Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

2017 International Conference on Military Technologies (ICMT)

May 31 – June 2, 2017, Brno, Czech Republic

Mathematical Model of Aircraft Ground Dynamics


Hristina Georgieva*, Vladimir Serbezov*
*Faculty of Transport, Technical university of Sofia, Sofia 1000, bul. Kl. Ohridski 8, e-mail: hgeorgieva@tu-sofia.bg
*
Faculty of Transport, Technical university of Sofia, Sofia 1000, bul. Kl. Ohridski 8, e-mail: vlados@tf10109-1.tu-sofia.bg

Abstract—This paper presents a mathematical model with 3 safety, are also determined by factors such as the runway
degrees of freedom developed in Matlab/Simulink environment surface, weather conditions and tire wear. From a commercial
for numerical simulation of aircraft ground dynamics. As a point of view, the velocity at which taxiing maneuvers are
reference aircraft Airbus A319 midsize commercial passenger performed is important, e.g. the reduction of time spent taxiing
aircraft has been chosen. The model takes into account the forces improves efficiency of operations at airports.
and moments generated by the aircraft
(aerodynamics and thrust) and the forces generated in the Computer modelling has been an invaluable tool in
landing gear tires. The model was validated against a simplified studying the aircraft ground dynamics due to the high cost of
takeoff run model derived from aircraft flight records. For real ground tests. Modelling and simulation have been used for
qualitative model assessment two cases of curvilinear aircraft the analysis of existing aircraft to perform different turning and
ground motion were simulated and analyzed. In these simulations maneuvering characteristics. An important aspect of the aircraft
the aircraft weight and center of gravity position were varied as a modelling is the nonlinearity that is introduced via the tires, the
parameter of operational relevance. steering system, the geometry and dynamic surfaces.
Since nonlinearities are known to play an important role in the
Keywords: aircraf ground dynamics, aircraft ground movement dynamics of a given system, it is important to fully incorporate
modelling, aircraft weigt distrubution, aircraft center of gravity, them as part of the model and the analysis. A number of
steering angle
nonlinear models are developed to study the aircraft ground
dynamics [4-13].
I. INTRODUCTION
Studies [4-7] proposed a mathematical model with six
The larger air traffic over the last two decades has involved degrees of freedom to investigate the aircraft dynamics on the
many challenges in face of the researchers. The aircraft ground ground. A five degrees of freedom model for aircraft in
motion is one of the areas where the pilot's skills are the key ground situation is presented in [8]. To control the yaw
factor in safe motion. The pilot has to control the thrust, dynamics of the aircraft and its maneuvering on the ground the
the rudder deflection, the brake pressure and the nose wheel authors of [9] proposed a mathematical model with two
steering angle. Moreover, the aircraft ground dynamics varies degrees of freedom. The study of a yaw rate control of the
with external conditions (runway contamination, wind, etc.). aircraft-on-ground to improve the aircraft ground handling
As results of the study conducted by International Civil qualities is discussed in [10]. A mathematical model with
Aviation Organization (ICAO) during the Global runway 5 degrees of freedom to investigate the path fallowing control
safety symposium in 2011, the runway excursions are the most problem of aircraft-on ground is developed in [11].
common type of accident reported annually, in the European In work [12] a mathematical model with 2 degrees of freedom
region and worldwide and their numbers have not decreased is present in order to improve the on-ground control system of
in more than 20 years [1]. G. W. H. van Es [2] performed aircraft. Aircraft control on the ground is one of the few areas
an analysis of the accidents data and identified the factors in in which automation has not been employed. The design of
the runway excursions. The most dominate factors are runways controllers to automate ground operations is heavily reliant on
condition and crosswind, nose wheel steering, aborting takeoff computer modelling and a greater understanding of ground
above the velocity V1, wrong takeoff weight, tire failure, maneuvers in general. J. Kardos [13] developed a
asymmetric power etc. To avoid the incidents due to the mathematical model with three degrees of freedom for
runway excursions and to improve the safety during the airport AutoTaxi control system. This system will allow fuel saving,
operation optimal aircraft ground trajectory needed to be minimization of nose pollution in the airport area by
determined. Recent studies focused on overall reduction of improving the way aircraft are operated on the ground. Such
pollutants and especially on ground noise minimization and automation system will also allow decreasing ground
exhaust gas emissions. In [3] has been evaluated the separation standards and to increase of throughput at airports
performance of different aircraft taxiing strategies like full in metropolitan areas.
engine taxiing or single engine taxiing and also electrical nose
gear implementation in order to determine the percentage The goal of this work is to present a mathematical model of
reduction in fuel consumption and emissions achieved by each ground movement of a commercial passenger aircraft that is
taxiing strategy. based on the experience of vehicle trajectory modeling [14]
and is extended to account for the aircraft aerodynamics and
The aircraft ground dynamics are governed by many landing gear geometry. As a base aircraft the Airbus A319
different aspects of its design, loading and operational midsize commercial passenger aircraft has been chosen.
practice. The handling qualities, which play a crucial part in The model has 3 degrees of freedom and the aircraft is

c
978-1-5090-5666-8/17/$31.00 2017 IEEE 514
modeled as a road vehicle on the ground. The model is realized & = F +F
in Matlab programming environment. To investigate the mVx xTL xTR - FxL - FxR - FxN cos(δ ) - FyN sin(δ ) - D
aircraft ground dynamics we use a lane-change maneuver & = F + F + F cos(δ) - F sin(δ)
mVy yL yR yN xN . (1)
widely used in automotive industry for the evaluation of the
vehicle handling performance. The reference parameters of the & z = l yR FxR - l yL FxL - l xR FyR - l xL FyL + l xN FyN cos(δ) -
Izω
model are the aircraft taxi weight, aircraft velocity, thrust, nose - l xN FxN sin(δ) + l yT FxTR - l yT FxTL
steering angle and center of gravity position.
The overall motivation for this work is deeper where the dimensions l* are given in Tab. 1. The parameter
understanding of the aircraft ground dynamics in order to center of gravity (CG) is defined as percentage along the mean
inform operational practice, the future aircraft design and the aerodynamic chord lmac, taken from the leading edge.
design of ground automation systems. The velocities along each of the aircraft axes (X, Y) are given
by Vx, Vy and the rotation velocity about axis Z by ωz.
II. MATHEMATICAL MODEL The steering angle applied to the nose gear is denoted δ.
The tires forces are denoted Fx*, Fy*, where * indicates nose
The aircraft mathematical model presented on Fig. 1 has a (N), main right (R) or main light (L) landing gear.
tricycle configuration and the nose gear is used for steering. The aerodynamics forces are denoted by D for the drag force
The aircraft is modeled as a rigid body with three degrees of and L for the lift force. The thrust forces are denoted by FxTL
freedom: two translational degrees of freedom and one and FxTR.
rotational degree of freedom. To simplify the model several
assumptions are made: To plot the aircraft ground movement the following
equations are used (2):
• The landing gear is rigid, i.e. the oleos aren’t included
in the model;
X = Vx cos (ψ) - Vy sin (ψ)
• The pitch and roll angle are small, i.e. the normal load . (2)
on the left and right main landing gears is identical: Y = Vx sin (ψ ) - Vy cos (ψ)
FzMR = FzML; ψ& = ω z
• There are two tires per gear on the aircraft;
In order to identify the causes of the different aircraft
• No incident wind during the ground motion; behavior on the ground, we developed a hierarchy of the
model. The idea is to capture the relevant dynamics is a simple
• No breaking forces applied to the tire. a model as possible. The goal of the analysis is to see what are
the effects of the tires and aerodynamics on the aircraft
dynamic behavior.

A. Tire Modeling
The tire behavior is highly complex due to the varying
pressure distribution, non-constant friction coefficient and etc.
The most representative models are based on empirical
functions, designed and tuned to fil measure data. The tire
model takes vertical Fz, longitudinal Fx, and lateral Fy effects
into account and they are calculated with a tire model proposed
in [14]. The idea of this modeling is to make a link between the
tire study in automotive vehicles and aircraft dynamics.
At lower velocities the forces generated by the tire have a
dominant effect over aerodynamics forces on the aircraft
motion.
• Longitudinal force:

Fx = Fz μ R , (3)

Figure 1. Aircftaft mathematical model. where the parameters Fz is the resultant vertical force with
respect to the aircraft weight and wing lift, acting at the center
The equations of motion (1) of the aircraft on ground are of gravity, and µ R is defined in Tab. 3.
given as: • Lateral force:

FyN = -C N α N
FyM = -C M α M , (4)

515
l xN ψ z + Vy III. SIMULATION RESULTS
α N = δ - arctan
Vx A. Takeoff Run Simulation
, (5)
l xM ψ z - Vy As a preliminary model verification a part of a takeoff run
α M = arctan simulation was compared with data taken from real aircraft
Vx
flight records, recorded on Digital Flight Data Recorder.
Assuming that the modeling of the aerodynamic forces and
where the geometric parameters lxN, lxM are defined in Tab. engine thrust is correctly verified in [15], the acceleration at
1. The slip angle (αN, αM) and the equivalent cornering stiffness takeoff run would be indicative for the correctness of the tire
(CN, CM) were obtained according to the [14]. modeling. The flight data is taken from [15]. A 30 second time
interval of the takeoff run was taken as a reference.
B. Modeling the Aerodynamics The interval was chosen to start from the moment the engine
The aerodynamic forces are defined with respect to the thrust was stabilized at its takeoff value and to end shortly
aerodynamic center of the aircraft. The forces are given by: before aircraft rotation at takeoff.
The conditions of the recorded takeoff run were set in the
1 simulation. The aircraft mass was 62 200 kg, the wing highlift
D = ρVa2 SC x devices set at position 1F (position 1.5 in the aircraft model),
2 , (6)
the airport elevation was 25 m above sea level, the deviation
1 2
L = ρVa SC y from standard temperature was +5.9 degrees K. At the start
2 moment the aircraft velocity was 26.4 m/s and engines fan
velocity – 82 %. The results of the simulation, compared with
where S is the wing area (see Tab. 1), Va is the aircraft the flight data is presented in Tab. 2.
velocity and ρ is the air density. The air density is calculated
using a standard atmosphere model described in [15]. TABLE II. COMPARISON OF SIMULATION RESULTS WITH FLIGHT DATA
The Cx and Cy are receptivity the drag and lift coefficient.
V, (m/s)
These coefficients are also obtained from a model
t=0s t = 15 s t = 30 s
developed in [15].
Flight data 26.4 58.7 80.6
Model 26.4 57.05 85.6
C. Thrust Modeling Difference - -1.65 5.0
For engines thrust simulation a simplified model of Difference, (%) - -2.81 % 6.2 %
CFM 56-5 turbofan engine was used. The model was
developed in Matlab for aircraft flight modeling [16]. The The results show good coincidence of the aircraft velocity
required thrust in the aircraft ground movement model is set as in the simulation and the recorded data especially in the first
a percentage of the maximum thrust available. The thrust interval. At the end of the second interval the difference
forces FxTL and FxTR are assumed to act parallel to the x-axis of between the velocity in the model and the record is much
the aircraft. higher, but still much less than 10 %. It should be noted that in
the second interval, because of the higher velocity the
D. Aircraft Data Source aerodynamic forces prevail over the tire forces.
A preliminary conclusion can be made that the model is good
For numerical simulation of the proposed mathematical
model the midsize commercial passenger aircraft Airbus for low and medium velocity ground movement simulations.
A319-100 has been chosen as a reference aircraft. Further experiments could be made to improve further the
The aircraft parameters are listed in Tab. 1 [17]. accuracy of the model in all velocity ranges.

TABLE I. GEOMETRIC PARAMETERS OF THE AIRCFRAT B. Lane Change Maneuver Study


Parameter Value To test the effectiveness of the proposed model in area of
Wing span, b (m) 34.1000 aircraft ground problems like a taxi maneuvering, two variants
Reference area, S (m2) 122.6000 of a simulation are studied. We define:
Mean aerodynamic choard, MAC (m) 4.1935
Aircraft length, d (m) 33.8400 • Heavy and light mass cases as 62 400 kg and
Wheel track, W (m) 7.5900 49 920 kg representing values of the maximum takeoff
Wheel base, B (m) 11.0400 weight and receptivity 80 % from maximum takeoff
x-distance from CG to the nose gear, lxN (m) 9.0688 weight;
x-distance from CG to the main gears, lxM (m) 1.9712
y-disatance from CG to the main gears, lyM (m) 1.9712 • Forward (FWD) and rear weight distribution (RWD),
y-distance from C.G. to the engines, lyE (m) 5.7500 representing forward position of center of gravity at
21 % and rear position of center of gravity at 36 %
along the aircraft mean aerodynamics chord taken from
the leading edge;

516
• Different constant thrust force during the aircraft
ground motion, receptivity T = 18.2 % of the
maximum thrust for the heavy aircraft case and
T = 14.6 % of the maximum thrust for the light
aircraft case.
Each simulation study is defined and characterized by two
parameters (Tab. 3):
• The initial velocity Vx at the corresponding thrust T;
• Steering angle δ. From the initial condition, for which
the aircraft is travelling in a straight line at a steady
velocity, the steering angle is ramped up from 0 to the
desired value, according to the maneuvering.
The higher thrust level is required in the heavy aircraft
case to achieve similar velocity to the light aircraft case.
Combining the results from these two computations into a
single plot is an effective way of representing the lane-change
maneuver dynamics.
Figure 2. Modeled longitudinal velocity of aircraft.
TABLE III. PARAMETERS OF STUDY
Parameter Value Fig. 3 shows the computed result about the aircraft lateral
Aircraft taxi weight, m (kg) 49 920 / 62 400 velocity for the case of light and heavy aircraft for two
Thrust level, % 14.6 / 18.2 different position of center of gravity. At the initial point
Steering angle applied to the nose gear, δ (degrees) 10 / -10 where δ = 0o, the mathematical model has a lateral velocity of
Coeficient of rolling resistance, µ R 0.02 Vy = 0 m/s. When the steering angle is increased to a desired
Initial aircraft velosity, Vx (m/s) 10 value δ = 10o the lateral velocity increase too. When the
steering angle returns to δ = 0o the lateral velocity decrease to
Figures (2-6) show the results of lane-change maneuver for zero. Therefore, for FWD case the computed lateral velocity
the case of light and heavy aircraft, each at different position of Vy ≈ 0.3 m/s is higher throughout compared with the RWD
of center of gravity and constant thrust level. case Vy ≈ 0.2 m/s. This means that the region of lateral
On the Fig. 2 at the initial point where δ = 0, instability could be approached more suddenly for the FWD
the mathematical model has a longitudinal velocity of case.
Vx = 10 m/s. When the steering angle is increased to the
desired value δ = 10o the aircraft velocity decrease due to the
turning maneuver. When the maneuvering finished the
longitudinal velocity starts to increase again. The solutions of
modeled longitudinal velocity for the two cases of aircraft
weight distributions and for the heavy and light mass cases are
qualitatively the same.

Figure 3. Modeled lateral velocity of aircraft.

The mathematical model has a yaw velocity of


ωz = 0 deg/s at the initial point where δ = 0o (Fig. 4). When the
steering angle is increased to a desired value δ = 10o the yaw
velocity increase too and its maximum value of ωz ≈ 9 deg/s is
realized during the turning. When the maneuvering finished

517
the yaw velocity down to 0 deg/s. The trend for the change in of gravity is shown on Fig. 6. At the initial point where δ = 0o,
the aircraft handling angle similar for all of the simulated the aircraft travels in a straight line with Vx = 10 m/s.
causes: two cases of aircraft weight distributions and heavy When the steering angle is increased to a desired value δ = 10o
and light mass cases. the aircraft will attempt to follow a smaller radius turning.
The trajectory computed with the two cases agrees very
closely during the simulations. The simulation results for the
FWD case are qualitatively the same, as compared with the
RWD case. The difference in the results is in the traveled
distance of the light aircraft which is longer when compared
with the heavy aircraft case.

Figure 4. Modeled yaw velocity of aircraft.

Fig. 5 presents the results obtained for aircraft handling


angle for the case of light and heavy aircraft for two different Figure 6. Modeled XY motion of aircraft.
position of center of gravity. At the initial point where δ = 0o,
the mathematical model has a handling angle ψ = 0. The simulation results Fig. (2-6) indicate that in the light
When the steering angle is increased to a desired value δ = 10o aircraft case the size of directional instability increase much
the handling angle increase too. After ending the steering more when compared with the heavy aircraft case and more
input the aircraft handling angle decrease to zero. The change dramatically for the forward position of center of gravity.
in the heading angle over the entire maneuver is around 62o.
The all simulated cases again exhibit qualitatively the same IV. DISCUTION AND CONCLUSION
behavior.
In this paper we present a mathematical model of a typical
mid-sized passenger aircraft. The model has been validated
against an existing model that was used in a previous study.
Specifically, a comparison between the takeoff velocities of the
two models revealed a consistent and accurate agreement over
the low and medium velocities ground movement simulation
for a particular configuration of the aircraft.
The mathematical model presented here allows the
investigation of aircraft ground dynamics in dependence on
operational and design parameters. The aircraft weight and the
center of gravity position served as the main parameters for
computed results. This is an effective way of analysis of
aircraft ground and handling dynamics in hopes of improving
the performance of the aircraft. As we demonstrate for the light
and heavy aircraft, there are qualitative changes in the solution
when the position of aircraft center of gravity is changed.
From a practical point of view, the region of lateral instability
increases in size more rapidly when the aircraft has forward
weight distribution for the light and heavy aircraft case.
Future work will involve a more extensive investigation of
Figure 5. Modeled handling angle of aircraft. aircraft ground dynamics, in particular the influence of runway
condition, braking force into the tire models, asymmetric
The aircraft movement in (X, Y) ground plane for the case loading on the landing gears during ground maneuvers and etc.
of light and heavy aircraft for two different position of center

518
REFERENCES
[1] European action plan for the prevention of runway excursions, Edition
1.0, Eurocontrol, January 2013.
[2] G. W. H. van Es, “A study of runway excursions from a european
perspective”, NLR-CR-2010-259, May 2010.
[3] M. I. MD Ithnan, T. Selderbeek, W. W. A. Beelaerts van Blokland, and
G. Lodewijks, “Aircraft taxiing strategy optimization”, TRAIL Internal
Congress, Netherlands, November 2013.
[4] E. Coetzee, B. Krauskopf, and M. Lowenberg, “Nonlinear aircraft
ground dynamics”, International Conference on Nonlinear Problems in
Aviation and Aerospace, Budapest, Hungary, June 21-23 2006.
[5] J. Rankin, “Bifurcation analysis of nonlinear ground handling of
aircraft”, A dissertation submitted to the University of Bristol in
accordance with the requirements of the degree of Doctor of Philosofy in
the Faculty of Enginerreing, 2010.
[6] J. Rankin, B. Krauskopf, M. Lowenberg, and E. Coetzee, “Operational
parameter study of aircraft dynamics on the ground” in ASME, Journal
of computation and nonlinear dynamics, 2010; 5(2):021007-021007-11.
doi:10.1115/1.4000797.
[7] J. Rankin, B. Krauskopf, M. Lowenberg, and E. Coetzee, “Nonlinear
analysis of lateral loading during taxiway turns”, Journal of Guidance,
Control and Dynamics, vol. 33 (6), pp. 1708-1717,
Noveber-Decemer 2010.
[8] M. C. Turbuk, and P. Paglione, “Aircraft ground dynamics modeling”,
VI National congress of mechanical engineering, Campina Grande,
Paraiba, Brazil, August 18-21 2010.
[9] D. Lemay, Y. Chamaillard, M. Basset, and J. P. Garcia,
“Gain-Scheduled Yaw Control for Aircraft Ground Taxiing”, IFAC
Proceedings Volumes, 18th IFAC World Congress, vol. 44 (1),
pp. 12970-12975, January 2011.
[10] J. Duprez, F. Mora-Camino, and F. Villaumé, “Robust control of the
aircraft-on-ground lateral motion”, 24th Congress of International
Council of the Aeronautical Sciences, 29 August - 3 September 2004,
Yokohama, Japan, Paper ICAS 2004-7.2.1.
[11] B. Chen, J. Yao, and Sh. S. Ge “Aircraft-on-ground path following
control by dynamical adaptive backstepping” in Chinese Journal of
Aeronautics, vol. 26 (3), pp. 668–675, June 2013.
[12] C. Roos, J. M. Biannic, S. Tarbouriech, Ch. Prieur, and M. Jeanneau,
“On-ground aircraft control design using a parameter-varying anti-
windup approach” in Aerospace Science and Technology, vol. 14 (7),
pp. 459-471, October 2010.
[13] J. Kardos, “AutoTaxi system design for aircfraft”, Excel@FIT, 2015.
[14] H. Georgieva, and L. Kunchev, “Vehicle trajectory modelling unter the
influence of lateral sliding”, III International symposium of young
researchers, TP’14, June 2014, Katowice, Poland.
[15] V. Serbezov, “Boosted Diesel Engine as an Auxiliary Power Unit for
Commercial Transport Aircraft” (PhD thesis), TU-Sofia, 2012.
[16] V. Serbezov, “Synthesis Of Simplified Sub Model Of A Turbofan
Engine For Aircraft Flight Modeling”, BulTrans-2010 Proceedings,
Sozopol, September 2010.
[17] Airbus-AC-A319, Aircraft Characteristics - Airport And Maintenance
Planning. AIRBUS S.A.S. Customer Services, Technical Data Support
and Services, 31707 Blagnac Cedex, FRANCE, 2016.

519

You might also like