International Journal of Intelligent Computing and Cybernetics

Flight control iron bird based on passive electric-hydraulic servo loading system
Yong Tian Tian Li
Article information:
To cite this document:
Yong Tian Tian Li, (2011),"Flight control iron bird based on passive electric-hydraulic servo loading system",
International Journal of Intelligent Computing and Cybernetics, Vol. 4 Iss 1 pp. 111 - 137
Permanent link to this document:
http://dx.doi.org/10.1108/17563781111115822
Downloaded on: 23 June 2016, At: 01:14 (PT)
References: this document contains references to 7 other documents.
To copy this document: [email protected]
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

The fulltext of this document has been downloaded 406 times since 2011*

Access to this document was granted through an Emerald subscription provided by emerald-srm:203785 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.

*Related content and download information correct at time of download.

 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1756-378X.htm

Flight control
Flight control iron bird based iron bird
on passive electric-hydraulic
servo loading system
111
Yong Tian
School of Aeronautic Science and Engineering, Received 5 April 2010
Beijing University of Aeronautics and Astronautics, Beijing, Revised 18 November 2010,
People’s Republic of China, and 3 December 2010
Accepted 9 December 2010
Tian Li
Shenyang Aircraft Design Research Institute of AVIC, Shenyang,
People’s Republic of China and
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

School of Aeronautic Science and Engineering,

Beijing University of Aeronautics and Astronautics, Beijing,
People’s Republic of China

Abstract
Purpose – The purpose of this paper is to develop a novel type of full-size flight control iron bird
based on a passive electronic hydraulic servo loading system.
Design/methodology/approach – On the basis of mathematical modeling of passive loading
system math model, the detailed design process of the flight control iron bird is presented.
Subsequently, the system digital simulation and physical verification are also given.
Findings – Experimental results show that the proposed approach can reduce the redundant forces
and improve the system dynamic and force-tracking accuracy.
Practical implications – This newly-developed flight control iron bird system has been
successfully applied in the flight control system design of some fighters.
Originality/value – The proposed approach for flight control iron bird is new and significant for the
design of fighter flight control systems.
Keywords Flight control, Dynamic loading, Aircraft, Simulation
Paper type Technical paper

1. Introduction
During the design process of fighter flight control system, a general method for
aerodynamic load simulation is by means of the dynamic load simulation system (Parker,
1999; Gerkens et al., 2004; Cieslak et al., 2010). Among the dynamic load simulation system,
the electric-hydraulic servo loading system is widely used at the present stage (Karpenko
and Sepehri, 2006). It is a sophisticated machine-electric-hydraulic composite servo
system, which involves the comprehensive content with system dynamics, hydraulic
servo drive, control theory, computer control technology and other disciplines. International Journal of Intelligent
The flight control iron bird console electric-hydraulic servo loading system is used to Computing and Cybernetics
Vol. 4 No. 1, 2011
simulate the rudder surface aerodynamic force load. It provides the realistic simulation pp. 111-137
environment for the flight control actuators in the tests such as flight control system q Emerald Group Publishing Limited
1756-378X
rudder loop test and flight control system closed-loop test to make the results DOI 10.1108/17563781111115822
 IJICC of the tests more reliable (Patel et al., 2006). It becomes an indispensable part in
4,1 environment of flight control system physical tests. In tests, the main function of the
system is to achieve the aerodynamic hinge moment of rudder surface simulation
according to the aerodynamic load spectrum of aerodynamic rudder under different fight
state, and test the effect to the flight control system dynamics quality (Tan et al., 2010).
This paper proposed a new approach for a type of fighter flight control iron bird,
112 which is based on passive electronic hydraulic servo loading system. On the basis of
the mathematical modeling of passive loading system math model, the detailed design
process of the flight control iron bird is presented. The system digital simulation and
physical verification are also given. The detailed experimental results show that the
proposed approach can reduce the redundant forces and improve the system dynamic
and force tracking accuracy.

2. Passive loading system math model

Fight control iron bird console fighter rudder surface aerodynamic simulation loading
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

system is the passive electric-hydraulic servo loading system. Its structure principle
indicates that if the forced object is a position system whose movement is determined
mainly by system itself and the loading system only plays the role to apply load, this
kind of loading system is called passive electric-hydraulic servo loading system which is
also called position disturbance loading system. The load simulator shown in Figure 1
belongs to this kind. Right half of the figure is the loaded object, usually the booster of
fighter or missile and systems such as rudder and surface. Left half is the load simulator,
which is used for loading for position system in the right part, such as simulation
aerodynamic force and so on. The fight control iron bird console fighter rudder surface
aerodynamic force simulation loading system designed and used in this paper is this
kind of passive electric-hydraulic servo loading system (Tian, 2010). The characteristic
of this kind system is to apply every kind of load to the position system, but the
movement of the position system is according to various laws. The movement of loading
hydraulic cylinder piston is determined according to the join forces of load given by itself
and control force given by position system. This kind of loading system is a complex
form of electric-hydraulic servo system. The external interference (redundant force)

Aerodynamic load Aircraft

simulation
Disturbance

Compensating
element

Force
command Force Forced
Controller Actuator
agency object
–
Figure 1.
The fighter rudder
surface aerodynamic Force
loading principle feedback
structure diagram Load system Iron bird console
 existing in the system is the key factor to affect the loading system performance quality. Flight control
So, it is needed that analyze this system carefully, and establish the correct mathematical
model and control compensating measures. And adjust and improve the loading system
iron bird
control compensating parameters in the simulation and test, which eliminate the
redundant force effectively and improves the system-tracking accuracy.
In the loading process, the movement of loaded object is determined by itself rather
than driven by loading system. Such loading system is called passive loading system 113
(also called position disturbance force system). For example, in the fighter rudder
surface aerodynamic force simulation loading process, the displacement of forced object
is changing which is controlled by the rudder given signal. This system is much more
complicated than no disturbance loading system, which is because that the movement of
the forced object will cause a great disturbance – redundant force in the loading system.
If improper handle, the magnitude of the redundant force will be much larger than the
required maximum Fomax of the force function. It may also appear the phenomenon that
the loading force function F cannot track the command force function F0. The fight
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

control iron bird console electric-hydraulic servo loading system in this paper is the
passive electric-hydraulic servo loading system mentioned above. The work principle
structure diagram is shown in Figure 1.
Suppose that the translation quality of the loaded object is m. The elasticity
coefficient is Ks. The displacement is Yf. The displacement of the loading actuating
cylinder piston stick is Y (Figure 1). Therefore, the following equations can be list
(ignore the friction force and damping):
9
mY € ¼ Ff þ F =
ð1Þ
Ff ¼ K s ðYf 2 YÞ ;

where Ff is the control force of the position system (the output force of the rudder).
The Laplace transform of the equations above are as follow:
)
m ys2 ¼ Ff þ F
Ff ¼ Ks ðyf 2 yÞ
Solve the following equation:
K sY f þ F
y¼ ð2Þ
m ys2 þ Ks
The loading flow equations of the servo valve are as follows:
QL ¼ Kx X 2 KP PL ð3Þ
X ¼ Wo U ð4Þ
where:
QL ¼ Load flow, m3/s.
X ¼ Valve core window opening, m.
PL ¼ Load pressure, Pa.
KX ¼ Flow gain, m3/s.
KP ¼ Flow-pressure coefficient, m5/S · N.
 IJICC Wb ¼ Servo valve transfer function.
4,1 U ¼ Output voltage of correction and amplifying elements, V.
The load flow equation of the hydraulic loading actuating cylinder is as follows:
V
QL ¼ ASy þ Ct PL þ SPL ð5Þ
114 4 be
where:
A ¼ Area of loading actuating cylinder piston, m2.
Ct ¼ Total leakage coefficient, m5/S · N.
V ¼ The total volume of pipe between load loading actuating cylinder and servo
valve, m3.
be ¼ Liquid elasticity modulus coefficient, Pa.
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

y ¼ The displacement of the loading actuating cylinder piston, m.

The correction element equation is as follows:
U ¼ Ky W2 ðFo 2 KF FÞ ð6Þ
where:
Ky ¼ Amplifier gain.
W2 ¼ Correction element transfer function.
KF ¼ Force-electric conversion coefficient.
F ¼ Loading force function.
Fo ¼ Force command function (command signal).
Combine the equations (3) and (5) to obtain the following solution:
V
KX X 2 K P P L 2 C t P L 2 SPL
V 4 be
KX X 2 KP PL ¼ ASy þ Ct PL þ SPL y ¼
4be AS

 
1 KX KP þ C t VS ð7Þ
¼ X2 þ PL
S A A 4 be · A

 
1 KX VS Kt
¼ ·X 2 þ PL
S A 4be · A A
where:
Kt ¼ The total flow pressure coefficient, Kt ¼ KP þ Ct , m5/S · N.
Obtain the load system dynamic equation combining by servo valve and load actuator
form equation (7):


KX VS Kt
ys ¼ X2 þ APL
A 4be A2 A2
 That is: Flight control
KX
· x 2 ys iron bird
APL ¼ A ¼F
V·S Kt
þ
4be A2 A2
Multiply the numerator and denominator by (A2/Kt): 115
AKX A2 AKX A2
X2 ys X2 y·s
K Kt K Kt
F¼ t ¼ t ð8Þ
V·S TQ S þ 1
þ1
4be Kt
where:
V
TQ ¼
4be Kt
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

Therefore, the transfer function block diagram of the passive loading system (position
disturbance electric-hydraulic loading system) obtained by equations above is shown in
Figure 2.
Combine equations (2), (4), (6) and (8) to obtain the closed loop transfer function that
system output force F to the input F0. Then, make yf ¼ 0. Because that bandwidth of
correction element and servo valve is much larger than that of load system, make
W2 ¼ 1 and W0 ¼ K0 and FðsÞ ¼ ððFÞ=ðF0 ÞÞ:


m 2
S þ 1 K y Ko K x A
Ks
¼

mv 3 m   V A2  
S þ Kt þ Ky Ko Kx AKF S2 þ þ S þ Kt þ Ky Ko Kx AKF
4 b e KS Ks 4be Ks
ð9Þ
The transfer function that system output force F to the position disturbance Yf is as
follows (make F0 ¼ 0):

yf K2S S
mS2 + KS mS2 + KS

D y· C

A2

F0 1
Kt W2 W0KZA Kt F
Figure 2.
– TQ S + 1
The transfer function
block diagram of
position disturbance
electric-hydraulic
KF loading system
 IJICC F A2 S
FðsÞ ¼ ¼
 ð10Þ
4,1 Yf mv 3 m   V A2  
S þ Kt þ Ky Ko Kx AF S2 þ þ S þ Kt þ Ky Ko Kx AF
4be KS Ks 4be Ks
Combine equations (2) and (8) to determine the math model of the passive loading
system which is shown as follows:
116 AKX A2 Ks YfðsÞ þ FðsÞ
· XðsÞ 2 S
K Kt ms2 þ KS
FðsÞ ¼ t
TQ S þ 1
Obtain F(s) after shifting and merging:


AKX m 2 A2
· XðsÞ S þ1 2 SYfðsÞ
K Ks Kt
FðsÞ ¼ t

  m 2 A2 S
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

TQ S þ 1 S þ1 þ
Ks Ks Kt
Put TQ ¼ (V/4beKt) into the equation above and arrange:


AKX m 2 A2
· XðsÞ S þ1 2 sYfðsÞ
Kt Ks Kt
FðsÞ ¼
 ð11Þ
mv m 2 V A2
S3 þ S þ þ Sþ1
4 b e KS Kt Ks 4 b e Kt Ks Kt
It is can be seen from the equation once again that:
A2
SyfðsÞ
Kt
is the redundant force which is proportional to the square of the loading hydraulic
cylinder area A, inversely proportional to the total pressure flow coefficient Kt, and
proportional to the motion velocity of the forced object.
In Figure 2, it is shown that the system is the first-order no difference system to the
position disturbance Yf. So, the loading function is unrelated to Yf after stability. But, if
solve the transfer function of:
F
;
_
Yf
it is gotten that:
F A2
FðsÞ ¼ ¼

_f
Y mv 3 m   2 V A2  
S þ Kt þ Ky Ko Kx AKF S þ þ S þ Kt þ Ky Ko Kx AKF
4 b e KS Ks 4be Ks

It is the zero-order no different system. When S ! 0:

F A2
¼ –0
_ f Kt þ Ky Ko Kx AKF
Y
 _f
It is obvious that the velocity of loaded object will affect the output. The exchange of Y Flight control
will lead to the exchange of loading force F. The so-called redundant force can be
generated as follows:
iron bird
FfðsÞ _
F¼ YfðsÞ
S
Because the system is affected by the load parameters, the frequency band is low
generally and the redundant force is usually large. The larger redundant force must lead
117
that the loading precision greatly decrease, which makes the point to point tracking
different to achieve, even the phenomenon “top cow” which makes the loading different
to achieve. The compensating and corrective measures must be applied into the control
system to reduce or eliminate the redundant force.

3. Certain type of fighter fight control iron bird console electric-hydraulic

servo loading system
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

3.1 Main function

The main function of loading system is that in the flight control physical simulation test,
according to the aerodynamic load spectrum of every aerodynamic rudder surface under
various flight conditions, achieve the simulation for the fighter rudder surface aerodynamic
hinge moment to detect the effect to the fight control system dynamics quality.
The fight control iron bird console loading system uses the all-digital, distributed,
multi-modal-integrated control technology. Loading computer control system real-time
collects fight simulation system, and calculates the rudder surface aerodynamic load from
the fighter motion parameters, such as flight altitude, airspeed, rudder angle and so on, as
the force control command signal. It also real-time collects displacement and loading force
feedback signal as the control compensating parameters and sends commands to the
electric-hydraulic servo valve through the servo amplify after the loading computer
control system integration and control law calculation to control the real-time force
tracking of the loading actuator (Tian, 2010). At the same time, it follows the motion of iron
bird console rudder surface to complete the simulation the aerodynamic hinge moment on
the fighter rudder surface. The load system control principle is shown in Figure 3.

3.2 System basic components

The fighter iron bird console loading system is composed with the elements of load
control system, including computer and controller, the force and displacement sensors,
loading actuator and connection structure and subsystems. The structure of the fighter
iron bird console loading system is shown in Figure 4, and the pictures of the exact
system are shown in Figure 5.

3.3 The system basic configuration requirement

3.3.1 The loading computer control system. The loading computer control system is
composed with a monitoring computer, real-time controller, reflective memory optical
fiber real-time net card and so on.
(1) Monitoring computer. The monitoring computer is the user interface layer of
loading control system. The advanced PC machine (DELL company, commercial
grade) can be used as the monitoring computer. Its main function is to achieve the test
resource management, including sensor calibration, servo control parameters setting
(servo valve current, mechanical zero bias), test parameters setting, test execution
 IJICC
4,1
Feedback

118 Person feel

Computer
Servo
Booster
Aircraft
system valve simulation

Loading computer Loading Force

Downloaded by Ohio University At 01:14 23 June 2016 (PT)

control system actuator sensor

Position
Figure 3. Xi disturbance
The loading system Force feedback
control principle diagram
Aircraft motion parameters

Flight simulation
Monitor computer computer Printer
Hydraulic control panel

Ethernet/fiber

Real-time controller

Security control line

Figure 4.
The fighter iron bird
console loading
system structure Hydraulic oil Hydraulic oil
Hydraulic actuators road source/sub-station

and monitoring and test data analysis and processing, all of which are operated by
monitoring computer.
The functions are as follow:
.
System fault diagnosis: fault detection and system failure treatment.
.
Hardware of all connections in system management.
.
All tests management and new construction.
.
Activation the selected test.
.
The feedback signal calibration.
 Flight control
iron bird

119
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

(a) (b)

Figure 5.
Iron bird system
(c)

.
The real-time communication with the iron bird console simulation computer
establishment.
.
Channel distribution: choose the test console control channel random.
.
The corresponding relationship between the logical channel of the test and the
physical channel of the hardware establishment.
. Every channel setting, including channel sensor range, control mode (force,
displacement), PID parameters (maximum and minimum current, zero bias,
polarity), channel safety limit (force, displacement, absolute limit, relative limit,
velocity limit and so on).
.
Loading channel command value setting: special performance function, script
program and difference calculation with table.
.
Data collection setting: collecting mode and sample rate.
 IJICC .
Test operation: soft panel mode, with which users can edit the panel form
4,1 themselves.
.
When starting the test, make the command value equal to the feedback value
achieve the effect of smooth starting.
.
Test monitoring: real-time show the command and feedback signal (value, curve
graph or bar chart) and real-time monitor the test state.
120 .
Breakpoint data storage: the test data can be saved at any time according to user
needs, and processed by word processing software, spreadsheet software,
MATLAB software and so on.
(2) Real-time controller. The real-time controller is composed with the elements of data
collection, resolving, control regulation, signal conditioning, servo driven amplifier,
and so on. It collects the rudder surface aerodynamic loads calculated by fighter
simulation computer through the reflective memory optical fiber real-time net card, and
sends the loading force control commands though digital closed loop control
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

compensation and PID regulation to complete five channels loading control. And, it
maintains the communications with the monitoring computer, and receives commands
sent by monitoring computer to take security measures for loading system.
The real-time controller uses the Versa Module Europe zero-slot controller (embedded
computer). It can real-time schedule with multi-task and complete the closed loop
control. Its operating system is the real-time operating system VxWorks with high
performance and scalable performance, and its resolving control period is , 1 ms.
3.3.2 Configuration of electric-hydraulic servo valve. Among the five channels of the
loading system, two channels of hirao select the United States MOOG servo valve. Every
loading channel configures the electric-hydraulic servo valve according to itself needs
and loading channel actuating cylinder velocity requirements. The flow is calculated as
follows:
Qmax ¼ Amax £ Vmax £ 60=1; 000 l=min

Then, select the servo valve rated flow up a block. The best configuration of the servo
valve pre-opening is determined by test commissioning after calculation. The dynamic
quality of servo valve should meet the system designing requirement and the bandwidth
large enough.
3.3.3 Sensor selection. According to the system requirement, configure seven pull
pressure sensors and seven displacement sensors whose accuracy is not lower than
0.05 per cent. Two channels of hirao select foreign high-quality sensors.
3.3.4 Loading control cabinet. The loading control cabinet is composed with
computer chassis, signal conditioner template chassis, model-driven template chassis,
power chassis, control panels and so on.
3.3.5 Hydraulic oil source. For the loading system, larger oil source pressure and flow
is useful to improve the dynamic quality of the loading system. According to the motion
velocity and load requirements of loading system and considering the hydraulic
components standard, the selected hydraulic oil source pressure is 21.0 MPa and the
total flow is 500 l/min.
3.3.6 The designing option of hydraulic actuating cylinder. The designing parameters
of the hydraulic actuating cylinder directly affect the dynamic performance and
redundant force of the loading system. The effective area and volume of the loading
 actuating cylinder piston should be as small as possible at the condition that satisfy Flight control
the static index after calculation. iron bird
3.3.7 System function requirement.
(a) System simulation. The loading system should have the system digital simulation
function for system self-test, self-calibration and commissioning. And the system can
real-time simulate the work condition of the electric-hydraulic servo loading system.
(b) Security protection. The loading system should have the security protection 121
measures. During the test, if abnormal or overload, the system should be able to
automatic alarm, and take the security protection measures in time.
(c) Monitoring condition. The loading system should have the real-time monitoring
interface and a good man-machine interface which is facilitate for test personnel to control
the loading system and set the initial conditions. At the same time, the system may accept
the remote control, and receive “prepare”, “start”, “reset”, “stop”, “emergency” or other
control instructions from the test console. Then, give the response information in time.
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

4. System digital simulation and physical verification

The system digital simulation is the important component both in the system dynamic
analysis and control integration (Tian, 2010). It is also the important part of the system
designing certification:
.
Determine the crucial components of system designing and the type and
parameters of the electric-hydraulic servo valve through system simulation.

The servo valve, which is an important component for loading system, has crucial
influence for the loading performance. The electric-hydraulic servo valve is including
pressure valve and flow valve two kinds. In this section, the digital simulation and the
test commissioning are caught on as the same technical requirements to the loading
systems using the selected pressure valve and flow valve. And then, determine the type
and pre-opening parameters. In the system, there are the system parameters only for
pressure valve, which is the same as the condition of flow valve. So, the system
optimization parameters are determined for them, respectively, before simulation.
Because of the comparison, non-linear factors are not considered at this time:
.
Construct the system model and control structure, and carry on the digital
simulation for the system control compensating scheme.

Through the digital simulation analysis, determine appropriate control parameters for
the compensating scheme of the loading system. Then evaluate the control scheme,
system dynamic performance and accuracy:
.
Amend the system control parameters by ground physical commissioning,
and validate the control compensating scheme.

Based on the control system scheme and parameters determined by digital simulation,
amend the system control parameters by ground physical commissioning. And analyze
and compare the effect achieving by the selected control scheme and control parameters
used in the practical engineering application. Then, determine the appropriate
control program and control parameters, which make the system achieving the
satisfactory effect.
 IJICC 4.1 Modeling and simulation analysis for flight control iron bird loading system
4.1.1 System work principle and math model. Take the flight control iron bird console
4,1 aileron channel loading system of a certain type of fighter as the object for simulation.
The work principle of the loading system is shown in Figure 6.
The loading actuating cylinder load flow equation is:
Vt
122 QL ¼ A1 SY1 þ Kce PL1 þ SPL1 ð12Þ
4be
The motion equation of loading actuating cylinder is:
F1 2 F* ¼ m1 S2 Y1 ð13Þ
where, F * is positive when it is pressure, and F1 ¼ A1PL1.
The transformation of the force sensor is:
F* ¼ Ks ðiy2 þ y1 Þ ð14Þ
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

where, i ¼ 1.502 is the transmission ration, Ks is the stiffness of the force sensor.
The known parameters are:
.
The maximum velocity of loading actuating cylinder (the piston stick) is
200 mm/s. The diameter of the actuating cylinder piston is D ¼ 52 mm. The
diameter of the actuating cylinder piston stick is d ¼ 26 mm. The effective area
of actuating cylinder is A ¼ 15.92 cm2.
. The maximum flow according to the maximum velocity required above can be
expressed as following:
60
Qmax 1 ¼ 20 £ 15:93 £ ¼ 19:12 L=min
100

Amplifier
F* F*
Control correction
Servo vavle

Force 1.056L
y1 sensor
F1
PL1
L
Amplifier

Figure 6. y2
The principle diagram Rudder
F2
of aileron channel
loading system
 According to system designing requirement, the movement of loading actuating Flight control
cylinder is sinusoidal movement whose frequency is 2 Hz and amplitude is the 0.2 times
as the maximum displacement. That is y1 < ð5 £ 0:2Þ · Sin4pt and y_ ¼ 4p · Cos4pt.
iron bird
The maximum flow according to the maximum velocity required is:
60
Qmax 2 ¼ 4p £ 15:93 £ ¼ 12:01 L=min:
100
Considering various factors such as bearing flexible gap, the maximum flow of loading 123
valve should be much larger than the maximum flow required, according to the experience,
to meet the dynamic performance requirements. Among domestic pressure valve,
the DYSF-3P valve whose maximum flow is 60 L/min is applied in the system which
pressure is 210 kg/cm2. So, the following analysis and calculation is according to the
DYSF-3P valve.
4.1.2 The loading simulation system with DYSF-3P pressure valve.
(1) The transfer function and block diagram of the valve. The simplified transfer
function of the valve is shown in Figure 7. With it, it is explained that the transfer function
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

must be derived from the pressure valve whose output is pressure together with the
loading actuating cylinder. It is shown that loading actuating cylinder parameters Tc and
jc which also derived when Vt is a certain value have been included, for example the
volume of the loading actuating cylinder. Therefore, coefficients in Figure 7 depend on not
only DYSY-3P pressure valve itself, but parameters of the loading actuating cylinder.
For the aileron channel, the value of Vt is:
Vt ¼ 1:1 £ distance £ effective area ¼ 1:1 £ 1:96 £ 16 ¼ 160 cm3
The relevant coefficients are:
 
 ›PL  Vt 2:2 £ 160
TP ¼  ·
 ¼ ¼ 1:26 £ 1022 Sec
›QL 4be 4 £ 7 £ 103
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Tb þ T p 1:03 £ 1022 £ 1:26 £ 1022
Tc ¼ ¼ ¼ 2:05 £ 1023 Sec
Kc 31
1
fc ¼ ¼ 77:6 perimeter=s
2pTc
Tb þ Tp 1:26 £ 1022 þ 1:03 £ 1022
jc ¼ ¼ ¼ 0:18
2Tc K1c 2 £ 2:05 £ 1023 £ 31
Other parameters only relate to the parameters of the valve itself. So, the transfer
function of the pressure valve is shown in Figure 8.
(2) The loading principle block diagram is shown in Figure 9. It should be noted that:


Vt
S þ Kce PL1
4 be
in Equation (12) has been incorporated into other block in the block transformation
Vt
process. That is to say that there is not the separate block S+Kce in Figure 9. Also, the
4be

responses of Vt, Kce in the system are reflected to Tc, jc in Figure 7. The value of QL in the
block diagram is also related to parameters of the valve transfer function.
 IJICC Considering that the stiffness Ks of the force sensor itself is very large and y2 and y1
4,1 are approximately equal, the effect of
1
can be ignored. In addition, the quality of
Ks
the loading actuating cylinder piston, including the stick among the force sensors, is
smaller, so the effect of m1s2 also can be ignored.
124
∂PL/∂QL
QL (S)
∂P
Wb (S) × L
∂x

Figure 7.
The simplified transfer –
I Ke
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

function of the valve Kt PL (S)

Tc2S2 + 2xcTc S + 1

QL (S)
9.11 × 10–5 × (1 + 0.01S)
cm3/sec

Figure 8.
i(s) 7.4 × 102
The transfer function of 5 PL (S)
the pressure valve A (2.05 × 10–3)2 S2 + 2×0.18 × (2.05 × 10–3) S + 1 kg/cm2

y2 – + 1
i
KS

y1
A1S
m1S2
Kw Wf
QL

Kp (1 + Tb s)

K1 + – KC – F*
K (1+ ) Ka Kt A1
– S – TC2 S + 2xC TC S + 1
Figure 9. a1
The loading principle
block diagram a2 Kf
 In Figure 9, the part surrounded by dashed line is the computer control part which is Flight control
operated by discrete signals. In this paper, the effect of discrete signals is not
considered. Make the two weighted feedback branches a1 and a2 emerging into one
iron bird
branch a2 for consideration. The simplified system simulation block diagram is shown
in Figure 10.  K

(3) System integration. In Figure 11, Wf and K 1+ 1 are computer control



S
125
element, where Wf is the feed-forward compensation according to the structure
invariance principle discussed above. That is:
iðAI SÞ £ KP ð1 þ Tb SÞ 1:052 £ 16 S £ 9:11 £ 1025 ð1 þ 0:0103 SÞ
Wf ¼ 2 ¼
K a Kt 0:005

¼ 20:3067 Sð1 þ 0:0103 SÞ



K1
Downloaded by Ohio University At 01:14 23 June 2016 (PT)


And K 1+ whose role is necessary is the PI regulation in the main channel.


S
Otherwise, the system is the difference system. The method to determine the
y2
1.052

16S

–0.3067S (1 + 0.0103S)
3345
4.8
–9.11×10–5 (1 + 0.0103S)

+ – 740
 K 
K 1+ 1 0.005 S2 S 16
 S  + 0.36 +1
– (488)2 (488)
Figure 10.
The simplified system
1 simulation block diagram

dB

–20

20lg 59.2×KS
w = 480
0 w = 604
lgw
0.01

0° –40 Figure 11.

The amplifying factor
–90° KB of a second
40° order open-loop
–180° transfer function
 IJICC parameters K and K1 of the PI regulation is as follow: suppose that the PI regulation
block is a simply amplifying element KB. This KB can be selected first. The principle is
4,1 to ensure that the major loop has sufficient stability margin and the open-loop
amplifying factors in low- and high-frequency areas are larger at the same time.
Consider that this open-loop transfer function is a second order element. So, make
L(v) and 0 decibel line connect at the point that the phase margin is 408. According to
126 this, the amplifying factor KB is determined:
v
2 £ 0:18 £
Make /GðjvÞ ¼ 21408 ¼ 2tg 21 488
v
2
12
488
v
0:36 £
That is tgð21408Þ ¼ 0:8391 ¼ 2 488
v
2
12
488
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

Obtain v ¼ 604 from the equation. Put this value into the logarithmic
amplitude-frequency characteristics and make it equal to 0. It is gotten that:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

v
2 2 v
2
LðvÞ ¼ 20 lg 59:2 KB 2 20 lg 12 þ 0:36 ¼0
488 488
From the equation, obtain KB ¼ 0.0117.
By this way, the phase margin is guaranteed. When v is larger, the phase-frequency
characteristics tend to 2 1808 and at this time, the logarithmic amplitude-frequency
characteristics have been lower than 0 dB early, so the amplitude margin is not have to
be calculated. This is clearly shown in Figure 11. Then, parameters K and K1 can be
determined after KB obtained. The principle is that the stability indexes between them
are the same, and in the low- and middle-frequency area, the magnification can be
increased greatly.
Select the PI regulation parameters which make the open-loop logarithmic
amplitude-frequency characteristics curve. At the same time, when v is larger, the
effect to the phase-frequency characteristics should be as little as possible. Then, the
stability margin is unchanged. So, it is can be gotten that:

 "
2 #1=2
K K1
20 lg þ 20 lg 1 þ ¼0
0:0117 488

Suppose that K1 ¼ 488, so:

1
K ¼ 0:0117 £ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

2 ¼ 0:00836
K1
1þ
488
The final simulation block diagram obtained from the control compensating
parameters determined above is shown in Figure 12.
4.1.3 The loading simulation system with the FF106A-103 flow valve. The maximum
flow of the valve is 63 L/min.
 (1) The transfer function of the valve. According to the sample list: Flight control
Q Kq
iron bird
¼
2
I S 2j
vn þ v n S þ 1

where: 127
Q ¼ The output flow, L/min.
I ¼ The input current, mA.
Kq ¼ The flow gain of valve, which is (63/15) ¼ 4.2 ((L/min)/(mA)) when zero
opening, and 8.4 ((L/min)/(mA)) when pre-opening.
vn ¼ The frequency corresponding to 908 lag. The curve shows that vn < 50 Hz.
j ¼ 0.5 according to the production list recommendation.
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

(2) The loading principle block diagram. Through the derivation and simplify
similar to the process of pressure valve, the loading principle block diagram is shown
in Figure 13.
y2 cm
1.052
–
16S

–0.3067S (1 + 0.0103S)
697
–9.11 × 10–5 (1 + 0.0103S)

+ – 740
+
 488  F* (kgf) Figure 12.
0.00836 1 + 0.005 S2 S 16
–  S  2
+ 0.36 +1 The aileron channel
(488) (488)
loading system designing
simulation block diagram
1 with the 3P pressure valve

y2
i
–
AL S
Wf
Kw

+ Kq – AL
 K 
K 1+ 1 Ka S 2 Vt
+   S+1

2x
S  + Kce Figure 13.
–  wn   wn  4beS
The loading principle
block diagram of the
1 flow valve
 IJICC In Figure 13, Kce is the leakage coefficient combined by servo valve and actuating
4,1 cylinder. For the loading system with the pre-opening servo valve, Kce is mainly the
flow-pressure coefficient:
 
›QL L=min
›PL kg=cm2
128 of the servo valve. So, if ignore the actuating cylinder internal leakage, it is considered that:
 
 › QL 
Kce ¼  :
›P L 
For the pre-opening valve, the load flow QL into the actuating cylinder through the valve is:
sffiffiffi sffiffiffi
1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Q L ¼ Cd w ðD þ XÞ Ps 2 PL 2 Cd w ðD 2 XÞ Ps þ PL
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

r r

where:
Cd ¼ The flow coefficient.
w ¼ The valve core area gradient.
D ¼ The pre-opening value.
PL ¼ The load pressure difference.
rffiffiffi rffiffiffi
1 1
Cd w ðD þ XÞ £ ð21Þ Cd w ðD 2 XÞ £ ð1Þ
› QL r r
¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
›P L 2 P1 2 PL 2 Ps 2 PL
At the neutral position, PL ¼ 0 and X ¼ 0. So:
rffiffiffi rffiffiffiffiffi
1 Ps
2Cd w ·D Cd w ·D
› QL r r Flow at the neutral position
¼ pffiffiffiffiffi ¼2 ¼
›P L Ps Ps oil source force Ps
Suppose that the pre-opening of the flow valve is 15 per cent, which is the same as that
of pressure valve. At the neutral position:
›QL 0:15Q 9:45 L=min L=min
¼ ¼ 2
¼ 0:045
›P L Ps 210 kg=cm kg=cm2
In the block diagram, Kce takes this value above, and the loading actuating cylinder
still takes Vt ¼ 160 cm3 and Al ¼ 160 cm2, and:
Al 16 355:5
¼ ¼
Vt 160 60 S
S þ Kce £ £ S þ 0:045 1 þ
4be 4 £ 70; 000 1; 000 131:2
The principle block diagram of the loading system with flow valve is shown in Figure 14.
 y2 Flight control
1.052
iron bird
0.96S
Wf
697
129
8.4 355.5
 K  Figure 14.
K 1+ 1  S 2
+   +1
 S S
– S  +1 The principle block
314 314 131.2 diagram of the loading
system with flow valve

(3) System integration. According to the structure invariance principle, which is the
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

same as the pressure valve, Wf approximately takes the value ((1.052 £ 0.96 S)/(8.4)).
That is Wf ¼ 0.12 S.
The main loop is series of the second order oscillating element and the inertial
element. The open-loop frequency characteristics curve is shown in Figure 15.
The approach to determine the PI regulation control parameters is the same as that of
loading simulation system with pressure valve. The principle is that the stability margin
of amplitude-frequency characteristics is 10 dB. First, determine KB, which makes
L(v) ¼ 10 dB at the point v ¼ 314. After obtaining KB, select the regulation parameters
K and K1. The principle is that make L(v) and the line of 2 20 dB connect at the point
v ¼ 131.2, which is shown as the dashed line in Figure 15.
The simulation block diagram of the loading system with the flow valve is shown in
Figure 16.

dB

–20

314
0 131.2
lgw
0.01
Amplitude
margin is 10 dB
–60
0°

–90° Figure 15.

The open-loop frequency
characteristics curve
–180°
 IJICC y2
1.052
4,1 –
0.96S

0.12S
697
130
8.4 – 355.5
 131 
+
–2.85 × 10–4 × 1 +  S  2 +  S  +1 S
 S  +1
Figure 16. – 314 314 131.2
The simulation block
diagram of the loading
system with the flow valve 1.052
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

5. Experimental results
(1) The simulation result of the loading system with the pressure valve:
.
The frequency of the rudder is 2 Hz, the vibration amplitude is 0.2 £ maximum
input amplitude of the sinusoidal movement. That is y ¼ sin4ft.
The simulation curve and the frequency response curve are shown in
Figure 17(a) and 17(b), respectively. In Figure 17(a), solid line is for the input,
dotted line for output and dashed line for error. The peak value of the error is 56.
. The input value of Y2 is the maximum motion velocity. That is Y ¼ 4ft. The
simulation curve is shown in Figure 18. In the figure, solid line is for the input,
dotted line for output and dashed line for error. The value of error is 40.
(2) The simulation result of the loading system with the flow valve (require that the
condition is the same as that of the pressure valve):
.
The system response curve with the sinusoidal input and the frequency
response curve are shown in Figure 19(a) and 19(b), respectively.
In Figure 19(a), solid line is for the input, dotted line for output and
dashed line for error. The peak value of the error is 160.
.
The system response curve with ramp input is shown in Figure 20.
In Figure 20, solid line is for the input, dotted line for output and dashed line
for error. The value of the error is 220.
. The system redundant force assessment.
.
The redundant force conditions without and with the compensating element
are shown in Figure 21(a) and 21(b), respectively. Comparing two figures, the
effect of restraining the redundant force with Wf is obviously better.
.
The interfaces of the developed flight control iron bird.

The interfaces of the developed flight control iron bird are shown with Figure 22,
which is developed by Visual Cþ þ .
 Flight control
N iron bird
800

600

131
400

200

–200
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

–400

–600

–800 S
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
(a) The simulation curve with the pressure valve

dB Bode diagram
100

50
Magnitude (dB)

–50

–100

–90
Phase (deg)

–135

–180
w
10–1 100 101 102 103 104
Frequency(rad/sec) Figure 17.
(b) The frequency response curve The simulation curve
and the frequency
response curve
 IJICC
4,1

132

M
7,000
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

6,000

5,000

4,000

3,000

2,000

1,000 40

–1,000 s
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Figure 18.
The maximum motion
input tracking curve
 Flight control
iron bird
M
800

600 133
400

200

–200
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

–400

–600

–800

–1,000 s
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
(a) The system response curve with sinusoidal input

dB Bode diagram
100
Magnitude (dB)

50

–50

–90

–135
Phase (deg)

–180

–225

–270 w
10–1 100 101 102 103
Frequency (rad/sec)
(b) The frequency response curve
Figure 19.
The simulation curve
and the frequency
response curve
 IJICC
4,1

134

M
7,000
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

6,000

5,000

4,000

3,000

2,000
220
1,000

–1,000 s
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Figure 20.
The system response
curve with ramp input
 Flight control
iron bird

N
1,500

1,000 135
500

500

–1,000
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

–1,500

–2,000

–2,500

–3,000 s
0 0.5 1 1.5 2
(a) The redundant force curve without compensating element Wf
N
600

400

200
55
0

–200

–400

–600 s
0 0.5 1 1.5 2
(b) The redundant force curve with compensating element Wf

Figure 21.
The redundant
force curves
 IJICC
4,1

136
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

(a) The interface of (left flap) force control for Channel 1

(b) The interface of (left rudder) force control for Channel 4

Figure 22.
The interfaces
of the developed
flight control iron bird
 6. Conclusions Flight control
This paper proposed a new approach for a type of fighter flight control iron bird, which
is based on passive electronic hydraulic servo loading system. The simulation and
iron bird
physical certification show that the proposed approach can reduce the redundant
forces and improve the system dynamic and force-tracking accuracy. This newly
developed flight control iron bird system has been successfully applied in the flight
control system design of some fighter. 137
References
Cieslak, J., Henry, D. and Zolghadri, A. (2010), “Fault tolerant flight control: from theory to
piloted flight simulator experiments”, IET Control Theory and Applications, Vol. 4,
pp. 1451-61.
Gerkens, T., Fritsch, G., Seyffert, C., Büttner, J., Mugtussidis, I. and Morrish, I. (2004), “The iron
bird for the Fairchild-Dornier 728”, Aerospace Science and Technology, Vol. 8, pp. 231-43.
Karpenko, M. and Sepehri, N. (2006), “Hardware-in-the-loop simulator for research on fault
Downloaded by Ohio University At 01:14 23 June 2016 (PT)

tolerant control of electrohydraulic flight control systems”, 2006 American Control

Conference.
Parker, K. (1999), “Is it a bird? Is it a plane? [pilot training simulator]”, IEE Review, Vol. 45,
pp. 178-9.
Patel, A.V., Patel, V.V. and Deodhare, G. et al. (2006), “Flight control system clearance using
static tests at iron bird”, AIAA Guidance, Navigation, and Control Conference, Vol. 2,
pp. 1308-18.
Tan, C.F., Chen, W. and Boomen, V.D. et al. (2010), “Application of automation for low cost
aircraft cabin simulator”, 2010 International Conference on Control Automation and
Systems (ICCAS), pp. 44-7.
Tian, Y. (2010), “Design and analysis of flight control iron-bird test platform’s electro-hydraulic
servo load system”, PhD thesis, Beihang University, Beijing.

About the authors

Yong Tian received the BS in Hydraulic Control Engineering from Shenyang
University of Technology in 1983, MS of Computational Mathematics from
Dalian University of Technology in 2000, and PhD degrees in fluid dynamics
from Beijing University of Aeronautics and Astronautics in 2010. He is currently
a Research Fellow with the Flight Control Department of Shenyang Aircraft
Design Research Institute. His research interests include flight control
simulation, advanced flight control theory and applications. Yong Tian is the
corresponding author and can be contacted at: [email protected]
Tian Li received the BS in Engineering Mechanics from Tsinghua University in
1963. He is academician of the Chinese Academy of Sciences. He is currently an
Associate General Designer at Shenyang Aircraft Design Research Institute of
AVIC. He is also a Professor and PhD Supervisor at Beijing University of
Aeronautics and Astronautics. His current research interests include flight
aerodynamics, advanced aircraft configuration design and flight simulation.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints