energies

Article
Reducing Fuel Consumption in Hydraulic
Excavators—A Comprehensive Analysis
Milos Vukovic, Roland Leifeld * and Hubertus Murrenhoff
Institute for Fluid Power Drives and Controls (IFAS), RWTH Aachen University, 52062 Aachen, Germany;
[email protected] (M.V.); [email protected] (H.M.)
* Correspondence: [email protected]; Tel.: +49-24180-27528

Academic Editor: Andrea Vacca

Received: 29 March 2017; Accepted: 9 May 2017; Published: 12 May 2017

Abstract: Mobile machines, especially excavators, still consume considerable amounts of fuel during
their operating lifetimes. This is not only undesirable in economic terms but also adversely affects
our environment. The following paper discusses methods to lower fuel consumption by conducting
a comprehensive analysis of the components comprising a hydraulic excavator and the cycles these
machines perform. One of the main aims is to emphasise that a design centred on the standard
definitions of efficiency, especially hydraulic efficiency, can be rather misleading. A new approach
using a novel fuel consumption model, based on the Willans approximation, coupled with the
concepts of fixed and variable fuel consumption is introduced and validated using real test data
obtained from an 18 t excavator. The new methodology can be used to help uncover simpler methods
to improve today’s machines.

Keywords: hydraulic excavators; hybrid systems; fuel consumption; energy recovery; engine
downspeeding; independent metering

1. Introduction
Hydraulic excavators are responsible for approximately 60% of the CO2 emissions produced by
construction machinery [1]. This is in part due to the sheer number of machines in use and also to
their extremely low efficiencies of around 10% [2]. Despite their immense impact on our environment
many aspects regarding the exact reasons for their high fuel consumption remain misunderstood.
The hydraulic systems used to power these machines are often unjustly blamed for the majority of
the losses. As a result, much research has gone into the development of more efficient hydraulic
architectures capable of lowering so-called throttling losses and enabling energy recovery. Much less
attention has been paid to the machine as a whole.
The following work aims to clarify many issues and myths surrounding hydraulic excavators by
guiding the reader through a comprehensive analysis of the whole machine. The losses occurring in
components and subsystems are explained and a detailed discussion of measurement data, obtained
from field tests with an 18 t machine, is presented. These thoughts ultimately lead to the introduction
of a novel fuel consumption model, capable of describing and predicting the fuel consumption of a
machine for all duty cycles. Instead of using the widespread definitions of machine and hydraulic
efficiency, the concepts of fixed and variable fuel consumption are introduced. An important aspect of
the research is the validation of the methodology using real measurement data.
The authors hope to provide a tool that can be used by engineers, during the initial design
phase, to easily evaluate the fuel saving potential of different architectures before running complex
system simulations. As an example, solutions involving independent metering, displacement control
and hybridization are discussed. Along with a brief summary, the paper concludes with an outlook
concerning future work.

Energies 2017, 10, 687; doi:10.3390/en10050687 www.mdpi.com/journal/energies

Energies 2017, 10, 687 2 of 25

Energies 2017, 10, 687 2 of 25

Energies 2017,
2. Review of 10, 687
Components and Subsystems 2 of 25
2. Review of Components and Subsystems
Figure 1ofillustrates
2. Review the typical layout and individual components of a state of the art machine.
Figure 1 Components
illustrates theand Subsystems
typical layout and individual components of a state of the art machine.
One or several hydraulic pumps, powered by a diesel engine, provide pressurised flow to the system.
One or several
Figure hydraulicthe
1 illustrates pumps, powered
typical layout byanda diesel engine,
individual provide pressurised
components flow
of a state of thetoart
themachine.
system.
Using
Using joysticks in in thethe
cab, thethe
operator controls a series of directional valves located in a manifold block,
One orjoysticks
several hydraulic cab,
pumps, operator
powered controls a series
by a diesel of directional
engine, valves
provide pressurised located
flow in a manifold
to the system.
often referred
block, to as
often referredthe main control valve (MCV). These allow an intuitive and precise distribution
precise of
Using joysticks in the to asthe
cab, theoperator
main control
controlsvalve (MCV).
a series These allow
of directional an located
valves intuitiveinanda manifold
thedistribution
incoming pump flow to the individual actuators of the implement structure (boom, arm, bucket,
block, oftenofreferred
the incoming
to as pump flowcontrol
the main to the individual
valve (MCV).actuators
Theseof the implement
allow structure
an intuitive and (boom,
precise
swing)
arm, and travel
bucket,
distribution swing)
of drive.
the andAtravel
incoming wheeled
pump machine
drive.
flow A to the with
wheeled a bucket
machine
individual attached
with a bucket
actuators asimplement
of the a tool is shown
attached a toolinisthe
asstructure figure,
shown
(boom,
butin various
arm,thebucket,other
figure, swing) arrangements
but various with tracks
other arrangements
and travel drive. A wheeled and other
withmachine specialised
tracks and other
with attachments,
specialised
a bucket such
attachments,
attached as hydraulic
as a tool issuch
shown as
hammers
hydraulic and scissors,
hammers
in the figure, canscissors,
and
but various be found.
other can be found.with tracks and other specialised attachments, such as
arrangements
hydraulic hammers and scissors, can be found.

Figure 1. Layout of a typical hydraulic excavator.

Figure 1. Layout of a typical hydraulic excavator.
Figure 1. Layout of a typical hydraulic excavator.
The flow of power through the machine and location of the individual losses are depicted in
The 2.
Figure flow
Theofconversion
power through of energythe machine one and location of the individual losses are depicted in
The flow of power through the from
machine formlocation
and into another,
of the as performed
individual by the
losses are engine
depictedand
in
Figure 2. The
pump(s), conversion
is the of energy
first source fromThe
of losses. one form into another, as throttling
performedlosses,
by theisengine and
Figure 2. The conversion of energy from onesecond,
form into referred
another,to as
as performed by a direct
the engine and
pump(s), is
consequencethe first source
of using of losses. The second, referred to as throttling losses, is a direct consequence
pump(s), is the first valves
sourcetoofdistribute hydraulic
losses. The second, power. Finally,
referred to asthethrottling
ancillary drives
losses,required for
is a direct
of vital
usingfunctions,
valves of
consequence tosuch
distribute
using asvalveshydraulic
steering, brakingpower.
to distribute and Finally,power.
cooling,
hydraulic the
are ancillary
Finally,drives
considered losses
the required
as they
ancillary for vital
consume
drives functions,
energy
required for
such
andas steering,
do not braking
directly and
contributecooling,
to the are considered
execution of thelosses as
required they consume
task. energy
vital functions, such as steering, braking and cooling, are considered losses as they consume energy and do not directly
contribute
and do notto the execution
directly of the
contribute to required task.of the required task.
the execution

Figure 2. The flow of power through the machine and the losses along the path.
Figure 2. The flow of power through the machine and the losses along the path.
Figure 2. The flow of power through the machine and the losses along the path.
Energies 2017, 10, 687 3 of 25

The exact reasons for these losses and their importance can be derived by analysing the individual
machine subsystems and their interactions with one another in more detail. Let us begin by discussing
the engine.

2.1. Engine
In theory, an engine can reach an efficiency of up to 100%, as the chemical input and mechanical
output energies are both ordered and of the same quality, i.e., have the same exergy [3,4]. Unfortunately,
converting chemical energy directly into mechanical work proves difficult and the fuel must first
be ignited in an unconstrained chemical reaction generating heat and losses. As a result, engines
specifically designed and optimised to work at a specific operating point can push their peak efficiency
into a region of 45–50% but the typical units found in construction machinery display maximum
efficiencies of just over 40% [5]. This value changes considerably depending on the load, i.e., torque,
and speed at which the engine operates. An exemplary internal combustion engine (ICE) efficiency
map is shown in Figure 3a.
The map is bounded by the full load line Tmax (n), which corresponds to the maximum torque or
power the ICE can generate at each rotation speed. If the ICE is operated along this line, no additional
acceleration torque is available. Load torques above this value lead to a deceleration of the output
shaft. Only above the minimum rotation speed nmin of about 800 rpm can an engine generate enough
mechanical power to overcome its internal losses. Typical units provide a substantial acceleration
torque at rotation speeds larger than 1000 rpm [6]. As the speed increases and more power can be
generated, the efficiency starts improving. However, higher speeds also lead to increased friction
losses and allow less time for the reactants to mix during the combustion process. As a result, efficiency
first increases and then begins decreasing with speed. The so-called sweet spot in the contour plot is
therefore usually located in the lower to mid-speed range just below the full load line.
Efficiency maps are characterised by curves of constant power, hyperbolas. Not every power
curve passes through the optimal operating region, indicating that the unit can only function efficiently
within a certain power range. When selecting an engine for a specific application, in this case an
excavator, the designer must take into account the peak power demand of the hydraulic system driven
by the engine. In most engines, the rotation speed nPowMax , at which maximum power is available, is
considerably higher than the speed nopt at which maximum efficiency is attained. In order to deliver
peak power and thereby avoid having to change the engine speed during a working cycle, standard
excavators are frequently operated at a fixed engine speed, namely nPowMax . During such a cycle,
the load pressure and pump displacement will vary constantly, leading to fluctuations in the load
torque shown in grey in Figure 3a. This results in frequent part loading and therefore inefficient engine
utilisation [7].
Thinking in terms of absolute efficiency can be misleading, as fuel consumption is, in fact,
the quantity that really matters. The Willans approximation, illustrated in Figure 3b, is another method
of plotting engine performance and reveals some interesting aspects that are not so evident from the
.
contour plotted efficiency map [8]. The fuel consumption V Diesel of an engine at a constant speed
increases approximately linearly with its output power. For most diesel engines, this proportionality
factor takes on a value of about 0.22 L/kWh [9]. In other words, each additional kW of output power
results in the same increase in the fuel consumption rate, regardless of the current operating point.
What this actually means is that the engine’s differential efficiency is actually a constant with a value
of 42%. The Willans lines also show that an engine delivering no power still consumes fuel due to its
.
parasitic losses. This is referred to as the idle fuel consumption V 0 and largely depends on the engine
size VICE , the number of pistons/cylinders and rotation speed nICE . In general, larger engines and
higher rotation speeds lead to higher parasitic losses and therefore an increase in idle fuel consumption.
Energies 2017, 10, 687 4 of 25
Energies 2017, 10, 687 4 of 25

Figure 3. (a) Typical diesel engine efficiency map; and (b) Willans approximation of fuel consumption
Figure 3. (a) Typical diesel engine efficiency map; and (b) Willans approximation of fuel consumption
adapted from [8].
adapted from [8].

According to Rohde-Brandenburger, in the case of a turbocharged diesel engine, the Willans

Accordingcan
approximation to Rohde-Brandenburger,
be expressed as followsin [8]:the case of a turbocharged diesel engine, the Willans
approximation can be expressed as follows [8]:
VDiesel l h = V0 l h + 0.22 l kWh · PICE kW (1)
. l . l l
with V Diesel = V0 + 0.22 · PICE [kW] (1)
h h kWh

with VICE, 0 = VICE · (8·10-8 · nICE 2 + 7.5·10-5 · nICE + 0.1) (2)

.
= VICE
V ICE,0 why
This is in fact the reason cars 10−8 ·engine
·(8·with 7.5·10−5 ·nsystems
nICE 2 +start–stop ICE + 0.1reduce
) (2)
fuel consumption.
Similarly, this explains the real advantage of the concepts referred to as engine downsizing and
This is in fact the reason why cars with engine start–stop systems reduce fuel consumption.
downspeeding. It is not due to the widely cited claim that the thermodynamic efficiency of the engine
Similarly, this explains the real advantage of the concepts referred to as engine downsizing and
improves when operating at higher loads and smaller displacements, but rather that the idle fuel
downspeeding. It is not due to the widely cited claim that the thermodynamic efficiency of the engine
consumption of a downsized or downsped engine is lower [9,10]. This concept reveals invaluable
improves when operating at higher loads and smaller displacements, but rather that the idle fuel
information regarding how much the fuel consumption of a vehicle can be improved without making
consumption of a downsized or downsped engine is lower [9,10]. This concept reveals invaluable
changes to the engine. For example, in the case of an average passenger vehicle with a four cylinder
information regarding how much the fuel consumption of a vehicle can be improved without making
engine, the idle fuel consumption is responsible for up to 50% of the total fuel consumption in the
changes to the engine. For example, in the case of an average passenger vehicle with a four cylinder
new European driving cycle (NEDC). This would mean that the maximum fuel reduction achievable,
engine, the idle fuel consumption is responsible for up to 50% of the total fuel consumption in the
by designing a new vehicle with theoretically zero mass and zero wind resistance but without any
new European driving cycle (NEDC). This would mean that the maximum fuel reduction achievable,
changes to the engine and gearbox, is only 50% [9]. This highlights the importance of using smaller
by designing a new vehicle with theoretically zero mass and zero wind resistance but without any
engines operating at lower speeds.
changes to the engine and gearbox, is only 50% [9]. This highlights the importance of using smaller
These observations have important consequences for the energy management strategies in
engines operating at lower speeds.
hybrid vehicles. When an engine is already running, the decision to charge the battery or accumulator
These observations have important consequences for the energy management strategies in hybrid
using the engine can be made without regarding the current operating point because the differential
vehicles. When an engine is already running, the decision to charge the battery or accumulator using
efficiency is constant [11]. Only when the engine is off does the decision to charge lead to additional
the engine can be made without regarding the current operating point because the differential efficiency
losses due to idling.
is constant [11]. Only when the engine is off does the decision to charge lead to additional losses due
Reducing engine losses has not been one of the main concerns of excavator manufacturers in
to idling.
recent years. In the late 20th century, studies proved that the particulate matter (PM), carbon
Reducing engine losses has not been one of the main concerns of excavator manufacturers in
monoxide (CO) and nitrous oxides (NOX) formed during the combustion of diesel fuel, are all
recent years. In the late 20th century, studies proved that the particulate matter (PM), carbon monoxide
detrimental to human health [12–14]. In an attempt to improve the air quality in cities and towns,
(CO) and nitrous oxides (NOX ) formed during the combustion of diesel fuel, are all detrimental to
governments introduced a set of stringent diesel emission standards in the 1990s. The initial TIER 1–
human health [12–14]. In an attempt to improve the air quality in cities and towns, governments
3 standards (in Europe Stage I–III), phased in between 1998 and 2008, were met using advanced
introduced a set of stringent diesel emission standards in the 1990s. The initial TIER 1–3 standards
engine designs and basically no or very simple exhaust gas after-treatment methods. With the
(in Europe Stage I–III), phased in between 1998 and 2008, were met using advanced engine designs
introduction of the TIER 4 interim and final (European Stage IIIB and IV) standards, which were
and basically no or very simple exhaust gas after-treatment methods. With the introduction of the
phased in from 2008 to 2015, manufacturers first agreed to decrease PM and then NOX emissions by
TIER 4 interim and final (European Stage IIIB and IV) standards, which were phased in from 2008 to
a further 90%. Meeting these standards was considerably more difficult.
Today’s excavators and other off-highway machines use expensive after-treatment systems with
diesel oxidation catalysts (DOC) and diesel particulate filters (DPF). This has led to increased costs
Energies 2017, 10, 687 5 of 25

and a further reduction in the available installation space around the engine. As explained by Filla,
2015,
the waymanufacturers first agreed
in which emissions tests to
aredecrease
currently PM and then has
conducted NOalso
X emissions
given OEMsby a further
(Original90%. Meeting
Equipment
these standards was considerably more difficult.
Manufacturers) little or no incentive to introduce efficient downspeed engines with hybrid
Today’s [15].
technologies excavators and other
The official non-roadoff-highway
transient machines use expensive
cycle (NRTC), after-treatment
used to evaluate systems
engine emissions,
with diesel oxidation catalysts (DOC) and diesel particulate filters (DPF).
uses a predetermined set of engine speeds and torques, which may not even match the actual engineThis has led to increased
costs and in
operation a further reduction
a machine. in thea available
As a result, machine with installation spaceengine
an advanced aroundmanagement
the engine. system
As explained
would
by Filla, the way in which emissions tests are currently conducted
be no different to a standard unit in regard to the current engine emissions regulations.has also given OEMs (Original
Equipment
All theManufacturers)
aspects mentioned littleabove
or no relate
incentive to introduce
to quasi-static efficient
engine downspeedAs
performance. engines
shownwith hybrid
in Figure 4
technologies
the [15]. Theengine
effect of dynamic officialloading
non-road transient
should cyclecannot
not and (NRTC), used to evaluate
be neglected. In any engine
typical emissions,
excavator
uses acycle,
duty predetermined
the engine setloadoffluctuates
engine speedsrapidlyand torques,
and which
the engine maybenot
must ableeven match the
to respond actualtoengine
quickly these
operation in a machine. As a result, a machine with an advanced
load changes to avoid stalling or excessive drops in rotation speed [16]. engine management system would
be noDepending
different toon a standard
the extentunit in regard
of these to the current
transients, engine emissions
fuel consumption regulations.
can fluctuate considerably from
the quasi-static descriptions in Figure 3. Research has, in fact, shown that up to As
All the aspects mentioned above relate to quasi-static engine performance. 50% shown in Figure
of emissions are4
the effect of dynamic engine loading should not and cannot be neglected.
caused by transient loads [17,18]. As a result, newer TIER 4 engines do not have the same responseIn any typical excavator
duty the
that cycle, the engine
older TIER 3load fluctuates
engines had. rapidly
This is and the engine
especially mustatbelower
evident able toengine
respond quickly
speeds andto these
is an
load changes to avoid stalling or excessive drops in rotation
important issue that will affect the design of newer mobile hydraulic systems.speed [16].

Figure
Figure 4.
4. Measured
Measured engine
engine load
load and
and rotation
rotation speed during aa typical
speed during typical excavator
excavator cycle.
cycle.

The mechanical power generated by the engine, must now be distributed to the individual
Depending
hydraulic on the
actuators. Thisextent
takesof
usthese
to thetransients,
machine’sfuel consumption
hydraulic system.can fluctuate considerably from
the quasi-static descriptions in Figure 3. Research has, in fact, shown that up to 50% of emissions are
caused
2.2. by transient
Hydraulic Systemloads [17,18]. As a result, newer TIER 4 engines do not have the same response that
the older TIER 3 engines had. This is especially evident at lower engine speeds and is an important
issueThe
that way in which
will affect the hydraulic
the design of neweractuators interact systems.
mobile hydraulic with the external environmental load is
particularly
The mechanical power generated by the engine, mustdemands
complex. Not only are the force and velocity of each actuator
now be distributed to the completely
individual
different, they also vary independently of each other depending
hydraulic actuators. This takes us to the machine’s hydraulic system. on the operator’s commands. Some
actuators may require high force and low velocity (high pressure, low flow) while others require low
force and high
2.2. Hydraulic velocity (low pressure, high flow). Figure 5 illustrates the load situation for the
System
actuators making up the implement structure. The x-axis shows the flow required by each actuator
The way in which the hydraulic actuators interact with the external environmental load is
(QL ) and can be interpreted as the operator’s input to the system. The y-axis shows the force or load
particularly complex. Not only are the force and velocity demands of each actuator completely
pressure (pL ) experienced by the actuator and is a direct consequence of the surroundings, cf.
different, they also vary independently of each other depending on the operator’s commands. Some
Equations
actuators may(3)–(5).
require high force and low velocity (high pressure, low flow) while others require low
ARod
force and high velocity (low pressure, pLhigh flow).
= pPiston - Figure·5pillustrates the load situation for the actuators
(3)
APiston Rod
making up the implement structure. The x-axis shows the flow required by each actuator (QL ) and can
be interpreted as the operator’s input to the system. The y-axis shows the force or load pressure (pL )
QL = QPiston (4)
experienced by the actuator and is a direct consequence of the surroundings, cf. Equations (3)–(5).
PAct = FL ·x = pL APiston ·x = pL QPiston = pL QL (5)
ARod
In the case of the linear actuators, pL = the
pPiston
force− acting on· pthem (3)
APiston Rod is due to both the weight of the
attached structure and the external forces present during digging and other operations. Inertial forces
caused during acceleration play a less important role. Depending on the movements, each actuator
Energies 2017, 10, 687 6 of 25

QL = QPiston (4)
. .
PAct = FL · x = pL APiston · x = pL QPiston = pL QL (5)

In the case of the linear actuators, the force acting on them is due to both the weight of the
attached structure and the external forces present during digging and other operations. Inertial forces
Energies 2017, 10, 687 6 of 25
caused during acceleration play a less important role. Depending on the movements, each actuator
experiences
experiences either
either aa resistive
resistive force
force opposing
opposing its
its motion
motion (Quadrants
(Quadrants II and
and III)
III) or
or an
an assistive
assistive force
force
aiding
aiding its
its motion
motion (Quadrants
(Quadrants IIII and
and IV).
IV). Consequently,
Consequently, inin quadrants
quadrants II and
and III,
III, the
the actuator
actuator must
must bebe
actively supplied with power, while, in quadrants II and IV, the actuators can actually supply
actively supplied with power, while, in quadrants II and IV, the actuators can actually supply power power to
the system.
to the system.

Figure 5. Load quadrants experienced by actuators of the implement system.

Figure 5. Load quadrants experienced by actuators of the implement system.

Due to the kinematic arrangement and large weight of the implement structure, the boom
Due to
cylinders the kinematic
almost exclusivelyarrangement
operate in loadand large weight
quadrants of the
I and II. implement
In contrast, structure, the boom
the magnitude and
cylinders almost exclusively operate in load quadrants I and II. In contrast,
direction of the load acting on the arm and bucket cylinders varies a great deal, causing operation inthe magnitude and
direction of the loadThe
all four quadrants. acting on the motor
hydraulic arm and bucket
driving cylinders
the varies a great
swing experiences deal,
four causing
quadrant operationbut
operation, in
all
in contrast to the linear actuators the inertial forces dominate here, meaning that the load pressure in
four quadrants. The hydraulic motor driving the swing experiences four quadrant operation, but is
contrast to the linear actuators the inertial forces dominate here, meaning
mainly due to the acceleration of the superstructure and not caused by external forces. that the load pressure is
mainly duepoint
Each to theinacceleration of therepresents
the pL /QL plane superstructure
a stateandofnot caused by external
quasi-stationary forces. in which the
equilibrium,
Each point in the pL /QL plane represents a state of quasi-stationary equilibrium, in which the
pump flow rate is proportional to the operator’s joystick displacement, system pressure is determined
pump flow rate is proportional to the operator’s joystick displacement, system pressure is determined
by the load, and the engine torque and pump torque are equal. As the actuator operating points move
by the load, and the engine torque and pump torque are equal. As the actuator operating points move
through the different load quadrants, the system’s power demand changes. To maintain a stable
through the different load quadrants, the system’s power demand changes. To maintain a stable engine
engine speed, every change in demand must be closely followed by a change in supply. This can be
speed, every change in demand must be closely followed by a change in supply. This can be tricky for
tricky for two reasons.
two reasons.
Firstly, the engines in most machines cannot even deliver the same amount of power as can be
Firstly, the engines in most machines cannot even deliver the same amount of power as can be
demanded by the pump. This may sound strange, but has to do with dimensioning. The pump size
demanded by the pump. This may sound strange, but has to do with dimensioning. The pump size is
is selected to meet the maximum speed/flow rate Qmax requirements of all the actuators
selected to meet the maximum speed/flow rate ( Qmax ) requirements of all the actuators (implements
(implements + travel drive) when operating at the rated engine speed. System pressure is a function
+ travel drive) when operating at the rated engine speed. System pressure is a function of the load
of the load and can reach values up to pmax = 380 bar, but is typically around 200 bar. The pump’s
and can reach values up to pmax = 380 bar, but is typically around 200 bar. The pump’s corner power
corner power is, therefore, considerably
is, therefore, considerably greater than the powergreater than the power
required duringrequired
standardduring standard
operation. operation.
Installing an
Installing an engine with the same corner power capabilities as the pump in
engine with the same corner power capabilities as the pump in order to cover these infrequent cases order to cover these
infrequent
would cases would
be expensive andberequire
expensive
more and require
space. Asmore space.
a result, it isAsquite
a result, it is quite
common common
to find to find
machines, in
machines, in which the corner power of the pump is two or three times greater
which the corner power of the pump is two or three times greater than the maximum engine power. than the maximum
engine power. Full actuator speed can only be maintained for average system pressures around
180 bar, thereafter a power limiting controller swivels the pump back preventing the engine from
stalling.
Secondly, as the operator adjusts his joystick commands and the load changes, an actuator’s state
can move through the load plane rather rapidly. Unfortunately, most engines cannot react so quickly
to such rapid load changes and are prone to stalling. The pump controller and valves must assume
Energies 2017, 10, 687 7 of 25

Full actuator speed can only be maintained for average system pressures around 180 bar, thereafter a
power limiting controller swivels the pump back preventing the engine from stalling.
Secondly, as the operator adjusts his joystick commands and the load changes, an actuator’s state
can move through the load plane rather rapidly. Unfortunately, most engines cannot react so quickly
to such rapid load changes and are prone to stalling. The pump controller and valves must assume
this role, and ideally work in sync with the EECU, to ensure the dynamic power demand and supply
are well matched. Achieving a well-tuned engine-pump interface is one of the major challenges in
Energies 2017, 10, 687 7 of 25
today’s mobile hydraulic systems [19].
In
In contrast
contrastto tomost otherother
most mobilemobile
machines, excavators
machines, are capableare
excavators of rotating
capable their
of superstructure
rotating their
relative to the undercarriage by using the swing drive. The
superstructure relative to the undercarriage by using the swing drive. The way way in which the acceleration
in which and the
deceleration of the swing is regulated largely determines the machine’s
acceleration and deceleration of the swing is regulated largely determines the machine’sperformance. Not only is this
drive used approximately
performance. Not only is this60%drive
of theused
time,approximately
it is also critical
60%to of
safety because
the time, it isthe operator’s
also critical tofield of
safety
vision changes as the machine turns. As a result, swing motion must be precise
because the operator’s field of vision changes as the machine turns. As a result, swing motion must and have priority
over other and
be precise functions. Each OEM
have priority over has itsfunctions.
other own specificEachsolution,
OEM has to its
attain
ownthe required
specific swingtomotion
solution, attain
without affecting the other actuators negatively. The single circuit flow sharing
the required swing motion without affecting the other actuators negatively. The single circuit (SC-FS) system studied
flow
in this work,
sharing which
(SC-FS) also represents
system studied inthethis
mostwork,
commonly
whichusedalsohydraulic
representssystem
the for
mostwheeled excavators,
commonly used
is shown insystem
hydraulic Figurefor6. wheeled excavators, is shown in Figure 6.

Figure 6. Single Circuit Flow Sharing (SC-FS) circuit used in European wheeled excavators, including
Figure 6. Single Circuit Flow Sharing (SC-FS) circuit used in European wheeled excavators, including
downstream pressurecompensators
downstream pressure compensators (PCs)
(PCs) forfor load
load independent
independent control,
control, a torque
a torque controlled
controlled (TC)
(TC) swing
swing with upstream compensation and LS-pressure bypass on the
with upstream compensation and LS-pressure bypass on the boom (LSB ). boom (LSB).

The implement cylinders are controlled using downstream stream pressure compensated
The implement cylinders are controlled using downstream stream pressure compensated valves.
valves. Controlling the swing’s large inertia with such a flow controlled valve, causes high pressures
Controlling the swing’s large inertia with such a flow controlled valve, causes high pressures during
during acceleration as the load pressure is primarily determined by inertial forces resulting from
acceleration as the load pressure is primarily determined by inertial forces resulting from acceleration
acceleration and not by external load forces from the environment [20]. If the acceleration is not
and not by external load forces from the environment [20]. If the acceleration is not controlled
controlled appropriately, the system pressure can rise excessively leading to large pressure
appropriately, the system pressure can rise excessively leading to large pressure differences among
differences among the various actuators, which cause unnecessary throttling and forces the pump to
the various actuators, which cause unnecessary throttling and forces the pump to swivel back thereby
swivel back thereby delivering less flow in order to avoid overloading the engine [21]. Such issues
delivering less flow in order to avoid overloading the engine [21]. Such issues are detrimental to
are detrimental to performance. To overcome these issues, manufacturers have come up with various
performance.
solutions. To overcome these issues, manufacturers have come up with various solutions.
Instead
Instead of regulating the
of regulating the flow
flow rate to the
rate to the swing,
swing, so-called
so-called torque
torque control
control is
is used.
used. In
In these
these circuits,
circuits,
the
the operator’s joystick command determines the maximum load sensing pressure of the swing drive
operator’s joystick command determines the maximum load sensing pressure of the swing drive
and is used to directly regulate the swing torque and not the swing speed. To
and is used to directly regulate the swing torque and not the swing speed. To ensure the swing alwaysensure the swing
always
receivesreceives
exactlyexactly the it
the flow flow it demands,
demands, regardless
regardless of the
of the other
other actuatormovements,
actuator movements,an an upstream
upstream
pressure compensator with a lower spring pressure differential setting than the
pressure compensator with a lower spring pressure differential setting than the pump controller’s pump controller’s
∆p is used
LS is
∆pLS used [22].
[22]. This
This gives
gives thethe swing
swing priority
priority over
over the
the other
other actuators. This is
actuators. This is an
an important
important safety
safety
function as the operator’s field of view changes during swing operation and unexpected obstacles
may suddenly appear.
An additional feature of the circuit is the boom load sensing pressure bypass. During fast lifting
operations, the load sensing line is directly connected to the boom piston side using a bypass throttle,
Energies 2017, 10, 687 8 of 25

function as the operator’s field of view changes during swing operation and unexpected obstacles may
suddenly appear.
An2017,
Energies additional
10, 687 feature of the circuit is the boom load sensing pressure bypass. During fast lifting 8 of 25
operations, the load sensing line is directly connected to the boom piston side using a bypass throttle,
causingTo the
discuss
boomthe losses occurring
pressure within thepressure
to be the dominant hydraulic system,
signal wesent
that is must takepump.
to the a closer look
This at the
ensures
pumps,
that valvespressure
the swing and actuators.
does not exceed the boom pressure, which results in minimal throttling and
maximum pump flow, improving cycle times and efficiency [20]. Some manufacturers also offer an
2.2.1. Pump
energy Losses
efficient boom float function in which the boom down motion does not require any pump
flow [23]. These features
The hydraulic pump, represent
almost the state of the
exclusively art in
of the today’s
axial wheeled
piston type, isexcavators.
responsible for converting
To discuss the losses occurring within the hydraulic
the mechanical output power from the engine into hydraulic power, system, we inmust takeofa flow
the form closer
andlook at the
pressure.
pumps,
A typicalvalves
pump and actuators.
efficiency map for a constant rotation speed is shown in Figure 7a. Depending on the
displacement setting and pressure, the efficiency can vary from as low as 60% to peak values of up
2.2.1. Pump Losses
to 91% at higher displacement settings and pressure levels [24]. As with the engine, thinking in terms
The hydraulic
of efficiency can bepump, almost
misleading. exclusively
The leakage and of the axial piston type,
hydro-mechanical is responsible
losses do not changefor converting
depending
the
on mechanical outputsetting,
the displacement power from the engine
they actually into hydraulic
remain power,In
fairly constant. inreality,
the form of flow
only and pressure.
the output power
Achanges
typical leading
pump efficiency map
to a higher for a constant
efficiency. A pump rotation speed
operating at is showndisplacement
a higher in Figure 7a. Depending on the
actually consumes
displacement
more energy as setting and pressure,
its power output isthe efficiency can vary from as low as 60% to peak values of up to
higher.
91% at higher
The displacement
mismatched cornersettings
powersand of pressure
the engine levels
and [24].
pump Asalso
withcreate
the engine,
some thinking
additionalineffects.
terms
of efficiency can be misleading. The leakage and hydro-mechanical losses do not
Pump operation in the upper right hand region is not even possible, meaning that the pump is forcedchange depending
on the displacement
to operate setting, theysettings
at lower displacement actuallyand remain fairly constant.
efficiencies In reality,
when system only
pressure is the
high.output power
Studies have
changes leading
shown that to a higher
depending efficiency.
on the cycle theA pump
pumpsoperating at amobile
in a typical higherhydraulic
displacement actually
system consumes
are responsible
more energy as between
for dissipating its power10%output
andis15%
higher.
of the mechanical power supplied by the engine [25,26].

(a) (b)
1
Displacement Set ting [-]

Displacement
0,91 PPump,max act uator and
0,90 cont roller
PICE,max pLS
0,89

0,87
0,5
0,85

0,8
0,75
0,7 Damping
0 ... 1 Orifice
Standard
Operating
Region
0 Act ual
180 360
pump
Pressure [bar]

Figure7.7.(a)
Figure (a)Characteristic
Characteristic efficiency
efficiency map
map for
for aa200
200cc
ccaxial
axialpiston
piston unit
unit[24];
[24];and
and(b)
(b)aaschematic
schematicof
ofaa
pumpwith
pump withaaload
loadsensing
sensingpressure
pressurecontroller.
controller.

In the
The past couple
mismatched of years,
corner powersoneoffurther aspect
the engine andconcerning
pump alsopump create operation has become
some additional effects.known.
Pump
operation in the upper right hand region is not even possible, meaning that the pump is the
The values shown in typical pump efficiency maps are taken from measurements, in which pump
forced to
controller is inactive and the displacement actuator has been mechanically
operate at lower displacement settings and efficiencies when system pressure is high. Studies have locked thereby not
allowing
shown thatthe pump swash
depending on theplate
cycletothe
vibrate.
pumpsThese are notmobile
in a typical realistic boundary
hydraulic conditions
system and do for
are responsible not
represent how such units really operate in a machine. The pump’s
dissipating between 10% and 15% of the mechanical power supplied by the engine [25,26]. controller is, in fact, constantly
adjusting
In thethe swash
past couple plate and regulating
of years, one further theaspect
flow entering
concerningthe system.
pump operation has become known.
The values shown in typical pump efficiency maps are taken hydro-mechanical
This causes two additional loss mechanisms. First, the from measurements, controller
in whichusually has
the pump
various damping
controller orifices
is inactive and the(Figure 7b), which
displacement createhas
actuator additional leakage. The
been mechanically second
locked loss mechanism
thereby not allowingis
due to the dynamic high frequency oscillation of the swash plate [27]. As
the pump swash plate to vibrate. These are not realistic boundary conditions and do not represent the pump piston barrel
how such units really operate in a machine. The pump’s controller is, in fact, constantly adjustingplate
rotates, a strong vibrating torque acts on the swash plate. In the case of tests, in which the swash the
is mechanically locked, this vibrating torque
swash plate and regulating the flow entering the system. is counteracted by pump’s end stop. In reality, the pump
controller must feed the swash plate with pressure in order to balance these torque oscillations,
thereby increasing the energy consumed by the controller. Together these controller losses are by no
means negligible and can decrease efficiency by up to ten percentage points [27,28].
Just like in a diesel engine, the concept of efficiency can be a bit misleading. A variable
displacement pump operating in standby at near 0 displacement still consumes power. For example,
Energies 2017, 10, 687 9 of 25

This causes two additional loss mechanisms. First, the hydro-mechanical controller usually has
various damping orifices (Figure 7b), which create additional leakage. The second loss mechanism
is due to the dynamic high frequency oscillation of the swash plate [27]. As the pump piston barrel
rotates, a strong vibrating torque acts on the swash plate. In the case of tests, in which the swash plate
is mechanically locked, this vibrating torque is counteracted by pump’s end stop. In reality, the pump
controller must feed the swash plate with pressure in order to balance these torque oscillations, thereby
increasing the energy consumed by the controller. Together these controller losses are by no means
negligible and can decrease efficiency by up to ten percentage points [27,28].
Just like in a diesel engine, the concept of efficiency can be a bit misleading. A variable displacement
pump operating in standby at near 0 displacement still consumes power. For example, a 210 cc pump
operating at 1800 rpm and a standby pressure of 28 bar will consume around 4 kW of power. These
parasitic losses increase with rotation speed and cannot be neglected as they increase the machine’s
idle fuel consumption. A Willans representation of pump efficiency has yet to published, but would
surely be extremely valuable.

2.2.2. Valve Losses

The flow leaving the pump(s) is distributed to the actuators using valves. The fundamental
physics explaining the flow of fluid through a valve can be described using the orifice equation:

∆p
p
Q = KV y (6)

In order for a flow Q to pass through a valve a certain pressure difference ∆p, in other words
a driving force, must be present. The amount of pressure needed depends on the valve’s geometry,
described by the coefficient KV , and spool position y. In summary, for a valve to function, part of the
hydraulic power entering the valve must be dissipated as heat. These so-called pressure or throttling
losses can be expressed as follows
PThrottle = Q∆p (7)

In literature, there seems to be a number of misconceptions concerning throttling. Yes, due to

their hydraulic resistance valves will always generate losses when supplying flow to an actuator, but if
correctly sized, that is with a large enough value for KV , these losses can be kept low, down to only a
couple of bar. The extreme throttling losses attributed to valves and so often cited in literature, are due
to completely different reasons. These are worth mentioning.
The first and major cause is directly related to the nature of the hydraulic architecture. Multiple
actuator systems, in which valves are used to distribute flow, delivered by a single pump are a classic
example. Each actuator has its own independent pressure level, determined by the load it is currently
subjected to. In order to deliver flow to each actuator the pump must supply the system with a pressure
level greater than that of each actuator. Unfortunately, no matter how well the cylinder areas and
kinematics are designed, there will always be instances, during which the individual actuator pressure
levels are substantially different. This creates a complete mismatch between the pump pressure and
actuator pressures. Unfortunately, valves can only operate according to the orifice equation and are
not capable of performing a lossless transformation of the pump pressure down to a lower actuator
pressure level. Consequently, the already existing pressure difference is used to generate the required
flow, resulting in considerable losses, especially when the flow is large, see Equation (7).
Other than distributing flow, the valves must enable the operator to precisely regulate the
movement of an actuator in all four load quadrants. This is done by regulating both the flow to
the actuator using an inlet metering edge and the flow leaving the actuator using an outlet metering
edge. The inlet edge is needed to control the actuator speed as it determines the amount of flow
entering the actuator. The outlet edge, on the other hand, fulfils a different function. In quadrants
II and IV, it is used as a type of brake and prevents runaway loads. In quadrants I and III, it is
used to marginally throttle the flow leaving the actuator, thereby maintaining a pressure level of
Energies 2017, 10, 687 10 of 25

approximately 5 to 10 bar in the outlet chamber [29]. This ensures that both the inlet and outlet
chambers are always pressurized and act as springs, guaranteeing a higher natural frequency and
better system response [30]. For actuators undergoing four quadrant operation, finding an optimal
outlet geometry
Energies 2017, 10, 687 involves compromise. The outlet resistance must be large enough to prevent 10 ofthe
25
overrunning loads in quadrants II and IV, but preferably should not be too large to cause unnecessary
2.3. Ancillary
throttling Drives
in the other two quadrants I and III.
In summary,
Apart from the it is hydraulic
important system,
to differentiate between
the engine alsothe following
provides causes
power to of
a throttling:
number of smaller
subsystems,
• referred
Throttling acrosstothe
asinlet
ancillary
edgedrives,
in orderwhich basically
to supply only consume
the actuator energy and do not directly
with flow
contribute in the generation of useful work, but without them the machine could not function. A
• Throttling across the outlet edge to maintain controllability and prevent runaway loads
typical setup is shown schematically in Figure 8.
• Throttling to equalise mismatched supply and actuator pressures
These include:
2.3. Ancillary Drivesregulating the machine’s 24 V electrical system;
The alternator
 The
Apartairfrom
conditioning compressor
the hydraulic unit;
system, the engine also provides power to a number of smaller subsystems,
referred
A number of hydraulic gear pumps
to as ancillary drives, which basicallyproviding pressure
only consume for and
energy the joysticks as well
do not directly as for the
contribute in
steering, braking and cooling systems. It should be noted that crawler type excavators
the generation of useful work, but without them the machine could not function. A typical setup is do not
shownhave a separateinsteering
schematically Figure 8.system and therefore do not require a pump supplying steering
pressure.

Figure 8. Ancillary drives in a wheeled excavator.

Figure 8. Ancillary drives in a wheeled excavator.

Figure 9 shows measurements of engine output power in an excavator at idle for different engine
speedThese include:
settings. To keep all the components on the shaft, including the ancillary drives and main
pump,
• Therotating up to
alternator 30% of the
regulating theengine’s full24
machine’s output powersystem;
V electrical is consumed.
• The air conditioning compressor unit;
• A number of hydraulic gear pumps providing pressure for the joysticks as well as for the steering,
braking and cooling systems. It should be noted that crawler type excavators do not have a
separate steering system and therefore do not require a pump supplying steering pressure.

Figure 9 shows measurements of engine output power in an excavator at idle for different engine
speed settings. To keep all the components on the shaft, including the ancillary drives and main pump,
rotating up to 30% of the engine’s full output power is consumed.
In mathematical terms, the idle power demand can be described as a linear function of the
engine speed:
PIdle
Figure 9. Measured engine load and = m0 ·speed
rotation nICE during a typical excavator cycle. (8)

In order to ensure no damage occurs if the machine is operated at low speeds, the ancillary drives
In mathematical terms, the idle power demand can be described as a linear function of the engine
are dimensioned to be fully functional at lower engine speeds of 800 rpm. Therefore, the additional
speed:
ancillary power produced at higher engine speeds is totally unnecessary and can be considered as an
additional loss term. PIdle = m0 · nICE (8)
In order to ensure no damage occurs if the machine is operated at low speeds, the ancillary drives
are dimensioned to be fully functional at lower engine speeds of 800 rpm. Therefore, the additional
ancillary power produced at higher engine speeds is totally unnecessary and can be considered as an
additional loss term.
 Figure 8. Ancillary drives in a wheeled excavator.

Figure 9 shows measurements of engine output power in an excavator at idle for different engine
Energies 10, 687 To keep all the components on the shaft, including the ancillary drives and main
2017, settings.
speed 11 of 25
pump, rotating up to 30% of the engine’s full output power is consumed.

Figure 9. Measured engine load and rotation speed during a typical excavator cycle.
Figure 9. Measured engine load and rotation speed during a typical excavator cycle.
In mathematical terms, the idle power demand can be described as a linear function of the engine
3. Cycle Analysis
speed:
Energies 2017, 10, 687 11 of 25
Hydraulic excavators, especially in the case PIdleof
=mthose up to 25 tons, are used for a variety of (8)
0 · nICE different
tasks3.not only
Cycle for digging
Analysis and moving earth. As a result, engineers have to deal
In order to ensure no damage occurs if the machine is operated at low speeds, the ancillary drives with the complex
task ofarehaving to design
dimensioned to beafully
Hydraulic excavators,
machine without
functional
especially at the
in
knowledge
lower engine
case
ofup
speeds
of those
itsof
exact
to 800
future
rpm.
25 tons,
application.
Therefore,
are used for
Fortunately,
thea additional
variety of
collecting detailed
ancillary
different power information
tasks notproduced on fuel
only foratdigging
higher consumption
engine
and speedsearth.
moving and Asother
is totally important
aunnecessary
result, and system
engineers can states,
be considered
have to is becoming
deal with asthe
an
easieradditional
with the loss term.
introduction of fleet management systems. These allow companies
complex task of having to design a machine without knowledge of its exact future application. to track and record
the movements
Fortunately,as well as detailed
collecting operation of all their
information on machines [31]. Without
fuel consumption and othera doubt,
important thissystem
new technology
states,
is becoming
will help change easier
the way with the introduction
machines of fleetasmanagement
are designed, engineers learn systems.
howThese
to mineallowandcompanies to
extract valuable
track and
knowledge record
from thesethemassive
movements dataasquantities.
well as operation
In 2013,ofLiebherr
all their machines
published [31].
someWithout a doubt,results
interesting this for
new technology
a wheeled excavator will help change
obtained the way
using their LiDatmachines
systemare designed,
[32]. as engineers
These showed learn
that the how to spent
machine mine 25%
and extract valuable knowledge from these massive data quantities. In 2013, Liebherr published some
of its operating time idling, 15% travelling between worksites and the remaining 60% performing actual
interesting results for a wheeled excavator obtained using their LiDat system [32]. These showed that
earth moving tasks. Using a similar fleet management system, researchers in Wuppertal collected
the machine spent 25% of its operating time idling, 15% travelling between worksites and the
data for 3733 machines
remaining over a period
60% performing of sixmoving
actual earth monthstasks.[33]. Using
For wheeled
a similarexcavators,
fleet managementthey found that, on
system,
average,
researchers in Wuppertal collected data for 3733 machines over a period of six months [33]. For and
these machines spent 30% of their operating time idling, 20% travelling, 10% grading
40% digging. Tracked excavators
wheeled excavators, they foundspent 15%
that, on at idle,
average, 10%machines
these travelling,spent15%
30% grading
of their and the remaining
operating time
60% of their20%
idling, operating time10%
travelling, digging.
grading and 40% digging. Tracked excavators spent 15% at idle, 10%
travelling, 15% grading and the remaining 60% of their operating time digging.
3.1. Characteristics of Typical Duty Cycles
3.1. Characteristics of Typical Duty Cycles
Although not exhaustive, a survey of research conducted in this field suggests that dig and
Althoughand
dump, trenching, notgrading
exhaustive, a survey
are the three of research
most commonconducted in thisperformed
duty cycles field suggests that dig and
by excavators [21,33].
dump, trenching, and grading are the three most common duty cycles performed by excavators
These are illustrated in Figure 10. An analysis of measurement data, collected from an 18 t excavator
[21,33]. These are illustrated in Figure 10. An analysis of measurement data, collected from an 18 t
performing these three cycles, shows some typical aspects of excavator operation, which can help
excavator performing these three cycles, shows some typical aspects of excavator operation, which
explain the main cause of losses and identify potential for improvement.
can help explain the main cause of losses and identify potential for improvement.

Figure 10. Typical excavator duty cycle.

Figure 10. Typical excavator duty cycle.
The first important characteristic to take note of is the large fluctuation in actuator power
demand for all three
The first important cycles. Tableto1take
characteristic summarises
note of isthese fluctuations
the large usinginthree
fluctuation parameters:
actuator zP
power demand
for alldescribes the ratio
three cycles. of the1 mean
Table and maximum
summarises actuator power
these fluctuations demand
using during
three the cycle, zzPQ describes
parameters: describes the
the ratio of the mean and maximum pump flow rate during the cycle and finally zp describes the ratio
of the mean and maximum pump pressure during the cycle. The peak power required for each of the
cycles is at least five greater than the mean power output. Similarly, the peak values for the pump
flow rate and pressure are approximately two times greater than the mean values.

Table 1. Comparison of average and maximum power, flow and pressure for each cycle.
Energies 2017, 10, 687 12 of 25

ratio of the mean and maximum actuator power demand during the cycle, zQ describes the ratio of the
mean and maximum pump flow rate during the cycle and finally zp describes the ratio of the mean
and maximum pump pressure during the cycle. The peak power required for each of the cycles is at
least five greater than the mean power output. Similarly, the peak values for the pump flow rate and
pressure are approximately two times greater than the mean values.

Table 1. Comparison of average and maximum power, flow and pressure for each cycle.

Cycle zP [−] zQ [−] zp [−]

Dig & Dump 0.20 0.60 0.51
Trenching 0.18 0.50 0.50
Grading 0.08 0.45 0.44

These observations should come as no surprise because a machine like an excavator can, in fact,
Energies 2017, 10, 687 12 of 25
only perform cyclic tasks. Every time the boom is raised, it must eventually be lowered. Similarly,
every acceleration of the swing
These observations shoulddrive
come must be followed
as no surprise becauseby its deceleration.
a machine Everycan,
like an excavator period of high
in fact,
poweronly
demand is followed by a period in which energy is no longer required
perform cyclic tasks. Every time the boom is raised, it must eventually be lowered. Similarly, or can actually be
everyfrom
recovered acceleration of the swing
the actuator. drive exception
The only must be followed
to this by
ruleits isdeceleration.
travelling Every periodduring
operation, of highwhich
power
constant power demand is followed
is required for aby a period in
prolonged whichofenergy
period is noalllonger
time. For otherrequired
cycles, theor can
poweractually
demand be will
recovered from
fluctuate considerably. the actuator. The only exception to this rule is travelling operation, during which
constant power is required for a prolonged period of time. For all other cycles, the power demand
An important consequence of the fluctuating power is that all the components in the system must
will fluctuate considerably.
be designed Anfor peak requirements
important consequence of butthewill spend most
fluctuating powerofistheir time
that all theactually
components operating at part load.
in the system
As will be discussed
must be designedlater on in
for peak this section,
requirements butpart
will loading
spend most is associated
of their timewith lowoperating
actually efficiency at values
part for
both the engine
load. As willand pump. later on in this section, part loading is associated with low efficiency values
be discussed
The second
for both property,
the engine and typical
pump. for multi-actuator hydraulic systems, is illustrated for each cycle
The second property,
and actuator in Figure 11. Due typical
to thefor kinematic
multi-actuator hydraulic systems,
arrangement of the isexcavator
illustratedarm
for each
the cycle
load force
and actuator in Figure 11. Due to the kinematic arrangement of
acting on an actuator not only varies during a duty cycle but is also different compared tothe excavator arm the load force
the load
acting on an actuator not only varies during a duty cycle but is also different compared to the load
force experienced by the other actuators. This means that each actuator has its own unique power
force experienced by the other actuators. This means that each actuator has its own unique power
demand profile, which can be expressed in terms of load pressure and required flow rate ( p , QL ).
demand profile, which can be expressed in terms of load pressure and required flow rate pL , QL L.
As discussed previously,
As discussed using
previously, valves
using valvesto distribute
to distribute flow
flowfrom
froma asingle
singlepump
pumpto to multiple actuators with
multiple actuators
mismatched power profiles
with mismatched power is inefficient.
profiles is inefficient.

Figure 11. Load pressure over flow required for each actuator and cycle.
Figure 11. Load pressure over flow required for each actuator and cycle.
Before continuing, an important aspect must be mentioned. Yes, for the majority of applications
power demand does fluctuate considerably and the average engine output power is lower than its
rated power; nevertheless, a machine must be capable of supplying the rated power continuously for
prolonged periods of time. For example, when travelling at full speed (30–35 km/h) in a wheeled
excavator or for tasks requiring the use of a special attachment like a drill, during which a
continuously high pump flow rate is demanded. This must be considered when designing any new
Energies 2017, 10, 687 13 of 25

Before continuing, an important aspect must be mentioned. Yes, for the majority of applications
power demand does fluctuate considerably and the average engine output power is lower than its
rated power; nevertheless, a machine must be capable of supplying the rated power continuously for
prolonged periods of time. For example, when travelling at full speed (30–35 km/h) in a wheeled
excavator or for tasks requiring the use of a special attachment like a drill, during which a continuously
high pump flow rate is demanded. This must be considered when designing any new system.

3.2. Energy Recovery in Excavators

The energy available for recovery in an excavator is found in two forms, either gravitational
potential energy
Energies 2017, 10, 687 or mechanical kinetic energy. When lowering a body of mass m by a distance 13 of∆h
25
through a gravitational field of strength g the energy Epot , described by Equation (9), is released.
through
Similarly,a when
gravitational field mass
a body, with of strength
m andgrotational
the energy Epot ,J,described
inertia traveling by
at aEquation
speed v, (9), is released.
or rotating at a
Similarly, when a body,
speed ω, decelerates with mass
its kinetic m and
energy Ekin rotational
(Equation inertia
(10)) isJ,made
traveling at a speed v, or rotating at a
available.
speed ω, decelerates its kinetic energy Ekin (Equation (10)) is made available.
Epot = mg∆h (9)
Epot = mg∆h (9)
1 12 11 22
Ekin,=lin 2=mvmv, 2E,kin,rot
Ekin, lin Ekin,rot== 2Jω Jω (10)
(10)
2 2
It is
It is worth
worth taking
taking aa moment
moment to to appreciate
appreciate the the similarities
similarities and
and differences
differences in
in these
these two
two equations.
equations.
Both are linearly dependent on the mass or inertia. Similarly, a change in height affects the potential
Both are linearly dependent on the mass or inertia. Similarly, a change in height affects the potential
energy linearly, but a change in speed affects the kinetic energy quadratically. Consequently,
energy linearly, but a change in speed affects the kinetic energy quadratically. Consequently, changes changes
in mass or height have less of an influence compared to
in mass or height have less of an influence compared to changes in speed.changes in speed.
Due to
Due to its
its substantial mass and
substantial mass and large
large changes
changes in in height,
height, the the boom
boom structure
structure has has the
the greatest
greatest
amount of recoverable potential energy, Figure 12a. On the other hand,
amount of recoverable potential energy, Figure 12a. On the other hand, due to the low speed of due to the low speed of the
the
centre of mass its kinetic energy is negligible. Similarly, potential energy can be
centre of mass its kinetic energy is negligible. Similarly, potential energy can be recovered from the recovered from the
arm and
arm and bucket
bucket actuator,
actuator,but
butconsiderably
considerablyless lesscompared
compared toto
thetheboom.
boom.The swing
The swing drives possesses
drives no
possesses
potential energy, but due to its large rotational inertia J, a considerable amount
no potential energy, but due to its large rotational inertia J, a considerable amount of kinetic energy of kinetic energy is
generated as the superstructure is accelerated (Figure
is generated as the superstructure is accelerated (Figure 12b). 12b).

Figure 12. (a) Boom potential energy; and (b) swing kinetic energy.
Figure 12. (a) Boom potential energy; and (b) swing kinetic energy.

If the energy released during lowering and braking cannot be reused immediately or stored, it
mustIfbethe energy released
dissipated and leave during lowering
the system and
in the braking
form cannot
of heat. In a be reusedvalve
standard immediately or machine,
controlled stored, it
must
the be dissipated
recoverable andisleave
energy first the system into
converted in the form of heat.
hydraulic power,In in
a standard
the form valve controlled
of flow machine,
Q and pressure p,
the recoverable energy is first converted into hydraulic power, in the form of flow Q and pressure
which is then throttled to tank pressure thereby generating heat and causing an unnecessary increase p,
which is then throttled
in oil temperature to tank
(Figure 13). pressure thereby generating heat and causing an unnecessary increase
in oil temperature (Figure 13).
 If the energy released during lowering and braking cannot be reused immediately or stored, it
must be dissipated and leave the system in the form of heat. In a standard valve controlled machine,
the recoverable energy is first converted into hydraulic power, in the form of flow Q and pressure p,
which is
Energies then
2017, 10, throttled
687 to tank pressure thereby generating heat and causing an unnecessary increase
14 of 25
in oil temperature (Figure 13).

Figure 13. (a) Energy dissipation during: boom lowering; and (b) swing braking.
Figure 13. (a) Energy dissipation during: boom lowering; and (b) swing braking.

Numerous papers and patents discussing ways to recover boom and swing energy have been
published in recent years [34–39].
Unfortunately, few of these publications answer the following four fundamental questions:

1. Which actuator has the highest amount of recoverable energy?

2. Should this energy be stored or can it be reused immediately?
3. How difficult, or in other words feasible, is it to recover energy from each of these actuators?
4. How much of the energy sent to the various actuators can be recovered?

It is difficult to give a universally valid answer to all four questions. The amount of recoverable
energy depends on the duty cycle and the size of the excavator. To accurately determine these quantities
it would be necessary to take a range of differently sized machines and conduct extensive tests. Such a
study would be invaluable but has yet to be published.
Table 2 summarises measurement data regarding recoverable energy for an 18 t excavator.
Four different cycles were analysed. To begin with a simple 90◦ and then 180◦ swing acceleration and
braking motion, followed by a 90◦ and 180◦ dig and dump cycle. The data reveals some interesting
facts. During a 90◦ swing motion, the machine reaches only seventy per cent of the maximum swing
rotational speed as the operator is forced to already start braking around the 60◦ mark in order to
come to a stop after 90◦ . During a 180◦ motion, the full swing speed is reached. Due to the quadratic
relationship between speed and energy (Equation (9)), a 90◦ rotation has approximately only half as
much kinetic energy as a 180◦ rotation. Furthermore, the potential energy x that can be recovered when
lowering the boom in a typical dig and dump cycle (D&D) is twice as large as the kinetic energy of
the swing at full speed. Despite this, the swing should not be disregarded because, during each D&D
cycle, it is accelerated and decelerated twice while the boom is only lowered once. Therefore, in the
case of a 180◦ D&D cycle, the swing and boom have the same amount of recoverable energy. For the
same cycle, measurements show that ten times less energy can be recovered from the arm actuator.

Table 2. Recoverable actuator energy for an 18 t excavator for various duty cycles compared to the
recoverable energy x of the boom drive in a typical dig and dump cycle.

Swing 90◦ Swing 180◦ Boom D&D Arm D&D Swing 90◦ D&D Swing 180◦ D&D
0.25 x 0.5 x x 0.1 x 0.5 x x

For larger machines the amount of boom and swing energy will increase, but it is difficult to
judge how their ratio is affected. In the case of the boom, the weight, mg, and the height, ∆h, of
the attachment structure increase with the size of the machine. This is not the case for the swing
Energies 2017, 10, 687 15 of 25

drive as the rotational inertia, J, increases with operating weight but the maximum rotational speed,
ω, decreases.
Figure 14 answers the second question. In a D&D cycle, during swing braking and boom lowering
no other actuator requires this amount of power at exactly the same moment. Therefore, the only way
to take advantage of the recoverable energy is to store it and then reuse it at a later stage. This takes us
to the third question, feasibility. In general, storing energy is not a problem, but the ability to reuse it
efficiently and cost effectively determines whether recovery is feasible. If the energy is to be stored
hydraulically, two factors are important. First, the design and sizing of the storage device is much
Energies 2017, 10, 687 15 of 25
simpler if the actuator pressure during recovery is fairly constant and known beforehand. Secondly,
if the recoveryhigher.
considerably pressure level is greaterdespite
Consequently, than thethe
pressure level during
large amount the reuse energy
of potential phase, the
thatstored
can oil
be
can be sent directly to the actuator, making things easier.
recovered, reusing it is far more difficult and costly in comparison to the swing drive.

Figure 14. Positive and negative work during dig and dump cycle.
Figure 14. Positive and negative work during dig and dump cycle.

Finally, to answer the last question, what portion of the energy delivered to the actuator can be
This explains
recovered? If this why
ratio recovering swing
is small, then it energy
does not is feasible.
make sense The topressure
invest during
in suchbraking is not only
technologies. An
constant but is also higher than the maximum pressure during acceleration.
integration of the areas under the curves in Figure 14 reveals that approximately half of the energy As a result, brake energy
can easily
used be storedand
to accelerate andmove
usedthe to assist the acceleration
actuators can in fact be phase. The boom
recovered. Put is far more complex.
differently, Due to
if all the energy
the kinematic arrangement and wide range of lowering speeds, the pressure
from the actuators could be recovered and reused without any losses, the engine would only have level during boom down
to
operation is not constant. Furthermore, the pressure level during boom
supply half as much energy. We can conclude that an excavator is very well suited for systems with up operation is usually
considerably
integrated higher. circuits.
recovery Consequently, despite the large amount of potential energy that can be recovered,
reusing it is far more
In summary, thedifficult and costlyconditions
typical operating in comparison can beto described
the swing as drive.
follows [40]:
Finally, to answer the last question, what portion of the energy delivered to the actuator can
1. Average power
be recovered? If thisrequirements
ratio is small,are then
considerably
it does not lowermakethansense
peak power
to investrequirements.
in such technologies.
An integration of the areas under the curves in Figure 14 reveals that approximatelydecelerating).
2. Actuator power can be positive (lifting, accelerating) or negative (lowering, The
half of the energy
peak negative power can reach levels similar to the rated engine power.
used to accelerate and move the actuators can in fact be recovered. Put differently, if all the energy
3.
fromDemand profiles
the actuators (pressure
could and flow
be recovered and rate) of allwithout
reused actuators anyvary independently
losses, of eachonly
the engine would other. Some
have to
actuators require high pressure and low flow rate, while others may
supply half as much energy. We can conclude that an excavator is very well suited for systems with require a low pressure and
high flow
integrated rate. circuits.
recovery
4. Idling is common.
In summary, the typical operating conditions can be described as follows [40]:
Now that the basic concepts regarding how an excavator works and the tasks it performs are
1. Average power requirements are considerably lower than peak power requirements.
clear, a couple of definitions and methods to compare losses can be introduced.
2. Actuator power can be positive (lifting, accelerating) or negative (lowering, decelerating).
The
4. Defining peak negativeand
Efficiency power can reach levels
Quantifying Lossessimilar to the rated engine power.
3. Demand profiles (pressure and flow rate) of all actuators vary independently of each other. Some
Figure 15 shows a Sankey plot of the energy flow for a 90° dig and dump cycle. The results are
actuators require high pressure and low flow rate, while others may require a low pressure and
comparable with those published by Sturm for a similar machine [25]. As expected, more than 60%
high flow rate.
of the energy contained in the fuel is lost through the combustion process. A further 11.7% consumed
4. Idling is common.
by the ancillary drives and pump. Approximately half the energy supplied by the pump, EPump , is
lost through
Now thatthrottling
the basic and otherregarding
concepts hydraulichowlosses, such as pipe
an excavator worksfriction.
and theIntasks
total,itonly aboutare
performs 12% of
clear,
the total input energy is delivered to the actuators,
a couple of definitions and methods to compare losses EAct,can . Half this energy
pos be introduced. actually performs useful
work on the objects in the surroundings, while the other half EAct,neg is used to raise and accelerate
the implement structure. Instead of recovering this portion during lowering and braking actions, it
is throttled to tank and is dissipated as heat.
In summary, the diagram reveals that approximately a third of the energy supplied by the engine
is consumed by the ancillary drives and pump, a third is lost through throttling in the hydraulic
system and only the last third actually reaches the actuators.
Energies 2017, 10, 687 16 of 25

4. Defining Efficiency and Quantifying Losses

Figure 15 shows a Sankey plot of the energy flow for a 90◦ dig and dump cycle. The results are
comparable with those published by Sturm for a similar machine [25]. As expected, more than 60% of
the energy contained in the fuel is lost through the combustion process. A further 11.7% consumed
by the ancillary drives and pump. Approximately half the energy supplied by the pump, EPump , is
lost through throttling and other hydraulic losses, such as pipe friction. In total, only about 12% of
the total input energy is delivered to the actuators, EAct, pos . Half this energy actually performs useful
work on the objects in the surroundings, while the other half EAct,neg is used to raise and accelerate
the implement structure. Instead of recovering this portion during lowering and braking actions, it is
throttled to 10,
Energies 2017, tank687and is dissipated as heat. 16 of 25

Figure 15. Sankey

Figure 15. Sankey plot
plot of
of energy
energy flow through an
flow through an 18
18 tt excavator
excavator during
during 90
90°◦ dig
dig and
and dump.
dump.

4.1. System Efficiency

In summary, the diagram reveals that approximately a third of the energy supplied by the engine
Using the
is consumed bydiagram, the drives
the ancillary hydraulic
andsystem
pump, aefficiency is defined
third is lost throughas:throttling in the hydraulic system
and only the last third actually reaches the actuators. EAct, pos
ηHyd = (11)
EPump
4.1. System Efficiency
For the dig and dump cycle a hydraulic efficiency of approximately 49% is given, which is
Using the diagram, the hydraulic system efficiency is defined as:
actually a very good value for a valve controlled hydraulic system with four actuators and only one
supplying pump. The hydraulic system efficiencyEAct, depends
pos on the number of pumps used in the
ηHyd = (11)
system and how the valves are used to distribute the
EPump flow among the actuator. Consequently, a
typical single circuit Load Sensing system has an efficiency of up to 50%, while dual circuit positive
Forsystems
control the dig and
can dump cycle a hydraulic
reach efficiencies aroundefficiency
60%. of approximately 49% is given, which is actually
a very good value for a valve controlled hydraulic
Total system efficiency can be defined in two differentsystem with ways.four actuators
If we and
are only only oneinsupplying
interested knowing
pump. The hydraulic system efficiency depends on the number of pumps
how much of the fuels energy entering the system, EDiesel , is actually delivered to used in the system and how
the actuators
the
EAct,valves are used to distribute the flow among the actuator. Consequently, a typical single circuit
pos , then the total gross efficiency is the relevant quantity:
Load Sensing system has an efficiency of up to 50%, while dual circuit positive control systems can
EAct, pos
reach efficiencies around 60%. ηTot, Gross = (12)
EDiesel
Total system efficiency can be defined in two different ways. If we are only interested in knowing
If the fact
how much that
of the energy
fuels can
energy be recovered
entering and reused
the system, EDiesel , is considered, then the
actually delivered total
to the net efficiency
actuators is,
EAct, pos
relevant:
then the total gross efficiency is the relevant quantity:
EAct, pos - EAct,neg
EAct, pos EAct, net
ηTot, Netη= = = (13)
(12)
Tot, GrossEDiesel
EDiesel EDiesel
Defining system efficiency in these terms conceals the fact that a machine still consumes fuel
when just idling because the individual efficiencies would be calculated to 0% with EAct, pos = 0
disregarding the amount of EDiesel . Instead of discussing efficiencies solely, an evaluation of absolute
losses and fuel consumption is more expedient.
Energies 2017, 10, 687 17 of 25

If the fact that energy can be recovered and reused is considered, then the total net efficiency
is relevant:
EAct, pos − EAct,neg E
ηTot, Net = = Act, net (13)
EDiesel EDiesel
Defining system efficiency in these terms conceals the fact that a machine still consumes fuel when
just idling because the individual efficiencies would be calculated to 0% with EAct, pos = 0 disregarding
the amount of EDiesel . Instead of discussing efficiencies solely, an evaluation of absolute losses and fuel
consumption is more expedient.

4.2. Absolute Losses and Fuel Consumption

A far more useful approximation of a machine’s fuel consumption can be derived by combining
the Willans approximation (Equation (1)) with the idle power demand relation (Equation (8)) and
the definition
Energies of hydraulic efficiency (Equation (11)). Because the idle pump losses are already
2017, 10, 687 17 of 25
included in the term PIdle , it is necessary to introduce the differential pump efficiency ∆ηPump , which
displacement
describes howsetting, measurements
much extra engine power show must that be∆η cantobe
supplied
Pump
theroughly
pump to approximated using ahydraulic
produce additional constant
power. Although
value of 0.9. the pump’s differential efficiency changes slightly depending on the pressure and
displacement setting, measurements show that ∆ηPump can be roughly approximated using a constant
value of 0.9. PPump, AVE
VDiesel l h = VICE, 0 + 0.22 · PICE = VICE,0 + 0.22 · PIdle +
. h i . . ∆ηPump P 
V Diesel hl = V ICE, 0 + 0.22 · PICE = V ICE, 0 + 0.22· PIdle + Pump, AVE
∆ηPump
. h i .  PAct,pos, AVEP 
VDieselVlDiesel l Act,pos, AVE
= V = V
+ 0.22 ·
ICE, 0 + m 0.22
n · +m n
0 ICE + ∆ηPump ηHyd
(14)
(14)
.
h h h ICE,0
i .
0 ICE
∆ηPump ηHyd
V Diesel hl = (V ICE, 0 + 0.22 ·m0 nICE ) + ∆ηPump 0.22
ηHyd · PAct,pos, AVE
l 0.22
VDiesel h = VICE,0 + 0.22 · m0 nICE + ·PAct,pos, AVE
To show how useful the equation is, consider an 18∆η t Pump ηHyd with
excavator a six-litre diesel engine
.
operating at 1800 rpm as an example. V ICE, 0 , determined using Equation (2), is approximately 2.9 L/h
To show how useful the equation is, consider an 18 t excavator with a six-litre diesel engine
and a typical value for m0 is 0.0167 kW/rpm. Assuming an average hydraulic system efficiency of
operating at 1800 rpm as an example. VICE,0 , determined using Equation (2), is approximately 2.9 L/h
between 30% and 60% results in the following relation
and a typical value for m0 is 0.0167 kW/rpm. Assuming an average hydraulic system efficiency of
between 30% and 60% results in the following . relation  l 
9.5 + 0.4074 · PAct,pos, AVE < V Diesel, 1800 rpm < 9.5 + 0.8148 · PAct,pos, AVE (15)
9.5 + 0.4074 · PAct,pos, AVE < VDiesel, 1800 rpm l hh < 9.5 + 0.8148 · PAct,pos, AVE (15)
Figure 16
Figure 16 shows
shows this
this inequality
inequality graphically.
graphically.

Figure 16. Fuel consumption model (engine speed 1800 rpm).

Figure 16. Fuel consumption model (engine speed 1800 rpm).

This means that even when just idling and conducting no work, the machine will still consume
9.5 L of fuel per hour. The total amount of fuel consumed due to this phenomenon will be referred to
as the fixed fuel consumption VFixed .

VFixed = VDiesel dt (16)

In the case of a grading cycle with a low average actuator power demand of around 10 kW the
fuel consumption only increases by 4.89 L/h compared to idle. This additional amount of fuel
Energies 2017, 10, 687 18 of 25

This means that even when just idling and conducting no work, the machine will still consume
9.5 L of fuel per hour. The total amount of fuel consumed due to this phenomenon will be referred to
as the fixed fuel consumption VFixed .
Z .
VFixed = V Diesel dt (16)

In the case of a grading cycle with a low average actuator power demand of around 10 kW the fuel
consumption only increases by 4.89 L/h compared to idle. This additional amount of fuel consumed
to actually perform the cycle will be referred to as the variable fuel consumption VVariable . The variable
fuel consumption is not affected by the engine operation as the engine’s differential efficiency is
constant according to Willans. As shown in Figure 17 only for cycles with average actuator power
demands
Energies greater
2017, 10, 687 than 20 kW does the variable fuel consumption surpass the fixed fuel consumption.
18 of 25
Energies 2017, 10, 687 18 of 25

Figure 17. Concept of fixed and variable fuel consumption.

Figure 17.
Figure Concept of
17. Concept of fixed
fixed and
and variable
variable fuel
fuel consumption.
consumption.

4.3. Model Validation

4.3. Model
4.3. Model Validation
Validation
The very simplified model behind Equations (14) and (15) does not consider dynamic effects as
The very
The very simplified
simplifiedmodel
modelbehind
behind Equations(14) (14)and
and(15)
(15) does not consider dynamic effects
frequent transient loading conditions,Equations
which further does
increase not
fuel consider dynamic
consumption. effects
Despite as
this,
as frequent
frequent transient
transient loading
loading conditions,which
conditions, whichfurther
furtherincrease
increase fuel
fuel consumption.
consumption. Despite
Despite this,
this,
measurements conducted on a real machine correlate well with the predicted theoretical values.
measurements conducted
measurements conducted on on aa real
real machine
machine correlate
correlate well
well with
with the
the predicted
predicted theoretical
theoretical values.
values.
Figure 18 proves the good comparability of the measured fuel rate and the introduced fuel
Figure
Figure 18 proves
18 proves the good comparability of the measured fuel rate and the introduced fuel consumption
consumption modelthe good comparability
(Equation of the measured
(14)) with a hydraulic efficiency offuel
49%,rate and was
which the calculated
introduced forfuel
the
model (Equation
consumption (14))(Equation
model with a hydraulic
(14)) efficiency
with a of 49%,
hydraulic which of
efficiency was calculated
49%, which for the
was dig andfor
calculated dumpthe
dig and dump cycle in Section 4.2.
cycle
dig in Section
and 4.2. in Section 4.2.
dump cycle

Figure 18. Measurements of load sensing (LS) system compared to fuel consumption model.
Figure 18. Measurements of load sensing (LS) system compared to fuel consumption model.
Figure 18. Measurements of load sensing (LS) system compared to fuel consumption model.
Fuel rate fluctuations around the idealized Willans line are caused by a varying hydraulic
Fuel rate
efficiency fluctuations
during around
the cycle and the idealized
the mentioned Willans
transient line are
loading caused
of the by whereby
engine, a varying hydraulic
more fuel is
efficiency during the cycle and the mentioned transient loading of the engine, whereby more
consumed to prevent the engine from stalling. Averaged values of fuel rate and positive actuator fuel is
consumed
power for theto prevent the engine
dig and dump cyclefrom
were stalling. Averaged
calculated values
and are also of fuel
plotted rate and
in Figure 18.positive
The lowactuator
average
power for the dig and dump cycle were calculated and are also plotted in Figure
power demand compared to the maximum power demand becomes apparent. This analysis 18. The low average
was
power demand compared to the maximum power demand becomes apparent. This analysis was
Energies 2017, 10, 687 19 of 25

Fuel rate fluctuations around the idealized Willans line are caused by a varying hydraulic
efficiency during the cycle and the mentioned transient loading of the engine, whereby more fuel
is consumed to prevent the engine from stalling. Averaged values of fuel rate and positive actuator
power for the dig and dump cycle were calculated and are also plotted in Figure 18. The low average
power demand compared to the maximum power demand becomes apparent. This analysis was done
for a variety
Energies 2017, 10,of687
different load cycles (see Figure 19). 19 of 25

Figure 19. Validation of fuel consumption model.

Figure 19. Validation of fuel consumption model.

Most of the cycles reach hydraulic efficiencies in the range of 30–50%. In contrast to that the air
Most
grading of the
cycle, whichcycles reach hydraulic
simulates a commonefficiencies
grading cycle in the range of
(cf. Figure 10)30–50%.
but withoutIn contrast to that
any external the
loads,
air grading
reveals cycle, which
a hydraulic simulates
efficiency a common
of 22% and a grading cycleefficiency
total gross (cf. Figure of10)
5%.butThis
without
resultanyis external
a direct
loads, reveals a hydraulic efficiency of 22% and a total gross efficiency
consequence of the single circuit hydraulic system used in wheeled excavators. At one stage of 5%. This result is aduring
direct
consequence of the single circuit hydraulic system used in wheeled excavators.
this cycle the boom is raised and simultaneously the arm cylinder is extended rapidly. Hence the At one stage during
this cycle
system the boom
pressure is by
is set raised
the and
boom’s simultaneously
piston pressure, the arm
which cylinder is extended
is considerably rapidly.
higher Hence the
(approximately
system pressure is set by the boom’s piston pressure, which is considerably
110 bar) than the required pressure to supply the arm cylinder. In combination with a high flow higher (approximately
110 bar) than
demand of thethe
arm required pressure 200
(approximately to supply
L/min) the arm cylinder.
throttling losses ofInaboutcombination
35 kW can with
be aobserved
high flow at
demand of the arm (approximately 200 L/min) throttling losses of about 35
the inlet edge of the arm control valve. In the case of a real grading cycle the load and accordingly kW can be observed at the
inlet edge of the
the pressure, whicharmiscontrol
neededvalve. In thethe
to supply case
arm of drive,
a real grading
increasescycle the load
leading and throttling
to lower accordingly the
losses
pressure,
and finally which is needed
to a higher to supply
hydraulic the armofdrive,
efficiency 30%.increases
Based onleading to lower
these results, anthrottling
averagedlosses and
hydraulic
finally to a higher hydraulic efficiency of
efficiency of 40% is assumed hereafter, see Equation 17.30%. Based on these results, an averaged hydraulic efficiency
of 40% is assumed hereafter, see Equation 17.
VDiesel, 1800 rpm, η = 0,4 l h = 9.5 + 0.6111 · PAct,pos, AVE (17)
Hyd
. l
V Diesel, 1800 rpm,ηHyd = 0,4 = 9.5 + 0.6111 · PAct,pos, AVE (17)
h
5. Predicting Fuel Consumption Improvements
5. Predicting Fuel Consumption Improvements
Now that the loss mechanisms occurring in an excavator are clear, we can take the next step and
Nowways
consider that the loss mechanisms
to reduce fuel consumptionoccurringandinto anpredict
excavator the are clear,
level we can take the
of improvement next
with thestep
helpand
of
consider ways to reduce fuel
this simple model using Willans lines. consumption and to predict the level of improvement with the help of
this simple model using Willans lines.
5.1. Saving Fuel by Lowering Hydraulic Losses
5.1. Saving Fuel by Lowering Hydraulic Losses
The majority of research in the field of energy efficient excavators has dealt with the
The majority
development of research
of methods in the hydraulic
to lower field of energy
losses. efficient
A number excavators has dealt with
of publications have the development
investigated the
of methods to lower hydraulic losses. A number of publications have investigated
ability of different oils to improve hydraulic efficiency [1,41,42]. However, what most of these studies the ability of
different oils to improve hydraulic efficiency [1,41,42]. However,
have failed to discuss is the direct impact of these changes to the hydraulic systems on fuel what most of these studies have
failed to discuss
consumption. is theEquation
Using direct impact of these
(18), the changes to thereduction
fuel consumption hydraulicofsystems
a machine on fuel consumption.
operating with a
Using Equation (18), the fuel consumption
theoretically lossless hydraulic system can be studied: reduction of a machine operating with a theoretically
lossless hydraulic system can be studied:
VDiesel, 1800rpm, η = 1 l h = 9.5 + 0.2444 · PAct,pos, AVE (18)
Hyd

Figure 20 reveals that even if it were possible to design a system capable of completely
eliminating throttling, fuel consumption could be reduced by no more than 30% for a dig and dump
cycle.
Energies 2017, 10, 687 20 of 25

 
. l
V Diesel, 1800 rpm,ηHyd = 1 = 9.5 + 0.2444 · PAct,pos, AVE (18)
h

Figure 20 reveals that even if it were possible to design a system capable of completely eliminating
Energies 2017, 10, 687 20 of 25
throttling, fuel consumption could be reduced by no more than 30% for a dig and dump cycle. 20 of 25
Energies 2017, 10, 687

Figure 20. Maximum theoretical fuel consumption reduction with a lossless hydraulic system.
Figure 20. Maximum theoretical fuel consumption reduction with a lossless hydraulic system.
Figure 20. Maximum theoretical fuel consumption reduction with a lossless hydraulic system.
For other cycles with a lower average positive actuator power this theoretical value is even
lower. For other cycles with a lower average positive actuator power this theoretical value is even
lower. Forhydraulic
A other cycles with capable
circuit a lower average positive
of reducing actuatorlosses
throttling powerto this
a theoretical
minimum value is even lower.
is a displacement
A hydraulic
A hydraulic
controlled circuit
system. circuit capable
Insteadcapable of reducing
of usingof reducing
valves throttling losses
throttling
to control actuator to a
losses minimum is
to aa single
motion, minimuma displacement unitcontrolled
is a displacement
hydraulic supplies
system.
controlled Instead
system. of using
Instead valves
of using to control
valves to actuator
control motion,
actuator a single
motion, a hydraulic
single
each actuator with the required power individually. The actuators velocity is controlled by regulating unit
hydraulic supplies
unit each
supplies
actuator
each with the required
actuatordisplacement.
the pump’s with the requiredpower individually.
power individually.
Unfortunately, The actuators
The actuators
the considerable velocity
higher is
velocity controlled
costs as is controlledby regulating
well as thebylow the
regulating
system
pump’s
the
damping displacement.
pump’s Unfortunately,
displacement.
associated the
Unfortunately,
with these systems haveconsiderable
the prevented higher
considerable them costs
higher as well
fromcosts as the
beingasanwell low system
as the and
economic damping
low easy
system
to
associated
damping with
control alternative these
associatedto systems
with these
valve have prevented
systems have
controlled systems. them
prevented from
Another,them being an economic
frompromising,
more being an economicand easy
approachand to control
to easy to
reduce
alternative
control
throttling to valve
alternative
losses is controlled
independent systems.
to valve controlled
meteringAnother,
systems.
valves,more promising,
Another,
which more
decouple approach
promising,
the meter to reduceandthrottling
inapproach the to losses
reduce
meter out
is independent
throttling
edges. As losses metering
mentioned valves,
is independent
before the which
metering
outlet decouple
edge valves, thewhich
meterto
is necessary inmaintain
and thethe
decouple meter
meteroutinedges.
controllability andand As even
the mentioned
metermore
out
before
edges.
importantthe outlet
As mentioned edge is
to preventbefore necessary
runaway to
the outlet maintain
loads. edge controllability and
is necessary to controllable
An independently even more
maintain controllabilityimportant
meter out edge to
and even prevent
morea
enables
runaway loads.
important
reduction of An independently
tothrottling
prevent runaway controllable
losses to aloads.
minimum meter out on
An independently
depending edge enables
actual aload
controllable
the reduction
meter of edge
out
pressure throttling losses
of theenables a
actuator.
to a minimum
reduction of depending
throttling on
losses the
to a actual
minimum load pressure
depending of the
on actuator.
the actual
With such a system hydraulic efficiencies of about 70% can be achieved. However, as shown in With
load such
pressurea system
of the hydraulic
actuator.
efficiencies
With
Figure such aofsystem
21, the about 70%
reduction can
hydraulic beefficiencies
in fuel achieved.
consumption However,
of about
for the asdig
70% shown
can andbe in Figure
achieved.
dump 21,
cycle thethereby
However,
is reduction in fuel
as limited
shown in
to
consumption
Figure for
21, the 17%.
approximately the dig and dump cycle is thereby limited to approximately
reduction in fuel consumption for the dig and dump cycle is thereby limited to 17%.
approximately 17%.

Figure 21. Theoretical fuel consumption reduction of a hydraulic system with 70% efficiency.
Figure 21. Theoretical fuel consumption reduction of a hydraulic system with 70% efficiency.
Figure 21. Theoretical fuel consumption reduction of a hydraulic system with 70% efficiency.
Solutions solely focused on the hydraulic subsystem, without any consideration of the fixed fuel
Solutions are
consumption, solely focusedvery
therefore on the hydraulic subsystem, without any consideration of the fixed fuel
limited.
consumption, are therefore very limited.
5.2. Saving Fuel by Lowering Idle Losses
5.2. Saving Fuel by Lowering Idle Losses
Although it is not often mentioned explicitly, the concept of idle consumption is actually behind
manyAlthough it is
of the fuel not often
saving mentioned
methods explicitly,
discussed the concept
in literature. of idle
Engine consumptionengine
downspeeding, is actually behind
downsizing
many of the fuel
and start–stop savingall
systems methods
improve discussed
machine in literature.
efficiency byEngine downspeeding,
lowering idling losses. engine downsizing
Downspeeding not
Energies 2017, 10, 687 21 of 25

Energies 2017, 10, 687 21 of 25

Solutions solely focused on the hydraulic subsystem, without any consideration of the fixed fuel
consumption, are therefore very limited.
(Equation (2)), but also lowers the parasitic power demanded by the ancillary drives (Equation (8)).
For
5.2. example,
Saving Fuelusing the sameIdle
by Lowering simplified
Losses model from above, by reducing engine speed from 1800 rpm
down to 1200 rpm, the fuel consumption relation becomes:
Although it is not often mentioned explicitly, the concept of idle consumption is actually behind
VDiesel, l literature.
many of the fuel saving methods discussed
1200rpm in h = 6.2 + 0.489 · PAct,pos,
Engine downspeeding,
AVE (19)
engine downsizing
and start–stop systems all improve machine efficiency by lowering idling losses. Downspeeding
Figure 22 shows the relation graphically. Especially for cycles with low average power demands,
not only lowers the friction losses in the engine itself, which are a quadratic function of the speed
this measure is very effective, more so than any optimisation of the hydraulic system. The efficiency
(Equation (2)), but also lowers the parasitic power demanded by the ancillary drives (Equation (8)).
improvements resulting from downspeeding have been shown in numerous studies [20,43,44]. What
For example, using the same simplified model from above, by reducing engine speed from 1800 rpm
these studies have also shown is that a reduced engine speed also lowers productivity because the
down to 1200 rpm, the fuel consumption relation becomes:
engine can no longer supply the same amount of power as at the higher speed.
Downsizing, i.e., using. a smaller engine l with less power, is also discussed frequently, but is very
 
V Diesel, 1200rpm = 6.2 + 0.489 · PAct,pos, AVE (19)
difficult to implement in an excavator. For hduty cycles like travelling and drilling with an attachment,
in which full power is demanded for prolonged periods of time a downsized engine will just not have
Figure 22
the required showsAlthough
power. the relation
thesegraphically.
cycles areEspecially for than
less frequent cycles with low
others, average power
the machine must be demands,
capable
this measure is very effective, more so than any optimisation of the hydraulic
of performing them. Start–stop systems are particularly interesting, as the engine is completelysystem. The efficiency
improvements
switched resulting
off meaning from downspeeding
absolutely have This
no fuel is consumed. beenisshown in numerous
especially efficient forstudies [20,43,44].
cars in city traffic
What these studies have also shown is that a reduced engine speed also lowers productivity
that are frequently stopping at traffic lights. Implementing this technology in construction machinery because
theconsiderably
is engine can no longer
more supply the same amount of power as at the higher speed.
difficult.

Figure 22. Theoretical fuel consumption reduction possible with engine downspeeding.
Figure 22. Theoretical fuel consumption reduction possible with engine downspeeding.

Other methods to lower idle power demand, which are not directly related to the engine, include
Downsizing,
removing ancillaryi.e., using
drives a smaller
from engine
the engine withand
shaft lessusing
power, is also discussed
decentralised electricfrequently, but
drives, for is very
example
difficult
for to implement
the cooling fan pump. in an excavator. For duty cycles like travelling and drilling with an attachment,
in which full power is demanded for prolonged periods of time a downsized engine will just not
haveSaving
5.3. the required power.
Fuel through Although
Energy these cycles are less frequent than others, the machine must be
Recovery
capable of performing them. Start–stop systems are particularly interesting, as the engine is completely
Various methods of recovering and reusing both boom and swing energy have been proposed
switched off meaning absolutely no fuel is consumed. This is especially efficient for cars in city traffic
in literature and can now be found in various series machines available on the market [45]. An
that are frequently stopping at traffic lights. Implementing this technology in construction machinery
important issue to address is the potential of energy recovery methods to lower fuel consumption.
is considerably more difficult.
To do so the fuel consumption model is extended to include negative actuator power. If a recovery
Other methods to lower idle power demand, which are not directly related to the engine, include
system with an efficiency of ηRec were installed the engine would be required to supply less power
removing ancillary drives from the engine shaft and using decentralised electric drives, for example
and thecooling
for the averagefan
fuel consumption would be
pump.
VDiesel, 1800 rpm l h = 9.5 + 0.6111 · (PAct,pos, AVE - ηRec · PAct,neg, AVE ) (20)

Measurements show that the average negative actuator power is always less than or equal to
50% of the average positive actuator power. For cycles such as dig and dump the maximum of 50%
is reached but in the case of trenching this value goes down to approximately 30%. This can be
expressed with the following inequality:
Energies 2017, 10, 687 22 of 25

5.3. Saving Fuel through Energy Recovery

Various methods of recovering and reusing both boom and swing energy have been proposed in
literature and can now be found in various series machines available on the market [45]. An important
issue to address is the potential of energy recovery methods to lower fuel consumption. To do so the
fuel consumption model is extended to include negative actuator power. If a recovery system with an
efficiency of ηRec were installed the engine would be required to supply less power and the average
fuel consumption would be
 
. l
V Diesel, 1800 rpm = 9.5 + 0.6111 · ( PAct,pos, AVE − ηRec · PAct,neg, AVE ) (20)
h

Measurements show that the average negative actuator power is always less than or equal to
50% of the average positive actuator power. For cycles such as dig and dump the maximum of 50% is
reached but in the case of trenching this value goes down to approximately 30%. This can be expressed
with the following inequality:
Energies 2017, 10, 687 22 of 25
PAct,neg, AVE ≤ 0.5 · PAct,pos, AVE (21)

Inserting Equation (20) into EquationPAct,neg, (21),

AVE ≤assuming
0.5 · PAct,pos,
a AVE
recovery efficiency of 100%, gives(21)
the
minimum theoretical fuel consumption of a perfect recovery system, capable of recovering boom,
Inserting Equation (20) into Equation (21), assuming a recovery efficiency of 100%, gives the
swing and arm energy (see Figure 23).
minimum theoretical fuel consumption of a perfect recovery system, capable of recovering boom,
swing and arm energy (see l Figure 23).
 
.

V Diesel, 1800 rpm,ηrec = 1 = 9.5 + 0.6111 · 0.5 · PAct,pos,AVE = 9.5 + 0.3056 · PAct,pos,AVE (22)
h
VDiesel, 1800 rpm, η = 1 l h = 9.5 + 0.6111 · 0.5 · PAct,pos,AVE = 9.5 + 0.3056 · PAct,pos,AVE (22)
rec

Figure 23. Maximum theoretical fuel consumption reduction possible with energy recovery.
Figure 23. Maximum theoretical fuel consumption reduction possible with energy recovery.

5.4. Saving Fuel through Holistic Approach

5.4. Saving Fuel through Holistic Approach
The only way to drastically lower fuel consumption is to follow a holistic approach [40,46]. This
The only way to drastically lower fuel consumption is to follow a holistic approach [40,46].
means improving the hydraulics, enabling energy recovery and lowering idle fuel consumption. For
This means improving the hydraulics, enabling energy recovery and lowering idle fuel consumption.
a machine operating at 1200 rpm, with a lossless hydraulic system and optimal energy recovery, the
For a machine operating at 1200 rpm, with a lossless hydraulic system and optimal energy recovery,
minimum theoretical fuel consumption can be calculated as follows:
the minimum theoretical fuel consumption can be calculated as follows:
VDiesel, 1200 rpm, rec, min l h = 6.2 + 0.2444 · 0.5 · PAct,pos,AVE = 6.2 + 0.1222 · PAct,pos,AVE (23)
. l

V Diesel, 1200 rpm, rec, min = 6.2 + 0.2444 · 0.5 · PAct,pos,AVE = 6.2 + 0.1222 · PAct,pos,AVE (23)
As shown in Figure 24, h this results in both a parallel displacement and a change in gradient of
the consumption line. Theoretically, consumption can be reduced by up 59% for the dig and dump
As shown in Figure 24, this results in both a parallel displacement and a change in gradient of the
cycle. In practice, this value cannot be reached but it is helpful to have a theoretical limit.
consumption line. Theoretically, consumption can be reduced by up 59% for the dig and dump cycle.
In practice, this value cannot be reached but it is helpful to have a theoretical limit.
minimum theoretical fuel consumption can be calculated as follows:

VDiesel, 1200 rpm, rec, min l h = 6.2 + 0.2444 · 0.5 · PAct,pos,AVE = 6.2 + 0.1222 · PAct,pos,AVE (23)

As shown in Figure 24, this results in both a parallel displacement and a change in gradient of
the consumption
Energies 2017, 10, 687 line. Theoretically, consumption can be reduced by up 59% for the dig and dump
23 of 25
cycle. In practice, this value cannot be reached but it is helpful to have a theoretical limit.

Figure 24. Maximum theoretical fuel consumption reduction using engine downspeeding, lossless
Figure 24. Maximum theoretical fuel consumption reduction using engine downspeeding, lossless
hydraulics and energy recovery.
hydraulics and energy recovery.

6. Conclusions
6. Conclusions
A detailed examination of measurement data shows that the fuel consumption of an excavator
A detailed examination of measurement data shows that the fuel consumption of an excavator can
can be divided into a fixed and variable component. The fixed component quantifies the amount of
be divided into a fixed and variable component. The fixed component quantifies the amount of fuel
fuel the machine consumes if it was just left at idle and did not move for the entire cycle time. The
the machine consumes if it was just left at idle and did not move for the entire cycle time. The variable
variable component specifies the additional amount of fuel consumed in order to actually perform
component specifies the additional amount of fuel consumed in order to actually perform the task.
Depending on the cycle the ratio of fixed to variable fuel consumption changes. In cycles with low
average power demands, for example grading, the fixed consumption is actually greater than the
variable consumption. Consequently, lowering the total fuel consumption of a machine, requires a
consideration of both the fixed and variable terms.
The model introduced in this paper helps assess the potential of certain measures regarding
their ability to lower fuel consumption a priori. An interesting result is that a hydraulic system with
100% efficiency can only reduce the fuel consumption during grading by approximately 25%. This
is because improvements in the hydraulic system efficiency only lower the variable consumption,
not the fixed consumption. Similarly, a machine capable of recovering all available boom potential
and swing kinetic energy can theoretically only reduce fuel consumption by a maximum of around
30%. Consequently, thinking only in terms of hydraulic efficiency is rather misleading. A more holistic
approach, considering engine operation and other important factors, is far better suited.

Author Contributions: Milos Vukovic and Roland Leifeld performed the measurements and analysed the data.
All authors contributed to the writing of the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Abekawa, T.; Tanikawa, Y.; Hirosawa, A. Introduction of Komatsu Genuine Hydraulic Oil KOMHYDRO HE;
Komatsu Technical Report; Yumpu: Komatsu, Japan, 2010; Volume 56, No. 163.
2. Ohira, S.; Suehiro, M.; Ota, K.; Kawamura, K. Use of emission rights for construction machinery to help
prevent global warming. Hitachi Rev. 2013, 62, 123–130.
3. Bloomfield, L.A. How Things Work: The Physics of Everyday Life; John Wiley & Sons: Inc.: Hoboken, NJ,
USA, 2013.
4. Pragmatic Efficiency Limits for Internal Combustion Engines, Readout Guest Forum. 2014. Available
online: http://www.horiba.com/publications/readout/article/pragmatic-efficiency-limits-for-internal-
combustion-engines-31871/ (accessed on 7 October 2016).
5. Hanlon, M. Most Powerful Diesel Engine in the World, Gizmag. Available online: http://www.gizmag.
com/go/3263/ (accessed on 7 October 2016).
Energies 2017, 10, 687 24 of 25

6. Frei, B. Regelung Eines Elektromechanischen Getriebes für Hybridfahrzeuge. Ph.D. Thesis, Technical
University Chemnitz, Chemnitz, Germany, 2006.
7. Yang, J.; Quan, L.; Yang, Y. Excavator energy-saving efficiency based on diesel engine cylinder deactivation
technology. Chin. J. Mech. Eng. 2012, 25. [CrossRef]
8. Liebl, J.; Lederer, M.; Rohde-Brandenburger, K.; Biermann, J.-W.; Roth, M.; Schäfer, H. Energiemanagement im
Kraftfahrzeug; Springer: Wiesbaden, Germany, 2014.
9. Rohde-Brandenburger, K. Was bringen 100 kg Gewichtsreduzierung im Verbrauch? In ATZ; Springer:
Wiesbaden, Germany, 2013; Volume 115.
10. Guzzella, L.; Sciaretta, A. Vehicle Propulsion Systems; Springer: Berlin/Heidelberg, Germany, 2013.
11. Wilde, A. Eine Modulare Funktionsarchitektur für Adaptives Und Vorausschauendes Energiemanagement
in Hybridfahrzeugen. Ph.D. Thesis, Technical University Munich, München, Germany, 2008.
12. Smil, V. Energy: A Beginner’s Guide; Oneworld Publications: London, UK, 2006.
13. Chiatti, G.; Chiavola, O.; Falcucci, G. Soot Morphology Effects on DPF Performance; SAE Technical Paper
2009-01-1279; SAE: Warrendale, PA, USA, 2009. [CrossRef]
14. Falcucci, G. A lumped parameter model for diesel soot morphology evaluation and emission control.
Proc. Instit. Mech. Eng. Part D J. Automob. Eng. 2012, 226, 987–998. [CrossRef]
15. Filla, R. Getting Real about Test Cycles; SAE International Off-Highway Engineering: Warrendale, PA, USA,
2013; ISSN: 1528-9702.
16. Terdich, N. Impact of Electrically Assisted Turbocharging on the Transient Response of an Off-Highway
Diesel Engine. Ph.D. Thesis, Imperial College London, London, UK, 2015.
17. Hanson, P.A.; Lindgren, M.; Nordin, M.; Pettersson, O. A methodology for measuring the effects of transient
loads on the fuel efficiency of agricultural tractors. Appl. Eng. Agric. 2003, 19. [CrossRef]
18. Lindgren, M.; Hansson, P.-A. Effects of transient conditions on exhaust emissions from two non-road diesel
engines. Biosyst. Eng. 2004, 87, 57–66. [CrossRef]
19. Forschungsdialog: Erfolgsprojekt STEAM, O+P Ölhydraulik und Pneumatik 07/08; Vereinigte Fachverlage: Mainz,
Germany, 2016.
20. Kunze, G.; Göhring, H.; Jacob, K. Baumaschinen: Erdbau-Und Tagesbaumaschinen; Vieweg Verlag: Wiesbaden,
Germany, 2002.
21. Holländer, C. Untersuchungen zur Beurteilung und Optimierung von Baggerhydrauliksystemen.
Ph.D. Thesis, TU Braunschweig, Braunschweig, Germany, 1988.
22. Herfs, W. LUDV-Steuerungen für Bagger. Europäische Mobiltagung, Lohr a. Main, Mannesmann Rexroth AG;
Presentation (RD 00 207/10.97); Rexroth: Lohr, Germany, 1997.
23. N.N. Caterpillar M318D Wheeled Excavator Brochure. Available online: http://www.cat.com/
en_US/products/new/equipment/wheel-excavators/wheel-excavators/18115489.html (accessed on 7
October 2016).
24. Kawasaki K3VL Axial piston pumps for open circuits in mobile. In Industrial and Marine Applications;
Datasheet: Kawasaki, Japan, 2014.
25. Sturm, C. Bewertung der Energieeffizienz von Antriebssystemen mobiler Arbeitsmaschinen am Beispiel
Bagger. Ph.D. Thesis, Karlsruhe Institute of Technology, Karlsruhe, Germany, 2015.
26. Vukovic, M.; Sgro, S.; Murrenhoff, H. STEAM—Hydraulic design for engine integration. O + P : Fluidtechnik
für den Maschinen- und Anlagenbau 2015, 59. Available online: https://publications.rwth-aachen.de/record/
466125 (accessed on 29 March 2017).
27. Achten, P. Dynamic high-frequency behaviour of the swash plate in a variable displacement axial piston
pump. J. Syst. Control Eng. 2013, 227, 1–12. [CrossRef]
28. Lux, J.; Murrenhoff, H. Experimental loss analysis of displacement controlled pumps. In Proceedings of the
10th International Fluid Power Conference, Dresden, Germany, 8–10 March 2016.
29. Zhang, J. Pressure compensated metering valve for EH systems. In Proceedings of the National Conference
on Fluid Power, Las Vegas, NV, USA, 12–14 March 2008.
30. Murrenhoff, H. Servohydraulik—Geregelte Hydraulische Antriebe, 4th ed.; Shaker: Aachen, Germany, 2012.
31. LiDAT the Complete Comprehensive Data Collection and Tracking System from Liebherr; Liebherr Product Brochure;
Liebherr: Kirchdorf, Germany, 2016.
32. Bolz, G. Konzepte zur Erhöhung der Energieeffizienz von Mobilbaggern. In Proceedings of the 5th
Fachtagung Baumaschinentechnik, Dresden, Germany, 20–21 September 2012.
Energies 2017, 10, 687 25 of 25

33. Fecke, M. Standardisierung definierter Lastzyklen und Messmethoden zur Energieverbrauchsermittlung

von Baumaschinen. In Proceedings of the 5th Fachtgaung Baumaschinentechnik, Dresden, Germany,
17–19 September 2015.
34. Steindorff, K. Energierückgewinnung am Beispiel eines ventilgesteuerten hydraulischen Antriebs.
Ph.D. Thesis, TU Braunschweig, Braunschweig, Germany, 2010.
35. Sgro, S.; Murrenhoff, H. Energierückgewinnungssysteme für Baggerausleger—Eine Systematische Übersicht der
Vorhandenen Lösungsmöglichkeiten; O+P 10/2010; Vereinigte Fachverlage: Mainz, Germany, 2010.
36. Amrhein, J.; Neumann, U. PRB—Regeneration of potential energy while boom-down. In Proceedings of the
8th International Conference on Fluid Power, Dresden, Germany, 26–28 March 2012.
37. Thompson, B.; Yoon, H.; Kim, J.; Lee, J. Swing Energy Recuperation Scheme for Hydraulic Excavators;
SAE Technical Paper 2014-01-2402; SAE: Warrendale, PA, USA, 2014. [CrossRef]
38. Hießl, A.; Scheidl, R. Energy consumption and efficiency measurements of different excavators—Does
hybridization pay? In Proceedings of the ASME/BATH Symposium on Fluid Power and Motion Control,
Chicago, IL, USA, 12–14 October 2015.
39. Joo, C.; Stangl, M. Application of power regenerative boom system to excavator. In Proceedings of the 10th
International Conference on Fluid Power, Dresden, Germany, 8–10 March 2016.
40. Vukovic, M.; Sgro, S.; Murrenhoff, H. STEAM—A holistic approach to designing excavator systems.
In Proceedings of the 9th International Fluid Power Conference, Aachen, Germany, 24–26 March 2014.
41. Korane, K. How Hydraulic Fluids Affect Energy Efficiency, Mobile Hydraulic Tips. Available
online: http://www.mobilehydraulictips.com/how-hydraulic-fluids-affect-energy-efficiency/ (accessed on
7 October 2016).
42. Michael, P.W.; Mettakadapa, S. Bulk Modulus and Traction Effects in an Axial Piston Pump and a Radial
Piston Motor. In Proceedings of the 10th International Fluid Power Conference, Dresden, Germany, 8–10
March 2016.
43. Forche, J. Antriebsstrangmanagement eines Hydraulikbaggers. Ph.D. Thesis, Technical University
Braunschweig, Braunschweig, Germany, 2007.
44. Ng, F.; Harding, J.A.; Glass, J. An eco-approach to optimise efficiency and productivity of a hydraulic
excavator. J. Clean. Prod. 2016, 112, 3966–3976. [CrossRef]
45. Inderelst, M. Efficiency Improvements in Mobile Hydraulic Systems, Ph.D. Thesis, RWTH Aachen University,
Aachen, Germany, 2013.
46. Leifeld, R.; Vukovic, M.; Murrenhoff, H. Hydraulic hybrid architecture for excavators. ATZoffhighway Worldw.
2016, 9, 44–49. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
 Reproduced with permission of copyright owner.
Further reproduction prohibited without permission.