CFD ASSESSMENT OF AERODYNAMIC

DEGRADATION OF THE TU-154M PLANE

Glenn Arthur Jørgensen

 wego rozdzielonych dystansem 47m lotu samolotu, jakie

Abstract
The purpose of this work is to determine the change in zarejestrowały rejestratory samolotu.
the aero dynamical forces and moments of a TU-154M Słowa kluczowe – CFD, zniszczenie skrzydła, beczka
plane exposed to a left wing loss of 5.5m and 10m samolotu, Katastrofa Smoleńska, Tu-154M.
respectively with and without interaction of the right aileron
and right outer interceptor surface. This is done through the
computational fluid dynamics technique (CFD) using 1. INTRODUCTION
validated CFD++ software by Metacomp Inc. and by use of
a detailed model of the TU-154M aircraft. Results presented On the 10th of April 2010 a Tu-154M plane crashed into
here correlate well with the manufacturers data for the the ground close to the Smolensk Airfield killing all 96
plane in cruise and landing modes, thereby confirming the people on board including the polish president, the top
models and method and bringing a level of assurance that
the CFD is being solved consistently. The lost lifting force NATO generals of the Polish Air Force, navy and army. In
associated with a wing loss of 5.5m is found to be less than the final report by the Russian Interstate Aviation
8.6% of the original lifting force. Including the effect of the Committee (MAK Report, [1]), it is claimed, that the crash
increased airflow at the inner portion of the wings caused
by the engine flow will further reduce the loss associated was a result of the 150o left roll of the plane in low height
with the loss of the wing tip. The results of this work are caused by the loss of 5.5 m of the left wing tip after its
important in understanding the case of the crash of the collision with a 30 - 40 cm thick birch tree about 5 m above
presidential plane in Smolensk in April 2010. The results the ground and about 340 m prior to the site of crash. The
undermine the official explanation and support a hypothesis
of a significantly higher wing loss than claimed in the measured roll angle is recorded by the planes data recorders
official accident investigation report. The hypothesis is (black boxes) the first 1.6 seconds after loss of the wing area
furthermore supported by the recording of two distinct and until the system reaches some (unexplained) cut-off
large drops in the vertical acceleration signal separated limitation of about 62o. The measured speed of roll is linked
with about 120 m flying distance as logged by the planes
QAR data recorder. to the lost wing area, and from this the estimated wing loss
can be found. Loss in wing area on one side during normal
Keywords - CFD, wing damage, roll, Smolensk
Catastrophe, TU-154. flight of a symmetric plane will result in the loss in lifting
power of this side and a moment of rotation about the planes
Streszczenie
W pracy ustalono zmiany w siłach i momentach length axis driving a roll of the plane. Two models for
aerodynamicznych działających na samolot TU -154M calculating the final trajectory and roll angles for this case
wywołane utratą odcinka lewego skrzydła o długości have been published (Kowaleczko 2013 [2]),
odpowiednio 5.5 m i 10 m zarówno w przypadku zadziałania (G.A.Jørgensen 2013 [3]). These are very different in
jak i bez zadziałania prawej lotki i zewnętrznej powierzchni
prawego interceptora. Obliczenia wykonano w technice nature. Both models predict similar results of plane roll,
komputerowej mechaniki płynów (CFD). Wykorzystano when presented for the same input in form of the associated
zwalidowane oprogramowanie CFD ++ firmy Metacomp loss in lifting force or more correct the driving moment of
Inc. przy użyciu szczegółowego modelu samolotu TU -154M. roll caused by the loss of the left wing tip. In the CFD work
Uzyskane wyniki dobrze zgadzają się z danymi producenta
samolotu zarówno w konfiguracji rejsowej jak i w of Kowaleczko the associated loss in the lifting force at 8 o
konfiguracji do lądowania. Potwierdza to słuszność angle of attack was first estimated to be 14 % [4] then later
przyjętych modeli i metody obliczeń i zwiększa poziom reduced to 12.5 % of the initial lifting force. This work was
pewności, że uzyskane rozwiązanie CFD jest prawidłowe.
Stwierdzono, że utrata 5.5 m skrzydła powoduje spadek siły based on an highly inaccurate model of the TU-154M plane
nośnej mniejszy niż 8.6 % oryginalnej siły nośnej skrzydła. geometry not including the correct double slotted flap
Uwzględnienie efektu zwiększonego przepływu powietrza na configuration, wrong twist of wing etc. and showed a 44 %
części skrzydła przylegającej do kadłuba wywołanego higher change in the lifting force ratio than found here for
działaniem silnika dodatkowo zredukuje spadek siły
wywołany utratą końcówki skrzydła. Wyniki pracy są ważne the angles of attack of interest. By correcting the main
dla zrozumieniu przypadku katastrofy prezydenckiego discrepancies of the model used by Kowaleczko the results
samolotu w Smoleńsku w kwietniu 2010. Podważają one found by him match those found in this work [4].
oficjalne wyjaśnienie przyczyny katastrofy i potwierdzają
hipotezę o znacząco większej utracie skrzydła, niż In the early work of the author this loss was estimated to
stwierdzona w oficjalnym raporcie z badania wypadku. be about 8%-9% based on two different methods including a
Hipoteza ta jest ponadto poparta przez zarejestrowanie correction of the main errors in the model of the TU-154M
dwóch wyraźnych i dużych spadków przyspieszenia piono- plane used in the CFD work of Kowaleczko [4]. In the work
presented here and based on a very detailed model of both
Ms. Sc. Mech. Eng. Glenn Arthur Jørgensen (e-mail: gaj@xtern- the TU-154M and TU-154B this loss is found to be less than
udvikling.dk).

1
Glenn Arthur Jørgensen

8.6%. The work presented here clearly supports the sections, where the Tu-154B features triple-slotted flaps.
hypothesis of a significantly higher loss of total wing length One piece ailerons are located outboard of the flaps. The
than claimed in the official report. outer wing spoilers are flight spoilers and assist the ailerons
for roll control. The hinge line is located ahead of the
2. MODEL DESCRIPTION leading edge so that a gap appears when the spoilers are
2.1. Geometry deployed, hereby optimizing the airflow and preventing
pressure fluctuations. Each outer wing has two prominent
The standard notation for describing the motion of, and
the aerodynamic forces and moments acting on, a flight boundary layer fences. The fences are attached in line with
vehicle is indicated in Fig. 1. The variables x, y, and z the inboard ends of the No2 LE slat sections and the outer
represent coordinates, with origin at the center of mass of ends of the outboard flaps. The wing utilizes TsAGI high-
the vehicle. speed airfoils, P-56M2-12 at the roots, P-35M3-11 at mid-
span and P-35M3-10 at the tips. In the absence of these
airfoil coordinates and based on the overall description of
the airfoil main parameters found in T.N.Nnrym, the NACA
23012 airfoil is chosen for the root section, the Boeing
B737b-il (11%) airfoil is chosen for the mid span and the
Boeing B737b-il (10%) airfoil is chosen for the tip section.
The wings have zero sweepback inboard of the main landing
gear fairings. Sweepback at quarter-chord is 35°, the
anhedral angle 1°10', incidence at root is +3° and -1° at the
tip giving a total wing twist of 4°, aspect ratio 7.85, taper 3.5
and MAC 5.285 m. The plane has a cantilever swept T-tail
featuring symmetrical airfoils. Sweepback at quarter-chord
is 45°, aspect ratio 1.0, taper 1.88 and thickness/chord ratio
11%. The symmetrical airfoil S9032-il (11 %) is chosen for
the tail profile (both root and tip). The plane in cruise mode
is modelled without landing gear and with retracted flaps
Fig. 1. Definition of the coordinate system. The figure is and slats. See Fig. 2 and Fig. 3.
borrowed from [5].
The x-axis lies in the symmetry plane of the vehicle
(Nelson 1989, [6]) and points toward the nose of the vehicle.
The z-axis also is assumed to lie in the plane of symmetry,
perpendicular to the x-axis, and pointing approximately
down. The y-axis completes a right-handed orthogonal
system, pointing approximately out the right wing. The
variables u,v, and w represent the instantaneous components
of linear velocity in the directions of the x-axis, y-axis, and
z-axis, respectively. The variables FX, FY, and FZ represent
the components of aerodynamic force in the directions of the
x-axis, y-axis, and z-axis, respectively. The variables p, q,
and r represent the instantaneous components of rotational
velocity about the x-axis, y-axis, and z-axis, respectively.
The variables L (or CMx), M, N represent the components
of aerodynamic moments about the x, y, and z axes,
respectively. The variables Φ, Θ, ψ, represent the angular
rotations, relative to the equilibrium state, about the x, y, Fig. 2. Top view of the 3D model of the TU-154 in cruise mode.
and z axes, respectively (roll angle, pitch angle and yaw
angle). Thus
d , d , d .
 q r (1)
dt dt dt
3D models of the TU-154 plan are created in CreoTM
Release 2.0 and imported into CFD++ by Metacomp. Minor
corrections are added by Metacomp to make the model
water tight. In some instances, the geometry is slightly
altered to enable quality mesh generation. The alterations
made to the models are extremely small and prevent mesh
singularities that would affect the CFD simulation. Where
possible the geometry is based on the data given by
B.Bextnp et Al. [7]. Additional data is found in TU-154 by
Dmitriy Komissarov [8]. According to Komissarov the Tu-
154M features double-slotted flaps and redesigned LE slates Fig. 3. Bottom view of the 3D model of the TU-154 in cruise
defected 22° with no gap between the inner/outer wing mode.

2
 CFD ASSESSMENT OF AERODYNAMIC DEGRADATION OF THE TU-154M PLANE

In landing mode the plane is modelled with deployed any pilot counteraction of the roll of the plane is examined
landing gear. One version is modelled as the Tu-154B with by setting both the right wing outer interceptor and the right
triple-slotted flaps and rounding of the wing tips (see figures aileron to their full travel (45°) and (20°) respectively.
Fig. 4, Fig. 5 and Fig. 6),
Note the extended slats towards the tip whereas the Tu- 2.3. Flight conditions
154M is modelled with double slotted fowler flaps and the Each plane mode is evaluated for two or three angles of
slats are extended further outboard compared to the TU- attack from α = 3° to α =13°. The free stream air
154B with a more square wing tip (see figures Fig. 7 - Fig. temperature is T = 275K, the free stream air pressure p =
9). 99325 Pa and the free stream velocity V = 75 m/s. The
2.2. Pilot interaction reference area is set to S = 180 m2 for easy comparison to
The effect of no pilot counteraction (free stick) is the manufacturers data. The reference length is set as the
examined by setting both the right wing outer interceptor Mean Aerodynamic Chord L = 5.285 m.
and the right aileron to their neutral positions. The effect of

Fig. 4. Top view of a 3D model of the TU-154B in landing Fig. 7. Top view of a 3D model of the TU-154M in landing
mode. mode.

Fig. 5. Bottom view of a 3D model of the TU-154B in landing Fig. 8. Bottom view of the 3D model of the TU-154M in landing
mode. Note the triple slotted fowler flaps. mode. Note the double slotted fowler flaps.

Fig. 6. Close view of the triple slotted fowler flaps of the TU- Fig. 9. Close view of a 3D model of the TU-154M in landing
154B in landing mode. mode. Note the double slotted fowler flaps and extended slats
towards the tip.

3
Glenn Arthur Jørgensen

3. COMPUTATIONAL METHOD. drop by more than 6 orders of magnitude in the first 600
iterations.
3.1. Numerical methodology used
3.4. Alpha shift
Steady state viscous simulations are conducted using
For the case of the plane in landing mode with 36° flaps,
Reynolds Averaged Navier Stokes equations (RANS)
and 22° slats extension the overall lifting curve slope and
methodology. This approach allows a quick and reliable
shape correlates extremely well with the manufacturers data,
estimate of the pressure distribution on the body as well as
but a small alpha shift of Δα=+2° is required to bring the
the skin friction in order to evaluate the various forces and
CFD results and manufacturers data completely in line. A
moment coefficients. The turbulence closure model used is
plane in landing mode produces a large amount of lift and
Spalart-Allmaras (SA). The solution is driven to a steady
circulation which can be affected by the presence of wind
state by relaxation using a pseudo time step. Convergence is
tunnel walls. Since the AoA for a 3-D finite wing is reduced
accelerated via multigrid.
because of the induced AoA. It is the effect of this induced
3.2. Generation of Grid AoA that is different from tunnel to free flight and thus can
All meshes discussed in this report have been generated cause a shift in alpha. Such alpha shift is therefore not
by Metacomp’s Multipurpose Intelligent Meshing uncommon when comparing CFD results with wind tunnel
Environment (MIME) pre-processing package. The data or actual flight data, and it is without importance to the
geometry considered is somewhat complex and maintains conclusions drawn here.
several parts of the actual aircraft (fairings, flaps, slats,
fences). The total number of cells in the mesh for the 4. RESULTS
various modes varies between 35 million to 56 million
volume cells (see Fig. 10 and Fig. 11). To enhance accuracy, 4.1. Lifting coefficient and moment of roll
curvature refinements as well as several local refinement The overall lifting coefficient found in this work for the
boxes were used to ensure proper quality of the meshes. plane in cruise and landing configurations correlates well
with the manufactures data for the TU-154, thereby
confirming the models and method and bringing a level of
assurance that the CFD is being solved consistently. See Fig.
13.
The dimensionless pressure coefficient at the plane
boundary is shown in Fig. 12.

Fig. 10. One example of the grid generated by Metacomp's

Multipurpose Intelligent Meshing Environment. The total
number of volume cells vary between 35million and 56 million.

Fig. 12. The dimensionless pressure coefficient Cp for the TU-

154M plane in landing mode.
The local lift coefficient CL as a function of the span
coordinate y is shown in Fig. 15 and data are presented in
table 1 for the TU-154M plane in cruise mode for α = 3° and
α =13° and in landing mode for α = 5° and α =15°. An alpha
shift of +2° is observed for the plane in landing mode for
both the TU154M and TU154B2 planes. No "normalization
Fig. 11. Close look at the region of the aileron and outer techniques" are required to bring these data in agreement
interceptor of the right wing in the case of full pilot interaction.
with the reference data for the TU154M. Results presented
here correlate well with the manufacturers data for the plane
3.3. Problem setup in CFD++
in cruise and landing modes, thereby confirming the models
CFD work is performed using validated CFD++ software and method and bringing a level of assurance that the CFD
by Metacomp Inc. A steady RANS simulation is done to is being solved consistently. The total lost lift power of the
evaluate the flow around the body of the aircraft and using wing is found to be less than 8.6% for the TU154M for
the one equation Spalart-Allmaras turbulence closure. The (angle of attack, AOA or α) AOA's between 5° and 15. See
ratio of turbulent to laminar viscosity is set to 2.0. Steady Fig. 14. This is significantly lower than the 12.5% to 14%
state solution is obtained using the implicit solver and earlier claimed by Kowaleczko. The required average
ramping the Courant Friedrichs Lewy number (CFL) from 1 moment necessary to obtain a roll of the plane as logged by
to 120 in 100 steps. Residual convergence is very good on the plane's flight data recorders is earlier found by Jorgensen
the clean geometry (no flaps, no slats), and the residuals to be about CMxreq = 0.68 [3, 4]. In case of no pilot

4
 CFD ASSESSMENT OF AERODYNAMIC DEGRADATION OF THE TU-154M PLANE

interference the expected moment of roll is CMx = 0.297 or generally result in increasing AOA's in order to obtain the
44% of the required value, as seen in Tab. 2. In the case of necessary lift with the reduced remaining wing area, thus
full pilot interaction (full right aileron and outer interceptor) driving the moment of roll up.
the expected moment of roll drops to CMx = 0.117 or only The results clearly show, that a pilot interaction with full
17 % of the moment required to obtain the logged roll rate. right aileron and full right interceptor would almost entirely
In this case the pilots are nearly able to counteract the entire counteract the moment of roll associated with a wing loss of
loss of the wing tip of 5.5 m. A wing loss of about 10 m can only 5.5 m, i.e. the plane would only roll a minimum angle
generate the required moment of roll for AOA's between 10° and such loss would not lead to a crash [9]as claimed in
and 15° (see Tab. 1 and Tab. 2). Increasing wing losses will [1]. The actual values are listed in Tab. 1and Tab. 2.

Fig. 13. The overall lifting coefficients found in this work Fig. 14. Same as figure 13 but including the overall lifting
(circles) and compared to the Russian TU-154M data (lines) coefficients of the damaged wing. The loss of lift associated
[7] of the undamaged plane. The very fine correlation of both with a wing loss of 5.5 m is less than 8.6 %. The resulting
cases with and without flaps/slats confirm the model and moment of roll is less than half the value required to obtain
method and brings a level of assurance that the CFD is being the roll as recorded by the planes black boxes. For a loss of 10
solved consistently. m wing the moment of roll reaches a value, that perfectly
explains the recorded roll angle of the plane [9].

Fig. 15. The lifting distribution found in this work. The integral of area below each curve equals the total lifting coefficient for the
examined case. The full black lines correspond to the lift distribution of undamaged wing in neutral for three different angles of

5
Glenn Arthur Jørgensen

attack (5°, 10°, 15°). The dotted lines on the left side correspond to the same for the case of a loss of 5.5m wing length. The dotted
lines on the right side correspond to the same for the undamaged wing with outer interceptor and outer aileron at full extent. Note
the large simularity between the two sides, i.e. applying full intereptor and full right aileron could to a large degree cancel the
rotational moment of the loss of 5.5m wing tip. This effect is quantified in tab.1.

Tab. 1.The overall lifting coefficient found in this work, CLCFD++ , the value found by [7], CLREF , the angle of attack, α, and
the shift in α, αshift for both configurations without flaps/slats (cruise) and with 36° flaps and 22° slats (landing).

Tab. 2. The overall lifting coefficient, CL, and the moment of roll about the X-axis (CMx) is shown for the undamaged wing (Lost
span = 0m) and for a lost span of 5.5 m and 10.0 m respectively for three angle of attacks with and without full pilot interaction (by
right aileron and right outer interceptor). Note that for a wing loss of 5.5 m the pilots can nearly counteract the rolling moment,
i.e. prevent the rolling of the plane.

6
 CFD ASSESSMENT OF AERODYNAMIC DEGRADATION OF THE TU-154M PLANE

The results presented in this work do not include the simply by pulling a little more nose and thereby increasing
effect of the increased flow close to the fuselage as a result the angle of attack keeping this below the critical angle of
of the large engine thrust of motors at high speed. The effect αcritical = 17.4°.
of the increased flow around the inner portion of the wings
is to increase the lift produced in this region, and thereby VCritical, NP (G  1.25)  260km / hr , (8)
shifting more of the entire lift away from the tip and towards
the root of the wing. With other words the loss of lift 4.2.2. Case of full pilot interaction (full right aileron and
associated with a loss of say 5.5 m of the wing tip will be full right outer interceptor).
even lower than the 8.6 % found in this work when
In a similar fashion the critical speeds can be found for
including this effect. The 8.6 % is therefore a very
the case of full pilot interaction using full right aileron and
conservative maximum value, and further work including
full right outer interceptor, again in the case of a wing loss
the motor effect could be done to reveal whether this loss is
of ΔL = 5.5m. Here the reduction, χ, of the total lift
even below 7%.
coefficient caused by the activation of the right aileron and
outer interceptor must be taken into account. A good
4.2. Critical speed after wing damage estimate of χ can be found from the values of CL with and
From the results found in this work and presented in Fig. without pilot interaction for the loss of ΔL = 5.5m for α=15°
14 and Tab. 2 the planes critical speed (speed of stall) can be found in this work and listed in table 2
found for both cases "no-pilot interaction" and "full pilot
interaction" for a wing loss of ΔL = 5.5 m. 1.515
The basic lifting equation can be written as [7]
  0.952 . (9)
1.592
 airV 2 Then the maximum lifting coefficient for the case of full
Fz ,max  Cl max* S * , (2) pilot interaction can be found as
2
Where S is the aero dynamical reference area (S=180 m2 CL max L5.5m,FP  0.91*1.77 * 0.952  1.533 (10)
for the TU-154 [7]), ρair =1.272 kg/m3 and V is the speed of The critical speed in case of full pilot interaction is found
the plane relative to the air. as

4.2.1. Case of no pilot interaction (right aileron and outer M tot * g * 2 * G

interceptor in neutral). VCritical,FP (G )  , (11)
 air * S * CL max L5.5m,FP
The maximum lifting coefficient Clmax for a wing loss
of ΔL = 5.5 m is for the case of no pilot interaction found
from Fig. 14 VCritical,FP (G  1.0)  239km / hr , (12)

CL max L5.5m, NP  0.91* CL max L0 (3) VCritical,FP (G  1.25)  267km / hr . (13)
or From (12) and (13) it is shown, that the pilots with a
wing loss of just 5.5 m could continue the started upwards
CL max L5.5m, NP  0.91*1.77  1.611 , (4) acceleration similar to what they had prior to the wing loss
where ClmaxΔL=0 = 1.77 is the maximum lifting coefficient and at the same time nearly counteract the roll by pulling
of the undamaged wing [7] and the index "NP" denotes "No full right aileron and outer interceptor and a little more
Pilot Interaction" in contrast to "FP", "Full Pilot nose angle and thereby increasing the angle of attack still
Interaction". keeping this below the critical angle of αcritical = 17.4°.
The total required lift of the plane can be written as :
5. CONCLUSION
FZ  M tot * g * G , (5)
Accurate 3D Models of the Tu-154M have been
By inserting (4) and (5) in (2) the critical speed of the developed and CFD results based on these models are done
TU-154M after a wing loss of ΔL = 5.5 m can and be found using validated CFD software (CFD++). The CFD results
as show very good correlation to the Russian published lifting
data of the Tu154M for both cruise configuration, as well as
M tot * g * 2 * G
VCritical, NP (G )  , (6)
landing configuration, hereby verifying the models and CFD
 air * S * CL max L5.5m, NP method and bringing level of assurance, that the CFD is
being solved consistently.
where G is the factor of vertical acceleration. G=1 The loss in total lifting power associated with a wing loss
corresponds to a horizontal flight. Mtot = 78.600kg is the of 5.5 m is less than 8.6 % for the Tu-154M. This loss
estimated total weight of the plane [1] at the time of crash. correlates very well with CFD results of the work performed
For horizontal flight (G=1) the critical speed can hereby by Kowaleczko, when correcting his results for the many
be found as discrepancies between the 3D model used in his CFD work
and the actual Tu-154M as described in ref [4].
VCritical, NP (G  1.0)  233km / hr . (7)
Including the effect of the increased flow over the wing
From (8) it is shown, that the plane speed of V=275 roots caused by the influence of the three tail engines will
km/hr even would allow for a vertical acceleration of tend to even further reduce the loss associated with the loss
G=1.25, i.e. the pilots could continue the started upwards of the wing tip to below 8.6 %. Further work is required to
acceleration similar to what they had prior to the wing loss, show if the actual loss including this effect is below 7 %.

7
Glenn Arthur Jørgensen

For a wing loss of 5.5 m and free stick (no pilot would also like to acknowledge Senior Mech. Eng. Jan
interference), the main moment driving the roll is less than Andrzejewski for on going support and technical
50 % of the moment required to achieve a roll of the plane discussions.
as logged by the planes flight data recorders.
For full use of the right aileron and right outer
interceptor, the main moment driving the roll is less than References
20 % of the moment required to achieve a roll of the plane
as logged by the planes flight data recorders. [1] "Final Report Eng. Ver. Jan. 10th 2011". Interstate
A wing loss of effectively ΔL = 10 m provides the Aviation Committee, Air Accident Investigation
required roll moment to get good correlation to FDR data, Commission,
and leaves no possibility for the pilots to avoid a crash. [2] "Rekonstrukcja Ostatniej Fazy Lotu Samolotu TU-
The stall speed of Tu-154M with wing loss of 5.5 m is 154M." Prof. dr hab. inż. Grzegorz Kowaleczko, 31-
233 km/hr for neutral aileron and interceptor, and 239 km/hr dec-2013. (Published 04.01.2014).
for full right aileron and full right interceptor. Both speeds [3] "Reconstruction of Trajectories of Tu-154M in Smolensk
are well below the actual plane velocity of about 275 km/hr, During Last Seconds of Flight". Ms.Sc.Mech. Eng.
i.e. the plane could easily be flown and even accelerated Glenn A. Jørgensen, Materiały Konferencyjne,
upwards while keeping far from the stall limit. Konferencja Smolenska, 21-22.10.2013, ISBN 978-83-
With a wing loss of just 5.5 m the pilots could - in the 936018-1-3
time they had control over the plane - continue the started [4] "CFD results for TU-154M in landing configuration for
upwards acceleration similar to what they had prior to the an asymmetrical loss in wing length". Rev H.
wing loss and at the same time nearly counteract the roll by 06.02.2014 Ms.Sc.Mech. Eng. Glenn A. Jørgensen.
pulling full right aileron and outer interceptor and a little [5] "CFD Assessment of Aerodynamic Degradation of a
more nose angle and thereby increasing the angle of attack Subsonic Transport Due to Airframe Damage" Frink,
keeping this well below the critical angle of αcritical = N.T., Pirzadeh, S.Z., Parikh, Atkins, H.L.,Viken, S.A,
17.4°. Morrison, J.H.,The 48th AIAA Aerospace Sciences
Meeting, Orlando FL, January 2010
6. CONCLUDING REMARKS [6] " Flight Stability and Automatic Control", McGraw-Hill,
The results found in this work confirm that a loss of Inc., New York, 1989. Nelson’s book , Second edition,
5.5 m of the left wing of the TU-154M plane in Smolensk ISBN 978-0-07-046273-1
can not bring the plane to crash as claimed in the official [7] "ПРАКТИЧЕСКАЯ АЭРОДИНАМИКА САМО-ЛЕТА,
reports. Ту-154М", В. П. Бехтир, В. М. Ржевский, В. Г.
The moment of roll caused by the asymmetric wing as a Ципенко.
result of a loss of 5.5 m of the left wing is found to be less [8] "Tupolev Tu-154: The USSR's Medium-Range Jet
than half the value required to obtain the speed of roll of the Airliner (Aerofax). 2007, Dmitriy Komissarov ISBN-13:
plane as recorded by the planes flight recorders (black 978-1857802412.
boxes). This confirms earlier work in this field [3]. [9] "Reconstruction of Trajectories of Tu-154M in Smolensk
In the time the pilots had control of the plane they could During Last Seconds of Flight". Ms.Sc.Mech. Eng.
with such wing loss even accelerate upwards and at the Glenn A. Jørgensen, Materiały Konferencyjne,
same time apply the full use of the right aileron and right Konferencja Smolenska, 20.10.2014
outer interceptor without stalling.
Full use of right aileron and right outer interceptor would
to a large degree counteract the rolling moment associated
with the asymmetric wing.

7. ACKNOWLEDGMENTS
The CFD work was strongly supported by Metacomp
inc., to whom I am very great full. Metacomp has a
largeexperience within the field of CFD analysis. They work
together with several major aircraft industries using
Metacomp's validated state of the art software CFD++. I