Fluid Power System Dynamics

W. Durfee and Z. Sun

Department of Mechanical Engineering
University of Minnesota
April 22, 2009
A National Science Foundation
Engineering Research Center
c 2009
Forward
This minibook was created because system dynamics courses in the standard
mechanical engineering curriculum do not cover uid power, even though
uid power is essential to mechanical engineering and students entering the
work force are likely to encounter uid power systems in their job. Most sys-
tem dynamics textbooks have a chapter or part of a chapter on uid power
but typically the chapter is thin and does not cover practical uid power as
is used in industry today. For example, many textbooks conne their discus-
sion of uid power to liquid tank systems and never even mention hydraulic
cylinders, the workhorse of todays practical uid power.
The material is intended for use in an introductory system dynamics course
that would teach analysis of mechanical translational, mechanical rotary and
electrical system using differential equations, transfer functions and time and
frequency response. The material should be introduced toward the end of the
course after the other domains and most of the analysis methods have been
covered. It replaces or supplements any coverage of uid power in the course
textbook. The instruction can pick and choose which sections will be covered
in class or read by the student. At the University of Minnesota, the material
is used in course ME 3281, System Dynamics and Control.
This project is a result of the Center for Compact and Efcient Fluid Power
(CCEFP) (www.ccefp.org), a National Science Foundation Engineering Re-
search Center founded in 2006. The center conducts basic and applied re-
search in uid power with three thrust areas: efciency, compactness, and
usability. The center has over 50 industrial afliates and its research is ul-
timately intended to be used in next generation uid power products. To
ensure the material in the minibook is current and relevant, it was reviewed
by industry representatives and academics afliated with the center.
For questions about the minibook or about the Center for Compact and Ef-
cient Fluid Power please contact William Durfee, University of Minnesota,
[email protected]
Copyright and Distribution
Copyright is retained by the authors. Others may freely copy and distribute
this material for educational purposes, but may not sell the material for prot.
Contents
1. Introduction 1
1.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2. Fluid Power Examples . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3. Analyzing Fluid Power Systems . . . . . . . . . . . . . . . . . . 6
2. Basic Principles of Fluid Power 9
2.1. Pressure and Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2. Power and Efciency . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3. Hydraulic Fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4. Fluid Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4.1. Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4.2. Bulk Modulus . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4.3. Pascals Law . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4.4. High Forces . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5. Conduit Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5.1. Pressure Losses in Conduits . . . . . . . . . . . . . . . . . 20
2.6. Bends and Fittings . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.7. Orice Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3. Fluid Power Components 28
3.1. Cylinders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.2. Pumps and Motors . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.3. Control Valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.3.1. Dynamic Models for Valves . . . . . . . . . . . . . . . . . 33
3.3.2. Valve Symbols . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.4. Accumulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.5. Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.6. Reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.7. Hoses and Fittings . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4. Hydraulic Circuit Analysis 40
4.1. Fluid Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2. Fluid Capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.3. Fluid Inertance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.4. Connection Laws and States . . . . . . . . . . . . . . . . . . . . . 44
4.4.1. Connections . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.4.2. State Variables . . . . . . . . . . . . . . . . . . . . . . . . . 44
iii
iv Contents
4.5. Example Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5. Bibliography 47
A. Fluid Power Symbols 48
1. Introduction
1.1. Overview
Fluid power is the transmission of forces and motions using a conned, pres-
surized uid. In hydraulic uid power systems the uid is oil, or less com-
monly water, while in pneumatic uid power systems the uid is air.
Fluid power is ideal for high speed, high force, high power applications.
Compared to all other actuation technologies, including electic motors, uid
power is unsurpassed for force and power density and is capable of gen-
erating extremely high forces with relatively lightweight cylinder actuators.
Fluid power systems have a higher bandwidth than electric motors and can
be used in applications that require fast starts, stops and reversals, or that
require high frequency oscillations. Because oil has a high bulk modulus, hy-
draulic systems can be nely controlled for precision motion applications.
1
Another major advantage of uid power is compactness and exibility. Fluid
power cylinders are relatively small and light for their weight and exible
hoses allows power to be snaked around corners, over joints and through
tubes leading to compact packaging without sacricing high force and high
power. A good example of this compact packaging are Jaws of Life rescue
tools for ripping open automobile bodies to extract those trapped within.
Fluid power is not all good news. Hydraulic systems can leak oil at connec-
tions and seals. Hydraulic power is not as easy to generate as electric power
and requires a heavy, noisy pump. Hydraulic uids can cavitate and retain
air resulting in spongy performance and loss of precision. Hydraulic and
pneumatic systems become contaminated with particles and require careful
ltering. The physics of uid power is more complex than that of electric mo-
tors which makes modeling and control more challenging. University and
industry researchers are working hard not only to overcome these challenges
but also to open uid power to new applications, for example tiny robots and
wearable power-assist tools.
1.2. Fluid Power Examples
Fluid power is pervasive, from the gas spring that holds you up in the ofce
chair you are sitting on, to the air drill used by dentists, to the brakes in your
1
While conventional thinking was that pneumatics were not useful for precision control,
recent advances in pneumatic components and pneumatic control theory has opened up new
opportunities for pneumatics in precision control.
1
2 1. Introduction
Figure 1.1.: Caterpillar 797B mining truck. Source: Caterpillar
car, to practically every large agriculture, construction and mining machine
including harvestors, drills and excavators.
The Caterpillar 797B mining truck is the largest truck in the world at 3550
hp (Fig. 1.1). It carries 400 tons at 40 mph, uses 900 g of diesel per 12 hr shift,
costs about $6M and has tires that are about $60,000 each. It is used in large
mining operations such as the Hull-Rust-Mahoning Open Pit Iron Mine, the
worlds largest open pit iron mine, located in Hibbing MN and the Muskeg
River Mine in Alberta Canada.
2
The 797B uses uid power for many of its
internal actuation systems, including lifting the fully loaded bed.
Shultz Steel, an aerospace company in South Gate CA, has a 40,000-ton
forging press that weighs over 5.2 million pounds (Fig. 1.2). It is the largest
press in the world and is powered by hydraulics operating at 6,600 psi requir-
ing 24 700 hp pumps.
The Multi-Axial Subassemblage Testing (MAST) Laboratory is located at
the Universtiy of Minnesota and is used to conduct three-dimensional, quasi-
static testing of large scale civil engineering structures, including buildings,
to determine behavior during earthquakes (Fig. 1.3). The MAST system, con-
structed by MTS Systems, has eight hydraulic acuators that can each push or
pull with a force of 3910 kN.
The Caterpillar 345C L excavator is used in the construction industry for
large digging and lifting operations and has a 345 hp engine (Fig. 1.4). The
345C L operates at a hydraulic pressure of 5,511 psi to generate a bucket dig-
2
New Tech to Tap North Americas Vast Oil Reserves, Popular Mechanics, March 2007.
1.2. Fluid Power Examples 3
Figure 1.2.: 40,000 ton forging press. Source: Shultz Steel.
Figure 1.3.: MAST Laboratory for earthquake simulation. Source: MAST Lab.
4 1. Introduction
Figure 1.4.: Caterpillar 345C L excavator. Source: Caterpillar.
ging force of 60,200 lbs and a lift force of up to 47,350 lbs.
A feller buncher is a large forestry machine that cuts trees in place (Fig.
1.5).
Some of the fastest roller coasters in the world get their initial launch from
hydraulics and pneumatics. (Fig. 1.6). Hydraulic launch assist systems pump
hydraulic uid into a bank of accumulators storing energy as a compressed
gas. At launch, the energy is suddenly released into a hydraulic motor whose
output shaft drives a cable drum with the cable rapidly bringing the train
from rest to very high velocities. The Kingda Ka at Six Flags Great Adventure
uses this launch and reaches 128 mph in 3.5 s. The Hypersonic SLC at Kings
Dominion ups the ante with a compressed air launch system that accelerates
riders to 81 mph in 1.8 s.
Most automatic transmissions have hydraulically actuated clutches and
bands to control the gear ratios. Fluid is routed through internal passage-
ways in the transmission case rather than through hoses (Fig. 1.7)
The dental drill is used to remove small volumes of decayed tooth prior to
inserting a lling (Fig. 1.8). Modern drills rotate at up to 500,000 rpm using
an air turbine and use a burr bit for cutting. The handpiece can cost up to
$800. Pnuematic drills are used because they are smaller, lighter and faster
than electric motor drills. The compressor is located away from the drill and
pressurized air is piped to the actuator.
Hydraulic microdrives are used during surgery to position recording elec-
trodes in the brain with micron accuracy (Fig. 1.9). Master and slave cylinders
1.2. Fluid Power Examples 5
Figure 1.5.: Feller buncher. Souce: Wikipedia image.
Figure 1.6.: Hypersonic XLC roller coaster with hydraulic lanuch assist. Source:
Wikipedia image.
6 1. Introduction
Figure 1.7.: Mercedes-Benz automatic transmission model.
have a 1:1 ratio and are separated by three to four feet of uid lled cable.
1.3. Analyzing Fluid Power Systems
Analyzing the systemdynamics of uid power means using differential equa-
tions and simulations to examine the pressures and ows in components of
a uid power circuit, and the forces and motions of the mechanisms driven
by the uid power. For example, in an excavator, the engineer would be in-
terested in determining the diameter and stroke length of the cylinder that is
required to drive the excavator bucket and how the force and velocity of the
bucket changes with time as the valves to the cylinder are actuated. Because
uid power systems change with time and because uid power systems have
energy storage elements, a dynamic system analysis approach must be taken
which means the use of linear and nonlinear differential equations, linear and
nonlinear simulations, time responses, transfer functions and frequency anal-
ysis.
Fluid power is one domain within the eld of system dynamics, just as
mechanical translational, mechanical rotational and electronic networks are
system dynamic domains. Fluid power systems can be analyzed with the
Figure 1.8.: A dental handpiece. Source: Wikipedia image.
1.3. Analyzing Fluid Power Systems 7
Figure 1.9.: Hydraulic microdrive for neural recording electrode placement. Source:
Stoelting Co.
same mathematical tools used to describe spring-mass-damper or inductor-
capacitor-resistor systems. Like the other domains, uid power has funda-
mental power variables and system elements connected in networks. Un-
like other domains many uid power elements are nonlinear which makes
closed-form analysis somewhat more challenging, but not difcult to simu-
late. Many concepts from transfer functions and basic closed loop control
systems are used to analyze uid power circuits, for example the response of
a servovalve used for precision control of hydraulic pistons.
3
Like all system dynamics domains, uid power is characterized by two
power variables that when multiplied form power, and ideal lumped ele-
ments including two energy storing elements, one energy dissipating ele-
ment, a ow source element and a pressure source element. Table 1.1 shows
the analogies between uid power elements and elements in other domains.
Lumping uid power systems into elements is useful when analyzing com-
plex circuits.
3
See Transfer Functions for Mood Servovalves available on-line in the technical docu-
ments section of the Moog company web site.
8 1. Introduction
Table 1.1.: Element analogies in several domains.
Domain Power Power Storage Storage Dissipative
Variable 1 Variable 2 Element 1 Element 2 Element
Translational Force, F Velocity, V Mass, M Spring, K Damper, B
Rotational Torque, T Velocity, Inertia, J Spring, K Damper, B
Electrical Current, I Voltage, V Inductor, L Capacitor, C Resistor, R
Fluid Power Flow, Q Pressure, P Inertance, I
f
Capacitor, C
f
Resistance, R
f
2. Basic Principles of Fluid Power
2.1. Pressure and Flow
Fluid power is characterized by two main variables, pressure and ow, whose
product is power. Pressure P is force per unit area and ow Q is volume
per time. Because pneumatics uses compressible gas as the uid, mass ow
rate Q
m
is used for the ow variable when analyzing pneumatic systems.
For hydraulics, the uid is generally treated as incompressible, which means
ordinary volume ow Q can be used.
Pressure is reported several common units that include pounds per square
inch (common engineering unit in the U.S.), pascal (one newton per square
meter, the SI unit), megapascal and bar. Table 2.1 shows how pressure units
are related. For engineering, it is best to do cacluations and simulations in SI
units, but to report in SI and the conventional engineering unit.
Pressure is an across type variable, which means that it is always measured
with respect to a reference just like voltage in an electrical system. As shown
in Figure 2.1, one can talk about the pressure across a uid power element
such as a pump or a valve, which is the pressure differential from one side
to the other, but when describing the pressure at a point, for example the
pressure of uid at one point in a hose, it is always with respect to a reference
pressure. Reporting absolute pressure means that the pressure is measured
with respect to a perfect vacuum. It is more common to measure and report
gage pressure, the pressure relative to ambiant atmospheric pressure (0.10132
mPA, 14.7 psi at sea level). The distinction is critical when analyzing the
dynamics of pneumatic systems because the ideal gas law that models the
behavior of air is based on absolute pressure.
Pressure is measured with a mechanical dial type pressure gauge or with an
electronic pressure transducer that outputs a voltage proportional to pressure
Table 2.1.: Conversions between pressure units
pascal megapascal bar lbs-sq-in
(Pa) (Mpa) (bar) (psi)
1 Pa 1 10
6
10
5
145.04 10
6
1 Mpa 10
6
1 10 145
1 bar 10
5
0.1 1 14.5
1 psi 6895 6.895 10
3
0.06895 1
9
10 2. Basic Principles of Fluid Power
Figure 2.1.: Pressure is measured with respect to a reference.
(Fig. 2.2). Almost all pressure transducers report gage pressure because they
expose their reference surface to atmosphere.
Volume ow rate is reported in gallons per minute, liters per minute and
cubic meters per second (SI unit). Table 2.2 shows how ow rate units are
related.
Flow is a through type variable, which means it is volume of uid owing
through an imaginary plane at one location. Like current in an electrical sys-
tem, there is no reference point. Flow is measured with a ow meter placed
in-line with the uid circuit. One common type of ow meter contains a
turbine, vane or paddle wheel that spins with the ow. Another type has a
narrowed passage or an orice and ow is estimated by measuring the dif-
ferential pressure across the obstruction. A Pitot tube estimates velocity by
measuring the stagnation pressure. Figure 2.3 shows some common types of
ow meters. Because ow meters restrict the ow, they are used sparingly in
systems where small pressure drops matter.
2.2. Power and Efciency
The power available at any one point in a uid power system is the pressure
times the ow at that point
power = P Q (2.1)
Figure 2.2.: Dial pressure gauge and electronic pressure transducer.
2.2. Power and Efciency 11
Table 2.2.: Conversions between volume ow rate units
gallon/minute liter/minute cubicmeter/second
(gpm) (lpm) (m
3
/s)
1 gpm 1 3.785 6.31 10
5
1 lpm 0.264 1 1.67 10
5
1 m
3
/s 1.585 10
4
6 10
4
1
For the power available in a conduit, the pressure in Equation 2.1 is the
pressure relative to the pressure in the system reservoir, which is typically at
atmospheric pressure.
Example 2.2.1. The hose supplying the cylinder operating the bucket of a
large excavator has uid at 1000 psi owing at 5 gpm. What is the available
power in the line?
Solution: For most engineering examples, the reported pressure is a gage
pressure, which means the hose is operating at 1000 psi above atmospheric
pressure, the pressure of the resevoir. To calculate the power
1000 psi 6895 Pa/psi = 6.910
6
Pa
5 gpm 6.3110
5
m
3
/s/gpm = 31.610
5
m
3
/s
Figure 2.3.: Types of owmeters. Top row: turbine, digital paddle, variable area. Bot-
tom row is a dual-rotar turbine owmeter with a cutaway.
12 2. Basic Principles of Fluid Power
Power = 6.9 10
6
31.6 10
5
= 2180 watts = 2.9 horsepower
Components such as cylinders, motors and pumps have input and output
powers, which can be used to calculate the efciency of the component. For
example, pressured uid ows into a cylinder and the cylinder extends. The
input power is the pressure of the uid times its ow rate while the output
power is the compression force in the cylinder rod times the rod extension
velocity. Dividing output power by input power yields the efciency of the
component. The same can be done for components such as an orice. The
efciency of the orice is the output pressure divided by the input pressure
because the ow rate is the same on either side of the orice.
2.3. Hydraulic Fluids
The main purpose of the uid in a uid power system is to transmit power.
There are other, practical considerations that dictate the specic uids used
in real hydraulic systems. The uids must cool the system by dissipation of
heat in a radiator or reservoir, must help with sealing to prevent leaks, must
lubricate sliding and rotating surfaces such as those in motors and cylinders,
must not corrode components and must have a long life without chemical
breakdown.
The earliest hydraulic systems used water for the uid. While water is safe
for humans and environment, cheap and readily available, it has signicant
disadvantages for hydraulic applications. Water provides almost no lubri-
cation, has low viscosity and leaks by seals, easily cavitates when subjected
to negative pressures, has a narrow temperature range between freezing and
boiling (0 to 100

C), is corrosive to the steels used extensively in hydraulic

components and is a friendly environment for bacterial and algae growth,
which is why swimming pools are cholorinated.
Modern hydraulic system use petroleum based oils, with additives to in-
hibit foaming and corrosion. Petroleum oils are inexpensive, provide good
lubricity and, with additives, have long life. The brake and automatic trans-
mission uids in your car are examples.
2.4. Fluid Behavior
2.4.1. Viscosity
All uids, including oil and air have fundamental properties and followbasic
uid mechanics laws. The viscosity of a uid is its resistance to ow. Some
uids, like water, are thin and have low viscosity while others like honey are
2.4. Fluid Behavior 13
Figure 2.4.: Viscosity related hydraulic system losses.
thick and have high viscosity. The uids for hydraulic systems are a com-
promise. If the viscosity is too low, uid will leak by internal seals causing a
volumetric loss of efciency. If the viscosity is too high, the uid is difcult
to push through hoses, ttings and valves causing a loss of mechanical ef-
ciency. Figure 2.4 shows this tradeoff and indicates that a medium viscosity
uid is best for hydraulic applications.
The dynamic viscosity (also known as the absolute viscosity) is the shear-
ing resistance of the uid and is measured by placing the uid between two
plates and shearing one plate with respect to the other. The symbol for dy-
namic viscosity is the Greek letter mu (). The SI unit for dynamic viscosity
is the pascal-second (Pa-s), but the more common unit is the centipoise (cP),
with 1 cP = 0.001 Pa-s. The dynamic viscosity of water at 20

C is 1.00 cP
It is easier to measure and more common to report the kinematic viscosity of
a uid, the ratio of the viscous forces to inertial forces. The symbol for kine-
matic viscosity is the Greek letter nu (). Kinematic viscosity can be measured
by the time it takes a volume of oil to ow through a capillary. The SI unit for
kinematic viscosity is m
2
/s but the more common unit is the centistoke (cSt)
which is 1 mm
2
/s. The conversion is 1m
2
/s = 10
6
cSt = 10
4
stokes
If is the uid density, the kinematic and dynamic viscosity are related by
=

(2.2)
The kinematic viscosity of water over a wide range of temperature is 1 cSt
while common hydraulic oils at 40

C are in the range of 20-70 cSt. Some-

times hydraulic oil kinematic viscosity is expressed in Saybolt Universal Sec-
onds (SUS), which comes fromthe oil properties being measured on a Saybolt
viscometer.
14 2. Basic Principles of Fluid Power
Figure 2.5.: The viscosity of hydraulic oils varies with temperature.
Viscosity changes with temperature; as uid warms up it ows more easily.
One reason your car is hard to start on a very cold morning is that the engine
oil thickened overnight in the cold. The viscosity index VI expresses how
much viscosity changes with temperature. Fluids with a high VI are desirable
because they experience less change in viscosity with temperature. Figure 2.5
shows how viscostiy changes for typical hydraulic uids.
2.4.2. Bulk Modulus
In many engineering applications, liquids are assumed to be completely in-
compressible even though all materials can be compressed to some degree.
In some hydraulic applications, the tiny compressibility of oil turns out to be
important because the pressures are high, up to 5,000 psi. The bulk modulus
of the uid is the property that indicates the springiness of the uid and is
dened as the pressure needed to cause a given decrease in voume. A typical
oil will decrease about 0.5% in volume for every 1000 psi increase in pres-
sure. When the compressibility is signicant, it is modeled as a uid capacitor
(spring) and often is lumped in with the uid capacitance of the accumulator
(see Section 3.4).
When air bubbles are entrained in the hydraulic oil, the bulk modulus
drops and the uid becomes springy (Fig. 2.6). You may have experienced
this when the brake pedal in your car felt spongy. The solution was to bleed
the brake system, which releases the trapped air so that the brake uid be-
2.4. Fluid Behavior 15
Figure 2.6.: Fluid bulk modulous. Liquids are nearly incompressible (left), except
when there is trapped air, as shown on the right.
comes stiff again.
Another way that the uid can change properties is if the pressure fall be-
low the vapor pressure of the liquid causing the formation of vapor bubbles.
When the bubbles collapse, a shock wave is produced that can erode nearby
surfaces. Cavitation damage can be a problem for propellers and for uid
power pumps with the errosion greatly shortening the lifetime of compo-
nents.
The bulk modulus is dened as
=
P
V/V
(2.3)
where V is the original volume of liquid and V is the change in volume
of the liquid when subjected to a pressure change of P. Because V/V
is diminsionless, the units of are pressure. Water has a bulk modulus of
3.12x10
5
psi (2.15 GPa) while hydraulic oils have a bulk modulus between
2x10
5
and 3 10
5
psi (SAE 30 oil is 2.2 10
5
psi = 1.5 GPa), but can drop way
down if air is entrained.
Example 2.4.1. A piston pushes down on a volume of SAE 30 oil trapped in
a cylinder like the one shown in Figure 2.6. The cylinder bore is 5 in. and
the height of the unstressed column of oil is 4 in. The oil has a bulk modulus
= 2.2 10
5
psi. When a force of 10,000 lbs is applied to the piston, how
much does the piston displace?
16 2. Basic Principles of Fluid Power
Figure 2.7.: Pascal Law. Fluid at rest has same pressure everywhere and acts at right
angles against the walls.
Solution: Use Equation 2.3 to calculate the change in volume
V =
PV

=
(F/A)(Ah)

=
Fh

=
(1 10
4
) 4
2.2 10
5
= 1.8 cu. in.
The change in height is
h =
V
A
=
1.8
(/4)(5
2
)
= 0.092 in.
2.4.3. Pascals Law
Pascals Law states that in a conned uid at rest, pressure acts equally in
all directions and acts perpendicular to the conning walls (Fig. 2.7). This
means that all chambers, hoses and spaces in a uid power system that have
open passageways between them are at equal pressure so long as the uid is
not moving.
Pascals Law makes it easy to understand the operation of a simple hy-
draulic amplier. Figure 2.8 shows how it works. The piston on the right is
2.4. Fluid Behavior 17
Figure 2.8.: Hydraulic amplier. Pushing down with 10 lbs on the left piston raises
the 250 lb. weight on the right piston.
25 times the area of the piston on the left. Pushing down with 10 lbs on the
left piston will lift a 250 lb load on the right because the pressure of 10 psi is
the same everywhere.
2.4.4. High Forces
The crowning glory of uid power is the ability for small, light uid power
cylinders to produce extremely large forces. Consider the Caterpillar 345D
hydraulic excavator shown in Figure 2.9. The bore (diameter) of the cylinder
that drives the stick is 7.5 in. and the maximum working pressure is 5511 psi.
The peak force in the cylinder is therefore
F = P A = 5511 (7.5/2)
2
= 243, 468 lbs
or over 120 tons! It is impossible to produce this force using electric motors.
Figure 2.9.: A large excavator. Source: Caterpillar.
18 2. Basic Principles of Fluid Power
Figure 2.10.: Force in a cylinder acting at 60 psi. goes up with the square of the bore.
Force can be increased two ways, raising the pressure or increasing the bore
of the cylinder. Force is proportional to pressure but goes as the square of the
bore, which means small increases in cylinder size results in large increases in
force. For example, consider a small pneumatic cylinder with a 7/16 in. bore
running at 60 psi. This cylinder can push with 9 lbs of force. But, as shown
in Figure 2.10, force goes up as bore squared and a 4 in. bore cylinder at the
same 60 psi can push with 754 lbs of force.
Of course the extra force does not come free. A larger cylinder requires a
greater volume of uid to move the load than a smaller cylinder, the practical
result of the requirement that power be conserved.
2.5. Conduit Flow
In electromechanical actuation systems, power is carried to motors through
appropriately sized, low-resistance wires with negligible power loss. This is
not the case for uid power systems where the ow of oil through hydraulic
hoses and pipes can result in energy losses due to internal uid friction and
the friction against the walls of the conduit (Fig. 2.11). Designers size the
diameter and length of hydraulic hoses to minimize these losses. The other
cause of major losses in uid power systems is the orice drag of valves and
ttings. These will be discussed in Section 3 These loses are modeled as non-
linear resistances in the hydraulic circuit.
Conduit ow properties can analyzed using basic principles of uid me-
chanics. Typical ow patterns through pipes are shown in Figure 2.12. At
low velocities, the ow is smooth and uniform while at higher velocities the
2.5. Conduit Flow 19
Figure 2.11.: Hoses on an excavator.
ow turns turbulent. Turbulent ow can also be caused by sudden changes
in direction or when the area suddenly changes, conditions that are common
in hydraulic systems. Turbulent ow has higher friction, which results in
greater heat losses and lower operating efciencies. This is a practical concern
for designers of hydraulic systems because every right angle tting designed
into the system lowers the system efciency.
The Reynolds number, the non-dimensional ratio of inertial to viscous forces,
is commonly used to characterize the ow in pipes
Re =
V D
h

=
V D
h

(2.4)
where is the density (kg/m
2
), V is the mean uid velocity (m/s), D
h
is
the hydraulic diameter (m), is the dynamic viscosity (Pa-s) and is the
kinematic viscosity. For circular conduits, the hydraulic diameter is the same
as the pipe diameter. The more general formula, which accounts for non-
circular pipes and hoses is
D
h
=
4A
S
(2.5)
where A is the cross-sectional area of the conduit and S is the perimeter. For
a circle this formula reduces to diameter of the circle.
Figure 2.12.: Laminar and turbulent ow through a pipe.
20 2. Basic Principles of Fluid Power
For fully developed pipe ow, if Re < 2000 the ow is laminar, if 2000 <
Re < 4000 the ow is in transition, neither laminar or turbulent, and if Re >
4000 the ow is turbulent.
Example 2.5.1. A hydraulic hose with internal diameter of 1.0 in. is carrying
oil with kinematic viscosity 50 cSt at a ow rate of 20 gpm. Calculate the
Reynolds number and determine if the ow is laminar or turbulent.
Solution: Convert to SI units.
Q = 10 gpm = .00126 cu-m/s
D = 1.0 in. = .0254 m
= 50 cSt = 5.0 10
5
m
2
/s
Find the average velocity
V =
4Q
D
2
=
(4)(.00126)
(3.142)(.0254)
2
= 2.49 m/s
Calculate the Reynolds number
Re =
VD

=
(2.49)(.0254)
5.0 10
5
= 1265
Because Re is below 2000 the ow is laminar.
2.5.1. Pressure Losses in Conduits
The pressure losses in straight pipe and hoses contribute to the overall ef-
ciency of a uid power system. The losses can be estimated using the Darcy-
Weisbach equation
P = f
L
2D
V
2
(2.6)
where P is the pressure drop, f is the friction factor, L is the pipe length,
D is the pipe inside diameter and V
2
is the average ow velocity. There are
different formulas for the friction factor depending on the Reynolds number
and the surface roughness of the pipe or hose.
The experimental determination of friction factors are shown on the clas-
sic Moody Diagram
1
(Fig. 2.13) that plots the friction factor as a function of
1
Moody, L. F. (1944), Friction factors for pipe ow, Transactions of the ASME 66 (8): 671684
2.5. Conduit Flow 21
Figure 2.13.: Moody Diagram showing the experimental friction factor for pipes.
Reynolds number for a range of surface roughness of round pipes. Surface
roughness is dened as /D where is the mean height of the rougness of
the pipe. The diagram shows that for laminar ow, friction factor is inde-
pendent of roughness. For turbulent ow, the friction factor, and therefore
pressure losses, depend both on the Reynolds number and the roughness of
the pipe, except at high Reynolds numbers where losses depend only on the
roughness. Once the Reynolds number is determined, simple formulas can
be used to calculate pressure drop that avoid having to estimate numbers off
the Moody Diagram.
Laminar Flow For laminar ow, the friction factor depends only on the
Reynolds number
f =
64
Re
(2.7)
Using this equation and Equation 2.4 the pressure loss in a pipe with laminar
ow can be calculated as
P =
32L
D
2
V (2.8)
22 2. Basic Principles of Fluid Power
Converting average velocity to the more convenient uid ow rate (V =
4Q/D
2
) yields
P =
128L
D
4
Q (2.9)
This equation shows that for smooth ow, the pressure drop is proportional
to the ow rate Q. This is analagous to a linear resistor in an electrical system
and to a linear damper in a mechanical system.
Turbulent Flow, Smooth Pipes For turbulent ow, the friction factor de-
pends both on the Reynolds number and the surface roughness of the pipe,
however, in most uid power systems the pipes and hoses have smooth in-
teriors and the friction factor for smooth pipes can be used. Under these
conditions, an empirical approximation, dervied by Blasius and based on ex-
perimental data, can be used to calculate the friction factor
f =
0.316
Re
0.25
(2.10)
Combining with Equation 2.4 and converting from velocity to ow, the pres-
sure loss in a smooth pipe with turbulent ow is
P = 0.214

.25

.75
L
D
4.75
Q
1.75
(2.11)
The pressure loss is almost, but not quite, proportional to the square of the
ow rate, thus a pipe with turbulent ow acts like a nonlinear resistor.
Turbulent Flow, Rough Pipes If needed, the turbulent friction factor data
for rough piles in the Moody diagram can be modeled by the Colebrook
Equation
2
1

f
= 2 log
10
_
/D
h
3.7
+
2.51
Re

f
_
This equation must be solved iteratively for f, but a close-approximation di-
rect solution that can be used for hydraulic system modeling purposes comes
from the Swamee-Jain Equation
3
f =
0.25
_
log
10
_

3.7D
+
5.74
Re
0.9
_
2
2
Colebrook, C.F. (1938), Turbulent Flow in Pipes, J Inst Civil Eng 11:133.
3
Swamee, P.K.; Jain, A.K. (1976). Explicit Equations for Pipe-Flow Problems. Journal of the
Hydraulics Division, ASCE 102 (5): 657664
2.5. Conduit Flow 23
Example 2.5.2. Hydraulic oil ISO 68 is owing through a hydraulic line with
inside diameter 2.0 in. at a rate of 200 gpm. Find the pressure drop in psi for
a 10 ft length of hose.
Solution: Hydraulic oil ISO 68 has a density of 54.9 lb/cu-ft (880 kg/cu-m)
and a kinematic viscosity of 68.0 cSt at 104

F and 10.2 cSt at 212

F. For
this problem, assume the oil is at 104

F. First, convert to SI units.

Q =
6.31 10
5
cu-m/s
1 gpm
200 gpm = .0126 cu-m/s
D =
1 m
39.37 in.
2.0 in. = .0504 m
L =
1
3.281
10 = 3.048 m
=
1 m
2
/s
10
6
cSt
68.0 = 6.8 10
5
m
2
/s
= 880 kg/cu-m
= = (880)(6.8 10
5
) = .0598 Pa-s
Calculate the average velocity
V =
4Q
D
2
=
(4)(.0126)
(3.142)(.0504)
2
= 6.316 m/s
Calculate the Reynolds number
Re =
VD

=
(6.316)(.0504)
6.8 10
5
= 4681
This indicates turbulent ow. Using Equation 2.11 for turbulent ow and a
smooth pipe gives the pressure drop
P = 0.214
(.0598
.25
)(880
.75
)(3.048)
(.0504
4.75
)
(.0126
1.75
)
= 35, 980 Pa
=
1 psi
6895 Pa
35, 980
= 5.2 psi
Because most hydraulic systems operate we above 500 psi, the 5.2 psi drop
in the 10 ft. line is inconsequential. The pressure drop scales linearly with
line length so longer hoses, and certainly smaller hoses, could impact system
efciency.
24 2. Basic Principles of Fluid Power
Note: Calculations such as the one in the last example can be tedious and
prone to errors because it is easy to lose track of units. If you nd yourself
doing these calculations often, consider developing a spreadsheet macro that
takes care of the details. Software for hydraulic line calculations is also avail-
able
4
and uid power simulation software automatically caclulate line losses,
a big plus if you are designing a large system.
2.6. Bends and Fittings
Flow through long straight pipes is not common in practical uid power sys-
tem as typically ow goes through right-angle ttings and short sections of
bent, exible hose. For example, Figure 2.14 shows an SAE 90 deg elbow t-
ting made to SAE/JIC standards. The pressure drops in these pathways can
be signicant and must be estimated to analyze systemefciency. While uid
mechanics theory and computational uid dynamics simulations can be used
to generate precise values of pressure loss, the uid power engineer generally
uses tabulated values of dimensionless loss coefcients K for each component.
The pressure drop across the component is
P = K

2
V
2
= K

2A
2
Q
2
(2.12)
where A is the area of the tting. Note that the pressure drop is proportional
to the ow squared. Inverting Equation 2.12 gives the ow as a function of
pressure
Q =
A

K
_
2

P (2.13)
Loss coefcients can be found in vendor catalogs. Table 2.3 lists K for some
common ttings and ow path geometries.
5
For accurate modeling, loss
coefcients must be found experimentally.
2.7. Orice Flow
The third major pressure loss comes from the ow of uid through the re-
stricted orices found in valves and some ttings. Here the classic orice
equation, valid for steady, incompressible ow with Re >> 1, is used to
model the ow
Q = C
d
A

2P

(2.14)
4
For example, a free web-based calculator is at www.efunda.com/formulae/uids/calc pipe friction.cfm
and a comprehensive hydraulic sizing software tool that includes line loss calculations is avail-
able for a small cost from www.CompuDraulic.com.
5
For a 45 and 90 deg elbows, the loss coefcients depend on the ratio of the radius of the
bend to the diameter of the tting. Texts such as [1] and [2] have reference information.
2.7. Orice Flow 25
Figure 2.14.: SAE hydraulic elbow tting.
or rewriting for the pressure drop
P =

2A
2
C
2
d
Q
2
(2.15)
where A is is cross-section of the orice and C
d
is the discharge (or valve)
coefcient. The discharge coefcient can be a variable, changing with valve
position, however, an average value for C
d
of 0.62 is often used to simplify
calculations leaving area A to change with valve position.
The xed parameters can be lumped together to form a slightly different
version of the orice loss equation
Q = C
v
_
P
SG
(2.16)
or
P =
SG
C
2
v
Q
2
(2.17)
where SG is the dimensionless specic gravity for the uid (ratio of the den-
sity of the uid to the density of water) and C
v
is the valve coefcient. When
U.S. manufacturers state a value for C
v
in a valve data sheet, it assumes that
Table 2.3.: Loss coefcients K for geometric elements
Fitting K
90

deg elbow 0.2

45

deg elbow 0.15

Tee tting 0.9
Sharp-edged entrance 0.5
Rounded entrance 0.05
Sharp-edged exit 1.0
Rounded exit 1.0
26 2. Basic Principles of Fluid Power
Qis in gpm and P is in psi and that the temperature and viscosity are xed.
Typical hydrocarbon based oils have specic gravities between 0.85 and 0.95.
In all cases, note that ow is proportional to the square root of the pressure
drop across the valve, which makes it a nonlinear resistance.
Example 2.7.1. Find C
v
for a Sun Hydraulics DAAAsolenoid-operated on-off
valve when running hydraulic oil with a specic gravity of 0.864.
Solution: The Sun Hydraulics DAAA valve is a 2-position, 2-way cartridge
valve. When the solenoid is energized the valve switches from blocking the
ow to a fully open ow. The data sheet is at www.sunhydraulics.com. Some
of the key valve specs are: Max ow = .25 gpm (1 L/m), Max pressure = 5000
psi (350 bar), Response time = 30 ms. The gure below shows a photo of the
valve, the valve schematic and the experimental pressure vs. ow character-
istics for the valve in its open position. These are cartridge valves, designed
to screw into a manifold with the solenoid on top.
The following approximate (gpm,psi) data points can be read off the graph:
(0,0), (0.3,50), (0.5,100), (1,410). With the help of a spreadsheet the points were
2.7. Orice Flow 27
t to an approximate parabola, psi = 400gpm
2
. Using Equation 2.17
SG
C
2
v
= 400
C
v
=
_
SG
400
=
_
.864
400
= 0.046
3. Fluid Power Components
This section describes the most common components found in typical uid
power systems. Basic mathematical models for the consitutive properties of
components are presented, which are the building blocks for understanding
how to select components and for simulating uid power circuits. Standard
symbols for the schematic representation of ud power components are also
introduced. Fluid power symbols are referenced in ISO standard 1219-1:2006,
which is explained in Appendix A. While this section will not make you
an expert in practical hydraulic and pneumatic systems, it will build your
understanding of the fundamental principles underlying such systems.
3.1. Cylinders
Cylinders convert uid power pressure and ow to mechanical translational
power force and velocity (Fig. 3.1). Cylinders are linear actuators that can
push and pull, and when mounted around a joint, for example, as is done
in an ecavator, can actuate rotary motion. Cylinders come in single acting
(push only), single acting with spring return and double-acting (push-pull).
The rest of the section will focus on double acting cylinders, which are most
common in hydraulic applications.
1
A cutaway illustration of a typical double-acting cylinder used for indus-
trial applications is shown in Figure 3.2 and the ISO symbol is shown in Fig-
ure 3.3
Figure 3.1.: Hydraulic cylinder.
1
For a primer on cylinders see: www.hycocanada.com/cylinderintro.html or en.
wikipedia.org/wiki/Hydraulic\_cylinder.
28
3.1. Cylinders 29
Figure 3.2.: Cut-away view of a welded body, industrial double-acting hydraulic
cylinder. Source: Hyco International
The end of the cylinder where the rod emerges is called the rod end and
the other end is called the cap end. This distinction is important for modeling
because the rod side of the piston within the cylinder has less surface area
than the cap side of the piston. For the same pressure a double-acting cylin-
der can push with much greater force than it can pull. Examination of the
cylinder conguration in an excavator (e.g. Fig. 1.4 shows how designers can
take advantage of the larger pushing force. Cylinders have considerable fric-
tion, particularly around the piston because of its large circumference with
wrap-around seals. The rod seal tends to be even tighter than the piston seals
to prevent leaking of hydraulic oil, but because of the smaller circumference,
rod seals play less of a role in cylinder friction.
While the design of a cylinder is complex, the dynamic model used for most
simulations simply captures the pressure-force transformation and sometimes
includes the cylinder friction and leakage around the piston seal. The den-
ing equations for an ideal, friction-free, leakless cylinder are
F = PA (3.1)
V = Q/A (3.2)
The piston force depends on the difference in pressure across the piston, tak-
ing into account the area on each side. Referring to Figure 3.4 the piston force
is
F
P
= P
1
A
cap
P
2
A
rod
(3.3)
where all pressures are gauge pressures with respect to atmospheric pressure.
Figure 3.3.: ISO schematic symbol for a double-acting cylinder.
30 3. Fluid Power Components
Figure 3.4.: Forces on a cylinder.
2
and the piston velocity is
V =
Q
1
A
cap
=
Q
2
A
rod
(3.4)
The rod area is the piston annulus around the rod and is
A
rod
=

4
_
bore
2
roddia
2
_
Another consequence of the different areas on either side of the piston is that
the oil ow in one port will not be equal to the ow out the other port. If the
return line to the reservoir is long or the return line valve has small orices
then the pressure build up on the rod side of the cylinder when pushing can
be signicant and must be modeled.
The overall efciency of a cylinder is given by the ratio of the output me-
chanical power to the input uid power
=
FV
P
i
Q
i
(3.5)
where P
i
and Q
i
refer to either the cap or rod side depending on whether the
piston is pushing or pulling. Cylinder efciency can be split into two parts,
the force efciency

f
=
A
i
V
Q
i
(3.6)
and the volumetric efciency

v
=
F
P
i
A
i
(3.7)
2
For pneumatic systems where compressible gas laws dene cylinder behavior, pressures
are absolute, which means that the dening force equation is
F
P
= P
1
A
cap
(P
2
A
annulus
+P
atm
A
r
)
where A
annulus
is the area of the piston annulus around the rod and A
r
is the area of the piston
rod.
3.2. Pumps and Motors 31
with the overall efciency being the product =
f

v
. Using these relations,
Equations 3.3 and 3.4 can be modied for a non-ideal cylinder with friction
and leakage
F
P
= (P
1
A
cap
P
2
A
rod
)
f
(3.8)
V =
Q
1
A
cap

v
=
Q
2
A
rod

v
(3.9)
3.2. Pumps and Motors
Hydraulic pumps supply energy to the system, converting the torque and ve-
locity of an input shaft to pressure and ow of the output uid. Hydraulic
motors are the rotary equivalent of cylinders. Pressurized oil ows through
the motor and produces rotation of the output shaft. One way of looking at
a motor is simply as a pump driven in reverse (uid ow in, shaft rotation
out versus shaft rotation in and uid ow out) and indeed many devices can
act both as a pump and a motor, much the same way that a DC permanent
magnet motor can act both as a motor and a generator. The common types
of pumps and motors are gear and vane pumps and motors and axial and
radial piston pumps and motors motors. Fixed displacement devices rotate a
xed amount for a xed volume of uid. Variable displacement devices have
a mechanical or uid power control port that can be used to vary the rela-
tion between uid volume and shaft rotation, and are essentially a variable
transmission between uid and mechanical rotatary power. Pumps are gen-
erally driven by electric motors in industrial applications and diesel engines
in mobile applications.
3
The symbols for xed displacement, uni-directional
pumps and motors are shown in Figure 3.5.
Ideal pumps and motors are dened by the relations between uid pres-
sure P and ow Q and shaft torque T and velocity . For an ideal motor,
input and output power is conserved. If P is the pressure difference across
the pump or motor, then the power balance for a pump or motor is
Power = PQ = T (3.10)
If the volumetric displacement of the motor or pump is D
v
, typically ex-
pressed in cubic in./rad, then the relations relating uid to mechanical are
T = D
v
P (3.11)
Q = D
v
(3.12)
3
For descriptions of how various pumps and motors are work and their applications, see
these web resources:
http://en.wikipedia.org/wiki/Hydraulic_pump
http://www.hydraulicspneumatics.com/200/TechZone/HydraulicPumpsM/74
32 3. Fluid Power Components
Figure 3.5.: Pumps, motors and their schematic symbols. The difference between the
two symbols is the direction that the triangle is pointing; out for pumps,
in for motors
The parameter D
v
is the sole parameter that denes the operating charac-
teristics of an ideal pump or motor and in that respect is similar to the torque
constant of a DC motor. From the above equations, it can be seen that a pump
can be a constant ow source if its shaft is rotated at constant velocity and a
constant pressure source if the shaft torque is xed. In practice, neither is the
case as the shaft is rotated by an electric motor and the motor-pump combi-
nation results in pressure-ow performance curves. An example is shown in
Figure 3.6.
Real pumps and motors are not 100% efcient and, like cylinders, have an
overall efciency , which is made up of volumetric
v
and mechanical
m
Figure 3.6.: A small, packaged hydraulic power supply containing a 115V AC elec-
tric motor, hydraulic pump, relief valve, reservoir tank and lter. The
combined electric motor plus pump results in the output pressure-ow
characteristics shown in the data sheet to the right. Source: Oildyne.
3.3. Control Valves 33
Table 3.1.: Efciency equations for pumps and motors
Pump Motor
= PQ/T = T/PQ

v
= Q/D
v

v
= D
v
/Q

m
= D
v
P/T
m
= T/D
m
P
efciencies with
=
v

m
(3.13)
Table 3.1 shows the equations for the various efciencies for pumps and for
motors.
3.3. Control Valves
Control valves are essential and appear in all uid power systems. Valves are
sometimes categorized by function, which includes directional control valves
for directing uid ow to one or the other side of a cylinder or motor, pres-
sure control valves for controling the uid pressure at a point and ow con-
trol valves for limiting the uid ow rate in a line, which in turn limits the
extension or retraction velocities of a piston.
Valves are also characterized by the number of ports on the valve for con-
necting input and output lines and by the number of operating positions that
the valve can assume. For example, a 3-way, 2-position valve commonly
found in pneumatic systems has three ports for connecting supply line, ex-
haust or reservoir line and output line to the cylinder and two positions. In
one position the supply line connects to the cylinder line extending the piston.
In the other position the exhaust line connects to the cylinder retracting the
piston, assuming the piston has a spring return. On/off valves can only be
in the states dened by their positions while proportional valves are continu-
ously variable and can take on any position in their working range. A servo
valve is a proportional valve with an internal closed-loop feedback mecha-
nism to maintin precise control over the valve behavior. Example valves are
shown in Figure 3.7
3.3.1. Dynamic Models for Valves
For dynamic modeling purposes, valves are fundamentally variable orices
where the area of the orice depends on the valve position. For example, the
core dynamic model of a solenoid proportional valve has the area of an orice
as a nonlinear function of the command signal to the solenoid.
34 3. Fluid Power Components
Figure 3.7.: Types of control valves. Left to right: hand-operated directional valve
for a log splitter. On-off minature, solenoid acuated pneumatic valve.
Precision proportional pneumatic valves. High precision, apper-nozzle
hydraulic servo valve.
The basic equation for a valve is the orice equation introduced in Section
2.7 and repeated here
Q = C
d
A

2P

(3.14)
where C
d
is valve coecient, A is the area of the valve opening and P is the
pressure drop across the valve. For valves with internal spools and rectangu-
lar orice slots, the orice opening area is proportional to the valve position,
A = wx.
To simplify analysis, the orice equation can be linearized about a nominal
operating point at the x = 0 valve position with leakage ow Q
0
Q = Q
0
+K
q
x +K
c
P (3.15)
The linearized valve is characterized by two parameters, the ow gain coef-
cient
K
q
=
Q
x
= C
d

2P

A
x
= C
d
w

2P

(3.16)
and the ow pressure coefcient
K
c
=
Q
P
=
AC
d

2P
=
wxC
d

2P
(3.17)
3.3.2. Valve Symbols
Because there are so many types of valves, their symbols can be complex.
Some of the more basic symbols are covered in this section.
A valve has one square for each working position. The nominal or initial
valve position has connection points, or ways, to the valve ports. Thus a
three-way two-position (3/2) valve would have three connection ports and
two boxes as shown in Fig. 3.8. The ports are sometimes labelled with letters
with A,B,C, . . . indicating working lines, P indicating the pressurized supply
line and T or R indicating the return (tank) line connected to the reservoir.
3.4. Accumulators 35
Figure 3.8.: Basic symbology for uid power valves. The number of squares is the
number of valve positions. Image (a) represents a three-position valve
and (b) represents a two-position valve. The number of connection points
on the symbol is the number of ports (ways) for the valve. Image (c)
represents a 3-way, 2-position valve or 3/2 valve for short.
Lines with arrows inside the boxes indicate the path and direction of ow.
Pneumatic systems are indicated with unlled arrow heads. Examples are
shown in Figure 3.9.
The icons on the side of the symbol indicate how the valve is actuated.
Common methods include push button, lever, spring-return, solenoid (for
computer-control) and pilot-pressure line. Examples are shown in Figures
3.10 and 3.11. Also shown in Figure 3.11 is an example of a proportional
valve that can take on any position. An innite position, proportional valve
is indicated by parallel lines on top and bottom of the symbol.
3.4. Accumulators
Hydraulic accumulators are used for temporarily storing pressurized oil. The
oil enters a chamber and acts against a piston or bladder to raise a weight,
compress a spring or compress a gas. Accumulators are used to supply tran-
sient peak power, which reduces the ow rate requirement for the power
supply and to act as shock absorbers for smoothing out pressure wave spikes.
Figure 3.9.: A 4/2 valve with four connection points and two positions. In the nomi-
nal, unactuated state, supply line P connects to working line Aand work-
ing line B connects to return (tank) line T. In the actuated state, the valve
slides to the right and supply line P connects to working line B and work-
ing line A connects to return line T.
36 3. Fluid Power Components
Figure 3.10.: Valve actuation symbols. Left to right: push-button, lever, spring-
return, solenoid, pilot-line.
Accumulators are the equivalent to a capacitor in an electrical systemand to a
spring in a mechanical system. Bladder type accumulators, precharged with
nitrogen gas are the most common type for hydraulic systems (Fig. 3.12).
The capacity of a uid capacitor is dened by its change in volume divided
by its change in pressure
C
f
=
V
P
(3.18)
Change in volume per time is ow rate and change in pressure per time is the
derivative of pressure. This leads to the constituative law for a linear uid
capacitor
Q = C
f

P (3.19)
For a gas-lled accumulator, the capacitance C
f
will depend on the accumu-
lator pre-charge. The capacitance is the slope of the accumulator volume-
pressure curve, which is sometimes given in the manufacturers data sheet.
If the curve is nonlinear, the slope at the operating point should be taken for
C
f
.
Another type of accumulator is a cylinder with the uid pushing on one
side of the piston against a stiff spring on the other side of piston. For these
spring-loaded piston accumulators the capacitance is
C
f
=
A
2
K
(3.20)
Figure 3.11.: (a) Pushbutton hydraulic 4/2 valve. In the nominal position, the cylin-
der is held in the retracted postion because supply line P is connected
to the rod side. When the button is pushed, the rod extends because
supply line P is now connected to the cap side. When the button is
released, the valve spring returns the valve to the nominal position, re-
tracting the rod. (b) Solenoid 4/3 valve with center closed position and
spring return to center. Acomputer can control extension and retraction
of the cylinder by actuating the valve solenoids. (c) Same as b, but with
a continuously variable proportional valve.
3.5. Filters 37
Figure 3.12.: Accumulator and ISO symbol.
where A is the area of the piston and K is the spring constant.
3.5. Filters
During use, hydraulic oil picks up contaminating particles from wear of slid-
ing metallic surface that add to residual contaminants from the oil manufac-
turing process, rust from metal and polymer particles from seal wear. These
dirt particles are tiny grit that cause additional abrasive wear. Clumps of par-
ticles can clog tiny clearances in precision valves and cylinders and can lead
to corrosion.
All practical hydraulic systems require a lter in the circuit (Fig. 3.13). In-
line lters have a ne mesh media formed from wire, paper or glass ber,
formed to create a large surface area for the uid to pass through. The oil
lter in your car is an example of a hydraulic lter. Sometimes the lter is
included inside the reservior or is part of an integrated power supply unit
along with the motor, pump and reservoir. Selecting a lter is a tradeoff be-
tween a media that traps ne contaminants and one that passes uid with
minimal resistance.
The dynamic model for a lter is a nonlinear resistance P = f(Q) that can
be linearized about the nominal ow. If the pressure drop across the lter
is small compared to other pressure drops in the system, the effects of the
lter on the dynamic model can be ignored. Resistance values for simulation
models can be estimated from the manufacturers data sheet or from a lter
characterization experiment.
3.6. Reservoirs
The main function of the reservoir is to provide a source of room tempera-
ture oil at atmospheric pressure (Fig. 3.14). The reservoir is equivalent to
the ground in an electrical system. Conceptually, a reservoir is nothing more
38 3. Fluid Power Components
Figure 3.13.: Hydraulic lter and ISO symbol.
than an oil storage tank connected to atmosphere through a breather and hav-
ing pump and return lines to deliver and accept oil. In practice, a reservoir
has additional functions including de-arating and acting as a heat exhanger.
The dynamic model of a reservoir is to treat it as a ground, a source of zero
pressure.
3.7. Hoses and Fittings
The glue that connects the various components together are the hydraulic
hoses and ttings. As described in Sections 2.5, 2.6 and 2.7, they are modeled
as uid power resistors with with linear or non-linear pressure-ow charac-
teristics. Symbols for pipes and hoses are shown in Figure 3.15
Figure 3.14.: Hydraulic reservoir and ISO symbol.
3.7. Hoses and Fittings 39
Figure 3.15.: Symbols for uid power lines. Left to right: lines joined, lines crossing,
exible line
4. Hydraulic Circuit Analysis
A basic hydraulic circuit is shown in Figure 4.1. It contains a motor-driven,
xed-displacement hydraulic pump, a 4-way, 3-position, center off valve and
a double-acting hydraulic cylinder. A pressure relief valve is stationed be-
tween the pump output and the return line. This valve is required because
otherwise, with the valve shut and no place for the supply uid to move, the
output pressure of the xed displacement pump would quickly build up to
dangerous levels. A typical setting for the relief valve for a small system is
600 psi. The combination of the xed displacement pump and the relief valve
effectively turns the combination into a constant pressure supply, assuming
the pump owrate can keep up with the load demands. While the pump plus
relief valve combination is common, it is not efcient because when the sys-
tem is idling, the pump is wasting energy pushing ow through the pressure
valve drop back to the tank.
Hydraulic circuit static and dynamic analysis involves rst developing the
appropriate mathematical models for each component in the circuit and then
using Pascals Law(pressure same at all points for uid at rest) and conserva-
tion of ow to connect the components into a set of equations that describes
the complete system. For a static analysis, the set of equations will be alge-
braic while a dynamic analysis a set of differential equations.
For example, the circuit shown in Figure 4.1 could be modeled as a constant
source of pressure feeding into the valve because of the combination of the
positive displacement pump coupled to the relief valve. The pressure would
Figure 4.1.: A basic hydraulic circuit with xed-displacement pump, pressure relief
valve, 4/3 solenoid valve and cylinder.
40
4.1. Fluid Resistance 41
be the setpoint of the relief valve, for example 600 psi. The valve could be
modeled as a nonlinear resistance whose value depends on the position of
the valve spool as shown in Section 3.3.1. The pressure drop in the lines
would also be modeled as a nonlinear resistor as described in Section 2.5.1.
The cylinder is modelled as a transformer that converts pressure and ow to
force and velocity (Section 3.1). Not shown is the mechanical load that the
cylinder acts against, which would be some combination of inertia, friction,
damping and spring. The return lines from the cylinder through the valve
and back to the reservoir would be modeled as a resistors in series. If the lter
introduced signicant pressure losses, it would also be modeled as a resistor.
Because the pressures in typical hydraulic systems are so high (1000 to 3000
psi), small pressure drops in hoses and lters are often neglected or treated
in rule-of-thumb tables when sizing power supplies. Accurate models are
required, however, to understand efciencies and detailed system behavior.
A steady-state static analysis of the circuit would entail writing equations
for a network of nonlinear resistors and the force balance across the piston.
These equations can be solved to determine the pressures and ows at var-
ious points in the circuit. A dynamic analysis with differential equations is
needed if the cylinder pushed against a load with springs or inertias, if there
were an accumulator in the circuit or if uid capacitance were signicant.
For dynamic analysis, particularly for the purpose of designing a high per-
formance hydraulic control system, the system is generally linearized so that
linear control design methods can be used. The process for linearizing a con-
trol valve was described in Section 3.3.1.
The rest of this section covers the building blocks needed to develop dy-
namic models, resistance, capacitance, inertance and source elements, and
presents several modeling examples.
4.1. Fluid Resistance
Fluid resistors are any component that resists ow. Another way of looking
at uid resistors are any component that causes a pressure drop when uid
ows through the component. Fluid resistors include valves, lters, hoses,
pipes and ttings. Ageneralized linear uid resistor relates owand pressure
P = R
f
Q (4.1)
or
Q =
1
R
f
P (4.2)
As we saw in the sections on conduit, orice and valve ow, real uid
resistors are nonlinear and typically relate ow to the square root of pressure.
The general form of resistance can be written as
P = f
R
(Q) (4.3)
42 4. Hydraulic Circuit Analysis
Figure 4.2.: Symbol for a xed (left) or variable (right) restrictor valve. This symbol
is sometimes used to indicate a parasitic resistance, for example the ow
resistance in a pipe.
or
Q = f
1
r
(P) (4.4)
In uid power schematics, sometimes a general resistance, for example to
represent the ow resistance in a pipe, is indicated with the restrictor valve
symbol (Figure 4.2).
4.2. Fluid Capacitance
Fluid capacitors are one of two types of energy storing elements in uid
power systems. Capacitance in a uid power circuit comes from discrete ac-
cumulators (Section 3.4 but also from the uid itself if it is compliant. Fluid
compliance is essential to consider in pneumatic systems, but generally does
not play a signicant role in dynamic models of hydraulic systems unless
there is signicant trapped air causing spongy behavior.
The capacitance of the uid is captured by its bulk modulus property (Sec-
tion 2.4.2. The capacitance of a trapped section of compliant uid can be
determined as shown in the following example.
Example 4.2.1. Find the equivalent capacitance of the hydraulic ram for the
system described in Example 2.4.1
Solution: From the solution to Example 2.4.1 the uid volume changed 1.8
cu. in. for a pressure change of 10, 000/19.63 = 509 psi. The capacitance of
the uid trapped in the cylinder is
C
f
=
1.8
509
= 3.54 10
3
The idea of capacitance of uid trapped in a cylinder can be expanded to
estimate the capacitance of a plug of uid in a hose or pipe, which in turn can
4.3. Fluid Inertance 43
be used in a dynamic model. One application of a dynamic model involving
uid compression is to understand water hammer, which is impact loading
cause by sudden changes in ow, such as when a valve is switched from on
to off.
Equation 3.18 states that capacitance is the change in volume divided by
change in pressure while Equation 2.3 denes the bulk modulus as the
change in pressure divided by the normalized change in volume. These can
be combined into an expression for the capacitance of a known volume of
uid, for example the uid in a length of pipe.
C
f
=
V
P
(4.5)
=
V

(4.6)
Example 4.2.2. Find the uid capacitance of SAE 30 oil with bulk modulus
= 2.210
5
psi. owing through in a 20 inch length of 2 inch diameter hose.
Solution: The volume of the uid is
V =
LD
2
4
Using Equation 4.6 the capacitance is
C
f
=
V

=
LD
2
4
=
(20)(3.14)(4)
2.2 10
5
= .00114 in.
5
/lb.
In a uid power network, uid capacitance that comes from an accumu-
lator is referenced to ground (zero gauge pressure) while capacitance from
compressibility of the uid is referenced to the pressures at the two ends of
the uid plug.
4.3. Fluid Inertance
The second type of energy storing element is uid inertance. In mechanical
systems, mass and rotary inertia often dominate systembehavior and must be
44 4. Hydraulic Circuit Analysis
Figure 4.3.: Flow and pressure connection laws. The ows into a node must sum to
zero. The pressure drops around a loop must sum to zero. In the gure
on the right, P
i
indicates the pressure drop across component i.
modeled. In uid power systems, the inertia of the uid is generally insignif-
icant and usually ignored in dynamic system models. The reason is that in
hydraulic systems, pressures are so high that inertial forces can be neglected
and in pneumatic systems the mass of air is so lowthat inertial forces can also
be neglected.
When analyzing high frequency behavior of a system, for example with
sudden on off switching of valves that causes transients in uid ow, uid
inertance should be included in the model. The inertance of a plug of uid in
a hose is simply the mass of the uid
I
f
=
L
A
(4.7)
where is the uid density and L and A are the length and area of the hose.
In a network, the inertance of uid in a pipe is always modeled as being in
series with the ow resistance of the pipe.
4.4. Connection Laws and States
4.4.1. Connections
When hydraulic components are connected, conservation of ow and pres-
sure loop principles are used to write the equations that join components.
Figure 4.3 shows these simple rules.
4.4.2. State Variables
For generating systemequations, the state variables are the pressures P at the
nodes and the ows Qthrough the elements. These are completely analogous
to voltage V and current I, the states for electrical systems.
4.5. Example Systems 45
For dynamic system models, the order and the number of rst order dif-
ferential equations that describe the system is equal to the number of inde-
pendent energy storage elements. If there are no uid capacitances or uid
inertances in the systemthen the order of the systemis zero and the equations
will be purely algebraic with no derivatives.
4.5. Example Systems
Example 4.5.1. Write the state equations for the system shown in the gure.
There are two resistances, one is an orice the other is the pipe resistance. The
output of interest is the pressure at the accumulator.
Solution: The pressure at the reservoir is 0 and the pump is modeled as a
pressure source with output pressure P
S
. There is only one other pressure
in the system, P
A
, the pressure in the accumulator. Write the element and
connection equations.
Orice The behavior of the orice is described by Equation 2.14 or Equation
2.16. For this analysis, lump all constants into one paramter K
o
= C
v
_
1/SG
so that Q = K
o

P
orice
. From continuity, the ow through the orice is Q
S
P
orice
= P
A
P
S
Q
S
= K
o
_
P
A
P
S
(4.8)
Pipe Resistance Assume turbulent ow. The full expression for turbulent
ow in a smooth pipe is given by Equation 2.11. For this analysis, make the
approximation that the pressure loss is proportional to ow squared rather
46 4. Hydraulic Circuit Analysis
than to ow to the 1.75 power. Use K
P
for the porportional constant. The
ow through the pipe is the output ow Q
o
and the pressure across the pipe
is P = P
A
0 = P
A
. Thus, the pipe resistance is described by
Q
o
= K
P
_
P
A
(4.9)
Accumulator The accumlator is the only energy storage element in the sys-
tem and is described by Equation 3.19
Q
A
= C
f

P
A
(4.10)
where Q
A
is the ow into the accumulator.
Continuity Conservation of ow dictates that
Q
S
= Q
A
+Q
o
(4.11)
State Equation The goal is to nd a set of state equations (for this rst-
order example with one energy storage element there will be one equation)
with P
A
as the output and P
S
as the input. Using (4.10), (4.9) and (4.11) yields
C
f

P
A
= Q
S
Q
o
= Q
S
K
P
_
P
A
(4.12)
Using (4.8) and (4.12)
C
f

P
A
= K
o
_
P
A
P
S
K
P
_
P
A
(4.13)
Dividing by C
f
yields the state equation

P
A
=
K
o
C
f
_
P
A
P
S

K
P
C
f
_
P
A
(4.14)
Because the state equation is nonlinear, it cannot be solved directly but can
be easily simulated numerically in a package such as Simulink. A Simulink
block diagram for this example is shown below.
5. Bibliography
[1] Herbert Merritt, Hydraulic Control Systems. Wiley, 1967.
The classic text on hydraulic control systems. Still useful.
[2] Noah Manring, Hydraulic Control Systems. Wiley, 2005.
A modern rewrite of Merritts text. Highly recommended.
[3] Eaton-Vickers, Industrial Hydraulics Manual. Eaton Corportation, 2001.
A standard text for training uid power technicians. Contains good,practical
information. Used in the uid power lab course at the University of Minnesota.
[4] Arthur Akers, Max Gassman and Richard Smith, Hydraulic Power System
Analysis. Taylor & Francis, 2006.
Good text on hydraulic system analysis.
[5] Lightning Reference Handbook. Berendsen Fluid Power Inc., 2001.
Comprehensive reference information for hydraulics.
[6] Andrew Parr, Hydraulics and Pneumatics: A technicians and engineers
guide. Butterworth Heinemann, 1998.
Good introductory overview of uid power systems.
47
A. Fluid Power Symbols
Fluid power symbols are set by International Organization for Standardiza-
tion (ISO) standards, ISO 1219-1:2006 for uid power system and component
graphic symbols and ISO 1219-2:1995 for uid power circuits.
The following tables show the basic ISO/ANSI symbols for uid power
components and systems.
1
Tables will appear in the next version. For now, lists of symbols can be found at
these web locations:
http://www.hydraulicsupermarket.com/upload/db_documents_doc_19.
pdf
http://www.patchn.com/index.php?option=com_content&task=view&id=
31&Itemid=31
http://www.hydraulic-gear-pumps.com/pdf/Hydraulic\%20Symbols.pdf
http://www.hydrastore.co.uk/products/Atos/P001.pdf
http://www.scribd.com/doc/2881790/Fluid-Power-Graphic-Symbols
1
Microsoft VISIO has a library of symbols for generating uid power schematics, although
may not be in the latest ISO format.
48