34th INTERNATIONAL CONFERENCE ON
PRODUCTION ENGINEERING
28. - 30. September 2011, Niš, Serbia
University of Niš, Faculty of Mechanical Engineering
COMPARISON OF MODERN ELECTROHYDRAULIC SYSTEMS
Tadej TAŠNER1,2, Darko LOVREC2
HAWE hidravlika d.o.o., Petrovče, Slovenia
2
Production Engineering Institute, Smetanova ulica, Maribor, Slovenia
t.tasner@hawe.si, d.lovrec@uni-mb.si
1
Abstract: One of the main requirements of modern industrial systems is efficiency. Greater efficiency can
be achieved by minimizing required energy, improving reliability and eliminating break-downs needed
for maintenance. Due to increased usage of electrohydraulic systems in industrial applications, such
systems have to be improved. Most of the systems use motor coupled with variable displacement pump.
With cost-effective gear pumps driven by variable speed electric motor, another drive concept is
becoming more popular. There are some questions posed by system designers: “What are the advantages
and disadvantages of both concepts?” and “What would be the benefits when combining variable
displacement pump with variable speed motor?”. All three systems will be described and compared by
simulation results.
Key words: electrohydraulics, variable speed drive, constant pump, variable displacement pump
1.INTRODUCTION
Electrohydraulic systems are mostly used in machines or
production lines which require big forces and power in
order to operate as desired. Big forces result in big energy
consumption which can be optimized by lowering loses in
the electrohydraulic systems. Therefore, improved
efficiency and reduced energy consumption are two of the
main goals in modern electrohydraulic drive systems
design.
Hydraulic energy can be controlled in two main ways:
with throttling principle (by throttling on the directional
valve) or with volumetric principle (by adjusting the
pump displacement volume). The throttling principle has
good dynamic behaviour, but its energy losses are
substantial. The volumetric principle is energy friendlier,
but has worse dynamic response. [1] Due to better
efficiency, the volumetric principle is mostly used.
The hydraulic energy can be controlled by changing the
flow and consequently pressure, which is achieved by
adjusting the pumps’ displacement volume. This can be
done directly by using a variable displacement pump or
indirectly by using a constant displacement pump coupled
with a speed controlled motor. The second, indirect,
principle is becoming more and more popular due to low
prices of constant gear pumps and frequency inverters that
are used to control motor speeds. [2, 3] When thinking in
the way of improving efficiency of the electrohydraulic
system an idea of using variable speed motor and variable
displacement pump arises (Figure 1).
The main goals of this paper are to design a mathematical
model and compare all three mentioned drive concepts by
simulation results. The simulation results will give a
rough estimation of performance differences between the
drive concepts. The concepts will be evaluated by
comparing settling time and overshoot of step response
and ability to track a sine wave and a ramp.
Q
p
DC
Figure 1: Combined drive concept – variable displacement
pump coupled with variable speed DC motor
2.SIMULATION MODEL
The simulation model consists of a DC motor coupled
with a variable displacement vane pump. The vane pump
pumps the hydraulic fluid from a hydraulic tank through
small diameter flexible hoses (equiv. higher resistance),
which represents the load, and back to the tank. The flow
through the hoses causes a pressure drop which is
measured directly after the pump by an ideal pressure
sensor. The pressure after the pump is controlled by all
three different drive concepts to match the reference
pressure as precisely as possible.
To be able to perform the simulation all the components
used in the electrohydraulic circuit are modelled.
Table 2: Electrical-hydraulic analogy
2.1. Variable displacement vane pump
The simulation model uses relatively simplified model of
a variable displacement vane pump. The main parameter
of a hydraulic pump is displacement – volume of fluid
pumped in one revolution of the pump’s shaft. If the
displacement ( VD ) is multiplied by pump’s rotational
Electrical symbol
Electrical
equation
Hydraulic
equation
U  RI
p  RH  Q
speed ( n ) and pump setting ( k p   0,1 ), we get the
U  L
pump’s flow rate ( Q ) (1) [4].
Table 1: Variable displacement vane pump parameters
Displacement
VD = 6 cm3/rev.
Cut-off frequency
f = 50 Hz
 = 0.8
Pump efficiency
Because the flow cannot be changed as soon as the
pump’s setting changes, the pump’s delay is
approximated as a low pass filter with a cut-off frequency
( f ) and can be expressed in Laplace’s frequency domain
(2).

   f
s 1
(2)
torque ( TP ) which is proportional to it (3)
TP 
k p VD  p
2   
p
Q
CH
The flow produced by the pump creates a pressure
difference in the flexible hoses due to resistance between
hydraulic fluid and hose wall. When the pump is creating
pressure difference (  p ) the pump’s shaft is loaded with
(3)
2.2. Hydraulic tubing
The dynamic behaviour of the fluid in the pipeline can be
modelled in more different ways. The most exact model is
based on the Navier-Stokes equations and the law of mass
conservation, which results in a system of partial
differential equations which are too much time consuming
for such simulations.
Such an exact model of hydraulic pipeline is not needed,
because all three drive concepts will be tested on the same
pipeline system. Therefore more appropriate – discrete
model also known as model with concentrated parameters
was chosen. The discrete model is analogous to electrical
circuits used by electrical engineers, where the properties
of a circuit are represented by resistance, capacitance and
inductance. In hydraulics the properties of a pipeline
system are hydraulic resistance RH (pressure drop in a
1
 I dt
C
p
1
CH
 Q dt
Using the electrical symbols the hydraulic pipeline system
can be then represented by n segments as shown in
Figure 2. Each segment represents part of a pipeline with
a length of, l/n where l is total length of the pipeline. The
number of segments also equals the number of possible
pressure measurement points (ex. if the tube is modelled
as one segment, the pressure in the middle of the tube
cannot be calculated.)
1
1
dQ
H
dt
(1)
U 

p  L
dt
Q  k p  n  VD
D
dI
RH
LH
n
n
1 segment
n
Figure 2: Hydraulic tubing represented by electric symbols
Because the pressure will be measured only at the pump
outlet, the model will use a tube represented by only one
segment. Such segment can be described using two
differential equations; first one for hydraulic capacitance
(4) and second one for resistance and inductance (5). If
QC -flow through the capacitor is expressed out of the
first equation and inserted into the second equation, a
second order ordinary differential equation is obtained
(6). Converting that equation to Laplace’s frequency
domain and expressing p/Q yields the transfer function of
the hydraulic tubing represented by one segment.
p
1
CH
Q
c
dt
(4)
p  RH   Q  QC   LH


p  RH   Q  CH
p

dp 
d  Q  QC 
  LH
dt 


d  Q  CH
LH  s  RH
LH  C H  s  RH  C H  s  1
2
(5)
dt
dp 
dt


(6)
dt
(7)
tube due to flow), hydraulic capacitance C H (pressure
Q
drop in a tube due to tube volume increase/decrease) and
hydraulic inductance LH (pressure drop due to fluid
In the simulation a model of MINIMESS® flexible hose
with inner diameter of 2 mm was used. All other hose
parameters crucial for the simulation are described in
Table 3.
acceleration/deceleration).
The
analogy
between
electronics and hydraulics is presented in Table 2. [4]
Table 3: Flexible hose and oil parameters
Hose length
l =10 m
Hose inner diameter [5]
d i =2 mm
Hose outer diameter [5]
d i =5 mm
Pressure loss for mineral
oil with viscosity of
30
mm
RH  1
2
Armature current ( I a ) is controlled by controlling the
armature
E R =1 GPa
Density of oil
  900
3
Eoil  1 GPa
The hydraulic oil plays an important role in the hydraulic
transfer function. Its influence on the pipeline transfer
function is hidden in the hydraulic resistance, capacitance
and inductance, which can be calculated with the
following equations (8, 9). [6]
V0

'
Eoil
4
Eoil
l
 2, 41 106
l
bar (8)
Eoil
di

ER d a  d i
l
bar
LH 
 0, 080
2
l
  di
min 2
4
PI pressure
controller
DC motor with
speed and
current control
flexible
hose
displacement
error
(9)
pressure
setpoint
Figure 3: Control strategy of combined drive concept
In the electrohydraulic system model a brushed DC motor
was used. The DC motor was chosen because its model
and control is relatively simple. The motor parameters
used in the model are shown in Table 1.
Table 4: DC motor parameters
Armature resistance
Armature inductance
Back EMF constant
Torque constant
Rotor inertia
Friction coefficient
Ra = 2.5+2*0.38 
La = 0.3+2*1.5 mH
Ke = 0.0195 Vs/rad
Km = 0.0195 Nm/A
J = 9.87e-6 kg.m2
B = 1.42e-6 Nm s/rad
The dynamic behaviour of a DC motor can be split into
electrical (10) and mechanical (11) part. Both parts can be
modelled by differential equations (10, 11) which are
interconnected by torque (12) and back EMF constant
(13).
dI a
d
dt
variable
pump
pressure
PI pressure
controller
2.3. DC motor
TE  B    J 
is
Pressure is controlled by either DC motor in the indirect
concept (variable motor, constant pump), or variable vane
pump in the direct concept (constant motor, variable
pump) or by both motor and pump in the combined
concept (Figure 3). All controllers used in the simulation
are PI type with internal (integrator) and external limits.
PI controllers are known to be able to eliminate the steady
state error and are more invulnerable to noise than
controllers with the derivative (D) part. [8]
1
U a  I a  Ra  La 
current
3.IMPLEMENTED CONTROL STRATEGY
kg
m
CH 
armature
Therefore the flow through the hydraulic pump coupled
with the motor can be controlled by changing the voltage
applied to the motor.
l/min
Hose’s Young’s modulus
  di2
and
speed of the motor is therefore dependent on the motor
load ( TL ) and the voltage applied to the armature. [7]
s
Bulk modulus of oil
(Ua )
proportional to the electrical torque ( TE ). The rotational
MPa/m
[5]
voltage
dt
 Eb
 TL
(10)
(11)
TE  K m  I
(12)
Eb  K e  
(13)
3.1. DC motor controller
The control structure of the motor has a triple closed loop
or with other words cascaded control. The innermost loop
is current control loop which controls the armature current
by changing the armature voltage; the middle loop
controls rotational speed of the motor by setting a desired
current value to the current controller; and the outermost
controller controls pressure in the pipeline, by changing
the rotational speed.
Integral gain of the current controller is chosen to
compensate the poles of the electrical part and integral
gain of the rotational speed controller to compensate the
poles of the mechanical part.
3.2. Pump controller
The pump is controlled by a single closed loop and
controls pressure by changing the displacement. Integral
gain is chosen in the way that the controller’s zero
eliminates the pump’s pole.
4.SIMULATION RESULTS
Simulations were performed on simulation model
described above using Matlab/Simulink software.
Dynamics of each drive concept were compared by
responses to a combined cycle (sine, ramp and step) and a
step. Simulation results are presented in Figures 4 to 7.
5.CONCLUSION
The concept of variable motor coupled with constant
pump has the slowest dynamics. It is about 5 times slower
than the other two concepts. Other two concepts (variable
motor + variable pump and constant motor + variable
pump) are very hard to separate. On rising steps the
concept with variable motor and pump is a bit slower, but
on falling edges it’s sometimes slightly faster. The
following two systems are very hard to compare, because
the concept with variable pump and variable motor has
two controllers which sometimes interfere with each
other, resulting in poorer dynamics.
Figure 4: Pressure control performance – combined cycle
5.1. Future work
The control strategy for the concept with variable motor
and variable pump should be improved, to minimize
unnecessary changes (ex. lowering the rotational speed
and increasing pump’s displacement which cancel each
other out). Further task is to create a controller which will
maintain the maximum possible efficiency for current
operating conditions and comparison of the simulation
model with a real-world measurement.
6.ACKNOWLEDGEMENT
Figure 5: Rotational speed and pump setting – combined cycle
Operation part financed by the European Union, European
Social Fund. Operation implemented in the framework of
the Operational Programme for Human Resources
Development for the Period 2007-2013, Priority axis 1:
Promoting entrepreneurship and adaptability, Main type
of activity 1.1.: Experts and researchers for competitive
enterprises.
REFERENCES
Figure 6: Pressure control performance - step
Figure 7: Rotational speed and pump setting - step
[1]MAJUMDAR, S. R. (2003) Oil hydraulic systems:
principles and maintenance, McGraw-Hill.
[2]LOVREC, D., ULAGA, S. (2007) Pressure control in
hydraulic systems with variable or constant pumps,
Experimental techniques, Vol.31, pp 33-41.
[3]LOVREC, D., KASTREVC, M., HRIBERNIK, D.
(2005) Primerjava prilagodljivih elektrohidravličnih
napajalnih sistemov na primeru regulacije tlaka, Ventil,
Vol. 11, pp 153-160.
[4]LOVREC, D., HRIBERNIK, D., KIKER, E. (2002)
Model elektrohidravlične regulacije tlaka - osnova za
načrtovanje in optimiranje pogona, Ventil, Vol.3, pp 136146.
[5]MINIMESS®-Technical data on DN2 and DN4
microbore hose.
[6]MURRENHOFF, H. (1998) Grundlagen der
Fluidtechnik - Teil 1: Hydraulik, IFAS-RWTH, Aachen.
[7]KAMARUDIN, N., ROZALI, S. (2008) Simulink
implementation of digital cascade control DC motor
model - a didactic approach, Power and Energy
Conference 2008, pp 1043-1048.
[8]SCHLEICHER, M., BLASINGER, F. (2004) Control
Engineering; A guide for beginners, JUMO.
Corresponding author: TAŠNER Tadej, B.Sc.E.E., HAWE hidravlika d.o.o, Petrovče, Slovenia, t.tasner@hawe.si