Air North Brasilia Crash in Darwin (Merged)

[Ex FSO GRIFFO]

I'm with you Mr Framer...

I don't fly 'every day', and I don't fly twins as regularly now.
I have been 'back to the books' for a few of these posts, to rejuvenate the grey matter a little.
A sim practice, or a good safe practice at a safe alt. I know would be a whole lot better, but.....we do what we are able to do....

Cheers

[betaman]

V speeds - Wikipedia, the free encyclopedia

Betaman and the like, I hope you learnt something tonight.

For the record, V2 is a TOSS, nothing more. It's not a magic speed that will fix all ones worries in an asymmetric situation. It's just the minimum speed you can fly to achieve certified climb gradients.

Which is why it's called the take of safety speed

Ok GG, LS, et al,

Looks like I owe you an apology!

I have always used the 20.7.1b definition

"V2 means the initial climb out speed which is not less than the take-off safety speed;"

Which seems to imply that they are two different speeds.

But having blown the dust off the AFM of the last two turbo prop types I flew, my copy of Handling the big Jets & some old ATPL notes which quite plainly states that V2 is the TOSS.

Wish I could say that it was due to too many reds or stubbies but I was sober as judge, so yes I did learn or re learn something.

Incidently both AFM's don't mention anything about climbing at a speed higher than V2 but the current aircraft I fly does, so go figure.

Cheers

[compressor stall]

For those interested, there are a couple of iphone/ipad apps showing the FARs in a nice easy to navigate document that is also appropriately hyperlinked making regs that cross reference others simple. I know the stuff is free on the net, but the formatting does make such a difference to getting your head around some of the bits.

[The Green Goblin]

Glad we could help

Also good to see is someone who's mind is not a closed shop (like many in the industry).

Now it will become a frustrating exercise reducing your climb out speed to V2 to appease your checkie when you know it's a rediculous manuovre that could cause a loss of control for no reason other than ignorance.

Handling the big jets is pure gold.

[framer]

Also good to see is someone who's mind is not a closed shop (like many in the industry).

Spot on.It's a mark of maturity and character to wite a post like that. It used tbe the norm a decade or three ago, now....not so much.

[john_tullamarine]

Incidently both AFM's don't mention anything about climbing at a speed higher than V2 but the current aircraft I fly does, so go figure.

Probably a simple answer upon reflection -

(a) there is no imperative for the OEM to provide overspeed schedule data

(b) your turboprops, I guess, are short range machines (Dash, 340, and similar ?) so the presumption is that a lot of departures are from shorter runways where min speed schedules rule.. ie not much demand for overspeed schedules.

(c) jet ops typically involve longer runways at major aerodromes where the opportunity to squeeze a bit more payload out of a runway is significantly greater ergo more reason for overspeed scheduling


One very important point with increased V2 considerations which I don't recall seeing noted is - OEI, one should never attempt to increase V2 from a lower speed to get a better climb due to the distance penalty in the air - while the final climb will be better, the intermediate problem is that the aircraft will probably run under the planned NFP with obvious terrain clearance considerations.

The whole idea of the higher airspeed covers two situations -

(a) if the takeoff is planned with overspeed scheduled speeds, the higher V2 is targetted as part of the OEI takeoff

(b) if the initial takeoff didn't involve a failure, then the airspeed, in all likelihood, will be somewhat above V2 which puts you in a better part of the climb characteristic (gradient curve).

So, in the event of a post-V1 failure, hang on to whatever speed you end up with (after the failure is controlled) within a modest increment band (typically quoted as V2 to V2+20 or so) rather than pitching up to slow down to V2.

If, on the other hand, you are at V2+40, say, then, in the absence of general knowledge regarding your position on the climb characteristic, it is a better practice to pitch up slowly to come back within the AFM-cited overspeed range.

[Sarcs]

Without this diverging into a mixture vs closed throttle vs Flight idle vs zero thrust debate has anyone noticed the gradual decay of the quality of investigation reports coming out of the ATSB?

Besides the Hempel Inquest, where the ATSB appears to have abrogated all responsibility to investigate at all, the ATSB report into the Brasilia accident in Darwin is nothing short of totally spare in its conclusions!

There also appears to be no 'Safety Recommendations' generated from a training accident that I think we could have all learnt a lot more from.

Take a look at a couple of extracts from the report:

Terminology used in training and checking
The operator’s documentation did not contain any specific terminology for discontinuing a manoeuvre, but did provide clear instruction as to how control of an aircraft was to be changed between crew members.
To take over control from the pilot flying, or for the pilot flying to relinquish control to the other pilot in a multi-crew aircraft, very specific terminology was used. To avoid any confusion as to which pilot was manipulating the controls, the operator’s General Policy and Procedures Manual, section 4.7.2.2 Crew Communication - Handing Over and Taking Over stated:
The process of handing over control of the aircraft shall always be conducted in a positive manner. To minimise confusion or operational risk, the following terminology shall be used.
To assume control, the pilot monitoring shall call "taking over". To relinquish control, the pilot flying shall call "handing over".
Control of the aircraft cannot be handed over until the pilot monitoring has called "taking over"...
The term ‘disengage’ that was used by the PIC during this simulated engine failure was not standard phraseology. Other EMB-120 pilots reported that they had never heard the term ‘disengage’ used for any action other than deselecting the autopilot/yaw damper and had never heard it used to discontinue a manoeuvre.
They also reported that if a training or check pilot decided to discontinue a simulated engine failure procedure, they would expect that check pilot to restore power to the ‘failed’ engine. Alternately, if the training or check pilot wanted to assume control of the aircraft, they would expect to hear the term ‘taking over’.

Which is pretty basic stuff in a multi-crew aircraft....and then in regards to the Yaw Damper....

The operator’s flight operations manual for the EMB-120 stated that the yaw damper was not to be used for takeoff or landing, and that the minimum speed for its use during one engine inoperative (OEI) flight was 120 kts indicated airspeed (KIAS).

.....and then more in relation to the apparent Yaw Damper activation...

The use of the yaw damper during asymmetric flight was introduced to the simulator testing following consideration of the cockpit voice recording references to the PIC’s command ‘disengage’ and the pilot under check’s response, ‘yeah, disengaging’. It was assumed that the reference was to the yaw damper and not the autopilot because the chime that sounds when the autopilot was disengaged was not heard on the CVR recording. Additionally, the simulator instructor reported having previously observed pilots engage the yaw damper during simulated engine failures in the EMB-120 in response to pilots ‘overcontrolling’ rudder and aileron following a simulated engine failure.

All of the above is all good factual investigative methodology a lot of which points to a number of operational issues (company SOPs etc) and regulator oversight issues....right?? Wrong, take a look at the first paragraph of the ATSB conclusion.

No organisational or systemic issues that might adversely affect the future safety of aviation operations were identified as a result of this investigation.

Maybe this conclusion is a result of the regulator putting in place the Mandatory Simulator program and subsequent NPRM leading to the current NFRM, but does it excuse glossing over what was a particularly significant training accident event that, although tragic, we could all have learnt from!

ps ....and what gives with the no 'Safety Recommendations' issued!

http://www.atsb.gov.au/media/3546615/ao-2010-019.pdf

[Kharon]

Follow the steps

Look at the stellar job BASI did on the Seaview fatal.

Then look at the half hearted, puling effort ATSB made of Lockhart River.

Then read the pile of bollocks produced for Whyalla.

Compare the J* soft porn against the harsh fantasy of the Tigers tale.

Then read this provided by HMHB for the Hempel inquest.

Found the reference in the annual report
Civil Aviation Safety Authority - Stakeholder liaison
Complementing activities under the MoU, in 2008–09 CASA established an Accident Liaison and Investigation Unit to manage the agencies’ day-to-day interaction. The unit provides a contact point for the ATSB, reviews all ATSB occurrence reports and prepares responses to ATSB recommendations, and identifies opportunities for aviation safety improvement.

CASA has also established an Accident Investigation Report Review Committee, chaired by the Deputy Director of Aviation Safety, to review and agree on the method of implementation of any formal recommendations from the ATSB.

Then, when you join up all the dots; call a taxi for the Minister - they are his watch dogs after all.

Abogate - ab·ro·gate tr.v. ab·ro·gat·ed, ab·ro·gat·ing, ab·ro·gates
To abolish, do away with, or annul, especially by authority.

[Centaurus]

has anyone noticed the gradual decay of the quality of investigation reports coming out of the ATSB?

I agree. Of course we are not privy to the original draft by the investigating team which when all put together without the benefit (?) of lawyer input is probably a lot clearer than after it has been sanitised.
With all sorts of litigation flashing around as a reult of a serious accident, is it any wonder that ATSB are very careful in how they word the final document.

[Slippery_Pete]

standards limit the bank angle to "no greater than 5 degrees", the optimum is around 3 degrees

4Dogs, JTs comments are correct on this.

You are infact talking about two completely different things.

One is an attempt to minimise Vmca.
The other is where the aircraft performance (ie rate of climb) is best.

They are two separate objectives. Vmca will, in nearly all aircraft, continue to DECREASE a little more as bank increases above 5 degrees towards 8-10 degrees. As JT said, this is not practical, so certification places a restriction on the manufacturers that Vmca can not be at more than 5 degrees angle of bank.

It is also important as people have said, that during certification, speed is slowly and in a very controlled manner reduced back onto Vmca. The current type I fly at a minimum speed takeoff, requires aggressive and sudden application of full rudder and 60-70% aileron during rotation. Once at the required pitch attitude, these can be reduced very slightly. This is in complete contrast to a nice, slow and controlled walk back into Vmca.

Vmca could be considered somewhat of an irrelevant speed in flight operations (important conceptually, of course). When was the last time you had a V1 cut, and rotated the aircraft into the air with 5 degrees angle of bank?

All this discussion about min speeds is all well and good, but not particularly relevant. The main causes IMHO were:
1. Should have been in a sim
2. Autofeather failure with a V1 cut is a ridiculous scenario to train for. How many times has this actually happened vs. how many people have been killed training for it? ABSOLUTE waste of time.
3. The checkie made some fundamental errors when it turned to .
a) He should have been trained by CASA to have a strict set of criteria (a maximum heading change, speed change, controllability change) upon which he terminated the exercise immediately by no method other than advancing the power lever of the failed engine.
b) He should have been trainined by CASA that in the event of pushing the "failed" power lever forward, if things were still not improving, that the live engine should have been pulled back to ensure directional control, even if this required a temporary loss of altitude.

Both crew most probably could have given quite a good rundown in the briefing room on what Vmca is, under what conditions it is certified, and what that speed was for their aircraft.

But that knowledge meant nothing when the exercise was not immediately terminated when it went pear shaped, and when the power on the live engine was INCREASED, not reduced.

Anyone CASA approves to complete ME aircraft training needs to be extensively trained and examined on the following before being let loose:
1. Be absolutely confident in their head which engine they are going to pull and what actions will be required
2. Guard the "wrong" rudder just prior to the failure to provide protection against incorrect rudder application
3. Have a STRICT set of SMALL tolerances, beyond which they commit themsleves to IMMEDIATE termination of the exercise
4. Where exercise termination is required, do this by immediately advancing the "failed" engine
5. Where this does not resolve the issue immediately, reduce the live engine power to ensure directional control (sometimes at the expense of altitude)
6. Immediately takeover control of the aircraft.

Another two (from all reports) nice guys of aviation who have paid the ultimate price because of archaic, 1960s training ideologies & methods.

[aussie027]

Slippery Pete,
Very well said.
Your comments re strict criteria for control of and termination of the exercise are spot on. That is what is required in all training scenarios but especially something as critical assym work at the very edge of the envelope and close to the ground.
Again doing so in the sim is best and safest in these TKOF scenarios.
As you say it is never worth dying for to practice some extremely unlikely scenario that out of hundreds of millions of TKOFs worldwide over decades an EXTREMELY small number of flights have ever experienced.

Also for everyone out there who has seen the variety of FAA definitions of V1 that have been published over the last 30yrs here is the best one I have seen that is most accurate and easiest to remember--

V1

"The speed at which the takeoff must be continued, if the abort has not already been commenced."

V1 is NOT the latest speed for a stop decision, V1 is a GO speed, the difference is significant and critical. The failure must be recognized and the decision made before V1.

I have the article explaining those statements in a file, if anyone wants to see it just pm me.

[Wally Mk2]

I've always found the V1 speed thingy a total croc! No one can actually read/nominate this defined exact number whilst hurtling down a rwy sheez our speedo's aren't calibrated to such an accurate figure anyway!
By the time yr eyes see the V1 speed reach the line in the sand, yr brain processes it sends the req'd message to yr voice box the recipient processes that sound via his/hers ears into usable info & then via their CPU (brain) for consideration yr almost at TOC!:-) It's an arbitrary figure.

Aussie's comments are spot on re V1's def.

For me once the speed V1 has been called it's too late to stop (theoretically anyway) you go unless it's unflyable & even then it's an unknown as to the extent of major damage.

Wmk2

[Mach E Avelli]

But better to go off the end of the runway at 20 knots than to get airborne at 100 knots+ and go splat from 200 ft.
In training, where you don't have the luxury of a simulator, schedule the MAXIMUM V1 that the runway or aircraft performance allows, regardless of the real V1 for the weight at the time. Then you have a nice fat buffer for the lower typical training weight.
Scheduling V1 min for actual training weight will put many aeroplanes too close to Vmca for comfort.
The check pilot can always simulate a heavier aircraft weight to align with the high V1 by judicious use of less than max T/O power. This provides a further safety buffer against Vmca.
In fact, scheduling V1 max is not entirely a silly philosophy wherever you have a runway that will allow it. The longer you stay glued to the ground the better, because that's the best place to have an emergency if you really must have one.
How many runways do we have that we use for training that are really so accelerate-stop limiting that we must use V1 min?

[Sarcs]

I think K has a point here, take a look at the Seaview report from about pg 44 onwards:

http://atsb.com.au/media/24362/aair199402804_001.pdf

BASI (as they were known then) kicked the regulator so many times (13 directly) that their noses would have left a trail of blood from the accident site to Lord Howe and back to Canberra....all without stepping outside the boundaries of their remit!

No wonder the metamorphosed CASA promptly circled the wagons, battened down the hatches and have been in seige mode ever since!

[By George]

Interesting to see that some Airlines no longer call out "V1" anymore they call "Go". I'm sure Lufthansa is one and I think it's a good idea. V1 is called a decision speed by some which is not entirely correct.

[Kharon]

Now I'm curious

I find I am once again forced to read between the lines of an ATSB report. Technically it's spot on; for example, reading the time line analysis, there is a temptation to question the rudder v aileron input, however this is clearly resolved in the computer simulation graphics. Not having operated a Braz – there are a couple of points of interest which perhaps can be explained by someone who has.
Disengage ?? –could this refer to the Flight Director or is it the Yaw Damper ?. I note the AFM mentions –(paraphrased) FD Before take off (SET), expanded to PF select GA and check 7° pitch up; and, that the AP or YD may not be engaged during TOFF and LAND manoeuvres.

Has it been SOP for the PF to engage the YD as part of an EFATO or was this a recent innovation ?.

The V1, Vr and V2, V2 +10, etc. schedule. The AFM seems to be clear about the speed schedule and the acceleration to V2 +, then flaps up then Vfs (paraphrased). There appears to be a deliberate reduction from V2 + 4 (at – 23 seconds) to V2 (at -20.7 seconds). Is this a norm for the type or a new innovation ?.

Is the un monitored management of the OE, the over torque (124%) and the corresponding rudder/aileron excursions normal for the airborne exercise being conducted ?.

It is suggested by the ATSB final analysis that the BASI recommendations made after an investigation into the Flight Idle v Zero thrust (auto feather) scenario have been ignored. There is much documented proof that CASA have been enforcing 'black letter' CAO 40.1 requirements which conflict with both the AFM (see CAR 138) and a common sense approach to airborne EFATO exercises.

It is noteworthy that simulator based training had been recently introduced; and, conversely that Air North have safely, successfully conducted many 'in flight' simulated failures prior to the introduction of 'simulator' techniques. It is of concern that several things occurred which should give a check pilot the screaming heebie jeebies, airborne.

Perhaps we could ask of the ATSB to investigate 'in depth' the contributing factors to this situation occurring. We have the almost self evident facts of the accident, we have the why, but maybe it would be nice to know the wherefores.

P.S. Categorically not having a pot shot at the crew, the company or the simulator. Just seeking a satisfactory explanation of why and how this 'abnormal' chain of events occurred. If this was a new ME instructor and an initial twin conversion, perhaps this event may have occurred, but it wasn't – was it.

[Sarcs]

It is noteworthy that simulator based training had been recently introduced; and, conversely that Air North have safely, successfully conducted many 'in flight' simulated failures prior to the introduction of 'simulator' techniques. It is of concern that several things occurred which should give a check pilot the screaming heebie jeebies, airborne.

Perhaps we could ask of the ATSB to investigate 'in depth' the contributing factors to this situation occurring. We have the almost self evident facts of the accident, we have the why, but maybe it would be nice to know the wherefores.

A lot of what was covered in the ATSB report touched on the areas of operational concern and hinted to several differences in history where the Check Pilot had started to diverge from his 'norm'. This quote from page 54 of the report is significant:

Two of the pilots who were recently assessed by the PIC reported that he selected flight idle (zero torque) to simulate an engine failure after takeoff in their check flights. It was possible that the PIC had decided to deviate from the operator’s approved procedure in order to test the recognition by the candidate of the additional failure of the autofeather, before setting zero thrust.

However it isn't clear whether this 'divergence' from his 'norm' started after he had been to the simulator or before. If it was after then one may suggest that he was operating in a 'simulator induced complacency' manner i.e. it was proven that Flight idle (aircraft) or 'Autofeather Failed' engine failure (simulator) could be successfully recovered from while conducting a V1 cut.

This also appears to have been an area of concern for the regulator, as they used this accident as an example in Annex A of the NFRM for Mandatory Simulator, see here:

From CASA NFRM Mandatory Simulator training Annex A:

COMMENT 1.2
Some respondents proposed adding wording to allow training and checking to occur in the aircraft provided the exercise had been conducted by all pilots in a simulator in the preceding 12 months.
CASA Response
CASA is firmly of the view that where a qualified STD is available for aircraft of this size, this should be used for all training and checking activities. The ATSB has reported that the training captain of the aircraft involved in an accident in Darwin in March 2010 had undergone training and checking in a flight simulator, however the actions by the training captain in simulating an engine failure in the actual aircraft during the accident flight was not consistent with the training received during the simulator course. This suggests that doing one session of training and one check per year in an STD (with the subsequent session/check in an actual aircraft) does not satisfactorily address the risk of conducting non-normal exercises in an actual aircraft.

If the Check Captain was inducing this scenario (FI V1 cuts) prior to having gone to the simulator, whereas previously he always only induced a Zero Thrust EFATO scenario, then there must have been input from somewhere/someone to change him to suddenly start breaching the company T&C SOPs??

As 'K' suggests there has been many takes, ambiguity and debate..etc..etc..on the regulatory requirements of CAO 40.1.0:

There is much documented proof that CASA have been enforcing 'black letter' CAO 40.1 requirements which conflict with both the AFM (see CAR 138) and a common sense approach to airborne EFATO exercises.

Maybe there is an element of rogue FOI's, that lack the necessary industry experience, that insist on adhering to the letter of the law in CAO 40.1.0. Instead of applying practical safeguards and risk management to high risk training and checking scenarios!!

[john_tullamarine]

Vmca will, in nearly all aircraft, continue to DECREASE a little more as bank increases above 5 degrees towards 8-10 degrees

A LOT more. The nature of the beast is that

(a) for near wings level, Vmc varies sedately with a degree or two of bank

(b) the further we get away from wings level, the HIGHER the RATE at which Vmc varies with bank. Nothing sedate about it as the bank increases beyond 5 degrees. More importantly, if you are back near the real Vmc for the day (ie for 5 degrees) and you let the bank reduce or, worse, go the wrong way, the Vmc can ramp up VERY quickly and you might find yourself emulating the video linked a few posts ago.

When was the last time you had a V1 cut, and rotated the aircraft into the air with 5 degrees angle of bank?

However, if you are using a min speed schedule, especially if Vmca limited, and the configuration/conditions put you near to the real Vmc, then you had best make sure that you roll in the 5 degrees during the rotation flare ...


Mach E Avelli, who has been around for a LONG time and has a sound idea of what is what, provides good guidance in recommending the use of higher speed schedules within the RTOW ambit when departing at very light weights .. and not just for training sorties .. works the same way for line operations.

I would add that this is even more relevant in the case of a stiff crosswind takeoff .. why expose yourself to a potentially significant Vmcg consequence when, with a bit of lateral thinking, you can avoid it altogether ...

[sheppey]

The longer you stay glued to the ground the better, because that's the best place to have an emergency if you really must have one.

Interesting point of view which many will agree with. I guess we all have our personal opinions based often on personal experience.
I have always favoured a low V1 based on research papers that state most runway accidents are caused by a late abort when it was never necessary in the first place. A burst tyre is more likely at high speed than low speed and the general rule espoused by McDonnel Douglas aboout 30 years ago, was that it is safer to continue the take off from 15 knots below V1 rather than risk extended abort distance with a buggered tyre and associated brake loss that wheel. The statistical chances of hitting an obstacle after take off following engine failure after V1 are far less than the probability of a cocked up abort at high speed which is how some aircraft have crashed because of an over-run.

The higher the V1 the more the energy needed to stop with resultant hot brakes. Historically, overruns caused by an abort beyond V1 have occured because of indecision/uncertainty or even incompetency by the crew. If a late abort on a heavy weight take off occurs for whatever reason, there is the possibility of the brakes being beyond their energy limits and thus brake failure. For all these reasons this writer takes the view that a low VI if available, is a safer proposition in terms of the go decision rather than sweating out an unecessarily high V1 and the demonstrated dangers of a high speed abort. I would be much happier with a 20 knot spread between V1 and VR than having V1 =VR or maybe a couple of knots below.

[BPA]

Tend to agree that the standard of reports coming out of the ASTB is getting worse. Most reports these days a full of quotes etc about human factors from other sources and no real meat anymore. A lot of the reports they produce about incidents involving QF, Jetstar and Virgin are copies of the reports the safety departments from those airlines have produced ie they are letting the airlines do all the work and then take credit for it.

Have a read of some of reports the NSTB or the AAIB have produced such as the Q400 crash and they contain far more than just quotes on human factors and the like.

Perhaps it's time to start a new thread about the standard of the ATSB reports coming out.