Professional Documents
Culture Documents
Mathematical Model and Vibration Analysis of Aircraft With Active Landing Gears
Mathematical Model and Vibration Analysis of Aircraft With Active Landing Gears
Mathematical Model and Vibration Analysis of Aircraft With Active Landing Gears
net/publication/273337708
CITATIONS READS
11 3,694
2 authors, including:
Sivakumar Sivaprakasam
SRM Institute of Science and Technology
9 PUBLICATIONS 17 CITATIONS
SEE PROFILE
All content following this page was uploaded by Sivakumar Sivaprakasam on 23 December 2016.
Abstract
This paper deals with the study and comparison of the dynamic response of aircraft with passive and active landing gears
due to runway irregularities while the aircraft is taxying. This paper develops a detailed full aircraft mathematical vibration
model to describe an active landing gear system. The derived dynamic equations are used to analyze the active landing
gear system using proportional integral derivative (PID) controllers. The performance of this system is compared with
the passive landing gear system by numerical simulations. The active landing gear system is able to increase the ride
comfort and good track holding by reducing the fuselage acceleration, vertical fuselage displacement caused by landing
and runway excitations.
Keywords
Active landing gear, proportional integral derivative control, runway excitation
landing gear system significantly reduces the loads to 2. Full aircraft mathematical vibration
the airframe during landing and ground operations.
Investigations of development of a mathematical
model
model of a single active landing gear system with a In the full aircraft model the fuselage body or sprung
proportional integral derivative (PID) controller mass is free to roll and pitch. The sprung mass is con-
(HaitaoWang et al., 2008) has shown the improvement nected to the three unsprung masses which are front,
in performances of a passive landing gear system. rear left and rear right landing gears. They are free to
Active control technology has become popular in bounce vertically with respect to the sprung mass. The
recent years and has been applied to many systems full aircraft model contains three degrees of freedom
such as an automobile suspension system (Karnopp, (d.o.f) for the sprung mass (bounce, roll, pitch) and
1983), precision machine platform and building three d.f. for the vertical motions of the nose landing
structures. gear’s unsprung mass and the rear main landing gear’s
Figure 1 shows that the active landing gear system unsprung masses.
consists of a low pressure reservoir, a hydraulic pump,
a high pressure accumulator, a servo actuator and an
electronic controller. The passive system does not
2.1. Dynamic equations of motion
include a servo actuator, transducers and electronic Figure 2 indicates the six d.f. vibration model of the full
controllers. The transducers fitted in the landing gear aircraft. In this model u, , represents respectively the
send a signal to the electronic controllers depending on bounce, pitch and roll motion of the aircraft while
the impact conditions to actuate the servo system to u1 , u2 , u3 represents the displacement of the nose, left
supply hydraulic oil into the landing gears. The gener- and right main landing gears, a-distance from centre of
ation of active control energy is to attenuate the vibra- gravity (CG) to the nose landing gear, b-distance from CG
tions to improve the ride comfort. The above to the main landing gears, d-distance from CG to left main
theoretical and experimental studies considered the landing gear, e-distance from CG to right main landing
single active landing gear system, whereas in the present gear. Using Newton’s second law of motion, the second
work a mathematical model of the dynamics of a full order differential equations of motion describing dynam-
aircraft with all three active landing gears has been ics of the active landing gear system can be written as
developed and the equations of motion derived are For bounce motion of the sprung mass
used to demonstrate the behavior of an active landing
gear interaction with the aircraft subject to runway Mu€ þ ks1 p þ ks2 q þ ks3 r þ cs1 p_ þ cs2 q_ þ cs3 r_ þ Q1 ¼ 0
excitation. ð1Þ
Figure 2. Vibration model of full aircraft with active landing gear system subject to runway excitation.
The governing equation can be simplified as The flow quantity Qflow is calculated by
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
fu€ g ¼ ½M1 f F g ½M1 ½Cfu_ g ½M1 ½Kfug ð14Þ paccum pres
Qflow ¼ Cd wl ð16Þ
3. Full aircraft model parameters When the displacement l ðtÞ 4 0, the hydraulic oil
The model parameters of a Fokker airplane are taken would have positive flow from the accumulator in to
for numerical simulation to analyze the behavior of pas- the landing gear system and a positive control force
sive and active landing gear for runway bump input, FQ 4 0. When lðtÞ 5 0, oil is drawn from the landing
considering 20% of the aircraft’s body mass is trans- gear in to the LP reservoir so that FQ 5 0, where l ðtÞ
ferred to the nose landing gear and the 80% of the air- is the displacement determined from the controller as in
craft’s body mass is shared by the main landing gears. equation (16).
The parameters of stiffness and damping coefficient are
taken as linear evaluated from nonlinear behavioral
curves from Figures 3–8 for accurately analyzing the
5. Controller design
responses of passive and active landing gear for different A PID controller is a generic control loop feedback
runway inputs. The parameters are shown in Table 1. mechanism widely used in industrial control systems.
Assuming the damping ratio " ¼ 0.1, the damping of The PID controller designed (HaitaoWang et al.,
the structural part ofpthe unsprung mass is evaluated as
ffiffiffiffiffiffiffiffiffiffiffiffi 2008) controls the displacement lðtÞ of the servo valve
ct1 ¼ ct2 ¼ ct3 ¼ 2" kt2 m2 . of one landing gear and the error function is the differ-
ence between the reference signal and the feedback
signal as velocity signal. In this study it is applied for
4. Active control force all three landing gears. The controller design is defined
The active control force FQ is a function of the flow by
output of the servo valve. The servo valve displacement
Z t
lðtÞ is controlled by the PID controller. The controller deðtÞ
Gc ¼ kp eðtÞ þ ki eðtÞ þ kd ð17Þ
_ q,
actuates the servo valve by the signal p, _ r_ measured by 0 dt
the transducers. There is no exact relationship between
the active control force FQ and the flow quantity Qflow Gc is the current input from the controller. kp is the pro-
from the servo valve (Sharp, 1988). It is often deter- portional gain, ki and kd are the integral and derivative
mined through experiments or by empirical formula. It gain of the PID controller (Shinners, 1964; Hac, 1985;
is assumed that the active control force is described by Datta et al., 2000). r_ðtÞ represents a reference signal and
_ q,
p, _ and r_ are the feedback signal measured from the
FQ ¼ ka Qflow Qflow ð15Þ
350000
900000
300000
800000
250000
Spring force (N)
700000
Spring force (N)
600000 200000
500000
150000
400000
300000 100000
200000
50000
100000
0 0
0.0E+00 2.0E+02 4.0E+02 0.0E+00 2.0E+02 4.0E+02
Displacement (mm) Displacement (mm)
Figure 3. Main landing gear spring stiffness. Figure 4. Nose landing gear spring stiffness.
8.0E+06
6.0E+06
4.0E+06
Damping force (N)
2.0E+06
0.0E+00
–3.0E+04 –2.0E+04 –1.0E+04 0.0E+00 1.0E+04 2.0E+04 3.0E+04
–2.0E+06
–4.0E+06
–6.0E+06
–8.0E+06
–1.0E+07
Velocity (mm/s)
300000 300000
250000 250000
Spring force (N)
200000 200000
150000 150000
100000 100000
50000 50000
0 0
0 50 100 150 0 50 100 150
Displacement (mm) Displacement (mm)
Figure 6. Main landing gear tire stiffness. Figure 7. Nose landing gear tire stiffness.
sensors fitted in the nose landing gear, left main landing Similarly the displacement at the left and right main
gear and right main landing gear respectively. landing gear servo valves from the corresponding inde-
pendent PID controllers are given by
i:e, eðtÞ ¼ r_ðtÞ u_ a_ h_ u_ 1 ðtÞ ð18Þ
l2 ðtÞ ¼ kp r_ðtÞ u þ b_ d_ u_ 2 ðtÞ
The output signal of the controller at the nose land-
þ ki rðtÞ ½ðu þ b d u2 ÞðtÞ
ing gear gives the displacement of the nose landing gear
servo valve as þ kd r€ðtÞ ðu þ b€ d€ u€ 2 ÞðtÞ ð20Þ
l1 ðtÞ ¼ kp r_ðtÞ u_ a_ h_ u_ 1 ðtÞ l3 ðtÞ ¼ kp r_ðtÞ u þ b_ þ e_ u_ 3 ðtÞ
þ ki rðtÞ ½ðu a h u1 ÞðtÞ þ ki rðtÞ ½ðu þ b þ e u3 ÞðtÞ
þ kd r€ðtÞ u€ a€ h€ u€ 1 ðtÞ ð19Þ þ kd r€ðtÞ ðu þ b€ þ e€ u€ 3 ÞðtÞ ð21Þ
P 0.1 0 0
1.0E+07
PI 0.1 1.0 0
Damping force (N)
0.035
0.03
Height (m)
0.025
0.02
0.015
0.01
0.005
Figure 10. Runway bumps.
0
0 1 2 3 4 5 6 7 8 9 10
Time (s)
of height 60 mm for the right main landing gear and ug1 ¼ 40ð1 cos !tÞ 0:2 t 1:0 and
ð22Þ
100 mm for the left main landing gear. Duration of 0 otherwise
impulse is 0.8 s and frequency is 7.85 rad/s over which
the airplane travels, and is described by equations (23) ug2 ¼ 60ð1 cos !tÞ 2:6 t 3:4 and
and (24). ð23Þ
0 otherwise
Figure 13. The sprung mass bounce acceleration of the aircraft with passive and active landing gear.
Figure 14. The sprung mass pitch acceleration of the aircraft with passive and active landing gear.
Figure 15. The sprung mass roll acceleration of the aircraft with passive and active landing gear.
reduced by 49.7% and the settling time is also reduced to using the active landing gear system. The sprung mass
33% with the active landing gear system. Thereby the bounce acceleration for the active and passive landing
aircraft taxies more smoothly, crew/passenger comfort is gear when the aircraft is taxying above the runway with
improved and a better runway holding is achieved by half sine wave bumps in the frequency domain have been
Figure 16. The sprung mass displacement of the aircraft with passive and active landing gear.
Figure 17. The shock strut travel of passive and active landing gear.
compared with ISO-2631 human exposure to whole line, while the levels for the passive landing gear
body vibration curves in Figure 18. are below the 8 hr line only, indicating the higher
It can be noted that RMS acceleration levels in comfort level provided by the active landing gear
the case of active landing gears are below the 24 hr system.
Figure 18. Comparison of sprung mass bounce acceleration of the aircraft in frequency domain with the human exposure to
vibration ISO-2631 curves, when it is taxying on runway with the bumps.
Table 5. Road roughness standard deviation. 8. Dynamic response of the aircraft for
Road class ð103 mÞ ð0 Þ 106 m3 , 0 ¼ 1 a (rad/m) a random runway input
A (very good) 2 1 0.127 Apart from the sinusoidal bumpy runway input, a
B (good) 4 4 0.127 real runway surface taken as a random exciting func-
C (average) 8 16 0.127 tion is used as input to the aircraft. The random
D (poor) 16 64 0.127 road profile is generated by a shaping filter method
E (very poor) 32 256 0.127
(Goodall, 1983; Zhang et al., 2002; Giua et al., 2004). The
profile can be approximated by power spectral density
(PSD) distribution
2V 2
0.01 ð!Þ ¼ ð25Þ
!2þ 2 V2
0.005
where 2 denotes the road roughness variance (m2), V
Road profile signal (m)
Figure 21. The sprung mass acceleration of passive and active landing gear for random runway input.
Figure 22. The sprung mass displacement of passive and active landing gear for random runway input.
Figure 23. The shock strut travel of passive and active landing gear for random runway input.
where !(t) is a white noise process with the spectral given in Table 5. In the present case runway excitation
density ð!Þ. for a grade E road has been considered. The road profile
The random road generation is done through of the grade E road is as shown in Figure 20.
MATLAB/Simulink as in Figure 19. The road rough- The sprung mass acceleration, displacement and
ness standard deviations for various types of roads are as shock strut travel for the passive and active landing
Table 6. Root mean square (RMS) values of sprung mass acceleration, vertical displacement and shock strut
travel.
Figure 24. Comparison of fuselage bounce acceleration of the aircraft in frequency domain with the human exposure to vibration
ISO2631 curves when it is taxying on the grade E random runway.
gear systems for grade E random runway excitation et al., 1986; ISO 2631-1, 1997) human exposure to
are shown in Figures 21, 22 and 23 respectively. whole body vibration curves in Figure 24.
From the figures the rms values are taken for compari- It can be noted that rms acceleration levels for the
son of passive and active landing gears and are tabu- case of active landing gears are below the 24 hr line,
lated in Table 6. while those for the passive landing gear are below the
From Figures 21–23 and Table 6, for random input, 4 hr line only, indicating the higher comfort level pro-
it can be noted that the acceleration of the sprung mass vided by the active landing gear system.
has been reduced by 47.6%, which shows the improve-
ment in ride comfort. The sprung mass displacement
and suspension travel are also reduced, which will
9. Conclusions
improve the life of the landing gear system. Therefore A mathematical vibration model of an aircraft with
it can be concluded that the active landing gear system active landing gear system has been developed and its
has better performance capabilities over the passive performance simulated using MATLAB/Simulink. In
landing gear system. Also, the fuselage acceleration this system a PID controller has been used and the
for the grade E random runway excitation in the fre- gains of the PID controllers have been tuned using
quency domain is compared with ISO-2631(Griffin the Ziegler-Nichols method. The outputs of the
independent controllers are used to operate the servo Density of hydraulic fluid
control system which applies the control forces in the FQ Active control force
respective active landing gear. The fuselage acceleration ti Time integral
levels, displacement levels and the shock strut travel for td Time derivative
the active and passive landing gears have been com- kp Proportional gain
pared. The fuselage acceleration levels have also been ki Integral gain
compared with the ISO 2631 human exposure to whole kd Derivative gain
body vibration curves. It is observed that there is reduc- ug1 , ug2 , ug3 Ground excitation
tion in the magnitude of acceleration and the displace- u1 ,u2 ,u3 Sprung mass displacement
ment of the aircraft’s fuselage by the active landing gear V Aircraft speed
system when travelling over an uneven runway surface. Road roughness variance
Thus the active landing gear system improves crew and road surface type coefficient
passenger comfort and increases the fatigue life of the
aircraft structure and landing system.
Funding
This research received no specific grant from any funding
agency in the public, commercial, or not-for-profit sectors.
Nomenclature
References
M Sprung mass Bender EK and Beiber M (1971) A feasibility study of active
m1 Nose gear unsprung mass landing gear. AFFDL Technical Report-70-126, U.S.
m2 Rear left gear unsprung mass Airforce (available as AD 887451L).
m3 Rear right gear unsprung mass Catt T, Cowling D and Shepard A (1993) Active landing gear
ks1 Nose gear sprung mass stiffness rate control for improved ride quality during ground roll,
ks2 Rear left gear sprung mass stiffness rate AGARD Smart structures for aircraft and spacecraft,
ks3 Rear right gear sprung mass stiffness rate Conference Proceedings-531, Stirling Dynamics Ltd,
Bristol.
cs1 Nose gear sprung mass damper rate
Currey NS (1998) Aircraft Landing Gear Design: Principles
cs2 Rear left gear sprung mass damper rate and Practices. Washington: AIAA Education Series.
cs3 Rear right gear sprung mass damper rate Daniels JN (1996) A method for landing gear modeling and
kt1 Nose gear unsprung mass stiffness rate simulation with experimental validation. NASA
kt2 Rear left gear unsprung mass stiffness Contractor Report 201601.
rate Datta A, Ho MT and Bhattacharyya SP (2000) Structure and
kt3 Rear right gear unsprung mass stiffness Synthesis of PID Controllers. London: Springer.
rate Freymann R (1987) An experimental–analytical routine for
ct1 Nose gear unsprung mass damper rate the dynamic qualification of aircraft operating on rough
ct2 Rear left gear unsprung mass damper runway surfaces. AGARD Report-731.
Freymann R (1991) Actively damped landing gear system,
rate
Landing Gear Design Load Conference No. 20, AGARD
ct3 Rear right gear unsprung mass damper
Conference Proceedings-484.
rate Freymann R and Johnson W (1985) Simulation of aircraft
½ M Mass matrix taxi testing on the AGILE Shaker Test Facility, Second
½C Damping matrix International Symposium on Aero elasticity and Structural
½K Stiffness matrix Dynamics sponsored by Deutsche Gesellschaft fur Lufund
½F Force vector Raumfahrt e.V.in Aachen, W.Germany, April 1–3.
fu€ g Acceleration vector Giua A, Melas M, Seatzu C, et al. (2004) Design of predictive
fu_ g Velocity vector semi active suspension system. Vehicle Systems Dynamics
fug Displacement vector 41(4): 277–300.
Pitch angle Goodall RM (1983) Active controls in ground transportation
Roll angle – a review of the state-of-the-art and future potential.
Vehicle Systems Dynamics 12(4): 225–257.
" Damping ratio
Griffin MJ, McLeod RW, Maseley MJ, et al. (1986) Whole
Qflow Fluid flow quantity from servo valve body vibration and aircrew performance. Institute of
Cd Coefficient of discharge Sound and Vibration Research, Technical Report-132,
w Gradient area of servo valve University of Southampton, Southampton, England.
l Displacement of servo valve Hac A (1985) Suspension optimation of a 2-dof vehicle model
paccum High pressure in accumulator using a stochastic optimal control technique. Journal of
pres Low pressure in reservoir Sound and Vibration 100(3): 347–357.
HaitaoWang, Xing JT, Price WG, et al. (2008) An investiga- load-control system in the main landing gear. NASA
tion of an active landing gear system to reduce aircraft Technical Paper 1555.
vibrations caused by landing impacts and runway excita- Ross I and Edson R (1982) Application of active control
tions. Journal of Sound and Vibration 317: 50–66. landing gear technology to the A-10 aircraft. NASA CR-
Horta LG, Daugherty RH and Martinson VJ (1999) 166104.
Modeling and validation of an A6-Intruder actively con- Sharp JJ (1988) Basic Fluid Mechanics. London:
trolled landing gear system. NASA TP-1999-209124. Butterworths.
Howell WE, McGehee JR, Daugherty RH, et al (1991) Sheperd A, Catt T and Cowling D (1992) The simulation of
F-106B airplane active control landing drop test perfor- aircraft landing gear dynamics, 18th Congress of the
mance, Landing Gear Design Loads Conference No. 21, International Council of the Aeronautical Sciences,
AGARD Conference Proceedings-484, Povoa dc Varzim, Beijing, China, September 20–25, ICAS-92-1.7.1.
Portugal, October 8–12. Shinners SM (1964) Control System Design. NewYork: Wiley.
ISO 2631-1 (1997) Mechanical Vibration and Shock – Simulink (1997) Dynamic system simulation for MATLAB,
Evaluation of Human Exposure to Whole Body Vibration. version 2- using Simulink, The Math Works Inc.
Part 1: General Requirements. Geneva: International Wignot Jack E, Durup Paul C, Gamon et al. (1971) Design
Standards Organization. formulation and analysis of an active landing gear.
Karnopp D (1983) Active damping in road vehicle suspension AFFDL-Technical report 1: 71–80.
system. Vehicle Systems Dynamics 12(6): 291–316. Zhang LJ, Lee CM and Wang YS (2002) A study on non sta-
McGehee JR and Garden HD (1979) Analytical investiga- tionary random vibration of a vehicle in time. International
tion of the landing dynamics of a large airplane with a Journal of Automotive Technology 3(3): 101–109.