In 2005 an Embraer 145 overshotrunway 27L at Hanover Airport.According to the German FederalBureau of Aircraft AccidentInvestigation Report EX006-0/05which can be read in full on theirweb site (http://www.bfu-web.de):

The weather forecast indicatedthunderstorms and heavy showersof rain. The initial ATIS indicated

a wind of 150/5 kt and a visibilityof 8 km in light rain. The weatherdeteriorated during the approach,and at the time of landing thewind was 100/5 kt (a 3 kt tailwind)with a visibility of 2000 m in heavyrain. The runway was wet andshowed patches of standingwater. Approximately 15 minutesafter the occurrence the brakingcoefficients for the first third of therunway were between 0.4 and 0.7

and for the remainder of therunway were between 0.6 and 0.7(good); however, it had stoppedraining by this time, and so thesecoefficients were not consideredby the Investigation to be relevantto the accident.

A Boeing 737 that landed priorto the Embraer subsequentlyreported to the Investigation thatthe braking action was medium –but this was not reported at the

time to ATC. According to crewstatements the aircraft crossedthe threshold at 140 kt (Vref 131kt) and touched down in thetouchdown zone. The FDRshowed a threshold crossingheight of 62 feet as as opposed tothe correct 50 feet. The crew didnot experience any significantdeceleration of the aircraft even

pwk_0389 Produced by the Graphics and Media Team, BAE Systems Regional Aircraft, PrestwickTel: 01292 675825 Email: ragraphics@baesystems.com

© BAE Systems 2008. All rights reserved. The information contained within this newsletter is of the nature of general comment only and doesnot constitute a warranty or condition of any contract. BAE Systems accepts no liability for the misuse or misinterpretation of such information.

CONTINUED ON PAGE 02

Flight safety news for BAE Systems Regional Aircraft customers • November 2008

Spotlight> More info: see page 06

JETSETS

Don’t eat into yourmarginsHow to cope with wet runwaysand hydroplaning Above: wet runways can affect braking

An incident

Recent aviation articles discussing smoke/fumes occurrences and the emergencyprocedures for cabin high altitude warninghave highlighted a reluctance from certaincrews to don their masks and goggles. We ask the question...why?

Icing induced loss of controlThree incidents highlight thisproblem on Turboprop aircraftTurn to page 08

A report on fan blade icingand how to stay clear of it

See page 05

Avoid severe icing

JETSETS02 November 2008

Welcome>

JEt and Turboprop Support,Engineering,Training andSafety.

Welcome to the autumn (forthose of us in the northernhemisphere) issue of JETSETS.Here in Scotland summerseems to have passed us byand so, with winter nearly uponus, we look at some of theissues that face you, both in theair and on the ground. Althoughthe articles are slanted towardseither jet or turbo prop, I wouldlike to think that there arelessons for all of us in eacharticle.

Please also take the time tolook on our web site(www.regional-services.com) fora copy of Think Ice! 2007 it iswell worth reading if you’ve notseen it before, and even ifyou’ve read it there are a fewgood reminders for the comingwinter season.

Recent incidents (not only onour fleets) have shown thatthere is a marked reluctanceamong flight deck crews to donoxygen masks in the face ofsmoke or cabin high altitudewarnings, and this is discussedmore fully in a later article.

For the jet operators there isan article on fan blade icing,and a discussion onhydroplaning which is pertinentto all aircraft.

though the ground spoilers hadautomatically deployed aftertouchdown.

Both pilots attempted to brake,and shortly before the aircraftovershot the runway the pilot incommand activated the parkingbrake which is also theemergency brake. This resultedin deactivation of the anti-skidsystem, the wheels locked upand the ground spoilers retracted(because the wheels had locked).The aircraft came to a rest about160 m beyond the end of therunway, and suffered only minordamage.

Eyewitnesses stated that theaircraft touched down about 1000 m after the threshold, andthis was later refined from FDRdata to 849 m. All four tyresshowed traces of rubber reversionhydroplaning (Figure 1), and hadleft about 400 m long bright traceson the runway which weredefinitely caused by rubberreversion hydroplaning (runwaymarks left following rubberreversion hydroplaning look likethey might originate from steamblasting (Figure 2)). Furthermore,melted away rubber was found onthe runway (Figure 3).

The Investigation concludedthat based on the slowdeceleration after touchdown itwas highly likely that dynamicaquaplaning occurred in themiddle portion of the runwayfollowed by rubber reversionhydroplaning which occurredwhen the emergency brake wasactivated, the anti-skid deactivatedand the tyres locked. The aircrafttechnical log did not show anyirregularities regarding the brakesor tyres, and the tyres wereinflated correctly.

The worst case RLW calculation,which the Investigation thoughtmost likely to apply to this landing,gave a stopping distance with150 m runway remaining – therewas a longer runway availablebut the crew chose 27L because

FROM PAGE 01For the turboprops there is anarticle on Icing Induced loss ofControl.

Although not covered by articlesin this JETSETS, I would like todiscuss 2 further safety issues.

One of our turboprop operatorsrecently had an incident duringwhich a wheels-up approach wascarried out, and the bottom of thefuselage and one prop weredamaged. The captain, who wasPF, had called for theundercarriage, but had notchecked the undercarriageindicators after calling for itsselection because this check wasnot required by his company’sSOPs. The philosophy behind ourchecklists is based on theassumption that vital checks arecarried out as challenge andresponse, and that both crewmembers confirm the correctresult of the action. What are yourcompany SOPs? I personallyalways cross check suchselections during the flight. I lookat the gear, flaps and engines onshort finals to assure myself thatthey are in the correct positionsfor landing, and also check flapsetc., prior to increasing power ontake-off.

Another turboprop operatorsuffered a control restrictionduring the climb out which is likelyto have been caused by ice, andthere were four jet incidents overone weekend that were put down

to hail, after the aircraft hadbeen de-iced, causing elevatorproblems. Reading the incidentreports, I was reminded of anincident during which a cabincrew member was injured whenthe controls suddenly freed afterthe crew had applied a force tothe control column. Everyoneand everything should besecured before making anyvigorous attempts to free anyjammed control as it may freesuddenly leading to an abruptchange in the aircraft’s attitude.If the jam persists, rememberthat the secondary effects of thecontrols might help you to fly theaircraft. Thrust can be used tochange the pitch attitude, andrudder can provide roll control.The trim might still provide somecontrol if the elevator is jammed.Always make any inputs smallinitially until the effect isunderstood, and do not chaseairspeed altitude or attitude aslarge movements could result inpilot induced oscillations.I hope that you find this issueinteresting, and would welcomeany feedback. I would speciallywelcome any contributions fromany of you.

Colin Wilcock Product IntegrityPilot

How to cope with wetrunways and hydroplaning

Please contact:The Flight Operations SupportDepartmentBAE Systems Regional AircraftPrestwick International Airport,Ayrshire, Scotland, KA9 2RW.Tel: +44 (0)1292 675225 Fax: +44 (0)1292 675432E-mail: rafltops@baesystems.com

JETSETS is also available on ourweb site www.regional-services.com. Here it is split up to makeit easier for you to download anindividual article.

JETSETS 03November 2008

fluid film, and ensure wheel spin-up and spoiler activation.

• Immediately after touchdown, check the ground spoiler automatic deployment when thrust levers are reduced to IDLE.

• Lower the nose wheel positively, with forward pressure to assist traction and directional stability.

• Apply brakes with moderate-to -firm pressure, smoothly and symmetrically, and let the anti-skid do its job.

• If no braking action is felt, hydroplaning is probably occurring. Do not apply Emergency/Parking brake, as itwill cause the spoilers to close and cut the anti-skid protection.Maintain runway centre line and keep braking until airplaneis decelerated.

The crew do not appear to havefollowed some of the above advice.BAE Systems have received reportsof incidents occuring to BAe 146/Avro RJ aircraft where the anti-skidsystem was not given a chance toadapt to the conditions becauseyellow, green and then emergency

yellow were selected in quicksuccession. This does not allow theadaptive feature of the anti-skid tooperate correctly. You should notefrom the above that locked wheelsprovide very little stoppingassistance!

As can be seen, there were atleast five contributing factors. Thisincident may have been inevitablebecause the LDR following the longtouchdown might have beeninsufficient, but if the other factorshad not been present it is possiblethat an overrun would not haveoccurred. As stated at the end ofthe JETSETS article in the previousissue dated February 2008:

Don’t eat into your margins, and ifin doubt go around.

Lesson 2 - HydroplaningThe other lesson from this accidenthas to do with hydroplaning.Control of your aircraft on theground depends on the contactbetween the tyres and the surface,and on the friction provided by thatsurface. Whilst the above accident

of construction work on a taxiwayand the shorter distance to theterminal.

Lessons that can be learntI would like to use this incident toillustrate 2 lessons: firstly thefactors leading to an overrun, andsecondly to discuss hydroplaning.

Lesson 1 - The overrunThe last issue of JETSETS alsocovered overruns and we foundthat all overruns had more thanone contributory factor. From myreading of the report this overrun is no different in that there wereseveral factors as listed below:

1. Landed slightly long. The idealtouchdown point is 300 m from thethreshold - in this instance thetouchdown was at 849 m. Thisshortened the LDA.2. Landed slightly fast - but onlyby a few knots.3. Runway condition. The runwaywas wet, or even flooded, but it isnot clear whether the crew wereaware of this. The crew were notinformed of the runway conditionby ATC.4. Wind. There was a slight tailwind.

5. Incorrect braking technique.The Aircraft Operation Manual Part B included: When hydroplaningoccurs, it causes a substantial loss of tire friction and wheel spin-up may not occur.• The approach must be flown

with the target of minimising the landing distance.

• The approach must be stabilized, and landing on centre line in the touchdown zone.

• The touchdown should be firm to penetrate the contaminating

Above: an Embraer 145 similar to the aircraft that suffered an runway overshoot

Above: figure 1 - marks of rubber rever-sion hydroplaning on the tyre

Above: figure - marks of hydroplaningon the runway

Above: figure 3 - rubber dumped fromtyre

Conclusion

JETSETS04 November 2008

highlighted the braking problemscaused by hydroplaning, directionalcontrol can be equally affected,especially in strong crosswindsbefore the rudder becomes fullyeffective. As a tyre rolls along therunway it is constantly squeezingwater from the tread. Thissqueezing action generatespressure within the water that canlift part of the tyre off the runwayand reduce the amount of frictionthat the tyre can develop. Thissqueezing action is calledhydroplaning. Please note that icyrunways are a totally separateissue.

Three basic modes ofhydroplaning have been identified:dynamic, viscous and revertedrubber. Dynamic hydroplaning, which isalso called aquaplaning, is relatedto speed and tyre pressure. Highspeed and low tyre pressure arethe worst combination, giving thelowest aquaplaning speeds. Duringtotal dynamic hydroplaning the tyrelifts off the surface and rides on awedge of water like a water ski. Youhave probably experienced thiswhen driving through large puddleson the road, and felt the steering

lighten. Dynamic hydroplaning willoccur at speeds above 9 times thesquare root of your tyre pressure (inpounds per square inch). Forinstance the 146/RJ with tyres at 155psi would hydroplane at speedsabove 112 kt, the ATP (86 psi)above 84 kt, and the Jetstream 41(100 psi) above 90 kt Whendynamic hydroplaning occurs itmay lift the wheel off the runwayand prevent spin up or, if anti-skid isnot being used, cause the wheel tostop spinning. Once started thehydroplaning could continue tomuch lower speeds.Viscous hydroplaning occurs onall wet runways and describes thenormal slipperiness or lubricatingaction of the water. Viscoushydroplaning reduces the friction,but not to such an extent the spin upon touch down is prevented. Themost positive way to preventviscous hydroplaning is to providetexture to the surface – hencegrooved runways.Reverted rubber hydroplaning issimilar to viscous hydroplaning inthat it occurs with a thin film ofwater and a smooth runwaysurface. It often follows dynamic orviscous hydroplaning where the

wheels are locked. The lockedwheel creates enough heat tovaporise the underlying water filmthus forming a cushion of steamthat eliminates tyre to surfacecontact, and begins to revert therubber, on a portion of the tyre,back to its uncured state. Oncestarted, reverted rubberhydroplaning will persist down tovery low speeds – virtually until theaircraft comes to a stop. During theskid there is no steering ability.Indications of reverted rubberhydroplaning are distinctive whitemarks on the runway, and a patchof reverted rubber similar to theuncured state on the tyre. It is alsolikely that melted away rubber willbe found on the runway.

The increase in stoppingdistance as a result of hydroplaningis impossible to predict accurately,but it has been estimated toincrease it by as much as 700%.The reduced braking action on awet runway may prevent the aircraftfrom decelerating normally with theanti-skid system operational. But theanti-skid system will provideoptimum braking – switching it offwill most likely lead to wheel lockup and burst tyres.

I hope that this article has

provided some food for thought.

Our northern winter will soon be

here with wet weather and we

will be faced with conditions

similar, and worse to those

described above. Of course

other areas in the world suffer

similar conditions in monsoons

and heavy rain. I can do no

better than reiterate the advice

given in the previous issue:

• A good landing starts with a

good approach.

• Think before accepting a

downwind component.

• Land in the right place at the

right speed.

• If in doubt go around.

• Trust the systems and brake

for effect, not comfort.

Above: wet runway contamination can cause hydroplaning

Don’t eat into your margins

JETSETS 05November 2008

Report

final the ice seemed to havedissipated and we were able touse both engines in a normal way.

The ground report showed asurface temperature at the time of-1 deg C in light snow. The crewreported that the aircraftwing/cowl ant-ice had beenswitched on long before enteringcloud. With the first sign ofvibrations the crew also noticedsignificant elevator pulsing (4-5

both convinced that the vibrationswere caused by fan blade icingand not by damage, so wedecided to leave No 1 enginerunning at idle thrust instead ofshutting it down. During theapproach the vibrations reducedand the instruments showednormal values. We thenincreased thrust on the engine,but the VIB indication went upinto the ‘yellow’ range. On short

From the European RegionsAirline Association Air SafetyWork Group Safety TargetedAwareness Report 009 V1 -November 2007: The followingincident was reported by a CRJcrew on an early morning winterflight.

While holding in VMC we werecleared to descend to FL 150 and were IMC. Shortly afterentering the clouds aircraft icingbegan to develop rapidly, beingclearly visible at the wiper bladeposts and cockpit window frames.Engine vibration was feltimmediately. As we expected a 40minute hold we asked to climb toVMC, but initially we were onlycleared to FL160. Shortly after, theengine vibration increasedrapidly, and after a few moments

ice particles separated fromengine No 1, audible all over theaircraft. A more urgent request forclimb was successful but at thesame time there was a loud noise,the aircraft was vibrating heavilyand the VIB indication for No 1engine showed exceedance. Wereduced thrust on No 1 to idle anddecided to request priority.Vibrations also started on No 2and were increasing. We were

Fan blade icingAbove: Canadair Regional Jet (CRJ)

• First symptoms of icing can be detected by observing ice build-up on small objects like wiper blade posts.

• Sudden icing of the window frames is a sign of severe icing.

• Engine fan blade icing is often announced with engine vibrations.

cm fore and aft). There was nomechanical damage or fanimbalance on the engines, andthe incident was put down to fanblade icing.

Flying in severe icing is to beavoided where ever possible.When encountering suchconditions steps mustimmediately be made to moveout of that environment

Engine fan blade icing

• Due to ice shedding, the vibrations may be accompanied by sharp metallic noise as ice impacts the inside of the nacelle.

• Check the procedure for a an ENG VIBN caption

• In icing conditions, increase fan RPM in order to shed the ice. A higher fan speed reduces the risk of blade icing.

• Fan blade icing directly reduces the aerodynamic effectivity of the fan rotor, leading to significant performance degradation although the gas generator is at full thrust.

• Clean fan blades are less likely to pick up ice.

JETSETS06 November 2008

Why the reluctancy?The use of oxygen masks in the event of smokeor pressurisation failure

Above: smoke in the cabin, first priority is to ensure against crew incapacitation

Reading back numbers of the UKCAA Flight Operations DepartmentCommunications - I know, I shouldget out more! - I came across acommunication that discussedsmoke/fumes occurrences and theemergency procedures for cabinhigh altitude warning. Thesearticles struck me as timelybecause recently we have had acouple of incidents during whichthe crew seemed reluctant to dontheir masks and goggles.

Reviewing the reports has mademe ask: why are crew so reluctantto don a mask?

In one recent smoke/fumesincident the crew even had atherapeutic bottle and mask broughtup for them to use if they felt anyworse in spite of never having usedthe cockpit supply. There are twooccasions where it would beprudent, and possibly evenessential for life preservation, to useoxygen, and they are discussed inthis article. You have two sources ofoxygen available: there aretherapeutic sets located round theaircraft for use by one and all and,much more readily to hand, thereare the mask and goggles besideyour seat for your exclusive use. Iwould like to discuss the use of thelatter in the event of smoke orpressurisation failure.

We all need oxygen to survive.I am sure that we all remember,however vaguely, the reasons forneeding supplemental oxygen atheight, but I will cover them againbriefly. The atmosphere contains21% oxygen, 78% nitrogen and sometrace elements. We all need oxygento live, and the amount that we getat sea level is adequate for us. Asthe atmospheric pressure reduceswith altitude the proportion of gasesremains fairly constant, but the

amount of oxygen that we can takein with each breath reducesmarkedly with decreasingpressure. The amount of oxygen inthe air is usually given by thepartial pressure of the gas. At apressure of 1000hPa there will be210 hPa of oxygen (21% of 1000). At500 hPa (equivalent to about20,000ft) there will only be 105 hPaof oxygen and at 300 hPa(equivalent to about 30,000ft) therewill only be 63 hPa. So as you cansee there will not be enoughoxygen to keep us going at the

higher altitudes (your aircraftoxygen system should be able tosupply enough gas to maintain apartial pressure of around 122 hPa).Figure 1 gives times of usefulconsciousness, and is covered inmore depth later. Do rememberthat the symptoms ofhyperventilation, or over breathing,which is usually associated withintense stress or anxiety, can bevery similar to those of hypoxia. Smoke/fumesThe QRH instructs you to put onthe oxygen mask (with oxygen setto 100% - you do check that 100% isset on the first flight of the day don’tyou?) and goggles as part of thememory actions. This first action isto ensure that incapacitation does

not occur although somemanufacturers do allow some drillsbefore going onto oxygen. Thebasic premise is that the crew arenot best placed to diagnose thecontaminant/source of the smoke,and the safety of the aircraft mustbe paramount and so the safestoption has been chosen. Once on100% oxygen the crew areprotected from breathing anyimpurities, and should find that theirperformance is not impaired.Wearing the masks is lesscomfortable, and the mics are

noisy. However, many pilots in themilitary get used to wearing a maskfor every flight whereas for you it

will only be for the remainder of thecurrent flight. The smoke gogglesalso provide important protectionas some contaminants can causeheavy eye watering leading todifficulties in seeing out of thecockpit, and so if separate gogglesare provided they must also bedonned. Having donned the masksand goggles there will be no cue asto whether to remove them beforecoming to a safe halt, and so theyshould be worn for the remainder ofthe flight. Don’t be tempted to takethem off if the smoke seems to havedisappeared as there may still besome contaminant in the cockpit.

As we go to press a recent UKAir Accident Investigation BranchBulletin 04/2008(http://www.aaib.dft.gov.uk)contains a ‘smoke’ incident thatoccurred to an Embraer 145 duringa scheduled operation. The first

Above: Embraer 145

The table opposite (copied fromWikipedia, but there are manysimilar tables published) givesindications of times of usefulconsciousness without oxygen:

This table was probablycalculated on a young fit personsitting down. Someone walkingabout would have less time (cabincrew?), and other factors such asstress, fitness, smoking and fatiguewill also affect the time available toyou. Certainly some publishedtables give less time at the heightsabove 22,000 feet.

How long will it take you to donyour mask? For our turbo propsthere is a bit of leeway because theycruise at lower levels, but for the jetstime may be of the essence. Testshave shown that pilots often take upto 15 seconds to don a mask. If yougo unconscious you will probablyregain consciousness within 30seconds of oxygen being restored –but someone will need to be aroundto put your mask on for you. Thisshould be your first action, you mustnot delay putting the mask onbecause the cabin altitude is onlyabout 20,000 feet (as has happened)

couple of reports recently where thepacks were not selected ON and theaircraft continued the climb until thecrew received a warning to indicatethat they were not pressurised.

Not all of us are able toexperience hypoxia at first hand in apressure chamber or for real (manymight not want to!), and so we haveto make do with reading books orground training. Those of us whohave been in a chamber willremember how insidious the onsetof unconsciousness was. Mostly wewere given a simple task such aswriting down a multiplication tableor writing our name; when I wentthrough a chamber run, and my

oxygen was switched off, I felt allwas OK till I woke up when theoxygen had been restored andlooked at my pitiful attempt tocomplete the table and my scribbledwriting. The most important lessonfrom this exercise that I learnt was: Iwas not competent to judge mycondition. For this reason the firstaction on getting a cabin highaltitude warning must be to don youroxygen mask.

There have been some wellpublicised accidents that have beencaused by lack of pressurisation(and therefore inadequate oxygen) ataltitude. Two that spring to mind arePayne Stewart and the Boeing 737operated by Helios. You may recallthat the professional golfer PayneStewart was killed, along with 5others, when his chartered Learjetcrashed. Speculation centred on thepossibility that the accident mighthave followed a decompression earlyin the flight with all onboardbecoming incapacitated. Pilots ofmilitary aircraft who followed theLearjet after the crew stoppedresponding to ATC and climbed

above their assigned altitude of39,000 feet said that the aircraft’swindows were covered with ice, andthere was no sign of flight controlmovement. The Helios accident wasa result of continuing to climbunpressurised to an altitude abovewhich life could be supportedwithout supplemental oxygen.Apparently the flight crew did notdon their oxygen masks.Can’t happen to us? We’ve had a

JETSETS 07November 2008

I hope that this article has given

you something to think about,

and may help to save your life in

the event you suffer a problem.

These events are rare, and

preparation and education can

help us to reduce the possibility

of them becoming more than an

incident. I would be very

interested in any feedback from

you giving indications as to

whether you feel crews are

reluctant to don masks, and if

so giving an indication of the

problems - e-mail address at the

front of this issue.

since you might not get a secondchance!

Once you’ve got your masks onthe aircraft must be descended to alevel where the rest of the crew andpassengers can breathe normally –this is usually 10,000ft or MSA ifhigher. The emergency descentconfiguration given in your QRHshould ensure that the descent canbe made in less than 4 minutes.

Above: Boeing 737

officer was PF. Shortly after take-offa warning sounded, the captain’sdisplays went blank, and smokeappeared from the left side of hisseat. The flight deck crewdescribed this as a ‘smoke haze’and they smelt an ‘acrid burningsmell’. A return to the departureairfield was initiated and thecaptain ascertained from the cabin

crew that there was a strong smellat the front of the passenger cabin.The captain told the first officer thathe was happy to continue withoutmasks as there was only a littlesmell of smoke, and the first officerconcurred. At no time during theincident did the crew put on theiroxygen masks, instruct the cabincrew to put on their oxygen masks,

deploy the passenger oxygenmasks or refer to the QRH. TheQRH contained memory actionswhich included donning masks,and donning the goggles. Thecaptain commented that he did notput his oxygen mask on as therewas only a small amount of smoke.After the smoke cleared, andhaving discussed it with the first

officer, he did not want to put hisoxygen mask on as he wasconcentrating on monitoring thefirst officer, thought it might ‘hamperthings’, and did not want to causeundue concern to the passengersin the event of doing anannouncement with the mask on.

I leave you to draw your ownconclusions from this.

Conclusion

15,000 Indefinite

18,000 20 to 30 minutes

22,000 10 minutes

25,000 3 to 5 minutes

28,000 2.5 to 3 minutes

30,000 1 to 2 minutes

35,000 0.5 to 1 minute

40,000 15 to 20 seconds

43,000 9 to 12 seconds

Above: figure 1 - time of useful consciousness

Altitude ft AMSL Time

Spotlight >

Cabin high altitude warnings

Most crews associate ice with useof the engine de or anti-ice systemand looking out to try and gaugewhen their wings have acquiredenough ice to warrant the use ofthe boots. Unfortunately someicing encounters have also led tothe loss of control - which has not

always been regained. Thefollowing discussion covers threeincidents and tries to draw somelessons.Incident 1The Australian Transport SafetyBureau (www.atsb.gov.au) reportthat the SAAB 340 crewcommenced descent to MDA witha low power setting, and reportedthat the visibility alternatedbetween visual and instrumentflight conditions. During thedescent the engine anti-ice was onbut not the propeller de-ice, norhad the airframe boot de-ice

system been activated. The flightcrew noted ice accretion on thewindshield wiper but not on thewing. At MDA the autopilotcaptured the altitude and, as theairspeed was reducing due to alow power setting, commandedthe trim to progressively raise thenose of the aircraft to maintainaltitude. PF commanded the

autopilot to turn right to begintracking downwind. At about thistime PNF noticed that the airspeedwas decreasing and called‘speed’. As PF applied power tocompensate for the decreasingairspeed, the aircraft rolled to theleft and pitched down withoutwarning. During the recovery fromthe steep pitch and bank angles,the aircraft rolled to the right anddescended to 112 feet AGL. PFregained control of the aircraft andcarried out an uneventful landing.The aircraft’s aerodynamic stallwarning system did not activate

during the initial roll to the left, butthe autopilot disconnected duringthe subsequent roll to the right dueto activation of the stall warning.The investigation determined that,following altitude capture by theautopilot, the airspeed continuedto decrease, and the aircraftstalled. The stall occurred prior tothe stall warning operating. Theinvestigation found that it waspossible for the aircraft to stallprior to activation of the stallwarning system if ice hadaccumulated on the wings.

Incident 2Less lucky were the crew andpassengers on board an ATR overRoselawn Indiana USA asrecorded by the Aviation SafetyNetwork (http://aviation-safety.net).The aircraft had been in the holdfor over 30 minutes, and early on

the crew had selected flap 15 toreduce the nose up attitude of theaircraft for passenger comfort. Theaircraft was cleared to descendand the power was reduced toflight idle. The autopilot wasengaged and the aircraft wasdescending in a 15 degree rightturn. The flap overspeed warning

Lessons learnt sounded and so the flaps wereselected to zero; the AOAincreased and the ailerons begandeflecting to a right wing downposition. About half a second laterthe ailerons deflected rapidly andthe autopilot disconnected. Theaircraft rolled rapidly to the rightand the pitch attitude and AOAstarted to decrease. Withinseveral seconds of the initialaileron input the AOA was 3.5degrees, the ailerons were nearlyneutral and the aircraft attitudewas 77 degrees right wing down.The airplane began to roll to theleft towards a wings level attitudeand the elevator began moving ina nose up direction. Followingseveral unsuccessful attempts bythe crew to correct the airplane’sattitude the airplane impacted ina soya field killing all on board.The NTSB found that theprobable cause was loss ofcontrol attributed to a sudden,unexpected aileron hingemoment reversal that occurredafter a ridge of ice accretedbeyond the de ice boots while theaircraft was in the holdingpattern.

Incident 3The 1992 UK Air AccidentInvestigation Board report statedthat an ATP crew were en-route tothe Channel Islands, and theaircraft entered cloud about 3000feet below cruise height. Thecrew observed sleet and rain inthe cloud although the onlyindication of airframe ice was athin layer of what the crewdescribed as rime ice on the

Icing induced loss of controlexperienced by turboproppowered aircraft

Ice - bewareof the pitfalls

Above: a Saab 340 similar to the aircraft involved in an icing incident in Australia

Above: ice accretion on windshieldindicates severe icing conditions

JETSETS08 November 2008

JETSETS 09November 2008

leading edges of the wing, and3/8 of an inch of rime ice on thewindscreen wiper blade. Theengine and propeller iceprotection system had beenswitched on from take-off. About1000 feet before level off theaircraft speed reduced and therate of climb fell at times to zero.The commander requested alower level, but was notimmediately cleared to descenddue to other traffic. The aircraftwas being flown by the autopilot

in the heading mode with theattitude being controlled by theautopilot pitch wheel. Theaircraft began to experiencevibration which rapidly increasedin severity. Both pilots hadexperienced propeller icing andthe associated vibration, but onthis occasion they thought it to bemore severe. Shortly after theonset of the vibration the left wingdropped and the aircraft began todescend. The aircraft initiallypitched down approximately 15degrees and began a rollingoscillation. The commanderdisengaged the autopilot and feltthat the aircraft was slow torespond to aileron control inputsand large bank angles werereached, particularly to the leftwhere a single peak 68 degreesof bank was recorded. Theaircraft descended below thecloud and at about that time thevibration subsided. As thecommander stabilized the roll, the

aircraft pitch, which had oscillatedback toward zero following theinitial nose drop rapidly increasedto 10 degrees nose up under theinfluence of aft elevator trim whichhad been applied by the autopilotat the onset of the incident. Thecrew reported that at no timewere they aware of a warningfrom the stick shaker.

Three turbo-prop upset accidents/incidents out of many, and allseemingly caused by small

accumulations of ice. Loss ofcontrol has many definitions, butfor the purposes of this article lossof control will be assumed to bewhen an aircraft ceases torespond in the correct sense to a

control input. For instance if theaircraft starts to roll left and doesnot roll right in response to rightroll input then loss of control hasoccurred. This is specially true inpitch where, following a stall, itmay not be possible to raise thenose, and the aircraft can stabilisein a rapid descent. In the aboveincidents all the aircraft rolledinitially; two were subsequentlyrecovered but one was not. Themost likely explanation is that intwo instances the angle of attack(AOA) was reduced below that ofthe stall and control was regainedwhereas in the third instance it wasnot. Conventional stall trainingemphasises that the recoveryshould be made with minimumloss of height, and this minimumheight loss is emphasised by someauthorities. Simulator training usually requiresresponse to the first indication ofstall by applying power andmaintaining the pitch attitude so asto lose little or no height. Thus theaircraft accelerates level to ahigher speed and lower AOA.Upsets usually occur in IMC andthe recovery has to be made froman unusual attitude. An ice relatedstall will most likely have occurredat a lower than normal AOA,higher than normal airspeed withno stall warning. The only warningis the above defined loss ofcontrol.

Digging in the far recesses ofyour minds you may recall theStandard Stall Recovery that wastrained by your first instructor. Itgoes something like this:

Simultaneously• Control column centrally

forward until the buffet stops and:

• Apply full power.• Roll wings level. • Recover to level/climbing flight.Note that the forward movement is‘until the buffet stops’. In otherwords don’t keep pushing till youare pointed straight down! This isthe source of the minimum heightloss recovery training since in anincipient stall the buffet stopsalmost immediately the controlcolumn is moved forward and theaircraft can then be climbed.The most important action is thefirst one – you are unstalling theaerodynamic surfaces. We are notblessed as are many militaryaircraft with AOA gauges – if wewere then stall recovery would besimpler. Our only option is tomove the control column forwardto lower the nose, apply full powerto accelerate and then climb.Note that with autopilot engagedthese cues may be masked; this iswhy it is suggested that you handfly in heavy ice encounters –specially if there is an airspeedloss due to ice accretion.

Above: an ATR 42 aircraft suffered loss of control due to an unexpected aileron hingemoment reversal due to accreted ice

Above: an ATP was involved in an incident whilst en-route to the Channel Islands

JETSETS10 November 2008

Additionally, the autopilot mighttrim as ice is accumulated, thendisengage suddenly when itreaches its trim force limits, andso apply a sudden and substantialcontrol input. This unexpectedcontrol input can also occur whenthe pilot disengages the autopiloteven if it has not quite reached itslimits.

Recovery from an ice relatedstall should be possible if:• At the first sign of stall

(activation of stall warning device, uncommanded roll, buffet or other aerodynamic cues).

• Apply nose down pitch control.• Apply full power.• Initially try and stop further roll.• When roll control regained roll

wings level and recover to a climb.

• Monitor the airspeed to ensure it does not increase too much.

• If the nose cannot be lowered, selecting the first flap angle may help - don’t forget the flap airspeed limitation during the recovery.

In most cases the nose will dropdue to the stall, but this drop willusually not reduce the AOAenough, and further forward

movement of the pitch control willbe required. This pitch down maywell feel uncomfortable, and it isalso important not to bury thenose too deep.

There are other considerations.Ice will probably also affect theprops, and so there may not be asmuch power as usual, and ice willincrease drag (although the dragat Roselawn was only calculatedto have increased by 5 – 10%) .

Following the ATP incident itwas found that the Cl maxreduced from 1.6 to 0.9, and thestall speed increased from about110 knots to 140 knots.

Having covered wing stallswe’re not quite home and dry asthe tailplane can also give rise toproblems. A tailplane stall occurswhen the maximum AOA for thetailplane is exceeded and it stalls.Normally the tailplane creates adown force to balance the nosedown pitching moment of thewing and, under normalconditions without any iceaccreted, the forces on theelevator are roughly equal andproduce no hinge moment. Withice accreted on the tailplane flowseparation can occur on the lowersurface causing it to stall. Thisreduces the amount of down forceproduced by the tailplane and canlead to a nose down pitch. Theflow separation will also result in achange in the elevator force bycausing a change in the relativepressure distribution over theupper and lower surfaces with thelower surfaces sensing a lower

force. This can lead to sticklightening, or even forwardmovement of the control column.This forward force can be quitehigh, and the pull force needed toapply up elevator can be evenhigher. So in the worst case of atailplane stall the aircraft will pitchrapidly nose down and the controlcolumn will move rapidly andstrongly towards the instrumentpanel. Recovery from this is thereverse of a wing stall – pull backon the control column. Selectionof the flap up to the next setting isalso likely to assist the recovery.The tendency towards tailplanestall is exacerbated by speed andflap angle. A tailplane stall is mostlikely to occur at high flap anglesand at speeds around Vfe. This isbecause extending the flapsincreases the down wash anglefrom the wing and thus thetailplane AOA. For any flap anglethe AOA on the tailplane getsnearer the stall as speed isincreased. Therefore at higherflap angles the wing stall marginis increased but the tailplanemargin is reduced as airspeed isincreased. On some aircraft, theHS 748 for instance, the tailplanestall occurs at lower speedsnearer Vref.

From the above it is apparentthat those operating turbo propsin ice need to be aware of thepitfalls. However, jets are notimmune, and jet pilots shouldknow that ice may well have aneffect on the stall characteristicsof their aircraft.

ABBREVIATIONS

AGL Above Ground LevelAOA Angle Of AttackAMSL Above Mean Sea LevelATC Air Traffic ControlATIS Automatic Terminal Information ServiceCAA Civil Aviation AuthorityCL Co-efficient of Liftcm centimetre(s)CRJ Canadair Regional JetCWP Central Warning PanelECS Environmental Control SystemFDR Flight Data RecorderFL Flight LevelhPa hectoPascalsIMC Instrument Meteorological ConditionsKm kilometreKt knotLDA Landing Distance AvailableLDR Landing Distance Requiredm meter(s)MSA Minimum Safe AltitudeNTSB National Transport Safety Board (USA)PF Pilot FlyingPNF Pilot Not Flying (now Pilot Monitoring (PM))psi pounds per square inchQRH Quick Reference HandbookRLW Regulated Landing WeightRPM Revolutions Per MinuteSOP Standard Operating ProcedureVfe Flap limiting speedVIB Vibration warning captionVMC Visual Meteorological ConditionsVref Landing Reference Speed at threshold

JETSETS 11November 2008

Among the recent incidents that have been reported to us are the following:

Interesting accidents and incidents

BAe 146/Avro RJ• An RJ burst all 4 tyres on

landing during a proving flight into a short runway. BAE Systems are currently assisting the Authority in the investigation.

• An RJ lost a nose wheel on take-off and landed safely. The loss was identified as a bearing failure. The aircraft landed safely.

• A 146 turned off runway and skidded off taxiway. Left mainundercarriage collapsed. There were no injuries, and this incident is under investigation with BAE Systemsassisting the Investigators.

• A 146 was carrying out an approach to a runway where snow clearing had been carried out. During the flare to land the aircraft started

The UK AAIB has recentlypublished the following reportswhich can be read on their website (www.aaib.gov.uk/bulletins):

Formal Report 3/2008 reportson the Jetstream 32 wheels upapproach that I commented onin the editorial; Formal Report

6/2008 reports on an HS 748which landed in poor weather onGuernsey, and overran therunway.

The June 2008 Bulletin covers afumes incident that occurredduring a 146/300 positioning flight.There was an unusual smell and

both flight crew felt unwell. Thefumes were confirmed to comefrom a chemical in the forwardtoilet, and the August 2008 Bulletinreports on a tailstrike thatoccurred to an RJ/100 at LondonCity. The strike caused significantdamage to the rear fuselage

diverging to the left of the runway centre line, and it landed on the uncleared left hand side of the runway. Shortly after touchdown the nose landing gear collapsed. Under Investigation by the local Authorities, and BAE Systems are assisting the Investigators.

Jetsteam 31/32• A Jetstream 32 overran on a

slippery runway with no injuries. This incident has raised questions as to when to select full flap. To avoid destabilising the approach we recommend that the aircraft should be fully configured and stable by 1000 feet IMC and 500 feet VMC. Contrary advice in the Crew Manuals will be amended in due course.

• A Jetstream 31 landed short following an approach in turbulent conditions. This led to the collapse of both main undercarriage legs and the aircraft departed from the runway. There were no injuries.

• A passenger attempted to hijack a Jetstream 32 in New Zealand. Both crew were stabbed, but landed the aircraft safely.

Jetstream 41• A Jetstream 41 called MAYDAY

and diverted with control problems. The aircraft landed safely. This incident is the subject of an UK AAIB Investigation. BAE Systems are assisting this investigation.

ATP• An ATP suffered an inverter

failure with no CWP warning. During the flight, problems were encountered with the number 2 IAS and servo altimeter indications, and the number 2 auto-pilot. During the cruise, other faults occurred and the crew diagnosed an inverter fault. On checking the inverter output the number 2 inverter indicated a low voltage although no fault / failure was indicated. The inverter failure checklist was actioned with

the result that all systems were restored with only an amber XMON legend displayed on the number 2 PFD and the EFIS caption was illuminated. BAE Systemsare investigating this incident.

• An ATP suffered torque fluctuations of 20% which reduced slightly when the power lever was retarded. Selection of the ECS to OFF had no effect and, as the fluctuations increased, the engine was shut down. A connector to the EEC was found to be heavily contaminated. This connector was cleaned, and this rectified the fault..

• There have been 5 ATP power interrupts. These were characterised by the torque falling to (usually to 0)over a period of 1 to 3 seconds followed by a recovery between 2 and 19 seconds later. In one case the event was over in 2 seconds and in all other cases the propeller RPM maintained at its preset value of 80% or 82.5%. Ice ingestion is thought to be thecause, and BAE Systems areasking all ATP Operators to ensure that there are no large steps between adjoining panels in the intake.

JETSETS12 November 2008

FACT FILE REGIONAL AIRCRAFT IS RESPONSIBLE FOR THE WORLDWIDE SUPPORT OF APPROXIMATELY 900 AIRCRAFT

High flight

Car

toon

s: D

erek

Pau

l, G

raph

ics

and

Med

ia, P

rest

wic

k

Humour

Learn from the mistakes ofothers, you will not have enoughtime to make them all yourself.

There are 3 rules for making aperfect landing; unfortunately noone can remember what they are.

As a pilot you can do anythingyou want in flight. As long as itsright…and we’ll let you know ifits right after you’ve landed.

Always keep your words shortand sweet as you may have to eatthem.

Many years ago, when I waslearning to sail, the instructor askedme ‘when should you reef thesails?’ As I searched for an answer,and was about to guess, he said‘the first time you think about it’.

Basic flying rule; try and stay in themiddle of the air, and do not gonear to the edges of it. The edgescan be recognised by theappearance of ground, buildings,trees, sea and interstellar space. Itis much more difficult to fly therestoring dead batteries.

JETSNIPSA light hearted look at the aviation industry

Heard on the PA“Ladies and Gentlemen...Welcome to London... at this timein the flight we ask for volunteersto clean the toilets. If you want tohelp... all you have to do is standup before the fasten your seatbeltsign has been switched off.”

Oh, I have slipped the surly bonds of earth (1),And danced (2) the skies on laughter silvered wings;Sunward I've climbed (3), and joined the tumbling mirth (4)Of sun-split clouds (5) and done a hundred things (6)You have not dreamed of --Wheeled and soared and swung (7)High in the sunlit silence (8).Hov'ring there (9)I've chased the shouting wind (10) along and flungMy eager craft through footless halls of air (11).Up, up the long delirious (12), burning blueI've topped the wind-swept heights (13) with easy grace,Where never lark, or even eagle (14) flew;And, while with silent, lifting mind I've trodThe high untrespassed sanctity of space (15),Put out my hand (16), and touched the face of God.(from John Gillespie Magee Jr., “High Flight”)

NAA SUPPLEMENT to “High Flight”

1. Flight crews must insure that all surly bonds have been slipped entirely before aircraft taxi or flight is attempted.

2. During periods of severe sky dancing, flight crew and passengers must keep seatbelts fastened. Flight crew should wear shoulder harnesses if provided.

3. Sunward climbs must not exceed the maximum permitted aircraft ceiling.

4. Passenger aircraft are prohibited from joining the tumbling mirth.

5. Pilots flying through sun-split clouds must comply with all applicable visual and instrument flight rules.

6. Do not perform these hundred things in front of Flight Operations Inspectors.

7. Wheeling, soaring, and swinging will not be accomplished simultaneously except by pilots in the flight simulator or in their own aircraft on their own time.

8. Be advised that sunlit silence will occur only when a major engine malfunction has occurred.

9. "Hov'ring there" will constitute a highly reliable signal that a flight emergency is imminent.

10. Forecasts of shouting winds are available from the local meteorological office. Encounters with unexpected shouting winds should be reported by pilots.

11. Pilots flinging eager craft through footless halls of air are reminded that they alone are responsible for maintaining separation from other eager craft.

12. Should any crew member or passenger experience delirium while in the burning blue, submit an irregularity report upon flight termination.

13. Wind swept heights will be topped by a minimum of 1,000 feet to prevent massive airsickness-bag use.

14. Aircraft engine ingestion of, or impact with, larks or eagles should be reported to the Authority and the appropriate aircraft maintenance facility.

15. Air Traffic Control (ATC) must issue all special clearances for treading the high untresspassed sanctity of space.

16. Pilots and passengers are reminded that opening doors or windows in order to touch the face of god may result in the loss of cabin pressure.

“As you leave the aircraft pleasemake sure to gather all of yourbelongings. Anything left behindwill be distributed evenly amongthe cabin crew. Please do notleave any children or spouses.”

“Dad, why are there always two pilots?”“One has to prevent the other fromdoing stupid things.”“Which one is doing the stupid things?”

“And during the take-off I may fail one of the power units”