Index

MODELING AND PERFORMANCE IMPROVEMENTS OF A PRESSURE
COMPENSATED AXIAL PISTON PUMP
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY
BY
OKAN AYDOĞAN
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR
THE DEGREE OF MASTER OF SCIENCE
IN
MECHANICAL ENGINEERING
MAY 2018
Approval of the thesis:

submitted by OKAN AYDOĞAN in partial fulfillment of the requirements for the

degree of Master of Science in Mechanical Engineering, Middle East Technical
University by,
Prof. Dr. Halil KALIPÇILAR _________________

Dean, Graduate School of Natural and Applied Sciences
Prof. Dr. M. A. Sahir ARIKAN _________________

Head of Department, Mechanical Engineering
Prof. Dr. R. Tuna BALKAN _________________

Supervisor, Mechanical Engineering Dept., METU
Examining Committee Members
Assoc. Prof. Dr. Yiğit YAZICIOĞLU _________________

Mechanical Engineering Dept., METU
Prof. Dr. R. Tuna BALKAN _________________

Assoc. Prof. Dr. M. Metin YAVUZ _________________

Assist. Prof. Dr. Ulaş YAMAN _________________

Prof. Dr. Yücel ERCAN _________________

Mechanical Engineering Dept., TOBB ETÜ
Date: 31.05.2018
iii
I hereby declare that all information in this document has been obtained and
presented in accordance with academic rules and ethical conduct. I also declare
that, as required by these rules and conduct, I have fully cited and referenced
all material and results that are not original to this work.
Name, Last name : Okan AYDOĞAN
Signature :
iv
ABSTRACT

AYDOĞAN, Okan
M.Sc., Department of Mechanical Engineering
Supervisor: Prof. Dr. R. Tuna BALKAN
May 2018, 94 Pages
Analysis of internal dynamics of a pressure compensated axial piston pump has an

importance to predict the pump characteristics. In order to accomplish this mission
physical model is developed to simulate pump pressure-flow characteristics for a
given set of system parameters in Matlab/Simulink environment. The pump is
modeled as two main groups, namely, pumping section, and compensation section.
Pistons are orderly nested in cylinder block around shaft axis in the pump. In the
pumping section model, multi-piston models arranged with a phase shift according to
the piston number, 2/N. The model includes leakages from these pistons. Forces on
individual pistons are used to model the instantaneous hydraulic torque acting on the
swash plate. Individual flow rates generated by movements of the pistons determine
the instantaneous delivery flow of the pump. The pressure compensation section
consists of a three-way valve, a bias actuator, a control actuator, and a swash plate.
Dynamics of these components are also modeled in detail. Hence, the developed
v
model represents a handy computational tool to get pump dynamics for various
system parameters.
An experimental setup is built to measure discharge pressure and delivery flow of the
pump. In order to adjust discharge pressure of the pump, pump outlet port is loaded
with an adjustable needle valve which covers the entire working range. Simulation
results obtained by using the developed model are compared with experimental
results to verify the model. Using this model, various configurations are further
examined to obtain their effects on oscillations of pressure and flow.
The aim of this work is to obtain performance improvements related with geometric
dimensioning of pump internal parts. Three different configurations obtained by
setting the angle of kidney-shaped flow passage area on the valve plate are used in
the physical model to investigate their influence on the pump dynamics, especially
leakages, and oscillations in pressure, flow, and torque. Smoother pressure
transitions in piston chamber are obtained by utilizing line to line porting valve plate
as opposed to using the original valve plate. The improvement in piston chamber
pressure eliminates flow peaks when the piston aligns with the delivery port. As a
result of this improvement, the oscillation amplitude of delivery flow is reduced.
This improvement also reduces the hydraulic torque on swash plate due to pumping
action. Even, lower torques are obtained by utilizing a smaller trap angle than the
angles used in both original valve plate and line to line porting valve plate. Hence,
the control actuator areas can be reduced for the same working pressure, resulting in
lower package volumes for the pump.
Keywords: Swash Plate, Flow Oscillation, Torque Oscillation
vi
ÖZ
BASINÇ KOMPANZASYONLU EKSENEL PİSTONLU BİR POMPANIN

MODELLENMESİ VE PERFORMANS İYİLEŞTİRİLMESİ
AYDOĞAN, Okan
Yüksek Lisans, Makina Mühendisliği Bölümü
Tez Yöneticisi: Prof. Dr. R. Tuna BALKAN
Mayıs 2018, 94 Sayfa
Basınç kompanzasyonlu eksenel pistonlu bir pompanın iç dinamiklerinin analizi bu

pompanın karakteristiği öngörebilmek için öneme sahiptir. Bu misyonu başarmak
amacıyla set halinde verilen sistem parametrelerinin uygulandığı pompanın basınç–
debi karakteristiğinin benzetimi için Matlab/Simulink ortamında fiziksel bir model
geliştirilmiştir. Pompa, basınçlandırma bölümü ve dengeleme bölümü olmak üzere
iki ana grup olarak modellenmektedir. Pistonlar, pompanın içinde mil eksenin
etrafında silindir bloğu içinde düzenli yuvalanmaktadır. Basınçlandırma bölümü
modeli, piston sayısına göre bir faz farkı, 2/N, ile düzenlenmiş birçok piston
modelini içermektedir. Model bu pistonlara ait sızıntıları kapsamaktadır. Pistonların
üzerindeki kuvvetler, eğim plakasına uygulanan anlık hidrolik torkun modellenmesi
için kullanılmaktır. Pistonların hareketine bağlı meydana gelen ait tekil debiler,
pompa anlık debi karateristiğini belirlemektedir. Basınç dengeleme bölümü üç yollu
valf, değiştirme eyleyicisi, kontrol eyleyicisi ve eğim plakasından meydana
gelmektedir. Bu elemanların dinamikleri detaylı bir şekilde incelenmektedir.
vii
Dolayısıyla geliştirilen model, farklı sistem parametreleri için pompa dinamiğini elde
etmek için kullanışlı bir hesaplama aracını temsil etmektedir.
Pompanın çıkış basıncının ve çıkış debisinin ölçümü için bir test düzeneği
kurulmuştur. Çalışma basıncının ayarlanması için pompanın çıkış portu, tüm çalışma
aralığını ayarlayabilecek bir kısma valfiyle yüklenir. Geliştirilen modelden elde
edilen benzetim sonuçlarıyle test sonuçları modeli doğrulamak amacıyla
karşılaştırılmıştır. Ayrıca bu model kullanarak farklı konfigürasyonların basınç ve
debi salınımları elde edilerek farklılıkların etkileri incelenmiştir.
Bu çalışmanın amacı pompanın iç elemanlarını geometrik ölçülendirilerek

performans iyileştirmelerinin elde edilmesidir. Fiziksel modele valf plakası
üzerindeki kavisli kanal şeklindeki akış alanı açısı ayarlanarak elde edilen üç farklı
konfigürasyon uygulanarak bunların pompa dinamiğine, özellikle kaçaklar ve basınç,
debi ve tork salınımları, etkileri incelenmiştir. Piston odasındaki daha düzgün basınç
geçişleri, orijinal valf plakasının aksine uçtan uca valf plakasında elde edilir. Piston
basınç odasında elde edilen bu iyileştirme pistonun basma portu ile aynı eksene
geldiği anda oluşan debi tepesini yok etmiştir. Bu gelişmenin bir sonucu olarak, çıkış
debisi salınım genliği azalır. Bu iyileştirme ayrıca basınçlandırma hareketi sonucu
oluşan eğim plakası üzerinde hidrolik torku düşürür. Hatta, her iki uçtan uca valf
plakası ve orijinal valf plakasında kullanılan açılardan daha düşük geçiş açıları
uygulanarak daha düşük torklar elde edilir. Dolayısıyla kontrol eyleyici alanları aynı
çalışma basıncı için azaltılabilir; bu da pompa için daha düşük paket hacimlerine
olanak tanır.
Anahtar Kelimeler: Eğim Plakası, Debi Salınımı, Tork Salınımı
viii
To my family
ix
ACKNOWLEDGEMENTS
First of all, I would like to express my sincere appreciation to my thesis

supervisor Prof. Dr. R. Tuna BALKAN for his guidance throughout my thesis study.
I would like thank to Prof. Dr. Bülent E. PLATİN, Mr. Celal SELAMOĞLU,
and Mr. Şeref SET for their guidance and affords throughout the study. It was a
pleasure and honor to work with them.
The support of the Scientific and Technological Research Council of Turkey

(TÜBİTAK) should be acknowledged for funding me with National Scholarship
Program for MSc Students.
I wish to express my sincere thanks to my family.
Finally, I would like to express my gratitude to my beloved wife Ms. Elif

AYDOĞAN for her patience, support, and understanding throughout my Master’s
Science student life.
x
TABLE OF CONTENTS
ABSTRACT ........................................................................................................ v
ÖZ ..................................................................................................................... vii
ACKNOWLEDGEMENTS ................................................................................ x
TABLE OF CONTENTS ................................................................................... xi
LIST OF TABLES ........................................................................................... xiv
LIST OF FIGURES .......................................................................................... xv
NOMENCLATURE ......................................................................................... xix
CHAPTERS
1. INTRODUCTION .......................................................................................... 1
1.1 Background ........................................................................................... 1
1.2 Previous Researches.............................................................................. 3
1.3 Motivation and Objectives .................................................................... 3
1.4 Thesis Outline ....................................................................................... 4
2. VARIABLE DISPLACEMENT SWASH PLATE PUMPS .......................... 5
2.1 Types of Variable Displacement Swash Plate Pumps .......................... 6
2.1.1 Pressure Compensated Axial Piston Pump ............................. 6
2.1.2 Remote Pressure Compensated Axial Piston Pump ................ 8
2.1.3 Load-Sense Controlled Axial Piston Pump ............................ 8
2.1.4 Torque Controlled Axial Piston Pump .................................. 10
2.1.5 Proportional Controlled Axial Piston Pump ......................... 11
3. PRESSURE COMPENSATED AXIAL PISTON PUMP PRIMARY SUB-
COMPONENTS AND MODELING................................................................ 13
3.1 Definition of the Pump Primary Mechanisms ........................................ 13
3.2 Sub-component Kinematics and Pump Modeling .................................. 16
3.2.1 Kinematics of the Piston-Slipper Assembly .............................. 16
3.2.2 Leakage Flow Modeling of the Pump ........................................ 19
3.2.3 Valve Plate Geometry and Valve Plate Timing ......................... 25
3.2.4 Speed Limitations ...................................................................... 31
xi
3.2.5 Dynamics of the Three-Way Valve ........................................... 33
3.2.6 Dynamics of the Control Actuator ............................................. 37
3.2.7 Dynamics of the Bias Actuator .................................................. 38
3.2.8 Dynamics of the ith Pumping Piston........................................... 40
3.2.9 Dynamics of the Swash Plate Mechanism ................................. 42
3.2.10 Dynamics of the Pump ............................................................. 47
3.2.11 Dynamics of the Test Bench .................................................... 48
4. SIMULATIONS AND EXPERIMENTAL RESULTS ................................ 51
4.1 Pump Model ........................................................................................... 51
4.1.1 Piston Model .............................................................................. 51
4.1.2 Compensation Section Model .................................................... 52
4.2 Test Bench Model ................................................................................... 52
4.3 Internal Dynamics Simulations .............................................................. 52
4.3.1 Kinematic Analysis .................................................................... 52
4.3.2 Simulation of Single Piston Discharge Pressure Profile on Valve
Plate ............................................................................................... 54
4.3.3 Simulation of Single Piston Delivery Flow Profile ................... 56
4.3.4 Simulation of Single Piston Delivery Flow Profile at Rated
Pressure ............................................................................................... 57
4.3.5 Simulation of Single Piston Leakage Profiles ............................ 60
4.3.6 Simulation of Hydraulic Torque on Swash Plate Applied by
Single Piston ....................................................................................... 63
4.3.7 Simulation of Multi-Piston Pressure (Pd) ................................... 65
4.3.8 Simulation of Multi-Piston Flow (Qd) ....................................... 66
4.3.9 Simulation of Multi-Piston Flow, Qd, at Rated Pressure ........... 68
4.3.10 Simulation of Multi-Piston Hydraulic Torque on Swash Plate,
Ty ............................................................................................... 69
4.4 Simulations and Experimental Results ................................................... 72
4.4.1 Discharge Pressure Comparison of Test Results and Simulation
Results ............................................................................................... 72
4.4.2 Discharge Pressure-Delivery Flow Characteristics Comparison
of Simulation Results and Experimental Results ................................ 73
4.4.3 Discharge Pressure-Delivery Flow Characteristics Comparison
of Case Studies .................................................................................... 74
5. SUMMARY, CONCLUSIONS, AND FUTURE WORK ............................ 77
xii
5.1 Summary ............................................................................................ 77
5.2 Conclusions ........................................................................................ 78
5.3 Future Work ....................................................................................... 79
REFERENCES.................................................................................................. 81
APPENDICES
A. Matlab/Simulink Model of the Cylinder Block (Single Piston) ................... 87

B. Matlab/Simulink Model of the Compensation Section ................................ 88
C. Matlab/Simulink Model of the Test Bench .................................................. 89
D. Technical Properties of the Test Bench........................................................ 90
E. Data Acquisition System Hardware.............................................................. 94
xiii
LIST OF TABLES
TABLES
Table 1: A comparison of power densities of three different machines [2] ................. 2
Table 2: Geometric oscillation for an axial piston pump ........................................... 48
Table 3: The simulation cases .................................................................................... 60
xiv
LIST OF FIGURES
FIGURES
Figure 1: Swept volume of a pressure compensated axial piston pump while

discharge pressure is lower than the maximum full-flow pressure [9] ........................ 7

discharge pressure is equal to the rated pressure [9] .................................................... 7
Figure 3: The discharge pressure-delivery flow characteristic curve of a typical

pressure compensated axial piston pump [10] ............................................................. 8
Figure 4: Load-sense controlled axial piston pump configuration [9] ......................... 9
Figure 5: The discharge pressure-delivery flow characteristic curve of a typical load-

sense controlled axial piston pump [11]..................................................................... 10
Figure 6: The discharge pressure-delivery flow pressure characteristic curve of a

torque controlled axial piston pump [11] ................................................................... 11
Figure 7: Displacement volume vs input current of a pump [12] .............................. 12
Figure 8: General pump configuration [15] ............................................................... 14
Figure 9: A cross-sectional view of the pressure controlled axial-piston pump [17] 16
Figure 10: Piston geometry in the x-z plane [14] ....................................................... 17
Figure 11: Piston geometry in the x-y plane [14] ...................................................... 18
Figure 12: Control volume for the analysis of pressure development in the cylinder
[5] ............................................................................................................................... 21
Figure 13: Kidney port and manifold geometry [7] ................................................... 22
xv
Figure 14: Scheme for the determination of flow rate between slipper and swash
plate [5] ...................................................................................................................... 23
Figure 15: Ball and socket geometry (clearance exaggerated) [6] ............................. 25
Figure 16: Ball and socket movement [23] ................................................................ 25
Figure 17: A non-over centered valve plate showing different geometric zones [24]26
Figure 18: Trapped valve plate design [7] ................................................................. 28
Figure 19: Valve plate of Denison Model P6W-R5B-C10-00-M2 [25] .................... 29
Figure 20: Schematic diagram of an axial piston pump with pressure equalization
mechanism [27] .......................................................................................................... 31
Figure 21: Rexroth open-circuit machinery [29]........................................................ 33
Figure 22: Schematic of a three-way valve [30] ........................................................ 34
Figure 23: Flow forces on a spool valve due to flow leaving a valve chamber [31] . 36
Figure 24: Free body diagram of the control actuator ................................................ 37
Figure 25: Free body diagram of the bias actuator..................................................... 39
Figure 26: Forces acting on the ith piston in the x-direction [32] ............................... 42
Figure 27: Cylinder block and multi-piston [23] ....................................................... 43
Figure 28: Swivel torque force diagram [33] ............................................................. 43
Figure 29: Geometry of the swash plate [7] ............................................................... 45
Figure 30: Schematic of a pressure compensator of an axial-piston swash plate pump

[37] ............................................................................................................................. 47
Figure 31: Determination of the delivery line for modeling discharge pressure of the
pump ........................................................................................................................... 49
xvi
Figure 32: Circuit diagram of the pump test bench ................................................... 50
Figure 33: Pump test bench ........................................................................................ 50
Figure 34: Motion of the ith piston in a cycle while α=16°, η=2.5° ........................... 53
Figure 35: Motion of the ith piston in a cycle while α=0°, η=2.5° ............................. 54
Figure 36: The chamber pressure, Pi, of a piston in a cycle when Pd=100 bar .......... 55
Figure 37: The chamber pressure, Pi, of a piston in a cycle with original valve plate
and line to line porting design when Pd=100 bar ....................................................... 56
Figure 38: The delivery flow, Qdi, of a piston in a cycle on valve plate when Pd=100
bar............................................................................................................................... 57
Figure 39: The delivery flow, Qdi, versus angular position, θ, at the rated pressure on
the original valve plate ............................................................................................... 58
the line to line porting ................................................................................................ 59
Figure 41: Schematic of delivery flow direction for a piston on valve plate at the
rated pressure ............................................................................................................. 59
Figure 42: Leakage flow between the cylinder block and valve plate when P d=100
bar............................................................................................................................... 61
Figure 43: Leakage flow around a piston when Pd=100 bar ...................................... 61
Figure 44: Leakage flow through the slipper and swash plate when Pd=100 bar ...... 62
Figure 45: Ball and socket joint leakage flow when Pd=100 bar ............................... 63
Figure 46: Hydraulic torque on swash plate due to a piston when Pd=100 bar ......... 64
Figure 47: Hydraulic torque on swash plate due to pressure forces when Pd=100 bar
.................................................................................................................................... 64
xvii
Figure 48: One cycle pump pressure profile at 100 bar ............................................. 65
Figure 49: One cycle delivery flow rate, Qd, profile from start-up to 100 bar ........... 66
Figure 50: Individual delivery flows, Qdi, for each piston for various valve plate
geometries from start-up to 100 bar ........................................................................... 68
Figure 51: One cycle delivery flow rate, Qd, profile at rated pressure, 170 bar......... 69
Figure 52: Total hydraulic torque on swash plate from start-up to 100 bar ............... 70
Figure 53: Individual hydraulic torque on swash plate (N.m) for various valve plate
geometries at 100 bar ................................................................................................. 71
Figure 54: Discharge pressure of the pump at a cycle ............................................... 73
Figure 55: Pressure-flow characteristics of the pump ................................................ 74
Figure 56: Discharge pressure-delivery flow characteristics curves of case studies . 75
xviii
NOMENCLATURE
Distance between pistons spherical joint turning

a 2.8 mm
plane and swash plate in x-z plane
Ap Pumping piston area 23 mm2
Distance between swivel axis and swash plate in

b 2 mm
x-z plane
BSA Kidney area of cylinder block 7.6 mm2
Distance between pistons spherical joint turning

c 3 mm
plane and swash plate plane in x-y plane
crs Radial clearance between spool and sleeve 3.10-3 mm
Radial clearance between bias actuator and

crpc 5.10-3 mm
cylinder
Radial clearance between control actuator and
cspc 5.10-3 mm
cylinder
Cd Discharge coefficient 0.65 -
Cv Port velocity coefficient 0.98 -
Distance between swash plate axis and swash

d 2 mm
plate plane in x-y plane
db Diameter of bias piston 8.5 mm
dc Diameter of control piston 6.25 mm
dd Hole diameter through the piston 0.4 mm
dobr Hole diameter at side curved wall of rate piston 1.3 mm
dp Diameter of pumping piston 5.4 mm
dsp Diameter of spool land 1.84 mm
xix
Radial clearance between cylinder block and
hK 15.10-3 mm
piston
Radial clearance between cylinder block and

hB 6.10-3 mm
valve plate
Radial clearance between swash plate and

hG 6.10-3 mm
slipper
H Radial clearance between ball and socket 1.10-3 mm
Mass moment of inertia of yoke assembly with

I 3.8 kg.mm2
respect to the swivel axis
kb Stiffness of bias actuator spring 14 N/mm
ksp Stiffness of spool spring 20 N/mm
ld Length of the hole inside the pumping piston 5 mm
Distance of the axes of the bores of bias actuator

LR 15 mm
and control actuator from swivel axis
mb Mass of bias actuator 5 g
mc Mass of control actuator 7 g
mp Mass of pumping piston 2 g
msp Mass of spool 4 g
N Number of pumping pistons 9 -
od Initial overlap among control spool and sleeve 1 mm
Pa Atmospheric pressure 1 bar
xx
Pc Drain pressure 5 bar
Pr Reservoir pressure 1 bar
Ps Suction port pressure 5 bar
rbi Inner radius of the kidney port on cylinder block 9 mm
rbo Outer radius of the slot on valve plate 10.5 mm
rG Inner radius of slipper pad 1.8 mm
ro Outer radius of ball joint on pumping piston 3 mm
rp Metering radius of spool 1.3 mm
R Piston pitch radius of valve plate 10 mm
Rbi Inner radius of the slot on valve plate 9.5 mm
Rbo Outer radius of the kidney port on cylinder block 11 mm
RG Outer radius of slipper pad 3.3 mm
Vcb Minimum volume of the bias actuator 150 mm3
Vcc Minimum volume of the control actuator 50 mm3
Volume between delivery port of the pump and

Vmd 1 dm3
load
Maximum geometric displacement limit of bias
xbmax 4.7 mm
actuator
Maximum geometric displacement limit of
xcmax 4.7 mm
control actuator
𝛿 Indexing angle 0.18 rad
xxi
𝛿b Pre-compression of bias actuator spring 0.9 mm
𝛿sp Pre-compression of spool spring 1.6 mm
𝛿1 Half of the closed angle of the socket 2 rad

Half of the open angle of the ball joint to the
𝛿2 0.2 rad
socket inlet
 The bulk modulus of hydraulic fluid @ 35C° 1.42 GPa
α0 Swash plate angle at rest 0.28 rad
 Secondary swash plate angle 0.04 rad
 Dynamic viscosity of hydraulic fluid @ 35C° 0.017 Pa.s
 The mass density of the hydraulic fluid @ 35C° 850.5 kg/m3

The angle of a kidney-shaped flow passage on
 0.51 rad
the cylinder block
The angle of a kidney-shaped non-flow passage
′ 0.93 rad
on the cylinder block
 The angular speed of cylinder block rad/s
Viscous friction force due to the leakage on bias
fb N
actuator
Viscous friction force due to the leakage on
fcp N
control actuator
Fspring Bias actuator spring force N
FB The total force on the bias actuator N
FCP The total force on the control actuator N
FaK_i Inertia force acting on the ith pumping piston N
FAK_i The total force on the ith pumping piston N
FDK_i Pressure force acting on the ith pumping piston N

Viscous friction force due to the leakage at the
FTK_i N
pumping piston and cylinder
xxii
Length of the fluid column inside the cylinder
lk m
block
N The rotational speed of the shaft rpm
Qsi Volumetric suction flow at the ith piston m3/s
Qdi Volumetric delivery flow at the ith piston m3/s
Qd Total volumetric flow to the delivery manifold m3/s
Qs Total volumetric flow from the suction manifold m3/s

Volumetric leakage flow through the three-way
Qlt m3/s
actuator to the reservoir
Volumetric leakage flow through the clearance
Qlkb m3/s
between bias actuator and sleeve
Qlkc m3/s
between control actuator and sleeve
QL Volumetric flow through to the load m3/s
The volumetric flow between the suction port
Qsi m3/s
and ith cylinder
Volumetric leakage flow inward and outward
QSB direction between the cylinder block and valve m3/s
plate
QSG m3/s
between rate piston and cylinder
Volumetric leakage flow across the annulus

QSK m3/s
between pumping piston and cylinder block
Volumetric leakage flow across the ball and

QSPHERE m3/s
socket joint
Pd Pump discharge pressure Pa
Ppi ith piston chamber pressure Pa
Pccm Control actuator chamber pressure Pa
Pbcm Bias actuator chamber pressure Pa
xxiii
Vi The instantaneous volume of the piston chamber m3
xb Bias actuator displacement m
xc Control actuator displacement m
Displacement of the ith piston from the neutral

xi m
plane
xi̇ The velocity of the ith piston m
xï The acceleration of the ith piston m
xsp Spool displacement m
α Swash plate angle rad
xxiv
CHAPTER 1
1. INTRODUCTION
1.1 Background
Fluid power systems, which consist hydraulic systems and pneumatic systems,
become an indispensable part of the high-power control applications [1]. The power
density is the ratio of output power to the volume and the volume refers to the
physical space that devices take up on a shop floor [2]. Power densities of the various
powered devices are shown in Table 1. In order to compare the devices, output
power is selected as the same. As shown in Table 1, a piston pump can produce more
power than an AC Electric Motor and Diesel Engine for the same volume. Hydraulic
power is generally preferred when small package volume is the requirement of the
overall system, especially in aerospace industry, highway vehicle industry and in
many industrial applications. With a rough expression, small package volume means
less weight and inertia. The low inertia allows fast response to the system, which
means that higher gains and higher bandwidths are achieved by using hydraulic
actuators in servo circuits.
Maximum efficiencies of hydraulic fluid power devices (pumps and motors) are
between 85% and 95%. Although the efficiencies of hydraulic fluid power devices
are sufficient, the hydraulic fluid power systems work with lower efficiencies.
Hydraulic systems are generally powered by a single pump for supplying different
actuators and loads. If load potentials of those actuators are different, this condition
causes operation non-efficient. Another major reason of low efficiency is the
pressure drop across the valves.
1
Table 1: A comparison of power densities of three different machines [2]
Output Power Volume Power Density

Device
3
[kW] [m ] [W / m3]
Diesel Engine 375 1.500 250
AC Electric Motor 375 0.690 543
Piston Pump 375 0.055 6.818
By advancing hydraulic systems, hundreds of horsepower is controlled by signals in

the order of milliamp or millivolt. By the way, hydraulic systems are used under
various loading conditions at high frequencies. Heat is generated in all devices due to
the inefficiencies of systems. The hydraulic fluid transferred through the system
works also as cooling equipment for dissipating heat produced by the system in
hydraulic systems. In mobile applications, flexible routing of hydraulic power by
using hoses is a good benefit with respect to the engine driven systems. Lastly,
hydraulic fluid works as a lubrication element for the circuit components that
hydraulic fluid takes the maintenance duty spontaneously.
On the other hand, hydraulic systems have many handicaps to limit their application
areas. It is so hard to achieve zero-leak hydraulic systems. Because of this reason,
hydraulic systems may not be used for cleanroom applications. Secondly, filtration is
an important duty to protect valves or actuators against clogging and the hydraulic
equipment against wear. Working with hydraulic system needs careful handling. This
duty may be ignored by maintenance labor and the inspecting lists must be controlled
strictly not to use contaminated hydraulic fluid in the hydraulic system, especially for
aerospace industry. From manufacturing view, tight tolerances are needed to achieve
less leakage in the actuators or hydraulic components. These tight tolerances cause
high costs for the components. Lastly, hydraulic systems generally work with high
levels of noise.
2
1.2 Previous Researches
Variable displacement axial piston pumps are used generally in aerospace, military,
automotive, and mobile applications. The widespread usage of this pump in the
industry has attracted researchers’ attention about its dynamics in detail.
Lewis & Stern [3] described detailed approaches to formulate the transfer functions
of the hydraulic components and to get the dynamic responses of the components and
hydraulic circuits. Reethof [4] investigated the characteristics of positive-
displacement pumps and motors at steady state. Ivantysyn and Ivantysynova [5]
studied the principles, design performance, modeling, analysis, control, and testing of
the hydrostatic pumps and motors. They formulated dynamic leakage equations and
pressure distribution of a piston by integrating the Reynolds equation of lubrication.
Watton [6] presented flow through the various leakage gaps. Most of these flow
equations are related with a variable displacement pump components. In addition, he
investigated the flow losses and torque losses for an axial piston machine. Manring
[7] investigated the detailed design, analysis, and control of an axial piston pump. He
studied component level design fundamentals, different types of controlled axial
piston pumps, and performed their dynamic performance. Ma [8] presented
optimization of the cross angle on the dynamics of the pump. He put forward the
correlation between analysis and experimental results. Then he optimized key
dynamics (flow peak and pressure pulsation) by changing the cross angle on valve
plate.
In this study, a complete dynamic analysis is applied to project the discharge

pressure-delivery flow characteristics of the pump. Flow, pressure, and torque
characteristics are improved by applying various valve plate geometries.
1.3 Motivation and Objectives
The motivation of this work is to get performance improvements by geometric

dimensioning of the pump internal parts. Examination of the boundaries related with
working conditions is aimed in order to determine pressure and flow rate pulsation
3
limits. The objective of this thesis is to analyze the pump dynamics by producing
measured and non-measured operational variables; for example, the discharge
pressure peak, the pulsation of the discharge pressure, the flow peak, the torque peak
on the swash plate and affordability of instantaneous total torque by control actuator
are predicted by these simulations. The performances of the different pump
configurations are investigated by utilizing various parameters at the physical model.
1.4 Thesis Outline
The content of this thesis is an investigation of dynamics of the pressure

compensated axial piston pump including interior parts, such as bias actuator, control
actuator, swash plate, valve plate, etc. Firstly, leakage paths are examined in order to
determine the dominant path. Volumetric efficiency can be improved especially by
decreasing the dominant leakage path. Secondly, dynamics of the bias actuator, the
control actuator, the three-way valve, pumping pistons, the valve plate, and the
swash plate are inspected. Case studies are done to improve the performance of the
pump, such as torque peak, torque oscillations, pressure peak, pressure oscillations,
flow peak, flow oscillations, etc. A coupled model is developed in Matlab/Simulink
environment to model pump. Then, measured experimental results and simulated
results are compared to cross-check the model. Finally, case studies are inspected to
visualize the improvements on pressure-flow characteristics.
4
CHAPTER 2
2. VARIABLE DISPLACEMENT SWASH PLATE

PUMPS
Pumps or motors, whose working principle is related with displacement volume, are
called as displacement machines. These displacement machines work under different
phenomena. These displacement machines are classified as, piston machines, gear
machines, screw machines, vane machines, and other machines. These machines are
named via the component constitutes the displacement. The main distinction between
piston machines and other machines is the direction of the motion. The piston
machines working phenomena rely on translational motion of the displacement
component. The displacement component performs rotary motion at the gear
machines, screw machines, and vane machines. The piston type machines are
classified as, axial piston machines and radial piston machines. The difference
between axial piston machine and radial piston machines is motion of the pistons. In
axial piston machines, pistons work parallel to the each other. In radial piston
machines, the pistons are arranged around the driving shaft. Motion of each piston is
perpendicular to the shaft axis. The axial piston type machines are classified as,
swash plate machines and bent axis machines according to the principle of
generation of the piston stroke. The piston stroke is generally produced due to the
support of the piston on the swash plate machines [5].
In swash plate pump usage market, pressure compensated axial piston pumps are
utilized usually due to their simplicity and cost-effectiveness. In this thesis, the
internal dynamics and the pressure-flow characteristics of a pressure compensated
axial pump is examined in detail.
5
2.1 Types of Variable Displacement Swash Plate Pumps
Variable displacement pumps have an adjustable swept volume. Variable

displacement volume is useful for control systems. Employing a variable
displacement pump is more energy efficient because flow and pressure closely adapt
to the load. Favorable efficiency (higher than 90%) can be achieved by employing
swash plate pumps; which leads to low operating costs.
Swash plate angle, α, of a variable displacement axial piston pump is controlled by a

control actuator. This control actuator is powered by a signal that can be hydraulic,
mechanical or electrical. Desired displacement volume can be accomplished by
energizing the control piston.
There are several types of displacement/pressure control mechanisms. Some of these

mechanisms are explained in following sections.
2.1.1 Pressure Compensated Axial Piston Pump
Pressure compensator mechanism limits the discharge pressure to the rated discharge
pressure, like a relief valve. When discharge pressure is below the compensator
setting pressure, the pump is at full stroke in other words displacement volume is at
most as shown in Figure 1. When the hydraulic discharge pressure reaches the rated
discharge pressure, displacement volume is maintained to the minimum level as
shown in Figure 2. The discharge pressure-delivery flow characteristic curve of a
typical pressure compensated axial piston pump is shown in Figure 3. When the load
is at the minimum level, maximum delivery flow occurs. When the pump is
pressurizing the hydraulic system up to compensator setting pressure, flow decreases
due to volumetric losses. Volumetric losses increase with increasing discharge
pressure. It is possible to operate at various discharge pressure magnitude up to rated
pressure. Spring losses occur while the discharge pressure is between maximum full-
flow pressure and rated discharge pressure. The pump maintains rated discharge
pressure with the minimum delivery flow until the load drops.
6
discharge pressure is lower than the maximum full-flow pressure [9]

discharge pressure is equal to the rated pressure [9]
7
Figure 3: The discharge pressure-delivery flow characteristic curve of a typical
pressure compensated axial piston pump [10]
2.1.2 Remote Pressure Compensated Axial Piston Pump
Using a remote pressure compensator is another way to control the pump

displacement according to the actual need of the flow for various pressure
magnitudes which are determined by a remotely installed pilot valve. When the
system pressure is reached to the relief valve setting pressure, control piston adjusts
the swash plate angle to the minimum value. The pump maintains the relief valve
setting pressure until load drops. The difference between standard pressure
compensator and remote pressure compensator is that pressure limiter adjustment is
controlled with an external relief valve in remote pressure compensator.
2.1.3 Load-Sense Controlled Axial Piston Pump
Load-sense control is another way of controlling the adjustment of the swash plate
angle. Pump discharge pressure is synchronized to the system demand at a pressure
which is little above the system demand. This adjustment is triggered by sensing the
system maximum hydraulic load, then adjusting the displacement volume to cope
with the system maximum hydraulic load. When there is no load, pump works at full
8
displacement. When there is a load, the pump discharge pressure increases until it
overcomes the differential spring force and shifts the control actuator to adjust the
displacement volume as seen in Figure 4.
Figure 4: Load-sense controlled axial piston pump configuration [9]
The pump carries a displacement level, which keeps the pressure drop constant
across the hydraulic load. This pressure drop is equal to the differential spring load.
Discharge flow increases when the hydraulic load decreases, or vice versa. For a
unique hydraulic load, pump delivery flow does not vary with respect to the varying
driveshaft speed. Because load-sense control maintains a constant pressure drop
across the orifice, the load-sense pump will maintain the same delivery flow various
driveshaft speeds. When the hydraulic load reaches the maximum set pressure,
control actuator adjusts the displacement volume to the minimum. The discharge
9
pressure-delivery flow characteristic curve of a typical load-sense controlled axial
piston pump is shown in Figure 5.
Figure 5: The discharge pressure-delivery flow characteristic curve of a typical load-

sense controlled axial piston pump [11]
2.1.4 Torque Controlled Axial Piston Pump
Torque control is another way to control the adjustment of the swash plate angle.
This type of control is beneficial when the available power for the hydraulic system
is limited. The intent of this type of pump control is to use the available input power
most efficiently. Torque limiter control adjusts the displacement volume as the
hydraulic load changes, to maintain a constant required torque. The discharge
pressure-delivery flow characteristic curve of a particular torque controlled axial
piston pump is shown in Figure 6. Various delivery flows are seen for a particular
pressure (ex. 250 bar) with respect to the changing supply input power.
10
Figure 6: The discharge pressure-delivery flow pressure characteristic curve of a
torque controlled axial piston pump [11]
2.1.5 Proportional Controlled Axial Piston Pump
Another control option is the proportional displacement control. At this type of

control, displacement volume is proportional to the input voltage or input current.
Discharge pressure does not have an effect on the determination of the displacement
volume. Discharge pressure can be controlled by employing a pressure transducer
with the pump. A linear variable displacement transducer (LVDT) measures the
swash plate angle, so the displacement volume is measured with respect to this
feedback signal. A pressure compensator also exists in this type of pumps. This
pressure compensator limits the rated discharge pressure.
Proportional control gives chance to switch the pressure and tank ports by changing
the direction of the displacement volume. As seen in Figure 7, direction and
magnitude of displacement volume vary with respect to the changing direction and
magnitude of the current. These types of pumps are suitable to use for hydrostatic
drives in closed circuits. Displacement volume and shaft speed determine the flow
magnitude up to set pressures of relief valves. Pump protects itself by using two
relief valves on each pressure ports. This type of pumps can be controlled by
11
proportional hydraulic/electronic control units or mechanical servomechanisms. With
the help of electronic control units, discharge flow can be determined precisely by
well-adjusting displacement volume. By changing current direction and magnitude,
flow direction and flow magnitude can be determined so smoothly. EP3 and EP4
denote that various powering options.
Figure 7: Displacement volume vs input current of a pump [12]
12
CHAPTER 3
3. PRESSURE COMPENSATED AXIAL PISTON PUMP

PRIMARY SUB-COMPONENTS AND MODELING
In the previous chapter, general information is given about variable displacement

axial piston pumps. Mechanics and discharge pressure-delivery flow relations are
given in this chapter in component level.
3.1 Definition of the Pump Primary Mechanisms
Hydrostatic pumps are a component of hydraulic circuits that have been used widely
in the aircraft industry, industrial and mobile applications. Several researches are
made for sizing the critical components to design the response of the pump and
overall hydraulic system recently. Kim et al [13], Schoneau et al [14], and Manring
[7] have published documents related with the dynamics of the swash plate variable
displacement pumps.
Typical configuration of a pressure compensated axial piston pump is given in Figure

8. Several pistons, generally odd number, are nested in cylinder block around x-axis.
Pistons are connected to the slipper with ball and socket joint. This slipper-piston
assembly is connected to a circular shaped plate, retainer. Circular shaped plate is
connected to the swash plate with a connection that gives permission to the revolute
motion around the x-axis. Inclination degree of swash plate around z-axis, α, is a
variable to determine displacement volume. This angle is named as swash plate angle
that varies with respect to the system dynamics. Inclination degree of swash plate
around y-axis, η, is also a parameter to determine the displacement volume. This
inclination angle is named as secondary swash plate angle that is constant. The swash
13
plate-piston mechanism is assembled with cylinder block. This assembly gives
permission to the translational motion of the pistons. Cylinder block is located
against a valve plate and connected to the shaft with involute spline. As the cylinder
block is driven around x-axis, each piston is forced to rotate around x-axis by
cylinder block. Pistons cross over the intake port and discharge port. Because of the
swash plate angle, α, and the secondary swash plate, η, pistons do oscillatory motion
in and out of the cylinder block.
Figure 8: General pump configuration [15]
In order to explain the working mechanism of the pump in detail, a piston in the
cylinder block considered. When the fluid-filled chamber is at the maximum level in
cylinder block for a piston, this position is named as the top dead center, TDC. When
the fluid-filled chamber is at the minimum level in cylinder block for a piston, this
position is named as the bottom dead center, BDC. This means that piston is ready to
suck fluid to fill its volume up to TDC. After piston passes the TDC position, fluid-
filled chamber tends to decrease volume. The fluid, which is in the closed chamber,
is pressurized by this motion of the piston. Then, piston discharges fluid until it
reaches to the BDC. This process explains one complete cycle for one piston. This
motion is completed in every cycle, so pump working frequency can be calculated by
shaft frequency multiplied by the piston number.
14
The displacements of a piston at BDC and TDC positions are nearly equal when
swash plate angle, α, equals to the minimum and secondary swash plate angle, η,
becomes driver angle for displacement volume. This condition occurs when the
pump is working at rated discharge pressure. Because of the secondary swash plate
angle in the x-y plane as shown in Figure 11, piston gets minimum chamber volume
at the mid position of the delivery port in y plane.
Secondary swash plate angle, η, is used to limit the minimum displacement volume.
Pressure compensation mechanism needs a minimum volume (approximately 4% of
ΔVmax) to maintain a stable operation. Also, the pump needs this minimum volume
for self-lubrication [5]. This secondary swash plate angle, η, is provided to decrease
the excessive noise and to improve efficiency of the pump [16].
Swept volume is explained as the total displacement of the pumping pistons in

previous researches. This condition is valid when the pumping piston is connected to
the delivery port at BDC and when the piston is connected to the suction port at
TDC. If these conditions are not satisfied, the pump could not use suction and
discharge potentials efficiently. Although many researches have been published
related with pumps, little attention is mentioned about the importance of this
condition. Efficient use of suction and discharge potentials of the pump may be
satisfied in the experiments of these researchers, otherwise the definition of the swept
volume includes an error related with the potential swept volume and actual swept
volume. It is important to mention that when chamber volume starts to increase, the
piston must be closed to the delivery port to prevent backflow from delivery line.
Similarly, when chamber volume starts to decrease, the piston must be closed to the
suction port to prevent backflow to the suction line.
The pressure compensation mechanism consists of a three-way valve, bias actuator,

bias spring, and control actuator. The control mechanism of a typical pressure
compensated pump is shown in Figure 9. Swash plate angle, α, is controlled by the
forces of the control actuator and bias actuator. The displacement volume is altered
by the adjustment of the swash plate angle, α. When pump discharge pressure gets
higher than maximum full-flow pressure, valve spool moves towards the valve
15
spring. By this actuation, flow is oriented through the control actuator chamber.
Hence pressure is built up inside the control actuator chamber. This pressurized fluid
powers the control actuator to rotate the swash plate mechanism against the bias
actuator force constituted by the bias spring.
Figure 9: A cross-sectional view of the pressure controlled axial-piston pump [17]
The adjustment of pressure compensation mechanism tends to de-stroke swash plate

mechanism when the discharge pressure is higher than maximum full-flow pressure.
In contrary view, bias actuator tends to stroke swash plate mechanism when
hydraulic system pressure falls down lower than the maximum full-flow pressure.
3.2 Sub-component Kinematics and Pump Modeling
3.2.1 Kinematics of the Piston-Slipper Assembly
Displacement of a piston at x-direction is shown in x-z plane in Figure 10 and x-y

plane in Figure 11. The pump has N number pistons and these pistons are equally
spaced in the cylinder block. In Matlab/Simulink model, these pistons are modeled
completely similar to a phase shift. All variables (piston displacement, piston
16
velocity, flow, leakage, pressure, etc.) are similar for each piston for a particular
angle, θ, in y-z plane by assuming the pump is in the steady-state and components of
the pump are identical.
Figure 10: Piston geometry in the x-z plane [14]
Displacement of the ith piston in the x-direction is given in Equation 1 with respect to
the varying angular position of the piston, θ, in y-z plane and swash plate angle, α, in
the x-z plane. The perpendicular distance from piston-slipper ball joint to swash plate
is defined as “a” in the x-z plane. The perpendicular distance from swash plate to the
swivel axis is defined as “b” in the x-z plane. Pitch radius of the cylinder block is
defined as R.
a−b
xi = + R sin(θ) tan(α) (1)
cos α
Displacement of the ith piston in the x-direction is given in Equation 2 with respect to
the varying angular position of the piston, θ in the y-z plane and constant secondary
swash plate angle, η in the x-y plane. The perpendicular distance from swivel axis to
17
swash plate is defined as c in the x-y plane. The perpendicular distance from piston-
slipper ball joint to the swash plate is defined as d in the x-y plane.
Figure 11: Piston geometry in the x-y plane [14]
c−d
xi = + R sin(θ) tan(η) (2)
cos η
In order to express the total displacement of the ith piston in Equation 3, it is needed
to equalize these two motions phases. So phase difference, /2, between these two
motions are added into Equation 2 with considering the direction of the motion.
a−b c−d π
xi = + + R tan(α) sin(θ) + R tan(η) sin (θ + ) (3)
cos α cos η 2
The velocity of the ith piston in the x-direction is given in Equation 4 with respect to
the constant angular speed of the cylinder block, , varying angular position of
piston, θ, in the y-z plane, varying swash plate angle, α in the x-z plane and constant
secondary swash plate angle, η, in x-y plane. (θ̇ = ω)
18
sin(α) α̇ sin(θ) α̇
ẋ i = (a − b) 2
+ R tan(α) cos(θ) ω + R
cos (α) cos 2 (α)
π (4)
+ R cos (θ + ) tan(η) ω
2
Acceleration of the ith piston in the x-direction is given in Equation 5 with respect to
the constant angular speed of the cylinder block, , varying the angular position of
piston, θ, in y-z plane, varying swash plate angle, α in x-z plane and constant
secondary swash plate angle, η, in x-y plane. (θ̇ = ω)
1 + 2tan2 (α)α̇ 2 sin(α) α̈

ẍ 𝑖 = { + } (a − b)
cos(α) cos2 (α)
2Rα̇ 2 sin(α) Rα̈
+{ + + R tan(α)ω2 } sin(θ)
cos3 (α) cos 2 (α) (5)
2Rα̇ 2 ω π
+ 2
cos(θ) − R sin (θ + ) tan(η) ω2
cos (α) 2
3.2.2 Leakage Flow Modeling of the Pump
Different from the discharge flow of the pump, continuous leakage flows occur in the
pump while the pump is working. Although these leakage paths help to the pump for
self-lubrication, in analytical view these leakages decrease the pump volumetric
efficiency, ηv , as seen in Equation 6 [18].
Qd
ηv = (6)
Vd ω
where,
Qd: delivery flow (m3/s)
19
Vd: displacement volume (swept volume) (m3)
 : angular velocity of the shaft (rad/s)
While defining the leakage relations, following assumptions are made;
 The hydraulic fluid inertia is negligible,

 The hydraulic fluid is Newtonian,
 The hydraulic fluid is incompressible with nobody forces,
 Pressure gradient exists only along the direction of the flow,
 Eccentricity about piston and sleeve is neglected.
3.2.2.1 Leakage around a Piston
The leakage flow equation between pumping piston and cylinder block is given in
Equation 7 by considering steady flow in annulus between piston and sleeve, minus
flow related with the movement of the piston as shown in Figure 12. Piston and
cylinder block eccentricity is assumed as zero. Instantaneous leakage flow path
length is denoted as lk which is depended to the varying xi . Clearance between
piston and cylinder block is denoted as hK. Diameter of the pumping piston is
denoted as dp. Pressure inside the ith chamber and pump drain line is denoted as
Pi and Pc , respectively.
πdp h3K πdp hK ẋ i

QSK = (Pi − Pc ) − (7)
12μlk 2
Bergada et al. [19] made detailed research about leakages inside the pump in
previous studies. They presented leakage flow equation around a piston including the
effects of the grooves cut on the piston.
Blackburn [4] presented steady flow in the annulus by including eccentricity. The
flow rate is increased 2.5 times the value of the concentric cylinders, assuming the
same pressure drop.
20
Figure 12: Control volume for the analysis of pressure development in the cylinder
[5]
3.2.2.2 Leakage Flow between Cylinder Block and Valve Plate
Flow rate in radial flow conditions is defined in Equation 8 by integrating Reynolds

equation in radial direction. The fluid film thickness is shown as hB in Figure 12.
πhB 3
Qradial = (P − Pc ) (8)
6μ ln(rout ⁄rin ) i
Fully circumferential flow is considered through formulating Equation 8. While the

pump is operating, the kidney-shaped flow passage area on the cylinder block is
pressurized. Equation 8 is modified with respect to the angular dimension of the
kidney-shaped flow passage in Equation 9. Cylinder block slotted port angle, , is
denoted as by referencing cylinder block center line as shown in Figure 13. Inner
radii of kidney port on valve plate and cylinder block are denoted as Rbo and rbo,
respectively.
21
 h3B
Qopening angle = (P − Pc ) (9)
12μ ln(R bo ⁄rbo ) i
These leakages occur in radially inward and outward directions as seen in Figure 13.
Total leakage for the piston is defined in Equation 10. Inner radii of kidney port on
valve plate and cylinder block are denoted as rbi and Rbi, respectively. Although, the
film thickness varies with respect to the surface flatness and hydrodynamic bearing
conditions, for simplicity constant fluid film thickness is considered. This film
thickness is sufficient to prevent metal to metal contact between the cylinder block
and valve plate [20].
Figure 13: Kidney port and manifold geometry [7]
h3B 1 1
QSB = (Pi − Pc ) [ + ] (10)
12μ ln(R bo ⁄rbo ) ln(R bi ⁄rbi )
3.2.2.3 Leakage through the Slipper and Swash Plate
Flow rate in radial flow conditions is defined in Equation 8. Gap height between the
slipper and swash plate is defined as hG in Figure 14. Outer radius and inner radius
of the slipper is defined as R G and rG , respectively. The pressure inside the slipper is
22
defined as PG . Leakage flow rate through the slipper and swash plate is defined in
Equation 11.
πh3G
Qradial = (P − Pc ) (11)
6μ ln(R G ⁄rG ) G
Figure 14: Scheme for the determination of flow rate between slipper and swash
plate [5]
Steady flow through the pipe is defined in Equation 12. dd is the hole diameter inside
the piston. Hole length is defined as ld . PG is the pressure inside the slipper.
πd4d
Qpipe = (P − PG ) (12)
128μld i
Equation 11 and Equation 12 are substituted to present flow rate (QSG ) between
slipper and swash plate by considering zero leakage across the ball and socket joint
[5]. Slipper spin is not taken into account in Equation 13. Although hG varies
depending on the pump operating conditions, in this formulation this term is assumed
as constant for simplicity. Several researches have been done related with slippers. In
these researches, it is investigated that gap height is related with tilt angles, rotational
23
speed, facing surfaces flatness, machine working condition (pumping or motoring),
etc.
πh3G d4d
QSG = (Pi − Pc ) (13)
μ (6d4d ln( R G ⁄rG ) + 128h3G ld )
3.2.2.4 Ball and Socket Joint Leakage
Spherical journal bearing characteristics have been studied by different researchers.

These researches generally focus on friction torque characteristics of spherical
journal bearings. Meyer transformed Reynolds equation in spherical coordinates
[21]. This transformation gives permission to get pressure distribution on the
spherical bearing and fluid film thickness. Manring studied about volumetric flow-
rates and load carrying capacity characteristics of the various socket geometries
experimentally [22]. Although non-traditional spherical socket design, which is
explained in the article [22], and conical socket design permits higher load capacity
than classical design, measured bearing leakage at these various socket geometries is
higher than classical design.
Bearing leakage is defined in Equation 14. This equation is valid for non-rotational
motion between ball and socket, steady-state flow conditions, and ball and socket are
concentric [19]. H is the clearance between ball and socket geometry. The radius of
the ball is defined as ro as shown in Figure 15.
H3 H4
(Pi − Pc )π (ro + 12 )
6
QSPHERE = H tan(δ2 ⁄2)
(14)
μ (ro + ) ln( tan(δ ⁄2))
2 1
24
Figure 15: Ball and socket geometry (clearance exaggerated) [6]
In Equation 14, it is shown that variation in the clearance between ball and socket is
effective on the maximum limit of this bearing leakage to a high degree. For an
example, in order to limit piston leakage, Eaton firm suggests that maximum
translational motion of the piston with respect to the fixed shoe plate must not exceed
0.005 inches for industrial 220 series piston pump [23] as shown in Figure 16.
Figure 16: Ball and socket movement [23]
3.2.3 Valve Plate Geometry and Valve Plate Timing
A non-over centered valve plate is shown in Figure 17. These types of valve plates
are utilized when the direction of shaft rotation and direction of swash plate angle do
not change to the opposite direction. The valve plate shown in Figure 17 could be
used for clockwise rotation (view on the valve plate) due to slot geometries at zone 3,
25
and zone 8. Zone 4 and zone 5 show suction port. Zone 1 and zone 9 show discharge
port.
Differently, the suction port and the discharge port can twist at over center pumps by
the changing direction of swash plate angle. In this type of pumps, zone 4, and zone
5 are similar to the zone 1, and zone 9. The geometry shown at zone 1 and zone 9 is
named as Qweb. These webs are used to strengthen the valve plate. These webs are not
used in the y-axis (Figure 8) at the discharge port/suction port in order not to restrict
the flow while maximum flow occurs. Maximum flow occurs in this region because
the maximum translational motion of a piston occurs in this region related with the
swash plate angle, α.
Figure 17: A non-over centered valve plate showing different geometric zones [24]
Lastly, shaft rotation direction could be changed from clockwise to the

counterclockwise or vice-versa in bi-directional pumps. In addition to the over-center
pumps, zone 2 and zone 6 are similar to the zone 3 and zone 8 in this type of pumps.
The valve plate of a typical non-over center pump is shown in Figure 8. The angular
position of the ith piston is defined in Equation 15, where  is the angular speed of
cylinder block and N is the total number of pistons. Piston, where angular position, θ,
equals to zero over discharge port in the y-axis, is named as the first piston. The
26
angular position of the piston, , is defined with respect to the midpoint of the
delivery port.
(i − 1)2π
θ = ωt + (15)
N
Typical valve plate design is shown in Figure 18. While the pump is working,
cylinder port moves, and valve plate remains stationary. Discharge area, Ad, or
suction area, As, of a piston at an instantaneous angular position is given in Equation
16 with a rough approach. In this equation slot angle, φf, is considered as zero (no
slot geometry). Indexing Angle, , and angular dimension at BDC and TDC, , are
considered as same. Variation of the flow passage area is considered as linear at
transition regions.
PistonAd & As
π ′  π ′ 
Ad = BSA, As = 0 − + +δ+ <θ< − +δ−
2 2 2 2 2 2
π ′ 
( − +δ+ )−θ π ′  π ′ 
2 2 2
Ad = BSA , As = 0 − +δ− <θ< − +δ+
 2 2 2 2 2 2
π ′  π ′ 
Ad = 0, As = 0 − +δ+ <θ< + +δ−
2 2 2 2 2 2
π ′ 
(θ − ( + + δ − )) π ′  π ′ 
2 2 2
Ad = 0, As = BSA + +δ− <θ< + +δ+
 2 2 2 2 2 2
=
π ′  3π ′ 
Ad = 0, As = BSA + +δ+ <θ< − +δ− (16)
2 2 2 2 2 2
3π ′ 
(( − + δ + ) − θ) 3π ′  3π ′ 
2 2 2
Ad = 0 , As = BSA − +δ− <θ< − +δ+
 2 2 2 2 2 2
3π ′  3π ′ 
Ad = 0, As = 0 − +δ+ <θ< + +δ−
2 2 2 2 2 2
3π ′ 
(θ − ( + + δ − )) 3π ′  3π ′ 
2 2 2
Ad = BSA , As = 0 + +δ− <θ< + +δ+
{  2 2 2 2 2 2}
Indexing Angle, , is shown in Figure 18. The angular dimension between ports is
defined as  [8]. When  is bigger than , this design is named as trapped volume
design. When  is equal to the , this design is named as line to line porting. When
 is smaller than , this design is named as cross porting.
27
Figure 18: Trapped valve plate design [7]
A characteristic constant of the design is defined in Equation 17. Characteristic

constant, φ, can be positive, negative or zero in accordance with the design
parameters. When characteristic constant, φ, is positive, it means that piston is open
to the suction and discharge slots while it is crossing TDC and BDC for a zero-
indexed valve plate. Because of this reason, there is no possibility of cavitation for
positive values of characteristic constant, φ. On the other hand, when the
characteristic constant, φ, is negative, it means that piston is closed to the suction,
and discharge ports while it is crossing TDC and BDC for a zero-indexed valve plate.
For this reason, designs with negative characteristic constant, φ, may cause
cavitation. This condition can be seen at the trapped volume designed valve plates.
Manring suggested that characteristic constant, φ, should be higher than 0.25° [7].
 − 
φ = −δ (17)
2
In order to use swept volume efficiently, ⁄2 should be equal or higher than
(⁄2 − δ). If this condition is not satisfied, the piston will not be fully filled at
TDC. In contrary, piston chamber is closed to the delivery line while the end of the
28
kidney port is at (π + ⁄2 − δ). In order to empty the chamber, ⁄2 should be
equal or higher than (⁄2 − δ). If this condition is not satisfied, the piston will tend
to jeopardize the chamber until sweeping to the BDC. It is important to consider that
while piston starts to compress the chamber, the piston must be closed to the suction
port. Similarly, while piston chamber starts to increase, the piston must be closed to
the delivery port to prevent backflow.
In order to use swept volume efficiently and eliminate the cavitation, slots are
utilized at BDC and TDC as shown in Figure 19.
Figure 19: Valve plate of Denison Model P6W-R5B-C10-00-M2 [25]
Different types of slot geometries have been utilized to have smoother pressure rise,
and drop profiles. Common slots would be summarized as a combination of flat or
ramped geometries. The variation of the slot geometries is limited with the
production capabilities.
Manring studied to define pressure transition of various slot geometries (constant,

linear, and quadratic) at low displacements [15]. He investigated that if a short slot is
needed, linearly varying slot geometry is suitable. If maximum discharge area of the
slot is limited, the constant area slot geometry can be utilized. He figured out various
valve plate designs (line to line, cross porting, and trapped volume design) without
indexing for the same swash plate angle, α=9°, and rotational speed, n=2250 rpm, at
[7]. While piston pressure is decreasing to the inlet pressure, pressure transition (Ṗ
[GPa/s]) is about 95 for line to line porting, approximately 70 for cross porting, and
29
close to 95 for trapped volume design for the same pressure magnitude, 20 MPa, and
for overlap angle, Δ=3° for cross porting, and trapped volume design. While piston
pressure is increasing to the discharge pressure magnitude, pressure transition, Ṗ
[GPa/s]), is about 60 for line to line porting, approximately 50 for cross porting, and
close to 60 for trapped volume design for the same conditions. Pressure transition is
lower at cross porting comparing to the others. Lower pressure transition means
lower noise and vibration. It is important to mention that internal leakage occurs
from delivery port to the suction port across the pumping piston chamber in cross
porting design. This means that inefficiency occurs while the pump is reciprocating.
In addition to these researches, he made a comparison between trapped volume
design without slot and line to line design with flat bottom constant width slot
geometry [26]. This research shows that it is more efficient to use slot angular
volume as trapping volume. By the view of these researches, it is easy to say that
optimum valve plate design changes with respect to the applications.
Nafz et al analyzed a pressure equalization system with valve-controlled pressure

recuperation chamber (PRC) and it is shown that both the efficiency and the noise
emission reduction could be improved apparently [27]. Xu et al [27] investigated a
non-traditional approach to noise reduction method by utilizing pressure equalization
mechanism with a check valve and PRC as shown in Figure 20. Delivery flow peak,
swash plate torque oscillation, and efficiency of the pump are examined in this
detailed study. The major phenomenon of this study is to transfer hydraulic energy
stored in BDC to the TDC. This mechanism works while the first piston is at BDC
and its closed volume is jeopardized due to its shrinking volume. This energy is
transferred to the piston in TDC across check valves and PRC. This energy transfer
goes on one by one. The energy transfer from contrary direction is prevented by
utilizing check valves. Noise reduction is shown by considering flow peak reduction
and torque oscillation reduction. In addition, the total swept volume of the pump is
used effectively by employing this mechanism. Flow from BDC to the PRC and PRC
to the TDC decreases inefficiency because of the leakage of the piston around BDC.
The pressure at the piston, which is entering the discharge port, reaches to the
discharge pressure magnitude faster. Backflow from delivery line to the pump is
30
decreased by this way. Lastly, control actuator volume could be reduced by
decreasing required de-stroking torque by decreasing maximum torque exerted on
swash plate.
Figure 20: Schematic diagram of an axial piston pump with pressure equalization
mechanism [27]
3.2.4 Speed Limitations
Three well-known speed limitations related with axial piston machines are given in
the following sections. In addition to these limitations, tribology considerations
between piston and cylinder block, valve plate and cylinder block, and retainer and
swash plate might be examined.
3.2.4.1 Limitations Related with Cylinder Block
The pressure dynamics of a pump in a cycle is given in the previous sections. The
pumping pistons should not be run at the vacuum on whole valve plate in order not to
meet wear problems due to cavitation at low pressure.
Ivantysynova presented a pressure relief valve in cylinder-block to reduce pressure

peaks [5]. This relief valve limits the chamber pressure on the piston in the cylinder
block. By the way, pumping pistons are not jeopardized. This solution needs special
31
slots and space to implement valves inside the cylinder block. By the way, the force
on the cylinder block may be minimized.
Manring [28] investigated the analytical equation of cylinder block tipping away
from valve plate. He presented design criteria to eliminate this problem. The cylinder
block spring might be designed to satisfy this criterion. The speed limit of the pump
with no load condition is determined by inertial forces.
The cylinder block and the valve plate work facing one another with metal to metal
contact. It is suggested to produce these components from dissimilar materials. These
materials might have a different hardness to prevent adhesion. In general, the mutual
faces are coated to prevent wear. These materials might have also an ability to
dissipate heat in order to prevent softening.
Manring [7] presented the speed limitation equation due to the cylinder-block filling.
The flow passage area and suction port pressure of pump are examined in this
relation. The suction port pressure, Ps, or kidney area of the cylinder block, BSA,
might be increased in order to increase the maximum speed of the pump by
satisfying this equation. Decreasing the pitch radius, R, is another way to increase the
maximum speed.
3.2.4.2 Limitation Related with Socket Tipping
Similar to the cylinder block tipping, there exist slipper tipping restrictions. These
conditions can occur at high displacements (large inclination angles) and high shaft
speeds.
Manring [29] presented the analytical criteria for slipper tipping away from retainer
ring. Centrifugal inertia throws away the slippers due to rotational speed effects. The
inertial forces acting on the pistons and the slippers are included in this equation.
Speed scaling results of Rexroth open-circuit machinery is shown in Figure 21.

Scaling laws are utilized in order to determine speed and displacement volume
limitations.
32
Figure 21: Rexroth open-circuit machinery [29]
3.2.5 Dynamics of the Three-Way Valve
Schematic of a three-way valve, which is utilized to control volumetric flow through

to the control actuator, is shown in Figure 22. This valve contains generally a spool,
sleeve, and a spring. The left side of the valve, where the spring is mounted, and left
chamber side of the spool are connected to the reservoir line. The discharge pressure
of the pump, Pd, is used as a hydraulic signal to control spool position, xsp. This valve
limits flow through to the control actuator while discharge pressure is lower than
maximum full-flow pressure. Discharge pressure works against the spring force to
overcome initial overlap, od, towards the left chamber connected with reservoir line.
When the discharge pressure is higher than the spring force, the fluid flows through
to the control chamber.
Although the initial overlap is constant due to the design of the valve, pre-
compression of the spring is adjustable. The adjusting component of the valve is
shown in Figure 4. Maximum full-flow pressure is determined by adjusting this
component.
33
Figure 22: Schematic of a three-way valve [30]
In common usage, sealing equipment is not used to eliminate sticking forces related
with the sealing equipment. This type of valve configuration causes internal leakages
across the valve while discharge pressure is lower than maximum full-flow pressure.
Leakage equation between the right chamber and control chamber is considered as a
steady flow in the annulus between circular shaft and cylinder with no eccentricity
[4]. This yields the leakage as where Pccm is the pressure in the control actuator
chamber, dsp is spool diameter, crs is clearance between the sleeve and spool land (od-
xsp) is column length. Equation 18 is valid while od is larger than xsp. This leakage
decreases volumetric efficiency while pump discharge pressure is below maximum
full-flow pressure. Leakage term related with the spool velocity is neglected.
3
(Pd − Pccm )πdsp crs
Qsp = (18)
12μ(od − xsp )
By considering spool is zero-lapped, the flow rate is derived in Equation 19 while

discharge pressure is between maximum full-flow pressure and rated pressure. Flow
rate between the right chamber and control chamber is governed using the classical
orifice equation where AD is the opening area between spool and sleeve. Geometric
relation of AD is given in Equation 20 where θd is the half angle at the center of the
metering port opening of the spool at the right chamber as shown in Figure 22.
34
2(Pd − Pccm )
Qsp = AD Cd √ (19)
ρ
rp2 ( θd − sin(2θd )⁄2) xsp < (od + rp )

AD = { 2 }
rp ( θd + sin(2θd )⁄2) xsp > (od + rp )
(20)
Leakage flow from control actuator chamber to the reservoir port through the
clearance between sleeve and spool, Qlt, is valid while xsp is bigger than od, as given
in Equation 21. This yields the leakage as where spool diameter is dsp, the clearance
between spool and sleeve is crs, spool displacement is xsp, and initial overlap between
spool and sleeve is od. The direction of this leakage is shown in Figure 22. While od
is bigger than xsp, opening area, AR, at the left side of the metering port is big enough
to discharge leakage flow coming from the right chamber. When the discharge
pressure is lower than maximum full-flow pressure, the leakage flow regime at the
left metering port is negligible for the analysis for a zero lapped valve.
3
(Pccm − Pr )πdsp crs
Qlt = (21)
12μ(xsp − od )
The volumetric flow into the control actuator, Qcc, is shown in Figure 22 by
summing flows across the metering port. The relation is expressed in Equation 22.
Qcc = Qsp − Qlt (22)
Steady state flow produces axial force, F1, which tends to reduce the valve opening
area as shown in Figure 23. Axial force exerted on the spool is defined as
35
F1 = 2 Cd Cv AD (Pd − Pccm )cosθj (23)
where Cd is discharge coefficient, Cv is velocity coefficient, AD is orifice area, and j

is jet angle across the orifice.
Figure 23: Flow forces on a spool valve due to flow leaving a valve chamber [31]
The equation of motion of the spool shown in Figure 22 is defined as
Pd Asp − k sp δsp − F1 = msp ẍ sp + csp ẋ sp + k sp xsp (24)
where Asp is spool area, ksp is spring stiffness, 𝛿sp is initial pre-compression of the
spring, msp is mass of the spool, xsp is spool displacement, csp is damping coefficient.
Damping coefficient of the system is defined in Equation 25. The equation shows
that the system works like a simple mass-spring system and it is considered as
critically damped.
csp = 2√msp k sp (25)
36
3.2.6 Dynamics of the Control Actuator
The geometry and the flow configuration of the control piston are shown in Figure
24. Flow against to the control piston, Qcc, is given in Equation 22.
Leakage flow around the control piston is given in Equation 26. Leakage length, lc0,
is constant due to pump design. This yields the leakage as where pressure behind the
control piston is defined as Pccm, the diameter of control piston is dc, and the radial
clearance between control piston and sleeve is ccp, with considering no eccentricity.
Body forces of control piston mass are neglected.
Figure 24: Free body diagram of the control actuator
3
πdc ccp
Qlkc = (P − Pc ) (26)
12μlc0 ccm
The pressure of the control actuator chamber is given in Equation 27. Displacement
of control piston is denoted as xc. The bulk modulus of fluid is defined as  and
minimum volume of the control actuator piston is defined as Vcc.
37
β
Ṗccm = 2 (Qcc − Qlkc − (π⁄4) dc 2 ẋ c ) (27)
Vcc + dc xc (π⁄4)
The viscous force, fcp, acting on the control piston is given in Equation 28 according
to the given relation in [14] where ẋ c is the velocity of the control piston. The total
force acting on the control piston is defined in Equation 28.
μπ(dc /2)lc0 ẋ c
fcp = (28)
ccp
FCP = (π⁄4) dc 2 (Pccm − Pc ) − fcp (29)
3.2.7 Dynamics of the Bias Actuator
The geometry and the flow pattern of the bias actuator are shown in Figure 25.
Similar to the control actuator, pressure behind the bias piston is defined as Pbcm and
leakage around the piston is defined as Qlkb. The bias piston diameter is defined as
db. The leakage length, lb0 - xb, is variable depending on the displacement of the bias
piston, xb. Positive displacement direction of the bias piston is accepted as opposite
direction with respect to the direction shown in Figure 25. While the pump is
working below maximum full-flow pressure, displacement of the bias piston, xb, is at
the maximum limit. This yields the leakage as where the clearance between bias
piston, the guide is defined as crpc, the diameter of the bias piston is db, and the bias
actuator chamber pressure is denoted as Pbcm. Body forces of bias piston mass are
neglected. Leakage flow around the bias piston, Qlkb, is expressed in Equation 30.
Additionally, there is a hole on the side wall of this bias piston. This hole helps the
fluid to flow out from the bias piston chamber while unloading the pump. Flow
through this hole, Qor, is defined in Equation 31 where the diameter hole at the side
wall is denoted as dobr.
38
Figure 25: Free body diagram of the bias actuator
3
πdb crpc πdb crpc ẋ b
Qlkb = (Pbcm − Pc ) + (30)
12μ(lb0 − xb ) 2
πdobr 2 2(Pbcm − Pc )
Qor = Cd √ (31)
4 ρ
The pressure of the bias piston chamber is given in Equation 32, where the minimum
volume of the bias actuator is defined as Vcb.
β
Ṗbcm = 2 (−(π⁄4) db 2 ẋ b − Qlkb − Qor ) (32)
Vcb + (π⁄4) db xb
The bias piston is at maximum displacement, xb_max, while the pump is not working.
The viscous force, fb, acting on the bias piston is given in Equation 33 according to
the given relation in [14]. A compression spring is utilized inside the bias actuator
chamber as shown in Figure 9. While the pump is not working, this spring, kb, has a
39
pre-compression 𝛿b. The spring force, Fspring, on the bias piston is given in Equation
34 and total force acting on the control piston is defined in Equation 35.
μπ(db /2)(lb0 − xb )ẋ b

fb = − (33)
crpc
Fspring = k b (δb + (xbmax − xb )) (34)
FB = Fspring + (π⁄4) db 2 (Pbcm − Pc ) + fb (35)
3.2.8 Dynamics of the ith Pumping Piston
Figure 12 shows the control volume analysis of pressure development in a cylinder.

Qd is defined as instantaneous discharge flow out of a piston in Equation 36, where
Ad is defined in Equation 16. The discharge pressure of the pump is defined as Pd. Qs
is defined as instantaneous suction flow through a piston in Equation 37, where As is
defined in Equation 16. The pressure of the pump suction line is defined as Ps.
Leakage terms are defined in the previous sections. Qr represents the flow into and
out from the piston chamber through the valve plate opening [32]. Definition of Qr_i
for the ith piston is given in Equation 38.
2(Pi − Pd )
Qd_i = Cd Ad √ (36)
ρ
2(Ps − Pi )
Qs_i = Cd As √ (37)
ρ
40
Qr_i = Qd_i − Qs_i (38)
Instantaneous volume, Vi, is given in Equation 39, where Vm is maximum cylinder

volume. Instantaneous displacement of the ith piston is denoted as xi and piston area
is denoted as Ap.
Vi = Vm − Ap xi (39)
Individual chamber volumes might be defined for each piston, which is assembled in
a circular array inside the cylinder block at equal intervals around the x-axis. The
pressure in the cylinder is derived in Equation 40, according to the control volume
analysis.
β
Ṗi = (A ẋ − Qr_i − QSB_i − QSK_i − QSG_i − QSPHERE_i ) (40)
Vi p i
The angular position of the ith piston is given in Equation 15. Pressure with respect to
the angular position of the cylinder is given in Equation 41.
dPi β
= (A ẋ − Qr_i − QSB_i − QSK_i − QSG_i − QSPHERE_i ) (41)
dθ ωVi p i
The viscous force, FTK, acting on the ith piston is given in Equation 42 where ẋ i is the
velocity of the ith piston. The positive direction of the viscous force is considered as
the direction against the positive direction of ẋ i, which is shown in Figure 26. The
engaged length between piston and cylinder block is shown li when xi equals to zero.
μπ(dp /2)(li + xi )ẋ i

FTK_i = (42)
hk
41
Figure 26: Forces acting on the ith piston in the x-direction [32]
The inertia force, FaK, acting on the ith piston center of the piston-slipper group is
given in Equation 43 where the mass of a piston and slipper group is defined as mp.
FaK_i = mp ẍ i (43)
The pressure force, FDK, acting on the ith piston is given in Equation 44 where the
area of a barrel piston is defined as AP multiplied by differential pressure across the
barrel piston.
FDKi = AP (Pi − Pc ) (44)
The total force acting on the ith piston in the x-direction is given in Equation 45.
Centrifugal force, FK, acts in the y-direction on the piston.
FAK_i = FaK_i + FDK_i + FTK_i (45)
3.2.9 Dynamics of the Swash Plate Mechanism
Swivel torque due to pumping piston pressure is obtained by summation of the forces
on each piston which are nested in the cylinder block as shown in Figure 27.
42
Figure 27: Cylinder block and multi-piston [23]
Swivel torque occurs due to the forces acting on each pumping piston and actuators.
Figure 28 shows a diagram of forces exerted on the swash plate by a single piston
[33]. Cylinder block center axis in x-axis and swivel axis of the swash plate (parallel
to the y-axis) do not intersect as shown in Figure 28. The moment arm is defined in
Equation 46. The axial total force in the x-direction is shown in Equation 45. The
reaction of the ith piston, FSK, is defined in Equation 47.
Figure 28: Swivel torque force diagram [33]
Rsin(θ) + (a − b)sin(α)
Moment_Arm = (46)
cos(α)
43
FAK_i
FSK_i = − (47)
cos(α)
The expression for torque applied to the swash plate, Ty_i, by the ith piston is defined
in Equation 48. The total torque applied to the swash plate in the y-direction by the
pumping pistons is simply the summation as defined in Equation 49.
FAK_i (Rsin(θ) + (a − b)sin(α))

Ty_i = FSK_i Moment_Arm = − (48)
cos(α) cos(α)
Ty = ∑ Ty_i (49)
i=1
Control actuator has to have the ability to overcome total torque applied to the swash
plate in the y-direction in order to control the swash plate angle. Control actuator
area has to be designed by considering this restriction.
Torque on the swash plate around x-direction and z-direction must be carried by
swash plate bearings. The definition of these torques is explained with details in [5].
The force formed by the pressure, FDK , is overpowering the other force sources at
high pressures. The pressure in a barrel is being negative in a region on valve plate in
a cycle. The total hydraulic torque applied to the swash plate in y-axis by the
pumping pistons can be negative because of the pressure variations in a cycle. In
order to have smooth operation Inequality 50 must be satisfied while discharge
pressure is lower than the maximum full-flow pressure. Force constituted by the bias
actuator, FB, is shown in Equation 35 and moment arm is defined as LR. By
satisfying this condition, the total torque on the swash plate is being always positive.
This is very important to eliminate oscillations of swash plate in a cycle while
discharge pressure is lower than maximum full-flow pressure.
44
FB ∗ LR > abs(min(0, Ty )) (50)
Manring explained the various designed swash plate geometries in [7]. In this paper
design of a trunning mounted, transverse servo swash plate is investigated as shown
in Figure 29. Moment arms are the same for bias actuator and control actuator in this
type of pumps.
Figure 29: Geometry of the swash plate [7]
Schoenau et al [14] presented a complete pump model that includes individual

torques on swash plate. An effective mass moment of inertia is defined which
includes parameters apart from yoke assembly and variables related with operating
conditions. This relation is handled by substitution of the input torques into complete
pump model and grouping similar terms. In addition to the definition of effective
mass moment of inertia, complete pump model coefficients are determined.
Manring and Johnson [34] presented a dynamic model of the pumping system. In this
research, the equation of motion of the swash plate is presented by summing the
moments at the pivot axis. By considering the results of this research, it is eligible to
say that effects of the swash plate inertia and damping are negligible compared to the
stiffness of the control actuator on system characteristics. This means that motion of
the swash plate is highly dependent on the working conditions.
45
Manring [35] examined the control and containment forces acting on the swash plate.
Total containment force acting on the swash plate bearings is shown analytically. As
shown in the paper, the frequency of the forces is designated the number of pistons
multiplied by pump shaft frequency. It is shown that frequency of the torque on the
swash plate is twice of the force-frequency. Frequencies of the alternating forces and
torques must be considered in order to determine right bearings, housings, and
driving motor shaft.
Zhang et al presented a new swash plate damping model for hydraulic axial-piston
pump [36]. An analytical expression is given which is a function of discharge
pressure and swash plate angle.
The equation of motion of the swash plate mechanism around y-axis [34] is defined
in Equation 51. Mass moment of inertia of the swash plate is defined as I.
Iα̈ = (FB − FCP)LR + ∑ Ty (51)

i=1
The translational motion of the bias actuator and control actuator is in the opposite
directions as shown in Figure 30. Motions of bias actuator and control actuator are
defined in terms of swash plate angle in Equation 52 and Equation 53, respectively.
xb = αLR (52)
xc = −αLR (53)
46
Figure 30: Schematic of a pressure compensator of an axial-piston swash plate pump
[37]
3.2.10 Dynamics of the Pump
Pistons, which are coupled with ball and socket geometry, are arranged on the swash
plate plain surface as shown in Figure 9 and the fundamental working principles are
explained in Section 3.1. While a piston crosses over the suction port, it tends to
increase fluid volume of the pumping piston chamber up to TDC by sliding on the
inclined plane. The instantaneous suction flow of the pump is defined in Equation 54,
where Qsi is defined in Equation 37.
Qs = ∑ Qsi (54)
i=1
As a piston passes over the delivery port, it tends to decrease the fluid volume of the
pumping piston chamber up to BDC by sliding on the inclined plane. This process
occurs N times for each cycle of the pump. The instantaneous delivery flow of the
pump is defined in Equation 55, where Qdi is defined in Equation 36.
47
N
Qd = ∑ Qdi (55)
i=1
Watton [6] presented flow rate change of pump with respect to the angular position
of a piston. It has been seen that flow rate has unique oscillation that resembles half
sine wave at twice of the pump frequency (N x shaft frequency). Definition of flow
rate change is shown for both odd and even piston pumps. Geometric oscillations for
an axial piston pump for various piston numbers are shown in Table 2, where Vs is
swept volume for a piston. Flow oscillation for a pump, whose piston number is odd,
is the same for twice of this odd piston number. Because of this reason, an odd
number of pistons are utilized in axial piston pumps generally.
Table 2: Geometric oscillation for an axial piston pump
N 3 4 5 6 7 8 9 10
2 π δQ
0.14 0.325 0.05 0.14 0.0253 0.078 0.0153 0.05
N Vs 
3.2.11 Dynamics of the Test Bench
Ivantysynova presented various circuit diagrams in order to determine displacement

volume in open circuit, displacement volume in closed circuit, efficiency test of a
pump in open circuit, efficiency test of a pump in closed circuit and response time of
a pressure controlled pump in [5].
Functional requirements and qualification tests are explained in detailed in [10]. In

order to qualify a pump; dimensional check of principal parts, proof pressure test,
determination of response time test, pressure oscillation test, cavitation test, etc. must
48
be done. In addition to these qualification tests, this document also consist
acceptance tests.
The discharge pressure of the pump depends on the load condition of the hydraulic
circuit. Flow through the orifice in the direction of the reservoir is defined in
Equation 56, where reservoir pressure is defined as Pr and orifice opening area is
defined as Aod.
2(Pd − Pr )
QL = Cd Aod √ (56)
ρ
The discharge pressure of the pump is modeled using control volume approach. The
control volume is defined as Vmd, which covers from pump outlet to the needle valve.
The discharge line of the pump is shown in Figure 31. The discharge pressure is
given in Equation 57. Leakage term covers various flow paths, which are explained
in Section 3.2.2. These flow paths are given for individual pistons. Total internal
leakage of the pump is simply summation of these terms with leakage due to pressure
compensator mechanism.
Figure 31: Determination of the delivery line for modeling discharge pressure of the
pump
β
Ṗd = (Q − QL − Leakage) (57)
Vmd d
49
As seen from Figure 32, a needle valve is utilized in order to load pump. Opening
area, Aod, of this valve is controlled in order to control the discharge pressure of the
pump. The hydraulic line between the test pump and the needle valve is constructed
with a rigid pipeline to eliminate compensation of pressure oscillations on hoses. The
pump is driven by a simple DC Motor at a constant speed. Technical details of the
test bench hydraulic components are given in Appendix D. Pump test setup is shown
in Figure 33.
Figure 32: Circuit diagram of the pump test bench
Figure 33: Pump test bench

50
CHAPTER 4
4. SIMULATIONS AND EXPERIMENTAL RESULTS
The simulation model is built by considering physical relations given in the previous
sections. Matlab/Simulink environment is used in order to model nonlinear equations
given in the previous sections numerically.
The simulation and the experimental results of the pressure compensated axial piston
pump are examined in this section. The evaluation of the performance of the pump is
made in time-based.
It is important to mention that fixed-step ODE5 (Dormand-Prince) is used as solver

in Matlab/Simulink simulations. The model is run several times by decreasing step
size till achieve converged solutions in order to determine critical step size. Same
results are handled by using step size equal or less than the critical value. It is found
that critical step size is 10e-08 for these coupled equations by trial. Step size is used as
5e-08 in simulations in order to handle converged solutions.
4.1 Pump Model
4.1.1 Piston Model
The piston model uses the swash plate kinematics, discharge pressure, and angular
position of piston, θ, as shown in Appendix A. The swash plate kinematics and
angular position of piston, θ, determines the piston kinematics. Instantaneous
delivery port area and suction port area are calculated due to angular position of
piston. Delivery flow rate and suction flow rate are calculated by considering these
areas, discharge pressure, and instantaneous pressure of the piston chamber. These
51
flows, piston kinematics, instantaneous swash plate angle, angular position of piston,
instantaneous piston pressure are used in order to calculate delivery flow rate,
hydraulic torque, leakages, and instantaneous piston pressure. The pump model
contains multi-piston models with a phase difference, 2π/N.
4.1.2 Compensation Section Model
The compensation section uses discharge pressure and hydraulic torque as shown in
Appendix B. The discharge pressure powers the control actuator due its magnitude.
Mechanical torque on the swash plate is calculated due to discharge pressure. Total
torque on the swash plate is accumulated in the torque model section in order to
model the swash plate kinematics. The bias actuator and the control actuator
kinematics are also determined by Equation 52 and Equation 53 in the torque model
section.
4.2 Test Bench Model
The pump test bench contains the pump and a needle valve as shown in Figure 32.
This needle valve is used to adjust pump discharge pressure at various magnitudes.
The load flow, QL, across the needle valve versus discharge pressure is measured
experimentally. From these measurements, the opening area, Aod, versus the
discharge pressure is calculated for the needle valve used in the test bench as shown
in Appendix C. The test pressure is used as input to the simulation model by
considering this relation.
4.3 Internal Dynamics Simulations
4.3.1 Kinematic Analysis
The kinematic relation of the displacement, velocity and acceleration of the ith
pumping piston are given in Equation 3, Equation 4 and Equation 5 respectively.
Normalized kinematics of the ith piston is given in Figure 34 for the case where
52
swash plate angle, α, is 16°, and secondary swash plate angle, η, is 2.5°. These angles
are obtained values from the pump geometry.
The angles are given in Figure 34 by considering θ equals to the zero on discharge
port in y-plane as shown in Figure 8. It is important to mention that BDC (where
minimum cylinder volume occurs) is at 81.4° and TDC (where maximum cylinder
volume occurs) is at 261.1° as seen in Figure 34. The angle between the axis of
symmetry and BDC is named as the cross angle. The cross angle is 8.6° for this
pump. These BDC and TDC points are so important to determine the indexing angle,
δ, and the angle of kidney-shaped flow passage area on valve plate, , that are
explained in Section 3.2.3. The variation of these points from quarter angles is due to
superimposing of swash plate angles. This condition is valid when the discharge
pressure is lower than rated pressure.
Figure 34: Motion of the ith piston in a cycle while α=16°, η=2.5°
The normalized motion of the ith pumping piston at rated pressure is shown in Figure
35. Swash plate angle, α, is equal to the minimum due to compensation of pressure
mechanism and secondary swash plate angle, η, is equal to 2.5°. This secondary
swash plate angle does not vary with the pump working conditions. As explained in
53
Section 3.1, the displacement of the ith piston at rated pressure is designed in order
for self-lubrication of the pump, decreasing excessive noise, improving the efficiency
of the pump, and the flow requirement of pressure compensation mechanism.
Because of the piston motion, extreme points are carried out near the middle of the
suction port and discharge port, only half of this displacement volume can turn into
flow. After the piston passes the extreme point at the discharge port approximately
360°, this piston is ready to suck fluid from pump discharge line. In other words,
while the piston passes the extreme point at suction port approximately 180°, this
piston is ready to discharge fluid to the reservoir line.
Figure 35: Motion of the ith piston in a cycle while α=0°, η=2.5°
4.3.2 Simulation of Single Piston Discharge Pressure Profile on Valve

Plate
The pressure profile of a piston on original valve plate at 100 bar in a cycle is shown
in Figure 36. Saturation toolbox is used to determine the least pressure in the cylinder
54
block. 1e-12 Pa is used as minimum pressure limit. This pressure magnitude is
obtained in a laboratory by researchers [38].
Figure 36: The chamber pressure, Pi, of a piston in a cycle when Pd=100 bar
It is seen that instantaneous pressure in trapping motion reaches through 368 bar
while the pump/piston is working at 100 bar. It must be mentioned that the
mechanical components are forced more than needed due to this jeopardization
action.  is measured 53.5° and indexing angle, δ, is measured 10.5° in this pump.
In order to decrease this pressure peak, a case study is done. In this study, the valve
plate is designed in accordance with line to line porting where  equal to  and the
indexing angle, δ, equal to zero.
It is simulated that pressure peak can be minimized by utilizing line to line porting
valve plate and changing indexing angle, δ, from 10.5° to the 0° as shown in Figure
37. The base method to control pressure peaks of the pump is optimizing dimensions
of kidney ports on valve plate.
55
Figure 37: The chamber pressure, Pi, of a piston in a cycle with original valve plate
and line to line porting design when Pd=100 bar
4.3.3 Simulation of Single Piston Delivery Flow Profile
The flow rate of a piston in a cycle on valve plate is shown in Figure 38 while the
discharge pressure is at 100 bar. As seen from Figure 38, an initial flow peak occurs
when the piston aligns with the discharge port due to over-compression between
TDC and the angle that delivery flow begins. Then the flow rate decreases to a
degree in accordance with the piston velocity. As shown in Figure 34, the maximum
piston velocity happens when the piston is at the position where θ is between 340°
and 350° during discharge pressure is lower than maximum full-flow pressure. The
flow rate is the highest magnitude when the piston is at the maximum velocity which
is shown in Figure 38, by ignoring flow peak. It is mentioned in Section 3.2.3 that
Qweb might not be used around the region where maximum flow occurs.
56
Figure 38: The delivery flow, Qdi, of a piston in a cycle on valve plate when Pd=100
bar
The case study is done in same conditions with the pressure profiles are given in
Section 4.3.2. It is seen that the decrement in the pressure peak comes into account as
a smooth flow rate variation in the discharge port.
The flow peak causes pressure fluctuations and vibration on the delivery line. This is
very harmful to hydraulic pipes. According to the statistics, almost half of the
reported failures of hydraulic systems on aircraft were due to the fracture of
hydraulic pipes [39].
4.3.4 Simulation of Single Piston Delivery Flow Profile at Rated Pressure
The flow rate of a piston in a cycle on original valve plate is shown in Figure 39
while the discharge pressure is at rated pressure, 170 bar. As seen from Figure 39, an
initial flow peak occurs when the piston aligns with the discharge port, due to the
high pressure inside the piston. Then the flow rate decreases to the zero in
accordance with the piston velocity. After piston crosses maximum displacement
57
point as shown in Figure 35, pumping piston chamber tends to increase while
discharge area of the piston is open to the delivery line. Due to this reason, there
exists backflow from delivery line to the pumping piston chamber.
the original valve plate
The flow rate of a piston in a cycle on line to line porting valve plate is shown in
Figure 40 while the discharge pressure is at rated pressure, 170 bar. As seen in Figure
40, a backflow peak occurs when the piston tends to decrease the pumping piston
chamber. Then delivery flow corresponds to piston velocity as shown in Figure 35.
Schematic of delivery flow, Qdi, the direction of a piston at rated pressure is given in
Figure 41 in order to visualize the motion of the hydraulic fluid. Flow through to the
inward plane means that hydraulic fluid in delivery line is fed by pumping piston. In
adverse, flow through to the outward plane means that hydraulic fluid in the chamber
of pumping piston is fed by pressurized delivery line.
58
the line to line porting
Figure 41: Schematic of delivery flow direction for a piston on valve plate at the
rated pressure
59
4.3.5 Simulation of Single Piston Leakage Profiles
In this section, leakage flow sources are investigated related with a pumping piston.
The definitions of these flows are shown in Section 3.2.2. The simulations of these
leakage flows are studied in this section. In addition to original valve plate
simulation, the leakage flow forms of the case studies are inspected. By eliminating
pressure peak in the cylinder block chamber, the line to line porting designated valve
plate improves the leakage flow to be smoother than other cases. Simulations show
that leakage flow between the cylinder block and valve plate, and ball and socket
joint leakage are the dominant part of the total leakage.
In addition to the simulation conditions explained in previous sections, a trapped

design is studied. The simulation cases examined in this study are shown in Table 3.
Table 3: The simulation cases
The angle of the The angle of the kidney-

kidney-shaped port shaped closed section Indexing
Cases between ports on valve
on cylinder block, Angle (δ )
Ψ plate, ‫׳‬
Original Valve
29.5° 53.4° 10.5°
Plate
Trapped Design 29.5° 40.1° 3.5°
Line to Line
29.5° 29.5° 0°
Porting
4.3.5.1 Simulation of Leakage Flow between Cylinder Block and Valve

Plate (QSB)
The leakage flow between the valve plate and cylinder block (QSB) is explained in
Section 3.2.2.2. As shown in Figure 42, this leakage flow around piston acts like
pressure profile as shown in Figure 37. This result is valid while the gap between the
valve plate and cylinder block, hB, is constant.
60
Figure 42: Leakage flow between the cylinder block and valve plate when Pd=100
bar
4.3.5.2 Simulation of Leakage Flow around a Piston (QSK)
The leakage flow around a piston, QSB, is explained in Section 3.2.2.1. As shown in
Figure 43, this leakage flow acts like pressure profile as shown in Figure 37. The
shape of the leakage flow likes a half circle as shown in Figure 43. The center of half
circle is located at a point which has an offset from the center point of axis system.
This offset occurs due to the variation of instantaneous leakage flow path length, lk .
When the piston moves to the BDC, this flow path term increases. The leakage flow
rate is inversely proportional to this term.
Figure 43: Leakage flow around a piston when Pd=100 bar

61
4.3.5.3 Simulation of Leakage Flow through the Slipper and Swash Plate
(QSG)
The leakage flow through the slipper and swash plate, QSG, is explained in Section
3.2.2.3. As shown in Figure 44, this leakage flow acts like pressure profile as shown
in Figure 37. This leakage flow varies only with pressure variations. Although hG
varies depending on the pump operating conditions, in simulation hG is assumed to
be constant. The other terms in Equation 13 do not change with working conditions.
Figure 44: Leakage flow through the slipper and swash plate when Pd=100 bar
4.3.5.4 Simulation of Ball and Socket Joint Leakage Flow (QSPHERE)
The leakage flow outward of the ball and socket joint, QSPHERE, is explained in
Section 3.2.2.4. As shown in Figure 45, this leakage flow across the ball and socket
joint is similar to the pressure profile shown in Figure 37. The magnitude of this flow
changes with the pressure variations. The other terms in Equation 14 are related with
joint design.
62
Figure 45: Ball and socket joint leakage flow when Pd=100 bar
4.3.6 Simulation of Hydraulic Torque on Swash Plate Applied by Single

Piston
The sources of hydraulic torque on swash plate are explained in Section 3.2.9. The
direction of the hydraulic torque changes when the piston passes swivel axis as
shown in Figure 46 while the discharge pressure is 100 bar. This figure is given for
original pump, whose maximum full-flow pressure is higher than 100 bar. The swash
plate swivel axis is generally located away from the shaft axis in order to decrease
the negative torque. This negative torque tries to decrease swash plate angle, α.
The torque peak is eliminated by employing line to line porting. This is very
beneficial while the control actuator works for decreasing swash plate angle, α.
The torque sources are examined to determine dominating part of the total torque.
The equation of pressure force acting on swash plate is defined in Equation 44. The
torque due to the pressure force is shown in Figure 47. By considering maximum
torque values of both valve plate designs, the torque due to pressure is the most
dominant part of the total torque on the swash plate. The inertial and friction forces
are equal for both conditions and negligible with respect to the pressure forces.
63
Figure 46: Hydraulic torque on swash plate due to a piston when Pd=100 bar
Figure 47: Hydraulic torque on swash plate due to pressure forces when Pd=100 bar
64
4.3.7 Simulation of Multi-Piston Pressure (Pd)
The multi-piston model is simulated by employing the single piston model with a
phase shift. The phase shift, 2π/N, is determined by the number of pistons. A pump,
which has nine pistons, is simulated in this study.
The discharge pressure profile of multi-piston model from start-up to 100 bar in a
cycle for various valve plates is shown in Figure 48. The pressure oscillations
magnitude is maximum on original valve plate. This pressure oscillations magnitude
is decreased by utilizing trapped design. In line to line porting type valve plate, this
magnitude is reduced to nearly zero.
Although the frequencies of pressure oscillations are different as shown in Figure 48,
this is misleading. The second flow peak in 40° degrees travel of a piston at original
valve plate is not seen clearly. The pump shaft frequency is multiplied by two times
N (number of pistons) in order to determine pump discharge pressure frequency.
Figure 48: One cycle pump pressure profile at 100 bar
Gao utilized an attenuator in order to decrease the pressure and the flow oscillations
in an aircraft pump [17]. It is shown that the maximum pressure fluctuations can be
65
reduced down to 80%. The relation between spring stiffness (attenuator component)
and flow fluctuation amplitude is shown in this paper. Employment of the attenuator
is similar to utilizing an accumulator on pressure line in industrial hydraulics. The
amplitude attenuation rate is dependent on the ratio between the maximum flow rate
of the pump and the attenuator volume.
4.3.8 Simulation of Multi-Piston Flow (Qd)
One cycle delivery flow rate profile is shown in Figure 49 with same conditions
given in Section 4.3.7. It is shown that instantaneous variations from the mean value,
14 lpm, are from 4 lpm to -2 lpm for the original valve plate. These instantaneous
variations are reduced to 0.3 lpm in excessive flow and -0.4 lpm in insufficient flow
in trapped design. In line to line porting type valve plate, these variations are reduced
to 0.2 lpm in the excessive flow and -0.1 lpm in the insufficient flow.
Figure 49: One cycle delivery flow rate, Qd, profile from start-up to 100 bar
By the results of these simulations, it is shown that the delivery flow frequency is
equal to the discharge pressure frequency. Similar to the discharge pressure profile
66
simulations, it seems that delivery flow frequency for original valve plate is different
from others. It could be said that flow peak due to pressure peak is overpowering the
other sources for original valve plate.
In order to deeply investigate the delivery flow, the individual delivery flow for each
piston is plotted in Figure 50. Sharp variations exist in trapped valve plate at the
angular displacement of a piston around 45°, 85°, 125°, etc. as shown in Figure 49.
The reason of this result is shown in Figure 50 to examine individually contribution
of delivery flow by each piston. The sharp variation in delivery flow for a piston
increases the amplitude of delivery flow. The oscillation of instantaneous total
delivery port area might be minimized in order to decrease pump delivery flow
variation amplitude. The instantaneous total delivery port area is simply the
summation of each pumping piston delivery area.
Ma et al [8] investigated a single cylinder and multi-cylinder model in order to

present effects of cross angle and pre-compression angle on pressure peak and flow
peak. This study is done on a non-over centered valve plate. The cross angle is
defined as the angle between symmetry z-axis of valve plate as shown in Figure 8
and TDC by referencing cylinder block axis. Similarly, the pre-compression angle is
defined as the angle between symmetry z-axis of the valve plate and BDC by
referencing cylinder block axis. Model is run for various cross angles and pre-
compression angles, in order to achieve best results. Although this study is very
useful to optimize key parameters of the valve plate to minimize pressure peak and
flow peak of the pump, no attention is paid on dynamics at rated pressure by
variation of these parameters.
67
Figure 50: Individual delivery flows, Qdi, for each piston for various valve plate
geometries from start-up to 100 bar
4.3.9 Simulation of Multi-Piston Flow, Qd, at Rated Pressure
One cycle delivery flow rate profile is shown in Figure 51 while the discharge
pressure is considered as the rated pressure.
It is shown that directions of flow peaks are different. The reason of this variation
comes from different sources of the flows. Flow peaks at original valve plate and
trapped design are related with the excessive pressure in the chamber of pumping
piston. In contrary, flow peak at line to line porting occurs due to low pressure in the
chamber of pumping piston with respect to the pressurized delivery line pressure at
the closed system. Mean flow is nearly zero at each valve plates.
68
Figure 51: One cycle delivery flow rate, Qd, profile at rated pressure, 170 bar
4.3.10 Simulation of Multi-Piston Hydraulic Torque on Swash Plate, Ty
Multi-piston model is simulated in order to calculate total hydraulic torque on swash

plate due to pumping pistons with the same conditions given in the previous sections.
There exists a tremendous difference in total hydraulic torque values as shown in
Figure 52. Maximum hydraulic torque is shown as 9.87 N.m in the original valve
plate. This value is decreased to the 2.65 N.m in valve plate with line to line porting.
Minimum hydraulic torque is obtained in trapped designed valve plate as 2.53 N.m in
these simulation conditions. In spite of the previous results, the best solution is
reached at the trapped designed valve in given simulation conditions.
Minimizing the greatest hydraulic torque gives chance to minimize the control
actuator area. Secondly, the total hydraulic torque on swash plate is always positive
in Figure 52 when the discharge pressure is 100 bar. If hydraulic torque on swash
plate is always positive for whole working range, this means that Equation 50 is
satisfied without considering bias actuator force and pumping pistons are always
trying to increase swash plate angle by itself. The bias actuator area and the control
69
actuator area can be decreased in order to reduce the total pump package volume, by
optimization of hydraulic torque on the swash plate.
Figure 52: Total hydraulic torque on swash plate from start-up to 100 bar
In the view of these simulations, it is observed that hydraulic torque frequency is

equal to the discharge pressure frequency. Similar to the discharge pressure profile
simulations, it seems that hydraulic torque frequency at original valve plate is
different from line to line porting and trapped design. Similar to the previous
sections, torque peak is superior to other torque sources in 40° degree travel of a
piston on original valve plate.
In order to deeply examine the hydraulic torque on swash plate, the individual
torques of each piston are shown in Figure 53. There exists compensation of
hydraulic torques on swash plate due to the mutual torque of pistons. Compensation
is minimum at original valve plate, because of the individual maximum hydraulic
torque is maximum at original valve plate. Although individual maximum hydraulic
torque is minimum at valve plate with line to line porting, maximum total hydraulic
70
torque is minimum at trapped designed valve plate. This result is formed due to
various valve timings. Valve timing and the length between the axis of the cylinder
block and swash axis might be optimized in order to design hydraulic torque on
swash plate.
Figure 53: Individual hydraulic torque on swash plate (N.m) for various valve plate
geometries at 100 bar
Manring [7] investigated that utilizing various indexing angle, δ, to alter carry-over
angle. Carry-over angle, γ, is defined as the angle between z-axis on valve plate as
shown in Figure 8 and beginning of discharge port at TDC with\without relief notch.
The contribution of carry-over angle is presented in Equation 6.23 in [7] by
averaging steady state operation variables. According to the Equation 6.23 in [7], the
additional torque due to carry-over angle is linearly proportional to the magnitude of
the carry-over angle. For this reason, the small values of carry-over angles are
demanded. The negative indexing angle, δ, is utilized in order to reduce the carry-
over angle, γ. Additional to this study, it is shown that transient torque peaks are
71
superior to determine pumping torque on swash plate in Figure 52 by using transient
solution.
4.4 Simulations and Experimental Results
The simulations and the experiments are done in order to get the discharge pressure-
delivery flow characteristics of the pump. The pump is modeled as explained in the
previous sections. In order to obtain discharge pressure and delivery flow data on the
entire working range of the pump, resistance on the delivery line is increased up to
discharge pressure equal to the rated pressure by decreasing opening area magnitude
of the needle valve. Experimental tests are done in test bench as shown in Figure 33.
Similar to the simulations, the opening area of the needle valve is manually
decreased down to get rated pressure on the delivery line.
4.4.1 Discharge Pressure Comparison of Test Results and Simulation

Results
This test is performed with the hydraulic test circuit as shown in Figure 33. The
pump suction line is pressurized to 5 bar with a booster pump. The discharge
pressure data is gathered with a data acquisition at a sampling frequency of 50 kHz.
The details of the data acquisition system are given in Appendix E. In this
experiment and simulation, the volume between the delivery port of the pump and
the load, Vmd, is constructed as 300 cm3. The chamber is decreased from 1000 cm3 to
300 cm3 to have less capacitance to visualize pressure oscillations at the delivery line
clearly.
According to experimental results, the discharge pressure of the pump is

approximately 48±15 bar. Oscillations of the discharge pressure are nearly at the
same amplitude for both results as shown in Figure 54. The number of pressure peaks
in a cycle is two times the number of piston. Although the first pressure peak is
higher than the second peak in 40° degrees travel in simulation results, this condition
is opposite in experimental results. The difference about discharge pressure trends is
constituted due to differences in leakages related with the piston as explained in
72
Section 3.2.2. Although the shaft speed changes with respect to the load conditions,
the shaft speed is assumed to be constant during the simulations. Due to this
assumption, there exists a difference in the period of oscillations.
Figure 54: Discharge pressure of the pump at a cycle
4.4.2 Discharge Pressure-Delivery Flow Characteristics Comparison of

Simulation Results and Experimental Results
The simulated and experimental results given in Figure 55 and Figure 56 are filtered
for a window size, which is equal to the sampling frequency, 50 kHz. According to
the simulation results, the maximum full-flow pressure of the pump is observed as
157.2 bar and the rated pressure is 172.8 bar. While the discharge pressure is lower
than the maximum full-flow pressure, delivery flow oscillates about ±0.2 lpm in the
simulation results.
Internal leakage due to increasing pressure is seen as a decrement in discharge flow

rate in both cases. Additional to the pressure increment, the wear of the internal parts
causes an increase of the clearances. As seen from Figure 55, internal leakages are
73
higher in test pump. This condition is probably formed due to wear of the internal
parts. In adverse, model is run by using nominal manufacturing values.
Figure 55: Pressure-flow characteristics of the pump
Although rated pressure values are approximately same in both cases, there exist
differences in the maximum full-flow pressure values related with the determination
of damping torque of swash plate or determination of the pressure compensation
mechanism parameters.
4.4.3 Discharge Pressure-Delivery Flow Characteristics Comparison of

Case Studies
In order to obtain the discharge pressure-delivery flow characteristics of various

valve plated pumps, individual cases are implemented to the pump model. The
attenuations of the delivery flow oscillation amplitudes for a unique discharge
pressure magnitude are obtained by utilizing various cases as shown in Figure 56.
Flow rate oscillations are decreased by utilizing less trapping angle with respect to
the original valve plate. This oscillation amplitude decreases to ±0.01 lpm in line to
74
line porting valve plates. The minimum leakage is achieved in line to line porting by
obtaining smoother flow behavior as explained in Section 4.3.5. The decrease in the
leakage leads to a higher flow rate at the same magnitude of discharge pressure.
Figure 56: Discharge pressure-delivery flow characteristics curves of case studies
75
76
CHAPTER 5
5. SUMMARY, CONCLUSIONS, AND FUTURE WORK
5.1 Summary
In this thesis, dynamics of a pressure compensated axial piston pump is examined by

deriving their linear and non-linear equations by modeling the pump in
Matlab/Simulink environment. In this study, temperature variations in the hydraulic
fluid are not considered. In order to obtain the discharge pressure-delivery flow
characteristics of the pump, a test bench is constructed. The experimental results are
compared with the simulation results. Various case studies are examined with
different valve plate configurations in simulations. As a result, it is shown that the
dynamic behavior of the pump can be improved by utilizing line to line porting type
valve plate.
A detailed pump model is done in order to visualize the contribution of the functional
components on the dynamics of the pump. Kinematics of the pumping pistons is
modeled to determine the hydraulic torque on the swash plate. The hydraulic torque
on the swash plate is studied by consideration of the pressure forces, the inertial
forces, and the frictional forces. It is shown that dominant part of these sources is the
pressure forces on the piston. The instantaneous pressure inside the pumping piston
chamber is calculated by considering the piston kinematics, and the flow into and out
of this volume. Within the pressure calculations, the compressibility of the hydraulic
fluid is also taken into consideration.
Moving parts of the three-way valve are modeled as a simple spring-mass system.
The valve activates the control spool by allowing flow into the control chamber when
77
the discharge pressure is higher than the maximum full-flow pressure. This flow
allowed by the control spool increases the pressure in the control chamber. The rise
in the pressure activates the control actuator which increases the negative torque
exerted on the swash plate. When the resultant torque on the swash plate becomes
negative, the swash plate tends to decrease the swash plate angle. By this motion,
swept volume of the pump is minimized which results in a decrease in the delivery
flow. In other words, pressure build-up capacity is limited by decreasing the delivery
flow.
Leakage paths related with the pumping pistons, the three-way valve, the bias
actuator, and the control actuator are modeled to obtain total internal leakage of the
pump. The most significant leakage path is determined in order to handle the internal
leakage problem more effectively.
5.2 Conclusions
As a result, it is possible to obtain the discharge pressure-delivery flow

characteristics and internal dynamics of a newly designed pump. Different
configurations can be implementable to the simulation model to analyze the effects
of these variations on discharge pressure-delivery flow characteristics and internal
dynamics of a new pump. The simulation model gives chance to a designer to project
the pump dynamics for a given set of parameters, for example, the loads on the
swash plate bearings, loads on the cylinder block bearings, and required input torque
from the prime mover. By inverse approach, it is possible to determine the design
parameters for a set of operational requirements. It is easy to design geometry of
internal parts by using the parametric model.
It is shown that the amplitudes of the pressure oscillations are similar in simulated
and experimental results. The simulation results show that utilizing various valve
plate configurations can improve pressure oscillations in the delivery line as
expected. The attenuation of the discharge pressure oscillations minimizes the
oscillations amplitude of the delivery flow rate. In some applications where the
control valve is mounted on the pump, minimizing the amplitude of the discharge
78
pressure oscillations is essential to control the actuator position precisely. In
applications where the control valve is connected to the pump by a long piping, the
discharge pressure oscillations cause fatigue in the pipes. This condition acts as a
source of failure.
Although the rated pressure of the original pump is approximately 170 bar for the
test pump, the instantaneous pressure of the pumping piston chamber can be much
higher than the rated pressure. This study shows that the maximum pressure on the
pumping piston can be nearly four times of the discharge pressure at original pump
design. The peak pressure can be eliminated by utilizing line to line porting
designated valve plate. Pressurizing piston chamber above the rated pressure causes
wear problems especially on the cylinder block and pumping pistons.
Another outcome of the simulation results is that the resultant hydraulic torque on
swash plate can be decreased by utilizing trapped or line to line porting designated
valve plate. Reducing the maximum hydraulic torque on swash plate decreases the
required control torque which changes swash plate angle. Smaller required control
torque gives a possibility to employ smaller control actuator area which enables the
designer to use a smaller volume for the pump.
In this study, discharge pressure oscillations of the pump are decreased by employing
line to line porting valve plate instead of utilizing original valve plate. Internal
leakage of the pump is decreased by smoother pressure regime acting on the
components of the pump. This improvement increases the magnitude of the delivery
flow rate for thr entire working range.
5.3 Future Work
Although the pump has been modeled down to the component level, the experiments
have been performed on the pump as a whole. As a continuation of this work, the
subcomponents of the pump can be experimentally verified in order to verify the
entire simulation model.
79
The simulated case studies regarding the valve plate configurations in this thesis can
be physically manufactured and tested. Experimental work on the case studies will
give chance to cross-check the simulation results considering the whole pump model.
For example, the pressure in the pumping piston chamber can be measured to
compare with the simulation results. A single piston test bench can be constructed in
order to measure various leakage paths to verify the simulation models.
The simulation model can be enhanced to include various numbers of pumping

pistons, like 7, 11, 13.
80
REFERENCES
[1] Yücel, E., "Akışkan Gücü Kontrolü Teorisi", 1st Ed., Gazi University, Ankara,
1995.
[2] Manring, N.D., "Designing the Shaft Diameter for Acceptable Levels of Stress
Within Axial-Piston Swash-Plate Type Hydrostatic Pump," Journal of
Mechanical Design, vol. 122, pp. 553-559, 2000.
[3] Lewis, E., Stern, H., Design of Hydraulic Control, 1st Ed., McGraw-Hil
Company Inc, New York & London, 1962.
[4] Blackburn, J.F., Reethof, G., and Shearer, J.L., Fluid Power Control, 1st Ed.,
MIT Press and John Wiley & Sons Inc., New York & London, 1960.
[5] Ivantsyn, J., Ivantsynova, M., Hydrostatic Pumps and Motors, 1st English Ed.,
Akademia Books International, New Delhi, 2001.
[6] Watton, J., Fundamentals of Fluid Power Control, 1st Ed., Cambridge
University Press, Cambridge, 2009.
[7] Manring, N.D., Fluid Power Pumps & Motors, 1st Ed., Mc Graw Hill
Education, New York, 2013.
81
[8] Ma, J., Fang, Y., Xu, B., Yang, H., "Optimization of cross angle based on the
pumping dynamics model," Journal of Zhejiang University-Sci A (Appl. Phys.
& Eng.), vol. 3, pp. 181-190, 2010.
[9] Parker Company, "Hydraulic Pump Basics," 02.02.2018.
[10] "IS0 8278 - Aerospace - Hydraulic, pressure compensated, variable delivery

pumps - General requirements". 1986.
[11] Parker Hannifin, Pump & Motor Division Europe, "Axial Piston Pumps (HY30-
32-3245/UK)," Chemnitz, Germany, 20215.
[12] Bosch Rexroth AG Mobile Applications, "Axial Piston Variable Pump A4VG,"
Elchingen, 2012.
[13] Kim, S.D., Cho, H.S., Lee , C.O., "A Parameter Sensitivity Analysis for the
Dynamic Model of a Variable Displacement Axial Piston Pump," Proceedings
of the Institution of Mechanical Engineers, vol. 201, pp. 235-243, 1987.
[14] Schoenau, G., Burton, R., Kavanagh, G., "Dynamic Analysis of a Variable
Displacement Pump," Journal of Dynamic Systems, Measurement, and Control,
vol. 112, pp. 122-131, 1990.
[15] Manring, N.D., "Valve-Plate Design for an Axial Piston Pump Operating at Low
Displacements," Journal of Mechanical Design, vol. 125, pp. 200 - 207, 2013.
[16] Manring, N. D., Dong, Z., "The Impact of Using a Secondary Swash-Plate
Angle Within an Axial Piston Pump," Journal of Dynamic Systems,
82
Measurement, and Control, vol. 126, pp. 65-74, 2004.
[17] Gao, F., Ouyang, X., Yang, H., Xu, X., "A novel pulsation attenuator for aircraft
piston pump," Mechatronics, pp. 566-572, 2013.
[18] Manring, N. D., "Measuring Pump Efficiency: Uncertainty Considerations,"

Journal of Energy Resources Technology, vol. 127, pp. 280-284, 2005.
[19] Bergada, J.M., Kumar, S., Davies, D.LI., Watton, J., "A complete analysis of
axial piston pump leakage and output flow ripples," Applied Mathematical
Modelling, pp. 1731-1751, 2012.
[20] Bergada, J.M., Davies, D.J., Kumar, S., Watton, J., "The effect of oil pressure
and temperature on barrel film thickness and barrel dynamics of an axial piston
pump," Meccanica, pp. 639-654, 2012.
[21] Meyer, D., "Reynolds Equation for spherical bearings," vol. 125, pp. 203-206,
2003.
[22] Manring, N. D., Wray, C.L., Dong Z., "Experimental Studies on the
Performance of Slipper Bearings Within Axial-Piston Pumps," Journal of
Tribology, vol. 126, pp. 511-518, 2004.
[23] Eaton, "220 Piston Pump Service Manual E-PUMC-RR001-E2," February 2014.
[24] Kavanagh, G.P., "The Dynamic Modelling of an Axial Piston Hydraulic Pump,"
University of Saskatchewan, 1987.
83
[25] Likos, W. E., "Hydraulic Test Loop For Nonflammable Hydraulic Fluid,"
Southwest Research Institute, San Antonio, 1998.
[26] Manring, N. D., "The Improved Volumetric-Efficiency of an Axial-Piston Pump

Utilizing a Trapped-Volume Design," Journal of Dynamic Systems,
Measurement, and Control, vol. 123, pp. 479 - 487, 2001.
[27] Xu, B., Zhang, J., Yang, H., "Simulation research on distribution method of
axial piston utilizing pressure equalization mechanism," Journal of Mechanical
Engineering Science, vol. 227, pp. 459-469, 2012.
[28] Manring, N.D., "Tipping the Cylinder Block of an Axial-Piston Swash-Plate

Type Hydrostatic Machine," Journal of Dynamic Systems, Measurement, and
Control, vol. 122, pp. 216 - 221, 2000.
[29] Manring, N. D., "Scaling the Speed Limitations for Axial-Piston Swash-Plate
Type Hydrostatic Machines," Journal of Dynamic Systems, Measurement, and
Control, vol. 136, pp. 1-8, 2014.
[30] Mandal, N.P., Saha, R., Mookherjee, S., and Sanyal, D., "Pressure Compensator
Design for a Swash Plate Axial Piston Pump," Journal of Dynamic Systems,
Measurement, and Control, vol. 136, pp. 1-12, 2014.
[31] Merritt, H. E., Hydraulic Control Systems, New York, London, Sydney: John
Wiley & Sons, Inc, 1967.
[32] Ivantysynova, M., "Design and Modeling of Fluid Power Systems Lecture 12,"
Indiana, 02.02.2018.
84
[33] Manring, N. D., Johnson, R. E., "Swivel Torque within a Variable-Displacement
Pump," in Proceedings of the 46th National Conference on Fluid Power,
Milwaukee, 1994.
[34] Manring, N.D., Johnson, R.E., "Modeling and Designing a Variable-

Displacement Open-Loop Pump," Journal of Dynamic Systems, Measurement,
and Control, vol. 118, pp. 267-271, 1996.
[35] Manring, N. D., "The Control and Containment Forces on the Swash Plate of an
Axial Piston Pump," Journal of Dynamic Systems, Measurements, and Control,
vol. 121, pp. 599-605, 1999.
[36] Zhang, X., Cho, J., Nair S.S., Manring N.D., "New Swash Plate Damping
Model for Hydraulic Axial-Piston Pump," vol. 123, pp. 463-470, 2001.
[37] Mandal, N.P., Saha, R., Sanyal, D., "Design and Performance Analysis of a
Pressure Compensator for a Swash Plate Type Variable Displacement Axial
Piston Pump," in Proceedings of the 37th National & 4th International
Conference on Fluid Mechanics and Fluid Power, Chennai, 2010.
[38] "Hypertextbook,"[Online].Available:
https://hypertextbook.com/facts/1998/RobertSaronson.shtml.[Accessed
11.02.2018].
[39] Changbin, G., Zongxia J., Shouzhan, H., "Theoretical study of flow ripple for an
aviation axial-piston pump with damping holes in the valve plate," Chinese
Journal of Aeronautics, vol. 27, pp. 169-181, 2013.
85
86
APPENDICES
A. Matlab/Simulink Model of the Cylinder Block (Single Piston)
Matlab/Simulink model of the cylinder block for a single piston is given in Figure
A.1.
Figure A.1: Matlab/Simulink model of cylinder block (single piston)
87
B. Matlab/Simulink Model of the Compensation Section
Matlab/Simulink model of the compensation section is given in Figure B.1.
Figure B.1: Matlab/Simulink model of the compensation section
88
C. Matlab/Simulink Model of the Test Bench
Matlab/Simulink model of the test bench is given in Figure C.1.
Figure C.1: Matlab/Simulink model of the test bench
89
D. Technical Properties of the Test Bench
A simple hydraulic test unit is used in order to pressurize the suction and drain line
of the pump. PRH082K10N coded pressure reducing valve (manuf. PARKER
company) is used to control the inlet pressure of the test pump. Technical details of
the needle valve are shown in Figure D.1. Hoses used in the test system are produced
according to the SAE 100R16. Dash size of the hoses is 5. 10-L fittings according to
the ISO 8434-1 are used in order to join the components. Stainless steel tubes are
used in delivery line whose material number is 1.4571. Outer and inner diameters of
the tubes are 10 mm and 8 mm, respectively. XPC10 coded pressure sensor (manuf.
TE Connectivity company) is used to measure discharge line pressure. Technical
details of the pressure sensor are shown in Table D.1. VS 10 coded flowmeter
(manuf. VSE company) is used to measure delivery flow. Technical details of the
flowmeter are shown in Table D.2. J02A2ZN coded needle valve (manuf. PARKER
company) is employed in order to load the pump. Technical details of the needle
valve are shown in Figure D.2.
Table D.1: Performance specifications of the pressure sensor (XPC 10)
90
Table D.2: Performance specifications of the flowmeter (VS 10)
91
Figure D.1: Performance specifications of the pressure reducing valve
(PRH082K10N)
92
Figure D.2: Performance specifications of the needle valve (J02A2ZN)
93
E. Data Acquisition System Hardware
For data acquisition, National Instruments (NI) products are used to measure data. It
is important to mention that measurements are saved at 50 kHz with NI components.
NI 9188 chassis tool is used to collect data taken from I/O modules. Maximum input
or output frequency of this tool is 1 MHz. This property is much higher than the
sampling frequency of experiments, 50 kHz. Timing resolution of this component is
12.5 ns and timing resolution is 50 ppm of sample rate. NI 9218 is built-in support
for voltage measurements. In the experiments, this module is used in order to gather
data taken from pressure sensor. This tool can collect data up to 13.1072 MHz. SEG
1060 tool, is produced by Hydrotechnik, is used to convert PWM data taken from
flowmeter, VS 10, to the voltage output. The scan rate of this tool is limited with 100
Hz. Because limitation of this component, flow measurements are taken up to 100
Hz. NI 9215 tool is preferred in order to measure data taken from SEG 1060.
94

Index

Uploaded by

Copyright:

Available Formats

Index

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Index

Uploaded by

Copyright:

Available Formats

MODELING AND PERFORMANCE IMPROVEMENTS OF A PRESSURE

COMPENSATED AXIAL PISTON PUMP

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

MODELING AND PERFORMANCE IMPROVEMENTS OF A PRESSURE

submitted by OKAN AYDOĞAN in partial fulfillment of the requirements for the

Prof. Dr. Halil KALIPÇILAR _________________

Prof. Dr. M. A. Sahir ARIKAN _________________

Prof. Dr. R. Tuna BALKAN _________________

Examining Committee Members

Assoc. Prof. Dr. Yiğit YAZICIOĞLU _________________

Prof. Dr. R. Tuna BALKAN _________________

Assoc. Prof. Dr. M. Metin YAVUZ _________________

Assist. Prof. Dr. Ulaş YAMAN _________________

Prof. Dr. Yücel ERCAN _________________

Name, Last name : Okan AYDOĞAN

MODELING AND PERFORMANCE IMPROVEMENTS OF A PRESSURE

M.Sc., Department of Mechanical Engineering

Supervisor: Prof. Dr. R. Tuna BALKAN

May 2018, 94 Pages

Analysis of internal dynamics of a pressure compensated axial piston pump has an

Keywords: Swash Plate, Flow Oscillation, Torque Oscillation

BASINÇ KOMPANZASYONLU EKSENEL PİSTONLU BİR POMPANIN

Yüksek Lisans, Makina Mühendisliği Bölümü

Tez Yöneticisi: Prof. Dr. R. Tuna BALKAN

Mayıs 2018, 94 Sayfa

Basınç kompanzasyonlu eksenel pistonlu bir pompanın iç dinamiklerinin analizi bu

Bu çalışmanın amacı pompanın iç elemanlarını geometrik ölçülendirilerek

Anahtar Kelimeler: Eğim Plakası, Debi Salınımı, Tork Salınımı

First of all, I would like to express my sincere appreciation to my thesis

The support of the Scientific and Technological Research Council of Turkey

I wish to express my sincere thanks to my family.

Finally, I would like to express my gratitude to my beloved wife Ms. Elif

A. Matlab/Simulink Model of the Cylinder Block (Single Piston) ................... 87

Table 1: A comparison of power densities of three different machines [2] ................. 2

Table 2: Geometric oscillation for an axial piston pump ........................................... 48

Table 3: The simulation cases .................................................................................... 60

Figure 1: Swept volume of a pressure compensated axial piston pump while

Figure 2: Swept volume of a pressure compensated axial piston pump while

Figure 3: The discharge pressure-delivery flow characteristic curve of a typical

Figure 4: Load-sense controlled axial piston pump configuration [9] ......................... 9

Figure 5: The discharge pressure-delivery flow characteristic curve of a typical load-

Figure 6: The discharge pressure-delivery flow pressure characteristic curve of a

Figure 7: Displacement volume vs input current of a pump [12] .............................. 12

Figure 8: General pump configuration [15] ............................................................... 14

Figure 9: A cross-sectional view of the pressure controlled axial-piston pump [17] 16

Figure 10: Piston geometry in the x-z plane [14] ....................................................... 17

Figure 11: Piston geometry in the x-y plane [14] ...................................................... 18

Figure 13: Kidney port and manifold geometry [7] ................................................... 22

Figure 16: Ball and socket movement [23] ................................................................ 25

Figure 18: Trapped valve plate design [7] ................................................................. 28

Figure 19: Valve plate of Denison Model P6W-R5B-C10-00-M2 [25] .................... 29

Figure 21: Rexroth open-circuit machinery [29]........................................................ 33

Figure 22: Schematic of a three-way valve [30] ........................................................ 34

Figure 24: Free body diagram of the control actuator ................................................ 37

Figure 25: Free body diagram of the bias actuator..................................................... 39

Figure 27: Cylinder block and multi-piston [23] ....................................................... 43

Figure 28: Swivel torque force diagram [33] ............................................................. 43

Figure 29: Geometry of the swash plate [7] ............................................................... 45

Figure 30: Schematic of a pressure compensator of an axial-piston swash plate pump