A new regulated method in cutter system of shield tunneling machine/Naujas skydines tunelines masinos pjovimo sistemos reguliavimo metodas.
Shi, Hu ; Gong, Guofang ; Liu, Huaiyin 等
1. Introduction
With the rapid urban development and the continuous extension of
transport systems, demands of underground space exploration and tunnel
construction have been considerably growing. Compared with the
conventional excavation methods, the shield machine, in which the cutter
system takes an important part, is famous for high efficiency and safe
operation environment. Pump controlled motors have been used for many
years in cutter system, taking the benefit of high power to mass ratio
and high efficiency [1]. When the shield machine proceeds forward and
encounters different layers of the earth, the displacement of the pump
changes to meet the requirement of the system.
There used to be two control methods for changing the displacement
of the pump when the shield machine is working in different conditions.
First one is to control the displacement of the pump according to
measured load pressure. The second one is to use load pressure directly
to control the displacement of the pump. A new control method is
proposed to change the displacement of the pump from one mode to another
mode as soon as possible when a sudden load changes take place. In order
to find the difference visually between the two methods when the shield
machine enters a different layer of earth, the models of these two
hydraulic systems are built in AMESim software. Using this kind of
programs it is possible to realize complex system models, starting from
simple sub-model already included in the program library in this system
[2]. Every component has to be constituted according to the parameters
of working condition, such as proportional relief valve, stroking
mechanism, constant output regulation mechanism. In order to get
realistic values to the parameters of different components, the related
component manuals are used. In order to validate the models, it is
necessary to compare simulation results as soon as possible to
experimental results. In this paper also the needed test rig is
described.
2. Description of the hydraulic system of cutter system
The diagram of the typical hydraulic system of a cutter of a shield
machine with an open loop control is shown in Fig. 1. The direction
control valve is used to change the rotation direction of the cutter,
the shuttle valve is used to select the higher pressure of the two sides
of the motors. This pressure and the rotation speed of the motors are
the input to the controller, which controls the displacement of the
pump.
[FIGURE 1 OMITTED]
When the shield machine works in a stable condition, the flow rate
of the system is determined. In some applications the controller of the
pump is so called constant power controller. The output flow rate of the
pump is Q = P/p, where Q is the flow rate, P is power, p is pressure of
the system. If the pressure doesn't exceed the critical pressure,
the flow rate is constant. But if the load changes and the pressure is
higher than the critical pressure, the flow rate of the system will
change according to this equation. Fig. 2 presents the block diagram of
the conventional control system of the cutter. The load pressure is
measured by the pressure sensor. When the shield machine faces sudden
load changes, the controller sets a new set value to the displacement of
the pump. If the measured load pressure is higher than the critical
pressure, the controller reduces the set value of the pump. If the
measured load pressure is less than the critical pressure the set value
of the pump stays constant. Considering the lag of the proportional
electro-magnet, pressure sensor, converting time between the signals, it
is time-consuming and can't meet the requirement of the system.
When the shield machine faces sudden load changes, it may result in
remarkable overloads to the system. So it is necessary to propose a new
method to simplify the system and make the control system more cost
effective, but still keep the system capable to response good enough
different working conditions. Fig. 3 presents the principle of the new
simplified control method. The proposed control method is pure
hydraulics without any electronics.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
3. Simulation software description
AMESim is the acronym for advanced modeling environment
simulations. This software is used to simulate engineering system [3].
The whole system is constituted of different submodels which can be
dragged directly from the software. Different submodels can make a
component through a regular arrangement and connection. When a simple
system is completed, several parameters must be defined, including
mathematical description of the interaction between components and their
characteristics. Then the simulation process can be started and it
produces graphical and numerical representation of the behavior of the
system. AMESim has different libraries, in order to get the wanted
system model, in this case the control mechanical and hydraulic library
are used. Sometimes the submodel can't meet all requirements, then
the model have to be completed by using the hydraulic component design
(HCD) library step by step, starting from the simpler components.
Important part is also to find correct values for the parameters of the
model. Only after verifying the results, it is possible to complete the
system model.
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
4. Models of components
In this chapter the models of some components of the cutter system
are presented.
4.1. Modelling of the pump
In the simulation, the two control methods of the pump
displacement, hydraulics and electronics (Fig. 4), is compared. The
parameters of the pump refer to the manual from HAWE corporation The
range of the control pressure is 0 - 32 MPa.
The electric and hydraulic controlled pumps have the same
parameters and structure. Based on AMESim presentation method the
hydraulic control system of the pump is presented in Fig. 5. a and
electric control system in Fig. 5. b. The Table 1 illustrates the
parameters of the pump.
4.2. Modelling of the sensor
Considering the pipe in the shield machine is long and big, it
can't be ignorant about the delay of transmission and loss when the
oil transmits from one point to another point. The connection pipe and
the chamber make up a pressure transmission system. On the assumption
that [p.sub.i] is the wanted pressure which lies besides a point A,
[p.sub.v] is the pressure that directly acts on the membrane of the
sensor. Now the membrane of the sensor is simplified to a piston with
centralized mass. So the model of the sensor can be treated like what
Fig. 6 illustrates.
In this model, input is [p.sub.i] and output [p.sub.v]. [p.sub.i]
and [p.sub.v] conform to the following equation [4]
[4[rho]lV/[pi][[beta].sub.c][d.sup.2]]
[[d.sup.2][p.sub.v]/[dt.sup.2]] + [128
[micro]lV/[pi][[beta].sub.c][d.sup.4]] [[dp.sub.v]/dt] + [p.sub.v] =
[p.sub.i] (1)
[FIGURE 6 OMITTED]
Laplace transforms
[P.sub.V]/[P.sub.i] = [[1/4 [pi]lV/[pi][[beta].sub.c][d.sup.2]]
[s.sup.2] + [128 ulV/[pi][[beta].sub.c][d.sup.4]] s + 1 (2)
where d is inner diameter of the pipe (m); [[beta].sub.c] is bulk
modulus of the liquid (N/[m.sup.2]); V is volume of chamber ([m.sup.3]);
[micro] is absolute viscosity of the liquid (Ns/[m.sup.2]); l is pipe
length (m); [rho] is density of the liquid (kg/[m.sup.3]); t is time
(s).
4.2.1. Modelling of the proportional magnet
In the hydraulic system, it can't be ignored of the influence
of the electro-magnet, because it is an important part of pump and
proportional valve. When the signal forces the electro-magnet to act,
the dynamic force equilibrium is achieved.
Dynamic characteristic of coil current [5]
[u.sub.s](t) = [L.sub.d] [di(t)/dt] + i(t)R(s) + [K.sub.v]
[dy(t)/dt] (3)
where [K.sub.v] is coefficient of the reverse EMF; [R.sub.s] is
resistance of the coil and the proportional magnifier; [L.sub.d] is
dynamic inductance of the coil.
4.2.2. Dynamic characteristic of output
The equation of the dynamic force is
[F.sub.d](t) = [K.sub.Fi] i(t - [[tau].sub.d]) - [f.sub.M]
sgn[di(t)/dt] + [F.sub.r] (4)
where [K.sub.Fi] is gain of the current; [[tau].sub.d] is time
delay; [f.sub.M] is delayed force of the electromagnetism; [F.sub.r] is
Columbic friction.
4.2.3. Dynamic characteristic of displacement
According to the force balance of the system
[F.sub.d](t) = m [[d.sup.2]y(t)/dt] + c [dy(t)/dt] + [K.sub.s]y(t)
(5)
where m is equivalent mass related to proportional magnet; c is
damping ratio related to proportional magnet; [K.sub.s] is spring rate
related to proportional magnet.
In the system of the electro-magnet, Fig. 7 illustrates the
transfer function and the control principal of the system.
[FIGURE 7 OMITTED]
4.2.4. Modelling of the 4-way 3-position directional valve
This valve, which is used to change the flow direction of the
liquid to change the direction of the cutter, is a 4-way 3- position
electric controlled valve. It is a digital valve, normally closed,
normally open or normally middle. According to the parameters of the
valve from the manual, the valve model is shown in Fig. 8.
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
Table 2 illustrates the parameters of the valve. In this
simulation, the supply flow rate is 120 L/min, the load is 50 bar.
Control force and pressure in the chamber for the valve is shown in
Figs. 9, 10.
[FIGURE 10 OMITTED]
5. AMESim model of the two control methods
This paragraph describes the models of system with the two control
methods. Fig. 11 illustrates the control system with electric control
and Fig. 12 with hydraulic control, in which a hydraulic control valve
replaces the electric control.
[FIGURE 11 OMITTED]
Fig. 13 shows the torque of the cutter system. First it is 6000 Nm
and then at 3 s it changes step-wise to 27000 Nm. It can be seen from
Fig. 14 that the pressure of the system with electric control is 75 bar
at first and then jumps to 250 bar with a little vibration, while the
pressure of the system with the hydraulic control jumps from 75 bar to
250 bar smoothly. The following Fig. 15 illustrates the result of the
response of the flow rate when the pressure suddenly changes. The
response time of the hydraulic control method is less than 0.1 s, while
the other is about 0.3 s. It can be seen that the hydraulic control
method responds faster than the other one and the change curve is
smoother without too much vibration.
[FIGURE 12 OMITTED]
[FIGURE 13 OMITTED]
[FIGURE 14 OMITTED]
[FIGURE 15 OMITTED]
When the oil transmits through the pipe connected between the
outlet of the shuttle valve and the inlet of stroking mechanism, the
response of the system may have remarkable delay. Fig. 16 illustrates
control principal of the hydraulic controlled system. The control oil
transmits through the long line and reaches the valve. Then the oil
transmits through another pipe to control the stroking mechanism to
change the displacement. In order to analyze the effect of the long pipe
in delaying the transmission of the signal, a long pipe model which is
referring to the pipe used in the shield machine is built in AMESim. In
this model, the compressibility and friction effects of the fluid are
taking into account. The pipe is characterized by the diameter of 20 mm
and length which ranges from 0 to 40 m. In the simulation condition, the
batch run is operated by cutting the length into 5 parts in AMESim
software. The simulation results are illustrated in the following
diagrams. Fig. 17 illustrates the torque of the system which the system
encounters in the simulation condition. At 3s, the torque jumps from
3000 Nm to 27000 Nm, which simulates different working conditions.
[FIGURE 16 OMITTED]
[FIGURE 17 OMITTED]
Fig. 18 illustrates different response time of the system by
inspecting the change of the pressure when the length of the pipe
between the shuttle valve and the stroking mechanism is different. It
can be seen from Fig. 18 that when the length of pipe ranges from 0 to
40 meters, the response time ranges from 0.1 s to 0.7 s. The longer the
pipe is, the slower the response of the system is. It can be seen from
Fig. 19 that when the length is different, the dynamic response of the
flow rate of the system is different. When the length is small, the flow
rate of the system has a little vibration. When the length becomes
bigger, the response of the flow rate is smoother. The longer the pipe
is, the bigger the damping ratio of the system is, so it will affect the
stability and the response of the system. Compared with the proceeding
electric controlled method, when the length of the pipe is small, the
hydraulic control method is advantageous in the aspect of the response
time. When the length of the pipe is longer than about 20 meters, the
response of the hydraulic control method may be slower than the other
one.
[FIGURE 18 OMITTED]
[FIGURE 19 OMITTED]
6. Experimental test rig for the characterize of the system
To validate the simulation models, it is important to compare
simulations with experimental data provided by appropriate test rigs.
The test rig has been built, which include a pump with different control
methods, proportional relief valve, pressure sensor and flow meter, etc.
In this system, a motor, whose ports are connected with the proportional
relief valve, is used to simulate the load the shield machine may
encounter. By regulating the pressure of the proportional relief valve,
which ranges from 0 to 315 bar, we can simulate different working
torque. Fig. 20, a illustrates the valve group of the system and Fig.
20, b illustrates the motor and the pump which connect through an
inertia wheel.
[FIGURE 20 OMITTED]
7. Conclusions
1. Based on the working principal of the cutter system of the
shield machine, a new control method for the change of the displacement
of the pump is proposed. When the pump works in the constant output
regulation mode, the simulation result demonstrates that it is possible
to control the displacement by the hydraulic control and is advantageous
and swift to fit the flow rate with the needs of the system.
2. The length of the pipe between the shuttle and the pump is also
taken into account in affecting the response time of the system. The
simulation results show that when the pipe is shorter than a definite
value, the response time of hydraulic control method is faster than the
other one. Also, the new method will cut down the cost of the system,
because it will reduce a lot of conversion devices which change the
hydraulic signal to electric signal. Through the simulation, it proves
feasible to use the method to change the flow rate of the system in
order to avoid the overloading of the motors and improve the response of
the pump to act. In the future, the control effect will be testified and
compared with the conventional control methods in the experiment.
Received March 11, 2011
Accepted June 13, 2012
References
[1.] Xing Tong; Li Xilian; Gong Guofang; Yang Huayong. 1987.
Contribution to modeling and simulation of an electro-hydraulic control
system in shield machine, IEEE Translated J. Fluid Power, China, 2:
740-741, [Proceedings of the Fifth International Symposium on Fluid
Power Transmission and Control].
[2.] Bdlforte, G.; Carello, M.; Ferraresi, C.; Martinelli, M.;
Pescarmona, F.; Viktorov, V. 2008. Modeling and identification
methodology of components for train braking system, IEEE Translated J.
Fluid Power in motion. Dresden, 5: 393-405, [Digest 6th International
Fluid Power Conference. Fluid Power in motion Dresden, p. 301, 2008].
[3.] AMESim User Manual, Version 4.2, 2004.
[4.] Huang Changyi; Yan Puqiang 2005. Testing Basis of Mechanical
Engineering, China Machine Press, 207-209.
[5.] Wu Genmao; Qiu Xiumin; Wang Qingfeng; Wei Jianhua; Kong
Xiaowu; Fu Xin, 2006. Electrohydraulic Proportional Technique in Theory
and Application, Zhejiang University Press, 92-93.
Hu Shi, Corresponding author, Zhejiang University Hangzhou,
Zhejiang, 310027, China, E-mail: tigershi@zju.edu.cn
Guofang Gong, Zhejiang University Hangzhou, Zhejiang, 310027,
China, E-mail: gfgong@zju.edu.cn
Huaiyin Liu, Zhejiang University Hangzhou, Zhejiang, 310027, China,
E-mail: lhykm1985@163.com
Lintao Wang, Zhejiang University Hangzhou, Zhejiang, 310027, China,
E-mail: jixiezhixing2005@163.com
cross ref http://dx.doi.org/10.5755/j01.mech.18A2321
Table 1
Parameters of the pump
Names of parameters Value
spool diameter, mm 10
rod diameter, mm 5
mass, kg 0.02
higher displacement limit, mm 0.01
left spring stiffness, N/mm 600
spring force at zero displacement, N 30
coefficient of viscous friction, N/(m/s) 1000
right spring rate, N/m 300000
piston diameter, mm 10
rod diameter, mm 3
left equivalent orifice diameter, mm 1.3
right equivalent orifice diameter, mm 0.5
Table 2
Parameters of the directional valve
Name Value
spring force at zero 20
displacement, N
spring stiffness, N/mm 10
volume of chamber, [cm.sup.3] 10
piston diameter, mm 10
rod diameter, mm 5
mass, kg 0.03
coefficient of viscous 1000
friction, N/(m/s)