning occurs during the takeoff roll, reject the takeoff. If it occurs during initial climb, call “Windshear, toga”, apply toga thrust and follow srs orders. If it occurs during approach, fly a normal go-around. Configuration may be changed as long as the windshear is not entered.

If a caution occurs during approach, use conf 3 to optimise go-around climb gradient and consider increasing VAPP; up to a maximum of VLS+15 may be used.

If positive verification is made that no hazard exists and providing that the reactive windshear is serviceable the crew may downgrade the warning to a caution.

[ fctm ]

2.3 Unreliable airspeed (memory item)

The fmgcs normally reject erroneous adr data by isolating a single source that disagrees significantly with the other two sources. It is possible, however, that all adr data is simultaneously compromised or that the fmgcs reject a single remaining good adr because the other two sources are erroneous in a sufficiently similar way.

The first problem, then, is recognition of the unreliable airspeed condition. The key to this is an instinctive understanding of the relationship between power, attitude and airspeed in all flight phases. Other clues that may aid identification include fluctuations or rapid changes in displayed airspeed, abnormal behaviour of the automatics, high speed buffet, low aerodynamic noise and stall warnings.5

Even with the condition correctly diagnosed by the crew, until the unreliable adrs are disabled the aircraft systems will continue operating on the basis that the erroneous airspeed data is valid. Flight envelope protections may activate, leading to pitch inputs from the flight computers that cannot be overridden with the sidesticks. In this case, immediately switch off any two adrs; this causes the flight computers to revert to Alternate Law with no protections, and thus allows control of the aircraft to be regained.

In unaccelerated flight, for a given weight and wing configuration, a given angle of attack will result in a specific airspeed. In level flight, angle of attack is a fixed offset from attitude, the offset being dependent on wing geometry and thus configuration. The upshot of this is that, for any given combination of weight and configuration, indicated airspeed in unaccelerated flight may be inferred from attitude alone. This allows both for recovery of a good adr if it has been rejected, or, if necessary, for completion of the flight without airspeed data.

The initial goal, then, is to stabilise the aircraft in level flight at a safe altitude with an attitude for which the associated indicated airspeed is known. These attitude vs. airspeed relationships are tabulated at the start of the qrh nav unreliable speed indication procedure.

If the aircraft is not yet at a safe altitude, a climb must be manually flown without reference to airspeed data. The required power settings and attitudes for this are6:

Condition Thrust Pitch
Below Thrust Reduction Altitude toga 15°
Below fl100 clb 10°
Above fl100 clb 5°

If configured conf Full, select conf 3, otherwise flap/slat configuration should be maintained. The gear and speedbrake should be retracted.

Transitioning to or maintaining level flight requires an altitude reference, and it is possible that whatever caused the unreliable airspeed is also causing erroneous altimeter readings. The most likely root cause in this case would be blocked static ports, which should be fairly obvious as the altimeters will be frozen. In general, if all the adrs are giving sensible and near identical altimeter readings and they are in the ballpark of the gps altitude, reliable barometric altitude data can be assumed and the fpv, vsi and barometric altimeters are available. If barometric altitude is considered erroneous, a constant gps altitude can be flown. If gps altitude is also unavailable, the only option is to fly a constant attitude with a sensible power setting and wait for the resulting phugoid to damp; this can take a long time, results in significant altitude excursions and will only approximate level flight, but should be sufficient for diagnostic purposes.

The initial thrust setting needs to be chosen so that the resultant airspeed in level flight will be neither too close to the stall nor at risk of overspeeding any high lift surfaces that are still deployed. Either use a rule of thumb (e.g. 50% + fl/10 for two engines) or refer to the table at the start of the qrh nav unreliable speed indication procedure. Sensible adjustments will need to be made if there are extra sources of drag such as radome damage or stuck gear. Thrust should then be adjusted until level flight at the target attitude is achieved, at which point the actual indicated airspeed can be read from the table.

With the indicated airspeed now known independently of the adrs, any working adrs may be identified and used to complete the flight.

If no working adrs can be identified, the flight will need to be completed by inferring airspeed from attitude data. To achieve this, dependent on airframe, you may have access to, from most advanced to least, Digital Backup Speed (dbus), reversible Backup Speed Scale (buss), standard Backup Speed Scale or none of these.

dbus equipped aircraft provide a significant amount of assistance. Data from angle of attack sensors is augmented with load factor and GPS information to provide normal looking speed and altitude scales, albeit with degraded accuracy.7 These can be used to operate the aircraft in a relatively normal manner. In addition, the computed speed is available as a fourth source to the computers in identifying compromised adrs, and the computers are able to identify when an adr becomes reliable again and provide appropriate ecam procedures.

buss equipped aircraft provide more rudimentary assistance. The reversible version provides a button to turn the system on and off, whereas the older version requires all three adrs to be turned off to activate it. When activated, the speed scale is replaced with a display showing a bar with green, amber and red sections, the aim being to fly the green. It is basically a thinly disguised angle of attack display: the green band can represent a significant range of speeds, but does not change size, it should not be used for dynamic manoeuvres such as levelling off, and it’s insufficiently accurate to use above fl250. To clean up, accelerate until the speed is towards the top of the green before selecting the next configuration; to configure, reduce speed towards the bottom of the green before selecting next configuration.

When neither dbus nor buss are available, pitch vs. airspeed tables are provided in the qrh which can be used to complete the flight by inferring airspeed from attitude. The advice when using these tables is to only change one of altitude, airspeed or configuration at a time; this means that clean up and configuration are both done in level flight, stabilising at the attitude associated with each new configuration before selecting the next. A 3° ILS is highly recommended for the approach, as this is the angle that the final approach attitude (and hence airspeed) are predicated on.

[ qrh aep.nav, fcom pro.aep.nav, fctm pro.aep.nav ]

2.4 Evacuation

Evacuation should be carried out in accordance with the emergency evacuation checklist. The easyJet procedure is for cm1 to call for the checklist and then send a Mayday message to atc before commencing the checklist.

The first two items confirm the rto actions of stopping the aircraft, setting the parking brake and alerting the cabin crew. The next item confirms atc has been alerted.

The next four items prepare the aircraft for evacuation. If manual cabin pressure has been used, cm2 checks cabin diff is zero, and if necessary manually opens the outflow valve. cm2 then shuts the engines down with their master switches, and pushes all the fire buttons (including the apu). Confirmation is not required before carrying out these actions. In response to the next checklist item, “Agents”, cm1 decides if any extinguishing agents should be discharged and instructs cm2 to discharge them as required. Engine agent 2 will not be available. Agents should only be discharged if there are positive signs of fire.

Finally, order the evacuation. This is primarily done with the pa “Evacuate, unfasten your seat belts and get out”, with the evacuation alarm being triggered as a backup.

[ eomb 3.80.1, fcom pro.aer.misc, fctm pro.aer.misc ]

2.5 Forced Landing (inc. Ditching)

There are two scenarios where off field forced landing or ditching would be considered: either you have insufficient energy to reach a suitable airfield (e.g actual or impending fuel exhaustion, catastrophic failure of both engines), or you have insufficient time to do so (e.g. uncontained fire).

Ditching and off field forced landing without power are discussed in Section 10.1. Support for such landings is provided by the eng all engines failure ecam and the eng all eng fail qrh procedure.

Support for ditching and forced landing with power is provided by the qrh misc ditching and qrh misc forced landing procedures respectively. Necessarily implicit in these checklists is the assumption that the aircraft is fully serviceable, which is unlikely to be the case. There is also an assumption that plenty of time is available for extensive preparation of cabin and cockpit. It is highly likely that these checklists will need adapting to the situation.

The fundamentals are the same with or without power. Ditching will be gear up with a target pitch attitude of 11° and minimal vertical speed, landing parallel to the swell unless there are strong crosswinds, in which case an into wind landing is preferred. Forced landing will be gear down with the spoilers armed.

The aircraft should be depressurised for the landing, with the ditching pb pushed for the ditching case. For forced landings or for “without power” ditching this is achieved using ram air. For “with power” ditching, depressurise by turning off all bleeds, which provides a slightly more watertight hull.

The main difference between the with and without power cases is that max available slats and flaps are used in the former case, wheras conf 2 is mandated for the latter. Approach speeds must also be high enough to prevent rat stall (i.e. >140kt) if it is being relied upon. The combination of these factors means that much lower landing speeds can be achieved if power is available.

[ qrh aep.misc ]

2.6 Overweight landing

A landing can be made at any weight, providing sufficient landing distance is available. In general, automatic landings are only certified up to mlw, but the fcom specifies that, for the A319 only, autoland is available up to 69000kg in case of emergency.

The preferred landing configuration is conf Full, but lower settings may be used if required by qrh or ecam procedures. qrh misc overweight landing also specifies conf 3 if the aircraft weight exceeds the conf 3 go around limit; this will only ever be a factor for airfields with elevations above 1000ft.

Packs should be turned off to provide additional go around thrust.

If planned landing configuration is less than conf Full, use conf 1+F for go-around.

It is possible that S speed will be higher than VFE next for conf 2. In this case, a speed below VFE next should be selected until conf 2 is achieved, then managed speed can be re-engaged.

Airbus specifies that, in the final stages of the approach, speed should be reduced in order to cross the threshold at VLS. This implies manual thrust since use of autothrust requires a 5kt increment. Airbus also specifies that v/s at touchdown should be minimised, this being achieved by an increase in flare height. It is worth considering whether these requirements are compatible, particularly if dealing with gusty crosswinds etc. — if not, an option would be to add 5kt (or more) ΔVPilot in the landing performance calculation and use a standard landing technique.

Apply max reverse as soon as the main gear touches down. Maximum braking can be used after nosewheel touchdown.

After landing, switch on the brake fans and monitor brake temperatures carefully. If temperatures exceed 800°C, tyre deflation may occur.

[ qrh aer.misc, fcom pro.aer.misc, fctm pro.aer.misc ]

2.7 Engine failure in cruise

Engine out ceiling is highly dependent on weight; isa deviation also has a modest effect. It will generally lie between fl180 and fl350.

The first action will be to select both thrust levers to mct so as to allow the autothrust its full engine out range. If the n1 gauges indicate a thrust margin exists, then the aircraft is below engine out ceiling; descent may be appropriate to increase the available thrust margin, but there is no immediate threat. If, however, the n1 gauges indicate that the autothrust is commanding mct, and the speed is still decaying, then the aircraft is above engine out ceiling and prompt execution of a drift down procedure is required.

Drift down with autopilot engaged in op des is preferred. Engagement of this vertical mode normally results in the autothrust commanding idle thrust, which is not what is desired. Thus, having set the thrust lever to mct, the autothrust is disconnected. The prog page provides a rec max eo flight level to use as an altitude target.

If the speed decay is modest, it may be possible to alert atc before initiating the descent, but in-service events have shown that speed decay is often very rapid, requiring descent initiation to be prioritised.

Once drift down has been initiated, a decision needs to be made about speed.

If obstacles are a concern, the lowest drift down rate and highest ceiling are achieved at green dot. Airbus refers to drifting down at green dot as “Obstacle strategy”.

Flying at green dot reduces the chance of the fadecs automatically relighting the failed engine as the engine will be windmilling more slowly. Therefore, if obstacles are not a concern, M.78/300kt is flown, a speed that will always fall within the stabilized windmill engine relight envelope; Airbus refers to this as “Standard Strategy”.

If obstacles remain a problem, mct and green dot speed can be maintained to give a shallow climbing profile. Once obstacles are no longer a problem, descend to lrc ceiling (use v/s if <500 fpm descent rate), engage the autothrust and continue at lrc speed.

[ fctm pro.aep.eng.efdc ]

2.8 Bomb on board

The primary aim is to get the aircraft on the ground and evacuated asap.

The secondary aim is to prevent detonation of the device. This is achieved by preventing further increases in cabin altitude through the use of manual pressure control and by avoiding sharp manoeuvres and turbulence.

The tertiary aim is to minimise the effect of any explosion. This is achieved by reducing the diff to 1 psi. The method is to set cabin vertical speed to zero using manual pressurisation control, then descend to an altitude 2500ft above cabin altitude. As further descent is required, cabin vertical speed should be adjusted to maintain the 1 psi diff for as long as possible. Automatic pressure control is then reinstated on approach.

Low speeds reduce the damage from an explosion but increase the risk of a timed explosion occurring whilst airborne; a compromise needs to be found. The aircraft should be configured for landing as early as possible to avoid an explosion damaging landing systems.

In the cabin, procedures are laid down for assessing the risks of moving the device and for moving the device to the lrbl at door 2R.

[ qrh aer.80, fcom pro.aer.misc ]

2.9 Single engine circling

It may not be possible to fly level in the standard circling configuration of conf 3, gear down. This can be ascertained by checking the table in the qrh misc one engine inoperative – circling approach procedure.

If affected, plan a conf 3 landing and delay gear extension until level flight is no longer required; anticipate a l/g not down ecam warning below 750ft (which can be silenced with the emer canc pb) and a gpws “Too Low Gear” aural alert below 500ft ra.

[ qrh aep.misc ]

2.10 Stall recovery (memory item)

Aerofoil stall is always and only an angle of attack issue. It is not possible to directly prove an unstalled condition from attitude and airspeed data. The flight recorders from the December 2014 Air Asia accident recorded an angle of attack of 40° (i.e. around 25° greater than critical angle) with both pitch and roll zero and speeds up to 160kt. Importantly, it is perfectly possible to be fully stalled in the short term unreliable airspeed configurations described in Section 2.3. Identification of a fully stalled condition is thus largely dependent on identifying a high and uncontrollable descent rate that does not correlate with normal flight path expectations for the attitude and thrust applied.

To recover from a fully stalled condition, the angle of attack of the aerofoils must be reduced to below critical. The generic stall recovery is therefore simply to pitch the nose down sufficiently to break the stall and level the wings. In normal operations, the velocity vector of the aircraft is around 3° below the centerline of the aircraft (i.e. an attitude of around 3° is required to fly level). In a stalled condition, the velocity vector may be 40° or more below the centerline of the aircraft. Thus the amount of pitch down required to recover a fully stalled aircraft can be 30° or more.

In two recent Airbus accidents involving stalls, the lack of physical cross coupling of sidesticks was a major factor. If one pilot elects to hold full back sidestick, the aircraft cannot be recovered by the other pilot unless the takeover button is used. With all the alarms, it would be easy to miss “Dual Input” warnings, so always press the takeover button.

The aircraft’s thrust vector helps to accelerate the aircraft during the recovery, and increasing speed along the aircraft’s centerline acts to reduce the stalled angle of attack. Thus, while thrust is not a primary means of recovery, it does help. Unfortunately, Airbus have determined that due to the pitch couple associated with underslung engines, there may be insufficient longitudinal control authority to pitch the aircraft sufficiently to recover from a stall if toga is selected. It may therefore be necessary to initially reduce thrust to allow the primary recovery technique to be applied; this is extremely counterintuitive.

Once there are no longer any indications of the stall, smoothly recover from the dive, adjust thrust, check speedbrakes retracted and if appropriate (clean and below 20,000ft) deploy the slats by selecting conf 1. The load factor associated with an overly aggressive pull out can induce a secondary stall; on the flip side, once reattachment of the airflow occurs, drag rapidly diminishes and exceedance of high speed airframe limitations becomes a threat. A balance needs to be found.

If a stall warner sounds on takeoff it is likely to be spurious since you are almost certainly in normal law. The procedure in this case is essentially to initially assume unreliable airspeed and fly toga, 15°, wings level until it can be confirmed that the warning is spurious.

A stall warning may occur at high altitude to indicate that the aircraft is reaching αbuffet. In this case simply reduce the back pressure on the sidestick and/or reduce bank angle.

[ fcom pro.aer.misc ]

2.11 Computer reset

Abnormal computer behaviour can often be stopped by interrupting the power supply of the affected computer. This can be done either with cockpit controls or with circuit breakers. The general procedure is to interrupt the power supply, wait 3 seconds (5 seconds if a cb was used), restore the power, then wait another three seconds for the reset to complete. qrh aer.reset details the specific procedures for a variety of systems.

On the ground, almost all computers can be reset. moc can usually supply a reset procedure if nothing applicable is available in the qrh. The exceptions are the ecu and eiu while the associated engine is running and the bscu when the aircraft is not stopped.

In flight, only the computers listed in the qrh should be considered for reset.

[ qrh aer.system reset ]

2.12 Abnormal V Alpha Prot

If two or more angle of attack vanes become frozen at the same angle during climb, a Mach number will eventually be reached such that the erroneous angle of attack data indicates an incipient stall. When this happens, Normal Law high angle of attack protection will activate. The flight computers’ attempt to reduce angle of attack will not, however, be registered by the frozen vanes, leading to a continuous nose down pitch rate which cannot be overridden with sidestick inputs.

Indications of this condition are available from the αprot and αmax strips. If the αmax strip (solid red) completely hides the αprot strip (black and amber) or the αprot strip moves rapidly by more than 30kt during flight manoeuvres with ap on and speed brakes retracted, frozen angle of attack vanes should be suspected.

The solution is to force the flight computers into Alternate Law where the protection does not apply. This is most conveniently done by turning off any two adrs. Once in Alternate Law, the stall warning strip (red and black) becomes available. Since stall warning data also comes from the angle of attack vanes, erroneous presentation is likely.

2.13 Overspeed Recovery

In general the response to an overspeed should be to deploy the speedbrake and monitor the thrust reduction actioned by the autothrust. Disconnection of the autopilot will not normally be required. If autothrust is not in use, the thrust levers will need to be manually retarded.

It is possible that the autopilot will automatically disengage and high speed protection will activate, resulting in an automatic pitch up. In this case, smoothly adjust pitch attitude as required.

At high altitude, there is a threat of over-correction caused by the lethargic response of the speedbrake when commanded to stow. In the worst case, a descent may be required to recover speed. This threat can be mitigated by promptly cancelling the speedbrake as soon as the overspeed condition ceases.

[ fctm pro.aer.misc ]

2.14 Volcanic Ash Encounter

Volcanic ash clouds are usually extensive, so a 180° turn will achieve the quickest exit.

Air quality may be affected, so crew oxygen masks should be donned with 100% oxygen to exclude fumes. Passenger oxygen may also need to be deployed.

Probes may become blocked with ash, so be prepared to carry out the unreliable speed procedure.

Disconnect the autothrust to prevent excessive thrust variations.

To minimise the impact on the engines, if conditions permit thrust should be reduced. Turn on all anti-ice and set pack flow to high in order to increase bleed demand and thus increase engine stall margin. Wing anti-ice will need to be turned off again before attempting relight in case of flameout.

If engine egt limits are exceeded, consider a precautionary engine shutdown with restart once clear of volcanic ash. Engine acceleration may be very slow during restart. Since compressor and turbine blades may have been eroded, avoid sudden thrust changes.

Damage to the windshield may necessitate an autoland or landing with a sliding window open.

[ qrh aep.misc, fcom pro.aep.misc, fctm pro.aep.misc ]

2.15 Incapacitation

Take control, using the stick priority button if necessary. Contact cabin crew asap. They should strap the incapacitated pilot to his seat, move the seat back, then recline it. If there are two cabin crew available, the body can be moved. Medical help should be sought from passengers, and the presence of any type rated company pilots on board ascertained.

[ fctm pro.aep.misc ]

1According to Airbus, structural damage may be suspected if there has been a “loud bang” or there is a high cabin vertical speed. When limiting descent speed due to suspected structural damage, it is ias rather than Mach that is relevant.

2In UK airspace it is recommended to stay on the airway or descend on current track.

3There may be insufficient oxygen to cover the entire emergency descent profile if the oxygen masks are left set to 100%.

4 “Sufficient runway remaining” is actually Boeing advice – Airbus offers no guidance for the case where there is insufficient runway available to stop nor to rotate at normal speeds.

5Stall warning, available in alternate or direct law, is based on alpha probes, which will likely be giving valid data

6Airbus have recently (2022) clarified that if you transition through thrust reduction altitude you should adopt clb 10°, and if you transition through fl100 you should adopt clb 5°.

7Displayed airspeed is accurate to ±15kt. Displayed altitude is derived from GPS data which is both less accurate and fundamentally different from barometric altitude. The degraded accuracy is indicated by two amber lines through the last digit of the speed scale and the last two digits of the altitude.