energies

Article
Achievement of Fuel Savings in Wheel Loader by
Applying Hydrodynamic Mechanical Power
Split Transmissions
Xiaojun Liu ID
, Dongye Sun *, Datong Qin and Junlong Liu
State Key Laboratory of Mechanical Transmissions & School of Automotive Engineering, Chongqing University,
Chongqing 400044, China; hschl@cqu.edu.cn (X.L.); dtqin@cqu.edu.cn (D.Q.); liujunlong12388@126.com (J.L.)
* Correspondence: dysun@cqu.edu.cn; Tel.: +86-188-83876783

Received: 5 July 2017; Accepted: 21 August 2017; Published: 25 August 2017

Abstract: The fuel economy of wheel loaders is deeply affected by the efficiency of their propelling
transmissions, however, the torque converter (TC) in existing propelling transmissions is a
low-efficiency component and leads to excessive energy consumption. Accordingly, this paper
replaces the TC with a hydrodynamic mechanical power split transmission (HMPST) for improving
the fuel economy of wheel loader. Based on probability similarity theory, the typical operating mode
for the vehicles is constructed, which is used to evaluate the energy consumption performance of the
selected solutions. The four reasonable solutions, which can be initially applied to wheel loaders,
are selected from among the HMPSTs using the lever diagram. Furthermore, the comparisons on
efficiency and loading characteristics between these four solutions and a prototype TC are conducted.
The design optimization for all the four solutions is implemented, in order to find the optimal fuel
saving solution relative to the prototype TC, and only one solution with pure power split can meet the
constraints. Finally, a simulation model of the wheel loader powertrain is established for validating
the effectiveness of this optimal solution. The results show that the optimized solution can effectively
improve the fuel economy of wheel loaders compared to the prototype TC and provides a novel
substitute for current technology.

Keywords: wheel loader; hydrodynamic mechanical power split transmissions (HMPSTs); torque
converter (TC); fuel economy; design optimization

1. Introduction
Wheel loaders are indispensable machines for the construction of infrastructure such as highways
and in coal mining. The wheel loader operation is pretty repetitive and the vehicle is frequently
exposed to large loads, leading to high fuel consumption [1]. Taking a five-ton class application as
an example, the total fuel consumption each year can be as much as 25.6 tons (assuming the working
times per day and per year are 8 hours and 200 days, respectively). Today the most common wheel
loader powertrain arrangement is as displayed in Figure 1 [2]. The engine is required to supply energy
to both the working hydraulics system and propelling transmission system, which allow the bucket to
be manipulated and thrust the vehicle forward or backward, respectively.
In this architecture, the torque converter (TC) is a key element since it has the excellent
vibration-damping and thoroughly self-adaptive performance which enable the engine to work
robustly even though the vehicle is frequently subjected to variable loads. However, owing to the
double energy conversion in the TC and high slip between two terminals that performance inevitably,
results in high losses. The low efficiency of the TC is the main reason for the excessive fuel losses of
wheel loaders [3]. With the increasing emphasis on fuel savings and emission reduction worldwide,

Energies 2017, 10, 1267; doi:10.3390/en10091267 www.mdpi.com/journal/energies

Energies 2017, 10, 1267 2 of 20

Energies 2017, 10, 1267 2 of 20

these devices may hardly meet the future development trends [4]. Therefore, a substitute for the TC
may hardly meet the future development trends [4]. Therefore, a substitute for the TC that improves
that improves the fuel economy of wheel loaders is essential.
the fuel economy of wheel loaders is essential.

Working Hydraulics System

Hydraulic HP
System
3
Wheel Wheel
TC
Engine
MG 2 1

Final Drive 4 Final Drive

Propelling Transmission System

Wheel Wheel

HP‒Hydraulic pump TC‒Torque converter MG‒Mechanical gearbox

Figure 1. Wheel loader powertrain architecture.

Figure 1. Wheel loader powertrain architecture.
The hydrodynamic mechanical power split transmission (HMPST) can be a desirable choice,
sincehydrodynamic
The in it a part of themechanical
power is transferred
power split by the mechanical (HMPST)
transmission branch andcan the berestaisdesirable
transferred choice,
through the TC [5,6]. With this special architecture, the HMPST
since in it a part of the power is transferred by the mechanical branch and the rest is transferredsimultaneously has a higher
efficiency
through the TCand retains
[5,6]. Withmostthis of the advantages
special architecture, of atheTC.HMPST
Furthermore, there is a possibility
simultaneously has a higher thatefficiency
the
HMPST can enable the engine to work in the economical areas due to a wide variety of configurations
and retains most of the advantages of a TC. Furthermore, there is a possibility that the HMPST can
available for HMPSTs. Therefore, in this paper, the HMPST is selected as the research subject.
enable the engine to work in the economical areas due to a wide variety of configurations available for
HMPSTs comprise a TC and a planetary gear train (PGT). Many references have examined the
HMPSTs. Therefore, in of
basic characteristics this paper, providing
HMPSTs, the HMPST is selected
comparative as the for
analyses research
different subject.
configurations, as well
HMPSTs
as the design methods for HMPSTs. Wang et al. elaborated the operational principle ofhave
comprise a TC and a planetary gear train (PGT). Many references a newexamined
HMPST the
basicincharacteristics
a roadheader’s of cutting
HMPSTs, providing
unit. A dynamic comparative
model foranalyses for differentwas
such a transmission configurations,
constructed. as Thewell as
the design methods
simulation forshowed
results HMPSTs. thatWang
both the et al. elaborated
vibration of thethe operational
drive motor and principle
the dynamicof a new HMPST in a
engagement
force of the
roadheader’s gears unit.
cutting are substantially
A dynamiclessened model for compared
such a to those of the was
transmission prototype transmission
constructed. when
The simulation
encountering pulse loads [7]. Jo et al. suggested a new input torque coupled
results showed that both the vibration of the drive motor and the dynamic engagement force of the HMPST concept, where
gearsthe
arehydraulic elements
substantially were placed
lessened compared at both to the mechanical
those and hydraulic
of the prototype branches. when
transmission To evaluate the
encountering
usefulness of the designed mechanism relative to a prototype TC (TC), the torque ratio and efficiency
pulse loads [7]. Jo et al. suggested a new input torque coupled HMPST concept, where the hydraulic
for all six schemes were analyzed using the performance locus diagram [8]. Linares et al. proposed a
elements were placed at both the mechanical and hydraulic branches. To evaluate the usefulness
thorough methodology for designing a continuously variable power split transmission with several
of thetypes
designed
of power mechanism
flow suitable relative to a prototype
for an agricultural tractor. TCThe(TC), the torque
objective ratioisand
of this study efficiency
to determine thefor all
six schemes were analyzed
design parameters usingtype
for a general theofperformance
this transmission locussuchdiagram [8]. Linares
as the transmission et of
ratios al.PGT,
proposed
and a
thorough methodology for designing a continuously variable power
the lockup point, which is the transition point between the hydromechanical transmission and thesplit transmission with several
typespurely
of power flow suitable
mechanical for an[9].
transmission agricultural
Achtenova tractor. The objective
et al. developed a softwareof thisforstudy is tohighly
designing determine
efficient
the design HMPSTs with
parameters for several
a general mechanical
type of thisspeeds, which aimed
transmission at obtaining
such continuouslyratios
as the transmission variable
of PGT,
transmissions
and the lockup point, with which
wide speed
is theratio ranges. Each
transition pointspeed was the
between thecombinations
hydromechanical of a few transmission
PGTs, and its and
mechanical
the purely efficiency
mechanical was calculated
transmission byAchtenova
[9]. the matrix method. The advantages
et al. developed of this code
a software are the high
for designing highly
integration and computational efficiency [10].
efficient HMPSTs with several mechanical speeds, which aimed at obtaining continuously variable
Other studies have concentrated on improving the performance of construction machinery using
transmissions with wide speed ratio ranges. Each speed was the combinations of a few PGTs, and its
some types of HMPSTs. Zhao et al. implemented the design optimization for the crawler bulldozer
mechanical
equipped efficiency
with a pure waspower
calculated
split typeby the matrixfor
of HMPST method.
increasingTheboth
advantages
the dynamic of this code are and
performance the high
integration and computational efficiency [10].
fuel economy of the vehicle under a typical operating mode. These two objectives are achieved by
Other studies
the smaller TC have concentrated
diameter on improving
and less vehicle the performance
quality. However, of construction
the TC adopted machinery
in the HMPST is notusing
somebased
typesonoftheHMPSTs.
prototypeZhao et al.is,implemented
TC. That the improvement theofdesign optimization
performance for the crawler
is not necessarily caused by bulldozer
the
powerwith
equipped splita structure
pure power [11].split
Wang type et of
al.HMPST
proposed fora increasing
power reflux bothhydrodynamic
the dynamic transmissions
performance and
(PRHTS) applied
fuel economy to the wheel
of the vehicle underloader.
a typicalFor constructing
operating mode. the optimal
Thesematching between
two objectives aretheachieved
diesel by
the smaller TC diameter and less vehicle quality. However, the TC adopted in the HMPSTonis not
engine and the PRHTS, the design optimization for the PRHTS was conducted, which was based
based on the prototype TC. That is, the improvement of performance is not necessarily caused by the
power split structure [11]. Wang et al. proposed a power reflux hydrodynamic transmissions (PRHTS)
Energies 2017, 10, 1267 3 of 20

applied to the wheel loader. For constructing the optimal matching between the diesel engine and the
PRHTS, the design optimization for the PRHTS was conducted, which was based on the radar chart
and genetic algorithm. However, a simulation model for the wheel loader powertrain system was not
constructed, so whether the performance of the vehicle is increased needs to be justified. Moreover,
such a transmission is only one type of HMPST and its rationality relative to other schemes requires to
be verified as well [12].
The objective of this paper is to improve the fuel economy of a wheel loader by applying the
HMPST concept, which is rarely researched in previous works. In view of this, a design methodology
for such a transmission is necessary. Moreover, considering the relatively complex construction
of HMPSTs, the simplest form, which consists of a TC with three components and a single PGT,
is adopted here.
In this paper, there are two critical issues to be dealt with. One is the preliminary selection of
the the HMPST options. Since HMPSTs have a number of different schemes, if a design optimization
for all the schemes is conducted, it is bound to cause a high calculation cost. The subsequent
comparison analysis between each scheme can be hardly justified. This issue will be solved in
Section 3. Furthermore, the TC in the HMPSTs has to be based on the prototype, i.e., only the effective
diameter of the TC can be altered. Only in this way a comparison between HMPSTs and a prototype
TC can make sense.
The other is the construction of an optimization model. Although the HMPST efficiency is the key
factor influencing the fuel consumption of a wheel loader, it is not the only one. The engine efficiency is
also an option. The ideal fuel saving solution provided by HMPSTs is that they have a higher efficiency
and can enable the engine to work in the most economical area. However, the HMPST is a self-adaptive
mechanism and thus it cannot actively control the engine to work in a specified areas. On the other
hand, the wheel loader operation is highly repetitive, that means the general operating mode of wheel
loaders can be predicted and modeled [3,13]. Considering these two aspects, this paper optimizes
HMPSTs using some features of the typical operating mode, in order to find the optimal fuel saving
potential solution for HMPSTs.
The rest of paper is arranged as follows: an operating mode for a wheel loader is constructed in
Section 2. Section 4 implements the design optimization for the selected HMPSTs, where the objective
function is the fuel saving potential relative to the prototype TC. After the results are obtained,
a simulation to determine the optimal solution is conducted in the Matlab/Simulink environment,
in order to verify the effectiveness of the constructed optimization model.

2. Operating Mode and Instantaneous Fuel Consumption

The total fuel consumption that a wheel loader expends in a task is the time integral of its
instantaneous fuel consumption. The instantaneous fuel consumption is the product of engine
output power, the fuel consumption rate of the engine and unit time [14]. In view of this, first,
an operating mode that presents the power requirements for the typical working manner of wheel
loader is constructed. Second, the variables which influence the instantaneous fuel consumption of
vehicle are analyzed. Third, the loading characteristics of the TC is derived, which is used to determine
the working points of the TC and engine under external loads.

2.1. Operating Mode

To evaluate the performance of a wheel loader such as the working time cost, loading
material volume and fuel consumption, a definite operating mode is required to describe the power
requirements for the typical working manner of the wheel loader [3,13]. The most common working
manner for a wheel loader is the V-type route that consists of five segments [15,16]. That is, the operator
implements the task repeatedly according to the V-type route. In the first segment, the vehicle drives
towards the material piles until it reaches the destination. During the second segment, the bucket is
inserted into the material piles by the great driving force provided by the propelling transmission
Energies 2017, 10, 1267 4 of 20

system, then the operator manipulates the bucket and boom in an almost synchronous and continuous
manner. After this process is repeated several times, the bucket can be filled up. In the third segment,
the vehicle returns back to the initial position. In the fourth segment, the operator turns the vehicle
towards the2017,
Energies transport
10, 1267 while lifting the bucket to a certain height. The unloading of the material 4 of 20 is
mainly completed by
Energies 2017, 10, 1267
the tilt action, which controls the rotation motion of the bucket. After the material
4 of 20
the material
is delivered is mainly
to the completed
transport, by the tilt
the vehicle goesaction,
backwhich controls
to the initialthe rotationaccording
position motion of the bucket.
to the route of
After the material is delivered to the transport, the vehicle goes back to the initial position according
the material
Segment is mainly completed
4. One complete procedurebyforthethis
tilt working
action, which controls
manner the rotation
is shown motion
in Figure 2. of the bucket.
to the route of Segment 4. One complete procedure for this working manner is shown in Figure 2.
After the material is delivered to the transport, the vehicle goes back to the initial position according
to the route of Segment 4. One complete procedure for this working manner is shown in Figure 2.

5
4
3

MaterialMaterial
4 2
Wheel
Loader 3

Piles Piles
2
Wheel
1
Loader

Figure 12. V-cycle of wheel loader.

Figure 2. V-cycle of wheel loader.
Figure 2. V-cycle of wheel loader.
A-five-ton class wheel loader is the objective vehicle in this study, as shown in Figure 3. The
A-five-ton class wheel
basic parameters of thisloader
vehicleisare
thelisted
objective vehicle
in Table 1. Forinobtaining
this study, theasreal-time
shown inworking
Figure state
3. The ofbasic
A-five-ton class wheel loader is the objective vehicle in this study, as shown in Figure 3. The
vehicle,of
parameters testing
this transducers
vehicle are are placed
listed in at the 1.
Table designated
For positions,
obtaining the which are indicated
real-time working bystate
the Arabic
of vehicle,
basic parameters of this vehicle are listed in Table 1. For obtaining the real-time working state of
numerals
testing in Figure
transducers are 1. The measurement
placed at variables of the transducers are the inputby and output angular
vehicle, testing transducers arethe designated
placed positions,
at the designated which are
positions, indicated
which the
are indicated Arabic
by numerals
the Arabic
speeds
in Figure of
1. Thethe TC, pressure of the hydraulic pump, rotation angles of the boom and bucket cylinders,
numerals in measurement variables ofvariables
Figure 1. The measurement the transducers are the input
of the transducers are theand output
input angular
and output speeds of
angular
and output angular speed of the mechanical gearbox. The vehicle is manipulated by an expert driver,
the TC, pressure
speeds of theof thepressure
TC, hydraulic pump,
of the rotation
hydraulic pump,angles of the
rotation boom
angles andboom
of the bucket
andcylinders, and output
bucket cylinders,
and its working manner is based on the V-type route shown in Figure 2. The objective material piles
angularandspeed
outputof angular speed of the gearbox.
the mechanical mechanicalThe gearbox.
vehicleTheisvehicle is manipulated
manipulated by anby an expert
expert driver,
driver, and its
are raw soil, and the working cycle times of the vehicle are fifty. Since the fifty sets of data collected
workingand its working
manner manner
is testing
based on is based on
the V-type the V-type
routemany route
shown shown
in Figurein Figure 2. The objective material piles
directly by the transducers contain noise signals2. andThe objective
also materialthe
cannot present piles are raw
power
are raw
soil, and soil, and the working cycle times of the vehicle are fifty. Since the fifty sets of data collected
requirements of the wheel loader, they have to be further converted into the operating mode of wheel by
the working cycle times of the vehicle are fifty. Since the fifty sets of data collected directly
directly by the testing transducers contain many noise signals and also cannot present the power
the testing
loader,transducers
which can becontain
used formany noise simulation
the off-line signals and andalso cannot present
evaluation the power
of the wheel loader. requirements of
requirements of the wheel loader, they have to be further converted into the operating mode of wheel
the wheel loader, they have to be further converted into the operating mode of wheel loader, which can
loader, which can be used for the off-line simulation and evaluation of the wheel loader.
be used for the off-line simulation and evaluation of the wheel loader.

Figure 3. Objective wheel loader.

Figure 3. Objective wheel loader.

Generally, a data processing Figure method3.based on probability
Objective similarity theory is used to construct
wheel loader.
the operating mode of wheel loaders due to the fact that it can extract the feature data from raw data
Generally, a data processing method based on probability similarity theory is used to construct
correctly while reproducing the energy consumption features in both the simulation environment
Generally,
the operating a data
modeprocessing method
of wheel loaders duebased on probability
to the fact similarity
that it can extract theory
the feature is from
data usedrawto construct
data
and experiments [17,18]. Therefore, the operating mode of a wheel loader, which consists of the
correctly while
the operating modereproducing
of wheel loadersthe energy consumption
due to the fact that features in both the
it can extract the simulation
feature data environment
from raw data
vehicle velocity vw , thrusting resistance Ft , whole vehicle mass mw , and load torque of hydraulic
and while
correctly experiments [17,18]. the
reproducing Therefore,
energy the operating mode
consumption of ainwheel
features loader,
both the which consists
simulation of the and
environment
pump
vehicle Tvelocity
h , is derived
v , through this
thrusting method.FMoreover,
resistance , whole both the
vehicle lateral
mass m dynamics
, and load oftorque
vehicle ofand wheel
hydraulic
experiments [17,18]. Therefore,w the operating t mode of a wheel loader, w which consists of the vehicle
slippage T are ignored in this derivation, since this study focuses on thedynamics
longitudinal dynamics of the
pump
velocity vw , thrusting resistance Ft , whole vehicle mass mw , and load torqueofofvehicle
h , is derived through this method. Moreover, both the lateral hydraulicand wheel
pump Th ,
vehicle. Similar assumptions can be found in [3,19]. Figure 4a–d present the typical operating mode
slippage
is derived are ignored
through this in this derivation,
method. Moreover, since
both this
the study focuses
lateral on theoflongitudinal
dynamics vehicle and dynamics
wheel of the are
slippage
of this wheel loader. It can be seen that based on the different working states of the wheel loader, the
vehicle.
ignored Similar assumptions can be found in [3,19].
on Figure 4a–d present the typical operating mode
loadinmodes
this derivation,
of its engine since
can this study
be classifiedfocuses
into three the longitudinal
types, which are the dynamics
light, theofmedium
the vehicle. Similar
and the
of this wheel loader. It can be seen that based on the different working states of the wheel loader, the
assumptions
heavy type. canAtbe thefound in [3,19].
light load mode, Figure
the engine4a–d present
outputs thepower,
a low typical operating
which is used mode of thisthe
to overcome wheel
load modes of its engine can be classified into three types, which are the light, the medium and the
driving resistance. In medium load mode, additional power is required to shift the working unit. In
heavy type. At the light load mode, the engine outputs a low power, which is used to overcome the
the heavy mode, the engine almost outputs the rated power, which is used to both overcome the
driving resistance. In medium load mode, additional power is required to shift the working unit. In
the heavy mode, the engine almost outputs the rated power, which is used to both overcome the
Energies 2017, 10, 1267 5 of 20

loader. It can be seen that based on the different working states of the wheel loader, the load modes of
its engine can be classified into three types, which are the light, the medium and the heavy type. At the
light load mode, the engine outputs a low power, which is used to overcome the driving resistance.
In medium load mode, additional power is required to shift the working unit. In the heavy mode,
the engine
Energiesalmost outputs the rated power, which is used to both overcome the insertion resistance
2017, 10, 1267 5 of 20 and
lift the bucket with full loads. The corresponding load mode of each segment is depicted in Figure 4a.
insertion resistance and lift the bucket with full loads. The corresponding load mode of each segment
Figure 4e presents the division of engine power between the propelling transmission system and
is depicted in Figure 4a. Figure 4e presents the division of engine power between the propelling
working hydraulics system. The 1 indicates that the engine outputs power to both branches, and the
transmission system and working hydraulics system. The 1 indicates that the engine outputs power
0 indicates only the hydraulic branch absorbs energy from the engine.
to both branches, and the 0 indicates only the hydraulic branch absorbs energy from the engine.

10 Middle Load Light Load

vw (km/h)

Mode Mode
0 Light Heavy Middle
Load Mode Load Mode Load Mode
-10
0 10 20 30 40 45
(a)
×105
1
Ft (N)

0.5

0
0 10 20 30 40 45
(b)
22
mw (t)

20
18
16
0 10 20 30 40 45
×102 (c)
5
Th (N ⋅ m)

0
0 10 20 30 40 45
(d)

0
0 10 20 30 40 45
Time (s)
(e)

Figure 4. Operating mode of wheel loader. (a) Vehicle speed; (b) Thrusting resistance; (c) Whole
Figure 4. Operating mode of wheel loader. (a) Vehicle speed; (b) Thrusting resistance; (c) Whole
vehicle mass; (d) Load torque of hydraulic pump; (e) Division of engine power between propelling
vehicle mass; (d) system
transmission Load torque of hydraulic
and working pump;
hydraulics (e) Division of engine power between propelling
system.
transmission system and working hydraulics system.
Table 1. Basic parameters of the wheel loader.
Table 1. Basic parameters of the wheel loader.
Parameters Value
Vehicle mass (kg) 16,800
Parameters
Working loads (kg) Value
5000
Rolling
Vehicle radius
mass (kg)(m) 0.75
16,800
Rolling
Workingresistance
loads (kg)coefficient 0.0335000
Engine rated
Rolling radius power
(m) (kW) 1630.75
Engine
Rolling maximumcoefficient
resistance torque (N·m) 8720.033
TC effective
Engine diameter
rated power (kW)(m) 0.34163
EngineTC torque ratio
maximum at stall(N
torque state
·m) 2.52872
Mechanical
TC effective gearbox
diameter speed
(m) ratio 4.278/2.368/1.126/0.648
0.34
TCMechanical
torque ratio gearbox
at stallefficiency
state 0.922.52
Final drive
Mechanical gearbox speed
speed ratio
ratio 4.625
4.278/2.368/1.126/0.648
Finalgearbox
Mechanical drive efficiency
efficiency 0.950.92
Wheel
Final reducer
drive speed speed
ratioratio 4.8754.625
Wheel
Final reducer
drive efficiency
efficiency 0.950.95
Wheel reducer speed ratio 4.875
2.2. Analysis of Instantaneous
WheelFuel Consumption
reducer efficiency 0.95
According to the above, the instantaneous fuel consumption of wheel loader g f is:
Energies 2017, 10, 1267 6 of 20

2.2. Analysis of Instantaneous Fuel Consumption

According to the above, the instantaneous fuel consumption of wheel loader g f is:

g f = Pe be t (1)

where Pe is the engine output power, be is the fuel consumption rate of engine, t is the unit time, equal
to one second.
During the operation of the vehicle, the engine output power is divided into two parts. One is
transferred to the working hydraulics system, which controls the working unit to perform shoveling
operations. The other is transferred to the propelling transmission system, which is used to overcome
the driving resistance. The ratio of engine speed to hydraulic pump speed is 1 in the introduced wheel
loader. According to the transfer path of engine output power and dynamics of vehicle propelling
transmission system, the following expressions can be derived:


 Pe = Pp + Pw

Pp = Tηvaω
ηh
v (2)

Tw ωv i a

Pw = Tw ωe =

ih

where Pp and Pw are the input power of the propelling transmission system and working hydraulics
system, respectively, Tv and Tw are the input torque of wheels and working hydraulics system,
respectively, ωe and ωv are the angular velocities of engine and wheels, respectively, ηa and i a are the
total efficiencies and speed ratios of mechanical gearbox, final drives and wheel reducers, respectively,
ηh and ih are the efficiency and speed ratio of TC, respectively.
The fuel consumption rate of engine is the inverse proportional function of engine efficiency [20].
Moreover, the engine efficiency is the function of output torque Te and engine speed ne :

 be = k
ηe
(3)
 η = f (T , n )
e 1 e e


 Te = Tv
ηa i a Kh + Tw
(4)
 n = 30ωv i a
e πin

where ηe is the engine efficiency, k is the fuel constant, Kh is the torque ratio of TC.
The instantaneous fuel consumption is rearranged by substituting Equations (2)–(4) into Equation (1):

kTv ωv t kTw ωv i a t
gf = + (5)
η a η h ηe ηe i h

Equation (5) shows that the instantaneous fuel consumption of vehicle is a function of the engine
efficiency, speed ratio and efficiency of TC under the constructed operating mode. Therefore, the higher
the three variable factors, the lower the instantaneous fuel consumption of the vehicle.

2.3. Loading Characteristics of Torque Converter

Equation (4) shows that the working points of the engine only depends on the torque ratio and
speed ratio of the TC. Moreover, the TC itself is the key component dominating the fuel economy of
a wheel loader. Hence, in this section, the loading characteristics of the TC are analyzed. The pump
torque capacity coefficient of TC c p is [21]:

np 1
cp = p = q (6)
Tp λ p ρgD5
Energies 2017, 10, 1267 7 of 20

np 1
cp = = (6)
Tp λp ρgD5
Energies 2017, 10, 1267 7 of 20

where Tp and np are the pump torque and speed, respectively, λp is the pump torque coefficient,
where
ρ is theTpfluid n p are the
and density, pump
g is torque and speed,
the gravitational respectively,
acceleration, and D λ is the
ispthe pumpdiameter
effective torque coefficient, ρ is
of TC circuit.
the fluid density, g is the gravitational acceleration, and D is the effective
Similarly, the turbine torque capacity coefficient of TC ct can be defined as: diameter of TC circuit.
Similarly, the turbine torque capacity coefficient of TC ct can be defined as:
n ih np ih c p
ct = n t t = i h n p = i h c p (7)
ct = √ T = pK T = K√ (7)
Tt t Khh Tpp h Kh

where TTtt and

andnt naret are
thethe turbine
turbine torque
torque andand speed,
speed, respectively.
respectively.
Based on the turbomachinery principles,
Based on the turbomachinery principles, all the torque all the torque ratio,
ratio, efficiency
efficiency andand pump
pump torque
torque
coefficients of
coefficients of the
the TCTC are
are univariate
univariate functions
functions ofof the
the speed
speed ratio
ratio [22].
[22]. As
As aa result,
result, the
the turbine
turbine torque
torque
capacity coefficient is a univariate function of the speed ratio as well. Equation
capacity coefficient is a univariate function of the speed ratio as well. Equation (7) expresses the (7) expresses the
relationship between speed ratio and external loads effected on the output terminal
relationship between speed ratio and external loads effected on the output terminal of the TC, which of the TC, which is
treated
is asasthe
treated theloading
loadingcharacteristics
characteristicsofofthe theTC.
TC.The
Thesolid
solid curve
curve inin Figure
Figure 55 displays the loading
displays the loading
characteristics of a prototype TC. It can be observed that the speed ratio increases
characteristics of a prototype TC. It can be observed that the speed ratio increases monotonically with monotonically with
the increase of turbine torque capacity coefficient. As long as the external loads
the increase of turbine torque capacity coefficient. As long as the external loads are known, both the are known, both the
speed ratio
speed ratioofofthe
theTC TC and
and working
working points
points of the
of the engine
engine will will be uniquely
be uniquely determined.
determined. The whole
The whole speed
speed ratios of TC for the constructed operating mode are calculated,
ratios of TC for the constructed operating mode are calculated, and subsequently imported into and subsequently imported
into Figure
Figure 5. figure
5. This This figure
showsshows
that atthat
theat the heavy
heavy engineengine load mode,
load mode, the corresponding
the corresponding speed speed ratios
ratios of the
of the
TC areTC are considerably
considerably low, which
low, which matchesmatches
with awithlowaefficiency
low efficiency
area area
of the of TC.
the TC.
As aAs a result,
result, its
its instantaneous fuel consumption is the highest among the
instantaneous fuel consumption is the highest among the three types of modes. three types of modes.

Light load mode

0.8
Speed ratio

0.6
Medium load mode
0.4

0.2 Heavy load mode

0
0 50 100 150 200 250 300 350
1/2
Turbine torque capacity coefficient (r/(min ⋅ (N ⋅ m) ))

Figure 5. Loading characteristics of torque converter (TC).

Figure 5. Loading characteristics of torque converter (TC).

From the above, it can be concluded that the reasonable load characteristics of a TC should be
From
matched thewith
well above,
theitexternal
can be concluded that the the
loads. Otherwise, reasonable
externalload
loadscharacteristics of atoTC
will not be able should
load be
the TC
matched well with the external loads. Otherwise, the external loads will not be able to load
effectively, and consequently this TC cannot be applied in a wheel loader. The TC indicated with the the TC
effectively,
dotted curveand consequently
in Figure this TC cannot
5 is unreasonable, sincebethere
applied in working
are no a wheel loader. Thethis
points on TCTCindicated with the
for the Segment
1dotted
and 5curve
of thein Figure 5 is operating
constructed unreasonable, since there are no working points on this TC for the Segment
mode.
1 and 5 of the constructed operating mode.
3. Initial Selection and Analysis for the HMPSTs
3. Initial Selection and Analysis for the HMPSTs
Since a lot of different HMPST configurations exist, the analysis and design for all these
Since a lot of different HMPST configurations exist, the analysis and design for all these
configurations would be a cumbersome task for an industrial designer. To avoid such a problem, this
configurations would be a cumbersome task for an industrial designer. To avoid such a problem,
section firstly selects the most reasonable schemes from within the possible HMPSTs which can be
this section firstly selects the most reasonable schemes from within the possible HMPSTs which can
initially applied to a wheel loader. Then the comparisons between the basic characteristics of each
be initially applied to a wheel loader. Then the comparisons between the basic characteristics of
each solution and that of the prototype TC are carried out, in order to get a clear understanding of
these solutions.
 Energies 2017, 10, 1267 8 of 20
Energies 2017, 10, 1267 8 of 20

solution and that of the prototype TC are carried out, in order to get a clear understanding of these
solutions.
3.1. Initial Selection
3.1. Initial
The HMPST Selection
mainly consists of a TC and a single PGT. According to the different features for
the torqueThetransferred
HMPST mainlyby theconsists
two terminals of aaHMPST,
of a TC and theAccording
single PGT. HMPST can bedifferent
to the dividedfeatures
into twofortypes:
the input torque
the torque coupled by
transferred architecture and output
the two terminals torquethe
of a HMPST, coupled
HMPST architecture,
can be dividedas into
shown twoin Figure 6.
types:
The R, Cinput
the and Storque
denote the ring
coupled gear, the planet
architecture carrier
and output and coupled
torque sun gear, respectively.
architecture, The P, in
as shown T and G denote
Figure 6.
The R,the
the pump, C and S denote
turbine the ringrespectively.
and stator, gear, the planet carrier
In the inputand sun gear,
torque respectively.
coupled HMPSTs, The P, Tare
there andsix
G basic
denote
schemes the pump,
based on the the turbine and
connections ofstator, respectively.
the input In the input
shaft of HMPST andtorque coupled
the turbine toHMPSTs,
the threethere are
components
six basic schemes based on the connections of the input shaft of HMPST and the turbine
of the PGT. Considering the TC can be inversely placed at the hydraulic branch, six additional schemes to the three
components of the PGT. Considering the TC can be inversely placed at the hydraulic branch, six
are possible for this architecture. Similarly, the output torque coupled HMPSTs also have 12 schemes,
additional schemes are possible for this architecture. Similarly, the output torque coupled HMPSTs
so there are 24 HMPST schemes in total.
also have 12 schemes, so there are 24 HMPST schemes in total.

T P
C

G
S

(a)

C T P

(b)

Figure 6. Two types of HMPSTs. (a) Input torque coupled architecture; (b) Output torque coupled
Figure 6. Two types of HMPSTs. (a) Input torque coupled architecture; (b) Output torque
architecture.
coupled architecture.
Whether the HMPSTs can be applied to the wheel loader or not depends on both the working
form of the
Whether theTC and thecan
HMPSTs power flows intothe
be applied thesystem. Since the
wheel loader TC depends
or not can only onwork
bothwell
theinworking
the first form
quadrant, the schemes that cannot meet this condition will deteriorate the
of the TC and the power flows in the system. Since the TC can only work well in the first quadrant,working performance of
wheel loader and have to be eliminated. Furthermore, the schemes that exist following two types of
the schemes that cannot meet this condition will deteriorate the working performance of wheel loader
power flows have to be eliminated as well. One is that the power is transmitted from turbine to pump,
and have to be eliminated. Furthermore, the schemes that exist following two types of power flows
the other is that the input power of TC is higher than that of system. The efficiencies of these two
have types
to be are
eliminated as well. One is that the power is transmitted from turbine to pump, the other is
always lower than that of the prototype TC, which results in them lacking any practical
that the input
value. power of TC is higher than that of system. The efficiencies of these two types are always
lower than The that of the prototype
reasonable TC, which
HMPST schemes results in
are selected bythem lacking
the lever any practical
diagram, value.
since it can visualize both
The reasonable
the power transferHMPST schemes
path of the are selected
mechanism by the lever
and the kinematic diagram,between
relationships since it each
can link
visualize
based both
on the transfer
the power torque equilibrium
path of theof mechanism
all terminals and
and conservation
the kinematic of energy [23,24]. The
relationships lever diagram
between each link forbased
Figure 6a is shown in Figure 7a, where the RC is the reference coordinate system,
on the torque equilibrium of all terminals and conservation of energy [23,24]. The lever diagram for and the circle and
Figurearrow
6a isindicate
shownthe in direction
Figure 7a, of the torque
where theTRC andisspeed ω , respectively.
the reference It cansystem,
coordinate be observed
andthat
the as the and
circle
speed of the ring gear increases from the stall state, and the three components of the PGT rotate
arrow indicate the direction of the torque T and speed ω, respectively. It can be observed that as
simultaneously in the same direction, which means that the TC always works in the first quadrant.
the speed of the ring gear increases from the stall state, and the three components of the PGT rotate
However, the power transferred by planet carrier is the sum of that transferred by the ring gear and
simultaneously in the same direction, which means that the TC always works in the first quadrant.
sun gear, which indicates that the sum of HMPST input power and the recirculating power passes
However,
throughthethepower transferred
TC with by planet
low efficiency. carrier isthis
Consequently, thescheme
sum ofshould
that transferred by the
be rejected. The ring gear and
discriminant
sun gear,
methodwhichfor theindicates
scheme in that the 6b
Figure sum of HMPST
is presented input power
in Figure 7b. and the recirculating power passes
through the TC with low efficiency. Consequently, this scheme should be rejected. The discriminant
method for the scheme in Figure 6b is presented in Figure 7b.
Energies 2017, 10, 1267 9 of 20
Energies 2017, 10, 1267 9 of 20

Energies 2017, 10, 1267 9 of 20

T ω

T ω
R Output
R Output

T C
T C

P S Input
P S Input

- + - +
- + - +
RC RC
RC (a)
(a)
RC

P P R R

Output
Output
T T C C

S Input
S Input
- + - +
- + - +
RC RC
RC (b) RC
(b)
Figure 7. Lever diagram. (a) Discriminant method for the scheme in Figure 6a; (b) Discriminant
Figure 7. Lever diagram. (a) Discriminant method for the scheme in Figure 6a; (b) Discriminant
Figure method for the
7. Lever scheme in
diagram. (a)Figure 6b.
Discriminant method for the scheme in Figure 6a; (b) Discriminant
method for the scheme in Figure 6b.
method for the scheme in Figure 6b.
The ring gear rotates in the opposite direction relative to the planet carrier while the planet
carrier
The speed increases from zero to a certain value, whichtomeans that the pumpwhile
and turbine rotate
The ring
ringgear
gearrotates
rotatesinin
thethe
opposite
oppositedirection relative
direction relativethe
toplanet carrier
the planet the planet
carrier while carrier
the planet
speed inincreases
opposite from
directions,
zero and
to a the TC value,
certain works in the second
which means quadrant.
that the As a result,
pump and this scheme
turbine rotatehas
in to be
opposite
carrier speed increases from zero to a certain value, which means that the pump and turbine rotate
rejected as well. Similarly, the remaining 22 schemes are judged by this method. Figure 8 shows the
directions,
in opposite and the TC works
directions, and thein the
TC second
works quadrant.
in the secondAs aquadrant.
result, thisAsscheme hasthis
a result, to be rejected
scheme hasasto
well.
be
four ultimately reasonable solutions for HMPSTs.
Similarly,
rejected asthe remaining
well. 22the
Similarly, schemes are judged
remaining by this
22 schemes aremethod. Figure
judged by 8 shows Figure
this method. the four8 ultimately
shows the
reasonable
four solutions
ultimately for HMPSTs.
reasonable solutions for
R HMPSTs. R

P T P T
R C C R

G G
P T P T
C S C
S
G (a) G
(b)
S
R S
R
(a) (b) T P
C
C T P
R G
R G

S S T P
C
C T P
(c) (d)
G
G
Figure 8. Four reasonable solutions for HMPSTs. (a) Solution a; (b) Solution b; (c) Solution c; (d)
S
Solution d. S
(c) (d)
3.2. Basic Characteristics Analysis
Figure 8. Four reasonable solutions for HMPSTs. (a) Solution a; (b) Solution b; (c) Solution c; (d)
Figure 8. Four reasonable solutions for HMPSTs. (a) Solution a; (b) Solution b; (c) Solution c;
Solution d.
(d) Solution d.

3.2. Basic Characteristics Analysis

Energies 2017, 10, 1267 10 of 20
Energies 2017, 10, 1267 10 of 20

As the
3.2. Basic TC is embedded
Characteristics in the HMPST, each of above four solutions is equivalent to a certain
Analysis
type of TC. The basic characteristics of the HMPSTs comprise efficiency ηm and loading
As the TC is embedded in the HMPST, each of above four solutions is equivalent to a certain type
of TC. The basiccmt
characteristics . The derivation
characteristics forHMPSTs
of the these basic characteristics
comprise areηbased
efficiency m and on the steady
loading state and cthe
characteristics mt .
The
PGTderivation
efficiencyfor these
treats asbasic characteristics
the constant [25,26]. areAccording
based on the to steady state equilibrium
the torque and the PGT of
efficiency treats
PGT, energy
as the constant
conservation and[25,26].
known According
power flowto the
of torque equilibrium
each solution, of PGT,
the basic energy conservation
characteristics and known
of each solution can
power flow
be derived as:of each solution, the basic characteristics of each solution can be derived as:

ηmηm== ff2m (τ,i)

(
2 m ( τ, i )
 (8)
(8)
cmt
cmt== ff 3 m (τ, D,i)
3m ( τ, D, i )

where m
where m ((mm ==a,b,c,d ) denotes
a,b,c,d) denotesthe thefour
foursolutions
solutionsininFigureFigure8,8,ττ is is the
the structural
structural parameter
parameter of of the
the PGT,
PGT,
which is the ratio of the numbers of ring gear teeth to that of sun gear
which is the ratio of the numbers of ring gear teeth to that of sun gear teeth, i is the HMPST speed ratio teeth, i is the HMPST speed
ratio
and equaland equal
to thetoratio
the of
ratio of output
output speedspeed to input
to input speed.speed.
Equation
Equation (8) shows that the variable factors influencing
(8) shows that the variable factors influencing the the basic
basic characteristics
characteristics of of each
each solution
solution
are
are the
the structural
structural parameter
parameter of of the
thePGTPGT ττ and andthe theeffective
effectivediameter
diameter of of the
the TC
TC circuit
circuit D. D. By
By further
further
analysis, it is found that the efficiency varies strictly monotonically with
analysis, it is found that the efficiency varies strictly monotonically with τ and D. Therefore, the two τ and D . Therefore, the
two
boundaryboundaryvaluesvalues
for thefor the a τspecific
τ and and avaluespecific value
for the D are foradopted
the D are adoptedthe
to examine tobasic
examine the basic
characteristics
characteristics
of each solution. of each solution.
Figure
Figure 99 shows
shows thethe efficiency
efficiency of of each
each solution.
solution. The term pp denotes
The term denotes thethe prototype
prototype TC. TC. For
For solution
solution
aa,, its
its efficiency
efficiencycan cangain
gaina asignificant
significant advantage over the prototype when the
advantage over the prototype when the speed ratio is greater speed ratio is greater
than
than about 0.65. Figures 4 and 5 show that the engine works in the
about 0.65. Figures 4 and 5 show that the engine works in the light load mode in this range. However, light load mode in this range.
However,
in the lowin the low
speed speed
ratio range,ratio
therange,
engine theoutputs
engine higher
outputspower.
higher That
power. Thatthe
means means the proportion
proportion of fuel
of fuel consumption
consumption for lowfor low speed
speed ratio range
ratio range are greater
are greater than for
than that thathigh
for high
speed speed
ratioratiorange range
among amongthe
the
totaltotal
fuelfuel consumption.
consumption. DueDue to theto oversized
the oversized ica ,deficiency
ica , the the deficiency of efficiency
of efficiency in theinlow thespeed
low speed
ratio
ratio
rangerangecan hardlycan hardly
be made be upmade up by advantage
by advantage of efficiency
of efficiency in the high in speed
the high speed
ratio range. ratio
Forrange.
solution Forb,
solution
its efficiencyb, itsisefficiency
much lessisthan much less
that ofthan that ofinprototype
prototype the entirein the entire
range of τ. Forrange of τ .c,For
solution thesolution
efficiencyc,
integral
the area of
efficiency such a area
integral solution
of suchis slightly
a solution greater than that
is slightly of thethan
greater prototype.
that of the icc decreases
Theprototype. Thewith
icc
the increase of τ, which facilitates the improvement of fuel economy
decreases with the increase of τ , which facilitates the improvement of fuel economy of vehicle in the of vehicle in the low speed ratio
range.
low As for
speed ratiotherange. As d,
solution foritthe
displays
solution a similar situation
d, it displays to solution
a similar a. to solution a.
situation

1 1
icc icc
0.8 0.8
ica ica
Efficiency
Efficiency

0.6 0.6
p
0.4 a 0.4
b
c
0.2 d 0.2

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Speed ratio Speed ratio
(a) (b)
Figure 9. Efficiency of each solution. (a) τ = 4/3; (b) τ = 5.
Figure 9. Efficiency of each solution. (a) τ = 4/3; (b) τ = 5.

Figure 10 shows the loading characteristics of each solution. For the solution with characteristic
curveFigure
above10the
shows the loading
prototype TC, itcharacteristics
can achieve larger of each solution.
speed ratio For the same
at the solution with characteristic
abscissa values. That
means the engine can work at a smaller speed, and consequently the engine efficiency isThat
curve above the prototype TC, it can achieve larger speed ratio at the same abscissa values. means
increased.
the engine the
However, canchange
work atrate
a smaller speed,
of speed ratioand consequently
for this solution isthe engine
greater efficiency
than is increased.
for the prototype. AsHowever,
a result,
the change rate of speed ratio for this solution is greater than for the prototype. As
the working points of the engine will fluctuate in a larger range with the variation of external a result, the working
loads.
points of the engine will fluctuate in a larger range with the variation of external
Excessive transient conditions of engine are not conducive to improving the engine efficiency. loads. Excessive
transient conditions
Moreover, of engine
the solutions a and d are not be
cannot conducive
applied tototheimproving
wheel loaderthe engine
as the τefficiency. Moreover,
increases nearly to 5,
since they cannot match well with the external loads. In a word, the ideal curve of each solution
Energies 2017, 10, 1267 11 of 20

the solutions a and d cannot be applied to the wheel loader as the τ increases nearly to 5, since they
Energies 2017, 10, 1267 11 of 20
cannot match
Energies 2017, 10, well
with the external loads. In a word, the ideal curve of each solution should approach
1267 11 of 20
that of the prototype and make the engine work in the economical areas simultaneously.
should approach that of the prototype and make the engine work in the economical areas Figure 11
presents
should the engine
approach efficiency
that of map
the for this
prototype study.
and make the engine work
simultaneously. Figure 11 presents the engine efficiency map for this study. in the economical areas
simultaneously. Figure 11 presents the engine efficiency map for this study.
1 1
1 1
0.8 0.8
0.8 0.8
p
ratioratio

ratioratio
0.6 0.6
ap
0.6 0.6
Speed

Speed
ba
0.4 0.4
Speed

Speed
cb
0.4 dc 0.4
0.2 d 0.2
0.2 0.2
0 0
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400
0 0
0
Turbine 50 100
torque 150coefficient
capacity 200 250(r/(min ⋅ m)1/2
300⋅ (N350 400 )) Turbine
0 torque
50 100capacity 200 250(r/(min
150coefficient ⋅ m)1/2
300 ⋅ (N350 400 ))
(a)
Turbine torque capacity coefficient (r/(min ⋅ (N ⋅ m)1/2 )) (b) (r/(min ⋅ (N ⋅ m)1/2 ))
Turbine torque capacity coefficient
(a) (b)
Figure 10. Loading characteristics of each solution. (a) τ = 4/3, D = 0.3m; (b) τ = 5, D = 0.3m.
Figure 10. Loading characteristics of each solution. (a) τ = 4/3, D = 0.3 m; (b) τ = 5, D = 0.3 m.
Figure 10. Loading characteristics of each solution. (a) τ = 4/3, D = 0.3m; (b) τ = 5, D = 0.3m.
900
900
42.1%
700 42.1%
41.5%
41.5%
41%
m) m)

700
41% 40.6%
(N· (N·

500 40.6%
torque

500 40.2%
torque

39.3%40.2%
Engine

39.3% 38.5%
300
Engine

300 38.5%
37.4%
37.4%36.1%
100 36.1%
33.9%
100
0 33.9%
800 1200 1600 2000 2400
0 −1
800 1200 Engine1600 （r ⋅ min
speed 2000） 2400
Engine speed（r ⋅ min −1）
Figure 11. Engine efficiency.
Figure 11. Engine efficiency.
Figure 11. Engine efficiency.
4. Design Optimization of Each Solution
4. Design Optimization of Each Solution
4. Design
In thisOptimization of Each
section, the design Solution for each preliminary solution derived from the above is
optimization
In this section,
implemented, which the design
aims at optimization
minimizing thefor each
fuel preliminarycaused
consumption solutionbyderived fromand
the engine the above
HMPST is
In this section, the design optimization for each preliminary solution derived from the above
implemented,
losses. which
At this point, aims
first at
the atminimizing
optimization the fuel consumption caused by the engine and HMPST
is implemented, which aims minimizingproblems are put forward,
the fuel consumption caused where theengine
by the objective
andfunctions
HMPST
losses.
are At this point,potential
first theof optimization problems are putprototype
forward, where the these
objective functions
losses. At this point, first the optimization problems are put forward, where the objective functions are
the fuel saving each solution relative to the TC. Then problems are
are the by
solved fuelthesaving potentialalgorithm.
enumeration of each solutionBoth the relative
design to the prototype
optimization TC. Then
problem and these problems
technique for are
each
the fuel saving potential of each solution relative to the prototype TC. Then these problems are solved
solved by
HMPST the enumeration
solution arealgorithm.
the same algorithm.
since isBoth the design to aoptimization problem and technique for each
by the enumeration Both it the equivalent
design optimization certain type of
problem TC.technique
and for each HMPST
HMPST solution are the same since it is equivalent to a certain type of TC.
solution are the same since it is equivalent to a certain type of TC.
4.1. Optimization Problems
4.1. Optimization Problems
4.1. Optimization Problems
4.1.1. Objective Functions
4.1.1. Objective
4.1.1. Objective Functions
Functions
The fuel consumption of a wheel loader, which was discussed rigorously in the previous section,
The fuel
The fuel consumption
is dominated only by the engine
consumption of
of aa wheel loader,
efficiency,
wheel loader, which
speed
which ratiowas
was discussed
and efficiency
discussed rigorously in
in the
of the HMPST.
rigorously the previous section,
At any moment,
previous section,
is dominated
if
is each of these
dominated only by
by the
onlythree engine
factors
the engine efficiency,
obtained
efficiency, speed
byspeed ratio
ratio and
the HMPSTs andisefficiency
greater of
efficiency the
than
of HMPST.
thethe At
At any
any moment,
corresponding
HMPST. factor
moment,
if each
obtained of these three
by thethree
if each of these prototypefactors
factors TC, obtained
then the
obtained by the by the HMPSTs
instantaneous
HMPSTs is greater is greater
fuel consumption than the corresponding
will be naturally
than the corresponding factor
factor reduced.
obtained
obtained
The
by theabove by the prototype
conception
prototype can be
TC, then TC, then
theembodied the
instantaneous instantaneous
by the fuelfollowing
consumptionfuel consumption
expression: will be naturally
will be naturally reduced. The above reduced.
The above conception
conception can be by
can be embodied embodied
the followingby theexpression:
following expression:
ηep (tij )ηhp (tij ) ηep (tij )ihp (tij )
ηinterm1 (tij ) = η (t )η (t ) + η (t )i (t ) (9)
ηinterm1 (tij ) = ηei (tij )ηhi (tij ) + ηhi (tij )ihi (tij )
ep ij hp ij ep ij hp ij
(9)
ηei (tij )ηhi (tij ) ηhi (tij )ihi (tij )
Energies 2017, 10, 1267 12 of 20

ηep (tij )ηhp (tij ) ηep (tij )ihp (tij )

ηinterm1 (tij ) = + (9)
ηei (tij )ηhi (tij ) ηhi (tij )ihi (tij )
where ηep (tij ) is the engine efficiency for wheel loader using prototype TC at the moment of tij , ηhp (tij )
and ihp (tij ) are the efficiency and speed ratio of prototype TC, respectively, is the engine efficiency for
wheel loader using each solution of HMPSTs at the moment of tij , ηhi (tij ) and ihi (tij ) are the efficiency
and speed ratio of each solution of HMPSTs, respectively.
Furthermore, the transmission types of engine output power have to be weighed, since a portion
of the engine output power is used to thrust the vehicle itself while the rest is taken to implement the
working task. In view of this, the two weighting coefficients on the two legs of power are added to
Equation (9). At any moment, the fuel saving potential for the wheel loader using each solution of
HMPSTs relative to that using the prototype TC can be derived:

ηep (tij )ηhp (tij ) ηep (tij )ihp (tij )

ηinterm2 (tij ) = wi1 + wi2 (10)
ηei (tij )ηhi (tij ) ηhi (tij )ihi (tij )

where wi1 and wi2 are the weight coefficients for the propelling transmission system and working
hydraulics system, respectively.
Existing studies show that the ratio of average input power of the propelling transmission system
to that of the working hydraulics system is about 1.5 [15,18], so the relationships between wi1 and wi2
are given as follows:
wi1 + wi2 = 1
(11)
wi1 : wi2 = 3 : 2
In every working segment, the total fuel saving potential for the HMPSTs is the time accumulation
of Equation (10). In order to display the pros and cons of each solution of HMPSTs relative to the
prototype TC clearly, the further normalization is conducted:
!
1 Ni ηep (tij )ηhp (tij ) ηep (tij )ihp (tij )
Jmi = ∑
Ni j=1
wi1
ηei (tij )ηhi (tij )
+ wi2
ηhi (tij )ihi (tij )
(12)

where Ni is the number of sampling points for each working segment (i = 1, 2, . . . 5).
Figures 4 and 5 present that each load mode of engine differs greatly in the output power level,
so in the light load mode, the engine output power is much lower than the rated power. In the heavy
load mode, the engine works near the rated power point so that the considerable inserting resistance
can be overcome. Therefore, the contribution of each load mode to the total fuel consumption is
different. By incorporating the corresponding weighting coefficient into Equation (12), the total fuel
saving potential for the HMPSTs under general working manner is obtained as below:

5
Jm = ∑ wi Jmi (13)
i =1

where wi is the weighting coefficient for each working segment. Here, is subjected to the lucid
assumptions:


 ω1 + ω2 + ω3 + ω4 + ω5 = 1

 ω :ω :ω =1:3:2

1 2 3
(14)


 ω1 = ω5

ω3 = ω4

Energies 2017, 10, 1267 13 of 20

4.1.2. Design Variables

Considering this study focuses on improving the fuel economy of a wheel loader using a HMPST,
the design variables for the optimization problem are the structural parameter of the PGT and the
effective diameter of the TC circuit, which satisfy the following inequalities:

Xm = [ τ D ] (15)

4/3 ≤ τ ≤ 5 (16)

0.2 m ≤ D ≤ 0.34 m (17)

4.1.3. Constraints
The engine output torque has to meet the requirements of both the driving and the working as
the wheel loader performs a task. Accordingly, the sum of the input torques for these two systems
should not exceed the torque of the engine external characteristics Tex :

Tpr (ne ) + Thr (ne ) ≤ Tex (ne ) (18)

where Tpr and Thr are the input torques of the propelling transmission system and working hydraulics
system, respectively.
Moreover, the torque load characteristics of the HMPST’s input terminal at zero speed ratio is
important since the wheel loader has to overcome a heavy load at stall state. The intersection between
the torque load curve of the HMPST’s input terminal at zero speed ratio and the engine external
characteristic curve should be located between the point of maximum torque Temax and that of engine
rated power Pemax :
nex ( Temax ) ≤ nex (λhp0 ) ≤ nex ( Pemax ) (19)

where nex is the engine speed corresponding to engine external characteristic curve, λhp0 is the
equivalent pump torque coefficient of the HMPST at zero speed ratio.

4.2. Optimization Algorithm

As for the design optimization for each HMPST solution, the number of objective functions and
design variables is one and two, respectively. That means the optimization problems can be solved by
a less-intelligent algorithm. In this study, the enumeration algorithm is used to solve these problems.
The concrete solving flow chart for each solution is presented in Figure 12. All the implementation of
these procedures are realized in Matlab 2015b (The MathWorks, Natick, MA, USA).
Figure 13 shows the solutions for scheme c while the other three schemes cannot find any solution
less than 1, which indicates schemes a, b, and d cannot show fuel saving advantages over the prototype
TC and have to be eliminated. It can be seen that the solutions mainly gather in the areas where
the effective diameter of the TC circuit D is large. The reason is that the solution c with small D can
achieve a low speed ratio relative to the prototype for the same turbine torque capacity coefficient.
Even if at the same speed ratio of low speed ratio area, the efficiency of solution c is less than that of
the prototype, so the lower speed ratio will reduce the efficiency of solution c further. At this point,
the HMPST cannot gain fuel saving advantages over the prototype. The optimal solution for scheme c
is 0.93 and the corresponding design variables are [2.73 0.31].
Figures 14 and 15 show the architecture and basic characteristics of solution c with the optimal
structural parameters, respectively. Compared with the prototype TC, the optimized solution c can
gain advantage in the global speed ratio range due to the fact that its higher efficiency in the high
speed ratio range, and its improvement of engine efficiency can counteract the deficiency of efficiency
in the low speed ratio range.
Energies 2017, 10, 1267 14 of 20
Energies 2017, 10, 1267 14 of 20

Energies 2017, 10, 1267 14 of 20

Energies 2017, 10, 1267 Start 14 of 20
Start Engine output torque
Typical operating
Start mode
Engine output torque
Typical operating mode
Engine output torque
Output torqueoperating
Typical and speedmode
for each No
solution of HMPSTs Tpr (ne ) + Thr (ne ) ≤ Tex (ne )
Output torque and speed for each No
Outputsolution of HMPSTs
torque and speed for each No
Tpr (ne ) + Thr (ne ) ≤ Tex (ne )
D = Dmin of
solution : ΔHMPSTs
D : Dmax Tpr (ne ) + Thr (nYes
e ) ≤ Tex ( ne )

D = Dmin : ΔD : Dmax
Engine outputYes
torque
τ ==τD
D min : Δ
:ΔτD
: τ:max
D and speed
No
min max
Yes
Engine output torque
τ =τ min : Δτ : τ max and speedtorque
Engine output
No τ =τ min : Δτ : τ max Engine
andefficiency
speed
No nex (Te max ) ≤ nex (λhp 0 )
Engine efficiency
nnex ((λThp 0 ) ≤) ≤nexn( P(eλmax ))
ex e max ex hp 0 Engine efficiency
Calculating Jm
λehpmax
nex (T 0 ) )≤≤nn λe max
exex( (P hp 0 )
Calculating J m
n (λ ) ≤ nYes( P )
ex hp 0 ex e max
No Calculating J m
Speed ratio and efficiency
Yes for each Jm < 1
solution of HMPSTs No
Yes for each
Speed ratio and efficiency Next stepJ <Yes 1
m
solution
Speed ratio and of HMPSTsfor each
efficiency No J <
Next step m Yes1
Input torqueofofHMPSTs
solution propelling Min J m
Next step Yes
transmission system and engine speed
Input torque of propelling Min J m
transmission systemofand
Input torque engine speed
propelling End J m
Min
transmission system and engine speed
End
Figure 12. Solving flow chart.
Figure 12. Solving flow chart. End
Figure 12. Solving flow chart.
Figure 12. Solving flow chart.

Figure 13. Solutions for scheme c.

Figure 13. Solutions for scheme c.
Figure 13. Solutions for scheme c.
RFigure 13. Solutions for scheme c.
R
CR T P

C T P
C T G P

S G
G
S
Figure 14. Architecture of solution c.
S
Figure 14. Architecture of solution c.
Figure 14. Architecture of solution c.
Figure 14. Architecture of solution c.
Energies 2017, 10, 1267 15 of 20
Energies 2017, 10, 1267 15 of 20

1 1

0.8 0.8

Speed ratio
Efficiency

0.6 0.6
Solution c
0.4 0.4
Prototype TC

0.2 0.2

0 0
0 0.2 0.4 0.6 0.8 1 0 50 100 150 200 250 300 350
Speed ratio Turbine torque capacity coefficient (r/(min ⋅ (N ⋅ m)1/2 ))
(a) (b)
Figure 15. Basic characteristics of solution c. (a) Efficiency; (b) Loading characteristics.
Figure 15. Basic characteristics of solution c. (a) Efficiency; (b) Loading characteristics.
5. Simulation and Analysis
5. Simulation and Analysis
5.1. Simulation Model
5.1. Simulation Model
In order to verify the effectiveness of solution c with the optimal structural parameters, it is
essential
In ordertotobuild
verifya simulation model thatofcan
the effectiveness imitate cthe
solution behavior
with of a wheel
the optimal loader according
structural to the it is
parameters,
constructed
essential to buildoperating mode.model
a simulation As thisthat
study canis imitate
solely interested
the behaviorin theoffuel consumption
a wheel loader of wheel to
according
loaders, the operating
the constructed detailed working
mode. state of each
As this study element such
is solely as the engagement
interested in the fuelprocess of the clutch
consumption of wheel
during shifting and the boom rotating angle, can be ignored while constructing the simulation model.
loaders, the detailed working state of each element such as the engagement process of the clutch during
For that reason, the backward programming method is adequate for such a job [19,24]. The principle
shifting and the boom rotating angle, can be ignored while constructing the simulation model. For that
of the backward programming method is calculating the input torque and speed of each element
reason, the backward
according programming
to the known output method is adequate
power variables. for such a job the
By implementing [19,24]. The principle
calculation from theof the
backward programming method is calculating the input torque and speed of
downstream to the upstream step by step, the total fuel consumption can be derived. The modellings each element according
to theforknown
both theoutput powerusing
wheel loader variables. By implementing
the solution c of HMPSTs andthe thatcalculation fromTC
using prototype the
aredownstream
performed to
the upstream step by step,environment.
in the Matlab/Simulink the total fuel consumption can be derived. The modellings for both the
wheel loaderThe simulation
using the procedures
solution c for both models
of HMPSTs andarethat
stated below.
using Firstly, the
prototype TC operating mode of
are performed in the
wheel loader is divided
Matlab/Simulink environment. into two types: the mechanical (the vehicle speed, thrusting resistance and
whole
The vehicle mass)
simulation and thefor
procedures hydraulic (the load
both models torque below.
are stated of hydraulic
Firstly,pump). All the mechanical
the operating mode of wheel
components in the propelling transmission system are treated as the inertia element. The output
loader is divided into two types: the mechanical (the vehicle speed, thrusting resistance and whole
torque and speed of mechanical gearbox are calculated from the mechanical variables based on the
vehicle mass) and the hydraulic (the load torque of hydraulic pump). All the mechanical components
vehicle system dynamics. Secondly, the input torque and speed for the solution c of the HMPSTs or
in theTC
propelling
are derivedtransmission
by the shiftingsystem
controlare treatedand
strategy as the
theloading
inertia characteristics.
element. The outputFinally, torque and speed
the hydraulic
of mechanical gearbox
variable is added areinput
to the calculated
torque offrom the mechanical
the solution variables
c, so the engine torquebased on theare
and speed vehicle system
obtained.
dynamics. Secondly,
Thus, the the input torque
whole simulation and speed for the solution c of the HMPSTs or TC are derived by
is completed.
the shifting control strategy and the loading characteristics. Finally, the hydraulic variable is added
to the5.2. Results
input Analysis
torque of the solution c, so the engine torque and speed are obtained. Thus, the whole
simulation is completed.
In this section, the simulations for the wheel loader using solution c for the HMPST and that
using the prototype TC are implemented, respectively. Figure 16a,b present the load power of the
5.2. Results
engine Analysis
and working hydraulics system for the prototype TC and the solution c of the HMPST,
respectively. As for the wheel loader equipped with TC, the engine speed depends primarily on the
In this section, the simulations for the wheel loader using solution c for the HMPST and that using
loading characteristics of the TC. The principle of cooperation between the driving and working is to
the prototype TC are implemented, respectively. Figure 16a,b present the load power of the engine
ensure that the hydraulics work robustly and consume less energy, simultaneously. The load power
and working hydraulics
of the engine system forTCthe
for the prototype is prototype TC and
slightly greater thanthe
thatsolution c of thec HMPST,
of the solution when therespectively.
vehicle
As forspeeds up nearly to the peak. The reasons are that both schemes work at the high speed on
the wheel loader equipped with TC, the engine speed depends primarily the
ratio loading
areas,
characteristics of the TC. The principle of cooperation between the driving and
and the efficiency of solution c is higher than that of the prototype TC in these areas.working is to ensure
that the hydraulics work robustly and consume less energy, simultaneously. The load power of the
engine for the prototype TC is slightly greater than that of the solution c when the vehicle speeds up
nearly to the peak. The reasons are that both schemes work at the high speed ratio areas, and the
efficiency of solution c is higher than that of the prototype TC in these areas.
Energies 2017, 10, 1267 16 of 20
Energies 2017, 10, 1267 16 of 20
Energies 2017, 10, 1267 16 of 20

160 160 160

160
Engine
Engine Engine
Engine
Hydraulics
Hydraulics Hydraulics
Hydraulics
120 120 120
120

(kW)
Power(kW)
Power (kW)
Power (kW)

80 80

Power
80 80

40 40
40 40

0 0
0 0 5 10 15 20 25 30 35 40 45 00 5 10 15 20 25 30 35 40 45
0 5 10 15 20 Time
25(s) 30 35 40 45 0 5 10 15Time20(s) 25 30 35 40 45
Time(a)
(s) (b)Time (s)
(a) (b)
Figure 16. Load power of engine and working hydraulics system. (a) Prototype TC; (b) Solution c.
Figure 16. Load power of engine and working hydraulics system. (a) Prototype TC; (b) Solution c.
Figure 16. Load power of engine and working hydraulics system. (a) Prototype TC; (b) Solution c.
Since the engine efficiency, speed ratio and efficiency of the HMPSTs or TC determine the fuel
Since the engine
consumption
Since the engine efficiency,
of a wheel loader,speed
efficiency, ratio
ratio and
the comparisons
speed and efficiency
on these three
efficiency of the
of HMPSTs
factors
the between
HMPSTs or TC
orthe
TCtwodetermine the
the fuel
schemes are
determine fuel
conducted.
consumption ofFigure
a wheel17aloader,
shows thethe values of speedon
comparisons ratio of solution
these three c minusbetween
factors that of the
theprototype.
two All are
schemes
consumption of a wheel loader, the comparisons on these three factors between the two schemes are
the speed
conducted. ratios of solution
Figure c are greater than those of of
thesolution
prototype when the
c minus engine operates at the
conducted. Figure 17a17ashows
showsthe thevalues
valuesofofspeed
speedratio
ratio of solution c minus thatthat
of the prototype.
of the prototype.All All
the
speednon-idle
ratios state,
of which
solution facilitates
c are the
greater display
than of the
those fine
of fuel
the economy
prototype of this
when architecture
the engine relative to
operates theat the
the speed ratiosThe
prototype. of solution
speed ratio c are greater
surplus than
at the those
heavy of mode
load the prototype
can narrow when the engine
the efficiency operates
gap betweenat the
non-idle state,
non-idle state, which
which facilitates
facilitates the
the display
display of of the
the fine
fine fuel
fuel economy
economy of of this
this architecture
architecture relative
relative to
to the
the prototype and the solution c, while in the light and medium load mode, the speed ratio surplus the
prototype.
prototype. The
The cspeed
speed ratio
ratiosurplus
surplus atatthe
the heavy
heavy load
loadmode
mode can
cannarrow
narrow thethe
efficiency gapgap
efficiency between
betweenthe
of solution can increase the efficiency benefit relative to the prototype further.
prototype andand
the prototype the the
solution c, while
solution in the
c, while light
in the andand
light medium
medium load mode,
load thethe
mode, speed ratio
speed surplus
ratio of
surplus
solution c can increase thethe 0.04
efficiency benefit relative to the prototype further.
of solution c can increase efficiency benefit relative to the prototype further.

0.04
0.03
Speed ratio

0.02
0.03
Speed ratio

0.01
0.02

0
0.01 0 5 10 15 20 25 30 35 40 45
Time (s)
(a)
0.1
0
0.08
0 5 10 15 20 25 30 35 40 45
Time (s)
0.06 (a)
Efficiency

0.1
0.04
0.08
0.02
0.06
Efficiency

0
0.04
-0.02
0 5 10 15 20 25 30 35 40 45
0.02 Time (s)
(b)
0

-0.02
0 5 10 15 20 25 30 35 40 45
Time (s)
(b)

Figure 17. Cont.

Energies 2017, 10, 1267 17 of 20
Energies 2017, 10, 1267 17 of 20

0.015

0.01

Engine efficiency
0.005

-0.005

-0.01

-0.015
0 5 10 15 20 25 30 35 40 45
Time (s)
(c)

Figure 17.
Figure 17. Difference
Differencebetween
betweenthe
thesolution
solutionc cand
andthe prototype
the TC.
prototype (a)(a)
TC. Speed ratio;
Speed (b) (b)
ratio; Efficiency; (c)
Efficiency;
Engine efficiency.
(c) Engine efficiency.

These phenomena are reflected in Figure 17b, where the solution c efficiency is slightly less than
These phenomena are reflected in Figure 17b, where the solution c efficiency is slightly less than
the prototype efficiency only at the start stage of wheel loader and a few parts of the heavy load
the prototype efficiency only at the start stage of wheel loader and a few parts of the heavy load
mode. Moreover, in the light and medium load mode, the solution c shows its prominent efficiency
mode. Moreover, in the light and medium load mode, the solution c shows its prominent efficiency
advantages over the prototype. Taken together, the solution c can attain an efficiency advantage over
advantages over the prototype. Taken together, the solution c can attain an efficiency advantage over
the prototype in the global working range, which helps to reduce the fuel consumption. Figure 17c
the prototype in the global working range, which helps to reduce the fuel consumption. Figure 17c
presents the results for the engine efficiency of the wheel loader using solution c minus that of wheel
presents the results for the engine efficiency of the wheel loader using solution c minus that of wheel
loader using the prototype. All of these differences are less than 0.015, which indicates the solution c
loader using the prototype. All of these differences are less than 0.015, which indicates the solution
cannot fundamentally change the engine working areas of each load mode. The reasons are that the
c cannot fundamentally change the engine working areas of each load mode. The reasons are that
HMPST is a self-adaptive mechanism and cannot control the engine to work in specified areas. In the
the HMPST is a self-adaptive mechanism and cannot control the engine to work in specified areas.
heavy load mode, the engine efficiency of solution c is slightly higher than that of the prototype. This
In the heavy load mode, the engine efficiency of solution c is slightly higher than that of the prototype.
is because the solution c efficiency is less than the prototype efficiency in this mode, which results in
This is because the solution c efficiency is less than the prototype efficiency in this mode, which results
a larger engine output power for the same power requirements, so the working points of the engine
in a larger engine output power for the same power requirements, so the working points of the engine
have to move to the lower speed and higher efficient areas. On the whole, the solution c can gain an
have to move to the lower speed and higher efficient areas. On the whole, the solution c can gain an
engine efficiency advantage over the prototype in the global working range, which helps to reduce
engine efficiency advantage over the prototype in the global working range, which helps to reduce the
the fuel consumption.
fuel consumption.
From the above, it is concluded that compared with the prototype TC, HMPST solution c can
From the above, it is concluded that compared with the prototype TC, HMPST solution c can
reduce the fuel consumption effectively in the global range. Figure 18 shows the total fuel
reduce the fuel consumption effectively in the global range. Figure 18 shows the total fuel consumption
consumption for these two architectures. The fuel consumption of solution c is obviously less than
for these two architectures. The fuel consumption of solution c is obviously less than that of the
that of the prototype after the wheel loader implements heavy loads. Table 2 presents the total fuel
prototype after the wheel loader implements heavy loads. Table 2 presents the total fuel consumption
consumption for the 50 repetitive typical operating modes. By applying the optimal HMPST solution
for the 50 repetitive typical operating modes. By applying the optimal HMPST solution c, the fuel
c, the fuel economy of the wheel loader improved by 3.38%. Similarly, the total fuel consumption for
economy of the wheel loader improved by 3.38%. Similarly, the total fuel consumption for the
the sub-optimal solution c is listed in Table 2, and its design variables are [3.33 0.315]. The sub optimal
sub-optimal solution c is listed in Table 2, and its design variables are [3.33 0.315]. The sub optimal
solution c presents a fuel savings of 2.9%. Therefore, the TC can be substituted by this type of HMPST
solution c presents a fuel savings of 2.9%. Therefore, the TC can be substituted by this type of HMPST
in wheel loader applications.
in wheel loader applications.
Table 2. Comparison results for total fuel consumption.
Table 2. Comparison results for total fuel consumption.
Objective Fuel Consumption (L) Savings (%)
Prototype TC
Objective 10.35 (L)
Fuel Consumption -
Savings (%)
Optimal
Prototype TC solution c 10.3510.00 3.38 -
Sub
Optimal optimalc solution c
solution 10.0010.05 2.903.38
Sub optimal solution c 10.05 2.90
Energies 2017, 10, 1267 18 of 20
Energies 2017, 10, 1267 18 of 20

200

Optimal solution c

Total fuel consumption (g)

150 Prototype TC

100

50

0
0 5 10 15 20 25 30 35 40 45
Time (s)

Figure 18. Total fuel consumption.

Figure 18. Total fuel consumption.
5.3. Discussion
5.3. Discussion
A novel design methodology for HMPSTs is proposed in this paper, in order to find the optimal
A novelsaving
energy design methodology
solution relative tofor
theHMPSTs
prototypeisTC.proposed in this
The solution paper,
c has in orderoftoenergy
an advantage find the optimal
saving
over the schemes proposed in [7,12]. Although the optimal solution c is only verified
energy saving solution relative to the prototype TC. The solution c has an advantage of energy saving in the typical
over operating
the schemesmode, it can provide
proposed an energy
in [7,12]. savingthe
Although advantage
optimal over the prototype
solution TCverified
c is only in other operating
in the typical
modes as well. The reasons can be explained from two aspects. First, the design method proposed is
operating mode, it can provide an energy saving advantage over the prototype TC in other operating
based on the three load modes of engine, the two power flows of engine, and the five segments of
modes as well. The reasons can be explained from two aspects. First, the design method proposed
operating mode of wheel loader. These three aspects are the basic characteristics of a wheel loader.
is based on the three load modes of engine, the two power flows of engine, and the five segments of
Second, as analyzed in the Sections 3.2 and 5.2, the solution c can gain both efficiency and engine
operating mode
efficiency of the
over wheel loader.TCThese
prototype in thethree
globalaspects are the
range. Due basic
to the characteristics
relatively of a wheel
complex design methodloader.
Second, as analyzed in the Sections 3.2 and 5.2, the solution c can gain both efficiency
for HMPSTs and the article length restrictions, a sensitivity analysis of the weighting factors of the and engine
efficiency overmode
operating the prototype TC in theinglobal
are not considered range.Further
this paper. Due to the relatively
investigations oncomplex
this issuedesign
shouldmethod
be
carried out.
for HMPSTs and the article length restrictions, a sensitivity analysis of the weighting factors of the
operating mode are not considered in this paper. Further investigations on this issue should be
6. Conclusions
carried out.
In this paper, a HMPST was used to replace the existing TC for improving the fuel economy of
6. Conclusions
wheel loaders. The initial selection for the HMPSTs was conducted based on the lever diagram. It
was found that only four solutions can be applied to a wheel loader. Furthermore, the basic
In this paper, a HMPST was used to replace the existing TC for improving the fuel economy of
characteristics of these four solution were analyzed, and the results show all four solutions cannot
wheel loaders. The initial selection for the HMPSTs was conducted based on the lever diagram. It was
show efficiency advantages over the prototype TC at all speed ratio points, which makes a design
foundoptimization
that only four solutions
of each solution can be applied to a wheel loader. Furthermore, the basic characteristics
necessary.
of these four
The optimized results show thatand
solution were analyzed, onlythe results
solution show
c can all afour
display fuel solutions cannot
savings benefit show efficiency
compared with
advantages over the
the prototype TC.prototype
Despite the TC at all
large speed ratio
numbers points,
of possible which makes
configurations fora the
design optimization
HMPSTs, of each
the feasible
solution necessary.
solution replacing the TC is considerably less than expected. This situation may be due to the form
of the
The prototyperesults
optimized TC. If the
show basic characteristics
that only solution of cthe prototype
can display TC could
a fuel be altered,
savings benefitmore feasible with
compared
solutions could applied to wheel loader. The effectiveness of solution c with
the prototype TC. Despite the large numbers of possible configurations for the HMPSTs, the feasible the optimal structural
parameters
solution replacingis validated
the TC is inconsiderably
the Matlab/Simulation
less than environment.
expected. ThisThe simulation
situation may resultsbeshow
duethat theform
to the
solution c presents a fuel savings rate with 3.38% compared to the prototype TC. It is expected that
of the prototype TC. If the basic characteristics of the prototype TC could be altered, more feasible
the optimized HMPST solution c can be an ideal alternative to the existing TCs.
solutions could applied to wheel loader. The effectiveness of solution c with the optimal structural
parameters is validated
Acknowledgments: inwork
The the Matlab/Simulation
presented in this paper isenvironment. The simulation
funded by the National results
Natural Science show that
Foundation of the
solution c presents
China a fuel savings rate with 3.38% compared to the prototype TC. It is expected that the
(No. 51375505).
optimized HMPST solution
Author Contributions: c can
Xiaojun Liubeconceived
an idealthe
alternative to the
initial concept andexisting
developedTCs.the design methodology, as
well as wrote the paper. Dongye Sun produced the mathematical model and SIMULINK program. Datong Qin
and Junlong Liu The
Acknowledgments: workthe
conceived presented
structureinand
this paper direction
research is fundedofby
thethe National Natural Science Foundation of
paper.
China (No. 51375505).
Conflicts of Interest: The authors declare no conflict of interest.
Author Contributions: Xiaojun Liu conceived the initial concept and developed the design methodology, as well
as wrote the paper. Dongye Sun produced the mathematical model and SIMULINK program. Datong Qin and
Junlong Liu conceived the structure and research direction of the paper.
Conflicts of Interest: The authors declare no conflict of interest.
Energies 2017, 10, 1267 19 of 20

References
1. Fitriani, H.; Lewis, P. Comparison of Predictive Modeling Methodologies for Estimating Fuel Use and
Emission Rates for Wheel Loaders. In Proceedings of the Construction Research Congress 2014, Atlanta, GA,
USA, 19–21 May 2014.
2. Bohman, M. On Predicting Fuel Consumption and Productivity of Wheel Loaders. Master’s Thesis,
Lulea University of Technology, Lulea, Sweden, 2006.
3. Nilsson, T.; Fröberg, A.; Åslund, J. Development of look-ahead controller concepts for a wheel loader
application. Oil Gas Sci. Technol. 2015, 70, 159–178. [CrossRef]
4. Schneider, M.; Koch, O.; Weber, J. Green Wheel Loader–Improving Fuel Economy through Energy Efficient
Drive and Control Concepts. In Proceedings of the 10th International Fluid Power Conference (10. IFK),
Dresden, Germany, 8–10 March 2016; Volume 2, pp. 63–78.
5. Renius, K.T.; Resch, R. Continuously Variable Tractor Transmissions. In Proceedings of the ASAE Conference
on Agricultural Equipment Technology, Louisville, KY, USA, 14–16 February 2005; pp. 1–37.
6. Macor, A.; Rossetti, A. Optimization of hydro-mechanical power split transmissions. Mech. Mach. Theroy
2011, 46, 1901–1919. [CrossRef]
7. Wang, H.; Sun, D.; Qin, D. A new continuously variable transmission system applied to transmission system
of the roadheader’s cutting unit. Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 2016. [CrossRef]
8. Jo, H.; Lim, W.; Park, Y.; Lee, J. Prediction of the performance of a split/circulated power transmission.
Proc. Inst. Mech. Eng. Part D J. Automob. Eng. 1999, 213, 235–244. [CrossRef]
9. Linares, P.; Méndez, V.; Catalán, H. Design parameters for continuously variable power-split transmissions
using planetaries with 3 active shaft. J. Terramech. 2010, 47, 323–335. [CrossRef]
10. Achtenova, G.; Budinsky, T. Analysis and Guidelines for Design of Efficient Power-Split Systems; SAE Technical
Paper; SAE International: Detroit, MI, USA, 2012.
11. Zhao, K.; Bai, X.; Liu, Y.; Chen, W.; Li, G. Simulation study on reasonable matching of powertrain system for
crawler bulldozer. China J. Highway Transp. 2013, 26, 183–190. (In Chinese)
12. Wang, H.; Sun, D. Optimal matching between a diesel engine and a PRHTS transmission. J. Braz. Soc. Mech.
Sci. Eng. 2017, 39, 3375–3387. [CrossRef]
13. Nilsson, T.; Fröberg, A.; Åslund, J. Predictive control of a diesel electric wheel loader powertrain.
Control Eng. Pract. 2015, 41, 47–56. [CrossRef]
14. Sakaguchi, S.; Kimura, E.; Yamamoto, E. Development of an Engine-CVT Integrated Control System;
SAE Technical Paper; SAE International: Detroit, MI, USA, 1999.
15. Oh, K.; Kim, H.; Ko, K.; Kim, P.; Yi, K. Integrated wheel loader simulation model for improving performance
and energy flow. Automat. Constr. 2015, 58, 129–143. [CrossRef]
16. Nezhadali, V.; Frank, B.; Eriksson, L. Wheel loader operation—Optimal control compared to real drive
experience. Control Eng. Pract. 2016, 48, 1–9. [CrossRef]
17. Zou, N.; Huang, H.; Zhang, E.; Dai, Q. Establish oriented operating terminals wheel loader duty cycle.
Trans. Chin. Soc. Agric. Eng. 2015, 31, 78–85. (In Chinese)
18. Filla, R.; Ericsson, A.; Palmberg, J.O. Dynamic Simulation of Construction Machinery: Towards an Operator
Model. In Proceedings of the International Fluid Power Exhibition 2005 Technical Conference, Las Vegas,
NV, USA, 16–18 March 2005.
19. Pettersson, K.; Krus, P. Design optimization of complex hydromechanical transmissions. J. Mech. Des. 2013,
135, 1–9. [CrossRef]
20. Lei, Z.; Qin, D.; Liu, Y.; Peng, Z.; Lu, L. Dynamic energy management for a novel hybrid electric system
based on driving pattern recognition. Appl. Math. Model. 2017, 45, 940–954. [CrossRef]
21. Hrovat, D.; Tobler, W.E. Bond graph modeling and computer simulation of automotive torque converters.
J. Frankl. Inst. 1985, 319, 93–114. [CrossRef]
22. Jandasek, V.J. The Design of Single-Stage Three-Element Torque Converter; SAE Technical Paper;
SAE International: Detroit, MI, USA, 1961.
23. Wang, W.; Song, R.; Guo, M.; Liu, S. Analysis on compound-split configuration of power-split hybrid electric
vehicle. Mech. Mach. Theroy 2014, 78, 272–288. [CrossRef]
24. Son, H.; Park, K.; Hwang, S.; Kim, H. Design methodology of a power split type plug-in hybrid electric
vehicle considering drivetrain losses. Energies 2017, 10, 437. [CrossRef]
Energies 2017, 10, 1267 20 of 20

25. Macor, A.; Rossetti, A. Fuel consumption reduction in urban buses by using power split transmissions.
Energ. Convers. Manag. 2013, 71, 159–171. [CrossRef]
26. Mantriota, G. Performance of a parallel infinitely variable transmissions with a type II power flow.
Mech. Mach. Theory 2002, 37, 555–578. [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/).