WIN

TE

R 0

9

ISSUE 77

The official publication of the United Kingdom Flight Safety Committee ISSN 1355-1523

O N C O M M E R C I A L AV I AT I O N S A F E T Y

ContentsThe Official Publication of THE UNITED KINGDOM FLIGHT SAFETY COMMITTEE ISSN: 1355-1523 WINTER 2009

Editorial 2

Chairman’s Column 3

Do you really understand how your trim works? 4

by Alex Fisher – GAPAN

A Spanner in the Works? 7

by Ian Hay – Coplan Limited

Flying into the sea 8

by Linda Werfelmen

Seperation Anxiety 11

by Wayne Rosenkrans

At Last: The answer to the meaning of life, the universe and everything 15

by Major C. P. Evans, MoD Dars SO2 Engineering Rotary Wing

Using Weather Radar 19

by Erik Eliel

Fatigue and take-off data…

a lethal combination 22

Members List 24

Front Cover Picture: BLinks first Citation Mustang G-FBLK, on the ramp at Farnborough Airport.

FOCUS is a quarterly subscription journal devoted

to the promotion of best practises in aviation

safety. It includes articles, either original or

reprinted from other sources, related to safety

issues throughout all areas of air transport

operations. Besides providing information on safety

related matters, FOCUS aims to promote debate

and improve networking within the industry. It

must be emphasised that FOCUS is not intended

as a substitute for regulatory information or

company publications and procedures.

Editorial Office:

The Graham Suite, Fairoaks Airport, Chobham,

Woking, Surrey. GU24 8HX

Tel: 01276 855193 Fax: 01276 855195

e-mail: admin@ukfsc.co.uk

Web Site: www.ukfsc.co.uk

Office Hours: 0900 - 1630 Monday - Friday

Advertisement Sales Office:

UKFSC

The Graham Suite, Fairoaks Airport, Chobham,

Woking, Surrey GU24 8HX

Tel: 01276 855193 Fax: 01276 855195

email: admin@ukfsc.co.uk

Web Site: www.ukfsc.co.uk

Office Hours: 0900 - 1630 Monday - Friday

Printed by:

Woking Print & Publicity Ltd

The Print Works, St. Johns Lye, St. Johns,

Woking, Surrey GU21 1RS

Tel: 01483 884884 Fax: 01483 884880

e-mail: sales@wokingprint.com

Web: www.wokingprint.com

FOCUS is produced solely for the purpose of

improving flight safety and, unless copyright is

indicated, articles may be reproduced providing

that the source of material is acknowledged.

Opinions expressed by individual authors or in

advertisements appearing in FOCUS are those of

the author or advertiser and do not necessarily

reflect the views and endorsements of this journal,

the editor or the UK Flight Safety Committee.

While every effort is made to ensure the accuracy

of the information contained herein, FOCUS

accepts no responsibility for any errors or

omissions in the information, or its consequences.

Specialist advice should always be sought in

relation to any particular circumstances.

1focus winter 09

Threats and Challengesto Aviation Safety

EDITORIAL

Iam writing this editorial for the winter

edition of FOCUS on the aircraft

returning from this year’s International

Aviation Safety Seminar where we have

considering the current state of civil

aviation safety and, in particular, what

safety professionals should be tackling next

to get the air accident statistics from off

the plateau on the graph and moving

downwards once again. The conference was

presented with a wide variety of views and

solutions, some old and some new, both in

conference session and in the many sidebar

discussions which contribute to additional

value to these sorts of gatherings.

We were fortunate and fascinated to behosted by the Civil Aviation Administration ofChina (CAAC) who fielded several seniorspeakers who were able to provide a glimpseinto the impressive progress made in air safetyby their country, despite the challenges drivenby an incredible expansion in their commercialtransport sector of over 17% year on year anda mass migration from old Eastern Blocktechnology aircraft to modern Western fly-by-wire aircraft bristling with extensiveautomation. Equally impressive, the Chineseare opening a new airport every 4 months upuntil 2020! But in meeting this incredible paceof change, to their great credit they have doneso without a single accident in the past twoyears or more.

The philosophy expounded by the Administratorof the CAAC was interesting: the more you talkabout your achievements, the less you achieveand the more you talk about your problems, thefewer problems you encounter! Paradoxically,another IASS session reflected upon the impacton safety, both actual and potential, of therecession generally and the serious downturn inthe industry in the western world in particular.Unsurprisingly, the tragic Colgan accident inBuffalo earlier in the year was a major topic ofconversation and promoted training and fatigueto the top of the agenda particularly amongstthe large US contingent.

Another major theme of the conference wasthe introduction and implementation of SafetyManagement Systems across the globe. It isclear from evangelists and sceptics alike thatthe industry and its leaders are placing great

store on this approach to deliver the next majorstep improvement in aviation safety. There is asignificant body of opinion that believes this tobe the only game in town as far as getting theaccident rate moving down from the plateau ofrecent years. However, it was pleasing to notethat the other side of the SMS coin, thereporting culture required to support it, wasplaced firmly on the table by several speakers,including myself. An effective reporting regimeis absolutely fundamental to enable aproductive SMS to develop and prosper, but thisespecially difficult issue does not enjoyanything like the same level of consideration ordiscussion as the SMS process itself.

Development of an open reporting culturecannot be discussed without addressing thethreats to its success. One presenter, who is aleading Barrister in the UK specialising oncorporate manslaughter, made it perfectlyclear that as far as special pleading for thecommercial transport sector to be protectedfrom criminal prosecution was concerned, theGenie was well and truly out of the bottle! Buthe then went on to emphasise that, after anaccident involving death or injury, the targetfor such prosecutions was not necessarily thepilot or engineer, but the leadership of theorganisation employing them. In most criminalcases in the future, it would be the failure ofthe organisational support to the employeewhich would be cited as the main contributorto the event. Therefore, it would be the chiefexecutives and accountable managers of thesefailing organisations who would be brought tobook. Furthermore, the rather restrained andderisory fines of the past, which used to becomfortably met without large companiesbreaking step, are now being changed in law toamounts of up to 10% of annual global profitfor the worst offenders.

Our major concern now must be thatcriminalisation will increasingly featurealongside internal company discipline andanti-trust laws denying airlines fromexchanging information on the list of realthreats to any positive reporting culture.Certainly in the UKFSC, the concern must bethat any threat to free and open occurrencereporting could seriously undermine the vitalcontribution to aviation safety enjoyed in theUK over the past 2 decades. Arguably, with the

implementation of SMS now finally takingshape, this could not have happened at a worsetime since the gathering and analysis ofoccurrence reporting data is the very life bloodof a successful SMS. Without effective and freeoccurrence reporting within a confidentialwrapper, proactive action to deny accidentsand incidents is dead in the water!

So what must be done to maintain andencourage occurrence reporting in the UK, andbeyond, in light of this increasing threat. Onething is for certain; the idea of commercialaviation being placed above the law is a non-starter. The challenge for all aviation safetyprofessionals is to get engaged with the legalprofession and the judiciary at everyopportunity, to explain the essentialcontribution of occurrence reporting and safetyinformation sharing to improving futureaviation safety. Equally, we must explain thatwhilst negligent and wilful misbehaviour mustbe pursued through the court, the carefulhandling and protection of aviation safety datais required to reassure those reporting theirhonest and genuine mistakes, to which nohuman being alive is immune, can do so withoutfear of action being taken against them.

But that is not all we must do. There is anequally important challenge. We must alsoengage with the responsible media and seekthe trust of the general public. They must beeducated to understand that only throughaffording effective protection of safety dataand information and a confidential exchangebetween airline operators, regulators andservice providers can the lessons be learned forthe overall well-being and safety of the publicinto the future.

2 focus winter 09

CHAIRMAN’S COLUMN

Winter Thoughtsby Capt. Tony Wride, Monarch Airlines

Whilst sitting in my room on a

tropical paradise island in the

middle of the Indian Ocean with the

weather outside clear and very hot I

decided to write this article. Three days

ago when I flew the Company A330 into

here it was a different story completely

with strong blustery winds, heavy rain and

lots of cumulo-nimbus clouds around,

which at the end of a 10 ? hour flight

certainly concentrated the mind! At one

point it was looking likely that we would

have to divert as the wind, which was

straight across the runway, had increased

to over 33 knots and the visibility

decreased to 500m in heavy rain.

Fortunately a well timed gap appeared and

we were able to land with only a 15 knot

crosswind on a runway that was not

completely flooded. The cyclonic

conditions persisted for 2 more days, in

fact it actually got worse which caused a

number of aircraft to divert, and it’s only

now that the wind has abated and the sun

has come out.

Pilots constantly face the challenges that‘Mother Nature’ throws our way and as we allknow some of those challenges can beextreme. I watched an ‘Air AccidentInvestigation’ programme recently thatshowed a few serious accidents that had beencaused by severe weather.The first was a DC9that crashed in 1977 when it flew through themiddle of a thunderstorm and ended up withboth engines flamed out.The crew very nearlymade it down in one piece having landed ona highway but unfortunately the aircraft brokeup when it hit a power line and severalobstructions. In another incident a DC10landing in Texas encountered severewindshear as it passed through a microburston short finals resulting in a crash and a largenumber of fatalities. Both of these incidentswere as a direct result of the severe weatherassociated with thunderstorms a verycommon occurrence in the hotter parts of theworld and not that uncommon in the UK.

As we head into winter there are otherweather challenges that not only confront thepilot community but also challengeeverybody working to support the aircraft,particularly on the ramp. Winter brings awhole host of weather problems from severegales, to fog, ice and snow.The whole aviationcommunity suffers from the effects whether

you are sitting in Air Traffic Control trying towork out how to keep your airport operationalor working on the ramp loading the bags intoan aircraft in freezing conditions and sufferingfrom the effects of the cold. The pilots facethe challenge of making the decision, basedon well proven techniques, on whether it issafe to take off or land in the prevailingconditions. But the pilots rely heavily on somany other people doing their job correctly.The team that carries out the aircraft de-icingprior to departure, the team doing the runwaysnow clearing, the person doing the check onthe runway condition, and the person that didall the calculations for the contaminatedrunway performance to name but a few. All ofthese people are vital and any one of themcould cause an aircraft accident if they don’tperform their duty diligently.

You may remember the tragic crash thathappened in Birmingham a few years back toa private jet that had not been de-icedcorrectly, in fact I believe it hadn’t been de-iced at all.The aircraft tried to take off with icestill on one wing destroying the liftcharacteristics of that wing whilst the otherwing had been de-iced as a result ofenvironmental factors. The different liftcharacteristics caused the aircraft to rollshortly after getting airborne with fatalconsequences. The bottom line is that noaircraft should try taking off with anything ona wing, or other control surfaces, that couldaffect the lift generating properties of thatsurface.That’s easy I hear you say we just haveto make sure we always thoroughly de-ice theaircraft. However, in this climate of all airlinestrying to cut costs are we at risk of somecorners being cut?

You only have to read the papers to hear thatsome of the big airlines are loosing in excess of£1 million per day whilst the smaller airlines arealso making a loss. De-icing is a costly exerciseand airlines are actively trying to find cheapersuppliers in an effort to reduce the bottom linecosts. True, the Tony Wride De-Icing CompanyLtd using a 25 year old American built de-icingtruck with a team that were hired yesterdayand underwent a 1 day de-icing course, couldspray de-icing fluid onto an aircraft for 25% lessthan the well established professionals, butwould that aircraft be properly de-iced? In factin most cases the aircraft would be de-icedsufficiently but the risk that on occasions anaircraft could be de-iced incorrectly is greater.

Experience in any job, be it as an Air TrafficController, a Pilot, or a de-icing operator hasan unquantifiable value and it could be arguedthat the experience helps to reduce the risk.Equally, correct training and a robust, activeQuality System also helps reduce the risk.Now try selling that to a senior manager whois looking at the bottom line on everythingand deals in numbers not unquantifiablevalues! Add into the equation a general publicwho expect to fly several thousand miles in aseat with loads of leg room, get free drinksand Master Chef quality food all for theequivalent of a night out for 2 at aMacDonald’s restaurant and you wonder howon earth the commercial aviation industrycan survive safely.

Winter operations requires everybody toremain FOCUSED, (no pun intended but it is agood name for a Safety Magazine), on theirparticular job despite the adverse conditionsthat they may have to work in. Everybodyneeds to be professional and that may meanspending that little bit of extra timefamiliarising yourself with the relevantinformation on Winter operations so thatwhen it comes you are fully prepared. Thosepeople who are out working in the adverseconditions need to be looked after in terms ofproviding correct clothing and everyone needsto be more understanding if something takesa little longer to do.

As an industry we cannot afford to cut cornersand perhaps now is the time that we actuallyshould be re-educating the general public onthe costs involved in flying an aircraft safely.

3focus winter 09

focus winter 094

by Alex Fisher - GAPAN

Do you really understand how your trimworks? Many do not, and why it matters.

Picture yourself in a conventional

airliner, say a 737 of any generation.

You have to do a low level go-around,

perhaps because your fail passive Cat III has

just failed, er, passively. You apply GA

thrust, and the aircraft pitches up. If you are

low enough, you may already have some

extra helpful nose up trim applied thanks

to the ‘design feature’ that ensures that in

the event of AP failure at low level, the

aircraft pitches up not down, and so a few

units of nose up trim are applied late in the

approach.Your speed is low, about Vapp and

the thing is pitching firmly upward. You

need ample forward stick/elevator to

restrain it.You don’t want to carry this load

for long so you retrim. Question: if you run

the trim forward while maintaining

forward pressure on the wheel, what

happens? Hands up all those who think the

load reduces to zero. I see a lot of hands. My

unscientific polling to date suggests that

just about everyone is convinced that this

is what happens, but it doesn’t.

Nearly everyone of my generation trained ona Cessna 150 or a Piper PA28. You fly thoseaircraft by putting the attitude where youwant it, holding it there by holding the stickrigid and retrimming until the load goes tozero. In fact if you didn’t do that, but were tooquick and started trimming before the aircraftwas stable, the instructor would exhibit asevere sense of humour failure. Let’s justconsider what is going on. Starting from an‘in-trim’ state, fig 1(a) (just for illustration Ihave shown it as everything in the middle, butobviously this isn’t essential to the argument);then, fig 1(b), the column is held forwardmoving the elevator down. Moving the trimwheel, fig 1(c), in this case moves a trim tabwhich relieves the control load until it goes tozero; the column can again be released, and itstays forward where you left it. So in thisscheme, the control column stays forward forhigh speed and back at low speed. Although Ihave shown a tab operated system, the sameresult can be achieved without a tab bymeans of a spring in the control circuit or byaltering the neutral point of the feel system.Aircraft as diverse as the Tiger Moth, theL1011, and Concorde fly this way.

Now there is another class of aircraft thatworks totally differently. This group includesmost conventional transports, and even thenon conventional A320 series in direct law. Inthese, the tailplane is controlled directly bythe trim system, while the control wheelcontrols only the angle of the elevator relativeto the tailplane. Now starting again from theout of trim state we started from above (seefig 2), as the nose down trim is applied, thetailplane starts to move leading-edge up. Inorder to keep the force contributed by boththe tailplane and elevator constant (i.e. tomaintain attitude), the elevator angle has tobe reduced as the tailplane incidenceincreases (fig 2b). To do this, the column/wheel has to be moved back towards neutral.When the operation is complete, the column/wheel is back in the neutral position, which isthe only place it can be released withoutfurther movement (fig 2c); its position doesnot indicate the trim state of the aircraft. For

years Boeing manuals have said flatly that thecontrol wheel cannot be moved opposite tothe direction of trimming motion (the trimmotors cut out if it is)…. Wrong, it can, andindeed has to, be moved in the oppositedirection every time the trim is used; theaction is achieved by just relaxing thepressure on the column and allowing to driftback to neutral. It is true that if pressure isapplied to the column opposite to thedirection of trim, then the trim cuts out.

This behaviour (column always returns toneutral regardless of speed) is not necessarilylimited to aircraft with trimmable tailplanes;for example, if the column operates a servotab while the trimmer moves a separate trimtab, the effect would be the same (I believethe 146/RJ series works this way). Doubtlessthere are other combinations too, you reallyhave to study the systems carefully.

When I converted from a ‘conventional’trimming type (Trident) to a separatetrimming tailplane (757), not a word on thissubject appeared in the training notes, norwas anything ever said by any trainingcaptain. Many years later I did writesomething for the company Magazine andgeneric training manual, but apart from onereprint in the Far East it has not been widelycirculated. So how do people go through anentire career without realising things havechanged from the way they were first taught?I think it is because mostly any columnmovement is followed immediately by smallmovements of the trimmer, so large loads arenever allowed to develop and the reversecolumn movements are virtuallyimperceptible. In ‘normal’ flight operations,movements in pitch are mostly quite small,apart from two: rotate and go around; thelatter is relatively rare, while the former istransitory (if the take-off trim is roughly right(!) you can relax the load after lift off with theaircraft roughly at the right attitude).

So why does it matter? The chances are youwill fly more smoothly if you understandwhat is going on, but there are three broadcategories of error which are likely if thesesubtleties are not understood, I will citeexamples of each.

focus winter 09 5

1. Failure to understand the trim function (theprocess described earlier) itself. This isn’tdisastrous. Most pilots are in this category, butthey cope well anyway, by simply flying onthe trim. This isn’t how they were taught, but,well, it works. It begins to matter when thetrim changes are large. I have watched, in thesimulator, a 737 go-around from a Cat lll failpassive approach (as described above) with itsmarked pitch up; HP kept his arms lockedforward to contain the attitude whilstsimultaneously running the trim forward withthe thumb switch. I am sure he was expectingthe trim to reduce push needed and he eitherdidn’t know, or had forgotten, that it wouldn’t.We duly pitched straight back quickly into theground as the tailplane incidence ‘bit’. I can’tcite with certainty any accident that has beencaused by doing this, but I strongly suspectthis was a factor in the infamous Icelandairupset event at Oslo The aircraft went quickly

from +20 deg to -40 deg and was only savedfrom a CFIT by a 3.5g pull up, bottoming outat 360ft. Sadly, the report does not discussthe control inputs, nor does it contain anyFDR traces, so this trim confusion explanationmust remain speculation. I would beastonished, however, if there weren’t moreexamples of this error, particularly inunfamiliar situations.

2. Failure to realise that the tailplane,commanded by the trim system, is a totallyindependent pitch control; it will be availableif the primary control is inoperative orineffective. But if you only think of thetrimmer, wrongly, as a column-load reductiondevice, you may not think of its other usewhen needed. The following examplesillustrate the point; I am certain of the first,the others must remain speculation in theabsence of evidence.

(a) 747-400 Take-off incident . Just after liftoff the aircraft suffered an elevator hardover,uncommanded full nose down movement ofone elevator; the pitch attitude began toreduce. The crew’s reaction not unreasonably

was first to pull harder, then a lot harder,which succeeded in preventing an immediateaccident, but cannot be said to have trulyregained control. The anomaly lasted about 8secs until a spike in the hydraulic pressureduring the gear raising sequence allowednormal control to be resumed. No onethought of just blipping the trim button torestore order. Did thinking of the trim asmerely a load reducer blind them to thesimple solution? The incident report does notmention the alternative control available anddoes not discuss that part of the pitch controlsystem at all.

(b) THY DC-10 crash at Ermenonville in 1974.This was caused by an improperly securedcargo door which blew off; the floor above itcollapsed due to the pressurisation load,disrupting the controls and injecting a nosedown elevator input. Rumour, I admit quiteunsubstantiated, has it that it could have beenflown on the trim as there was still hydraulicpower to the tailplane (350 casualties).

(c) The BAC 1-11 flight test super-stall . Therewas insufficient elevator to recover, but theFDR trace shows that no attempt was madeto adjust the tailplane which would have beenmore powerful. It is pure speculation nowafter 40 years, but it is an intriguing thoughtthat it might have helped. There wouldcertainly have been no similar possibility forthe Trident that was lost during a pre deliverytest flight a year or so later as the trim and

6 focus winter 09

column both operated the tailplane and itsgeared elevator together (see fig 3).

3. Failure to appreciate that loss of control inpitch might be due to the independentoperation of the trim system. Several wellknown pitch upsets to A300s and A310s (seefor instance the TAROM upset at Orly Sep1994 , and the A300 at Nagoya, April 1994)have been caused by a tailplane movementwhich was not fully appreciated by the crew,and was all the more insidious preciselybecause there was NO change to the load onthe column. This is the reverse of thesituation in (2). None of these occurrenceswere technically trim runaways, so therewere no warnings and no indication to thecrew from the feel of the column. The firstincident started with the flap overspeedprotection system (the designers obviouslythought that putting in nose up trim wouldreduce the speed… well it will if youunderstand totally what is happening anddon’t override it); the second, a fatal accident,started with an inadvertent, and probablyunnoticed, GA selection.

A system where there are two independentmeans of control, has obvious safety benefits,but it also has pitfalls if it is not fullyunderstood.The lack of importance given to thetrim system in training seems extraordinary. Irecall asking for TC guidance during my 757conversion, to be told that there was nodifference to previous types; when I finallyconvinced His Eminence that there was, heblustered that it didn’t matter. I can find norelevant discussion in my edition of the Bible,Handling the Big Jets; I guess the Test Pilots justcope with anything they come across withoutpreconception, and perhaps don’t realise howmuch baggage the rest of us carry from ourbasic training.Accident investigators would alsodo well to ask themselves more often just howthe unfortunate pilots had been trained, andcover the likely rationale for the control inputsin their reports. The illustrations I have used areobviously very rare events, so it is very unlikelythat any one reading this will ever face theirlike. Engine cuts at V1 are pretty rare too, butthey get a lot more exposure in training thanthe basic control functions, odd, isn’t it.

Safe flying

Postscripts

This article was written for the UKFSC Focusmagazine in late 2007. Since then there hasbeen a spate of accidents and alarmingincidents in which ‘tailplane ignorance’ hasplayed a part. The 737 accident atAmsterdam, an as yet unpublicised 737incident in the Far East, and the PerpignanA320 crash all, in different ways, involved astall and unsuccessful or botched recovery.The shared feature is that in each case thetailplane had wound itself to a fully (aircraft)nose up position, as in (3) above; thecombination of pitch up, due to full power,and low speed, meant recovery was probablyimpossible using elevator alone, to get thenose down meant moving the tailplane backto a more normal position, which meansrunning the trim forward. The A320 accidentappears to be the result of an improper flighttest, but the two 737 cases occurred innormal line flying and illustrate howimportant it is to understand what thetailplane is doing, and how easy it is for it tofinish up somewhere unexpected; in boththese cases the trigger was an unnoticedAutothrottle failure on approach, the speedfell and the autopilot duly trimmedprogressively further back until it reached fullnose up and quit; recently, April 2009, the UKAAIB published a report into yet another 737near stall and upset and made the followingrecommendation:

Safety Recommendation 2009-045: It isrecommended that Boeing clarify the wordingof the approach to stall recovery QuickReference Handbook Non normalManoeuvres to ensure that pilots are awarethat trimming forward may be required toenhance pitch control authority.

The report contains the relevant Boeing OpsManual pages in an appendix, including this:

To recover from a stall, angle of attack mustbe reduced below the stalling angle. Nosedown pitch control must be applied andmaintained until the wings are unstalled.Application of forward control column (as

much as full forward may be required) and theuse of some nose-down stabilizer trim shouldprovide sufficient elevator control to producea nose-down pitch rate. It may be difficult toknow how much stabilizer trim to use, andcare must be taken to avoid using too muchtrim. Pilots should not fly the airplane usingstabilizer trim, and should stop trimming nosedown when they feel the g force on theairplane lessen or the required elevator forcelessen. (my emphasis)

The forces won’t lessen by themselves, so thatlast remark puzzles me – does the writer thinkthat the column load will go to zero as thetrim is run forward? It can certainly be readthat way, but if you have understood the restof this article you should be able tounderstand the subtle coordination requiredto bring the tailplane safely into play withoutcreating a worse nose-down problem. But youwill also appreciate that the bigger danger atthe moment may be that too many pilotsdon’t think about trimming at all in thissituation.

7focus winter 09

It’s a simple question which I ask when

visiting maintenance organisations:-

“How can you be absolutely sure that all

the tools you used on that airframe today

are back in your toolbox?”

The answer however, is not so simple and isquite disturbing.

Most aircraft and helicopter maintenanceoperations allow (and in some cases encourage)uncontrolled tooling to be used on airframe andpower plant maintenance. This is justified witha variety of statements including:-

■ Our engineers have their own tooling.■ It’s not workable.■ It costs too much.■ It’s not a legal requirement.■ Our engineers are careful.

Uncontrolled tooling is defined as tools thatare unaccounted for, either supplied by thecompany or brought into the hanger byengineering staff and usually stored in apersonal toolbox with hundreds of other tools.The problem is that if one tool is missing (leftin the airframe) the engineer/fitter is unlikelyto know and this ‘lost tool’ has the potential tocause a catastrophe!

The simple solution is to use a tried and testedtool control strategy to ensure ALL tools areaccounted for at any given time duringmaintenance. Simply housing the engineers’and fitters’ tools in twin coloured foam creates

a situation where just a glance at the tool boxcan confirm that all tools are present andaccounted for as any missing tools would behighlighted by the bright bottom layer offoam. This system is called Visual Tool ControlVTC and is a simple and economical first steptowards full tool control.

Once the decision to ‘control tools’ has beentaken the whole maintenance operation instantlybecomes safer, all unauthorised tools areremoved from the hanger and all the remainingtools are listed in a tool control manifest andthen housed in VTC foam (even the engineersown tools can be authorised and integrated intothe foam – excuse number one dealt with!).

Many organisations also have a variety ofspecialist tools usually kept in the bonded store,these tools should also be housed in VTC foamin order to establish tool control, anotherbenefit of using VTC foam is that it protectswhatever it holds and with very expensiveaircraft tooling that’s reassuring. VTC foamcomes in a wide variety of colours which can beused to further enhance safety and efficiency.For instance it can be used to provide a visualdifference between AF and metric tools or todefine tools assigned to a particular job orairframe. The foam is ‘closed cell’ so it does notabsorb liquids and is also resistant to allhydraulic fluids including Skydrol.

Tool control is vital to the safe and professionalmaintenance of all aircraft. It is defined by the

CAA, FAA and EASA as the responsibility of themaintenance organisation’s management toensure that acceptable standards andoperating procedures are put in place. In thesedays of corporate manslaughter and litigationany company not enforcing a well thought outtool control policy is leaving itself veryexposed. Imagine an accident (fatal orotherwise) where a ‘lost tool’ is proved to bethe cause; the subsequent accidentinvestigation would soon focus in on thecompanies standard operating procedures fortool control with devastating consequences.

*In fact just such an accident occurred atAviano AFB in Italy in 2007 where a USAirforce UH-60 Blackhawk helicopter crashedkilling six people and injuring five more. Theaccident investigation concluded the crashwas either caused by control failure or aforeign object lodged in the controls. Thefamilies have now filed a law suit against themaintenance contractor for negligence.*Source: Stars & Stripes, October 18th 2009.

Cost is always an issue and is the mostcommon excuse for doing nothing, in fact Iwas recently told by a major helicoptermaintenance organisation within the UK ‘Welove the idea but until there’s a fatal accidentand we’re forced into it, we won’t buy it!’ I wassimply stunned. However reality is oftendifferent to perception, the truth is, toolcontrol is very inexpensive especially whenweighed against inefficient working practices(looking for lost tools), delayed aircraft releasedue to tool loss (airlines) or the defence of alaw suit following an accident attributed totool FOD.

Tool Control – When you think about it,it’s common sense.

A Spanner in the Works?by Ian Hay - Coplan Limited

8 focus winter 09

Flying into the Seaby Linda Werfelmen

The crew of a CHC Scotia Aérospatiale

SA 365N Dauphin 2 lost control during

a night time approach to a gas platform in

the Irish Sea, overflying the landing site

and striking the water. The helicopter

disintegrated on impact and sank in the

Dec, 27, 2006, crash, killing the two pilots

and all five passengers.

The U.K. Air Accidents Investigation Branch(AAIB), in its final report on the accident, citedthree contributory factors, including the lackof a “precise” transfer of control from thecopilot to the commander after the copilotlost control of the helicopter during theapproach in poor weather conditions. Fourseconds elapsed after the copilot’s request forhelp before the commander took control ofthe helicopter, the report said.

“The commander’s initial actions to recoverthe helicopter were correct, but the helicoptersubsequently descended into the sea,” thereport said.

The AAIB also cited “the approach profileflown by the copilot, [which] suggests aproblem in assessing the correct approachdescent angle, probably… because of thelimited visual cues available to him.”

The third contributing factor was thecompany’s failure to use “an appropriatesynthetic training device,” although one wasavailable, the report said. “The extensivebenefits of conducting training and checking insuch an environment were therefore missed.”

The Report said that the helicopter haddeparted at 1800 local time from BlackpoolAirport, a base for helicopter support for gasoperations in the East Irish Sea, for a plannedeight-segment flight to offshore gasproduction platforms operated byHydrocarbon Resources Limited (HRL).

The crew had flown a similar multi-segmentflight earlier in the day and had completed thefirst two segments of the accident flightwithout incident. As they began the thirdsegment, from the Millom West platform, fivepassengers boarded. Plans call for a seven-minute flight to the North Morecambeplatform to pick up a passenger and somefreight before continuing to another platform.

The helicopter left Millom West at 1826,climbed to 500 ft and accelerated to 125 kt.The

automatic flight control system was engaged,and the helicopter was in the normalstabilization mode for flight, the report said.Thecommander, the pilot not flying, confirmed thatlights on the North Morecambe platform wereproperly illuminated.

“Shortly after the 4nm [7km] GPS [globalpositioning system] call made by thecommander, the crew became visual with therig, and the copilot said, ‘I got the deck now,’”the report said. “Allowing for the speed of thehelicopter at the time, this equates to a visualrange of about 6,800m [4mi]. The commanderthen completed before-landing checks, whichincluding arming the floats.”

The helicopter was at about 270 ft when thecopilot announced his sighting of the platformbut climbed to just over 400 ft and then begananother descent.

The helicopter’s combined voice and flightdata recorder (CVFDR), which records fivehours of data and one hour of audio from thecommander’s, copilot’s and cockpit areamicrophones, at 1832:21, recorded thecommander saying, “You get no depthperception, do you?”

The copilot replied, “Yeah, not on this one, nottonight, no.” During this part of the approach,there were “steady increases in the collective,tail rotor input, cyclic pitch and cyclic rollinput.” And radio height decreased, thenincreased, the report said.

At 1832:33 – with cyclic pitch and roll inputsincreasing and oscillating, the collectiveincreasing at an escalating rate and thehelicopter pitching nose down and rolling right– the commander asked,“You all right?” and thecopilot answered, “No, I’m not happy, mate.”

As the combined engine torques exceeded100 percent, the commander asked,“We goinground?” and the copilot replied, “Yeah, take…help us out.”

The report said, “This request was not initiallyunderstood by the commander, and thecopilot reiterated his request, saying, 'Help usout'. The commander took controlapproximately four seconds after the initialrequest for help and said,“I’ve got it, I’ve got it,I have got it, I have control, I have control” Atthe time, the helicopter's right bank angleincreased to 38 degrees, its nose was about 38degrees down, indicated airspeed (IAS) was 90kt and increasing, and radio altitude was 290ft, with a descent rate of 2,000 fpm.

A second after the commander took control,the report said, “a large left cyclic roll inputwas made, followed one second later by an aftcyclic pitch input”. The helicopter's bank angleshifted to 7 degrees left, and pitch attitudeshifted to 13 degrees nose-down; as thehelicopter descended through 180 ft, IASincreased through 100 kt. Over the next sixseconds, IAS continued to increase; verticalspeed, which initially had been reduced to1,320 fpm, increased to 1,690 fpm.

Route of Accident FlightNote: 1. The helicopter’s track was derived from its combined voice and flight data recorder.Source: UK. Air Accidents Investigation Branch

9focus winter 09

‘You All Right?’

“At 1832:45, the copilot uttered an expletive,as though disappointed, and the commanderasked, ‘You all right?’; the copilot said, ‘Yep…no,’ in a resigned manner,” the report said. At1832:47, the automatic voice alert device,which provided audio warnings of thehelicopter's height above the surface, soundeda "100 feet” call.

The report described cockpit communicationsas "calm" and said that there were noindications of other problems. The helicopterwas last recorded at 30 ft in a 12-degree nose-down attitude, a 20-degree right bank and anIAS of 126 kt.The recording ended at 1832:50.

Witnesses on the North Morecambe platformtold investigators that the helicopter“appeared to be on a standard approach untilit appeared to initiate a go-around, althoughit seemed faster and closer to the platformthan normal,” the report said. The helicopterthen banked right and disappeared intodarkness before the witnesses heard animpact with the water.

The fuselage broke apart on impact, and mostsections of the helicopter sank. Rescue boatsarrived 16 minutes after the crash from amultipurpose standby vessel that was near theplatform. Bodies of six of those in thehelicopter were recovered, but the seventhwas not found.

The commander, who had flown helicopters inthe Morecambe Bay gas field for 20 years, wasthe base chief pilot, a line training captain anda crew resource management instructor. Hehad an airline transport pilot license and aninstrument rating, and had accumulated 8,856flight hours, including 6,156 hours in type.Records showed he had completed 34instrument approaches and 37 night decklandings in the 90 days before the crash.

The copilot had received helicopter flighttraining in the British Army and had flownemergency medical services helicopters for 21/2 years. He had been working for CHCScotia for 13 months at the time of theaccident and had 3,565 flight hours, including377 hours in type. He had 467 hours of nightflight—three of which were recorded in thethree months prior to the accident. He had

completed nine instrument approaches andseven night deck landings in the 90 daysbefore the crash.

The helicopter was manufactured byAérospatiale (now Eurocopter) in 1985 andhad accumulated 20,469 airframe hours and13,038 cycles. Records showed that it hadbeen maintained in accordance with anapproved maintenance schedule and was incompliance with all applicable airworthinessdirectives. Maintenance records for the 12months preceding the accident showed nodefects had been reported that related to thecrash. A routine 50-hour maintenance checkhad been performed the day of the accident,and no problems were reported.

‘A Particularly Dark Night’

Weather at the time of the accident includedvisibility of 3 to 7 km (2 to 4 mi) in mist andlight rain or drizzle, scattered to broken cloudswith a base at 700 ft, broken to overcastclouds with a base at 1,200 to 1,500 ft andsurface wind from 130 degrees at 15 kt. Aweather observer on a platform near theaccident site said that conditions about 90minutes before the accident included 4,000 m(2.5 mi) visibility in rain and skies obscured; anaccurate assessment of the cloud base wasnot possible because the observer did nothave appropriate equipment to measure it.

The report said that, although there was a halfmoon, the clouds completely obscured anylight from the moon, and “it was a particularlydark night.”

Data from the helicopter's integrated healthand usage monitoring system (IHUMS), whichincorporated the CVFDR, showed that nosystem fault warnings were activated duringthe accident flight. Two main gearboxexceedances were recorded—the first, whenthe combined engine torque exceeded 100percent at an airspeed below 75 kt, and thesecond, after the commander took the flightcontrols, when the torque exceeded 94percent with the airspeed above 75 kt.

Data also showed that, during the accidentsegment of the flight, the autopilot headinghold, IAS hold, altitude hold and areanavigation (RNAV) modes were not used.

Aérospatiale SA 365N Dauphin 2The Aérospatiale (now Eurocopter) SA 365N, first flown in 1979, is a twin-engine helicopterdesigned to carry two pilots and up to eight passengers. It is equipped with Turbomeca Arriel 1Cgas turbine engines, each rated at 530 kW (710 shp).

Empty weight is 2,017 kg (4,447 lb) and maximum takeoff weight is 4,000 kg (8,818 lb).Maximum cruising speed at sea level is 140 kt, maximum rate of climb is 1,515 fpm, and serviceceiling is 15,000 ft. Maximum range, with standard fuel at sea level, is 475 nm (880 km).

Source: Jane’s All the World’s Aircraft, UK Accidents Investigation Branch

focus winter 0910

Two Distinct Phases

The report said that, because there was noevidence of any technical problem,investigators focused on human factors issues“to understand why two experienced pilotswere unable to stop a serviceable helicopter[from] flying into the sea.”

Investigators identified “two distinct phases”of the final approach. The first involved a“steady reduction in collective demand and asteady, positive change in pitch attitude,” thereport said. The second — which began afterthe commander's callout of “fifty-five,” areference to airspeed — involved a steadyincrease in collective demand as thehelicopter began to climb, suggesting “achange in the appreciation of the helicopter'sposition or motion relative to the deck,” thereport said.

“The approach was flown essentially byreference to visual cues. In dark, overcastconditions, it is likely that some cues weredegraded or absent. For example, without adistinct horizon, the assessment of pitchattitude and approach angle (by reference tothe depression of the deck below the horizon)would be compromised.”

The report noted that if recommended changesin helideck lighting had been implemented,better visual cues might have been available,perhaps enabling the crew to determine earlierin their approach that they had deviated froma safe approach path. The recommendations—to be mandated by the International CivilAviation Organization beginning in 2009—callfor installing green lights instead of yellowlights on helideck perimeters as a means ofenhancing pilot situational awareness. Furthertrials by the U.K. Civil Aviation Authority (CAA)have led to the development of other helidecklighting patterns now being tested on offshoreplatforms1.

The report said that judging the approach angleapparently had presented the crew with asignificant challenge that might have been metby minimizing the number of variablesinvolved— “by commencing the descent at aspecified height and range, and maintaining astable pitch attitude and a fixed relationship tothe intended landing area”—or by usinginstrument references in addition to thelimited visual cues. However, the radioaltimeter was not in a location that enabled it

to be conveniently included in the copilot'sinstrument scan, the report said, and thecockpit voice recorder indicated that the crewwas not “using range information to determinethe initiation of the descent or cross-checkingwith height, and except for the “fifty-five” calland one height call at 400 ft, the commanderdid not provide any information that may haveassisted the copilot.”

“The nature of the copilot's difficulty is opento conjecture; he may have commenced thedescent too early or initially too steeply; or hemay have used an inappropriate controlstrategy or inadvertently changed the pitchattitude. The underlying causes, however,most likely stem from the limited visual cuesavailable and the paucity of instrumentchecks. Inadequate monitoring of theapproach by the commander must also beregarded as a contributory factor.”

The report also said that the commanderappeared "ill-prepared" to take control of thehelicopter and that both the go-arounddecision and the subsequent transfer ofcontrol to the commander appeared to havebeen handled inappropriately.

“It is possible that more positive crewinteraction and a more active participation inapproach profile monitoring by the non-handling pilot may have resulted in a positiveoutcome,” the report said.

Monitoring the Approach

The report included a safety recommendationthat CHC Scotia review its standard operatingprocedures (SOPs) for helideck approaches “toensure that the non-handling pilot activelymonitors the approach and announces rangeto touchdown and height information toassist the flying pilot with his execution of theapproach profile.”

The recommendation said that the non-handling pilot’s assistance is especiallyimportant when an SA 365N copilot is flyingan approach in poor visual conditions “andcannot easily monitor a poorly positionedradio altimeter”.

A second recommendation to the operatorcalled for a review of all SOPs concerninghelideck approaches flown by all of its types“with the aim of ensuring safe operations.”

Another recommendation called on theEuropean Aviation Safety Agency (EASA) toensure the prompt completion of research intoinstrument landing systems that would aidhelicopter crews in monitoring approaches inpoor visual conditions to oil and gas platforms.

A second recommendation to the EASA saidthe agency should investigate methods ofincreasing the conspicuity of immersion suitsworn by flight crewmembers. Rescuers hadtold accident investigators that the yellowimmersion suits worn by passengers of theaccident helicopter were easier to see thanthe blue suits worn by the pilots.

The AAIB also recommended that the CAA ensurethat recurrent training and checking of JAR-OPS(Joint Aviation Requirements-Operations), Part 3approved operators be conducted in an approvedsynthetic training device.

A second recommendation to the CAA calledon the agency to ensure that personnel whoconduct weather observations from offshorefacilities are “suitably trained, qualified andprovided with equipment that can accuratelymeasure the cloud base and visibility.” Thereport noted that the employee whocompiled weather data on the evening of theaccident had not received formal training andhad no equipment to aid in his observations.

After the accident, the operator providedmore specific procedures and guidance foractions to be taken in the event of pilotdisorientation or incapacitation; developedgo-around procedures that included use ofthe autopilot coupler; developed andpublished a night circuit pattern; andcontinued development of its policy to trainall pilots in synthetic training devices.

This article is based on AAIB Accident ReportNo. 7/2008: Report on the Accident toAérospatiale SA 365N, Registration G-BLUN,Near the North Morecambe Gas Platform,Morecambe Bay, on 27 December 2006.

Note

1. CAA. Enhancing Oshore Helideck Lighting, CAA Paper

2004/01.

Reprinted with kind permission of AeroSafety

World Jan 2009.

11focus winter 09

by Wayne Rosenkrans

Separation AnxietyImminent approval of software upgrade promises safer TCAS II collision avoidance system logic

Traffic alert and collision avoidance

system (TCAS II) Version 7.1 – a

software upgrade developed by European

and U.S. specialists – is expected to clear

one of the last technical hurdles on its five-

year path to operational readiness during

April. Possibly by mid-2010, the upgrade

installed in new TCAS II equipment will fix

two serious problems in today’s collision

avoidance system logic and make other

minor improvements. Strategic decisions

on whether civil aviation authorities will

recommend or require retrofitting Version

7.1 logic are pending.

One problem is that Version 7.0 logic does notreverse some resolution advisories (RAs)when a reversal is required to resolve thethreat of collision between two equippedaircraft while both are climbing or descendingwithin a vertical distance of 100 ft of eachother. The other problem is flight crews withvertical speed TCAS II displays manuvering inthe wrong vertical direction after receivingone of four “Adjust Vertical Speed, Adjust”(AVSA) RAs. AVSA RAs, now consideredambiguous by many safety specialists, advisea pilot to reduce the aircraft rate of climb ordescent to 0, 500, 1,000 or 2,000 fpm forcollision avoidance, and they lack any upwardor downward aural annunciation.

AVSA RAs have accounted for nearly two-thirds of all RAs in European airspace,occurring mainly in geometries involvinglevel-off at 1,000-ft altitude increments asassigned by air traffic control (ATC). Pilottraining solutions – for example, re-emphasizing that the proper response to anyAVSA RA is a reduction in vertical speed whilemaneuvering toward level flight – alone havenot worked, European specialists say.1

Version 7.1 solves the first problem with asignificant software code change hatmonitors compliance with RAs and enhancesthe reversal logic, allowing reversals when theaircraft are vertically within 100 ft.Version 7.1solves the second problem by replacing AVSARAs with a “Level Off Level Off!” RA.Independent validations by computersimulations with actual air traffic data fromseveral European sources, Boston and NewYork have demonstrated safe and effectivesoftware performance.2

A Eurocontrol recommendation in July 2008urged the industry to aggressively pursue thissoftware upgrade when revised U.S. andEuropean technical standard orders (TSOs) forTCAS II take effect. “As TCAS II Version 7.1provides further significant reduction in therisk of midair collisions, it is therefore stronglyrecommended that TCAS II Version 7.1 isimplemented as rapidly as possible,”Eurocontrol said.3

The organization’s policy position is that untilall current TCAS II-equipped aircraft and newaircraft are Version 7.1 compliant, there will beno short term reduction in the unacceptablerisk of midair collision to the Version 7.0-compliant aircraft in Europe, a risk equivalentto one midair collision every three years.Forward fit plus retrofit delayed not morethan two years would reduce this risk by afactor of four (ASW, 10/08, p.53), Eurocontrolsaid.4 Some European specialists say that nohardware modifications should be necessary,and they have proposed that InternationalCivil Aviation Organization (ICAO) standardsrequire TCAS II Version 7.1 equipage by Nov.30, 2010, for new aircraft and by March 31,2013, for existing aircraft.5

In December 2008, John Marksteiner, the U.S.Federal Aviation Administration (FAA)representative to an ICAO aeronauticalsurveillance working group, said that it wouldbe premature for ICAO to consider a timelinefor mandatory worldwide carriage of TCAS IIVersion 7.1 without further study.6 He citedseveral issues as still unresolved, includingdifferent risk levels in the United States andEurope, possibly time needed formanufactures to develop new equipment andretrofit packages, and an unknown scope ofhardware upgrades.

He raised other questions to consider. Willstandard cost-benefit analyses show thatrequiring Version 7.1 retrofit is justifiableinstead of clarifying the meaning of AVSA RAsand improving pilot compliance with Version7.0 RAs through training? How effectivelycould midair collision risk be mitigatedwithout Version 7.1 compliance by trainingpilots to climb and descend at less than 1,500fpm in the last 1,000 ft before level-off at theassigned altitude/flight level? Would analysisof RAs, based on pilot reports and monitoringof downlinked Mode S RA data, enable civilaviation authorities to identify “RA hot spots”

in their airspace and mitigate the Version 7.0shortcomings – with procedural changes,for example?

After TSO revisions for the Versions 6.04a-to-Version 7.0 logic upgrade were issued in 1999,the European Joint Aviation Authorities inJanuary 2000 mandated TCAS II Version 7.0carriage by all civil turbine-engine aircraftwith more than 30 passenger seats ormaximum takeoff mass of more than 15,000kg (33,070 lb). U.S. Federal AviationRegulations (FARs) for these aircraft currentlyrequire Version 7.0 or equivalent logic butallow version 6.04A Enhanced if that logic wasinstalled before May 1, 2003, and can berepaired to conform to its original minimumoperational performance standards.7

U.S. and European TSO revisions expectedduring 2009 will establish the dates whennewly identified or manufactured TCAS IIequipment must be Version 7.1 compliant.Steve Plummer, designated federal officialrepresenting the FAA at the March 12 meetingof RTCA Special Committee 147 (SC-147),offered no details but said that the FAA is nowevaluating what the appropriate strategyshould be for implementing Version 7.1,working on harmonizing rule-making strategywith the European Aviation Safety Agency(EASA) and, like others, proposing Version 7.1-related language for ICAO standards andrecommended practices. The RTCA meetingincluded representatives of its counterpart onTCAS II standards, the European Organisationfor Civil Aviation Equipment (EUROCAE)Working Group 75 (WG-75).

Last-Minute Modification

The TCAS II Version 7.1 revision to minimumoperational performance standards waspublished by RTCA as RTCA/DO- 185B in June2008 and by EUROCAE as Document ED- 143in September 2008. A post-revision validationprocess led to a delay in completing the TSOs,however, when a minor discrepancy came tolight between the pseudocode8 and statecharts.9 In one multi-aircraft scenario—thatis, involving more than two aircraft—run on astandard computer simulation program, theRAs did not agree. This led to moredevelopment, testing, multi-site verificationand validation of modifications issued asChange 1 to this standard.

12 focus winter 09

Change 1 eliminates the corrective green arcin TCAS II display symbology for a weakeningRA for the aircraft in the middle of a multi-aircraft encounter, according to an SC- 147working group report presented by AndrewZeitlin of The MITRE Corp. Center forAdvanced Aviation Systems. Validations byEurocontrol and Massachusetts Institute ofTechnology (MIT) Lincoln Laboratoryconfirmed that the modifications were safeand effective, Zeitlin said.

On April 21, SC-147 is scheduled to approveChange 1 to RTCA/DO-185B. Probably later inthe second quarter, the RTCA ProgramManagement Committee is expected toapprove this change, in turn enabling the FAAto issue TSO C119c,“Traffic Alert and CollisionAvoidance System (TCAS II) AirborneEquipment, TCAS II With Optional HybridSurveillance.” Parallel work in Europe includedEASA's March 12 issuance of Notice ofProposed Amendment No. 2009-03 similarlyupdating European Technical Standard OrderETSO-C119b.

FAA Monitors RAs

The FAA has been deploying monitoringsystems at 20 U.S. sites that collect data onTCAS RAs for analysis of both safety and airtraffic management. As of March, the systemswere operational in Boston, Los Angeles, NewYork and Philadelphia, said Neal Suchy, theFAA’s TCAS program manager during Version7.1 development.

This FAA analysis first has focused on businessjets operating below Class B airspace and RAsoccurring during multi-aircraft encounters, hesaid. Three more California sites—Ontario,Long Beach and Oakland—are scheduled tobe deployed by the end of May, and the FAAalso expects to monitor TCAS II performancenear Louisville, Kentucky, using automaticdependent surveillance-broadcast technologyin the nation's first Next Generation AirTransportation System (NextGen)environment.

During development of Version 7.1,Eurocontrol contractors used TCAS IIcomputer simulations to validate theperformance of the AVSA RA-relatedenhancements.They first were compared withversion 7.0 using aircraft encounter data fromEurope. The effort comprised safety aspects,human factors aspects and operationalaspects.

After reviewing the European results, however,RTCA SC-147 specialists wanted to confirmthat AVSA related enhancements would notdisrupt FAA terminal control area operationsor induce a conflict with a third-party aircraftflying near a TCAS II-equipped aircraft, giventhe country’s dense mixes of air carrier andgeneral aviation traffic operating underdifferent flight rules. In response, aEurocontrol analysis identified 92 initial AVSARAs among a total 992 RA encounters fromBoston-area data recorded by MIT LincolnLaboratory, with 81 AVSA RAs suitable fordetailed study.

These RAs occurred during six months of2006 within a 60-nm (111km) radius ofBoston Logan International Airport, and theEurocontrol contractors received both FAAradar data and RAs downlinked by MIT from aMode S transponder sensor. About half of therecorded AVSA RAs involved two aircraft; theremainder involved three to seven aircraft inthe surrounding traffic.

This analysis found that the AVSA relatedchanges in TCAS II Version 7.1, assuming thatall aircraft in the airspace were equipped alike,would generate one “Level-off, Level-off” RAabout once every three days in the Bostonairspace compared with an average of 18 RAsof all types recorded every three days. Thenew “Level-off, Level-off” RA did not induce aconflict with any third-party traffic, and thelikelihood of such a conflict was deemed“extremely remote.”

Eurocontrol contractors next looked at threemonths of 2007 FAA radar data fromrecorded aircraft encounters that occurredwithin a 60-nm radius of John F. KennedyInternational Airport. They did not havedownlinked Mode S transponder dataavailable from this airspace, so RA data wereextrapolated based on an assumption that theaircraft were fitted with TCAS II operating inRA mode as required by current FARs.

TCAS II Version 7.1 Solution to Pilot ErrorTCAS = traffic alert and collision avoidance system; RA = resolution advisory; FL = flight level

Note: Current TCAS II logic allows only one climb/descend sense reversal, and reversing anongoing RA is not permitted while the aircraft are maneuvering within a vertical distance of 100ftof each other. The illustrated enhancement in the new Version 7.1 logic is that if the aircraft withthe red flight path decends contrary to a “Climb” RA, immediate reversal RAs will be generatedfor pilots of both aircraft.

Source: Eurocontrol Safety Issue Rectification Extension Plus Project

Pilot-Friendly Benefits

Eurocontrol, its research contractors, otherEuropean aviation organizations and the FAAexpect introduction of the “Level-off, Level-off” RAs in TCAS II Version 7.1 to be welcomedworld-wide. The Version 7.0 logic had beendesigned with an expectation that pilots ofconverging aircraft would becomecomfortable ensuring initial separation solelyby simultaneously modifying their presentclimb/descent rates rather than climbing,descending or leveling off. In such scenarios,however, todays TCAS II may direct one flightcrew to reduce climb rate from, say, 2,500fpm to 1,000 fpm in about three seconds,Eurocontrol noted. Unlike that scenario,intuitively simple “Level-off, Level-off” RAswill be of shorter duration and typicallyinvolve less altitude change.10

“In the same geometries, the Version 7.0 logiccan post increasingly stronger AVSA RAs,possibly up to a positive RA, in quicksuccession if the vertical convergence rate isnot decreasing as fast as expected, whichconstitutes a complex RA sequence.”said theEurocontrol report on New York airspace

simulations. “With [Version 7.1 logic], thiscomplex sequence can be replaced by a singleLevel-off RA, as it is more efficient in rapidlyreducing the vertical convergence.”

For ATC, one of the main safety benefits ofVersion 7.1 will be that pilots receiving RAswill not continue in the same verticaldirection, Eurocontrol said. A conclusion fromits analysis of Boston data was that “TCAS IIVersion 7.0 issued RAs that left both aircraftevolving in the same vertical direction and,despite appropriate pilot responses, the targetvertical separation of 350 ft was not achievedat their closest approach.”

Notes1. Arino, Thierry; Chabert, Stephan; Drevillon, Herve.

“Decision Criteria for Regulatory Measures on TCAS IIVersion 7.1.” Eurocontrol Mode S Programme, DSNAand Egis Avia Safety Issue Rectification Extension Plus(SIRE+) Project. July 17, 2008.

2. Arino; Drevillon. “Operational Performance of CP115 inNew York Airspace” Dec. 4, 2007.

3. Arino; Chabert; Drevillon.

4. Eurocontrol. “TCAS II Version 7.1” <www.eurocontrol.int/msa/public/standard_page_ACAS_Upcoming_Changes.html>.

5. Loscos, Jean-Marc; Mallwitz, Roland; Potier, Eric; Woods,Ray. “Update to ACAS Mandatory Carriage.” An ICAOworking paper presented in Montreal to theAeronautical Surveillance Panel Working Group of theWhole. Document ASP-WGW/1-WP/12. Nov. 24, 2008.

6. Marksteiner, John. “ACAS Carriage Considerations.” AnICAO working paper presented in Montreal to theAeronautical Surveillance Panel Working Group of theWhole. Document ASP-WGW/I-WP/17. Dec. 5, 2008.

7. Eurocontrol ACAS [Airborne Collision Avoidance System]Programme; Sofreavia; CENA. “European Maintenanceof TCAS II Version 7.0: Project EMOTION-7 FinalReport.” Report no. ACAS/03-003. January 2003.

8. Pseudocode is an informal structured language thatprogrammers use to convey to other people high-leveldescriptions of computer programming algorithms.

9. State Charts in RTCA/DO-185B are tables showing atransition, code location, trigger event, true/false statusof conditions and output action in the collisionavoidance system logic.

10. Arino; Drevillon. “Operational Performance of CP115 inBoston Airspace.” March 9, 2007.

Reprinted with kind permission of AeroSafety

World April 2009.

focus winter 09

Paymentsby PAYPALSubscription can now

be paid using PayPal. All

that is required is that

you advise us of your

email address, we will

then issue an invoice for

you to pay with your

credit card through our

PayPal Account.

You do not need

a PayPal Account

of your own.

There will be a

Transaction fee added

to the invoice amount.(approx 3.5%)

13focus winter 09

Paymentsby PAYPALSubscription can now

be paid using PayPal. All

that is required is that

you advise us of your

email address, we will

then issue an invoice for

you to pay with your

credit card through our

PayPal Account.

You do not need

a PayPal Account

of your own.

There will be a

Transaction fee added

to the invoice amount.(approx 3.5%)

focus winter 09 15

At Last: The answer to the Meaningof Life,The Universe and Everything…by Major C P Evans, MoD DARS SO2 Engineering Rotary Wing - Directorate of Aviation Regulation and Safety

The answer is, as some of you may

already know, ‘42’ however, that isn’t

terribly helpful. As it happens it wasn’t

terribly helpful to Arthur Dent either.

Despite everybody knowing the answer,

the types and numbers of incidents caused

by maintenance error are not decreasing.

As a result these continue to cause a

reduction in operational capability and

incur costs, both in terms of components

and rectification man-hours.

Maintenance Error and Regulation

- an Evolution

The DARS engineering team has been lookingat the data available covering maintenanceerror (The term ‘maintenance error’ has beenused generically to refer to all types ofaircraft-related error which occur while, ororiginate when, the aircraft is not beingoperated by qualifi ed aircrew, and includesmaintenance/servicing error and groundhandling error. It also includes errors made ororiginating on uninstalled aircraftcomponents). We have attended meetings ofthe UK Flight Safety Committee, the FlightSafety Working Group and the HumanFactors Group (Engineering) of the RAeS.

Additionally we have looked at a number ofdata sources: CHIRP MEMS - CAA MandatoryOccurrence Reports - US Federal AviationAdministration (FAA) Service Diffi cultyReports - Australian Transport Safety Board(ATSB) Air Safety Reports and our ownPANDORA database. It has become clear thatthe two major causes of both civilian andmilitary maintenance error are:

The data sets that are available identify thesesame issues but store and manipulate data indifferent ways. If progress is to be made onreducing and keeping down maintenanceerror a better way of collecting and using thisdata is essential. After an evaluationcompetition the contract for an all newAviation Safety Information ManagementSystem (ASIMS) has been let. This willsupersede all the current Signals basedoccurrence reporting systems. Therequirement also includes the need forcollection, correlation and use ofengineering/maintenance occurrence reports.These will utilize the Maintenance ErrorDecision Aid (MEDA) and support aMaintenance Error Management System(MEMS) as well as being compatible withfuture Error Management Systems.

Historically though, the immediate reactionto an incident or accident has been increasedregulation or procedures. However, there hasbeen little or no corresponding reduction inmaintenance errors. This pres criptiveapproach is now seen by the DARSEngineering Branch as counter-productive.There seems to be a corporate belief thatever-increasing procedures andcorrespondingly lengthy publications improvesafety. This is giving a false sense of security.An example to demonstrate the fallacy ofover-regulation is the Dutch town ofDrachten. In a road traffi c experiment theyhave removed the town’s traffi c lights, cyclelanes, pavement barriers, white lining etc. Theaccident rate has tumbled and congestion hasbeen eased. Drivers, cyclists and pedestriansare no longer blindly following the regulationsand have to think for themselves. Closer tohome, railings have been removed from somepavements along Kensington High Street. Thishas dramatically reduced pedestriancasualties. These outcomes are not surprising.Drivers and pedestrians were concentratingon signs and barriers to keep them safe; withtheir removal, they have to focus on whatother road users are doing.

1. Incorrect installation of components; often due to poor marking or ease of wrong assembly.2. Failure to follow laid-down procedures; often due to either incomprehensible or inaccurate publications, or the unavailability of the

correct publications.

Looks like we put the gearbox in back to front

focus winter 0916

There is a read across to MOD aviation. Withso many rules and procedures, no person canhope to be current on them all, so they tendto scan the sections that appear important.When they do need the laid-downprocedures, they are often not available at thework place, difficult to use, or inaccurate.Likewise they can never be reviewed oftenenough to keep them current. The requiredturnaround for MOD F765 publicationamendment is 3 months (iaw JAP 100A 01).This process can, in some cases, actually takeup to 3 years, with little or no feedback beinggiven to the originator.

Allocation of Resources

The enduring dilemma, or mismatch, ofengineering resources is illustrated below.This is particularly relevant to the SupportHelicopter (SH) and Air Transport (AT) fleetsdue to increased operational tasking:

Number of Aircraft Engineers x ProductiveWorking year = X Engineering Hrs

X Engineering hrs / Maint Man Hrs per FlyingHr (MMH/FH) = Y Airframe Hrs

However,

Number of Pilots x Currency or Taskingrequirement = Z Flying Hrs

Z is always much bigger that Y - i.e. SystemicEngineering undermanning.

As an example, the Apache was procured onan estimated 8-12 MMH/FH basis and REMEWorkshops were established against this figure. The recently rebalanced Forward/DepthREME Workshop establishments werepredicated on 17.8MMH/FH. By April 2008the actual figure had risen to very nearly30MMH/FH. While this figure has promptedsome action and mitigation, theestablishments remain extant and thesystemic undermanning stands.

Service personnel generally work in anenvironment of conflicting demands betweenmaintaining their skills as a tradesman andskills as a serviceman; which often takes

secondary importance. Working routines thatproperly accommodate training for Ops, foraircraft engineers, often conflict with theongoing requirement to provide serviceableaircraft. When a Sqn is training for Ops, theengineers are under considerable pressure,which continues on Ops and even intorecovery. The level of pressure, both actualand perceived, is often different depending onthe hierarchical position of the individualconcerned. Whilst some pressure is oftenneeded to achieve good levels of arousal,perceived pressure regularly results in amisplaced ‘Can Do’ attitude. Breaking thiscycle is vital; remembering that under-arousalcan be just as insidious in maintenance errors.

Perceived pressure is an often-quotedcontributory factor in ground handlingincidents, when Service personnel violate laid-down procedures to ‘get the job done.’Ground handling incidents are a major issuein the AT fl eets where many types of GSEattend the aircraft during turnarounds. Duringthe period Jul 05 to Oct 07 there were 75ground handling incidents at a particular ATunit. That is 3 incidents a month, indicatingthat perceived pressure, non-adherence toprocedures and a ‘can do’ attitude are veryreal issues.

Error Provocation

An important error-provoking factor, as aresult of systemic or even perceivedundermanning and reduced experience levels,is the creation of routines and norms –however unsafe – that support Ops. One ofthe differences between Ops and normalpeacetime workload can be distraction, giventhat in the 24/7 world of Ops all are heavilyfocused on the operation in hand.A stream ofdomestic, administrative, trivial and otherdistractions punctuate the normal peacetimeworking day. However, lack of in-theatre partscan be as concerning and lead tomaintenance errors, often from shortcuts, inan attempt to achieve the tasking.

The Effects of Operations on Rotary WingFlight Safety1, while concentrating to a largeextent on the Aircrew issues, is highly

pertinent to maintainers’ error-provokingfactors and, especially in light of recentaccidents, applies equally to fixed wing andmulti-engine aircraft.

European Helicopter Safety Analysis Team

(EHSAT)

DARS has been conducting a review ofmilitary helicopter accident Board of Inquiry(BoI) Reports, against the EHSATmethodology. This attempts detailedclassification of the organizational andsupervisory factors, as well as any unsafe actsor preconditions, of an accident; using theHuman Factors Analysis and Classifi cationSystem (HFACS).

The BoI Reports show, in addition to theactual cause of the accident, that there arealways many more contributory factors andobservations. These, in an open reportingculture, should have become Incident orOccurrence Signals and/or DCORS Reports;however these ‘near misses’ often only seemto surface during a BoI. The indication is thatwe are not operating in the open or reportingculture that we need.

Maintenance Error Management System

(MEMS)

Since Jan 03, JAR 145 (subsequently EASAPart 145) civilian organisations have beenrequired to ‘…establish an internal occurrencereporting system…’ The CAA’s MandatoryOccurrence Reporting (MOR) andConfidential Human Factors IncidentReporting Programme (CHIRP) both havemilitary equivalents, and yet civilianregulation goes further, mandating theadoption of a MEMS. Since Mar 07 the CAAhas sought, through Airworthiness Notice No71, ‘…to provide an environment in whicherrors may be openly investigated in orderthat the contributing factors and root causesof maintenance errors can be addressed usinga system that would complement, notsupplant, their current reporting systems.’ Inshort, a MEMS is now a mandatoryrequirement within civil aviation.

Apache Helicopter Depth Support Unit

(AHDSU) – Introduction of a MEMS

As a result of a major ‘near miss’ incident theAHDSU introduced a Maintenance ErrorManagement System (MEMS) in Nov 07. Thishas drawn together Self Assessment, HFtraining and the Maintenance Error DecisionAid (MEDA). Baines Simmons Ltd was thechosen provider. It should be noted howeverthat CAA Self Assessment tools and MoDsponsored HF training are available.Additionally MEDA is an open source (MSbased) investigation tool originally developedby Boeing, although investigatory techniquestraining is required to use this.

While many would hope for an Open and JustCulture, the MoD does not always have aReporting Culture, despite the availability ofmany open and confi dential systems. WhileDARS training and instruction refers to a‘misplaced’ can-do attitude, there seems tobe an almost institutionalized can-do attitude– which frequently leads to no-one reportingwhen a situation becomes unsafe. There havebeen very few DCORS (formerly Murphys,

CONDORS, Anymouse or Acorns) submittedto the DARS, from across Defence, this year;none of which have been raised by anengineer. In sharp contrast the AHDSU hasreceived over 180 FURBYs (In-housereporting scheme) since Nov 07.

Having delivered HF training to everybody inthe AHDSU, the Depth Support Managerbrought in a MEMS and his Policy Statementintroduced a just culture for the managementof maintenance error. This explicitly requires awillingness to report hazards and a process fordealing fairly with personnel, both civilian andmilitary. The AHDSU Culpability Chart deals,diagrammatically, with everything fromsabotage and malevolent damage – throughprogressively diminishing culpability – tosystem-induced and blameless errors.

With only 200 mixed civilian and militarypersonnel, the AHDSU reporting rate isexcellent. The formal MEDA investigationscarried out so far have produced some clearefficiency and safety results, including:highlighting the wasted effort in cannibalizingAH doors (as a result of the way they are

17focus winter 09

I’ve checked that the checker checked has been checked so can you check it now?

manufactured and fitted); and the easily-made error, and subsequent impact, of fittingan APU air valve 180? out-of-alignment, aswell as discovering that subsequent follow-ups are impossible. It is now hoped that thisinitiative will be rolled out across the AH forceand beyond.

ASG Study

The RAF ASG have carried out a review ofrecent maintenance-related RAF F765B FlightSafety Investigation Reports. This reviewhighlighted that significant improvementscould be made to the investigation process toenable a better understanding of the error

provoking conditions surrounding flight safetyincidents, especially at the organizational level.Additionally, where error-provoking conditionshad been identified, in around 50% of casesthese had not been addressed. This hashighlighted the need for a fully integrated ErrorManagement System within the RAFmaintenance organization. The RAF are nowdeveloping an Error Management System(EMS), utilizing the services of Baines SimmonsLtd. The ASG are working closely with DARS toensure that any solution offered is cognizant ofthe future intentions for MoD-wide policy.

The RAF’s HF (M)EMS programme is to berolled out simultaneously across most aircraftfleets at 12 MOBs2. This ambitiousimplementation strategy will be reviewedafter the results of the diagnostic phase3 areknown. DARS staff will continue to monitorthe implementation and support the RAFproject where necessary.

Conclusion

Aircraft engineers, across military aviation, areoverburdened with regulation. Where thereare resource, or other error-provokingconditions, maintenance errors becomeinevitable. The burden of regulation needs tobe addressed. The introduction of ASIMS andMEMS to the current procedures and someform of MEDA could and should improvematters. However; there is a long way to goand there are NO quick fixes.1 DARS 3/1/5 dated 3 Apr 07.

2 Marham, Lossiemouth, Leuchars, Coningsby,

Cottesmore/Wittering, Odiham, Benson, Waddington,

Kinloss, Brize Norton, Lyneham and possibly Northolt.

3 The diagnostic phase commenced in Sep 08 and

assesses the existing organizational and individual

attitudes to maintenance error.

This article originally featured in the DARS

Aviate Journal. Winter 2008/9

18 focus winter 09

19focus winter 09

Using Weather Radarto Counter FL Cellsby Erik Eliel

It’s what you can’t see that can hurt you.

Near LBL, [the crew] saw a patch of blue

sky to the right front and painted

nothing [on radar] in front of them. The

encounter occurred when a large buildup

appeared in front of the airplane with less

than two seconds notice. The NTSB report

goes on to say “the airplane experienced

airspeed excursions from about 275 knots

to 225 knots with an altitude loss of 500

feet. During the encounter, the airplane also

experienced small hail. Other aircraft in the

area reported no conditions greater than

light turbulence.”

Every now and then, a professional flight crewinadvertently penetrates the top of athunderstorm during cruise flight; in theforgoing incident, the pilots were cruising at FL370. With a little bit of knowledge, however,this is almost always avoidable.

To understand how these events can happen, abasic review of thunderstorm anatomy andweather radar capability is necessary.

As FAA Advisory Circular 00-24B warns, "Athunderstorm packs just about every weatherhazard known to aviation into one viciousbundle," including moisture in just about everystate of existence you can imagine. Althoughthere are no absolutes when it comes tothunderstorm anatomy, it simplifies thisdiscussion to separate the vertical structure ofa typical U.S. thunderstorm into two sections.

The lower and mid parts of such storms tendto be composed of "wet" droplets of adequatesize to react sufficiently with radar energy, solumping those two sections together theybecome the reflective portion of a cell. To apilot, this part of a cell will be seen on thedisplay as some combination of green, yellow,red and, for most business aircraft radardisplays, magenta. There is an imaginaryhorizontal line at the top of this area that isreferred to as the "radar top," because it is theupper vertical boundary of detectablemoisture by aircraft weather radar.

The topmost section—above the radar top—then, is different. Here, moisture also exists, butcommonly in a dry, frozen state that interactspoorly with aircraft radar energy. Do notconfuse the radar top with the actual top, sincethe latter normally will be several thousand

feet higher. The significance of this is that thesame hazards—turbulence, hail, lightning,etc.—exist in the upper part as in the lowerreflective part, but encounters typically catchpilots off guard because the radar is capable ofproviding little or no warning.

In order to successfully avoid thunderstorms,you have to see them. If flying in IMC or atnight, observing them out the windscreen maynot be possible, leaving radar as thepredominant tool. Unfortunately, the upperdry, frozen part of a thunderstorm typicallycoincides with the cruise altitudes of high-performance jet aircraft. Whether or not a cellwill be seen is a function of tilt management.Specifically, the antenna tilt must be set sothat radar energy is "aimed" at, or interactswith, the detectable moisture in the reflectivepart of the cell.

Frustrating Physics

The ability of a radar system to focus theenergy is governed by physics, specifically thesize of the antenna. Small antennas, like thoseon aircraft, complicate the situation becausethey broadcast a wider, more diffuse beam —an undesirable characteristic, but a limitationpilots are stuck with. Larger antennas focus theenergy better, which is desirable.

Almost every pilot who has flownradarequipped aircraft has probably heardabout the common tilt technique, handed

down through the generations: Lower the tiltuntil ground returns are displayed, and thenraise it just enough so they mostly disappear.For an aircraft cruising in the upper flight levels,this is a classic setup for penetrating the top ofa thunderstorm. To explain how this happens,let's use specific numbers.

Relatively speaking, all aircraft antennas aresmall, but the smallest of the antennas furtherexacerbate the situation. A common size forradars in today's business aircraft is 12 inches,which yields approximately an eight-degreebeam. At 100 miles this translates to a beamdiameter of about 80,000 feet. If the antennais properly aligned, using zero degrees tilt putsthe center of the beam approximately co-altitude (disregarding curvature of the earth)leaving about 40,000 feet worth of beamabove, and an equal amount below, theaircraft.

An aircraft cruising at FL 400 with a 100-milerange selected would typically observe a fewground returns at the perimeter of the display.Thunderstorms at 100 miles would be easilydetected, as radar energy would be interactingwith the reflective part in the lower to midaltitudes. The associated pattern of colors—green, yellow, red and possibly magenta—givethe active sections an unmistakable signature.However, if the tilt remains at zero degrees, theintensity of the cell may appear to weaken asthe aircraft approaches it. If that seemscounterintuitive, think about the bottom of

Traditional tilt settings that eliminate ground returns on longer display ranges are classic setupfor an inadvertent penetration while cruising in the upper flight levels. Here, the distant cellwould be displayed since the bottom of the beam is below the radar top. However, the closer cellis being over-scanned and would not be detected by the radar. By lowering the tilt, both cellswould be displayed and easily avoided.

20 focus winter 09

the beam. As the distance to the celldecreases, the bottom of the beam scansprogressively higher and higher in the cell.

Twenty miles from the cell, the total width ofthe beam would be only about 16,000 feet,and with the tilt still set to zero degrees, thebottom of the beam would be 8,000 feetbelow the aircraft, or about FL 320, and at adistance of 10 miles would be approximatelyFL 360. There is a high risk that these altitudeswill be at or above the radar top for manyhazardous cells, meaning they might appear aseither a weak green return or perhaps noreturn at all. Either way, it is deceptive andpenetration of either the frozen top or thehazardous area just above the actual top isprobable if it is in the aircraft’s flight path.

As pilots, we’ve been taught to remain at least20 miles away from storms identified as severeor that give an intense radar echo. Anincreased margin of safety is appropriate insome situations, so assuming the cell can beseen (visually or on radar) a turn to stay atleast 20 miles away should always be the goal.But in day-to-day professional flying,classroom theory and real world dynamicsoccasionally clash, and a pilot cruising in theupper FLs, looking at a weak green return onthe display may not recognize its significance.If it is part of a line of weather bracketed oneither side by more intense cells, forgingdirectly toward it might appear to be areasonably safe plan. Do not fall into this trap.

Minimizing the Risk

The solution is proper tilt management.The tiltmust be “aimed,” or lowered sufficiently so theradar energy interacts with detectablemoisture in the lower to mid sections of thecell as the distance to the cell decreases.However, with longer ranges selected, groundreturns may flood the display, and if there aretoo many, that can be a real nuisance.

But the benefit of having a solid band ofground returns on the outer portion of thedisplay is significant. Archie Trammell hastaught this technique to countless numbers ofpilots including me, and I have used it duringthousands of flying hours. It guards against oneof the biggest threats pilots face: failure toidentify severe attenuation caused byprecipitation. By having a solid band of groundreturns displayed —no more than about one-

third of the outer portion of the display—total attenuation can be readily identifiedwhen ground returns are absent from thearea directly behind a storm. This is a classicradar shadow. Failure to identify them andthe cells intense enough to cause them hasbeen causal in numerous aircraft accidentsover the past 30 years.

There are three absolutes when dealing withcells this intense. One, without exception, isthat they are very hazardous. Two, the trueshape and gradient of the cell may be grossly

misrepresented on the display because radarenergy is not completely penetrating thestorm. (If there do happen to be any weakweather returns in the shadow, the intensitywill be significantly understated.) Three, radarenergy is unable to make the two-way tripthrough the offending cell, and so anything inthe shadow area, where the most intenseweather may reside, can remain cloaked.

The solution to unwanted ground returns thatare displayed is learning how to read them.Operating the radar with a band of groundreturns on the outer portion of the display onclear weather days will provide a "mentalblueprint" of what normal ground returns looklike. Cities, mountains, the plains and bodies ofwater all have unique signatures. Just as adoctor knows what constitutes "normal" wheninterpreting an x-ray image, a pilot must knowwhat constitutes "normal" the ground returns.Once this level of proficiency is achieved,deviations from the norm — like abnormalitieson an x-ray — will be more readily identified.

So, even though cells might be embedded inground returns, the signature/profile of intenseweather is unique and difficult to miss. If thereis any confusion about which returns are fromthe ground vs. those from a thunderstorm, bepatient and alert. Thunderstorms will self-identify when they march out of the groundreturns on the display.

As already mentioned, a small antenna yields awider beam. Setting the tilt to zero or minusone degree may begin to flood the display withground returns as longer ranges are selected.The temptation will be to raise the tilt toeliminate them, but a risk managementapproach suggests doing otherwise. Someconsiderations follow.

Using a shorter selected range with sufficientdown-tilt is a good option when weather isanywhere near the aircraft. Details—anddetails are good—are easier to see on shorterdisplayed ranges. Keeping a band of groundreturns on the outer perimeter of the displaywill be more easily managed also. This reducesthe risk of overscanning close-in targets. Thedownside: When converted to time, shorterranges often don't provide sufficient warningof approaching weather. No one likes weathersurprises, but exclusively selecting a shorterrange will provide them.

Typical ground returns given a cruisingaltitude of FL 350, a 12-inch antenna and zerodegrees antenna tilt. Raising the tilt willreduce or eliminate the ground returns, butdoing so significantly increases the risk of over-scanning close-in cells – the ones capable ofcausing damage and injury. Strong return at10 o’clock and 70 miles is Las Vegas.

Taken at FL 370, return at 2 o’clock and 20miles appears weak because the bottom ofthe beam is approaching the radar top. Notethe tilt is set to zero degrees, 80-mile rangeselected, with no ground returns displayed.

focus winter 09 21

Exclusively using a longer range with a tiltsetting that eliminates ground returns maygive more notice of approaching weather if itis both tall and reflective, but generallyspeaking, it is a bad idea. Weather 150 milesaway can't hurt you or your aircraft, so eventhough knowledge of its existence isimportant, it is the close-in weather—the stuffthat can cause damage and injury—that takespriority. Using long ranges with the tilt set sothat no ground returns are displayed is a setupfor an inadvertent penetration.

Perhaps the most obvious technique then, is toperiodically alternate the tilt and range. This

may slightly increase pilot workload, but it isnot nearly as workload intensive as diverting tothe nearest field with aircraft damage andinjured passengers. The default setting whenweather is anywhere near the aircraft is ashorter range—say, 50 miles to 80 miles—and a tilt setting that puts ground returns onthe outer one-third of the display,approximately. As hazardous weather intrudesinside this distance, reduce the rangesufficiently to reveal adequate detail. But everynow and then, select a longer range with aslightly elevated tilt for a few sweeps to seewhat else is out there. This should providesufficient notice of approaching weather.

If you are lucky enough to have a configurationthat allows each pilot to independently controlthe radar tilt and gain for his or her respectivedisplay, then consider setting one side tolonger range with a tilt setting that does notflood the display with ground returns, and theother side to a shorter range and a lower tiltsetting as described previously. Providedweather is the biggest threat, thisconfiguration can provide the best of bothworlds. The downside is that the update rate isusually reduced to alternating sweeps. Withhazardous weather anywhere in the area,information should be updated every sweep.Also, if higher priority information should bedisplayed — terrain for instance — it may notbe a good idea to use more than one displayfor weather radar information.

Finding Your Strategy

How will you know which strategy will workbest for you or your flight department? Bytaking advantage of the learning opportunitiesthat present themselves on the easy days ascells approach and then pass harmlesslyabeam. Observe how cells look when they areembedded in the ground returns and then howthey eventually march out of the groundreturns. Learn to read the ground returns andobserve radar shadows. Note how a celllocated behind an intervening cell may beunderstated, weaker than its true intensity(radar energy had to make a two-way tripthrough the intervening cell).

Also, try using “traditional techniques” andnotice how hazardous cells will often weakenin intensity on the display as the bottom of thebeam begins to approach the radar top. Notehow lowering the tilt minimizes overscanning.All flights in good weather are analogous to ascientist's laboratory — an opportunity to trydifferent things, learning what does and doesnot work so well.

Successfully avoiding cells at night or in IMChinges on your learned skills. As with any otheractivity, good intentions don't guaranteesuccess. The only successful strategy forminimizing these weather risks will result fromtraining and practice. And even thoughmistakes will be made along the way, the key isto make them when the consequences areminimal and then to learn from them.Irrefutably, one of the biggest mistakes to bemade is to wait until the outcome of thedecision is critical — the threat to the aircraftor those within is at hand — before turningthe radar on and start learning.

Remember, in spite of the fact that convectiveweather can grow at a rate of several thousandfeet per minute, thunderstorms do not justsuddenly appear at cruise altitudes “with lessthan two seconds notice.” You just have to knowhow and where to look for them with the radar.

Reprinted with kind permission of Business &

Commercial Aviation June 2009.

View from the windscreen tells a different story. It was a very hazardous cell extending slightlyabove our cruising altitude and growing rapidly.

Lowering the tilt reveals the actual situation.The threat was not an isolated cell, but rather aline of hazardous weather masked by otherclouds. Even though a band of ground returns isnow visible at the outer perimeter of thedisplay, embedded weather is still visible (patchof yellow returns at 1 o’clock and 60 to 80miles). Note also how the cells self-identifywhen they march out of the ground returns.

22 focus winter 09

Fatigue and take-off data…a lethal combination

At 00.03 local time MK Airlines flight

1602 departed Windsor Locks-Bradley

International Airport (BDL) for a flight to

Zaragoza, Spain with a cargo of lawn

tractors. An intermediate stop was made

at Halifax (YHZ), where the aircraft landed

at 02.12 local time. At YHZ the aircraft was

loaded with 53,000 kilograms of lobster

and fish. After fueling, the total fuel load

was 89 400 kg.The planned take off weight

was 353 000 kg. The Boeing Laptop Tool

(BLT) was then used to calculate the take-

off speeds. Since the software was last

used before the take-off from Bradley, it

still contained those figures

The airport information and weather waschanged to Halifax, but somehow the take-offweight was not changed and remainedshowing 240000 kg. Take-off performancedata were generated, resulting in incorrect Vspeeds and thrust setting being transcribed tothe take-off data card. It is most likely that thecrew did not adhere to the operator’sprocedures for an independent check of thetake-off data card, so the erroneous figureswent unnoticed.

It was dark, but the weather was fine as theairplane was cleared to taxi to Runway 24 (8800ft / 2682 m long) for departure. After push back,the aircraft began to taxi, the flaps wereextended to 20°, and the horizontal stabilizerwas set to 6.1 trim units, where it remained forthe duration of the flight. The flight controlchecks were completed during the taxi. Theaircraft entered Runway 24 at Taxiway Deltaand backtracked to the threshold.

The aircraft then made a 180° turn to theright and, upon lining up with the runway, thethrust levers were advanced and a rollingtake-off was commenced at 06:53:22.

At the start of the take-off roll, the thrustlevers were smoothly advanced from groundidle thrust (approximately 1.0 EPR) to take-offpower with all final EPR settings indicatingbetween 1.3 and 1.33.The aircraft acceleratedthrough 80 KCAS (06:53:46) approximately1800 ft (550 m) from the threshold.

At 130 KCAS, the control column was movedaft to 8.4° to initiate rotation as the aircraftpassed the 5500-ft (1680 m) mark of Runway24 (3300 ft / 1010 m of runway remaining).The aircraft began to rotate.The pitch attitudestabilized briefly at approximately 9° nose-up,

with airspeed at 144 KCAS. Because the 747still had not lifted off the runway, the controlcolumn was moved a further aft to 10°, andthe aircraft responded with a further pitch upto approximately 11°; initial contact of thelower aft fuselage with the runway occurredat this time.The aircraft was approximately atthe 8000-ft (2450 m) mark and slightly left ofthe centreline. The control column wasrelaxed slightly, to 9° aft.

The pitch attitude stabilized in the 11° rangefor the next four seconds, and the lower aftfuselage contact with the runway endedbriefly. With approximately 600 ft (185 m) ofrunway remaining, the thrust levels wereadvanced to 92 per cent and the EPRsincreased to 1.60. With 420 ft (130 m)remaining, the lower aft fuselage contactedthe runway a second time.

As the aircraft passed the end of the runway,the control column was 13.5° aft, pitchattitude was 11.9° nose-up, and airspeed was152 KCAS.The highest recorded nose-up pitchof 14.5° (06:54:24) was recorded after theaircraft passed the end of the runway at aspeed of 155 KCAS. The aircraft becameairborne approximately 670ft (205 m) beyondthe paved surface and flew a distance of 325ft(100 m).

The lower aft fuselage then struck an earthenberm supporting an instrument landingsystem (ILS) localizer antenna. The aircraft’stail separated on impact, and the rest of theaircraft continued in the air for another 1200ft (370 m) before it struck terrain and burstinto flames.

Airline expansion

MK Airlines Limited has grown significantlyduring its relatively short history. Thecompany’s commercial success andsubsequent expansion increased demands inits infrastructure.

The addition of the B747 aircraft addedsignificantly to the Training Department’schallenge of meeting the demand forqualified flight crews.

At the same time, flight crew turnover wasincreasing as individuals found moreattractive employment elsewhere. Also, thecompany’s policy of recruiting from southernAfrica limited the pool of new potential crewmembers.

All these factors contributed to a shortage offlight crew required to meet the flying orproduction demand. This shortage of flightcrew increased the potential for increasedfatigue and stress among the personnel.

Rest, duty and flight time

Although the Operations Manual stated thatflights would not be planned beyond 24hours, the Crewing Department at MK AirlinesLimited routinely scheduled flights in excessof that limit.

There was no effective programme in place tomonitor how frequently these planningexceedences occurred, nor was there aprogram to detect and monitor exceedencesbeyond the planned duty days.

In the absence of adequate companycorrective action regarding theseexceedences, crew developed risk mitigationstrategies that included napping in flight andwhile on the ground to accommodate thelonger scheduled duty days. This routine non-adherence to the Operations Manualcontributed to an environment where someemployees and company management feltthat it was acceptable to deviate fromcompany policy and/or procedures when itwas considered necessary to complete a flightor series of flights.

There is a reasonable limit to the time a flightcrew can remain on duty before acute fatigue

23focus winter 09

begins to induce unacceptable humanperformance deficiencies. This is regardless ofthe crew composition and the adequacy ofthe rest facilities on board the aircraft.

Examination of the occurrence crew’swork/rest/sleep and duty history indicated thatthe operating crew would have been at theirlowest levels of performance because of fatigueat, or shortly after, their arrival in Halifax. Thisstate of fatigue would have made themsusceptible to taking procedural shortcuts andreduced their situational awareness.

This period of low performance would havebeen present when the take-off performancedata were calculated, the before-flightstandard operating procedures were notfollowed, and the inadequate take-offperformance was not recognized.

The company’s flight and duty schemeallowed flights to be scheduled up to 24 hourswith only three pilots required. This meantthat there would be either only one captain orone first officer in the crew. Because mostcrew members were only qualified to occupyeither the left or right pilot seat, two of theassigned pilots would have to be present forevery take-off, departure, arrival, and landingfor the entire route.

This resulted in the lone captain or first officerbeing subjected to a disproportionate amountof flight deck duty and, therefore, morevulnerability to fatigue. For this series offlights, the first officer was the critical crewmember in this respect.

The first officer had checked out of the hotelin Luxembourg at 0925 on October 13, but itis known that he was awakened earlier than0848, perhaps as early as 0630 or 0700. It isprobable that he was not in the cockpit for afew hours on the first flight, but it is unlikelythat he would have slept or had a good restbecause of circadian rhythm effects.

As other MK Airlines Limited flight crewsindicated, it was not easy to get rest on theflight to Bradley International Airport becauseof the time of day. The flight from Bradley toHalifax took 1 hour 9 minutes, and the firstofficer would have been in the cockpit duringthis flight. Therefore, he would likely havebeen the most fatigued pilot.

The aircraft was on the ground at HalifaxInternational Airport for 1 hour 42 minutes.

Twice during this time, it was noted byground personnel that the first officer wasnot in the cockpit, and it was common forflight crew to nap or rest if the turnaroundtime was long enough.

It is likely that he took a nap between the timethe take-off performance data was calculatedand when he was required to be back in thecockpit to prepare for the departure.

If the first officer had been sleeping while theaircraft was on the ground in Halifax, he wouldhave been susceptible to sleep inertia for 10 to15 minutes after waking up. As a result, hewould have been less alert than usual when hefirst entered the cockpit, the period when theperformance data would have been set fromthe take-off data card information.

In addition, if the captain had carried out someof the first officer’s pre-flight duties to allowhim to sleep, this would have further removedthe first officer from the cockpit environmentand decreased his situational awareness.

At the time of the occurrence, MK AirlinesLimited rest, duty and flight time scheme wasone of the least restrictive among ICAOsignatory states. The company’s increase ofthe maximum flight duty time for a heavycrew from 20 to 24 hours also increased thepotential for fatigue.

Findings as to causes and contributing factors:1. The Bradley take-off weight was likely used

to generate the Halifax Take-offperformance data, which resulted inincorrect V speeds and thrust setting beingtranscribed to the take-off data card.

2. The incorrect V speeds and thrust settingswere too low to enable the aircraft to take offsafely for the actual weight of the aircraft.

3. It is likely that the flight crew member whoused the Boeing Laptop Tool (BLT) togenerate take-off performance data did notrecognize that the data were incorrect forthe planned take-off weight in Halifax. It ismost likely that the crew did not adhere tothe operator’s procedures for anindependent check of the take-off data card.

4. The pilots of MKA1602 did not carry outthe gross error check in accordance with thecompany’s SOPs, and the incorrect take-offperformance data were not detected.

5. Crew fatigue likely increased theprobability of error during calculation ofthe take-off performance data, anddegraded the flight crew’s ability to detectthis error.

6. Crew fatigue, combined with the dark take-off environment, likely contributed to a lossof situation awareness during the take-offroll. Consequently, the crew did notrecognize the inadequate take-offperformance until the aircraft was beyondthe point where the take-off could be safelyconducted or safely abandoned.

7. The aircraft’s lower aft fuselage struck aberm supporting a localized antenna,resulting in the tail separating from theaircraft, rending the aircraft uncontrollable.

8. The company did not have a formaltraining and testing program on the BLT,and it is likely that the user of the BLT inthis occurrence was not fully conversantwith the software.

Reprinted with kind permission of Transport

Canada.

24 focus winter 09

Members List

FULL MEMBERS

ChairmanMonarch AirlinesCapt. Tony Wride

Vice-ChairmanEurocypriaSteve Hull

TreasurerAir ContractorsCapt. Anthony Barrett-Jolley

Executive Board CAA RepCAAMark Chesney

Non Executive Board MemberCTC Aviation Services LtdRobin Berry

Aegean AirlinesCapt. Vassilis Anadiotis

Aer ArannJoe Redmond

Aer LingusCapt. Henry Donohoe

Airbus S.A.SChristopher Courtenay

AirclaimsJohn Bayley

Air ContractorsCapt. Anthony Barrett-Jolley

Air MauritiusCapt. Francois Marion

Air SeychellesBen L’Esperance

Air Tanker Services LtdLee Carslake

ALAEIan Tovey

Astraeus LtdChris Barratt

AVISAPaul Chapman

BA CityflyerChris King

BAA LtdTim Hardy

BAE SYSTEMS Reg. A/CAlistair Scott

Baines SimmonsBob Simmons

BALPACarolyn Evans

Belfast Intl. AirportAlan Whiteside

BlinkDavid Summers

bmi regionalPeter Cork

British InternationalPhil Keightley

CargoLux AirlinesMattias Pak

Cathay Pacific AirwaysRick Howell

Charles Taylor aviationDavid Harvey

Chartis Ins. UK LtdJonathan Woodrow

CHC ScotiaMark Brosnan

CityJetJohn Kirke

Cranfield Safety &Accident Investigation CentreDr. Simon Place

CTC Aviation Services LtdRobin Berry

Cyprus AirwaysAndreas Georgiou

DHL AirGavin Staines

Eastern Airways UK LtdCapt. Jacqueline Mills

easyJetCapt. Chris Brady

EurocypriaCapt. Christis Vlademirou

European Aeronautical Group UKMax Harris

European Air Transport NV/SAHans Hoogerwerf

Flight Data Services Ltd Capt. Simon Searle

flybe.Neil Woollacott

FlyglobespanCapt. Steve Rixson

Gael LtdPaul Callaghan

GAMA AviationNick Mirehouse

GAPANCapt. Alex Fisher

GATCOShaneen Benson

GE AviationMike Rimmer

Goodrich Actuation Systems LtdGary Clinton

Gulf Air CoCapt. Paulo Fitze

Independent Pilots AssociationCapt. Peter Jackson

Irish Aviation AuthorityCapt. Harry McCrink

Jet2.comDavid Thombs

Libyan AirwaysEngr. Tarek Derbassi

LoganairRobin Freeman

London City AirportGary Hodgetts

Lufthansa Consulting GmbHIngo Luschen

Malaysia AirlinesOoi Teong Siew

Manchester Airport plcSimon Butterworth

Monarch AirlinesCapt. Tony Wride

Members of The United Kingdom Flight Safety Committee

focus winter 09

Panasonic AvionicsBob Jeffrey

PrivatAirJan Peeters

Qatar AirwaysTBA

QBE AviationJerry Flaxman

RTIDr Simon Mitchell

Rolls-Royce PlcPhillip O’Dell

RyanairCapt. George Davis

SBACMel JamesVic Lockwood - FR Aviation

ScotAirwaysNigel McClure

Shell Aircraft Intl.Tony Cramp

Superstructure GroupEddie Rogan

TAG Aviation (UK) LtdMalcolm Rusby

TAM Brazilian AirlinesCapt. Geraldo Costa de Meneses

Thomas Cook AirlinesKenny Blair

Thomson AirwaysMartin Ring

Titan AirwaysPavan Johal

Virgin Atlantic AirwaysRob Holliday

VistairStuart Mckie-Smith

GROUP MEMBERS

bmiDavid Barry

bmi Eng.Willam Taylor

bmi babyNicole Stewart

Bond Offshore HelicoptersTony Duff

Bond Offshore Helicopters (Maint)John Crowther

Bristow HelicoptersCapt. Derek Whatling

Bristow Helicopters Eng.John Parker

MODDARS Col. Steven MarshallWg.Cdr. Andrew G Tait

QinetiQFlt. Lt. Dominic Godwin

QinetiQ Eng.Andy Bruce-Burgess

RAeSPeter Richards

RAeS Eng.Jim Rainbow

CO-OPTED ADVISERS

AAIBCapt. Margaret Dean

CAASarah Doherty - Grp. Safety ServicesGraham Rourke - AirworthinessSimon Williams - Flight Operations PolicyGarth Gray – Flight Operations

CHIRPPeter Tait

GASCoJohn Thorpe

Legal AdvisorEdward SpencerBarlow Lyde & Gilbert

NATSKaren Bolton

Royal Met. SocietyRob Seaman

UK Airprox BoardAir Cdre. Ian Dugmore

ADVERTISING IN THIS MAGAZINE

Focus is a Quarterly Publication which has a highly targeted readership of32,000 Aviation Safety Professionals worldwide.

If you or your company would like to advertise in Focus please contact:

Advertisement Sales Office:

UKFSC, The Graham Suite, Fairoaks Airport, Chobham, Woking, Surrey. GU24 8HX.Tel: +44 (0)1276 855193 Email: admin@ukfsc.co.uk

25

We may not know much about aviation....

For more information please contact Andrew Kirk on

01483 884884 andrew@wokingprint.com

proud printers of Focus Magazine