This diagram illustrates profile drag with a green
line, lift induced drag with a blue line and total drag with a red line.
Profile drag (green line) increases by the square of the airspeed increase
i.e. doubling the airspeed gives four times the profile drag. The lift
induced drag (blue line) is not quite so simple to explain, but for the
purposes of this discussion all that is necessary to understand is that lift
induced drag is at a maximum at minimum airspeed and decreases with
increasing airspeed. That is a simplification, but serves for this
discussion. If we then add the values of the blue line to the values of the
green line we get the total drag (red line).
What use is all this stuff?
Let’s look at what we can learn from this simple graph. The airspeed at
which the blue and green lines cross gives the speed for minimum total drag
– which is the speed at which we will require minimum power in level
flight to maintain level flight. That, by definition, is our endurance speed
that will produce minimum fuel consumption per hour. This is not the same as
“range speed” which is the speed at which we consume minimum fuel
per mile. Stick with it, we are coming to the interesting bits.
If we are flying along at a speed greater than our endurance speed (which
is normal for cruising around) then we will be operating somewhere on the
total drag curve (red line) where the orange arrow points to the total drag
curve. If the aircraft now hits a little turbulence the airspeed will
decrease slightly and our operating point moves slightly to the left. Notice
that, with this decrease in airspeed, the total drag DECREASES. Thus we will
already have excess power applied and the airspeed will automatically
increase to return to the previously set speed. This is called “speed
stability.” Next let us have a look at what happens if we are flying
along at an airspeed LOWER than our endurance/min drag speed, as indicated by
the black arrow pointing to the total drag curve. If the aircraft now hits a
little disturbance and the airspeed decreases then the total drag will
INCREASE. This will cause the airspeed to decay further and it will continue
to decay right down to the stall until you either lower the nose or add more
power or do a bit of both. This is called “speed instability” and
it occurs whenever you are flying “on the wrong side of the drag
curve.”
The Drag Curve — Takeoffs &
Landings
When are you going to be flying on the wrong side of the drag curve? On each and
every take-off and approach to land. It is a basic rule of thumb that the approach to
land is flown at 1.2Vs. That is 1.2 times you basic stalling speed. If your
aircraft has a basic stalling speed of 50kts then you should fly the approach
at 60kts. If your Vs is 60kts then approach at 72kts – fairly easy to
remember! When did you last check the
stalling speed of your aircraft? Not what it says in the Manual but the
actual speed at the actual weight of your actual aircraft.
The important lesson to
remember from my little drag v. airspeed graph is that you will always be
“on the wrong side of the drag curve” and suffering from
“speed instability” when taking off and when approaching to land.
Remember your basic training that POWER controls your rate of descent on the approach
and ELEVATOR controls your airspeed. The trap for inexperienced pilots is to
get low and slow on the approach and instinctively raise the nose by applying
up elevator. This will only reduce your airspeed even further and precipitate
a sudden arrival in the undershoot or a ton of bricks thump onto the runway,
with yet another bent nose-leg and mutterings of “It must have been
windshear.”
Too much speed on the
approach is just as useless as too little speed. If you come racing down the
approach at 1.2Vs plus ten knots for the crosswind plus ten knots for the
wife and kids don’t be surprised when the sorely abused aircraft
strikes the runway a glancing blow with the nose wheel, folds the nose-leg
backwards and darts off into the far undergrowth. If it is extremely
turbulent on the approach it may be necessary to increase your airspeed by a
few knots and to select your touchdown point a little further into the runway
– or it may be wiser to go away and come back another day if you think
the conditions are beyond your abilities.
Finally let’s spend a
few moments thinking about the actual touchdown on the runway. It’s not
good practice to arrive a few feet above the ground, close the throttle and
switch your brain off. When I was doing my basic flying training on Harvards
my instructors drummed into my head “Remember, the landing is not
complete until the wheels have stopped turning.” That was vital on
Harvards as they could bite your bum at very slow speeds on the runway. It is
a very useful thought for any pilot in any aircraft.
Arrive at the selected
touchdown point at the correct airspeed in the correct attitude after a
stable approach. Continue to rotate the aircraft gently so that the main
wheels contact the ground with a gentle kiss. Continue to fly the aircraft
after touchdown. Lower the nose-wheel gently onto the ground and continue to
control the aircraft as you apply gentle braking. Leave your after landing
checks until AFTER the landing is complete. All the above may be applied to
any aircraft type.
And for goodness’ sake STOP CRASHING THE CESSNAS!
Basic
Mistake – Aquila AT01
The pilot planned to fly her Aquila AT01 from Heacorn to Lasham with a
passenger on board. She got a weather update for Lasham from the resident tug
pilot there. The surface wind was forecast to be 190/10kt with broken cloud
at 2,500ft and 20km visibility.
Lasham is operated by the Lasham Gliding Society and their Airfield Manual
has a section on VISITING LASHAM BY AIR that states:
”Runway
05/23 is the Medium Runway and runs north-east/south-west. It stands out well
from the air but the surface is rough and is not used for take-off or
landing. Visiting light aircraft will land on the grass centre triangle
formed by the crossing of the hard runways … If landing on the
south-westerly run or taking off to the north-east, turbulence van be
expected due to the line of trees that you cross on landing/take-off.
Visiting pilots should note that there is a great risk from
undershooting in both directions and should therefore aim to land well up the
airfield in this wind direction. If this runway is in use, the wind is likely
to be strong and so there will be a wind gradient and turbulence.”
The
visiting pilot commenced her final approach to Lasham at 500 ft and noted the
weather conditions were as forecast although the windsock indicated little
wind at the surface. During the latter part of the approach the pilot
assessed that the aircraft was going to land short of the grass landing area
and, with full flap selected, she raised the nose to extend the approach. The
aircraft stalled at about 15 ft above the ground and landed nosewheel first.
The nose gear collapsed and the propeller struck the ground. Neither occupant
was injured. She concluded that the accident was due to insufficient speed on
the final approach.
All the above information is taken from AAIB Report EW/G2007/09.22 which
source is gratefully acknowledged.
The following comments are not meant to represent the view of AAIB in
any way and are those of Gremline alone. They are added with the aim of
contributing to Flight Safety by encouraging other pilots to learn from
mistakes already made instead of making mistakes themselves.
The pilot’s candid assessment that the accident happened
because she had insufficient speed on final approach is almost correct, but
WHY did she have insufficient speed? Her own report gives the simple answer.
She found herself going low on the approach and raised the nose to
extend the approach. She had forgotten her basic training. You control your
approach slope with POWER, and you control you airspeed with ELEVATOR.
If you raise
the nose by applying up elevator without applying extra power the immediate
effect is to reduce the airspeed – which is exactly what happened in
this accident. During my basic training it was repeated over and over again
“NEVER STRETCH A GLIDE.” This was in reference to a glide
approach after an engine failure. Equally you cannot stretch your approach
path by raising the nose (unless you have considerable excess speed
already).
Pilots are recommended to spend a few minutes reading and
understanding the article
Slow Flying published in an earlier
issue of Gremline. It explains in simple terms how to avoid the trap that
caught the pilot involved in this accident.
Rushed
Approach to Kidlington? Socata TBM 700B
A Socata TBM 700B on short final
to land on Runway 01 at Oxford (Kidlington) Airport apparently rolled left
through some 360 degrees from low level and crashed about 100 metres
displaced from the runway threshold. The pilot and two passengers died on impact.
The Air
Accidents Investigation Branch Bulletin 5/2005 covers the extremely detailed
investigation into this accident under Reference EW/C2003/12/03. This report
may be viewed in total at
www.aaib.gov. The exhaustive investigation uncovered many
relevant facts but the inspectors decided no definite conclusion could be
reached as to why this crash happened. No technical evidence was found to
explain the uncontrolled roll but there were certain operational
possibilities. None could be fully supported without hard evidence, but loss
of control resulting from an unknown distraction, or during the application
of power for flight path adjustment or an attempted late go-around, were
considered as possibilities.
The facts in this report are based on the AAIB Report and that source
is gratefully acknowledged. Any further conclusions or comments are those of
the Technical Editor of Gremline and are not intended to reflect those of
AAIB.
The Socata TBM 700B
The TBM 700B is a single engine aircraft with six or seven seats in a
pressurised cabin. It is powered by one PT6A-64 free turbine engine producing
700 SHP and driving a four-bladed Hartzell constant speed propeller. The
aircraft is certified for single pilot operation. It has a MTWA of 2984 kg
and a maximum cruising speed of 300 KTAS at FL260. Large span coupled flaps
reduce the stalling speed to 61 KCAS and have three positions; up, takeoff
(10°) and land (34°). Roll control is by a combination of interconnected
ailerons and spoilers operating through a cable and pulley system. The Socata
TBM 700B also has a mechanical interconnection system that applies rudder
when roll control is applied and also applies roll control (aileron and
spoiler) when rudder is applied. This system allows coordinated turns to be
commanded by the control wheel without the pilot needing to apply rudder. The
engine air intake has an electrically operated ‘inertial
separator’ to protect the engine from ice and debris ingestion. This
‘inertial separator’ has two movable vanes that, when switched
on, rotate to cause the intake air to execute a sharp turn causing
centrifugal force to discharge any heavier particles overboard. The inertial
separator is normally activated as part of the ‘before landing’
checklist.
The Flight
The aim of the accident flight was to fly two passengers from Brussels
International Airport to Oxford Airport and to return them to Brussels on the
following day. The aircraft was registered in the USA and was not certified
by the FAA for ‘Commercial-on-demand’ operations as this flight
was subsequently categorised, although the commander recorded it as a
‘Private’ flight. The commander held a FAA Commercial
Pilot’s Licence with a total experience of 1,573 hours of which about
500 hours were on type. He had not flown the TBM 700 for three months prior
to the accident flight, although he had flown a light piston engine aircraft
in Florida in the previous month.
On the morning of the accident the pilot flew the TBM 700 from Liege
Airport to Brussels with another pilot on board. The accident pilot operated
the aircraft throughout this flight and landed at Brussels at 0840 hrs after
an autopilot ILS approach to 250 feet agl in a 300-foot cloud base. The
autopilot was disconnected at 250 feet agl and a manual landing was made at
about 85 kt using full flap. The aircraft was then refuelled to full and the
second pilot left before the two passengers arrived.
The aircraft left Brussels on an instrument departure at 1017 hrs and
climbed to FL240 without any unusual event being noted by ATC. A descent to
FL120 was begun at 1052 hrs and this level was maintained until a further
descent was begun at 1110 hrs. The aircraft had reached about 2000 feet amsl
by 1120 hrs and was requesting radar vectors towards Oxford from Brize Norton
Radar. The Brize controller turned the TBM 700 from its westerly heading
through about 270° left onto a northerly heading and cleared the aircraft to
2000 feet on 1029 mb. The turn was for separation reasons and left the TBM
700 heading directly towards the active Runway 01 at Oxford at 4 miles range.
The pilot reported visual contact and was transferred to Oxford Tower. The
pilot reported three miles visual and the controller cleared the aircraft to
land with the surface wind of 030/15 kt. Nothing further was heard from the
accident aircraft.
The Accident
Several experienced pilots were among the witnesses who watched the
aircraft during its approach. One witness said that the aircraft had
descended to about 50 feet agl when it began to roll to the left, with the
nose rising as the bank angle reached 60°. The roll continued and the nose
dropped with the roll continuing through almost 360° before he lost sight of
the aircraft as it struck the ground. Another witness had his attention drawn
to the approaching aircraft by a considerable increase in engine noise at it
passed directly overhead. It was rolling to the left with about 40° when he
saw it. The roll continued quickly to beyond 90° left before the roll
reversed to about 60° left. The aircraft turned to the left before losing
height and striking the ground in a nose low, left bank attitude. Most of the
witnesses thought the aircraft was beginning a go-around just as, or just
after, the aircraft started to roll to the left.
The sequence of events during the approach is reconstructed from the
evidence of a selection of witnesses, on the ground and in the air. Runway 01
at Oxford is 270 feet amsl with a Landing Distance Available of 1200 metres
and is 23 metres wide with an asphalt surface. The PAPIs are set to 3.5° and
are located left of the runway 128 metres from the threshold. Heathrow Radar
provided a final contact at 1122:29 hrs at a height between 1420 and 1520
feet agl and 2.6 nm from the threshold. The aircraft struck the ground
adjacent to the threshold at 1124:40 hrs giving an average groundspeed of 93
kt and an average rate of descent of 870 feet per minute throughout the
approach.
The aircraft was about 131 kg OVER the Maximum Takeoff Weight of 2984
kg with the CG at 32.2% when it took off from Brussels. At the time of the
accident the estimated weight of this TBM 700 was 2942 kg (just 42 kg below
MTOW) and the CG was at 32.5%, within the Flight Manual limits of 19.5% to
36%.
The Pilot’s Operating Handbook recommends that a minimum of 10%
torque be maintained until the landing is assured. This is to ensure positive
and rapid engine response to throttle movement. Normal approach is at full
flap and 80 KIAS. The stall speed at the accident weight and configuration at
idle power would be at 61 KIAS.
A Tentative Analysis
It is possible to reconstruct the approach path and profile of the accident
aircraft by reference to the combined evidence of the pilot’s last
radio transmission, radar recordings, of air traffic controllers and of
numerous witnesses on the ground close to the accident position. The aircraft
should have been flown down a 3.5° glideslope at 80 KIAS with power set not
below 10% torque. The surface wind during the approach was 030/15-25 kt and
the average recorded groundspeed of the TBM 700 for the whole approach was 93
kt, giving an average indicated airspeed of about 108 kt, twenty-eight knots
above the recommended approach speed. The aircraft was always some distance
above the ideal approach slope in that it was at 1730 feet above airfield
level at four miles range (about 5.5° approach slope), at 1470 feet QFE at
2.5 miles (about 6.4° approach slope) and at 540 feet QFE at 1mile range
(about 5.8° approach slope).
The significance of the combination of excessive airspeed and the
steep approach slope added to the heavy weight and aft CG of the aircraft
becomes clearer when we try to visualise the situation the pilot found
himself faced with in the late stages of the approach. The aircraft was fast,
high and heavy. It is likely that the pilot had the power lever fully
retarded. Oxford Airport had a NDB and a DME positioned close to the
threshold of Runway 01 but did not have an ILS for the pilot of the TBM 700
to use for an autopilot approach. This was apparently his first ‘manual’
approach in the TBM 700 for at least three months, having used the autopilot
to fly the earlier approach into Brussels International Airport. It appears
that the whole landing procedure may have been rushed and the aircraft never
achieved a ‘stabilised approach’ condition. The pilot may have
made a late decision to go around by beginning to apply up elevator at the
same time as he advanced the throttle. It is possible that the engine
response was slow because the inertial separator was operating (as is normal
during the approach) and because the pilot had selected less than the
recommended minimum of 10% torque in an attempt to reduce IAS and to lower
the glideslope angle. Post crash examination found the pitch trim consistent
with an airspeed between 111 kt and 114 kt. The aircraft would have been out
of trim at the normal approach speed of 80 KIAS and the control wheel would
have required a pull force of between 11 and 13 pounds to counter the
nose-down trim. Any abnormal flight condition could have been exacerbated by
this out-of-trim condition.
There was no obvious technical reason for a left roll at the final
stage of the approach so the AAIB investigation reviewed other possible
reasons for the loss of control and looked at evidence for and against these
possibilities. These included some sort of pilot incapacitation, a
distraction, fuel imbalance, icing, wing stall and loss of control during a
go-around. Despite an extensive investigation, no definite conclusion was
reached by AAIB as to why this aircraft crashed during a visual approach to
Oxford (Kidlington) Airport. There were certain operational possibilities but
none could be fully supported.
The aircraft would have stalled in normal landing configuration, with
idle power and wings level, at about 61KIAS. This would have resulted in a
left roll/wing drop as observed, but would require the airspeed to be some 19
kt below the normal approach speed. The recorded radar information indicates
an average airspeed of about 107 kt during the approach, some 27KIAS ABOVE
normal, but these radar recordings did not extend to the latter part of the
approach, so it is likely that the pilot found it necessary to make a
considerable reduction in airspeed while still attempting to reduce the angle
of glideslope towards the end of the approach. These two requirements are in
opposition, thus leaving the pilot with a very difficult control problem.
There is also a possibility that the pilot suffered some distraction at a
late stage on the approach.
It seems possible that the pilot failed to control the aircraft
because of the unstable nature of the approach, resulting in a departure at
an altitude so low that recovery was impossible. This conclusion is that of
the Gremline editor and not that of AAIB. It is the
opinion of the Gremline editor that specific type training is required for
the safe operation of such high performance aircraft.
Text and Photographs © 2008 Gremline & Hill House
Publications, unless otherwise stated.
landing page
about gremline
copyright/conditions/contact
information exchange
glossary
uk emergency diversions
uk links, chirp & gasco
global & misc links
forum
the gremline cockpit — index of
articles
the gremline bookshop
top of page
