Pump-Controlled Hydraulic Circuits for Single-Rod

Actuators: New Designs and Performance Evaluation

by

Ahmed Imam

A Thesis submitted to the Faculty of Graduate Studies of

The University of Manitoba
in partial fulfilment of the requirements of the degree of

Doctor of Philosophy

Department of Mechanical Engineering

University of Manitoba
Winnipeg, Manitoba, Canada

Copyright © 2020 by Ahmed Imam

 ABSTRACT

This thesis investigates the challenges in designing pump-controlled single-rod actuators and

proposes new solutions for them. Three different concepts to overcome the reported performance

problems are proposed. By applying the new concepts, eight different new designs are introduced,

among which, five are experimentally evaluated. These designs show good performance,

efficiency and energy recuperation ability as compared to previously reported circuits.

Hydraulic actuation systems are favored in many applications due to their fast response, high

power-to-weight ratio, and stability under variable loading conditions. Efficiency, however, is the

main disadvantage of conventional valve-controlled hydraulic actuators, which is typically around

25%. Pump-controlled hydraulic actuation systems are more efficient than valve-controlled ones,

as they eliminate the energy losses due to flow throttling in valves and require less cooling effort.

The existing common pump-controlled circuits for single-rod actuators experience a performance

problem at specific operating conditions. More specifically, circuits encounter oscillatory

performance at low loading conditions.

In this thesis, a detailed analysis to identify the low performance operating regions of a

commonly-used single-rod hydraulic pump-controlled circuit that utilizes two pilot operated check

valves is performed. Factors affecting these regions are discussed. The analysis is repeated to

evaluate performance of circuits with closed-center and open-center shuttle valves. Findings are

also validated by experimental results. Based on these findings, three different concepts to improve

circuit performance, specially, in the oscillatory zone are proposed. The proposed concepts are:

selective throttling over flow, selective oil leakage, and shifting of critical zone into lower loading

margin. Building upon these concepts, eight different new circuits are designed. Simulation studies

verify the enhanced performance of the proposed circuits as compared to the benchmark design.

ii
Five of the eight proposed circuits that utilize available off-the-shelf industrial components are

prototyped and tested using a newly constructed test rig. The test rig has been designed to facilitate

the implementation of different circuits and for various loading conditions. It is shown that the

proposed designs do not require additional electronic control components, which is desirable in

industrial applications. Furthermore, energy studies demonstrated superior efficiency of the

proposed circuits compared to the valve-controlled counterparts.

iii
 ACKNOWLEDGEMENTS

All praise and gratitude is due to Allah, Lord of the worlds.

I would like to express my deepest gratitude to Prof. Nariman Sepehri, my advisor, for his

continuous support, inspiration, motivation, and directions throughout this work. My thanks

extend also to my committee members Prof. Subramaniam Balakrishnan and Prof. Udaya

Annakkage for their valuable comments that assisted in enhancing this work.

Special thanks go to Moosa Rafiq whose encouragement, support and time were helpful to

complete this work; thanks extend also to my other colleagues at the Fluid Power and Tele-

Robotics Research Laboratory, Ehsan Jalayeri, Vikram Banthia, Ramhuzaini Abd.Rahman, Ali

Maddahi, Khurram Butt and Maziar Mardan for their support.

I thank the staff at the Department of Mechanical Engineering and the Faculty of Graduate

Studies.

I would like to express my sincere gratitude to the elementary foundations in my life, my family.

My parents, I only reached this moment because of you. My wife, I guess you earned this degree,

too. Finally, my precious sons, I overcome the hardships of life for you.

iv
 TABLE OF CONTENTS
ABSTRACT ............................................................................................................................ II
ACKNOWLEDGEMENTS ................................................................................................ IV
1. INTRODUCTION ........................................................................................................... 1
1.1 Statement of the problem................................................................................................ 1
1.2 Objective of this research ............................................................................................... 2
1.3 Methodology................................................................................................................... 3
1.4 Thesis outline.................................................................................................................. 3
2. LITERATURE REVIEW ............................................................................................... 4
2.1 Background on hydraulic actuation systems ................................................................... 4
2.2 Pump-controlled hydraulic actuation .............................................................................. 7
2.2.1 Four quadrants of operation......................................................................................... 7
2.2.2 Single versus double rod actuators .............................................................................. 9
2.2.3 Variable displacement versus variable speed pumps .................................................. 9
2.2.4 Conventional versus asymmetric pumps ................................................................... 10
2.2.5 Single versus multi-pump configurations .................................................................. 11
2.2.6 Closed-circuit versus open-circuit configurations ..................................................... 12
2.2.7 Pressurized versus non-pressurized reservoirs .......................................................... 13
2.3 Previous work ............................................................................................................... 13
2.4 Summary ....................................................................................................................... 18
3. EVALUATION OF SELECTED DESIGNS................................................................ 19
3.1 Circuit with two pilot operated check valves ................................................................ 19
3.1.1 Previous identification of poor performance zone .................................................... 22
3.1.2 Four quadrants of operation and switching zones ..................................................... 23
3.1.3 Modeling ................................................................................................................... 25
3.1.4 Discussions ................................................................................................................ 30
3.1.5 Simulation studies ..................................................................................................... 33
3.2 Circuit with closed-center shuttle valve ........................................................................ 39
3.2.1 Previous identification of poor performance zone .................................................... 40
3.2.2 Modeling ................................................................................................................... 41
3.2.3 Simulation studies ..................................................................................................... 42
3.3 Circuit with open-center shuttle valve .......................................................................... 47
3.3.1 Previous identifications of poor performance zone ................................................... 48
3.3.2 Modeling ................................................................................................................... 49
3.3.3 Simulation studies ..................................................................................................... 50
3.4 Comparison between the selected circuits .................................................................... 57
4. NEW SOLUTIONS ....................................................................................................... 59
4.1 Requirements for desirable operation of pump-controlled single-rod actuators .......... 59
4.2 Concept I- shifting of critical zone ............................................................................... 63
4.2.1 Implementation .......................................................................................................... 65

v
 4.2.2 Modeling ................................................................................................................... 67
4.2.3 Simulation studies ..................................................................................................... 69
4.3 Concept II- selective leakage ........................................................................................ 75
4.3.1 Implementation .......................................................................................................... 75
4.3.2 Modeling ................................................................................................................... 75
4.3.3 Simulation studies ..................................................................................................... 76
4.4 Concept III- selective throttling .................................................................................... 81
4.4.1 Implementation .......................................................................................................... 81
4.4.2 Modeling ................................................................................................................... 82
4.4.3 Simulation studies ..................................................................................................... 84
4.5 New designs .................................................................................................................. 88
4.5.1 Design 1..................................................................................................................... 88
4.5.2 Design 2..................................................................................................................... 91
4.5.3 Design 3..................................................................................................................... 94
4.5.4 Design 4..................................................................................................................... 96
4.5.5 Design 5..................................................................................................................... 99
4.5.6 Design 6................................................................................................................... 102
4.5.7 Design 7................................................................................................................... 104
4.5.8 Design 8................................................................................................................... 107
4.6 Summary ...................................................................................................................... 109
5. EXPERIMENTAL TEST RIG ................................................................................... 110
5.1 JD-48 Backhoe attachment and supporting structure ................................................. 112
5.2 Loading configurations .............................................................................................. 113
5.3 Hydraulic actuator ...................................................................................................... 114
5.4 Main pump unit .......................................................................................................... 115
5.5 Low-pressure charge unit ........................................................................................... 116
5.6 Measurement and control devices .............................................................................. 117
6. EXPERIMENTAL EVALUATIONS ........................................................................ 118
6.1 Evaluation of previous designs ................................................................................... 118
6.1.1 Constant loading experiments ................................................................................. 121
6.1.2 Variable loading experiments .................................................................................. 124
6.1.3 Summary ................................................................................................................. 130
6.2 Evaluation of proposed designs .................................................................................. 131
6.2.1 Evaluation of Design 5 ............................................................................................ 132
6.2.2 Evaluation of Design 7 ............................................................................................ 137
6.2.3 Evaluation of Design 8 ............................................................................................ 142
6.2.4 Evaluation of Designs 2 and 3 ................................................................................. 147
7. CONCLUSIONS .......................................................................................................... 153
7.1 Contributions of this thesis ......................................................................................... 153
7.2 Recommendations and future work ............................................................................ 156
REFERENCES .................................................................................................................... 158

vi
 LIST OF FIGURES

Figure 2-1 Typical valve-controlled hydraulic circuit. ................................................................... 5

Figure 2-2 Power losses in valve-controlled hydraulic circuits for different pump types: (a) fixed
displacement; (b) pressure compensated; (c) load sensing. ............................................................ 6

Figure 2-3 Single-rod cylinder: (a) sign convention; (b) quadrants of operation. .......................... 8

Figure 2-4 Defining quadrants in hydrostatic pumps: (a) sign convention; (b) quadrants of
operation. ........................................................................................................................................ 8

Figure 2-5 Pump-controlled actuator: (a) double rod; (b) single-rod. ............................................ 9

Figure 2-6 Pump-controlled actuator: (a) variable speed; (b) variable displacement. .................. 10

Figure 2-7 Asymmetric pump-controlled circuit: (a) layout; (b) ports plate. ............................... 11

Figure 2-8 Multi-pump circuits: (a) variable displacement controlled actuator; (b) variable speed
controlled actuator; (c) identical pump control of 1:2 area ratio cylinder; (d) utilization of hydraulic
transformer. ................................................................................................................................... 12

Figure 2-9 Pump-controlled actuator circuits: (a) closed circuit; (b) open circuit. ...................... 13

Figure 2-10 Circuit that utilizes 2-position 3-way shuttle valve, developed by Hewett [8]. ........ 14

Figure 2-11 Circuit with two pilot-operated check valves, by Rahmfeld and Ivantysynova [9]. . 15

Figure 2-12 Circuit that uses 3/3 closed-center shuttle valve, proposed by Wang et al. [11]. .... 16

Figure 2-13 Circuit that utilizes counterbalance valves proposed by Jalayeri et al. [5, 12]. ........ 16

Figure 2-14 A miniature pump-controlled circuit with counterbalance valves proposed by Altare
and Vacca [28, 29]. ....................................................................................................................... 17

Figure 2-15 Circuit with open-center 3-position shuttle valve, proposed by Caliskan et al. [13]. 18

Figure 3-1 Circuit that utilizes pilot operated check valve. .......................................................... 20

Figure 3-2 Simplified drawing of circuit that utilizes pilot operated check valves in the first
quadrant of operation; CH denotes charge system. ...................................................................... 21

vii
Figure 3-3 Actuator operating plane illustrating the four quadrants of operation, maximum
actuating forces, and maximum velocities in extension and retraction, and low-performance zone
according to [26]. .......................................................................................................................... 22

Figure 3-4 Flow patterns in four quadrants of operation (Q1, Q2, Q3, Q4) and switching zones
(S14, S43, S32, S21) of circuit that utilizes pilot operated check valve in the load-velocity plane.
....................................................................................................................................................... 24

Figure 3-5 An excavator stick in four quadrants of operation. ..................................................... 25

Figure 3-6 Schematic drawing of circuit that utilizes POCVs in the first quadrant of operation. 26

Figure 3-7 Pilot-operated check valve: (a) schematic drawing; (b) opening area versus effective
pressure. ........................................................................................................................................ 28

Figure 3-8 Construction of critical zones 5 and 6, taking into account the effect of transmission
line losses, Coulomb and viscous friction forces and valves cracking pressure. .......................... 31

Figure 3-9 Simulation results of circuit with two pilot-operated check valves given data in Tables
3-1 and 3-2: (a) construction of critical zone; (b)-(f) actuator velocity responses at test points (TP1-
TP5) given step control signal input of -4 V and assistive loads of 4500 N, 3150 N, 2750 N, 2500
N and 1500 N, respectively. .......................................................................................................... 35

Figure 3-10 Simulation results for performance of the circuit with two pilot-operated check valves
at low loading conditions given data in Tables 3-1 and 3-2: (a) input control signal; (b) variable
load of ± 3 kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pressure plane. ............................................................. 37

Figure 3-11 Simulation results for performance of the circuit with two pilot-operated check valves
at high loading conditions given data in Tables 3-1 and 3-2: (a) input control signal; (b) variable
load of ± 15 kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pressure plane. ............................................................. 38

Figure 3-12 Circuit that utilizes 3/3 closed-center shuttle valve in the first quadrant of operation.
....................................................................................................................................................... 39

Figure 3-13 Illustration of critical zone (hatched area) in circuit that utilizes closed-center shuttle
valve according to [11]. ................................................................................................................ 40

viii
Figure 3-14 3/3 closed-center shuttle valve: (a) schematic drawing; (b) opening area versus
effective pressure. ......................................................................................................................... 42

Figure 3-15 Simulation results of the circuit with closed-center shuttle valve, given data in Tables
3-1 and 3-3: (a) construction of critical zone; (b)-(f) actuator velocity response at test points TP1-
TP5 for step input signal of -4 V and assistive loads of 4500N, 3860N, 3200N, 2200N and 1400N,
respectively. .................................................................................................................................. 44

Figure 3-16 Simulation results for performance of the circuit with closed-center shuttle valve at
low loading conditions given data in Tables 3-1 and 3-3: (a) input signal; (b) variable load of ± 3
kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane. .................................................................................. 46

Figure 3-17 Simulation results for performance of the circuit with closed-center shuttle valve at
high loading conditions given data in Tables 3-1 and 3-3: (a) input control signal; (b) variable load
of ± 15 kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b (dotted
line); (e) load-velocity plane; (f) pressure plane. .......................................................................... 47

Figure 3-18 Circuit that utilizes an open-center shuttle valve. .................................................... 48

Figure 3-19 Critical zone in the circuit that utilizes open-center shuttle valve showing leakage
solution effectiveness according to Caliskan et al. [13]. .............................................................. 49

Figure 3-20 3/3 open-center shuttle valve: (a) schematic drawing; (b) openings areas versus
effective pressure. ......................................................................................................................... 50

Figure 3-21 Simulation results of the circuit with an open-center shuttle valve, given data in Tables
3-1 and 3-4: (a) construction of critical region; (b)-(f) actuator velocity response at test points TP1-
5 for step signal input of -4 V and loads of 4550 N, 3400 N, 2960 N, 2370 N and 1650 N. ....... 52

Figure 3-22, Simulation results of the circuit with an open-center shuttle valve with higher pre-
designed leakage, given data in Tables 3-1 and 3-4: (a) construction of critical region; (b)-(f)
actuator velocity response at test points TP1-5 for step signal input of -4 V and loads of 4380 N,
3580 N, 2800 N, 2200 N and 1350 N. .......................................................................................... 54

Figure 3-23 Simulation results of the circuit with open-center shuttle valve at low loading
conditions, according to the given data in tables 3-1 and 3-4: (a) input control signal; (b) applied

ix
load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane. ................................................................................... 55

Figure 3-24 Simulation results of the circuit with open-center shuttle valve at high loading
conditions, according to the given data in tables 3-1 and 3-4: (a) input control signal; (b) applied
load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane. ................................................................................... 56

Figure 3-25 Critical zones for circuits that utilize; (a) POCVs (solid line) and CC-SHV (dashed
line); (b) CC-SHV (dashed line) and OC-SHV (solid line) for comparable valve parameters. ... 57

Figure 4-1 Desired pressure plane of pump-controlled circuits given a variable load configuration
and a square input signal (shown in the inset). ............................................................................. 61

Figure 4-2 Desired performance of a pump-controlled circuit in the load-velocity plane. .......... 63

Figure 4-3 Construction of critical regions for, (a) conventional pump-controlled circuit as is
explained in Section 3-1 ; (b) circuit that applies shifting of the critical zone concept. ............... 64

Figure 4-4 Circuit with: (a) one charge pressure and two identical pilot operated check valves; (b)
two charge pressures and two different pilot operated check valves. ........................................... 65

Figure 4-5 Circuits with: (a) one charge pressure and a 3/3 closed-center shuttle valve (SHV); (b)
two charge pressures and a 4/3 closed-center biased SHV; (c) one charge pressure and a 3/3 open-
center SHV; (d) two charge pressures and a 4/3 open-center biased SHV. .................................. 66

Figure 4-6 Circuit that utilizes two charge pressures and a 4/3 closed-center biased shuttle valve
in first quadrant of operation......................................................................................................... 67

Figure 4-7 4/3 closed-center biased shuttle valve; (a) schematic drawing in center position; (b)
opening area as function of the effective pressure. ....................................................................... 68

Figure 4-8 Simulation results for circuit with biased closed-center shuttle valve and two charge
pressures, 1.17 MPa and 1.56 MPa: (a) construction of critical region; (b)-(f) actuator velocity
response at test points TP1-5 for step input signal of -4 V and assistive loads of 3700N, 2700N,
1500N, 940N and 500N, respectively. .......................................................................................... 71

Figure 4-9 Simulation results for circuit with biased closed-center shuttle valve and two charge
pressures under low loading conditions given data in Tables 3-1 and 4-1: (a) input control signal;

x
(b) applied load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pump pressure plane. ................................................... 73

Figure 4-10 Simulation results for circuit with biased closed-center shuttle valve and two charge
pressures under high loading conditions given data in Tables 3-1 and 4-1: (a) input control signal;
(b) applied load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pump pressure plane. ................................................... 74

Figure 4-11 Using leakage to stabilize pump-controlled circuits that utilize POCVs: (a) leakege in
main flow lines; (b) leakege in differential flow lines. ................................................................. 76

Figure 4-12 Simulation results for circuit with pilot-operated check valves and leakage valves,
given data in Tables 3-1 and 4-2: (a) velocity steady state responses at different loads and step
input signals on the FL- va plane; (b-d) velocity response at TP1-TP3 at beginning-middle and end
of critical zone for -4 V step input and loads of 4200N, 3000 N and 1700 N. ............................. 78

Figure 4-13 Simulation results for circuit with pilot-operated check valves and leakage control
under low loading condition, given data in Tables 3-1 and 4-2: (a) input control signal; (b) variable
load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane. ................................................................................... 79

Figure 4-14 Simulation results for circuit with pilot-operated check valves and leakage control
under high loading condition, given data in Tables 3-1 and 4-2: (a) input control signal; (b) variable
load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane. ................................................................................... 80

Figure 4-15 Proposed location for additional selective-throttling valve in the circuit that utilizes
two POCVs, (a) in the main flow lines; (b) in the differential flow lines..................................... 82

Figure 4-16 Schematic drawing of the circuit that utilizes POCVs equipped with limited throttling
valves, in the first quadrant of operation. ..................................................................................... 83

Figure 4-17 Simulation results for circuit with pilot-operated check valves and selective throttling
valves, given data in Tables 3-1 and 4-3; (a) steady state velocity responses around switching zone
S43 in the load-velocity plane; (b) velocity response at TP1 at load of 6202 N in the fourth
quadrant; (c) velocity response at TP2 at load of 6200 N in the third quadrant of operation....... 85

xi
Figure 4-18 Simulation results for circuit with pilot-operated check valves and selective throttling
valves at low loading conditions, given data in Tables 3-1 and 4-3: (a) input control signal; (b)
applied load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b
(dotted line); (e) load- velocity plane; (f) pressure plane. ............................................................ 86

Figure 4-19 Simulation results for circuit with pilot-operated check valves and selective throttling
valves at high loading conditions, given data in Tables 3-1 and 4-3: (a) input control signal; (b)
applied load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b
(dotted line); (e) load- velocity plane; (f) pressure plane of the circuit. ....................................... 87

Figure 4-20 Design 1: circuit with two pilot-operated check valves (POCVs) and two charge
pressures (CHs): (a) detailed circuit drawing; (b) simplified circuit. ........................................... 89

Figure 4-21 Flow patterns in the four quadrants of operation of Design 1................................... 90

Figure 4-22 Design 2: circuit with a 4/3 biased closed-center shuttle valve (CC-SHV) and two
charge pressures (CHs). ................................................................................................................ 92

Figure 4-23 Flow patterns in the four quadrants of operation of Design 2................................... 93

Figure 4-24 Design 3: circuit with a 4/3 biased open-center shuttle valve (OC-SHV) and two
charge pressures (CHs). ................................................................................................................ 94

Figure 4-25 Flow patterns in the four quadrants of operation of Design 3................................... 95

Figure 4-26 Design 4: circuit with two pilot-operated check valves (POCVs), a selective-throttling
valve (LTV) and two charge pressures (CHs); LTV flow area versus the effective pressure is shown
in the inset. .................................................................................................................................... 96

Figure 4-27 Flow patterns in the four quadrants of operation of Design 4................................... 97

Figure 4-28 Design 5: circuit with two pilot-operated check valves (POCVs), two counterbalance
valves (CBVs) and two charge pressures (CHs). ........................................................................ 100

Figure 4-29 Flow patterns in the four quadrants of operation of Design 5................................. 101

Figure 4-30 Design 6: circuit with two selective throttling valves (LTVs) and two charge pressures
(CHs); LTV flow area versus the effective pressure is shown in the inset. ................................ 102

Figure 4-31 Flow patterns in the four quadrants of operation of Design 6................................. 103

xii
Figure 4-32 Design 7: circuit with two sequence valves (SQVs) and two charge pressures (CHs).
..................................................................................................................................................... 105

Figure 4-33 Flow patterns in the four quadrants of operation of Design 7................................. 106

Figure 4-34 Design 8: circuit with a pilot-operated check valve (POCV), a sequence valve (SQV)
and two charge pressures (CHs). ................................................................................................ 107

Figure 4-35 Flow patterns in the four quadrants of operation of Design 8................................. 108

Figure 5-1 Test rig: (1) JD-48 backhoe attachment; (2) supporting structure; (3) actuator (4) load
configurations; (5) main pump unit; (6) charge pump unit; (7) control and monitoring station; (PS)
pressure sensors; (DS) displacement sensor. .............................................................................. 111

Figure 5-2 Experimental test rig; (a) schematic, (b) interface structure. .................................... 112

Figure 5-3 JD-48 backhoe attachment; (a) attached to a utility tractor; (b) retrofitted in the Lab.
..................................................................................................................................................... 113

Figure 5-4 Loading configuration of the JD-48; (a) weight plates; (b) sliding load................... 114

Figure 5-5 Hydraulic cylinder of the stick link ........................................................................... 114

Figure 5-6 The variable displacement swash-plate piston pump; (a) cross-section of an individual
pump; (b) dual pumps mechanically connected to electric motor and hydraulically connected to
the circuit components. ............................................................................................................... 115

Figure 5-7 Charge pump unit ...................................................................................................... 116

Figure 5-8 Pressure and displacement sensors installation; (a) actuator unit; (b) pump unit. .... 117

Figure 6-1 Test rig: (1) JD-48 backhoe attachment, (2) main pump unit, (3) charge pump unit, (PS)
pressure sensors, and (DS) displacement sensor......................................................................... 119

Figure 6-2 ReconFigured test rig used to test circuit at constant mass load; (a) a photo illustrates
the two load options; (b) a schematic drawing of load setup to test resistive extension, Q1, and
assistive retraction, Q4, quadrants; (c) a schematic drawing of load setup to test resistive retraction,
Q3, and assistive extension, Q2, quadrants................................................................................. 120

Figure 6-3 Circuit with two pilot-operated check valves in first quadrant of operation............. 121

xiii
Figure 6-4 Experimental identification of the critical zone (shown by hashed area) for circuit that
utilizes pilot-operated check valves (Figure 6-3)........................................................................ 123

Figure 6-5 Performance of the circuit with pilot-operated check valves at resistive extension and
assistive retraction (TP1 and TP2 shown in Figure 6-4) given 259 kg attached mass: (a) control
input signal; (b) actuator velocity; (c) pressures at pump port a (solid line) and port b (dotted line);
(d) pressures at actuator port a (solid line) and port b (dotted line). ........................................... 123

Figure 6-6 Circuits with shuttle valves: (a) 3/3 closed-center; (b) 3/3 open-center. .................. 124

Figure 6-7 Calculated static load at the actuator rod for 41 kg (dashed line) and 368 kg (solid line)
masses attached to the stick. ....................................................................................................... 125

Figure 6-8 Performance of the circuit with pilot-operated check valves for a 41kg attached mass:
(a) input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line) and b (dotted
line); (d) pump pressure plane. ................................................................................................... 126

Figure 6-9 Performance of the circuit that utilizes a closed-center shuttle valve for a 41kg attached
mass: (a) input signal; (b) actuator velocity; (c) pressures at pump ports a (solid line) and b (dotted
line); (d) pump pressure plane. ................................................................................................... 127

Figure 6-10 Performance of the circuit using an open-center shuttle valve (OC-SHV) for a 41kg
attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line)
and b (dotted line); (d) pump pressure plane. ............................................................................. 127

Figure 6-11 Performance of the circuit that utilizes pilot-operated check valves for a 368kg
attached mass: (a) input control signal; (b) actuator velocity; (c) pressures at pump ports a (solid
line) and b (dotted line); (d) pump pressure plane. ..................................................................... 129

Figure 6-12 Performance of the circuit that utilizes a closed-center shuttle valve for a 368kg
attached mass: (a) input control signal; (b) actuator velocity; (c) pressures at pump ports a (solid
line) and b (dotted line); (d) pump pressure plane. ..................................................................... 129

Figure 6-13 Performance of the circuit that utilizes an open-center shuttle valve for a 368kg
attached mass: (a) input control signal; (b) actuator velocity; (c) pressures at pump ports a (solid
line) and b (dotted line); (d) pump pressure plane. ..................................................................... 130

xiv
Figure 6-14 Performance of the circuit with two pilot-operated check valves and two
counterbalance valves in retraction and extension of a 245-kg attached mass: (a) control signal; (b)
actuator velocity; (c) pressures at pump ports a (solid line) and b (dashed line); (d) pressures at
actuator ports a (solid line) and b (dashed line). ......................................................................... 133

Figure 6-15 Performance of Design 5, for 41 kg attached mass; (a) input control signal; (b) actuator
velocity; (c) pressures at pump ports a (solid line) and b (dotted line); (d) pressures at actuator
ports a (solid line) and b (dotted line). ........................................................................................ 135

Figure 6-16 Performance of Design 5, for 368 kg attached mass; (a) input control signal; (b)
actuator velocity; (c) pressures at pump ports a (solid line) and b (dotted line); (d) pressures at
actuator ports a (solid line) and b (dotted line). .......................................................................... 135

Figure 6-17 Calculated hydraulic power delivered/received by the main pump for 368-kg attached
mass in the conventional circuit that utilizes pilot-operated check valves (dotted line) and Deign
5, circuit with extra counterbalance valves (solid line). ............................................................. 136

Figure 6-18 Performance of Design 7, circuit with sequence valves, for a 41kg attached mass: (a)
input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line) and b (dotted line);
(d) pump pressure plane. ............................................................................................................. 138

Figure 6-19 Performance of Design 7, circuit with sequence valves, for a 368kg attached mass: (a)
input control signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line) and b
(dotted line); (d) pump pressure plane. ....................................................................................... 139

Figure 6-20 Performance of Design 7 for a 368-kg attached mass: (a) joystick input signal; (b)
actuator velocity; (c) pressures at the pump ports a (solid line) and b (dotted line); (d) pump
pressure plane.............................................................................................................................. 140

Figure 6-21 Hydraulic power delivered/consumed by the pump in Design 7 (solid line) and power
consumed by the pump in a load-sensing circuit (dashed line) for experiment in Figure 6-20. . 141

Figure 6-22 Performance of Design 8, circuit with a pilot-operated check valve and a sequence
valve, for 41 kg attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump
ports a (solid line) and b (dotted line); (d) pump pressure plane. ............................................... 143

xv
Figure 6-23 Performance of Design 8, circuit with a pilot-operated check valve and a sequence
valve, for 368 kg attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump
ports a (solid line) and b (dotted line); (d) pump pressure plane. ............................................... 144

Figure 6-24 Input control signal (dashed line) and corresponding actuator velocity (solid line) of
Design 8, for attached mass of: (a) 41 kg; (b) 368 kg................................................................. 145

Figure 6-25 The power delivered/consumed in Design 8 circuit for experiment in Figure 6-23: (a)
measured electric (solid line) versus calculated hydraulic (dashed line); (b) calculated hydraulic in
the proposed circuit (dashed line) and a comparable load-sensing circuit (solid line). .............. 146

Figure 6-26 3/3 closed-center shuttle valve; (a) exploited view; (b) schematic drawing. .......... 149

Figure 6-27 Performance of Design 2, circuit with a biased closed-center shuttle valve, for a 368-
kg attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid
line) and b (dotted line); (d) pump pressure plane. ..................................................................... 150

Figure 6-28 performance of Design 3, circuit with a biased open-center shuttle valve, for a 368-kg
attached mass: (a) control signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line)
and b (dotted line); (d) pump pressure plane. ............................................................................. 151

xvi
 LIST OF TABLES

Table 3-1 Values of parameters of circuit with pilot-operated check valves shown in Figure 3-6.
....................................................................................................................................................... 33

Table 3-2 Values of parameters of pilot-operated check valves shown in Figure 3-7. ................ 34

Table 3-3 Values of parameters of closed-center shuttle valve shown in Figure 3-14. ................ 43

Table 3-4 Values of parameters of open-center shuttle valve in circuit in Figure 3-18. .............. 51

Table 3-5 Performance index of the simulated circuits ................................................................ 58

Table 4-1 Values of parameters of the 4/3 closed-center shuttle valve shown in Figure 4-7. ...... 70

Table 4-2 Leakage areas at different input signals to pump. ........................................................ 77

Table 4-3 Throttling areas at different input signals to pump ...................................................... 85

Table 5-1 Hydraulic actuator parameters .................................................................................... 114

Table 5-2 Main pump unit parameters. ....................................................................................... 115

Table 5-3 Charge unit parameters. .............................................................................................. 116

Table 5-4 Instrument parameters. ............................................................................................... 117

Table 6-1 Specifications of the test rig. ...................................................................................... 120

Table 6-2 Specifications of compensating valves used in previous designs .............................. 125

Table 6-3 Specifications of compensating valves used in the new designs ................................ 131

xvii
 CHAPTER 1

1. INTRODUCTION

1.1 Statement of the problem

Hydraulic actuation systems (HASs) use liquid fluids to transmit power. They are an essential

part of most modern industrial applications. Aviation, machining, mining, agriculture and

construction are few examples of these applications. HASs possess fast response, high stiffness

and stability under variable working conditions. Moreover, they have high power to weight ratio

and that is why they are frequently used in mobile and airborne equipment.

Most current applications employ valve-controlled hydraulic actuation systems. Low efficiency

is one main disadvantage of the valve-controlled HASs. The total efficiency of such systems is

typically from 20% to 30% [1, 2]. The efficiency of an industrial machine is crucial, not only

because of the uprising energy cost but also due to the pollution and other environmental concerns.

Improvements in the efficiency of HASs reduce the environmental impact of industrial activities

and save energy. Many improvements to the valve-controlled hydraulic circuits have been

achieved. Developing the pressure compensated (PC) and load sensing (LS) hydraulic pumps and

introducing the open center (OC) control valves have improved the total efficiency of these

1
systems [3]. Yet, the efficiency of fluid power systems is relatively low compared to the efficiency

of mechanical or electrical transmission systems [2]. A recent study on an LS excavator reported

that only 31% of the energy delivered to the hydraulic system had turned into mechanical work at

the actuator, while valves and pump wasted 35% and 29% of that energy, respectively [2].

One solution to improve the efficiency of HASs is to utilize pumps to control motion instead of

metering valves. These systems are recognized as pump-controlled actuators or throttle-less

hydraulic circuits. Pump-controlled hydraulic actuators (PCHAs) are more efficient as compared

to the conventional valve-controlled HASs. Researchers at Bath University [4] reported that their

pump-controlled system consumed only 11% of the energy consumed by a comparable valve-

controlled circuit during one representative mode of operation.

Pump-controlled hydraulic circuits for double rod cylinders are readily available [5, 6].

However, single-rod cylinders are used in at least 80% of the electro-hydraulic systems [7]. Many

throttle-less hydraulic circuits to control single-rod cylinders have been designed and tested [8, 9,

10, 11, 12, 13]. However, none of these designs have yet been widely used in the market. This

indicates that there is no ideal pump-controlled hydraulic circuit for single-rod cylinders yet [12].

Performance and stability issues are the main problems facing these designs. This research

addresses the above-mentioned issues.

1.2 Objective of this research

The objectives of this research are to: (i) explore the concept of pump-controlled hydraulic

actuators as one effective solution to improve the efficiency of hydraulic actuation; (ii) evaluate

the existing designs and identify challenges that encounter their application on single-rod

cylinders; (iii) develop new solutions that enhance efficiency and performance; (iv) validate the

new designs through both simulations and experiments.

2
1.3 Methodology

This research is built up on theoretical analysis, simulation and experimental work. Theoretical

analysis is used to determine the low-performance zones in the common pump-controlled circuits

for single-rod cylinders and the different factors affecting them. Simulation studies are conducted

using the Simulink-MATLAB platform to evaluate the performance of both existing and new

designs. Simulations are also performed to study the performances of the circuits at various load

configurations that cannot be performed at the lab due to safety precautions. Simulations also

create a controlled environment to study the effect of single variable isolated from others.

Experimental studies are carried out to evaluate the realistic performance and to prove the

practicality of the new designs. The test rig, on which all experiments are conducted, is unique in

the sense that it is made up of a John Deere 48 (JD-48) excavator attachment; it allows full range

of four quadrants operation of the stick link. It can also examine different valve-controlled as well

as pump-controlled circuits at different loading conditions. The physical part of the test rig is

interfaced to the user by means of data acquisition board and real-time operating system. Different

previously-designed and proposed hydraulic circuits are prototyped and evaluated using this rig.

1.4 Thesis outline

This thesis is outlined as follows. Chapter 2 presents a detailed literature review of the pump-

controlled single-rod actuators. Chapter 3 discusses the causes of the undesirable performance

regions in three previously-designed circuits. Proposed concepts and designs to improve

performance of the pump-controlled single-rod cylinders are introduced in Chapter 4. Layouts and

description of the test rig and its different components are shown in Chapter 5. Experimental results

of evaluation of selected circuits are presented in Chapter 6. Finally, the contributions and future

work are outlined in Chapter 7.

3
 CHAPTER 2

2. LITERATURE REVIEW

2.1 Background on hydraulic actuation systems

Fluid power is the use of pressurized fluids to transmit power. It can be classified into two

categories: hydraulic and pneumatic systems. Hydraulic actuation systems (HASs) are the systems

that use pressurized liquids as a working medium. A conventional valve-controlled HAS consists

of three main components: a hydraulic pump that pressurizes the hydraulic fluid, valves that

regulate fluid pressure, direction and rate of flow, and an actuator that actuates the load. Linear

motion is actuated by hydraulic cylinders, whereas rotary motion is actuated by hydraulic motors.

HASs incorporate many other components like tanks, accumulators, filters and transmission lines.

A typical valve-controlled hydraulic circuit is shown in Figure 2-1.

Valve-controlled HASs possess fast response, high stiffness, high power to weight ratio and

stability under variable loading conditions. HASs are an essential part of most modern industrial

applications. Aviation and aerospace, marine, machining and manufacturing, mining, agriculture

and excavation machines are examples of these applications. Low efficiency is one main

disadvantage of the valve-controlled HASs. The total efficiency of such systems is around 20% to

30% [1, 2]. Efficiency of an industrial machine is crucial because of the energy cost and

4
environmental concerns. Conventional mobile machines, like excavators, incorporate a centralized

hydraulic power unit that delivers the pressurized hydraulic fluid to the metering valves which in

turn control the actuators. Centralized hydraulic systems are not only bulky, heavy, and noisy, but

they also increase the weight and cost of the machine [14]. Improvements in the efficiency of

HASs reduce the environmental impact of industrial activities and save energy. Many

improvements to the conventional hydraulic circuit components have been achieved. Developing

the pressure compensated (PC) and load sensing (LS) hydraulic pumps have improved the total

efficiency of hydraulic systems [3].

Actuator
Load

Control valve
Pump
M

Relief
Filter valve

Tank

Figure 2-1 Typical valve-controlled hydraulic circuit.

Power utilization in valve-controlled HASs equipped with three different pump types is

illustrated in Figure 2-2. Systems that incorporate a fixed displacement pump have the greatest

power loss as the pump provides its maximum flow rate regardless of motion requirements (Figure

2-2 (a)). A relief valve controls maximum pressure in the system and direct extra flow to the tank.

PC pumps provide fluid only when the system pressure goes lower than a prescribed nominal

value, Ps, with flow rate just a little more than that required by the actuators. The extra amount of

5
flow, ΔQL, is attributed to internal leakage in the system components (Figure 2-2 (b)). LS pumps

maintain a constant pressure margin, ΔPLS, above the highest loaded actuator in the circuit, where

actuators 1, 2 and 3 control different loads in a multi-actuator system (Figure 2-2 (c)). Furthermore,

they provide the system with flow rate just a little more than what is required by the actuators.

Ps PsP Ps P
Losses ΔPLS
Losses Losses

Pressure
Pressure
Pressure

Losses

ΔQL
ΔQL

Actuator 2

Actuator 2
Actuator 2

Actuator 1
Actuator 3
Actuator1
Actuator 1

Actuator 3
Actuator 3

Flow Qmax Flow Qmax Flow Qmax

(a) (b) (c)

Figure 2-2 Power losses in valve-controlled hydraulic circuits for different pump types: (a) fixed
displacement; (b) pressure compensated; (c) load sensing.

Despite all the new developments in conventional hydraulic circuits, efficiency is still a big

concern. Consider the case of a popular excavator machine that is energized by a 36.5 KW diesel

engine. Only 31% of the energy delivered to the valve-controlled hydraulic system is used for

digging and lifting loads, while 35% and 29% of the energy are lost in valves and pump,

respectively [2]. Energy losses in pump is attributed to both volumetric and mechanical

efficiencies of its internal components [15]. Whereas losses in control valves is attributed to

hydraulic resistance in valve orifices and is considered as the product of pressure drop across valve

and flow [15]. Energy losses heat up the hydraulic oil, and, consequently, oil coolers that consume

more energy are needed to dissipate such excessive heat. On the other hand, the amount of

pollutants emitted by the machine in the form of carbon monoxide, nitrogen dioxide, and nitrogen

monoxide is considerable [7].

6
2.2 Pump-controlled hydraulic actuation

One significant solution to improve the efficiency of HASs is to design a system with no

metering valves. These systems are controlled through the pump and are recognized as pump-

controlled or throttle-less hydraulic systems [9]. In these circuits, pumps provide the exact amount

of flow required to achieve the specified actuator velocity. Accumulation of fluid in actuator

creates the exact pressure needed to perform the task, which leads to significant power savings

[16]. When a hydraulic cylinder is controlled by a variable displacement pump, the actuator is

variable displacement pump-controlled. However, when a variable RPM pump is utilized to

control the cylinder, it is variable speed pump-controlled or electro-hydrostatic actuator (EHA)

[12, 17, 18]. These systems do not incorporate a central hydraulic power unit. Instead, each

cylinder is generally controlled by a single pump. Pump-controlled hydraulic systems can be

classified in different ways. They can be classified according to the control methodology of the

pump displacement, cylinder type, circuit configuration and number of pumps in the system.

Before getting into the different classifications of the pump-controlled circuits, the quadrant of

operation of both actuator and pump are addressed.

2.2.1 Four quadrants of operation

There are two common concepts of the quadrants of operation in pump-controlled circuits:

actuator loading quadrants and pump mode quadrants. Figure 2-3 (a) shows sign convention where

cylinder velocity, va, is positive when the cylinder extends and the load force, FL, is positive when

acting against cylinder extension. Based on motion and load directions, the cylinder operates in

one of the four quadrants shown in Figure 2-3 (b). In the first and third quadrants, the cylinder

extends and retracts, respectively, against a resistive load. In both cases, energy is delivered from

7
the hydraulic circuit to the actuator to perform motion. In the second and forth quadrants, the

assistive load drives the cylinder extension and retraction, respectively.

va
va II. Assistive І. Resistive
FL
FL
III. Resistive IV. Assistive

(a) (b)
Figure 2-3 Single-rod cylinder: (a) sign convention; (b) quadrants of operation.

The quadrants of operation of the pump depend on the pump mode of operation (pumping or

motoring). First, the pressure difference across the pump is conventionally defined as

𝑃 = (𝑝𝑎 − 𝑝𝑏 ) where 𝑝𝑎 and 𝑝𝑏 are the pressures at the pump ports connected to the cap side and

rod side of the cylinder, respectively, Q is defined as the oil flow through the pump. Q is positive

if the flow direction is from port ‘b’ to port ‘a’ in the pump. Figure 2-4 illustrates the pump

operational quadrants according to the pump mode operation. When P and Q possess the same

sign, the pump works in pumping mode; it receives energy from the prime mover and transfers it

to the hydraulic circuit. When P and Q have different signs, the pump works in motoring mode; it

receives energy from the hydraulic circuit and delivers it to the prime mover.

pa Q
a II. Motoring І. Pumping

Q P= pa -pb

b III. Pumping IV. Motoring

pb

(a) (b)
Figure 2-4 Defining quadrants in hydrostatic pumps: (a) sign convention; (b) quadrants of
operation.

8
2.2.2 Single versus double rod actuators

There are two main types of hydraulic linear actuators; single-rod or asymmetric cylinders and

double rod or symmetric cylinders. A double rod hydraulic cylinder is larger than a comparable

single-rod cylinder. Therefore, they require larger space for installation and operation. Pump-

controlled hydraulic circuits for double rod cylinders are well developed and are already in use by

some critical applications including aviation. However, single-rod cylinders are used by at least

80% of the electro-hydraulic systems [7]. The challenge of implementing pump-controlled circuits

for single-rod cylinder is to find the proper way to compensate for the differential flow at both

sides of the cylinder and to maintain acceptable performance [3, 7]. Figure 2-5 shows simplified

illustrations of the pump-controlled circuit for double and single-rod cylinders. The focus of this

research is on the single-rod cylinders.

Motion Motion

Q Q+ ΔQ Q
Q
Flow Compensation
module

Q Q Q Q

M M

(a) (b)
Figure 2-5 Pump-controlled actuator: (a) double rod; (b) single-rod.

2.2.3 Variable displacement versus variable speed pumps

Two main strategies to control pump flow in pump-controlled circuits are variable speed and

variable displacement pump control. In variable speed pump circuits, a fixed displacement pump

is connected to a variable speed prime-mover (see Figure 2-6 (a)), while a variable displacement

pump is connected to a fixed speed prime-mover in variable displacement pump circuits (see
9
Figure 2-6 (b)). Variable displacement pump-controlled actuators possess better performance and

efficiency compared to variable speed pump-controlled actuators. Moreover, they are more

appropriate to run them on diesel engines fitted on conventional excavator machines. This research

focuses on variable displacement pump-controlled actuators.

Motion Motion

Q Q Q Q

Q Q Q Q

M M

(a) (b)
Figure 2-6 Pump-controlled actuator: (a) variable speed; (b) variable displacement.

2.2.4 Conventional versus asymmetric pumps

Generally, there are two holes in the port plate of an axial piston pump. These holes are

connected to the two conventional IN and OUT ports of the pump. New designs with more than

two holes in the port are now prototyped and being tested. Thus, according to number port plates

in an axial piton pump, two main types are used to compensate for the differential flow of a single-

rod hydraulic cylinders: (i) conventional pumps and (ii), asymmetric axial piston pumps [19].

Asymmetric pumps contain more than two flow ports conventional pumps have. They compensate

for the differential flow through the extra port(s). Figure 2-7 shows a typical hydraulic circuit

incorporating an asymmetric pump. Note that conventional hydraulic pumps are well developed,

affordable and available as compared to symmetric pumps. This research focuses on the use of

conventional pumps.

10
 Motion

pa pb
Q+ ΔQ Q

pa pb
Q+ ΔQ Q
M Pc
ΔQ
pc

(a) (b)
Figure 2-7 Asymmetric pump-controlled circuit: (a) layout; (b) ports plate.

2.2.5 Single versus multi-pump configurations

In pump-controlled circuits, an actuator is controlled by a single pump or multi-pump

configuration. Generally, in the two-pump configuration systems, each pump commonly controls

the flow of one side of the single-rod cylinder, see Figures 2-8 (a) and (b). However, there are

different configurations depending on the application requirements. Cleasby and Plummer [4]

designed a hydraulic flight simulator using a special hydraulic cylinder with 1:2 area ratio (see

Figure 2-8 (c)). They used a variable speed electric motor coupled to a tandem pump unit that is

formed by two identical fixed displacement pumps. Lodewyks [20] utilized the hydraulic

transformer to compensate for the differential flow of a single-rod cylinder (Figure 2-8 (d)). A

hydraulic transformer basically consists of two mechanically-connected hydraulic units, one of

them functions as a motor while the other functions as a pump. Similar to the function of an

electrical transformer, a hydraulic transformer converts an input flow at a certain pressure level to

a different output flow at the expense of the pressure level. Generally, the transformer ratio (flow

ratio between the two units) has to match the area ratio of the cylinder. Utilizing more than one

pump to control each hydraulic cylinder makes the machine heavy, noisy, and expensive.

Moreover, synchronizing the displacements of the two pumps to match the cylinder flow rate in

11
real-time operation is a challenging task [12]. The focus of this study is on circuits with a single

pump.

(a) (b)

Transformer
unit

(c) (d)
Figure 2-8 Multi-pump circuits: (a) variable displacement controlled actuator; (b) variable speed
controlled actuator; (c) identical pump control of 1:2 area ratio cylinder; (d) utilization of hydraulic
transformer.

2.2.6 Closed-circuit versus open-circuit configurations

In a closed hydraulic circuit, the pump does not have pre-defined high and low-pressure ports.

In an open circuit, the pump operates against high pressure only on one side [21]. From a

hydrostatic transmission viewpoint, a closed circuit is the one in which oil leaving the actuator is

immediately directed to the pump input, whereas in an open circuit it is directed to the reservoir

12
from where it is pumped again to the circuit [22]. Figure 2-9 shows schematic drawings of both

closed circuit and open circuit pump-controlled circuits.

Flow compensation Flow compensation

unit unit

(a) (b)
Figure 2-9 Pump-controlled actuator circuits: (a) closed circuit; (b) open circuit.

2.2.7 Pressurized versus non-pressurized reservoirs

Oversizing inlet port and pressurizing the low-pressure side of pump in a hydraulic circuit are

two main methods to avoid pump cavitation [23]. Both solutions are applicable for the open-

circuits. However, only the second solution is applicable for closed-circuits. In closed hydrostatic

circuits, a charging subsystem that provides low-pressure source (for example 200 psi) to the inlet

recommended port of the main pump is highly needed. It is required also for cooling and

lubricating the pump internal components. Further, it is needed to actuate the control mechanism

in variable displacement pumps. Low-pressure charging system improves pump protection and

reduces the contamination problems [23]. Increasing system cost and complexity, and consuming

extra energy are the main drawbacks of the charging system.

2.3 Previous work

Many pump-controlled hydraulic circuits to control single-rod hydraulic cylinders have been

designed and tested [8, 9, 10, 11, 12, 13]. However, none of these designs have been widely used
13
in the market which shows that there are no ideal pump-controlled single-rod cylinders, yet.

Hewett [8] patented the concept of displacement control of single-rod cylinders using the closed-

circuit shown in Figure 2-10. He utilized a 2-position 3-way shuttle valve (2/3 SHV) and two check

valves to compensate for the cylinder differential flow. The pump in this circuit can operate in the

four quadrants of operation.

2/3 shuttle valve

Figure 2-10 Circuit that utilizes 2-position 3-way shuttle valve, developed by Hewett [8].

Rahmfeld and Ivantysynova [9] introduced a closed-circuit pump-controlled single-rod cylinder

shown in Figure 2-11. Their design comprises a variable displacement piston pump and two pilot-

operated check valves (POCVs) to compensate for cylinder differential flow. The pump operates

in the four quadrants and recuperates energy during assistive load quadrants. Hippalgaonkar and

Ivantysynova [24] and Grabbel and Ivantysynova [25] applied the above circuit to a concrete pump

truck, a loader, and a multi-joint manipulator. Efficiency improvement and weight reduction were

reported in machines that utilized this design [26, 27]. However, Williamson and Ivantysynova

[28] and Wang et al. [11] reported that circuit with POCVs experience undesirable pump mode

14
and actuator velocity oscillations when lowering light loads at high speeds. Yuming et al. [29]

studied circuit instability during operation in switching zones via Lyapunov exponents.

Pilot-operated
check valves
M

Figure 2-11 Circuit with two pilot-operated check valves, by Rahmfeld and Ivantysynova [9].

Figure 2-12 shows the circuit proposed by Wang et al. [11]. In their design, A 3/3 shuttle valve

and two check valves to compensate for the cylinder differential flow are used. They implemented

two extra regulating valves they allow some oil leakage to deal with oscillatory behavior. An

additional control loop that requires real-time actuator pressures and position was needed to

operate the regulating valves.

Jalayeri et al. [5, 12] proposed controlling the load motion with the help of counterbalance

valves. They used On/Off solenoid valve and a check valve to compensate for the differential flow

of the single-rod cylinder. Figure 2-13 shows the schematic of their circuit. They reported that

their design, compared to the conventional circuits, is energy efficient and accurate enough for

applications such as tote dampers and elevators. However, this design cannot regenerate energy

and presented time lags of 0.4 s to 0.8 s [12].

15
 2 4

3
3/3 closed-center
shuttle valve
Regulating

2
valves

1
2

2
1

1
M
M

Figure 2-12 Circuit that uses 3/3 closed-center shuttle valve, proposed by Wang et al. [11].

Counterbalance
1
1

valves
3
3

2
2

.1
1
1
1

On/Off
valve

Figure 2-13 Circuit that utilizes counterbalance valves proposed by Jalayeri et al. [5, 12].

Similar to the Jalayeri et al design, Altare and Vacca [30, 31] proposed a compact electro-

hydraulic actuator shown in Figure 2-14. Their circuit comprises a miniature bi-directional external

gear pump, two counterbalance valves and a dual pressure valve (similar to shuttle valves) to

16
compensate for the differential flow. Simulations of their circuit showed that the counterbalance

valves guarantee load holding without energy consumption and improved system efficiency. The

overall size of the system was reported to be a 300 mm × 90 mm × 70 mm for a maximum force

of 2000 N, speed up to 20 mm/s and power up to 70 W.

LVDT

Counterbalance valves
1

1
3

2
2

Dual pressure valve

M
M

Controller

Figure 2-14 A miniature pump-controlled circuit with counterbalance valves proposed by Altare
and Vacca [30, 31].

In order to reduce complexity resulted from using the extra control circuit, Caliskan et al. [13]

proposed a modified version of the circuit that utilizes a 3/3 CC-SHV described in [11]. They

utilized a 3/3 open-center shuttle valve (OC-SHV) that compensates for the cylinder differential

flow in addition to stabilizing the circuit through oil leakage, as shown in Figure 2-15. They

reported that their circuit works best under certain actuator velocities. However, their experimental

work was limited to low loading conditions and lacks the effect of mass inertia.

17
 3/3 open-center
shuttle valve

2
1

1
M

Figure 2-15 Circuit with open-center 3-position shuttle valve, proposed by Caliskan et al. [13].

2.4 Summary

It is seen that pump-controlled hydraulic circuits possess superior efficiency but show less

performance compared to that of valve-controlled hydraulic circuits. Pump-controlled circuits for

double-rod cylinders have been well-developed. However, single-rod cylinders are commonly

used in at least 80% of the electro-hydraulic applications. In spite of the many initiatives, pump-

controlled single-rod cylinder solutions face performance issues during specific modes of

operation. This research focuses on pump-controlled single-rod cylinders due to their high

efficiency, energy regeneration ability, popularity and compact size. This study begins with

evaluating selected designs of existing pump-controlled single-rod actuators, recognizing their

performance challenges and developing new solutions that enhance circuits performance. The new

designs are then validated through both simulations and experimental work.

18
 CHAPTER 3

EVALUATION OF SELECTED DESIGNS

Based on the literature review, three promising circuits are selected in this thesis to be

thoroughly studied and evaluated in this chapter. These circuits are: (i) the circuit that utilizes two

pilot-operated check valves (POCVs) [9], (ii) the circuit that uses a closed-center shuttle valve

(CC-SHV) [11] and (iii) the circuit that utilizes an open-center shuttle valve (OC-SHV) [13]. First,

a detailed explanation of each circuit operation in all four quadrants and reasons for the undesirable

performance is provided. Identification of the undesirable performance region is conducted

through theoretical analyses and simulations. A comparison between the three circuits in terms of

critical zone size and characteristics of actuator oscillations are provided.

3.1 Circuit with two pilot operated check valves

Figures 3-1 and 3-2 illustrate the detailed and simplified schematics of the pump-controlled

circuit introduced by Ivantysynova and Rahmfeld to control single-rod actuator. The circuit

comprises a variable-displacement bi-directional swash-plate piston pump, two pilot operated

check valves (POCVs) and low-pressure charge system (CH). Typically, the main pump is

equipped with a servo-controlling mechanism, high pressure relief valves (HP-RV) and anti-

19
cavitation check valves at each port, as seen in Figure 3-1. Note that the main pump works as a

hydraulic pump in two operational quadrants while it works as a hydraulic motor in other two

quadrants. Thus, from a functionality perspective, it is described as pump/motor. However, in this

thesis, and for the sake of simplicity, it is called a pump that works in pumping mode or motoring

mode. Low-pressure charge system consists of low-pressure pump, accumulator and low-pressure

valve (LPRV).

Actuator
Pilot-operated
check valves

Low pressure
accumulator

Charge Low pressure Anti-cavitation

pump relief valve valve
1

2
1
2

High pressure
M
relief valve

Charge system

Main pump Control

M

M mechanism

Figure 3-1 Circuit that utilizes pilot operated check valve.

Figure 3-2 illustrates the simplified representation of the abovementioned circuit. POCVs are

used to compensate for the cylinder differential flow; they are opened by pilot signals from cross

pressure lines as the circuit operates in the four quadrants of operation. Consider extending the

20
actuator against the resistive external load, as shown in Figure 3-2. The pump delivers flow Q in

clockwise direction to the cap side of the cylinder through main transmission Line A. As the

pressure in Line A (𝑝𝑎 , 𝑝1 and 𝑝𝐴 ) builds up, it opens the cross pilot operated check valve,

POCVB. Consequently, the charge line is connected to Line B, which allows flow, 𝑄2 , to

compensate for the cylinder differential flow. In this case, the main pump works in pumping mode

and the actuator works in resistive mode. Detailed definitions of quadrants of operation of pump

and actuator have been discussed in Section (2.2.1).

F fr xa , va

pA FL
pB

QA QB
CH

p2
POCVA POCVB
pc Qc
Q1 Q2
p1 Line B

Line A

Qa Qb
pa pb

Figure 3-2 Simplified drawing of circuit that utilizes pilot operated check valves in the first
quadrant of operation; CH denotes charge system.

21
3.1.1 Previous identification of poor performance zone

Williamson and Ivantysynova to [28] observed a low performance region during specific mode

of operation in the circuit that uses the POCVs. More specifically, they reported that undesirable

pump mode and actuator velocity oscillations are experienced when lowering light loads at high

speeds (assistive retraction mode). Figure 3-3 illustrates the region where the circuit experienced

oscillatory performance on the actuator load-velocity 𝐹𝐿 -𝑣𝑎 plane. The maximum actuating force

and maximum actuator velocity in both extension and retraction are also shown in Figure 3-3.

𝑄𝑚𝑎𝑥 νa
𝛼𝐴𝐴

IV. Motoring

𝑄𝑚𝑎𝑥 I. Pumping
𝐴𝐴

𝐴𝐴 (𝑝𝐴𝑚𝑎𝑥 − 𝛼𝑝𝑐 )

𝐹𝐿
−𝐴𝐴 (𝛼𝑝𝐵𝑚𝑎𝑥 − 𝑝𝑐 )
𝐹𝑐𝑟 = (1 − α)𝑝𝑐

Critical zone II. Motoring

III. Pumping

Figure 3-3 Actuator operating plane illustrating the four quadrants of operation, maximum
actuating forces, and maximum velocities in extension and retraction, and low-performance zone
according to [28].

Maximum actuating force at the cylinder equals 𝐴𝐴 (𝑝𝐴𝑚𝑎𝑥 − 𝛼𝑝𝑐 ) and maximum velocity is
𝑄𝑚𝑎𝑥
on the right-hand side of the 𝐹𝐿 -𝑣𝑎 plane. Whereas, maximum actuating force at the cylinder
𝐴𝐴

𝑄𝑚𝑎𝑥
is 𝐴𝐴 (𝛼𝑝𝐵𝑚𝑎𝑥 − 𝑝𝑐 ) and maximum velocity is on the left-hand side of the 𝐹𝐿 -𝑣𝑎 plane. The
𝛼𝐴𝐴

𝐴
actuator area ratio 𝛼 = 𝐴𝐵 and maximum applied pressures at cap and rod sides of the actuator are
𝐴

22
𝑝𝐴𝑚𝑎𝑥 and 𝑝𝐵𝑚𝑎𝑥 , respectively. Thus, the actuator deals with higher forces at lower velocities on

the right-hand side of the 𝐹𝐿 -𝑣𝑎 plane compared to that on the left-hand side. It is also observed

that switching between the right and left-hand sides of the plane occurs at a biased force value,

𝐹𝑐𝑟 = 𝑝𝑐 𝐴𝐴 (1 − α). In order to better understand the operation of the pump-controlled circuit in

different loading conditions, a detailed explanation of the quadrants of operation and switching

zones between successive quadrants are presented in the next section.

3.1.2 Four quadrants of operation and switching zones

Figure 3-4 shows the different flow patterns in each of the four quadrants of operation for circuit

represented in Figure 3-2. The operational sequence represented in Figure 3-4 is as follows:

resistive extension (Q1), assistive retraction (Q4), resistive retraction (Q3) and assistive extension

(Q2). Apparently, the arrows between quadrants represent the switching zones. Switching zones

between each two successive quadrants are denoted as S14, S43, S32 and S21.

Observe that while the operational status (opened / closed) of the POCVs does not change at

zones S14 and S32, operation of the POCVs are switched at zones S43 and S21. Reconfiguration

of the compensating valves causes abrupt interruption to the system dynamics. These variations

are accompanied by changes in pump operating mode and in actuator velocity. Thus, S43 and S21

can be recognized as regions of potentially poor performance. Operation of the POCVs depends

on pressure values at circuit main lines (A and B). Accordingly, proper relation between the

pressures at both sides of the circuit is essential for desirable operation of the circuit. Pressure

plane graph that illustrates the relation between circuit’s main lines pressures is discussed in the

next sections. Note that two graphs are regularly used by researchers to illustrate performance of

the pump-controlled circuits and facilitate their analysis. The first graph is the circuit operating

plane that illustrates the actuator velocity versus external load 𝐹𝐿 -𝑣𝑎 plane.

23
Figure 3-4 Flow patterns in four quadrants of operation (Q1, Q2, Q3, Q4) and switching zones
(S14, S43, S32, S21) of circuit that utilizes pilot operated check valve in the load-velocity plane.

The 𝐹𝐿 -𝑣𝑎 plane is considered a combination of the actuator and pump operating planes shown

in Figures 2-3 and 2-4, respectively. The second graph is the pressure 𝑝𝑎 − 𝑝𝑏 plane that relates

the pressures at both sides of the circuit. It illustrates Line B versus Line A pressures. Pressures at

both sides of the circuit controls compensating valve(s) operation that highly affect circuit

performance.

24
3.1.3 Modeling

The mathematical model of the circuit with POCVs is developed to simulate performance of

the circuit and identify critical zone location. Figure 3-6 shows the schematic drawing of the circuit

with notations in the first quadrant of operation. Mathematical equations used to model the

different components in the circuit are described hereafter.

Figure 3-5 An excavator stick in four quadrants of operation.

3.1.3.1 Actuator model

The mathematical model of the actuator is represented by the piston equation of motion and

continuity equations at both chambers of the cylinder as follows (refer to Figure 3-6):

𝑚𝑣̇𝑎 = ( 𝑝𝐴 𝐴𝑎 − 𝑝𝐵 𝐴𝑏 ) − 𝐹𝑓 − 𝐹𝐿 (3-1)

𝐾𝑜𝑖𝑙
𝑝𝐴̇ = (𝑄𝐴 − 𝐴𝐴 𝑣𝑎 ) (3-2)
𝑉𝐴

𝐾𝑜𝑖𝑙
𝑝𝐵̇ = (−𝑄𝐵 + 𝐴𝐵 𝑣𝑎 ) (3-3)
𝑉𝐵

25
where 𝑚 represents the moving mass; 𝐴𝑎 and 𝐴𝑏 are piston effective areas at cap and rod sides,

respectively. 𝐾𝑜𝑖𝑙 denotes the effective bulk modulus of the system including compressibility of

the fluid and elasticity of the conduits and components [15]; 𝑉𝐴 = 𝑉0A + (𝐴𝐴 𝑥𝑎 ) and 𝑉𝐵 = 𝑉0B −

(𝐴𝐵 𝑥𝑎 ); 𝑉0A and 𝑉0B are the initial volumes of oil in the cap and rode sides of the actuator,

respectively. 𝐹𝑓 𝑎𝑛𝑑 𝐹𝐿 are frictions and external load forces at actuator.

F fr xa , va
AA AB FL
pA pB

QA QB

CH

p2
POCVA POCVB
P c Qc
Q2
Q1
p1 Line B

Line A

Qa Qb
pa pb

Figure 3-6 Schematic drawing of circuit that utilizes POCVs in the first quadrant of operation.

Friction force, 𝐹𝑓 , is assumed to be summation of the Coulomb and viscous friction forces.

Coulomb friction force in the internal oil seals of the hydraulic actuator is considered a function

of the seal preload force and pressures at both chambers of the actuator [32]. Friction force

equations are represented as follows:

26
 𝐹𝑓 = 𝐹𝐶 (1 + (𝐾𝑏 − 1) 𝑒 −𝑐𝑣|𝑣𝑎 | )𝑠𝑔𝑛(𝑣𝑎 ) + 𝑓𝑣 𝑣𝑎 (3-4)

𝐹𝐶 = 𝐹𝑝𝑟 + 𝑓𝑐𝑓𝑟 (𝑝𝐴 + 𝑝𝐵 ) (3-5)

where 𝐹𝐶 is the Coulomb friction force; 𝐾𝑏 and 𝑐𝑣 represent breakaway friction force increase and

velocity transition coefficients, respectively; 𝑓𝑣 and 𝑓𝑐𝑓𝑟 denote the viscous and Coulomb friction

coefficients, respectively; 𝐹𝑝𝑟 is the preload force needed to fit oil seals into place. 𝑠𝑔𝑛(𝑣𝑎 ) is

+1, 0 and − 1 for 𝑣𝑎 > 0, 𝑣𝑎 = 0, and 𝑣𝑎 < 0 , respectively.

3.1.3.2 Pump model

The pump internal flow leakage is considered zero and pump flow is considered a linear
function of the pump control voltage, pump flow is represented through the following equations:

𝑄𝑎 = 𝑄𝑏 = 𝑄𝑝 (3-6)

𝑄𝑝 = 𝐾𝑝𝑚𝑝 𝑉𝑒 (3-7)

where 𝐾𝑝𝑚𝑝 is the linearized coefficient for pump flow; Ve denotes input control signal to pump.

3.1.3.3 Flow balance in the circuit

The flow balance equations with regard to the pump and actuator flows at both sides of the

circuit are as follows:

𝑄𝐴 = 𝑄𝑎 − 𝑄1 (3-8)

𝑄𝐵 = 𝑄𝑏 − 𝑄2 (3-9)

3.1.3.4 Transmission line losses

Transmission losses between the pump and actuator ports are introduced using the lumped

resistance model. This model relates the pressure drop in a transmission line to resistance of the

viscous laminar flow in line and the local losses due to momentum change in fittings and different

restrictions. The pressure drop in a transmission line is represented as follows [33]:

27
 ∆𝑝 = 𝐶𝑇𝑣 𝑄 + 𝐶𝑇𝑙 𝑄2 (3-10)

where the combined viscous friction coefficient in transmission lines, 𝐶𝑇𝑣 , is calculated as follows:
128𝜇𝐿
𝐶𝑇𝑣 = is the combined viscous friction coefficient in transmission lines, and 𝐶𝑇𝑙 =
𝜋𝐷 4

8𝜌 𝜁
(𝜋2 ) ∑ 𝐷4 is the local drag coefficient, where 𝜇 is the fluid dynamic viscosity; 𝐿 and 𝐷 are the

transmission line length and diameter; fluid density is denoted as 𝜌; 𝜁 represents the local losses

coefficient.

3.1.3.5 Pilot operated check valves.

Figure 3-7 (a) illustrates a schematic drawing of the pilot operated check valve positioned in

the right-hand side of the circuit (POCVB). Due to similarity between both pilot operated check

valves, only the mathematical model of POCVB is deduced in this section.

charge line (pc) Line B 𝐴𝑐𝑣𝐵

𝐴𝑚𝑎𝑥𝐵

pilot signal
(p1) poppet
spring, ksv

𝐴𝑙𝑘
Avp
𝑝𝑐𝑟 𝑝𝑜𝑚𝑥 𝑝𝑒𝐵
Avs xv
(a) (b)
Figure 3-7 Pilot-operated check valve: (a) schematic drawing; (b) opening area versus effective
pressure.

According to the notations in Figure 3-6, the flow continuity through POCVB opening is

represented by the following equation [15]:

2
𝑄2 = 𝐶𝑑 𝐴𝑐𝑣𝐵 √ |𝑝𝑐 − 𝑝2 | 𝑠𝑖𝑔𝑛 (𝑝𝑐 − 𝑝2 ) (3-11)
𝜌

28
where 𝐶𝑑 is the flow discharge coefficient through the valve opening and 𝐴𝑐𝑣𝐵 represent the

opening area of POCVB.

The cracking condition of POCVs depends on the valve piloting and outlet pressures referenced

to the charge pressure [34]. The dynamics of the POCVs had slight effect on the overall circuit

dynamics, and, thus, it was neglected. Instead, the static force balance of the spool/poppet is used.

As an example, the static force balance of POCVB poppet is as follows:

𝐴𝑣𝑝 (𝑝1 − 𝑝𝑐 ) − 𝐴𝑣𝑠 (𝑝2 − 𝑝𝑐 ) − (𝐹𝑘0 + 𝑘𝑠𝑣 𝑥𝑣 ) = 0 (3-12)

where 𝐴𝑣𝑝 and 𝐴𝑣𝑠 are the pilot piston and valve poppet effective areas, respectively. 𝐹𝑘0 and 𝑘𝑠𝑣

represent the spring initial force and stiffness. The poppet displacement is denoted as 𝑥𝑣 .

By dividing the above-mentioned equation by the poppet effective area, 𝐴𝑣𝑠 , and considering

the critical opening condition of the valve, i.e. 𝑥𝑣 ≈ 0, the following pressure balance equation is

obtained:

𝐾𝑝𝑖𝑙𝑜𝑡 (𝑝1 − 𝑝𝑐 ) − (𝑝2 − 𝑝𝑐 ) = 𝑝𝑐𝑟 (3-13)

where 𝐾𝑝𝑖𝑙𝑜𝑡 is the piloting ratio of the valve that equal 𝐴𝑣𝑝 /𝐴𝑣𝑠 . For simplicity, the effective

opening pressure, 𝑝𝑒𝐵 , is introduced as the summation of the opining pressures of POCVB, i.e.

𝑝𝑒𝐵 = 𝐾𝑝𝑖𝑙𝑜𝑡 (𝑝1 − 𝑝𝑐 ) − (𝑝2 − 𝑝𝑐 ). Assuming that, valve opening area is linear function of the

effective opening pressure, 𝑝𝑒𝐵 , as shown in Figure 3-7 (b), then the opening area of the valve is

represented as follows:

𝐴𝑙𝑘 𝑝𝑒𝐵 ≤ 𝑝𝑐𝑟𝐵

𝐴𝑐𝑣𝐵 = { 𝑘1 (𝑝𝑒𝐵 − 𝑝𝑐𝑟𝐵 ) 𝑝𝑐𝑟𝐵 < 𝑝𝑒𝐵 < 𝑝𝑜𝑚𝑥𝐵 (3-14)
𝐴𝑚𝑎𝑥𝐵 𝑝𝑒𝐵 ≥ 𝑝𝑜𝑚𝑥𝐵

where 𝑘1 is the valve opening area coefficient. 𝑝𝑜𝑚𝑥𝐵 and 𝐴𝑚𝑎𝑥𝐵 represent the pressure required

to fully open valve and the maximum opening area, respectively.

29
3.1.4 Discussions

From the above POCVs governing equations, it is clear that the critical operating conditions in

this circuit occurs when pressures at both sides of the circuit are close to each other. In this case,

both POCVs operation is sensitive to operating conditions. In other words, it can be easily opened

closed with the slightest change in pressure at either side of the circuit. The effects of transition

lines losses, POCVs characteristics, and frictional force to the location and shape of the undesirable

regions (critical zones) are illustrated hereafter. To include the transmission lines losses effect, the

critical force, 𝐹𝑐𝑟 , is re-defined as the actuating force at the cylinder when both pressures at the

POCVs pilot ports are equal, i.e., 𝐹𝑐𝑟 = 𝐴𝐴 (𝑝𝐴 − 𝛼𝑝𝐵 )|𝑝1 =𝑝2 . Note that pressure values at the

cylinder ports and POCVs piloting ports are different due to pressure losses in transmission lines.

Transmission line losses is related to fluid flow rates through a quadratic relation as in Eq. (3-10).

Accordingly, the critical force regions are represented by a quadratic curve as shown in Figure 3-8.

Initial value of the critical force at zero velocity is 𝐹𝑐𝑟0 = 𝐹𝑐𝑟 |𝑣𝑎=0 = 𝐴𝐴 𝑝𝑐 (1 − 𝛼).

The width of the critical zone in the circuits with POCVs (difference between 𝐹𝐿10 and 𝐹𝐿20

and between 𝐹𝐿30 and 𝐹𝐿40 in Figure 3-8) depends on the cracking pressures of the POCVs and the

actuator piston areas. Let the force 𝐹𝐶𝑉 be defined as the corresponding force created at the

cylinder due to the extra pressure required for opening the POCV. Consider the case that the

pressures at both sides of the circuit are equal, then ∆𝑝𝑐𝑟 is the extra pressure at either side of the

circuit needed to pilot open the cross POCV. Note that the actuator force 𝐹𝐶𝑉𝐴 corresponding to

∆𝑝𝑐𝑟 needed to open POCVA, 𝐹𝐶𝑉𝐴 = ∆𝑝𝑐𝑟 𝐴𝐴 , is higher than the force 𝐹𝐶𝑉𝐵 needed to open

POCVB, 𝐹𝐶𝑉𝐵 = ∆𝑝𝑐𝑟 𝐴𝐵 . In pumping mode, the pump generates the required cracking pressure

∆𝑝𝑐𝑟 to guarantee proper configuration of POCVs. However, in the motoring mode, the external

load works to create this cracking pressure.

30
Figure 3-8 Construction of critical zones 5 and 6, taking into account the effect of transmission
line losses, Coulomb and viscous friction forces and valves cracking pressure.

To study the effect of the friction force components on the shape of the critical zones, the

actuator equation of motion is rearranged. Assuming low acceleration and ignoring the inertial

terms in Eq. (3-1), the external load, 𝐹𝐿 , in critical condition is defined as follows:

𝐹𝐿 = 𝐹𝑐𝑟 − 𝐹𝐶 𝑠𝑔𝑛(𝑣𝑎 ) − 𝑓𝑣 𝑣𝑎 (3-15)

The above equation shows that the friction force affects the shape of critical zone shape

differently in the upper and lower sections of the 𝐹𝐿 -𝑣𝑎 plane. In the upper portion of Figure 3-8,

at positive velocity, Coulomb friction component shifts the critical zones to the left-hand side,

while the viscous friction component bends this zone to the left with a rate related to the viscous

friction coefficient. These effects are reversed at negative velocities in the lower portion of graph.

Built upon the above analysis, Figure 3-8 shows the different limits describing the undesirable

performance regions. Regions 1, 2, 3 and 4 represent the good performance areas while the

performance deterioration occurs in regions 5 and 6. The mathematical representation of the

different limit lines can be shown as follows:

31
 𝐹𝐿1 = 𝐹𝑐𝑟 − 𝐹𝑓 (3-16)

𝐹𝐿2 = 𝐹𝑐𝑟 − 𝐹𝑓 − 𝐹𝐶𝑉𝐴 (3-17)

𝐹𝐿3 = 𝐹𝑐𝑟 + 𝐹𝑓 (3-18)

𝐹𝐿4 = 𝐹𝑐𝑟 + 𝐹𝑓 + 𝐹𝐶𝑉𝐵 (3-19)

where 𝐹𝐿10 = 𝐹𝑐𝑟0 − 𝐹𝐶 , 𝐹𝐿20 = 𝐹𝑐𝑟0 − 𝐹𝐶 − 𝐹𝐶𝑉𝐴 , 𝐹𝐿30 = 𝐹𝑐𝑟0 + 𝐹𝐶 , and 𝐹𝐿40 = 𝐹𝑐𝑟0 + 𝐹𝐶 +

𝐹𝐶𝑉𝐵 at zero velocity condition.

With reference to Figure 3-8, region 5 represents the area in which pump mode of operation

switches from motoring to pumping during actuator extension. Pressures at both sides of the circuit

are almost equal and less than the charge pressure which keeps both POCVs open. In this case, the

charge pump supplies both sides of the circuit with hydraulic flow and the actuator velocity is not

fully controllable. However, region 6 represents the area where pump mode of operation switches

from motoring to pumping during the actuator retraction. Pressures at both sides of the circuit are

almost equal and higher than the charge pressure and both valves, initially, are critically closed

[35]. Opening POCVB supports the motoring mode while motion decelerates due to less assistive

load. On the other hand, opening POCVA supports the pumping mode and motion acceleration.

Consequently, the pump mode of operation and POCVs configuration keep switching and the

actuator velocity oscillates. In this situation energy is displayed in potential, kinetic and hydraulic

forms. For instance, when velocity of lowered mass oscillates, kinetic energy oscillates while

potential energy decreases in variable rate. Switching of pump mode of operation from motoring

to pumping and vice versa leads to oscillations in hydraulic energy; hydraulic energy keeps fed in

and out of the circuit. A recent study [29] showed a clear limit cycle between velocity and pressure

difference across the pump in this type of circuits. The rest of this research focuses on studying

the undesirable oscillatory performance at region 6.

32
3.1.5 Simulation studies

Simulation studies are performed to evaluate circuit's performance at different operating

conditions and to identify the shape and position of the critical operating zone. A simulation

program of the circuit with two POCVs is developed in MATLAB. Parameters and values used in

simulations are listed in Tables 3-1 and 3-2. Simulation studies are done for two load-motion

scenarios. In the first scenario, the circuit responses are simulated for a constant load and a step

input control signal; this simulation is repeated at different operating points covering all four

quadrants of operation. The results are classified based on response quality, and, accordingly, the

poor performance zone is located on the FL-va plane. As for the second scenario, the circuit

responses are simulated for variable loading condition and a square input signal. This scenario

emulates a real motion of an excavator link.

Table 3-1 Values of parameters of circuit with pilot-operated check valves shown in Figure 3-6.

Parameter Definition Values

𝐴𝑎 Area of piston cap side 31.67 × 10−4 [m2 ]
𝐴𝑏 Area of piston rod side 23.75 × 10−4 [m2 ]
𝐹𝑝𝑟 Seal pre-loading force 500 [N]
𝐾𝑜𝑖𝑙 Effective bulk modulus 0.689 × 109 [Pa]
m3 V
𝐾𝑝𝑚𝑝 Linearized coefficient for pump flow 133 × 10 [−6
]
s
𝑃𝑐 Charge pressure for circuit 1.38 [MPa]
𝑉𝐴0 Initial oil volume at side A of the circuit 0.008 [m3 ]
𝑉𝐵0 Initial oil volume at side B of the circuit 0.006 [m3 ]
Ns
𝑓𝑐𝑓𝑟 Coulomb friction coefficient 0.0001 [ ]
Pa
Ns
𝑓𝑣 Viscous friction coefficient 10000 [ ]
m
𝛼 Actuator area ratio 0.75
𝑀𝑒𝑞 Equivalent mass 400 [kg]

33
Table 3-2 Values of parameters of pilot-operated check valves shown in Figure 3-7.

Parameter Definition Values

𝐴𝑙𝑘 Valve leakage area 0 [m2 ]
𝐴𝑚𝑎𝑥 Valve max flow area 25 × 10−6 [m2 ]
𝑃𝑐𝑟 Valve cracking pressure 0.2 [MPa]
𝑃𝑜𝑚𝑥 Valve maximum opening pressure 0.5 [MPa]

3.1.5.1 Constant loading simulations

Different simulation runs to evaluate circuit performance at different loads are performed.

At each simulation run, a voltage step input signal is applied to the system at a specific load and

the actuator velocity is recorded versus time. If velocity response possesses oscillations that

continue for over 0.6 s, such load-velocity operating point is identified as a critical operating point.

Critical zone is the area containing all the critical operating points. Note that settling time that

distinguishes desirable and undesirable responses is chosen as 0.6 s for demonstrative purposes.

However, different applications could have different undesirable performance criteria.

Figure 3-9 (a) shows the construction of the critical zone for the data listed in Tables 3-1

and 3-2. Circuit performance during critical zone is explained by comparing responses of five

selected operating points that are located before, within and after critical zone. All selected test

points have the same step signal input of -4 V but different load values. Figures 3-9 (b) through (f)

show the actuator velocity responses versus time at test points TP1 through TP5 for loads of 4500

N, 3150 N, 2750 N, 2500 N and 1500 N, respectively. Figures 3-9 (b) and (f) show the actuator

velocity responses at TP1 and TP5 located in motoring mode (Q4) and pumping mode (Q3),

respectively. It is clear from Figures that the system response is stable at both operating points

with settling time of about 0.25s. Figures 3-9 (c) and (e) show the actuator velocity responses at

TP2 and TP4 which are located at borders of the critical zone.

34
 0

Input signal = -4V

Velocity (cm/s) -12

TP4 TP2 TP1

TP5 TP3

-24 Input signal = -8V

-36
0 1 2 3 4 5 6
Load (kN)
(a)
0 0
TP2

Velocity (cm/s)
Velocity (cm/s)

TP1
-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(b) (c)
0 0
Velocity (cm/s)

Velocity (cm/s)

TP3 TP4
-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(d) (e)
0
Velocity (cm/s)

TP5
-6

-12

-18
0 0.5 1 1.5 2
Time (s)
(f)
Figure 3-9 Simulation results of circuit with two pilot-operated check valves given data in Tables
3-1 and 3-2: (a) construction of critical zone; (b)-(f) actuator velocity responses at test points (TP1-
TP5) given step control signal input of -4 V and assistive loads of 4500 N, 3150 N, 2750 N, 2500
N and 1500 N, respectively.

35
 The above Figures show that the system response possesses damped oscillations where the

velocity settling time is about 0.6s. Figure 3-9 (d) illustrates the actuator velocity response at TP3

that is located within the critical zone. The circuit response is oscillatory at this point where the

velocity mean value, oscillation amplitude and frequency are 9.8 cm/s, 3.5 cm/s and 7 Hz,

respectively.

3.1.5.2 Variable loading simulations

Circuit responses are simulated for variable loading condition and a square input signal; this

scenario emulates a real motion of an excavator link, which is referred to in Figure 3-5. Note that

during one complete operating cycle of the actuator, the motion of the mass at the end of the stick

generates different resistive and assistive loads that cover all four quadrants of operation. Load

pattern is approximated to be triangular-shaped in this simulation. The circuit performance is

simulated at both low and high loading conditions. Simulations at low loading conditions are

designed mainly to investigate stability issues and high loading simulations examine circuits

performance and efficiency.

Figures 3-10 and 3-11 show the simulation results of circuit performance at low and high

loading conditions, respectively. Figures 3-10 (a) and 3-11 (a) show the square input control

signals of values of ± 3 V and ± 4 V at low and high loading conditions. Figures 3-10 (b) and 3-

11 (b) illustrate the applied load patterns with maximum values of ± 3 kN and ± 15 kN,

respectively. The actuator velocities versus time are shown in Figures 3-10 (c) and 3-11 (c).

Operational quadrants and switching zones are shown in Figure 3-11 (c) for clarification. Figure

3-11 (c) shows that actuator velocity oscillates during switching from assistive to resistive modes

during actuator retraction (S43) at low loading conditions. However, velocity is oscillation-free at

high loading conditions as can be seen in Figure 3-11(c). Note that the position of velocity

36
 oscillation observed during switching zone S43 at low loading condition in Figure 3-11 (c) matches

that of the constructed critical zone shown in Figure 3-9 (a). Figures 3-10 (d) and 3-11 (d) illustrate

pressures at both sides of pump versus time. Similar to velocity responses, pressure oscillations

occurred during switching zone S43 at low loading condition.

4 4

2
Input signal (V)

Load (kN)
0 0

-2 -2

-4 -4
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(a) (b)
10 4
Q2 S32
Q1 Q1
Pressure (MPa)
Velocity (cm/s)

5
3
S32 S21 S32 S14 S43
0 S21
S14 . . .
S14
Q3 Q3
2
-5
Q4 S43 Q4 S43
-10 1
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(c) (d)
10 4 1.75
Q2 S21 S32
Q1
Velocity (cm/s)

5
pB (MPa)

3 1.25
S32 Q2 1.35 1.55
0 Q3
S14
-5 2
S43 Q4 S14
Q3 S43 Q4 S21
-10 Q1
1
-4 -2 0 2 4
1 1.5 2 2.5
Load (kN)
pA (MPa)
(e) (f)

Figure 3-10 Simulation results for performance of the circuit with two pilot-operated check valves
at low loading conditions given data in Tables 3-1 and 3-2: (a) input control signal; (b) variable
load of ± 3 kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pressure plane.

37
 Input signal (V) 6 20

3 10

Load (kN)
0 0

-3 -10

-6 -20
0 6 12 18 24 0 6 12 18 24
Time (s) Time (s)

(a) (b)
12 12
Velocity (cm/s)

Pressure (MPa)
6
8
0
4
-6

-12 0
0 6 12 18 24 0 6 12 18 24
Time (s) Time (s)
(c) (d)
12 12

6
Velocity (cm/s)

8
pB (MPa)

0
4
-6

-12 0
-20 -10 0 10 20 0 2 4 6 8
Load (kN) pA (MPa)

(e) (f)

Figure 3-11 Simulation results for performance of the circuit with two pilot-operated check valves
at high loading conditions given data in Tables 3-1 and 3-2: (a) input control signal; (b) variable
load of ± 15 kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pressure plane.

Figures 3-10 (e) and 3-11 (e) show circuit performance on the load-velocity FL-va plane. The

relation between pressures at both sides of the circuit at low and high loading conditions are shown

on the pump pressure plane (pa-pb) in Figures 3-10 (f) and 3-11 (f), respectively. It is clear from

the Figures that pressures at both sides of the circuit are almost equal and less than the charge

38
pressure at switching zone S21. The relative suction status at both sides of the circuit at S21 opens

both POCVs that in turn connect both sides of the circuit to the charge pressure. Limited pressure

drop at both sides of the circuit is noticed. On the other hand, pressures at both sides of the circuit

are almost equal and higher than the charge pressure at switching zone S43. In such status, pressure

oscillations at both sides of the circuit are noticed at low loading conditions as shown in Figure

3-10 (f). That indicates that both POCVs switch between open and closed status which supports

the theoretical discussion outlined in Section 3.1.5.

3.2 Circuit with closed-center shuttle valve

Figure 3-12 illustrates a simplified drawing of the circuit that utilizes 3/3 closed-center shuttle

valve (CC-SHV) in the first quadrant of operation.

F fr xa , va

FL
pA
pB
QA QB

p1 p2
Q1 Q2
2 4

3/3 Shuttle valve

Line A Line B
Qc Pc

CH

Qa Qb
pa pb

M
M

Figure 3-12 Circuit with 3/3 closed-center shuttle valve in the first quadrant of operation.

39
 The circuit uses a shuttle valve to compensate for the differential flow of the single-rod cylinder.

In this section, the mathematical model of the circuit is developed and the performance is

simulated. Low performance zone is identified and compared to the previously identified zone.

3.2.1 Previous identification of poor performance zone

Wang et al. [11] identified the critical operating conditions in circuits that utilize 3/3 CC-SHV

as the circuit operates around the critical load 𝐹𝑐𝑟 . The critical load is identified as the actuating

force when the pressure at both chambers of the actuator is equal to the charge pressure, i.e. 𝐹𝑐𝑟 =

𝐴
𝐴𝐴 𝑃𝑐 (1 − 𝛼) where the actuator area ratio 𝛼 = 𝐴𝐵 . Figure 3-13 shows an illustration of the critical
𝐴

zone in circuits utilizing closed-center shuttle valve according to [11].

𝑣𝑎

II. Motoring

I. Pumping

𝐹𝑐𝑟

0 FL

Critical zone
IV. Motoring

III. Pumping

Figure 3-13 Illustration of critical zone (hatched area) in circuit that utilizes closed-center shuttle
valve according to [11].

40
3.2.2 Modeling

The mathematical model of circuit with CC-SHV is similar to that of circuit with POCVs except

for the part concerning the compensating valve. Thus, the mathematical model of circuit with CC-

SHV is obtained by replacing governing equations of POCVs with that of the CC-SHV.

Figure 3-14 (a) and (b) shows the schematic drawing of a 3/3 closed-center shuttle valve and,

opening areas as function of the effective pressure acting on the spool, respectively. The flow

through the valve openings is represented by the following equations [15]:

2
𝑄1 = 𝐶𝑑 𝐴𝑠𝑣𝐴 √ |𝑝1 − 𝑝𝑐 | 𝑠𝑖𝑔𝑛 (𝑝1 − 𝑝𝑐 ) (3-20)
𝜌

2
𝑄2 = 𝐶𝑑 𝐴𝑠𝑣𝐵 √ |𝑃𝑐 − 𝑃2 | 𝑠𝑖𝑔𝑛 (𝑝𝑐 − 𝑝2 ) (3-21)
𝜌

where 𝐴𝑠𝑣𝐴 and 𝐴𝑠𝑣𝐵 represent opening areas of valve with transmission lines, Line A and Line

B, respectively (see Figure 3-12). The cracking condition of the shuttle valve depends on the

pressure difference between the two circuit lines, spool area and centering spring stiffness.

The SHV dynamics is neglected because of its low effect on the overall circuit dynamics.

Instead, the static force balance of the spool is employed. Assuming operation in the first quadrant

where 𝑝1 > 𝑝2 , static force balance of the spool at critical condition is represented as follows:

𝐴𝑝 (𝑝1 − 𝑝2 ) + 𝑘𝑠 (𝑥𝑣0 + 𝑥𝑣 ) = 0 (3-22)

where 𝐴𝑝 is the spool effective area; 𝑘𝑠 and 𝑥𝑣0 represent spring stiffness and initial compression

length, respectively. Spool displacement is denoted as 𝑥𝑣 .

By dividing the above equation by spool area, 𝐴𝑃 , and assuming operation at the critical opening

condition, i.e. 𝑥𝑣 ≈ 0, the following pressure balance equation is obtained:

41
 𝑝1 − 𝑝2 = 𝑝𝑐𝑟 (3-23)

For simplicity, effective opening pressure, 𝑝𝑒 , is introduced as the summation of the effective

opining pressures of SHV, i.e. 𝑝𝑒 = 𝑝1 − 𝑝2, Assuming that the valve opening area is linearly

proportional to the effective opening pressure, 𝑝𝑒 , as shown in Figure 3-14 (b). Then, the opening

area that connects valve to Line B can be represented as follows:

𝐴𝑙𝑘 𝑝𝑒 ≤ 𝑝𝑐𝑟
𝐴𝑠𝑣𝐵 = { 𝑘1 (𝑝𝑒 − 𝑝𝑐𝑟 ) 𝑝𝑐𝑟 < 𝑝𝑒 < 𝑝𝑜𝑚𝑥 𝐵 (3-24)
𝐴𝑚𝑎𝑥𝐵 𝑝𝑒 ≥ 𝑝𝑜𝑚𝑥𝐵

where 𝐴𝑙𝑘 represents valve leakage area; 𝑘1 is the valve opening area coefficient; 𝑝𝑜𝑚𝑥𝐵 and

𝐴𝑚𝑎𝑥𝐵 represent the pressure required to fully open the valve opening and maximum opening area,

respectively.

Line A, p1 Line B, p2 𝐴𝑠𝑣

pilot Line A pilot Line B
𝐴𝑚𝑎𝑥
𝐴𝑠𝑣𝐴 𝐴𝑠𝑣𝐵
AP

kS
𝐴𝑙𝑘
xv Charge line, pc 𝑝𝑜𝑚𝑥
−𝑝𝑐𝑟 𝑝𝑐𝑟 𝑝𝑒

(a) (b)
Figure 3-14 3/3 closed-center shuttle valve: (a) schematic drawing; (b) opening area versus
effective pressure.

3.2.3 Simulation studies

A simulation program of the circuit with CC-SHV has been developed in MATLAB

environment to evaluate the circuit performance and to identify the shape and position of the

critical operating zone. As mentioned in Section 3.1.6, simulation studies are done for two load-

motion scenarios. The first scenario is related to the constant load and the step input control signal.

42
The second scenario is about the variable loading condition and the square input signal. The

simulation parameters for the circuit are listed in Table 3-1 while the parameters for CC-SHV are

listed in Table 3-3.

Table 3-3 Values of parameters of closed-center shuttle valve shown in Figure 3-14.

Parameter Definition Values

𝐴𝑙𝑘 Valve leakage area 0 [m2 ]
𝐴𝑚𝑎𝑥 Valve max flow area 25 × 10−6 [m2 ]
𝑃𝑐𝑟 Valve cracking pressure 0.2 [MPa]
𝑃𝑜𝑚𝑥 Valve maximum opening pressure 0.5 [MPa]

3.2.3.1 Constant loading simulations

Similar to that simulation of the circuit with two POCVs in Section 3.1.5, different runs to

evaluate the circuit performance circuit with CC-SHV at different operating conditions are

performed. At each simulation run, a voltage step input signal is applied to the system at a specific

load; the actuator velocity is recorded versus time. The operating point is considered critical if its

velocity response oscillations continue for over 0.6 s. The critical zone is constructed of all critical

points. The circuit performance in switching zone is explained by tracking responses of 5 selected

operating points, as can be seen in Figure 3-15. Figure 3-15 (a) shows the construction of the

critical zone and position of the five tested points (TP1- TP5). All operating points have the same

step input signal of -4 V and different loading conditions. Figures 3-15 (b) through (f) show the

actuator velocity responses versus time at TP1 through TP5 for loads of 4500N, 3860N, 3200N,

2200N and 1400N, respectively. Figures 3-15 (b) and (f) show the actuator velocity responses at

TP1 and TP5 which are located in areas where pump runs in pure motoring and pumping modes,

respectively. It is clear from both Figures that the system responses are stable at both operating

points with settling time of about 0.25s.

43
 0

Velocity (cm/s) -12

TP2 TP1
TP3
TP5 TP4
-24

-36
0 1 2 3 4 5 6
Load (kN)
(a)
0 0
TP1 TP2

Velocity (cm/s)
Velocity (cm/s)

-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)

(b) (c)
0 0
TP3 TP4
Velocity (cm/s)
Velocity (cm/s)

-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)

(d) (e)
0
TP5
Velocity (cm/s)

-6

-12

-18
0 0.5 1 1.5 2
Time (s)

(f)
Figure 3-15 Simulation results of the circuit with closed-center shuttle valve, given data in Tables
3-1 and 3-3: (a) construction of critical zone; (b)-(f) actuator velocity response at test points TP1-
TP5 for step input signal of -4 V and assistive loads of 4500N, 3860N, 3200N, 2200N and 1400N,
respectively.

44
 Figures 3-15 (c) and (e) show the actuator velocity responses at TP2 and TP4 that are located

at borders of the critical zone. It can be seen from both Figures that the system response possesses

damped oscillations where the velocity performed 3 cycles before settling in about 0.6s.

Figure 3-15 (d) illustrates the actuator velocity response at TP3 that is located within the critical

zone. The circuit response at this point is oscillatory, where mean value, oscillation amplitude and

frequency of velocity response are 10.4 cm/s, 6 cm/s and 8 Hz, respectively.

3.2.3.2 Variable loading simulations

Simulation results of the circuit performance at low and high variable loading conditions are

shown in Figures 3-16 and 3-17, respectively. Figures 3-16 (a) and 3-17 (a) illustrate the square

input signals of values of ± 3 V and ± 4 V at low and high loading conditions. The applied load

patterns are shown in Figures 3-16 (b) and 3-17 (b), respectively. Figures 3-16 (c) and 3-17 (c)

show the actuator velocity versus time at low and high loading conditions. It is clear from the

velocity responses that velocity oscillations occur during switching zone S43 at both low and high

loading conditions. However, oscillations are more sever at low loading conditions. Figures 3-16

(d) and 3-17 (d) illustrate the pressures at both sides of the pump versus time, where pressure

oscillations are noticed during S43 at low and high loading conditions as well.

The velocity oscillations observed during S43 are shown also on the FL-va plane in Figures 3-

16 (e) and 3-17 (e). The relationship between pressures at both sides of the circuit at low and high

loading conditions are shown on the pa-pb plane in Figures 3-16 (f) and 3-17 (f), respectively.

Similar to the circuit with two POCVs, the pressures at both sides of the circuit are almost equal

and less than the charge pressure at switching zone S21. The relative suction status at both sides

of the circuit at S21 activates the CC-SHV that connects either sides of the circuit to the charge

pressure. However, an excessive pressure drop at both sides of the circuit is noticed.

45
 Contrarily, the pressures at both sides of the circuit are almost equal and higher than the charge

pressure at switching zone S43. Excessive pressure oscillations are noticed at low loading

conditions in Figure 3-16 (f) as compared to that of high loading condition in Figure 3-17 (f).

4 4

2 2
Input signal (V)

Load (kN)
0 0

-2 -2

-4 -4
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)

(a) (b)
8 4
Velocity (cm/s)

3
Pressure (MPa)

0
2
-8
1

-16 0
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)

(c) (d)
8 4

3
Velocity (cm/s)

0
pB (MPa)

2
-8
1

-16 0
-4 -2 0 2 4 0 1 2 3
Load (kN) pA (MPa)

(e) (f)

Figure 3-16 Simulation results for performance of the circuit with closed-center shuttle valve at
low loading conditions given data in Tables 3-1 and 3-3: (a) input signal; (b) variable load of ± 3
kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane.

46
 6 16

3
lnput signal (V)

Load (kN)
0 0

-3 -8

-6 -16
0 5 10 15 20 25 0 5 10 15 20 25
Time (s) Time (s)

(a) (b)
10
16
8

Pressure (MPa)
8
Velocity (cm/s)

6
0
-8 4

-16 2

-24 0
0 5 10 15 20 25 0 5 10 15 20 25
Time (s) Time (s)

(c) (d)
16 10
8 8
Velocity (cm/s)

0
pb (MPa)

6
-8 4
-16 2
-24 0
-16 -8 0 8 16 0 2 4 6 8
Load (kN) pa (MPa)

(e) (f)

Figure 3-17 Simulation results for performance of the circuit with closed-center shuttle valve at
high loading conditions given data in Tables 3-1 and 3-3: (a) input control signal; (b) variable load
of ± 15 kN; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pressure plane.

3.3 Circuit with open-center shuttle valve

Figure 3-18 shows the circuit that utilizes an open-center shuttle valve (OC-SHV), introduced

by Caliskan et al. [13]. This circuit uses the pre-designed leakage in the valve at center position

47
to dampen the system oscillations. However, Authors reported that their solution works best under

certain actuator velocities. Additionally, their experimental work is limited to low loading

conditions and lacks the effect of load mass inertia.

F fr xa , va

FL
pA pB
QA QB

p1 p2
Q1 Q2
2 4

3
Open-center
Line A shuttle valve Line B
Qc Pc

CH

Qa Qb

pa pb

M
M

Figure 3-18 Circuit that utilizes an open-center shuttle valve.

3.3.1 Previous identifications of poor performance zone

Caliskan et al. [13] specified two load limits (𝐹𝐿1 and 𝐹𝐿2 ) in the 𝐹𝐿 -𝑣𝑎 plane as the boundaries

of the critical zone for the circuit that utilizes OC-SHV, as shown in Figure 3-19. 𝐹𝐿1 and 𝐹𝐿2 are

the loads when the shuttle valve becomes fully open from the center position in both directions of

motion. The values of these limits depend on the shuttle valve operational pressures and the

actuator effective areas. Caliskan et al. [13] reported that their solution works best under a critical

48
velocity (vcr) shown in Figure 3-19. The value of such critical velocity depends mainly on the

maximum predesigned leakage area in the valve. However, higher leakage area reduces

controllability of the actuator velocity.

va

II. Motoring

I. Pumping

Zone where oscillations

are alleviated
𝐹𝑐𝑟

F L1 F L2 FL

Zone where oscillations

vcr are reduced
IV. Motoring

III. Pumping

Figure 3-19 Critical zone in the circuit that utilizes open-center shuttle valve showing leakage
solution effectiveness according to Caliskan et al. [13].

3.3.2 Modeling

Mathematical model of the circuit that utilizes OC-SHV is similar to that of the circuit with

CC-SHV except for the part concerning valve leakage. Therefore, the mathematical model of this

circuit is obtained by replacing the opening areas equations in CC-SHV model by their

counterparts in the OC-SHV. Figures 3-20 (a) and (b) show the schematic drawing of a 3/3 OC-

SHV and opening areas as function of the effective pressure acting on the spool, respectively.

The flow equations through the valve openings are similar to that of the CC-SHV described in

Eqs. (3.20) and (3.21). The main difference between the OC-SHV and CC-SHV is the value of

predesigned leakage area (𝐴𝐿𝑘𝑑 ) at the center position of the valve. 𝐴𝑠𝑣𝐴 and 𝐴𝑠𝑣𝐵 represent the

49
opening areas of the valve with Lines A and B, respectively. Areas are linearly proportional to the

effective opening pressure (𝑝𝑒 = 𝑝1 − 𝑝2), as shown in Figure 3-20 (b). Then, the opening area

that connects the valve with Line B of circuit is represented as follows:

𝐴𝑠𝑣𝑚𝑥 𝑝𝑒 ≥ 𝑝𝑜𝑚𝑥
𝐴𝑙𝑘𝑑 + 𝑘1 |𝑝𝑒 − 𝑝𝑐𝑟 | 𝑝𝑜𝑚𝑥 ≥ 𝑝𝑒 > 𝑝𝑐𝑟
𝐴𝑠𝑣𝐵 = 𝐴𝑙𝑘𝑑 𝑝𝑐𝑟 ≥ 𝑝𝑒 > −𝑝𝑐𝑟 (3-25)
𝐴𝑙𝑘𝑑 − 𝑘1 |𝑝𝑒 − 𝑝𝑐𝑟1 | −𝑝𝑐𝑟 ≥ 𝑝𝑒 > −𝑝𝑐𝑟1
{ 𝐴𝑙𝑘 −𝑝𝑐𝑟1 ≥ 𝑝𝑒

where 𝑝𝑜𝑚𝑥 and 𝐴𝑠𝑣𝑚𝑥 represent the pressure required to fully open the valve and the maximum

opening area, respectively. 𝐴𝑙𝑘𝑑 and 𝐴𝑙𝑘 are the designed and clearances leakage areas,

respectively. 𝑘1 is the valve opening area coefficient; the spool cracking pressure and pressure

critically needed to close either of the leakage openings which are denoted by 𝑝𝑐𝑟 and 𝑝𝑐𝑟1,

respectively.

Line A, p1 Line B, p2 𝐴𝑠𝑣

Pilot signal Pilot signal
𝐴𝑠𝑣𝑚𝑥
𝐴𝑠𝑣𝐴 𝐴𝑠𝑣𝐵
Q1 Q2 AP

kS 𝐴𝑙𝑘𝑑
𝐴𝑙𝑘
xv Charge line, pc 𝑝𝑐𝑟 𝑝𝑐𝑟1 𝑝𝑜𝑚𝑥 𝑝𝑒

(a) (b)
Figure 3-20 3/3 open-center shuttle valve: (a) schematic drawing; (b) openings areas versus
effective pressure.

3.3.3 Simulation studies

A simulation program is developed to evaluate the performance of the circuit with OC-SHV.

Similar to the previously-discussed circuits, simulation studies are done for constant and variable

load-motion scenarios. The simulation parameters for the circuit are listed in Table 3-1, whereas

50
the parameters for the OC-SHV are listed in Table 3-4. The initial leakage area of spool, 𝐴𝐿𝑘𝑑 , is

chosen to be comparable to that of the valve used by previous researchers [13].

Table 3-4 Values of parameters of open-center shuttle valve in circuit in Figure 3-18.

Parameter Definition Values

𝐴𝑙𝑘𝑑 Valve initial leakage area 2 × 10−6 [m2 ]
𝐴𝑚𝑎𝑥 Valve max flow area 25 × 10−6 [m2 ]
𝑃𝑐𝑟 Valve cracking pressure 0.2 [MPa]
𝑃𝑐𝑟1 Critical pressure needed to close either 0.3 [MPa]
of the leakage openings
𝑃𝑜𝑚𝑥 Valve maximum opening pressure 0.5 [MPa]

3.3.3.1 Constant loading simulations

Similar to previous simulated circuits with POCVs and CC-SHV, different runs at a constant

load are done for circuit with OC-SHV. Construction of critical zone is illustrated in Figure 3-21

(a). It is seen from Figure 3-21 (a) that undesirable performance area of the circuit with OC-SHV

is 33% smaller than that of the circuit with CC-SHV. Actuator velocity responses versus time at

test points TB1-TB5 for step signal input of -4 V are shown in Figures 3-21 (b)-(f). Figures 3-21

(b), (d) and (f) show damped velocity responses of the circuit at TP1, TP3 and TP5, respectively.

Figures 3-21 (c) and (e) show the actuator velocity responses at TP2 and TP4 located within the

critical zone. The circuit response at both points is oscillatory, where the mean value, oscillation

amplitude and frequency of velocity response are -9 cm/s, 4.5 cm/s and 9 Hz, respectively.

Figure 3-21 shows that the predesigned leakage in the circuit with OC-SHV reduced the circuit

oscillations compared to the circuit with CC-SHV in terms of the undesirable performance zone

size and oscillations severity. The initial leakage area of the valve, 𝐴𝑙𝑘𝑑 , is the main factor affecting

the system damping behavior.

51
 0

Velocity (cm/s) -12

TP1
TP2
TP5 TP3
TP4

-24

-36
0 1 2 3 4 5 6
Load (kN)
(a)

TP2
0 0
Velocity (cm/s)

TP2

Velocity (cm/s)
TP1
-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)

(b) (c)
0 0
TP3 TP4
Velocity (cm/s)
Velocity (cm/s)

-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(d) (e)
0
Velocity (cm/s)

TP5
-6

-12

-18
0 0.5 1 1.5 2
Time (s)
(f)
Figure 3-21 Simulation results of the circuit with an open-center shuttle valve, given data in Tables
3-1 and 3-4: (a) construction of critical region; (b)-(f) actuator velocity response at test points TP1-
5 for step signal input of -4 V and loads of 4550 N, 3400 N, 2960 N, 2370 N and 1650 N.

52
 To study the effect of increasing 𝐴𝑙𝑘𝑑 on the undesirable performance area, simulations are

done for 𝐴𝑙𝑘𝑑 that is double of the previous run. Simulation results of the higher initial leakage

area are shown in Figure 3-22. Figure 3-22 (a) shoes the construction of critical zone. Figures 3-

22 (b) through (f) show that velocity responses at different positions of the critical zone.

Comparing Figures 3-21 (a) and 3-22 (a) show that the critical area ratio is approximately 5: 4 for

the circuits with 2 mm2 and 4 mm2 initial leakage areas, respectively. A comparison between

Figures 3-21 (d) and 3-22 (d) shows that the actuator velocity oscillation amplitude is smaller in

the circuit with 4 mm2 initial leakage area compared to that of 2 mm2. However, the circuit with

higher initial leakage area encounters a non-gradual velocity gradient in the critical switching zone

S43. Figures 3-22 (b) through (f) show that velocity goes up from -8.4 cm/s at TP1 in the motoring

mode to -9 cm/s at TP2 and continues to rise up to -10.2 before TP3. However, velocity steadily

dropped down to -9 cm/s after TP3 before it continues increasing to -10.5at TP4, and finally to -

11.2 cm/s at TP5. That shows that as leakage increases, the actuator velocity switches from the

low to the high velocity modes in an unpredictable manner.

3.3.3.2 Variable loading simulations

Similar to the previous circuits, the circuit with OC-SHV is simulated for variable loading

conditions and a square input signal. Figures 3-23 and 3-24 show the simulation results of the

circuit performance at low and high loading conditions, respectively. Figures 3-23 (a) and 3-24 (a)

show the input control signals. Figures 3-23 (b) and 3-24 (b) show the applied loads.

Figure 3-23 (c) shows that the actuator velocity is oscillation-free at low loading conditions at

switching zone S43, while a small velocity drop is encountered. On the other hand, minor velocity

oscillations occurred at high loading conditions as shown in Figure 3-24 (c).

53
 0

Velocity (cm/s)
-12
TP1
TP2
TP4
TP5 TP3
-24

-36
0 1 2 3 4 5 6
Load (kN)
(a)
0 0
Velocity (cm/s)

Velocity (cm/s)
TP1 TP2
-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(b) (c)
0 0
Velocity (cm/s)
Velocity (cm/s)

TP3 TP4
-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(d) (e)
0
Velocity (cm/s)

TP5
-6

-12

-18
0 0.5 1 1.5 2
Time (s)
(f)
Figure 3-22, Simulation results of the circuit with an open-center shuttle valve with higher pre-
designed leakage, given data in Tables 3-1 and 3-4: (a) construction of critical region; (b)-(f)
actuator velocity response at test points TP1-5 for step signal input of -4 V and loads of 4380 N,
3580 N, 2800 N, 2200 N and 1350 N.

54
 Figures 3-23 (d) and 3-24 (d) illustrate the pressures at both sides of the pump versus time.

Similar to the velocity responses, the pressure responses are oscillation-free at all margins of

operation. However, a small pressure drop is noticed at both sides of the circuit at S21.

4 4

2
Input signal (V)

Load (kN)
0 0

-2 -2

-4 -4
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(a) (b)
8 4
Velocity (cm/s)

4
Pressure (MPa)
3
0
2
-4

-8 1
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(c) (d)
8 4

4
Velocity (cm/s)

3
pB (MPa)

0
2
-4

-8 1
-4 -2 0 2 4 1 2 3
Load (kN) pA (MPa)

(e) (f)

Figure 3-23 Simulation results of the circuit with open-center shuttle valve at low loading
conditions, according to the given data in tables 3-1 and 3-4: (a) input control signal; (b) applied
load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane.

55
 Figures 3-23 (e) and 3-24 (e) show the circuit performance on the FL-va plane where few

velocity oscillations are noticed in S43 at high loading conditions. The relation between pressures

at both sides of the circuit on the pa–pb plane are shown in Figures 3-23 (f) and 3-24 (f) at low and

high loading conditions, respectively.

6 16

3 8
Input signal (V)

Load (kN)
0 0

-3 -8

-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)

(a) (b)
16 10
8
Pressure (MPa)

8
Velocity (cm/s)

6
0
4
-8 2
-16 0
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)

(c) (d)
16 10
Velocity (cm/s)

8 8
pB (MPa)

6
0
4
-8
2
-16 0
-16 -8 0 8 16 0 2 4 6 8
Load (kN) pA (MPa)

(e) (f)

Figure 3-24 Simulation results of the circuit with open-center shuttle valve at high loading
conditions, according to the given data in tables 3-1 and 3-4: (a) input control signal; (b) applied
load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane.

56
3.4 Comparison between the selected circuits

In this section, the position and size of the critical zone as well as the characteristics of the

actuator velocity oscillation are compared in the aforementioned simulated circuits. Figures 3-25

(a) and (b) show comparisons of the construction of the critical zones of the circuits with POCVs

to that with CC-SHV and circuits with CC-HSV to that with OC-SHV, respectively. It is seen from

the load-velocity planes in Figure 3-25 that circuit with CC-SHV possesses a larger oscillatory

zone compared to the circuits with POCVs and OC-SHV. The ratios between oscillatory zone areas

are found to be 1: 2.15: 1.45 in the circuits with POCVs, CC-SHV and OC-SHV, respectively.

0 0
Velocity (cm/s)

Velocity (cm/s)

-12 -12

Circuit with
-24 -24 OC-SHV
Circuit with
CC-SHV Circuit with Circuit with
POCVs CC-SHV
-36 -36
0 1 2 3 4 5 6 0 2 4 6
Load (kN) Load (kN)
(a) (b)
Figure 3-25 Critical zones for circuits that utilize; (a) POCVs (solid line) and CC-SHV (dashed
line); (b) CC-SHV (dashed line) and OC-SHV (solid line) for comparable valve parameters.

According to Figures 3-9 (d), 3-15 (d) and 3-21 (c), the velocity oscillation characteristics of

all circuits are deduced and listed in Table 3-5. It is clear that the circuit with CC-SHV possesses

higher pressure ripples when compared to circuits with POCVs and OC-SHV. This can be

attributed to the coupling nature of the SHV that simultaneously creates a sudden dynamic change

at both sides of the circuit. However, the circuit with an OC-SHV possesses less velocity

oscillation amplitudes compared to the circuits with POCVs and CC-SHV, which is attributed to

the damping effect of the leakage. According to Figures 3-9 (a), 3-15 (a) and 3-21 (a), it is noticed

that the velocity gradient during switching zones S43 is smoother in the circuit that utilizes POCVs

57
compared to that of the circuits with SHVs. This feature provides more convenience for operator

and facilitates controller application. From above discussions it’s clear that circuits with POCVs

are more appropriate for pump-controlled applications.

Table 3-5 Performance index of the simulated circuits

Performance Index at Critical Circuit

Zone (S43) POCVs CC-SHV OC-SHV
Mean velocity (cm/s) -9.8 -10.4 -10.4
Oscillation amplitude (cm/s) 3.5 5.9 3
Oscillation frequency (Hz) 6 8 8
Critical zone size (ratio) 1 2.25 1.45

58
 CHAPTER 4

4. NEW SOLUTIONS

In this chapter, the requirements for desirable operation of pump-controlled circuits are

investigated. Three concepts to meet the requirements of improving the performance are proposed.

More specifically, these concepts reduce or alleviate the oscillations commonly seen in velocity

responses of the common pump-controlled circuits of single-rod cylinders. These concepts are: (1)

shifting of the critical zone into lower loading ranges, (2) applying controlled leakage and (3),

applying selective throttling. Theoretical analysis and simulation studies that demonstrate the

enhanced performance of the circuits as a result of applying the proposed methods are presented.

By employing these concepts, individually or collectively, eight different new pump-controlled

circuits are proposed and their operations are explained.

4.1 Requirements for desirable operation of pump-controlled single-rod actuators

It has been shown in the previous chapter that currently available pump-controlled circuits for

single-rod actuators possess acceptable performance when the pump operates in one of the four

quadrants of operation. However, these circuits encounter performance issues during operation at

switching zones. Three major phenomena are seen in different combinations at each of the

59
switching zones. Firstly, the pump mode of operation switches from one mode of operation to

another. Secondly, operation of compensating valves that switches charge line connection to both

sides of the circuit. Note that switching of the compensating valves is abrupt and causes a sudden

dynamic perturbation in the circuit. Thirdly, a sudden actuator velocity change happens due to the

occurrence of the first two phenomena. Whereas, switching zones S14 and S32 encounter only the

first circumstance, zones S21 and S43 face all three challenges, refer to Figures 3-3 and 3-4. In

this section, operation of the circuits in the pump pressure plane is discussed and desired

operational parameters for smooth operation in switching zones are deduced. Then, concepts that

assure maintaining proper parameters to alleviate or reduce the effect of the abovementioned

phenomena are proposed.

In circuits that utilize pilot operated check valves (referring to flow pattern in Figure 3-4), two

conditions are required to assure proper operation at resistive extension quadrant (Q1). Firstly, the

pressure at line A of the circuit has to be higher than the pressure at line B by a value 𝑝𝑡ℎ ( 𝑝𝑎 >

𝑝𝑏 + 𝑝𝑡ℎ ) in order to assure proper opening of POCVB and closing of POCVA. Secondly, the

charge line pressure has to be higher than the pressure at line B (𝑝𝑐 > 𝑝𝑏 ) in order to allow

compensating fluid to flow from charge line to line B of the circuit. Requirements are represented

on the pump pressure plane in Figure 4-1. Similarly, the conditions required to assure proper

operation in successive quadrant Q4 are 𝑝𝑎 > 𝑝𝑏 + 𝑝𝑡ℎ and 𝑝𝑏 > 𝑝𝑐 . Switching zone S14 (see

Figure 4-1) is situated between quadrants Q1 and Q4, where the cylinder completes resistive

extension and begins assistive retraction. During this transition, the pump operation changes from

pumping to motoring modes while the configuration of the POCVs stays the same. S14 is expected

to be oscillation-free zone. However, small oscillations in circuit pressures and actuator velocity

are expected if the input control signals contain abrupt changes (for example square input signals).

60
The same argument is applicable to switching zone S32. Generally, undesirable behavior in S14

and S32 can be alleviated by applying smooth input control signals, like sinusoidal signals for

instance.

u
Q3 Q4
Q1 Q4 t t

Figure 4-1 Desired pressure plane of pump-controlled circuits given a variable load configuration
and a square input signal (shown in the inset).

The conditions required to assure proper operation in quadrant Q3 are 𝑝𝑏 > 𝑝𝑎 + 𝑝𝑡ℎ and 𝑝𝑎 >

𝑝𝑐 as shown in Figure 4-1. In switching zone S43, the pump operation switches from motoring

mode in Q4 to pumping mode in Q3. In both quadrants, the extra fluid flows out of the circuit into

the charge system. During switching, the pressures at both main lines of the circuit are nearly equal

and higher than 𝑝𝑐 . Thus, both POCVs become closed and the motion is, momentarily, seized [35].

In order to continue the motion, the pump operates in pumping mode increasing the pressure in

line B, which ultimately opens POCVA, thus, accelerating the cylinder. Cylinder acceleration

accumulates flow that increases the pressure in line A. Accordingly, pump switches to motoring

mode and POCVB opens. Once the pump switches to motoring mode, pressure drops in line B

which leads to reduction in assistive load and consequently deceleration of the motion. As a result,

61
the pump mode of operation and POCVs functions keep switching causing pressure and velocity

to oscillate. In both quadrants Q2 and Q1, the balancing fluid flows from the charge system into

the circuit. Therefore, 𝑝𝑐 > 𝑝𝑎 and 𝑝𝑐 > 𝑝𝑏 in Q2 and Q1, respectively. As a result, pressures at

both sides of the circuit in zone S21 are very similar and smaller than the charge pressure, thus,

both POCVs open. Consequently, the charge system supplies fluid to both sides of the circuit

which reduces the actuator controllability [35]. Thus, the undesirable performance in S21 is not

necessarily due to oscillatory behavior and can be alleviated through proper sizing of POCVs and

installation of anti-cavitation valves in the circuit.

From the above discussion one can conclude that pump-controlled hydraulic circuits with

POCVs are stable and give acceptable responses when pump works, solely, in each of the four

quadrants of operation. At switching zones S14, S32 and S21 the circuit faces some challenges

which can be alleviated by applying smooth input control signals. However, operation in S43

faces oscillatory performance characterized by pump mode of operation and POCVs continuous

switching. Accordingly, actuator velocity experiences sever oscillations. In order to avoid

oscillatory performance during switching zone S43, switching is required to happen over a longer

and more controllable pressure margin. Figure 4-2 illustrates the desired circuit performance in

the load-velocity plane. It is clear from the Figure that switching is required to happen over a force

margin that is longer than the critical margin to allow smooth controllable switching. Moreover, it

provides longer period of time to apply controller corrective actions in case of implementing a

controller in the circuit. Three concepts to alleviate system oscillations and improve the

performance of pump-controlled circuits are proposed. These concepts are: (1) shifting of the

critical zone into lower loading ranges, (II) applying controlled leakage and (III) applying selective

throttling. A detailed discussion of each of these concepts is addressed in the following sections.

62
Figure 4-2 Desired performance of a pump-controlled circuit in the load-velocity plane.

4.2 Concept I- shifting of critical zone

With reference to Figure 4-3 (a), the undesirable zone forms two regions 5 and 6 located to the

left and right-hand sides of a vertical line passing through the initial critical load Fcr0. As explained

in Section 3.1, critical load is the applied external load that is equal to the actuator biased force

when the pressure at both sides of the cylinder is equal to the charge pressure i.e. 𝐹𝑐𝑟0 = 𝑝𝑐 𝐴𝐴 (1 −

α). It is also noticed that the critical zone is asymmetric around the zero-load axes, which results

in shifting the oscillatory critical region 6 to a higher load margin on the load axis. Oscillations at

higher load values, especially due to inertial loads, are more severe in amplitude and, consequently,

destructive on the machine [36]. This shifted position of the critical zone is attributed to creation

of the bias force at the actuator under zero control signal condition. In this section, different

strategies to shift the critical zone into a symmetric position around the zero-load vertical axes (see

Figure 4-3 (b)) are proposed. Simulations of a sample of these designs are carried out to show the

improved performance of the proposed concept.

63
 (a) (b)
Figure 4-3 Construction of critical regions for, (a) conventional pump-controlled circuit as is
explained in Section 3-1 ; (b) circuit that applies shifting of the critical zone concept.

Shifting of the critical zone into a symmetric position around the zero-load axes reduces the

load values seen by the circuit in region 6 and consequently reduces the magnitude and effects of

oscillations. Note that the undesirable region 5 may reach higher load values. However, this can

be tolerated since region 5 does not exhibit oscillatory behavior.

By ignoring the transmission lines losses, the critical load is defined as 𝐹𝑐𝑟0 = 𝐴𝐴 𝑝𝑐 (1 − 𝛼)
𝐴
where 𝛼 = 𝐴𝐵. There are two solutions to obtain zero-value critical load: (i) choosing 𝛼 = 1, which
𝐴

means utilizing a symmetric cylinder and (ii) reducing the charge pressure to zero gauge-pressure,

i.e. 𝑝𝑐 = 0 which means using a non-pressurized oil tank instead of the low-pressure charging

system in the circuit. Apparently, both solutions are not acceptable since this thesis is related to

single-rod asymmetric cylinders. Furthermore, a low-pressure charging system is needed in the

closed pump-controlled circuits in order to allow fast compensation for the differential flow of the

single-rod actuators, avoid pump cavitation and supply low-pressure flow to the pump case for

cooling and lubrication.

One feasible solution is to utilize two different charge pressures in the circuit. In this solution,

each side of the circuit is connected through the compensating valve to a separate charge system.

Pressure values of these charge systems are chosen such that, at zero-load, zero-velocity condition,

64
the pressure-induced forces at each side of the actuator are equal (𝐴𝐴 𝑃𝑐𝐴 = 𝐴𝐵 𝑃𝑐𝐵 ) and

consequently 𝐹𝑐𝑟0 = 0. In order to maintain proper operation of the circuit with two different

charge pressures, asymmetric or biased compensating valves are used.

4.2.1 Implementation

This concept can be implemented on different pump-controlled circuits, including those

utilizing two POCVs, CC-SHV and OC-SHV described earlier in Chapter 3. The implementation

can be done utilizing two different cracking pressures POCVs or biased shuttle valves.

Figures 4-4 (a) and (b) show the commonly used circuit with one charge pressure and two identical

pilot operated check valves and the proposed circuit with two charge pressures and two different

pilot operated check valves, respectively. Figure 4-5 illustrates circuits with one charge pressure

and conventional SHVs and proposed circuits with two charge pressures and biased SHVs.

CHA
CH POCVA (P crA)
P cA

POCVA (P cr)
Pc
CHB
POCVB (P crB)
POCVB (P cr ) P cB
M

M M

(a) (b)
Figure 4-4 Circuit with: (a) one charge pressure and two identical pilot operated check valves; (b)
two charge pressures and two different pilot operated check valves.

65
 3/3 Shuttle valve 4/3 Shuttle valve

2 4

ApA ApB
3 Ap KsA KsB
Ks P cA P cB

CH CHA CHB

M
M
M

(a) (b)

3/3 Shuttle valve 4/3 Shuttle valve

2 4

3 Ap Ap Ap
Ks Ks P cA P cB Ks

CH
CHA CHB
M

M
M

(c) (d)
Figure 4-5 Circuits with: (a) one charge pressure and a 3/3 closed-center shuttle valve (SHV); (b)
two charge pressures and a 4/3 closed-center biased SHV; (c) one charge pressure and a 3/3 open-
center SHV; (d) two charge pressures and a 4/3 open-center biased SHV.

66
4.2.2 Modeling

The mathematical model of the circuit with two charge pressures and a 4/3 closed-center biased

shuttle valve (CC-SHV) is first driven. Figure 4-6 shows the circuit with 4/3 closed-center biased

shuttle valve with parameters representation in the first quadrant of operation. Mathematical

models of the actuator, pump and flow balance in the circuit are similar to these driven in the

previous chapter. However, the mathematical model of the 4/3 biased CC-SHV is driven hereafter.

Note that a biasing effect can be obtained in a SHV by utilizing either two different balancing

springs or two different spool areas or both.

F fr xa , va

AA AB FL
pA pB

QB
QA

p1 p2
Q1 Q2

ApA
ApB
KsA P cA P cB KsB
QcA QcB Line B
Line A
CHA CHB

Qa Qb
pa pb
M

Figure 4-6 Circuit that utilizes two charge pressures and a 4/3 closed-center biased shuttle valve
in first quadrant of operation.

Figures 4-7 (a) and (b) show a schematic drawing of a representative 4/3 closed-center biased
shuttle valve in center position and areas of the valve openings as function of the effective pressure
acting on the spool, respectively. The flows through the valve openings are represented by the
following equations [15]:

67
 2
𝑄1 = 𝐶𝑑 𝐴𝑣𝐴 √ |𝑝1 − 𝑃𝑐𝐴 | 𝑠𝑖𝑔𝑛 (𝑝1 − 𝑃𝑐𝐴 ) (4-1)
𝜌

2
𝑄2 = 𝐶𝑑 𝐴𝑣𝐵 √ |𝑃𝑐𝐵 − 𝑝2 | 𝑠𝑖𝑔𝑛 (𝑃𝑐𝐵 − 𝑝2 ) (4-2)
𝜌

where 𝑃𝑐𝐴 and 𝑃𝑐𝐵 are the charge pressures connected to lines A and B of the circuit, respectively;

valve opening areas at sides A and B are denoted as 𝐴𝑣𝐴 and 𝐴𝑣𝐵 .

pilot line A p1 p2 pilot line B 𝐴𝑠𝑣

𝐴𝑚𝑎𝑥
Q1 Q2 𝐴𝑣𝐴 𝐴𝑣𝐵
APA APB

kSA kSB
𝐴𝑙𝑘
xv Charge line A Charge line B
𝑝𝑐𝑟 𝑝𝑚𝑥 𝑝𝑒
(P cA) (P cB)

(a) (b)
Figure 4-7 4/3 closed-center biased shuttle valve; (a) schematic drawing in center position; (b)
opening area as function of the effective pressure.

The cracking condition of a shuttle valve depends on the pressure difference between the two

circuit lines, spool side areas and centering springs stiffness. The SHV dynamics is neglected as

compared to the rest of the system components. Instead, the static force balance of the spool is

used, and this can be represented as follows:

𝐴𝑝𝐴 𝑝1 + 𝑘𝑠𝐴 (𝑥𝐴0 − 𝑥𝑣 ) − 𝐴𝑝𝐵 𝑝2 − 𝑘𝑠𝐵 (𝑥𝐵0 + 𝑥𝑣 ) = 0 (4-3)

In equation (4-3), the spool side areas, 𝐴𝑝𝐴 and 𝐴𝑝𝐵 , and the side balancing springs stiffness,

𝑘𝑠𝐴 and 𝑘𝑠𝐵 , are chosen such that the spool is balanced in center position when the piloting

pressures are equivalent to the charge pressures. By re-arranging the parameters in equation (4-3)

the following equation is derived:

68
 𝐴𝑝𝐴 𝑝1 − 𝐴𝑝𝐵 𝑝2 + 𝐹𝑘0 + 𝑥𝑣 (𝑘𝑠𝐴 − 𝑘𝑠𝐵 ) = 0 (4-4)

where 𝐹𝑘0 = 𝑘𝑠𝐴 𝑥𝐴0 − 𝑘𝑠𝐵 𝑥𝐵0 represents balancing springs initial force. By dividing the above

equation by area of side A of the spool, 𝐴𝑃𝐴 , and assuming operation around the critical opening

condition, i.e. 𝑥𝑣 ≈ 0, we obtain the following pressure balance equation:

𝑝1 − 𝛼𝑣 𝑝2 = 𝑝𝑐𝑟 (4-5)

where 𝛼𝑣 is the area ratio between side B and side A of the spool. Generally, it is chosen based on
𝑃
the ratio of the charge pressures 𝛼𝑣 = 𝑃𝑐𝐴. For simplicity, we introduce the effective opening
𝑐𝐵

pressure, 𝑝𝑒 , as the summation of the effective opining pressures of SHV, i.e. 𝑝𝑒 = 𝑝1 − 𝛼𝑣 𝑝2.

Assuming that valve opening area is linearly proportional to the effective opening pressure, 𝑝𝑒𝐴 ,

as shown in Figure 4-7 (b), then the opening area of the side B valve can be represented as follows:

𝐴𝑙𝑘 𝑝𝑒 ≤ 𝑝𝑐𝑟
𝐴𝑝𝑐1 = {𝑘1 (𝑝𝑒 − 𝑝𝑐𝑟 ) 𝑝𝑐𝑟 < 𝑝𝑒 < 𝑝𝑚𝑥 𝐴 (4-6)
𝐴𝑚𝑎𝑥 𝑝𝑒 ≥ 𝑝𝑚𝑥𝐴

where 𝐴𝑙𝑘 is the leakage area due to valve clearances; 𝑝𝑚𝑥𝐴 and 𝐴𝑚𝑎𝑥 denote the maximum

opening pressure and area at side A of the valve.

The mathematical model of the circuit with one charge pressure and 3/3 CC-SHV is deduced in

Section 3.2.2. However, it can be easily obtained from the above-mentioned 4/3 biased CC-SHV

model by considering two similar charge pressures and a similar and non-biased valve spool.

4.2.3 Simulation studies

Simulations are performed to show the effect of shifting the critical zone towards lower

loading values on the performance of the proposed circuits. Simulation parameters and values of

the different components of the circuit are listed in Table 3-1. Parameters for the 4/3 biased CC-

SHV and 3/3 CC-SHV are shown in Table 4-1. Similar to previous circuits, simulation studies are

done for two load-motion scenarios. The first scenario is related to the circuit responses for a

69
constant load and a step input control signal. The second scenario is related to the circuit responses

for variable loading condition and a square input signal.

Table 4-1 Values of parameters of the 4/3 closed-center shuttle valve shown in Figure 4-7.

Parameter Definition Values

𝐴𝑙𝑘 Valve leakage area 0 [m2 ]
𝐴𝑚𝑎𝑥 Max flow area in valve 25 × 10−6 [m2 ]
𝑃𝑐𝑟 Cracking pressure 0.2 [MPa]
𝑃𝑚𝑥 SHV maximum opening pressure 0.5 [MPa]
𝑃𝑐𝐴 Side-A charge pressure for circuit with 4/3 SHV 1.17 [MPa]
𝑃𝑐𝐵 Side-B charge pressure for circuit with 4/3 SHV 1.56 [MPa]

4.2.3.1 Constant loading simulations

Different simulation runs to evaluate circuit performance under different loading conditions are

done. At each simulation run, a voltage step input signal is applied to the system at specific loading

condition and actuator velocity is recorded versus time. Figure 4-8 shows characteristics of the

critical switching zone of the proposed circuit with the 4/3 biased CC-SHV and two charge

pressures. Figure 4-8 (a) shows the critical zone position in the 𝐹𝐿 -𝑣𝑎 plane. Compared to circuit

with 3/3 CC-SHV, critical zone of the proposed circuit is reduced and shifted about 1200 N

towards less load margin. Reductuion in load values where oscillation occurs reduced the adverse

effects of oscillationg load on the machine. Circuit performance during switching zone is shown

by tracking responses of 5 selected operating points that have the same step input signal and

different loading conditions. Figures 4-8 (b) to (f) show the actuator velocity responses at the test

points TP1 through TP5, representing loads of 3700N, 2700N, 1500N, 940N and 500N.

70
 0

-12
Velocity (cm/s)
TP1
TP5 TP3 TP2
TP4
-24

-36
0 1 2 3 4 5 6
Load (kN)
(a)
0 0
Velocity (cm/s)

Velocity (cm/s)
TP1 TP2
-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(b) (c)
0 0
Velocity (cm/s)

Velocity (cm/s)

TP3 TP4
-6 -6

-12 -12

-18 -18
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(d) (e)
0
Velocity (cm/s)

TP5
-6

-12

-18
0 0.5 1 1.5 2
Time (s)
(f)

Figure 4-8 Simulation results for circuit with biased closed-center shuttle valve and two charge
pressures, 1.17 MPa and 1.56 MPa: (a) construction of critical region; (b)-(f) actuator velocity
response at test points TP1-5 for step input signal of -4 V and assistive loads of 3700N, 2700N,
1500N, 940N and 500N, respectively.

71
 Figure 4-8 (d) illustrates the actuator velocity response at TP3 that possesses the most severe

oscillation in the critical zone. Velocity reponse at this point is oscillatory with mean value,

oscillation amplitude and frquency of 10.4 cm/s, 4.8 cm/s and 8 Hz, respectively. The comparison

between Figures 3-15 (d) and 4-8 (d) shows that velocity oscillation amplitude in the critical zone

of proposed circuit is reduced by 25% compared to that of conventional circuit with 3/3 SHV.

Reduction in cylinder actuation force and velocity oscillation amplitudes reduces the operator’s

inconvenience as well as the destructive effect on the machine. Although this solution does not

totally alleviate osillations, it reduces its effect and facilitaes the application of other solutions.

4.2.3.1 Variable loading simulations

Figures 4-9 and 4-10 show the simulation results of circuit performance under low and high

loading conditions, respectively. Figures 4-9 (a) and 4-10 (a) show the square input control signals.

Figures 4-9 (b) and 4-10 (b) illustrate the applied load patterns. The actuator velocity versus time

is shown in Figures 4-9 (c) and 4-10 (c). It is clear that velocity oscillation occurs during switching

zone S43 under low loading condition. Nevertheless, velocity is oscillation-free under high loading

condition. Figures 4-9 (d) and 4-10 (d) illustrate pressures at both sides of pump versus time.

Similar to the velocity responses, pressure oscillations occurred in switching zone S43 under low

loading condition. Figures 4-9 (e) and 4-10 (e) show the circuit performance on the FL-va plane.

Note that the velocity oscillations occurring in switching zone S43 under low loading condition in

Figures 4-9 (e) match those of the constructed critical zone shown in in Figure 4-8 (a). The

relationship between pressures at both sides of the circuit under low and high loading conditions

are shown on the pump pa-pb in Figures 4-9 (f) and 4-10 (f), respectively. The proposed circuit

with 4/3 CC-SHV shows improved performance when compared to that of the circuit with

conventional 3/3 CC-SHV.

72
 For instace, comparing velocity responses in Figures 3-16 (c) and 4-9 (c) shows that velocity

oscillation in critical zone of proposed circuit is greatly reduced compared to that of conventional

circuit with SHV. Similar improvements are obtained in all other performance parameters under

low and high loading conditions.

4 4
Input signal (V)

2 2

Load (kN)
0 0

-2 -2

-4 -4
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(a) (b)
8 4
Velocity (cm/s)

Pressure (MPa)

3
0
2
-8
1

-16 0
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(c) (d)
8 4
Velocity (cm/s)

3
0
pB (MPa)

2
-8
1

-16
0
-4 -2 0 2 4
0 1 2 3
Load (kN) pA (MPa)
(e) (f)

Figure 4-9 Simulation results for circuit with biased closed-center shuttle valve and two charge
pressures under low loading conditions given data in Tables 3-1 and 4-1: (a) input control signal;
(b) applied load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pump pressure plane.

73
 6 16

3 8
Input signal (V)

Load (kN)
0 0

-3 -8

-6 -16
0 5 10 15 20 25 0 5 10 15 20 25
Time (s) Time (s)
(a) (b)
16 10
8

Pressure (MPa)
8
Velocity (cm/s)

6
0
4
-8 2

-16 0
0 5 10 15 20 25 0 5 10 15 20 25
Time (s) Time (s)

(c) (d)
16 10
8
8
Velocity (cm/s)

pB (MPa)

6
0
4
-8 2

-16 0
-16 -8 0 8 16 0 2 4 6 8
Load (kN) pA (MPa)

(e) (f)
Figure 4-10 Simulation results for circuit with biased closed-center shuttle valve and two charge
pressures under high loading conditions given data in Tables 3-1 and 4-1: (a) input control signal;
(b) applied load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pump pressure plane.

74
4.3 Concept II- selective leakage

The concept of applying leakage to stabilize pump-controlled circuits was introduced by Wang

et. al. [11, 37] and Caliskan et. al. [13]. Both researchers applied this concept over circuits with

shuttle valves. Wang et. al. [11, 37] used two extra electrically-controlled regulating valves to

control leakage through an additional control loop. Caliskan et. al. [13] used an open-center shuttle

valve to incorporate leakage control together with flow compensation. However, as was reported

by [13] and as shown in Chapter 3, this solution does not cover the full undesirable margin and is

limited to a specific critical speed. It is seen from the comparative analysis in Section 3.4 that the

undesirable margin is smaller in size and exhibits lesser oscillation amplitudes in circuits with two

POCVs as compared to that in circuits with SHVs. Thus, applying the concept of controlled

leakage to the circuit with two POCVs is expected to give better results. In this section, applying

leakage to stabilize pump-controlled circuits that utilize POCVs is proposed.

4.3.1 Implementation

Leakage control in pump-controlled circuits that utilize POCVs is achieved by adding flow

control valves either in main flow lines or in differential flow lines, as can be seen in Figures 4-11

(a) and (b), respectively. Circuit with two controlled leakage valves next to the actuator ports

shown in Figure 4-11 (a) is used as proof of concept.

4.3.2 Modeling

The mathematical model of the circuit with leakage control is obtained by adding the

mathematical model of flow through a leakage valve to the model of the conventional circuit with

POCVs derived in Section 3.1.4. Leakage through flow throttling valves is described as follows:

75
 QLB
QLA

CH CH

POCVA (P cr ) POCVA (P cr ) POCVB (P cr )

Pc Pc

POCVB (P cr )
QLA QLB

M
M
M

(a) (b)
Figure 4-11 Using leakage to stabilize pump-controlled circuits that utilize POCVs: (a) leakege in
main flow lines; (b) leakege in differential flow lines.

2
𝑄𝑙𝐴 = 𝐶𝑑 𝐴𝑙𝐴 √𝜌 |𝑝𝐴 − 𝑃𝑐 | 𝑠𝑖𝑔𝑛 (𝑝𝐴 − 𝑃𝑐 ) (4-7)

2
𝑄𝑙𝐵 = 𝐶𝑑 𝐴𝑙𝐵 √ |𝑝𝐵 − 𝑃𝑐 | 𝑠𝑖𝑔𝑛 (𝑃𝑐 − 𝑝𝐵 ) (4-8)
𝜌

where 𝑄𝑙𝐴 and 𝐴𝑙𝐴 and 𝑄𝑙𝐵 and 𝐴𝑙𝐵 are flow and throttling areas in leakage valves at sides A and

B of circuit, respectively.

4.3.3 Simulation studies

Similar to the previous circuits, simulation studies are done for two load scenarios. The first

scenario is related to constant load and a step input control signal, and the second scenario is related

to variable loading condition and a square input signal. The simulation parameters for circuit are

listed in Table 3-1, while the selected valve leakage areas are listed in Table 4-2.

76
 4.3.3.1 Constant loading simulations

Similar to simulation of conventional circuit with two POCVs in Section 3.1.5, different runs

to evaluate proposed circuit performance under different conditions are done. Figure 4-12 shows

that the circuit performance is oscillation-free over the tested zone that includes the switching zone

S43. Figure 4-12 (a) shows velocity responses of the actuator at different loads and input control

voltages to the pump. The external load varies from 12000 N to 0 N with step of 10 N, while the

input voltage values are -3 V, -4 V, -6 V and -8 V. Figure 4-12 (a) shows that velocity increases

gradually when switching between the 4th and 3rd quadrants. The velocity dynamic responses at

three test points TP1, TP2 and TP3 are shown in Figures 4-12 (b) - (d). Test points are located at

the beginning, middle, and end of the switching zone S43 corresponding to external loads of 4200

N, 3000 N and 1700 N, respectively. Applied step input signal was -4 V. Figures 4-12 (b) - (d)

show the damped velocity responses for circuit with leakage control at all test points.

Table 4-2 Leakage areas at different input signals to pump.

Input signal (V) Leakage area in Leakage area in

Line A (mm2) Line B (mm2)
-3 0.5 0.45
-4 1 0.9
-6 2.1 1.8
-8 3.2 2.7

77
 0
-3 V
-4 V
-9
Velocity (cm/s)
TP1
TP2 -6 V
-18 TP3

-8 V
-27

-36
0 2000 4000 6000 8000 10000 12000
Load (N)
(a)
0 0
TP1
Velocity (cm/s)

TP2

Velocity (cm/s)
-5 -5

-10 -10

-15 -15
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s)
Time (s)

(b) (c)
0
TP3
TP3
Velocity (cm/s)

-5

-10

-15
0 0.5 1 1.5 2
Time (s)
(d)
Figure 4-12 Simulation results for circuit with pilot-operated check valves and leakage valves,
given data in Tables 3-1 and 4-2: (a) velocity steady state responses at different loads and step
input signals on the FL- va plane; (b-d) velocity response at TP1-TP3 at beginning-middle and
end of critical zone for -4 V step input and loads of 4200N, 3000 N and 1700 N.

78
 4.3.3.2 Variable loading simulations

Simulation results of circuit performance under low and high variable loading conditions are

shown in Figures 4-13 and 4-14. Figures 4-13 (a) and 4-14 (a) illustrate the square input control

signals under low and high loading conditions, respectively.

4 6
Input signal (V)

2 3

Load (kN)
0 0

-2 -3

-4 -6
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)

(a) (b)
8 6
Pressure (MPa)
Velocity (cm/s)

4
4
0
2
-4

-8 0
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)

(c) (d)
6
8
Velocity (cm/s)

4 4
pB (MPa)

0
2
-4

-8 0
-6 -4 -2 0 2 4 6 0 1 2 3 4
pA (MPa)
Load (N)

(e) (f)
Figure 4-13 Simulation results for circuit with pilot-operated check valves and leakage control
under low loading condition, given data in Tables 3-1 and 4-2: (a) input control signal; (b) variable
load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b (dotted line);
(e) load-velocity plane; (f) pressure plane.

79
 Load patterns are shown in Figures 4-13 (b) and 4-14 (b), respectively. Figures 4-13 (c) and

4-14 (c) show the actuator velocity versus time under both loading conditions. Figures 4-13 (d)

and 4-14 (d) illustrate pressures at both sides of the pump versus time. It is clear from the velocity

and pressure responses that the system is oscillation-free in all operating conditions.

6 16
Input signal (V)

3 8

Load (kN)
0 0

-3 -8
-6 -16
0 6 12 18 24 0 6 12 18 24
Time (s) Time (s)

(a) (b)
16 10
8
Pressure (MPa)
Velocity (cm/s)

8
6
0
4
-8 2
-16 0
0 6 12 18 24 0 6 12 18 24
Time (s) Time (s)
(c) (d)
16 10
Velocity (cm/s)

8
8
pB (MPa)

6
0
4
-8 2
-16 0
-16 -8 0 8 16 0 2 4 6 8
Load (kN) pA (MPa)
(e) (f)
Figure 4-14 Simulation results for circuit with pilot-operated check valves and leakage control
under high loading condition, given data in Tables 3-1 and 4-2: (a) input control signal; (b)
variable load; (c) actuator velocity versus time; (d) pressures at pump ports a (solid line) and b
(dotted line); (e) load-velocity plane; (f) pressure plane.

80
Load-velocity plane in Figures 4-13 (e) and 4-14 (e) and pressure plane in Figures 4-13 (f) and 4-

14 (f), in low and high loading conditions, further prove the improved performance of the proposed

circuit. Similar to the circuit with two POCVs, pressures at both sides of the circuit are almost

equal and less than the charge pressure at switching zone S21. The relative suction status at both

sides of the circuit at S21 activates both POCVs that connects both sides of the circuit to charge

pressure. Contrarily, pressures at both sides of the circuit are almost equal and higher than the

charge pressure at switching zone S43.

4.4 Concept III- selective throttling

The main idea behind the concept of applying selective throttling is to utilize flow throttling to

dampen system oscillations. Throttling is applied only in undesirable regions where responses are

prone to become oscillatory. In other operating regions, motion is throttle-less. Throttling of

hydraulic fluid creates pressure difference (Δp) across the valve orifice maintaining increased

pressure in cylinder chambers as compared to pump ports. This in turn contributes towards a stiffer

actuator [12, 35, 38]. The proposed concept allows circuits to have a comparable efficiency and

energy regeneration ability to the conventional pump-controlled circuits in high loading

conditions. It also achieves stability of the circuits with throttling valves under low loading

conditions.

4.4.1 Implementation

This concept is implemented by proposing circuits with new compensating valve(s) to perform

throttling along with flow compensation [39]. One other solution is obtained by adding a special

valve(s) to the previously-designed pump-controlled circuits [35]. Both solutions have to meet the

following requiremnts [35]: (1) applying the proper throttling over flow when the two pilot

81
pressures to the compensating valve(s) are close to each other; (2) allowing free flow when the

two pilot pressures are not close to each other and throttling is unnecessary. This concept is

applicable to previously designed circuits with pilot-operated check valves and those with shuttle

valves [35]. Figure 4-15 shows two different solutions to add the selective-throttling valves into

circuit that utilises two POCVs.

Δ pA QA QB Δ pB
CH
CH
Pc
POCVA
(P cr) POCVB
POCVA (P cr) POCVB (P cr) Δ p1 (P cr) Δ p2
Pc

Q1 Q2
M

M
M M

(a) (b)
Figure 4-15 Proposed location for additional selective-throttling valve in the circuit that utilizes
two POCVs, (a) in the main flow lines; (b) in the differential flow lines.

4.4.2 Modeling

The performance of a circuit that utilizes two POCVs and equipped with two throttling valves

next to the actuator ports (shown in Figure 4-16) is addressed here. The mathematical model is

obtained by adding throttling effect to mathematical model of the conventional circuit with POCVs

derived in Section 3.1.4. Figure 4-16 shows the circuit proposed in Figure 4-15 (a) with notations

in the first quadrant of operation. In this model two pressure variables pA1 and pB1 are added

82
downward of the throttling valves. The mathematical model of the flow through throttling valves

is described as follows:

2
𝑄𝐴 = 𝐶𝑑 𝐴𝑡ℎ𝐴 √𝜌 |𝑝𝐴1 − 𝑝𝐴 | 𝑠𝑖𝑔𝑛 (𝑝𝐴1 − 𝑝𝐴 ) (4-9)

2
𝑄𝐵 = 𝐶𝑑 𝐴𝑡ℎ𝐵 √ |𝑝𝐵 − 𝑝𝐵1 | 𝑠𝑖𝑔𝑛 (𝑝𝐵 − 𝑝𝐵1 ) (4-10)
𝜌

where 𝐴𝑡ℎ𝐴 and 𝐴𝑡ℎ𝐵 are throttling areas in valves at sides A and B of the circuit, respectively.

F fr va

FL
pA
pB

QA QB

pA1 CH pB1

p2
POCVA POCVB
Pc Qc
p1 Q2
Q1
Line B

Line A

Qa Qb
pa pb

Figure 4-16 Schematic drawing of the circuit that utilizes POCVs equipped with limited throttling
valves, in the first quadrant of operation.

83
4.4.3 Simulation studies

Simulations are used to show the damping effect of flow throttling on system performance.

Similar to the previous circuits, simulation studies are done for two load-motion scenarios. The

first scenario is related to constant load and a step input control signal, whereas the second scenario

is related to variable loading condition and a square input signal. Simulation parameters for the

circuit are listed in Table 3-1 , whereas the proposed throttling areas of valve at different input

signals to pump are listed in Table 4-3.

4.4.3.1 Constant loading simulations

Simulation is conducted under different loading conditions. At each simulation run, a voltage

step input signal is applied to the system at constant loading condition and actuator velocity versus

time is recorded. At each of the four input signals (-3 V, -4 V, -6V and -8 V), the applied load

varied from 0 N to 12,000 N with 2 N incremental step. Figure 4-17 (a) shows the simulation

results of the steady state actuator velocity at each of the different runs. Figure 4-17 (a) shows that

oscillatory switching zone is narrowed and almost turned into a line, while velocity abruptly

increased when switching between the 4th and 3rd quadrants. Figure 4-17 (b) and (c) show the

damped velocity responses at test points TP1 and TP2 located just before and after switching at

input signal of -4 V and loads of 6202 N and 6200 N, respectively. Abrupt changes in velocity

responses that are notices in S43 do not indicate any performance issues. Note that Figure 4-17 (a)

is generated by combining steady state velocity responses of different simulation runs. Simulations

that show realistic switching situation are illustrated in the second scenario in the following

section. Simulation results showed the enhanced performance of the proposed circuit compared to

that of a conventional circuit described in section 3.1. Note that the throttling areas that alleviate

oscillations at 4 different input signals are predicted based on the trial and error concept. Table 4-3

84
shows the input control signals to pump and corresponding throttling areas at lines A and B of the

circuit used in simulations.

Table 4-3 Throttling areas at different input signals to pump

Input signal Throttling area in Throttling area in

(V) Line A (mm2) Line B (mm2)

-3 9 6.75
-4 10 7.5
-6 20 15
-8 30 22.5

0
-3 V

-4 V
Velocity (cm/s)

-12 TP1
TP2
-6 V
-24
S43
-8 V
-36
0 2 4 6 8 10 12
Load (kN)

(a)
0 0
Velocity (cm/s)
Velocity (cm/s)

TP2
-5 TP1 -5

-10 -10

-15 -15
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)

(b) (c)
Figure 4-17 Simulation results for circuit with pilot-operated check valves and selective throttling
valves, given data in Tables 3-1 and 4-3; (a) steady state velocity responses around switching zone
S43 in the load-velocity plane; (b) velocity response at TP1 at load of 6202 N in the fourth
quadrant; (c) velocity response at TP2 at load of 6200 N in the third quadrant of operation.

85
 4.4.3.2 Variable loading simulations

Simulations are conducted as shown in Figures 4-18 and 4-19. Figures 4-18 (a) and 4-19 (a)

illustrate the square input control signals under low and high loading conditions, respectively. The

applied load patterns are shown in Figures 4-18 (b) and 4-19 (b), respectively. Figures 4-18 (c)

and 4-19 (c) show the actuator velocity versus time in both loading conditions.

4 6

2 3
Input signal (V)

Load (kN)
0 0

-2 -3

-4 -6
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)

(a) (b)
8 6
Pressure (MPa)

4
Velocity (cm/s)

4
0
2
-4

-8 0
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(c) (d)
8 6
Velocity (cm/s)

4
4
pB (MPa)

0
2
-4

-8 0
-6 -4 -2 0 2 4 6 0 1 2 3 4
Load (kN) pA (MPa)
(e) (f)
Figure 4-18 Simulation results for circuit with pilot-operated check valves and selective throttling
valves at low loading conditions, given data in Tables 3-1 and 4-3: (a) input control signal; (b)
applied load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b
(dotted line); (e) load- velocity plane; (f) pressure plane.

86
 Figures 4-18 (d) and 4-19 (d) illustrate the pressures at both sides of the pump versus time. It is

clear from the velocity and pressure responses that the system is oscillation-free under all operating

conditions. Figures 4-18 (e) and 4-19 (e) show circuit responses in the FL-va plane and Figures 4-

18 (f) and 4-19 (f) show circuit responses in pa-pb plane, at low and high loading conditions,

respectively, Figures further prove the improved performance of the proposed circuit.

6 16
Input signal (V)

3 8

Load (kN)
0 0

-3 -8

-6 -16
0 6 12 18 24 0 6 12 18 24
Time (s) Time (s)

(a) (b)
12 12
Pressure (MPa)
Velocity (cm/s)

6
8
0
4
-6

-12 0
0 6 12 18 24 0 6 12 18 24
Time (s) Time (s)
(c) (d)
12 12
Velocity (cm/s)

6
pB (MPa)

8
0
4
-6

-12 0
-16 -8 0 8 16 0 2 4 6 8
Load (kN) pA (MPa)
(e) (f)

Figure 4-19 Simulation results for circuit with pilot-operated check valves and selective throttling
valves at high loading conditions, given data in Tables 3-1 and 4-3: (a) input control signal; (b)
applied load; (c) actuator velocity versus time; (d) pressure at pump ports a (solid line) and b
(dotted line); (e) load- velocity plane; (f) pressure plane of the circuit.

87
4.5 New designs

Through applying the above-mentioned concepts, either, individually or collectively, 8 new

pump-controlled circuits for single-rod cylinders are proposed in this section. Note that the designs

are proposed by two charge pressures.

4.5.1 Design 1

In this design, shifting the undesirable performance region into lower load margin

(Concept I) is applied through utilizing two different charge pressures and two unidentical POCVs.

Figure 4-20 (a) and (b) show detailed and simplified schematic drawings of design 1, respectively.

The proposed circuit in Figure 4-20 (a) differs from the conventional circuit in Figure 3-6 in that

the two charging lines are independent from one another and fed by two different outputs of the

charging system. Charging line to POCVB is installed such that it is fed directly by the charging

pump, but charging line to POCVA is indirectly fed by the charging pump via a pressure reducing

valve. Figure 4-20 (b) shows a simplified drawing of Design 1 where the charge pressures are

represented by simple pressure source symbol. Lower and higher charging pressures feeding the

cap-side and rod-side of the actuator, respectively, are chosen such that the actuation force on

actuator is zero at zero input signal (neutral position). Feeding of POCV A by a lower charging

pressure than POCVB causes the critical operation zone to shift toward the origin of the FL-va plane

along the x-axis. Thus, lowering the load force range spanned by critical zone. Since the oscillation

in the hydraulic circuit occurs at lower loading values, the effective degree of vibration

experienced by the machine is less pronounced, thus improves the overall machine functionality.

The simulation studies in Section 4.2 further prove the enhanced performance of proposed circuits

compared to the original ones. Figure 4-21 illustrates the flow patterns of Design 1 in four

quadrants of operation.

88
 3 POCV A (PcrA)

Pressure
reducer
1 PcA
2
POCV B (PcrB)
PcB

M
Line A Line B

Charge
system

(a)

POCV A (PcrA)
PcA

CHA

POCV B (Pcr B)
PcB

CHB

(b)
Figure 4-20 Design 1: circuit with two pilot-operated check valves (POCVs) and two charge
pressures (CHs): (a) detailed circuit drawing; (b) simplified circuit.

89
 va

F fr va F fr va
FL FL
pA
pB
QA QB QA QB
POCVA (PcrA) p2
PcA
Q1 Q1
CHA CHA

Line A Line B
POCVB (PcrB)
PcB
Q2
p1

CHB CHB

Qa Qb
Qa Qb
pa pb
M

M
M
Q2 Q1

Q3 Q4 FL

F fr va F fr va
FL FL

QA QB QA QB

Q1
CHA CHA

p1 l l Q2

CHB CHB

Qa Qb Qa Qb
M

M M

Figure 4-21 Flow patterns in the four quadrants of operation of Design 1.

Note that the flow pattern of Design 1 is similar to that of the circuit with two POCVs explained

in Section 3.1. In the first operating quadrant Q1, the main pump provides flow to the cap-side of

the actuator through Line A. As the pressure in Line A builds up, sufficient pressurization of pilot

90
input p1 opens POCVB connecting the higher charge pressure to Line B through an unthrottled

connection, while Line A is closed off from the charging system.

In quadrants Q1 and Q2, the charge pump feeds lines LB and LA to balance the flow to the main

pump and actuator, respectively. In quadrants Q3 and Q4, the relief valve in the charging system

releases the extra flow from lines LA and LB, respectively.

4.5.2 Design 2

Similar to Design 1, Design 2 performs shifting of the critical zone to lower range on the load

force axis (concept I). In addition to the two different charge pressures, Design 2 utilizes a 4/3

biased CC-SHV. Figure 4-22 shows the schematic drawing of Design 2. Similar to Design 1,

Lower and higher charging pressures feeding the cap-side and rod-side of the actuator,

respectively, are chosen such that actuation force on actuator is zero at neutral position (zero

velocity of actuator). Two pilot inputs of the valve drives spool in opposing directions out of a

default center position against the resistance of respective springs. Valve biasing is achieved either

by using two different piloting areas (ApA and ApB) or balancing springs with two different stiffness

constants (KsA and KsB) at the piloting sides of the shuttle valve. Piloting areas and/or balancing

springs are designed such that valve spool is balanced in the default center position when piloting

pressures equal charging pressures. In center position, the valve spool blocks connections between

circuit’s main lines (Line A and Line B) and the charging system. Detailed simulation studies of

the circuit and the newly proposed 4/3 biased CC-SHV are performed in Section 4.2.

Flow patterns in the four quadrants of operation of Design 2 are shown in Figure 4-23. The

main difference in operation of Design 2 compared to Design 1 is that the movement of the valve

spool simultaneously controls the compensating flow to both sides of the circuit. In the first

operating quadrant Q1, the main pump provides flow to the cap-side of the actuator through Line

91
A. As the pressure in Line A builds up, sufficient pressurization of pilot input p1 shifts valve spool

to connect the higher charge pressure to Line B through an unthrottled connection, while Line A

is closed off from the charging system.

4/3 biased closed-

center shuttle valve

ApA ApB
KsA KsB
PcA PcB

CHA CHB

M
M

Figure 4-22 Design 2: circuit with a 4/3 biased closed-center shuttle valve (CC-SHV) and two
charge pressures (CHs).

92
 va
F fr va F fr va
FL pA FL
pB
Q
QA QA QB
B
Q1 Q1 Q2
p1 p2

ApA ApB
Line A KsA KsB Line B
P cA P cB

CHA CHB
CHA CHB

Qb Qa Qb
Qa
pa pb

M
M
M

FL
F fr va F fr va
FL FL

QB QA QB
QA
Q2
Q1

CHA CHB
CHA CHB
Qa Qb
Qa Qb

M
M

M
M

Figure 4-23 Flow patterns in the four quadrants of operation of Design 2.

93
4.5.3 Design 3

A schematic drawing of the proposed design comprises a 4/3 biased OC-SHV and two charge

pressures is shown in Figure 4-24. This design implements Concepts I and II to shift the

undesirable performance region into reduced load values and applies leakage control to dampen

potential system oscillations at this shifted critical region. Combining the two concepts reduces

the leakage needed to stabilize the system and saves energy. Figure 4-25 shows the flow patterns

in the four quadrants of operation of Design 3.

F fr va
FL
pA
pB
QA QB
4/3 biased open-center
shuttle valve
Q1 Q2

ApA ApB
KsA KsB
Line A P cA P cB Line B

CHA CHB

Qa Qb
pa pb

M
M

Figure 4-24 Design 3: circuit with a 4/3 biased open-center shuttle valve (OC-SHV) and two
charge pressures (CHs).

94
 va
F fr va F fr va
FL pA FL
pB

QA QB QA QB

Q1 Q1 Q2

Line A ApA ApB Line B

KsA KsB
P cA P cB

CHA CHB
CHA CHB
Qa Qb Qa Qb

M
M

M M

FL
F fr va F fr va
FL FL

QA QB QA QB

Q1 Q2

CHA CHB CHA CHB

Qa Qb Qa Qb
M
M

M M

Figure 4-25 Flow patterns in the four quadrants of operation of Design 3.

Operation of this design in the four quadrants is similar to that of Design II except for operation

around valve center position. In the default center position, OC-SHV provides a pre-designed

leakage flow between Line A and Line B and their corresponding charge pressures.

95
 4.5.4 Design 4

The design utilizes concepts I and III to shift the undesirable zone into a lower load margin and

throttle the flow in this zone. Figure 4-26 shows Design 4 that implements two POCVs and two

charge pressures for flow compensation besides a special pressure-operated selective-throttling

valve. The selective throttling valve, also called limited throttling valve (LTV), throttles flow only

in the undesirable region, but it allows throttle-less flow in other operating regions. Note that this

design can also be implemented using two throttling valves at each side of the circuit. Figure 4-27

shows the flow patterns in the four quadrants of operation of Design 4.

Valve flow area

Free flow zone

p2
Throttled flow zone LTV
P crA
0 P cr P cA
Effective pressure
CHA

P crB
CHB P cB
P1
M

Figure 4-26 Design 4: circuit with two pilot-operated check valves (POCVs), a selective-throttling
valve (LTV) and two charge pressures (CHs); LTV flow area versus the effective pressure is shown
in the inset.

96
 v va
F fr F fr va
a FL
FL
pA pB
QA QB QA
QB

P crA p2
P cA
CHA
Q1 CHA Q1
P crB
P cB
CHB CHB
p1
Q2 Line B
Line A
Qa Qb
Qb
Qa
M

M M

M
FL

F fr va F fr va
FL FL

QA
QA QB QB

CHA
Q1 CHA
Q2
CHB CHB

Qa Qb Qa Qb
M
M

M M

Figure 4-27 Flow patterns in the four quadrants of operation of Design 4.

97
 Design 4 differs from Design 1 in the addition of a selective-throttling valve (LTV).

Particularly, the illustrated valve is a 2-way/3-position dual-pilot valve installed in the first main

fluid line LA near the connection to the cap side of the actuator. The main purpose of this valve is

to dampen oscillations during critical operating conditions.

The pilot-operated inputs at opposing ends of the valve are activated via pilot paths from the

two pilot lines of the POCVs. Thus, fluid pressure from Line A drives the valve in one direction

out of a normally centered position, while fluid pressure Line B drives the valve in an opposing

direction. Once again, the motion of the valve in each direction out of center position is resisted

by a respective spring whereby the springs cooperate to normally center the valve. Variable flow

area of the valve is implemented through, for example, a spool-sleeve throttling configuration and

balance springs in order to achieve the flow-area profile shown in the inset of Figure 4-27

Inset of Figure 4-27shows that the open flow area of the valve is zero at its centered position

(zero-displacement). In each direction from the centered position, the flow area gradually

increases at a first rate denoted by the gradual slope rising slowly away from the origin of the

graph. The flow area rate dramatically rises at a predetermined point of displacement to a notably

steeper slope in the graphically represented profile. Within the displacement range between the

predetermined displacement points in the positive and the negative directions from center, the low

flow area of the valve performs a throttling action on the fluid passing therethrough. Beyond these

points the flow area of the valve quickly increases to a free-flow state allowing the fluid to pass

freely therethrough with no throttling action thereon.

The pre-set displacement points at which the valve transitions from their throttling condition to

their free-flowing state are set for a given circuit according to the pilot pressures at which the load

value FL has moved beyond the critical range. Thus, throttling of the fluid in the hydraulic circuit

98
is only performed in the critical zones to dampen the vibration/oscillation experienced therein,

while the free-flowing state of the valve avoids unnecessary throttling in all other regions, which

represent the majority of the overall operating conditions of the circuit. The energy inefficiencies

of throttling are therefore only exploited where needed, while efficient unthrottled operation of the

circuit is retained elsewhere.

4.5.5 Design 5

Figure 4-28 shows a circuit which employs the same concepts as Design 4, but uses readily

available off-the-shelf parts in place of the special throttling valve to provide similar selective-

throttling effect. In this implementation, counterbalance valves CBVA, CBVB are installed in Line

A and Line B, respectively, near the connections to cap and rod sides of the actuator. CBVs are

especially selected to apply selective throttling, only at low performance region. The flow patterns

in the four quadrants of operation of Design 5 are illustrated in Figure 4-29.

Generally, CBVs are throttling valves typically used for safety requirements through the whole

working range of the actuator. They throttle flow only in one direction (flow out of the actuator),

but they allow free flow through a non-return valve in opposite direction. CBVs have been used in

some pump-controlled applications [5, 12, 30, 31], but with no ability to regenerate energy [12].

Here, the CBVs are utilized to only restrict flow under low loading conditions to enhance the

performance while allowing free flow under high loading conditions to allow energy regeneration.

CBVA is operable by applying pressure at a respective pilot input port fed by a cross pilot line

connected to Line B, while CBVB is operable by respective pilot input port fed by Line A. In

addition to the cross-pilot line from the opposing main fluid line, the pilot input of each CBV is

also fed by a respective pilot path from the same main fluid line on which the valve is installed.

Each CBV is normally closed and is only opened with the sufficient pilot pressure applied from

99
either of its pilot sources. In its initial stages of opening, each CBV is only partially opened and

has a reduced flow area relative to the respective piloting pressure, and, thus, throttles the fluid

passing through it. However, as the respective pilot pressure increases due to load and

corresponding pressure increase, the CBV opens further, exposing an unrestricted flow area to

allow free, unthrottled flow therethrough. Hence, like the selective throttling valve of Design 4,

the CBV only throttles flow at low loading values, and, thus, limits throttling effect to the critical

zones shifted down to such lower loading ranges in the load-velocity plane.

F fr va
FL

QA QB
1

1
3

3
2

2
CBVA CBVB

P crA p2
P cA CHA
P crB
P cB
Line A CHB
p1
Line B

Qa Qb
M

Figure 4-28 Design 5: circuit with two pilot-operated check valves (POCVs), two counterbalance
valves (CBVs) and two charge pressures (CHs).

100
 va
Va F fr va
F fr va
FL FL
pA pB
QA QB QA QB

1
1
1
1

3
3

3
3
2
2

2
2
CBVA CBVB

Q1 Q1 P crA p2
CHA P cA CHA
P crB Line B
Line A P cB Q2
CHB CHB
p1

Qa Qb Qb
Qa
pa pb
M

M
M

FL
F fr va F fr va
FL
FL

QA QB QA QB

1
1
1
1

3
3
3
3

2
2
2
2

Q1
CHA CHA

Q2
CHB CHB

Qa Qb Qa Qb
M

M
M

Figure 4-29 Flow patterns in the four quadrants of operation of Design 5.

101
 4.5.6 Design 6

Figure 4-30 shows a schematic drawing of Design 6. This design utilizes Concepts I and III to

apply selective throttling on the differential flow of the circuit. It implements two charge pressures

and two selective throttling valves that perform flow compensation besides flow selective

throttling. Compared to Design 4, this design saves more energy since it throttles only the

differential flow which is about 25% of the main flows of lines A and B. The flow patterns in the

four quadrants of operation of Design 6 are shown in Figure 4-31.

Valve flow area

Free flow
PcrA
p2
Throttled flow P cA
0 Pcr Effective pressure
CHA PcrB
p1
P cB

CHB
M

Figure 4-30 Design 6: circuit with two selective throttling valves (LTVs) and two charge pressures
(CHs); LTV flow area versus the effective pressure is shown in the inset.

102
 va
F fr va F fr va
FL FL

pA pB
QA Q1 QB QA QB

PcrA
p2

P cB
Q2
CHA CHA PcrB
p1
P cB

CHB
CHB
Qa Qb Qa Qb

pa pb

M
M M
M

FL
F fr va
F fr va
FL
FL

QA Q1 QB
QA QB

Q2
CHA
CHA

CHB CHB
Qa Qb
Qa Qb

M
M

M
M

Figure 4-31 Flow patterns in the four quadrants of operation of Design 6.

103
 Design 6 employs Concepts I and III of shifting the critical zones and throttling the flow only

in the shifted critical zones. This design replaces each POCV in Design 1 with a respective 2-

way/2-position single-pilot selective-throttling valve. Like the 2-way dual-pilot selective-

throttling valve of Figure 4-26, each single-pilot throttling valve has a controllable variable flow

area that increases at a first rate during initial displacement, before increasing more rapidly under

further displacement. However, displacement out of the normal default position is only possible in

one direction. Like POCVs, the proposed 2/2 throttling valves are operated through cross pilot

lines connected to main fluid lines.

This design is more efficient than Design 5, as it only throttles the differential flow which is

only around 25% of the main flow. Consequently, this reduces the energy losses due to throttling.

It also reduces the number of components and complexity of the circuit required to accomplish

both critical zone shifting and vibration damping within the shifted critical zone.

4.5.7 Design 7

Figure 4-32 shows Design 7 that, similar to Design 6, accomplishes both critical zone shifting

functionality and selective-throttling functionality within the shifted critical zones. Design 7 uses

only two of off-the-shelf valves, which in this case are sequence valves (SQVs). SQVs are

especially selected to compensate for the cylinder differential flow and applies selective throttling

on it in the low-performance margin. The flow patterns in the four quadrants of operation of Design

7 are shown in Figure 4-33.

104
 SQVA
p2
PcrA
P cA

CHA
SQVB
p1
PcrB
P cB

CHB

M
M

Figure 4-32 Design 7: circuit with two sequence valves (SQVs) and two charge pressures (CHs).

The resulting effect of this design is similar to that of Design 6, as the normally closed sequence

valve closes off the respective flow line from the charging system. It throttles the fluid only during

an initial part of its opening stroke before fully opening its through-path to enable free unthrottled

flow between the charging system and the respective circuit line. Once again, only the differential

flow in the charging lines is throttled, not the main flow in the main flow lines.

105
 F fr va va
F fr va
FL
FL

QA Q1 QB QB
QA Q1

PcrA
p2

P cA
Q2
CHA CHA PcrB
p1
P cB

CHB CHB

Qa Qb Qa Qb

M
M
M

FL
F fr va F fr va
FL
FL

QA Q1 QB QA QB

Q2
CHA
CHA

CHB
CHB
Qa Qb
Qa Qb

M
M

Figure 4-33 Flow patterns in the four quadrants of operation of Design 7.

106
 4.5.8 Design 8

Figure 4-34 shows Design 8 employing a singular pilot-operated check valve POCVA installed on

Line A and connected to the lower pressure side charging system and a singular sequence valve

installed on Line B and connected to the higher pressure side of the charging system. The POCV

and the sequence valve are respectively operated through cross pilot lines whereby the circuit once

again provides both critical zone shifting (Concept I) and selective-throttling functionality

(Concept III). Figure 4-35 shows the flow patterns in the four quadrants of operation of Design 8.

P crA p2

P cA

CHA P crB
p1
P cB

CHB
M

Figure 4-34 Design 8: circuit with a pilot-operated check valve (POCV), a sequence valve (SQV)
and two charge pressures (CHs).

107
 va
F fr va F fr va
FL FL
pA
pB

QA QB QB
QA
Q1 Q1 P crA
p2

PcA
Q2
CHA CHA P crB
p1

PcB

CHB
CHB
Qb
Qa
Qa Qb
pa pb

M
M

M
FL

F fr va F fr va
FL FL

QA QB QA QB

Q1

Q2
CHA
CHA

CHB
CHB Qa Qb

Qa Qb

M
M

M
M

Figure 4-35 Flow patterns in the four quadrants of operation of Design 8.

108
4.6 Summary

In this chapter, the requirements for desirable operation of pump-controlled circuits were

provided. Moreover, three concepts to improve performance of pump-controlled circuits were

proposed. These concepts are: (1) shifting the critical zone into lower loading ranges, (2) applying

pre-designed leakage in critical zone, and (3) applying selective throttling over flow in critical

zone. Performance enhancement due to application of the aforementioned concepts was

demonstrated through simulation studies. Application of the second and third concepts totally

removed the whole critical zone while application of the first method reduced the area and effect

of oscillations. By applying the above-mentioned concepts, either individually or collectively, 8

new pump-controlled circuits for single-rod cylinders were proposed.

Each of the aforementioned designs uses valves that are exclusively pilot-operated (requiring

no electronic monitoring or control components). These valves not only perform the acceptable

switching necessary to accommodate differential flow to and from a single-rod actuator, but also,

they alleviate/reduce undesirable performance during critical operating conditions. Additionally,

the proposed designs have minimal effect on system efficiency since they apply corrective actions

(selective throttling or selective leakage) only in the critical loading zones, while allowing more

efficient throttle-less flow in the larger operational areas outside the critical loading zones.

Furthermore, four-quadrant operation is fully retained whereby motoring of the pump in two

quadrants can be used for energy regeneration purposes for optimal efficiency.

109
 CHAPTER 5

5. EXPERIMENTAL TEST RIG

A test rig was designed and constructed from ground up to perform systematic experimental

evaluation. This test rig was built up of a real John-Deere 48 excavator attachment. It possesses a

full range of operation in the four quadrants of the stick link of the excavator. Figure 5-1 shows

the different components of this test rig. It comprises a John-Deere backhoe attachment (JD-48),

supporting structure, loads, an electrically controlled variable displacement pump unit, a charge

pressure unit, instrumentations, and controlling station. Test rig design facilitates the

implementation of different pump-controlled hydraulic actuation circuits, as well as the valve-

controlled hydraulic circuit originally equipped in the backhoe.

Different pump-controlled hydraulic circuits are prototyped and their performances are

evaluated using the test rig. The previously designed (by other researchers) circuits with POCVs,

CC-SHV and OC-SHV, besides, five new proposed circuits (Designs 2, 3, 5, 7 and 8) are

prototyped and tested using the test rig. Different loading conditions are applied to the system,

covering all four quadrants of operation. The physical components of the test rig are interfaced to

the user through a data acquisition board and a real-time operating system. The results taken out

110
from this test rig are highly reliable due to the fact that it is made up of real commercial

components.

PS
3

DS
1

2 4
PS

5
6

Figure 5-1 Test rig: (1) JD-48 backhoe attachment; (2) supporting structure; (3) actuator (4) load
configurations; (5) main pump unit; (6) charge pump unit; (7) control and monitoring station; (PS)
pressure sensors; (DS) displacement sensor.

Figure 5-1 shows the entire test rig where motion of the attached mass at the end of the stick

generates the four quadrants of operation at the actuator. Furthermore, the external force at the

actuator varies proportionally to the actuator displacement. A schematic drawing of the test rig,

111
and the control structure of the experimental set up are shown in Figures 5-2(a) and (b),

respectively. Specifications of the different components of the test rig are explained in the

following sections.

Processor
x0
F fr va
DAQ
pA pB DS

QA QB

Control Signal
PS PS
Hydraulic circuit
Flow compensation
arrangement

Driver

PS PS
mg

M
M

(a) (b)

Figure 5-2 Experimental test rig; (a) schematic, (b) interface structure.

5.1 JD-48 Backhoe attachment and supporting structure

JD-48 backhoe is an attachment that is easily added to utility tractors to perform digging and

excavation tasks. It is compact and can be safely retrofitted and operated in the Lab. Also, its

hydraulic actuators are strong enough to perform all required tests in a clear and realistic manner.

The JD-48 degrees of freedom are mechanically restricted to only one which is the stick actuator.

Extension and retraction of the actuator with configured load creates the four quadrants of

operation in the circuit. A novel supporting structure has been designed, machined and assembled

to safely support the JD-48 and allows operation of the stick in its full range without need of the

112
heavy balancing mass. Figures 5-3 (a) and (b) show the JD-48 attached to a utility tractor and the

retrofitted unit in the Lab., respectively.

(a) (b)
Figure 5-3 JD-48 backhoe attachment; (a) attached to a utility tractor; (b) retrofitted in the Lab.

5.2 Loading configurations

Two different loading configurations have been designed and constructed to work with the JD-

48, weight plates and a sliding load. Where the different loading configuration are shown in Figure

5-4. Weight plates are standard 20.4 kg circular masses that can be easily attached to the end of

the stick. They are installed in groups of two, one at each side. Thus, masses are selected from

40.8kg to 367.5 kg. Such mass attachment provides proper loading conditions to run the tested

circuit in the four quadrants of operation. The second load type is a concrete block of 280 kg mass,

which is dragged back and forth on a sliding structure. The sliding load is attached to the stick

through a variable length chain to simulate jerky motions. It is well-known that excavators are

subjected to sudden load changes during the digging process or when are used in carrying or

dumping loads. These loads can be easily reconfigured and/or combined to meet the requirements

of the selected test.

113
 (a) (b)
Figure 5-4 Loading configuration of the JD-48; (a) weight plates; (b) sliding load.

5.3 Hydraulic actuator

Figure 5-5 shows the hydraulic actuator that controls stick motion. The main specifications of

the cylinder are listed in Table 5-1.

Figure 5-5 Hydraulic cylinder of the stick link

Table 5-1 Hydraulic actuator parameters

Parameter Value/Specification
Piston cap side area, Aa 31.67 cm2
Piston rod side area, Ab 23.75 cm2
Area ratio, α 0.75
Cylinder stroke 55.88 cm
Nominal pressure, Ps 17.5 MPa

114
5.4 Main pump unit

This unit comprises two tandem variable-displacement swash-plate piston pumps driven by a

fixed RPM electric induction motor. The unit is installed on a wheeled frame for mobility purposes.

Figure 5-6 shows the pump unit connected to the driver electric motor. Specifications of the main

pump unit are listed in Table 5-2.

Tandem pumps Motor

Coupling

Valves

Sensors

(a) (b)

Figure 5-6 The variable displacement swash-plate piston pump; (a) cross-section of an individual
pump; (b) dual pumps mechanically connected to electric motor and hydraulically connected to
the circuit components.

Table 5-2 Main pump unit parameters.

Parameter Specification /value

Pump type Sauer-Danfoss (42 series) variable displacement
swash-plate bi-directional axial piston Pump.
Pump displacement 0-28 cm3 ⁄Rev.
High pressure relief valve setting 17.5 MPa.
Pump control mechanism Electric displacement control unit.
Response time (max. to max.) 0.5 s.
Electromotor Toshiba (320 TC Frame), 1800 RPM, 230/460
volts, 37 kW induction motor.

115
5.5 Low-pressure charge unit

The low-pressure (LP) charge unit provides low-pressure fluid (typically 1-2 MPa) to the pump-

controlled circuit. Low-pressure flow is needed to compensate for the differential flow in the

circuit and to control the swash plate servo mechanism, lubricate and cool down the main pump

unit. Figure 5-7 shows the low-pressure pump unit connected to the driver motor. Specifications

of this unit are listed in Table 5-3.

Tank

LP Pump
Electric motor

Figure 5-7 Charge pump unit

Table 5-3 Charge unit parameters.

Parameter Specification/value
Pump type Northman (VPVC-F40-A1), variable vane pump.
Pump displacement 40 cm3 ⁄Rev.
Pump nominal RPM 1800 RPM
Pressure, 𝑃c 1-2 MPa, adjustable.
Electromotor type Toshiba (180-TC Frame), 1800 RPM, 2.2 kW, induction
motor.

116
5.6 Measurement and control devices

In order to measure and monitor the different system parameters, the circuit is equipped with

pressure transducers and displacement sensors. Four pressure transducers are installed in the

circuit, two of them at the pump ports and the others two at the actuator ports. Two angular

displacement sensors are installed at both the stick and boom rotational joints. Data acquisition

board and real-time operating system are incorporated to eliminate the time lag problem in sensing

and control implementation. Measuring instruments are shown in Figure 5-8 and their parameters

are detailed in Table 5-4.

Pressure sensors
Operated pump

Displacement sensors pressure sensors

(a) (b)
Figure 5-8 Pressure and displacement sensors installation; (a) actuator unit; (b) pump unit.

Table 5-4 Instrument parameters.

Parameter Specification/value
Pressure transducers Ashcroft, k1 series, 0-20 MPa, accuracy 0.5%.
Angular displacement sensor Novotechnik, P2500 series, resolution 0.01o.
Linear displacement sensor Bourns encoder, accuracy 8 μm.
Data acquisition I/O board Quanser ,Q8.
Host processor, operating system Desktop I7, Windows 7.
Target processor, operating system Desktop Dual core, QNX.
Pump driver Delta VC2124.

117
 CHAPTER 6

6. EXPERIMENTAL EVALUATIONS

Experiments are conducted to validate both theoretical analysis and simulation results and also

to demonstrate the practicality of the new designs developed in this thesis. Figure 6-1 shows the

test rig used in performing the experimental work and Table 6-1 summarizes its main

specifications. A detailed description of the test rig is also shown in Chapter 5. Different pump-

controlled hydraulic circuits are prototyped and tested using the test rig. Firstly, the previously-

designed circuits with POCVs, CC-SHV and OC-SHV (described in Chapter 3) are experimentally

evaluated and oscillatory performances are identified. Next, five new promising circuits (Designs

2, 3, 5, 7 and 8 out of the proposed eight circuits described in Chapter 4) are prototyped and

examined to prove the enhanced performance and practicality of the new designs.

6.1 Evaluation of previous designs

Three commonly-used circuits are tested to fully explore the performance of these circuits and

to validate the theoretical and simulation studies. The results are further used to highlight the main

differences in their responses in different operating conditions. The tested circuits are: (1) the

circuit that utilizes two POCVs; (2) the circuit that utilizes CC-SHV; (3) the circuit equipped with

118
OC-SHV. According to loading configuration, similar to simulation studies, experiments are

classified into two sets of tests. First, the circuit performance is tested at different operating points

covering all four quadrants of operation where load and input signal are kept constant.

Accordingly, the poor performance zone is verified and its location is identified in the circuit that

utilizes two POCVs. Tests are conducted using the reconfigured test rig shown in Figure 6-2.

Boom
DS

PS Stick

3 2
Weight plates

Figure 6-1 Test rig: (1) JD-48 backhoe attachment, (2) main pump unit, (3) charge pump unit, (PS)
pressure sensors, and (DS) displacement sensor.

The second sets of tests are performed on the full setup of the test rig shown in Figure 6-1. A

square input signal is applied to the main pump to obtain cyclic motion of the actuator. Note that

during one complete operating cycle of the actuator, motion of the mass at the end of the stick

generates different assistive and resistive loads that cover the four quadrants of operation. In this

case, undesirable performance transiently occurs when the circuit switches from assistive to

resistive quadrants during actuator retraction (around the neutral position, lowest point of the load).

119
Table 6-1 Specifications of the test rig.

Main components Specification

1 Actuator cap-side area, area 31.67 cm2, 0.75, 55 cm.
ratio and stroke
2 Main pump unit Sauer-Danfoss (42 series), 28 cm3/rev, electrically-
controlled variable swash-plate piston pump.
3 Charge pump unit Northman (VPVC-F40-A1), 1-2 MPa adjustable pressure
van pump.
PS Pressure transducer Ashcroft, k1 series, 0-20 MPa, accuracy 0.5%.
DS Displacement sensor Bourns encoder, accuracy 8 μm.

Testing load
for Q2 & Q3
Testing load
for Q1 & Q4

(a)

Q1 Q3

motion motion

Q4 Q2 mg
mg
(b) (c)
Figure 6-2 ReconFigured test rig used to test circuit at constant mass load; (a) a photo illustrates
the two load options; (b) a schematic drawing of load setup to test resistive extension, Q1, and
assistive retraction, Q4, quadrants; (c) a schematic drawing of load setup to test resistive retraction,
Q3, and assistive extension, Q2, quadrants.

120
6.1.1 Constant loading experiments

Figure 6-3 shows the commonly used circuit that utilizes two pilot-operated check valves

(POCVs) for motion control of a single-rod hydraulic actuator. Williamson at al. [28] studied the

performance of a skid-steer loader equipped with this circuit. They reported boom velocity

oscillations while lowering light loads at high speeds. Critical zone in such type of circuits is

experimentally identified in this section. The theoretical analysis and simulation studies of the

undesirable performance zone in circuits utilizing POCVs have been discussed in Section 3.1.

F fr xa , va
FL
pA pB
QA QB
CH

p2
POCVA POCVB
pc
Q1 Q2
Line B
p1

Line A

Qa Qb
pa pb

Figure 6-3 Circuit with two pilot-operated check valves in first quadrant of operation.

The first set of experiments is conducted to locate the position and size of the critical zone of

circuits that utilize POCVs and to validate the theoretical and simulation studies. Also, system

performance when running in critical zone is recorded and analyzed. System performance is tested

at different operating points where load and input control signal (pump swash plate angle) are

constant. The test rig is reconfigured to allow system operation at constant loading conditions.

121
Figure 6-2 (a) shows the reconfigured test rig, where the stick link is disconnected and the actuator

is connected to a mass through wires and pulleys. Figures 6-2 (b) and (c) show the schematic

drawings of load configuration used for circuit testing in the first and forth quadrants, as well as

second and third quadrants, respectively.

Different masses (0, 45, 91, 136, 177, 218 and 259 kg) are applied to the actuator and responses

are obtained at different control input signals (velocities) in each of the four quadrants. Figure 6-4

shows the results categorized and plotted in the 𝐹𝐿 -𝑣𝑎 plane based on the quality of performance.

Each vertical set of points in the Figure represents different actuator steady state velocities for the

same load value. The results make clear the non-oscillatory performance when the pump runs only

in single mode of operation in regions 1, 2, 3, and 4. However, operation in the switching regions

5 and 6 shows performance deterioration. Areas hatched with dashed lines are regions where the

pump switches mode of operation during actuator extension and retraction. The system response

becomes oscillatory in the lower switching region during actuator retraction, region 6. In the upper

switching zone during actuator extension, region 5, the system response does not experience

oscillations. However, velocity fluctuates in this region; it is not fully controllable.

Figure 6-5 shows a typical circuit performance at two operating points, covering regions 1 and

6, denoted by TP1 and TP2 in Figure 6-4. Note that, for convenience, the actuator is tested in both

extension and retraction strokes while applying the specified load in the same run. Figure 6-5 (a)

shows the input control signal to the circuit where load mass is 259 kg. The control input signal is

+3 V at TP1 in extension stroke (shaded portion) while -3.7 V is chosen at TP2 during retraction

stroke. As seen in Figure 6-5 (b), the actuator extends at 5 cm/s steady velocity during resistive

extension in the first (shaded) portion of the Figure. However, the second portion of the Figure

shows that the actuator velocity oscillates during assistive retraction stroke.

122
 TP1

TP2

Figure 6-4 Experimental identification of the critical zone (shown by hashed area) for circuit that
utilizes pilot-operated check valves (Figure 6-3).

4 10
Input signal (v)

Velocity (cm/s)

2 5
0
0
-5
-2 -10
-4 -15
7 9 11 13 7 9 11 13
Time (s) Time (s)
(a) (b)
3 3
Pressure (MPa)
Pressure (MPa)

2 2

1 1
7 9 11 13 7 9 11 13
Time (s) Time (s)
(c) (d)
Figure 6-5 Performance of the circuit with pilot-operated check valves at resistive extension and
assistive retraction (TP1 and TP2 shown in Figure 6-4) given 259 kg attached mass: (a) control
input signal; (b) actuator velocity; (c) pressures at pump port a (solid line) and port b (dotted line);
(d) pressures at actuator port a (solid line) and port b (dotted line).

123
 The velocity mean value, oscillation amplitude and frequency during retraction are -9 cm/s, 3

cm/s and 6 Hz, respectively. The first portions of Figures 6-5 (c) and (d) show that the pressure

response at pump ports and actuator ports are steady while creating the proper pressure difference

to actuate the load. Nevertheless, the pressures at actuator ports are almost equal and oscillate

around the value of 1.8 MPa (charge pressure value), with oscillation amplitude of 0.2 MPa in the

second portion of Figures. These experimental results clearly validate theoretical analysis

presented in Section 3.1.

6.1.2 Variable loading experiments

This experiment is designed to emulate the realistic operation of an excavator stick link. In this

set of tests, all three previously-designed circuits are tested to explore their performance. The

results are also used to highlight the main differences in their responses under different loading

conditions. Schematic drawings of the circuits are shown in Figures 6-3, 6-6 (a) and 6-6 (b). All

circuits are tested under low and high loading conditions where masses of 41 kg and 368 kg are

applied to the end of the stick link, as shown in Figure 6-1.

F fr x a , va F fr xa , va
FL FL
pA pB pA pB
QA QB QA QB
p1 Q1 Q2 p2
p1 Q1 Q2 p2 2 4
2 4
3
3 Line A Line B
Pc
Line A Line B
Pc
CH
CH

Qa Qb
Qa Qb
pb
pa pb
pa

M
M
M

(a) (b)
Figure 6-6 Circuits with shuttle valves: (a) 3/3 closed-center; (b) 3/3 open-center.
124
 All circuits are tested using the test rig shown in Figure 6-1. Specifications of the compensating

valves are listed in Table 6-2.

Table 6-2 Specifications of compensating valves used in previous designs

Compensating valves Model/Specification

Pilot-operated check valves Sun-hydraulics CKCB series, 0.2 MPa
cracking pressure; 3:1 pilot ratio.
Shuttle valve (closed-center) Sun-hydraulics DSCL series, 0.2 MPa
cracking pressure.
Shuttle valve (open-center) Parker series K04F3 series, 0.25 MPa
cracking pressure.

The calculated static loads at the actuator rod due to low and high weights, versus actuator

displacement are shown in Figure 6-7. Experiments at low loading conditions are designed mainly

to investigate stability issues and high loading tests are designed to examine performance and

efficiency. All tests are carried out using an open loop control mode.

12
Actuator load (kN)

-6

-12
0 0.1 0.2 0.3 0.4
Actuator displacement (m)

Figure 6-7 Calculated static load at the actuator rod for 41 kg (dashed line) and 368 kg (solid line)
masses attached to the stick.

Figures 6-8 to 6-10 show the responses of the circuits using POCVs, CC-SHV, and OC-SHV

in low loading conditions, respectively. Figures 6-8 (a) to 6-10 (a) illustrate the input control

signals to the corresponding circuits. A slight difference between input signals is attributed to the

human factor since all of them were manually applied for safety reasons. Figures 6-8 (b) and (c)

125
 through Figure 6-10 (b) and (c) show the actuator velocity trajectories and pressures at the pump

ports. Figure 6-8 (d) through Figure 6-10 (d) demonstrate the pressure plane of the aforementioned

circuits. Velocity and pressure plots show that there is no significant performance deterioration

during zone S14 (switching zone between 1st and 4th quadrants) and zone S32 (switching zone

between 2nd and 3rd quadrants) in all circuits. This is because switching occurs with the same

configuration of the valves, as mentioned earlier in Section 3.1. The performance in zone S21 is

not oscillatory, while a slight velocity drop is noticed in all circuits, especially in the circuit with

OC-SHV. In zone S21, a pressure drop is still observed in all circuits, particularly in circuit using

CC-SHV. This pressure drop is limited due to activation of the anti-cavitation valves that are

originally equipped into the pump housing. Figures 6-8 and 6-9 demonstrate the oscillatory

behavior in the circuits with POCVs and CC-SHV in the deteriorating performance zone S43.

4 9
S21
6 S14
Velocity (cm/s)

2
Input signal (V)

S32
3
0
0
-2
-3
-4 S43
-6
0 10 20 30 40 0 10 20 30 40
Time (s)
Time (s)
(a) (b)
3.5 3.5
S32 S32
Pressure (MPa)

2.8 2.8
S43
pb (MPa)

S14 S21
2.1 2.1 S43

1.4 1.4 S14

0.7 S21
0.7
0 10 20 30 40 0.7 1.4 2.1 2.8
Time (s) pa (MPa)
(c) (d)
Figure 6-8 Performance of the circuit with pilot-operated check valves for a 41kg attached mass:
(a) input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line) and b (dotted
line); (d) pump pressure plane.

126
 4 9

2 6
Input signal (V)

Velocity (cm/s)
3
0
0
-2
-3
-4 -6
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(a) (b)
4 4
Pressure (MPa)

pb (MPa)
3

2 2

1 1
0 10 20 30 40 1 1.5 2 2.5 3
Time (s) pa (MPa)
(c) (d)
Figure 6-9 Performance of the circuit that utilizes a closed-center shuttle valve for a 41kg attached
mass: (a) input signal; (b) actuator velocity; (c) pressures at pump ports a (solid line) and b (dotted
line); (d) pump pressure plane.

4 9
6
Velocity (cm/s)

2
Input signal (V)

3
0
0
-2 -3
-4 -6
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(a) (b)
4 4
Pressure (MPa)

3 3
pb (MPa)

2 2

1 1
0 10 20 30 40 1 1.5 p (MPa) 2 2.5
a
Time (s)
(c) (d)
Figure 6-10 Performance of the circuit using an open-center shuttle valve (OC-SHV) for a 41kg
attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line)
and b (dotted line); (d) pump pressure plane.

127
 The velocity oscillations in the circuit using POCVs show lower amplitude and frequency when

compared to the circuit using the CC-SHV. This is attributed to the coupled nature of the SHV that

affects both sides of the cylinder simultaneously leading to more severe dynamical changes. On

the other hand, Figure 6-10 shows non-oscillatory transition for the circuits with OC-SHVs in zone

S43, which is attributed to the damping effect of the valve pre-designed leakage.

Figures 6-11 to 6-13 illustrate the responses at high loading conditions for the circuits using

POCVs, CC-SHV, and OC-SHV, respectively. Figures 6-11 (a) to 6-13 (a) show the manually

applied control signals to the corresponding circuits. Figures 6-11 (b) and (c) to 6-13 (b) and (c)

demonstrate the actuator velocity and pressures at the pump ports for the corresponding circuits as

function of time. Figures 6-11 (d) to 6-13(d) show the pressure plane of the aforementioned circuits

and zoomed-in sections for switching zones S43 and S21. Few oscillations in the velocity and

pressure curves are noticed in all circuits in switching zones S14 and S32, which can be attributed

to the abrupt-change nature of the square input signal. Smoother input signals, e.g. sinusoidal

signal, do not display these oscillations. Zone S21 shows a non-oscillatory behavior that is similar

in all the circuits. This can be attributed to the activation of the two anti-cavitation valves in the

circuit. The velocity and pressure graphs demonstrate considerably smaller oscillatory responses

of all circuits at switching zone S43 compared to those seen in low loading conditions. However,

the responses show small, moderate and high ripples in the circuits with POCVs, CC-SHV and

OC-SHV, respectively. Higher ripples in circuits with SHVs is attributed to the global effect of

these valves on both sides of the circuit during switching, leading to vast dynamic changes.

128
 6 16

3 8

Velocity (cm/s)
Input signal (V)

0 0

-3 -8

-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)
(a) (b)
10 10
8 8
Pressure (MPa)

6 6

pb (MPa)
4 4
2 2
0 0
0 5 10 15 20 0 2 4 6 8
Time (s) pa (MPa)
(c) (d)
Figure 6-11 Performance of the circuit that utilizes pilot-operated check valves for a 368kg
attached mass: (a) input control signal; (b) actuator velocity; (c) pressures at pump ports a (solid
line) and b (dotted line); (d) pump pressure plane.
6 16
Input signal (V)

Velocity (cm/s)

3 8

0 0

-3 -8

-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)
(a) (b)
10 10
8 8
Pressure (MPa)

pb (MPa)

6 6
4 4
2 2
0 0
0 5 10 15 20 0 2 4 6 8
Time (s) pa (MPa)
(c) (d)
Figure 6-12 Performance of the circuit that utilizes a closed-center shuttle valve for a 368kg
attached mass: (a) input control signal; (b) actuator velocity; (c) pressures at pump ports a (solid
line) and b (dotted line); (d) pump pressure plane.

129
 6 16
Input signal (V)
3 8

Velocity (cm/s)
0 0

-3 -8

-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)
(a) (b)
10 10
8 8
Pressure (MPa)

pb (MPa)
6 6
4 4
2 2
0 0
0 5 10 15 20 0 2 4 6 8
Time (s) pa (MPa)
(c) (d)
Figure 6-13 Performance of the circuit that utilizes an open-center shuttle valve for a 368kg
attached mass: (a) input control signal; (b) actuator velocity; (c) pressures at pump ports a (solid
line) and b (dotted line); (d) pump pressure plane.

Figure 6-13 (b) shows that the damping effect of the predesigned leakage in the OC-SHV is

reduced at higher loads and velocities. Figures 6-11 to 6-13 demonstrate a relative improvement

in performance of the circuit using POCVs when compared to circuits using SHVs at high loading

conditions.

6.1.3 Summary

Performances of the circuits that utilize two pilot-operated check valves and those utilize

closed-center and open-center SHVs are experimentally investigated. Regions in the load-velocity

plane that show poor performance are experimentally validated for circuit with POCVs. Typical

velocity and pressure responses of the three abovementioned circuits are shown and analyzed. It

has been experimentally shown that circuits with shuttle valves encounter more challenges in all

switching zones, due to effect of valve switching on both sides of the circuit on the system

130
dynamics. The performance improvement in the circuits with open-center shuttle valves at low

loading conditions deteriorates at higher loads and velocities. On the other hand, circuits with pilot-

operated check valves experience performance issues under low loading conditions, while showing

an improved performance at higher loading conditions.

6.2 Evaluation of proposed designs

Experimental studies are performed for five new designs to evaluate their performances. These

designs are: (1) Design 5, circuit that utilizes POCVs and CBVs to apply selective throttling to

flow in the main transmission lines, shown in Figure 4-28; (2) Design 7, circuit that utilizes SQVs

to apply selective throttling to flow in the compensation transmission lines, shown in Figure 4-32;

(3) Design 8, circuit that utilizes one POCV and one SQV to apply selective throttling to flow in

one compensation transmission line, shown in Figure 4-34. (4) and (5) Designs 2 and 3, circuits

that utilize biased CC-SHV and biased OC-SHV to shift critical zone into lower loading margin,

shown in Figures 4-22 and 4-24, respectively; Specifications of the valves used in the new designs

are given in Table 6-3.

Table 6-3 Specifications of compensating valves used in the new designs

Valve Specification
Counterbalance valves (CBVs) Sun hydraulics model CAEK, 1:1 pilot ratio, 2.8-
10.5 MPa adjustable pressure.
Sequence valves (SQVs) Parker series B04H4 series, 0.2-8 MPa adjustable
cracking pressure.
Biased close-center shuttle valve Sun hydraulics model DSCL, 0.2 MPa cracking
(modified). pressure (0.14 MPa pressure offset).
Biased open-center shuttle valve Parker model: K04F3, 0.25 MPa cracking pressure
(modified) (0.14 MPa pressure offset).

131
6.2.1 Evaluation of Design 5

Figure 4-28 shows Design 5. This design is an implementation of Design 4 that utilizes special

throttling valves (LTVs) to apply selective throttling to main flow. Throttling is selectively applied

to hydraulic flow only in the undesirable performance zone while the flow is free to flow

unthrottled in all other operating zones in order to maintain system efficiency. Throttling of

hydraulic fluid dampens fluid pressure oscillations, which in turn reduces velocity oscillations.

Currently, LTVs with the required features are not commercially available. Therefore, Design 5

has been implemented by utilizing two POCVs for flow compensation and two CBVs to perform

selective throttling over main flow. To evaluate performance of Design 5, two sets of experimental

tests are performed. The first set of experiments is conducted in constant loading conditions to

make clear the performance improvements in critical operating conditions. The second set of tests

is performed to show the circuit performance and energy consumption during operation that covers

all four quadrants. Specifications of the basic components of the test rig and valves are listed in

Tables 6-1 and 6-3, respectively.

6.2.1.1 Constant loading experiments

The first set of experiments is designed to prove the oscillation-free response of the proposed

circuit in the critical zone. Those experiments were carried out on the reconfigured setup of the

test rig shown in Figure 6-2; they are similar to the experiments performed in Section 6.1.1. As

explained earlier, the system performance prone to be oscillative during the actuator retraction

when the two pressures at both sides of the circuit are close to each other. Figure 6-14 shows the

performance in a typical retraction-extension strokes of the actuator with a 245-kg mass load.

132
 6 20

Velocity (cm/s)
3 10
Input signal (v)

0 0

-3 -10

-6 -20
0 1 2 3 4 5 0 1 2 3 4 5
Time (s) Time (s)
(a) (b)
3.6 3.6

Pressure (MPa)
Pressure (MPa)

2.4 2.4
0.2 MPa

1.2 1.2
0 1 2 3 4 5 0 1 2 3 4 5
Time (s) Time (s)

(c) (d)
Figure 6-14 Performance of the circuit with two pilot-operated check valves and two
counterbalance valves in retraction and extension of a 245-kg attached mass: (a) control signal; (b)
actuator velocity; (c) pressures at pump ports a (solid line) and b (dashed line); (d) pressures at
actuator ports a (solid line) and b (dashed line).

Figure 6-14 (a) show the applied input signal to the pump to control actuator motion. Figure

6-14 (b) show the actuator velocity versus time during retraction and extension strokes. Pressures

at the pump ports and actuator ports are shown in Figures 6-14 (c) and (d), respectively.

The actuator velocity and pressure graphs are non-oscillatory in the retraction portion (negative

input signal) where the two pressures at the actuator ports are close to each other, as can be seen

in the first portion of Figure 6-14. The experiment is repeated different operating points where no

oscillations are experienced all over the tested operating regions.

133
6.2.1.2 Variable loading experiments

In the second set of experiments, circuit is tested in the four quadrants of operation using the

full test rig configuration shown in Figure 6-1. Procedures of the experimental tests are similar to

those shown in Section 6.2.1. In these tests two masses of 41 kg and 368 kg were applied

(distinctly) to the end of the stick link to investigate the circuit performance under low and high

loading conditions, respectively. Figure 6-7 shows the calculated static load at the actuator rod for

41 kg and 368 kg attached masses. Figures 6-15 (a) and 6-16 (a) show the square input signal that

is applied to the pump to control motion of the stick under low and high loading conditions,

respectively. Figures 6-15 (b) and 6-16 (b) show the actuator velocity versus time under low and

high loading conditions. Pressures at the pump ports and actuator ports are shown in Figures 6-15

(c) and (d) and Figures 6-16 (c) and (d) under low and high loading conditions, respectively. The

results show that the performance of the proposed circuit is smooth without any significant

oscillations during all switching zones. However, the proposed circuit consumes little more energy

than the circuit with only POCVs. Figure 6-17 shows the main pump delivered/received hydraulic

energies for the experiment shown in Figure 6-16. The extra energy consumed by the proposed

circuit is used to overcome the hydraulic resistance generated by the CBVs in critical zone. Note

that the amount of consumed extra energy decreases as the load increases. Figures 6-17 shows a

comparison in energy consumption between the circuits that utilize only the POCVs and that with

the CBVs. The delivered hydraulic energy from the pump to the circuit (and vice versa) is

calculated as the multiplication of the measured pressure differential across the pump and the flow

rate, 𝑊𝑝𝑚ℎ = (𝑝𝑎 − 𝑝𝑏 )𝑄. 𝑄 is calculated by multiplying the actuator measured velocity and the

piston effective area. Results show that both circuits consume energy when the load is resistive

and recuperate energy when the load is assistive.

134
 4 6
Input signal (v)

Velocity (cm/s)
2 3

0 0

-2 -3

-4 -6
0 10 Time (s) 20 30 0 10 Time (s) 20 30

(a) (b)
3 3
Pressure (MPa)

Pressure (MPa)
2.5 2.5

2 2

1.5 1.5

1 1
0 10 20 30 0 10 20 30
Time (s) Time (s)
(c) (d)
Figure 6-15 Performance of Design 5, for 41 kg attached mass; (a) input control signal; (b) actuator
velocity; (c) pressures at pump ports a (solid line) and b (dotted line); (d) pressures at actuator
ports a (solid line) and b (dotted line).

4 10

2 5
Input signal (v)

Velocity (cm/s)

0 0

-2 -5

-4 -10
0 10 20 30 0 10 20 30
Time (s) Time (s)
(a) (b)
10 10
8 8
Pressure (MPa)

Pressure (MPa)

6 6
4 4
2 2
0 0
0 10 20 30 0 10 20 30
Time (s) Time (s)

(c) (d)
Figure 6-16 Performance of Design 5, for 368 kg attached mass; (a) input control signal; (b)
actuator velocity; (c) pressures at pump ports a (solid line) and b (dotted line); (d) pressures at
actuator ports a (solid line) and b (dotted line).

135
 1

Pump power (kW)

0.5

-0.5

-1
0 10 20 30
Time (s)

Figure 6-17 Calculated hydraulic power delivered/received by the main pump for 368-kg attached
mass in the conventional circuit that utilizes pilot-operated check valves (dotted line) and Deign
5, circuit with extra counterbalance valves (solid line).

It is clear from the graph that the energy consumptions for both circuits are almost the same in

all operational quadrants except for low loading critical zone. The average delivered hydraulic

energies from the pump to the circuit are 145.1 W and 148.9 W for the circuit that utilizes only

POCVs and the proposed circuit, respectively. The average received (recuperated) hydraulic

energies from the circuit to pump are 108.7 W and 102.2 W for the circuit that utilizes only POCVs

and the proposed circuit, respectively. Note that, the charge pump in both circuits consumed about

49.5 W. The average delivered/received hydraulic energy for the circuit with POCVs only and the

proposed circuit, including the charge pump consumption, are 85.9 W and 96.2 W, respectively.

This means that the proposed circuit consumed 12% more energy compared to a commonly used

pump-controlled circuit that uses only POCVs under loading conditions prescribed in Figure 6-16.

Note that the percentage of consumed extra energy decreases as the load increases.

Summary

Experimental results of Design 5, circuit that utilizes two POCVs for flow compensation and

two counterbalance valves for system stability, demonstrated oscillation-free performance. The

energy studies showed that the proposed circuit consumed 12% more energy when compared to a

pump-controlled circuit that uses only pilot-operated check valves under a representative loading

136
condition. This extra consumed energy is used to stabilize oscillations in the circuit pressures and

actuator velocity, and it is inversely proportional to the load value. The proposed circuit has the

capability to recuperate energy. However, this capability is limited to system ability to deal with

the regenerated energy by transferring, reusing or storing.

6.2.2 Evaluation of Design 7

Design 7 shown in Figure 4-32 is an implementation of Design 6 that utilizes special

compensating LTVs to apply selective throttling over the compensating (supplementary) flow.

Similar to Design 5, throttling is selectively applied to hydraulic flow only in the undesirable

performance zone while the flow is free to flow in all other operating zones. Design 7 has been

implemented and experimentally tested using two sequence valves (SQVs). Among the different

types of SQVs, a bi-directional and externally piloted type is selected to replace the LTVs.

Performance has been evaluated and proven to be oscillation-free. It is also ready for energy

regeneration. SQVs are normally closed pressure-controlled valves that open when the pilot

pressure exceeds a predesigned cracking pressure, pcr1. They are typically used in the conventional

valve-controlled hydraulic circuits to assure proper sequence of operation for two or more

cylinders.

Similar to previous circuit, Design 7 is tested in the four quadrants of operation using the test

rig configuration shown in Figure 6-1. Figures 6-18 and 6-19 illustrate the circuit responses in low

and high loading conditions, respectively. Specifications for the basic components of the test rig

and SQV are listed in Tables 6-1 and 6-3, respectively.

Figures 6-18 (a) and 6-19 (a) show the manually applied control signals to the circuit at low

and high loading conditions, respectively. Figures 6-18 (b) and (c) and, 6-19 (b) and (c)

demonstrate the actuator velocity and pressures at pump ports as function of time. Figures 6-18 (d)

137
and 6-19 (d) show the pressure plane of the circuit with the corresponding load. Zoomed out

section for the switching zones S43 and S21 in Figure 6-19 (d) are also displayed. The velocity

and pressure graphs show that there is no significant performance deterioration in switching zones

S14 and S32 under low load condition. This is because switching occurs with the same

configuration of the valves, as mentioned earlier in Section 3.1. However, few oscillations in

velocity and pressure curves are noticed in high loading condition, which is attributed to the load

value and sharp nature of the control signal. Performance in zone S21 is not oscillatory, which is

attributed to the activation of the two anti-cavitation valves in the circuit.

4 9
6
Input signal (V)

2 Velocity (cm/s)
3
0
0
-2
-3
-4 -6
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(a) (b)
3 3
Pressure (MPa)

2.5 2.5
pb (MPa)

2 2

1.5 1.5

1 1
0 10 20 30 40 1 1.5 p (MPa) 2 2.5
Time (s) a

(c) (d)
Figure 6-18 Performance of Design 7, circuit with sequence valves, for a 41kg attached mass: (a)
input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line) and b (dotted line);
(d) pump pressure plane.

138
 6 16
Input signal (V)

Velocity (cm/s)
3 8

0 0

-3 -8

-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)
(a) (b)
9 9
Pressure (MPa)

6 6

pb (MPa)
3 3

0 0
0 5 10 15 20 0 3 6 9
Time (s) pa (MPa)
(c) (d)
Figure 6-19 Performance of Design 7, circuit with sequence valves, for a 368kg attached mass: (a)
input control signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line) and b
(dotted line); (d) pump pressure plane.

Furthermore, velocity and pressure graphs demonstrate the non-oscillatory responses of circuit

under low and high loading conditions at critical switching zone S43. This is attributed to the

damping effect of flow throttling in the SQVs. Additionally, pressure planes of the proposed circuit

show smooth pressure transitions at both sides of the circuit which leads to smoother motion of

the actuator.

To simulate realistic operation, the proposed circuit with SQVs is controlled through a joystick

input signal. One representative experiment for the circuit operation is shown in Figure 6-20.

Figure 6-20 (a) shows the joystick input control signal where a dead zone of amplitude ±1.8 V is

noticed. This dead zone is attributed to the response of the servo-control mechanism of the pump,

as referred to in the pump operation manual. Figure 6-20 (b) shows the actuator velocity response,

where good matching to the control signal and smooth switching are observed. Minor ripples are

139
noticed at seconds 5, 10, 17 and 23 of operation, which can be attributed to the dead zone effect.

The dead zone and the consequent ripples can simply be alleviated by using a joystick with a

special potentiometer that switches directly from +1.8 V to -1.8 V, or through simple controller.

Figure 6-20 (c) and (d) illustrate the pressures at both sides of the pump versus time and the

pressure plane of the circuit where smooth responses are observed. It is also noticed that there are

no pressure oscillations in zones S14 and S32 that appeared in Figures 6-18 and 6-19 due to the

relative smoothness of the input signal.

6 16

3
Input signal (V)

Velocity (cm/s)
0 0

-3 Joystick -8
dead zone
-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)
(a) (b)
9 9
Pressure (MPa)

6 6
pb (MPa)

3 3

0
0
0 5 10 15 20
0 3 6 9
Time (s) pa (MPa)
(c) (d)
Figure 6-20 Performance of Design 7 for a 368-kg attached mass: (a) joystick input signal; (b)
actuator velocity; (c) pressures at the pump ports a (solid line) and b (dotted line); (d) pump
pressure plane.

Based on the motion pattern in experiment in Figure 6-20, power consumption of the proposed

circuit is calculated and compared to that of a similar valve-controlled load sensing (LS) circuit,

as shown in Figure 6-21. Similar to the previous circuit, the net hydraulic power at the pump is

140
obtained by multiplication of the pressure differential across the pump and the flow rate. The

average delivered hydraulic power from the pump to circuit and vice versa are 258.2W and

127.8W, respectively. Thus, the total delivered hydraulic power from the pump to the circuit is

130.4W, while in the LS circuit the delivered power is 558.2W. The delivered hydraulic power

from an LS pump to the circuit has been calculated according to the following formula: 𝑊𝑝ℎ =

[(𝑝𝑎 − 𝑝𝑏 ) + ∆𝑝𝐿𝑆 ]𝑄 where, ∆𝑝𝐿𝑆 is the difference between the supplied pressure and the pressure

required by the highest load in the circuit, which typically ranges from 2 MPa to 3 MPa [16].

Conservatively, this value has been chosen to be as low as 2 MPa. Besides, the studied actuator is

typically the one that carries the highest load in a multi-actuator excavator.

Based on these assumptions, the hydraulic power delivered by the pump to the hydraulic circuit

in the proposed circuit is about 23% of that delivered by a LS pump. Note that the indicated power

here is not the total power consumed by the circuit, but the hydraulic power delivered to or received

by the pump. To calculate the total consumed electric power in this experiment, the efficiencies of

both the pump and electric motor should be considered.

2.4
Power (kW)

1.2

-1.2
0 8 16 24 32
Time (s)

Figure 6-21 Hydraulic power delivered/consumed by the pump in Design 7 (solid line) and power
consumed by the pump in a load-sensing circuit (dashed line) for experiment in Figure 6-20.

141
 Summary

Design 7 is a new pump-controlled hydraulic circuit that is equipped with selective throttling

(LTVs) compensating valves. The circuit is practically implemented using two adjustable

sequence valves. It is experimentally tested and the performance is analyzed. Experimental results

showed the enhanced performance of the proposed circuit in different loading conditions and

during realistic joystick operations. The energy studies proved the improved efficiency of the

proposed circuit when compared to a comparable LS circuit. The calculated hydraulic power

delivered by the pump to the actuator in the proposed circuit is about 23% of the power delivered

by an LS pump.

6.2.3 Evaluation of Design 8

Design 8 is shown in Figure 4-34. Two different types of flow compensation valves are used in

this circuit; a POCV is installed in Line A while a SQV is fitted in Line B. Design 8 is tested in

the 4 quadrants of operation using the test rig shown in Figure 6-1. The procedures of experimental

tests are similar to those in previous designs.

Figures 6-22 and 6-23 illustrate the circuit responses at low and high loading conditions,

respectively. Figures 6-22 (a) and 6-23 (a) illustrate the applied control signals to the circuit with

the corresponding load. Figures 6-22 (b) and (c) and 6-23 (b) and (c) show the actuator velocity

and pressures at the pump ports for the load as function of time. Figures 6-22 (d) and 6-23 (d)

demonstrate the pressure planes of the circuit at low and high loading conditions, respectively. The

velocity and pressure graphs show that there no significant oscillations during switching zones S14

and S32 at both loading conditions. However, slight damped oscillations in the velocity and

pressure curves are noticed at high loading conditions, which are attributed to the load mass inertia

effect and sharp nature of the input signal. Smooth control signals do not display these oscillations.

142
 6 16

3 8

Velocity (cm/s)
Input signal (v)

0 0

-3 -8

-6 -16
0 7 14 21 28 0 7 14 21 28
Time (s) Time (s)
(a) (b)
4 4
Pressure (MPa)

3 3

pb(MPa)
2 2

1 1
0 7 14 21 28 1 1.5 2 2.5
Time (s) pa (MPa)
(c) (d)
Figure 6-22 Performance of Design 8, circuit with a pilot-operated check valve and a sequence
valve, for 41 kg attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump
ports a (solid line) and b (dotted line); (d) pump pressure plane.

Performance in zone S21 is not oscillatory, which is attributed to the activation of the two anti-

cavitation valves in the circuit. Velocity and pressure graphs, also, demonstrate the non-oscillatory

responses of circuit at low and high loading conditions at switching zone S43. This is attributed to

the throttling effect of the SQV. Pressure planes of the circuit show smooth transition in pressures

at both sides of the circuit.

Design 8 is controlled through a joystick input signal to show more practicality of the circuit.

Two representative experiments for the circuit operation under low and high loading conditions

are shown in Figure 6-24. Figure 6-24 (a) and (b) show the joystick input control signal (dash line)

and the actuator velocity response (solid line) versus time for the low and high loading condition,

respectively.

143
 6 16

3 8
Input signal (V)

Velocity (cm/s)
0 0

-3 -8

-6 -16
0 7 14 21 28 0 7 14 21 28
Time (s) Time (s)
(a) (b)
9 9
Pressure (MPa)

6 6

pb (MPa)
3 3

0 0
0 7 14 21 28 0 2 4 6 8
Time (s) pa (MPa)
(c) (d)
Figure 6-23 Performance of Design 8, circuit with a pilot-operated check valve and a sequence
valve, for 368 kg attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump
ports a (solid line) and b (dotted line); (d) pump pressure plane.

Note that a dead zone of amplitude ±1.8 V is subtracted from the control signal in order to

demonstrate matching between actuator motion and input signal. The dead zone is attributed to

response of the servo-control mechanism of the pump, as denoted by manufacturer. However, the

effect of the dead zone is notable around times 8, 14, 19, 24, 29, 34 s and 7, 15, 22, 27 s in

experiments under low and high loading conditions, respectively. The dead zone and the

consequent ripples can simply be alleviated through simple techniques mentioned in previous

section. Both graphs show good matching of actuator velocity to the input signal. Absence of

system oscillations in zones S14 and S32 due to the relative smoothness of the control signal is

noticed. Also, no oscillations are experienced during operation in zones S43 and S21 due to the

dampening effect of flow throttling in SQV.

144
 Velocity (cm/s) 16 4 16 4

8 2

Velocity (cm/s)
Input signal (s)
8 2

Input signal (V)

0 0 0 0

-8 -2 -8 -2

-16 -4 -16 -4
0 10 20 30 40 0 10 20 30
Time (s) Time (s)

(a) (b)
Figure 6-24 Input control signal (dashed line) and corresponding actuator velocity (solid line) of
Design 8, for attached mass of: (a) 41 kg; (b) 368 kg.

Energy studies

Based on motion pattern and loading conditions in Figure 6-23, the power consumption of the

proposed circuit is measured by using a wattmeter (EKM Omni-Meter I v.3). Also, the main pump

consumed/generated hydraulic power is calculated for the same experiment and compared to the

measured power. Figure 6-25 (a) shows the measured (solid line) and calculated net hydraulic

power of the main pump (dashed line) in the prescribed experiment. Note that a minimum value

of 1860 W measured by the wattmeter that is consumed to overcome frictions and other losses in

the 37-kW motor/pump unit at pump null position is not shown in this Figure. This minimum value

is considerably high compared to the rated consumed power by the circuit which is attributed to

the oversized motor/pump unit. A 3-kW motor/pump unit would be suitable for this experiment

which roughly consumes about 10% of the aforementioned minimum value (180 W). The

hydraulic power at the pump is obtained by multiplying the pressure differential across the pump

by the flow rate. The average measured consumed electric power and delivered hydraulic power

from the pump to the circuit (pumping modes) are 331.6 W and 279.9 W, respectively. On the

other hand, the average measured generated electric power and received hydraulic power by the

pump from the circuit (motoring modes) are 75.1 W and 207.8 W, respectively. Thus, the net

145
consumed electric power and net delivered hydraulic power from the pump to the circuit are 256.50

W and 72.1 W, respectively. The above-mentioned values become 344.9 W and 160.5 W by adding

the 88.4 W consumed power by the charge pump.

2 2
Pump power (kW)

Pump power (kW)

1 1

0 0

-1 -1

-2 -2
0 7 14 21 28 0 7 14 21 28
Time (s) Time (s)

(a) (b)
Figure 6-25 The power delivered/consumed in Design 8 circuit for experiment in Figure 6-23: (a)
measured electric (solid line) versus calculated hydraulic (dashed line); (b) calculated hydraulic in
the proposed circuit (dashed line) and a comparable load-sensing circuit (solid line).

The pump hydraulic power of a similar valve-controlled load sensing (LS) circuit is calculated

and compared to that of the proposed circuit, as shown in Figure 6-25 (b). The net delivered

hydraulic power from the pump to the circuit is 160.5 W, while in the LS circuit the delivered

power is 795.7 W. The delivered hydraulic power from a LS pump to the circuit has been

calculated as follows: 𝑊𝑝ℎ = [(𝑝𝑎 − 𝑝𝑏 ) + ∆𝑝𝐿𝑆 ]𝑄 where ∆𝑝𝐿𝑆 is the difference between the

supplied pressure and the pressure required by the highest load in the circuit, which typically

ranges from 2 MPa to 3 MPa [16]. Conservatively, this value has been chosen to be as low as 2

MPa. Besides, the studied actuator is assumed typically the one that carries the highest load in a

multi-actuator excavator. Based on these assumptions, the hydraulic power delivered by the pump

to the hydraulic circuit in the proposed circuit is about 20% of that delivered by an LS pump. It is

worth mentioning that the indicated power here is not the total power consumed by the circuit but

146
is the hydraulic power delivered to or received by the pump. To calculate the total consumed

electric power in this experiment, efficiencies of pump and electric motor should be considered.

Summary

A new pump-controlled hydraulic circuit that utilizes one POCV and one LTV as compensating

valves is proposed. The circuit is practically implemented utilizing one POCV and one adjustable

SQV. The circuit is experimentally tested and the performance is analyzed. The velocity and

pressure graphs as well as the pressure planes for the circuit have shown good matching to the

theoretical study results. The experimental results showed the enhanced performance of the

proposed circuit under different loading conditions and during realistic joystick operations. The

energy studies proved the superior efficiency of the proposed circuit when related to a comparable

LS circuit. The calculated hydraulic power delivered by the pump to the actuator in the proposed

circuit, during one representative motion pattern, is about 20% of that delivered by an LS pump.

6.2.4 Evaluation of Designs 2 and 3

Experiments are performed to investigate the effect of shifting the critical zone toward the

center (zero-load axes) on the performance of selected circuits. This concept reduces the load

values where oscillation occurs and consequently reduces the destructive effect of oscillation.

Among the different designs proposed in Section 4.2 that applies the above-mentioned concept,

two are experimentally tested. These are: an implementation of Design 2, circuit with a biased

(offset spring) 3/3 CC-SHV and an implementation of Design 3, circuit with a biased (offset

spring) 3/3 OC-SHV. Currently proposed 3/3 biased shuttle valves are commercially unavailable

in the market. Hence, the circuit is tested utilizing a modified biased version of the same 3/3 CC-

SHV and 3/3 OC-SHV used to evaluate the conventional circuits with SHVs.

147
 The experiments apply variable loading conditions similar to previous circuits. In these tests,

two masses of 41 kg and 368 kg were applied (distinctly) to the end of the stick link to investigate

the circuit performance under low and high loading conditions, respectively. Figure 6-7 shows the

calculated static load at the actuator rod for 41 kg and 368 kg attached masses. Specifications for

the basic components of the test rig and the sequence valve are listed in Tables 3-1 and 6-3,

respectively.

Figure 6-26 shows a breakdown view and a schematic drawing of the modified 3/3 CC-SHV,

where an extra spring is added to the valve side where Line A of the circuit is connected.

Originally, this valve is equipped with one spring that acts against spool in both directions of

motion. The effective pressure difference across the spool (𝑝𝑒 = 𝑝1 − 𝑝2 ) that exceeds the 0.2 MPa

cracking pressure drives the spool motion in the corresponding direction. An extra spring with

2000 N/m stiffness and 6 mm initial compression is added to the side where Line A of the circuit

is connected. This spring creates 0.14 MPa offset in spool operation such that an effective pressure

difference of 0.06 MPa and 0.34 MPa are required to move the spool to the right and left directions,

respectively. Modification performed to the 3/3 OC-SHV is similar to that of the CC-SHV.

As shown in Figure 6-26, in this type of valves, the balancing spring is installed at one side of

the spool in a way that resembles the work of two springs at both sides of the spool. SHV is

modified by adding one extra spring to the free end of the spool that is connected to Line A of the

circuit in order to obtain biased position of the spool. The modified SHV possesses equal spool

piloting areas while it has different stiffness of balancing springs at both sides of the spool with
𝑘𝑠𝐵
ratio that is approximately equal to the actuator area ratio =𝛼=0.75.
𝑘𝑠𝐴

148
 (a) (b)
Figure 6-26 3/3 closed-center shuttle valve; (a) exploited view; (b) schematic drawing.

Experiments are performed on circuits equipped with the modified 3/3 CC-SHV and modified

biased 3/3 OC-SHV, under high loading conditions. Both experiments are performed with 200 psi

charge pressure and square input control signals. Figure 6-27 and Figure 6-28 show the responses

of modified spring-offset CC-SHV and modified OC-SHV, respectively.

In order to highlight the enhanced performance of circuits with modified offset-spring SHVs,

performance is compared to that of circuits with conventional valves. The performance of circuit

equipped with the modified 3/3 CC-SHV shown in Figure 6-27 is compared to that of the circuit

with conventional 3/3 CC-SHV shown in Figure 6-12. Figure 6-27 (a) and Figure 6-12 (a) illustrate

the input control signals to the corresponding circuits, which are manually applied for safety

reasons. Small differences between input signals are attributed to the human factor. Figure 6-27

(b) and (c) and Figure 6-12 (b) and (c) show the actuator velocity and pressure responses for the

circuit with the biased and conventional SHVs, respectively.

It is clear from velocity and pressure graphs that both circuits face damped oscillation in

switching zones S14 and S32 which is attributed to the sharp nature of the input control signal.

Compared to the circuit with conventional CC-SHV, the circuit with biased CC-SHV encounters

less oscillations in undesirable switching zone S43 at times from 4 to 6 seconds and from 16 to 18

seconds. The velocity graphs show that maximum oscillation amplitudes are 1.4 cm/s (from -8 to

149
-9.4 cm/s) and 2.4 cm/s (from -8.3 to -11.1 cm/s) in the proposed and conventional circuits,

respectively. Figures 6-24 (d) and 6-12 (d) illustrate the performance of the circuits with the

modified and conventional CC-SHVs, respectively. The graphs further show smoother transition

between motoring and pumping modes during actuator retraction in the critical zone S43.

6 16
S14 S32 S21
Input signal (v)

3 8

Velocity (cm/s)
0 0
S43
-3 -8

-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)
(a) (b)
9 9
Pressure (MPa)

pb (MPa)

6 6

3 3

0 0
0 5 10 15 20 0 2 4 6 8
Time (s) pa (MPa)
(c) (d)
Figure 6-27 Performance of Design 2, circuit with a biased closed-center shuttle valve, for a 368-
kg attached mass: (a) input signal; (b) actuator velocity; (c) pressures at the pump ports a (solid
line) and b (dotted line); (d) pump pressure plane.

Similarly, the performance of the circuit equipped with the modified 3/3 OC-SHV shown in

Figure 6-28 is compared to that of the circuit with the conventional 3/3 OC-SHV shown in Figure

6-13. Figures 6-28 and 6-13 show the responses of the circuits that utilize the modified and

conventional OC-SHV, respectively. Figures 6-28(a) and 6-13 (a) illustrate the manually applied

control signals to the corresponding circuits. Figures 6-28 (b) and (c) and Figures 6-13 (b) and (c)

show the actuator velocity and pressures of the circuit equipped with the biased and conventional

150
SHVs, respectively. It can be seen from the velocity and pressure graphs that both circuits

experience almost similar oscillations in switching zones S14, S32 and S21. However, it is

observed that the circuit with biased OC-SHV encounters less oscillations in the undesirable

switching zone S43 at times from 4 to 6 seconds and from 14 to 15 seconds.

6 16

Velocity (cm/s)
3 8
Input signal (V)

0 0

-3 -8

-6 -16
0 5 10 15 20 0 5 10 15 20
Time (s) Time (s)
(a) (b)
9 9
Pressure (MPa)

6 6
pb (MPa)

3 3

0 0
0 5 10 15 20 0 2 4 6 8
Time (s) pa (MPa)
(c) (d)
Figure 6-28 performance of Design 3, circuit with a biased open-center shuttle valve, for a 368-kg
attached mass: (a) control signal; (b) actuator velocity; (c) pressures at the pump ports a (solid line)
and b (dotted line); (d) pump pressure plane.

The velocity graphs show that the maximum oscillation amplitudes are 4.6 cm/s (from -6 to -

10.6 cm/s) and 6 cm/s (from -5 to -11 cm/s) in the modified and conventional circuits, respectively.

Figures 6-28 (d) and 6-13 (d) illustrate performance, on the pressure plane, of the circuits with the

modified and conventional CC-SHVs, respectively. The graphs further show smoother transition

between motoring and pumping modes in the proposed circuit during actuator retraction in the

undesirable switching zone S43.

151
 Summary

In this section, experimental work is done to validate a new concept to deal with the oscillations

issue of pump-controlled hydraulic circuits. Applying this concept, the critical zone is shifted

towards less loading values by utilizing asymmetric compensating valves. Two circuits that utilize

modified spring-offset (biased) CC-SHV and spring-offset (biased) OC-SHV are prototyped and

tested. Experimental results of the circuits with modified CC-SHV and OC-SHV showed 40% and

25% less velocity oscillation amplitudes when compared to the conventional circuits, respectively.

Consequently, the proposed concept reduces undesirable effect of oscillations on the machine and

facilitaes the application of other solutions.

152
 CHAPTER 7

7. CONCLUSIONS

7.1 Contributions of this thesis

Pump-controlled hydraulic actuation, as one promising solution to improve the efficiency of

hydraulic circuits, was investigated in this thesis. The drawbacks of existing pump-controlled

circuits for single-rod cylinders were examined and concepts to alleviate their effect were

proposed. Eight new designs were proposed and evaluated in simulations. Five designs were tested

in experiment. The experimental evaluations proved the enhanced performance and high efficiency

of the new designs. Details are highlighted below.

Currently, the existing pump-controlled solutions for single-rod cylinders exhibits undesirable

performance under certain operating conditions. Three previously-designed circuits to control

single-rod cylinders were thoroughly analyzed and their performances were evaluated. These

designs are the circuits that utilize two pilot-operated check valves (POCVs), a closed-center

shuttle valve (CC-SHV) and an open-center shuttle valve (OC-SHV). Firstly, a detailed

explanation of operation in the four quadrants and some proposed reasons for the undesirable

performance were offered. The critical zone is generated when the pump switches its mode of

153
operation under specific actuator assistive retraction conditions. The developed mathematical

models and simulation programs identified the position of such critical zone. The theoretical

studies have illustrated how transmission lines losses, flow compensating valves characteristics,

and frictional forces contribute to the location and shape of the critical zone. Performance

comparisons between the three circuits in terms of critical zone area and characteristics of

oscillations were made. It was shown that the circuit that utilizes POCVs possesses a smaller

oscillatory zone compared to the circuits with CC-SHV and OC-SHV. However, the circuit with

OC-SHV has a less oscillatory zone compared to the circuit utilizing CC-SHV and smaller velocity

oscillation amplitudes as compared to the circuit with POCVs.

Experimental work was performed to validate both theoretical analysis and simulation results

of the previously-designed circuits. A test rig that allows testing different circuits in the four

quadrants of operations was designed and constructed. The previously-designed circuits with

POCVs, CC-SHV and OC-SHV were experimentally tested and their performances were

evaluated. The experimental work verified the theoretical findings and proved the accuracy of the

mathematical models. Generally, circuit with POCVs showed better performance than circuit with

CC-SHV at all loading conditions and better performance than circuit with OC-SHV at high

loading conditions.

Based on the theoretical studies, requirements for desirable operation of pump-controlled

circuits were provided. Moreover, three concepts to improve the performance of pump-controlled

circuits were proposed; namely: (1) applying selective throttling, (2) applying pre-designed

leakage, and (3) shifting of the critical zone into lower loading ranges. Simulation studies indicated

that the new concepts alleviated the oscillation issue of the common pump-controlled circuits, and

improved their performance. The first and second concepts, in particular, were capable of

154
eliminating the whole critical zone. The third concept reduced the area of the critical zone in the

load-velocity plane and lessened undesirable effect of oscillations. Simulation studies further

proved the enhanced performance of circuits that applies these concepts. By applying the above-

mentioned concepts, whether individually or collectively, eight new pump-controlled circuits for

single-rod cylinders are proposed and their operations are explained.

Five proposed circuits were assembled and experimentally tested to prove the practicality and

enhanced performance of the new designs. The circuits were prototyped using off-the-shelf

commercial components. The tested circuits were: (1) Design 5, circuit that utilizes POCVs for

flow compensation and counterbalance valves (CBVs) to apply selective throttling to flow in the

main transmission lines, (2) Design 7, circuit that utilizes sequence valves (SQVs) to apply

selective throttling to flow in the compensating transmission lines, (3) Design 8, circuit that utilizes

one POCV and one SQV to apply selective throttling to flow in one of the compensating

transmission lines and (4), Designs 2 and 3, circuits that utilize biased shuttle valves to shift critical

zone towards lower loading areas.

Experimental work has validated theoretical analysis and has shown the improved performance

of all tested new designs. Energy studies proved the superior efficiency of all new circuits

compared to that of the conventional valve-controlled circuits. Designs 2 and 3 that apply the

concept of shifting of the critical zone into lower loading ranges showed reasonable performance

improvements. Designs 5, 7 and 8 that utilize the concept of applying selective throttling showed

similar performance improvements. Designs 7 and 8 implemented the concept of selective

throttling in simple and cost-effective designs. They showed better efficiency than Design 5 since

they apply throttling over less flow. Designs 7 and 8 displayed superior efficiency compared to

load-sensing (LS) conventional pump-controlled circuits. More specifically, the hydraulic power

155
delivered by the pump to the actuator in Designs 7 and 8 was calculated from experimental data

and was found to be about 23% and 20% of that delivered by a comparable LS pump, given the

same task. In general, we conclude that, new proposed circuits that apply the selective throttling

concept consumed about 10% extra energy when compared to previously-designed pump-

controlled circuits; this amount of energy is used to stabilize the system during critical operating

conditions. On the other hand, the new circuits with energy recuperation-ability consumed only

20-23% of energy consumed by a comparable load-sensing valve-controlled circuit.

7.2 Recommendations and future work

All tested designs are reliable and easy to implement as they prototyped of conventional

commonly-used hydraulic components with well-established technologies. Simulation and

experimental studies of Designs 7 and 8 showed acceptable performance and superior efficiency

as compared to valve-controlled hydraulic circuits. Compared to other proposed designs they are

more cost effective and easier to be implemented, as they contain less components. Previously-

designed pump-controlled circuits, presently, in service can be easily modified to new designs by

replacing the old compensating valves with the new proposed ones.

Testing of Designs 4 and 6 was not performed in this work due to the unavailability of the newly

proposed selective throttling valves. Detailed investigation of the proposed selective throttling

valves is recommended as a future work to allow prototyping and testing of Designs 4 and 6.

Designs with the proper selective throttling valves are expected to show better performance than

that of the comparable designs with available alternative valves. Generally, hydraulic systems are

favored in moderate and high-power applications (P ≥ 2 kW). Hydraulic actuation systems with

improved efficiency are able to compete with other power transmission systems in low power

156
applications. Thus, design of a compact pump-controlled actuator for relatively low power

applications is a promising future work.

157
 References

[1] G. F. Ritelli and A. Vacca, "Energetic and dynamic impact of counterbalance valves in
fluid power machines," Energy Conversion and Management, vol. 76, pp. 701-711,
2013.
[2] J. D. Zimmerman, M. Pelosi, C. A. Williamson and M. Ivantysynova, "Energy
Consumption of an LS Excavator Hydraulic System," in Proceedings of ASME
International Mechanical Engineering Congress and Exposition (IMECE), Seattle,
USA, 2007.
[3] A. A. Aly, F. A. Salem and T. O. S. Hanafy, "Energy Saving Strategies of an Efficient
Electro-Hydraulic Circuit (A review)," International Journal of Control, Automation
and Systems, vol. 3, no. 3, pp. 6-10, July 2014.
[4] K. G. Cleasby and A. R. Plummer, "A novel high efficiency electrohydrostatic flight
simulator motion system," in Proceedings of Fluid Power and Motion Control, Bath,
England, 2008.
[5] E. Jalayeri, A. Imam and N. Sepehri, "A Throttle-less Single Rod Hydraulic Cylinder
Positioning System for Switching Loads," Case Studies in Mechanical Systems and
Signal Processing, vol. 1, pp. 1-5, 2015.
[6] J. Li, Y. Fu, Z. Wang and G. Zhang, "Research on fast response and high accuracy
control of an airborne electro hydrostatic actuation system," in Proceedings of
International Conference on Intelligence, Mechatronics and Automotion, Changdu,
China, 2004.
[7] Z. Quan, L. Quan and J. Zhang, "Review of energy efficient direct pump controlled
cylinder electro-hydraulictechnology," Renewable and Sustainable Energy Reviews, vol.
35, pp. 336-346, 2014.
[8] J. A. Hewett, "Hydraulic Circuit Flow Control". US Patent 5 329 767, 19 July 1994.
[9] R. Rahmfeld and M. Ivantysynova, "Developing and control of energy saving hydraulic
servo drives," in Proceedings of Fluid Power Net International Symposium (FPNI-PhD),
Hamburg, Germany, 2000.

158
[10] K. Heybroek, J. Palmberg and J. Larsson, "Open circuit solution for pump controlled
actuators," in Proceedings of Fluid Power Net International Symposium (FPNI-PhD),
Sarasota, USA, 2006.
[11] L. Wang, W. J. Book and J. D. Huggins, "A hydraulic circuit for single rod cylinder,"
Journal of Dynamic Systems, Measurement, and Control, ASME, vol. 134(1), pp.
011019-1-16, 2012.
[12] Jalayeri, Ehsan; Imam, Ahmed ; Tomas, Zeljko; Sepehri, Nariman, "A Throttle-less
Single-rod Hydraulic Cylinder Positioning System: Design and Experimental
Evaluation," Advances in Mechanical Engineering, vol. 7, no. 5, pp. 1-14, 2015.
[13] H. Caliskan, T. Balkan and E. B. Platin, "A Complete Analysis and a Novel Solution for
Instability in Pump Controlled Asymmetric Actuators," Journal of Dynamic Systems,
Measurement, and Control, vol. 137, no. 1, pp. 1-14, 2015.
[14] S. Habibi., "Design of a new high-performance electrohydraulic actuator," IEEE/ASME
Transactions on Mechatronics, vol. 5 , no. 2, pp. 158-164, 2000.
[15] H. E. Merritt, Hydraulic Control Systems, New York: John Wily and Sons, 1967.
[16] M. Scherer, M. Geimer and B. Weiss, "Contribution on Control Strategies of Flow-On-
Demand Hydraulic Circuits," in The 13th Scandinavian International Conference on
Fluid Power, SICFP2013, Linköping, Sweden, June 3-5, 2013.
[17] K. Kang, M. Pachteri and C. Houpis, "Modeling and control of an electro-hydrostatic
actuator," in Proceedings of Aerospace and Electronics Conference,, Dayton, USA,
1995.
[18] I. 22072, "Electrohydrostatic actuator (EHA)," [Online]. Available:
https://www.iso.org/obp/ui/#iso:std:iso:22072:ed-2:v1:en. [Accessed 1 May 2020].
[19] Z. Quan, L. Quan and Zhang.Jinman, "Development of a dual-acting axial piston pump
for displacement-controlled system," Proceedings of Institution of Mechanical
Engineers, Part B: J. Eng. Manuf., vol. 228 (4), pp. 606-616, 2014.
[20] J. Lodewyks, "Differential cylinder in a closed hydrostatic transmission," PhD Thesis,
RWTH Aachen, Germany, 1994.

159
[21] J. Lillemets, M. Lugnberg and M. Ousbäck, "Evaluating a Pump Controlled Open Circuit
Solution," in Proceedings of International Exposition for Power Transmission, IFPE,
Nevada, USA, 2008.
[22] G. Costa and N. Sepehri, Hydrostatic Transmissions and Actuators: Operation,
Modelling and Applications, Chennai: John Wiley & Sons, 2015.
[23] B. Trinkel, "hydraulic reservoirs," [Online]. Available:
http://hydraulicspneumatics.com/other-technologies/chapter-6. [Accessed 1 May 2020].
[24] R. Hippalgaonkar and M. Ivantysynova, "A Series-Parallel Hydraulic Hybrid Mini-
Excavator with Displacement Controlled Actuators," in Proceedings of 13th
Scandinavian International Conference on Fluid Power, SICFP2013,, Linköping,
Sweden, 2013.
[25] J. Grabbel and M. Ivantysynova, "Model Adaptation for Robust Control Design of
Hydraulic Joint Servo Actuators," in Proceedings of 4th International Symposium on
Fluid Power Transmission and Control (ICFP 2003), Wuhan, China, 2003.
[26] R. Rahmfeld and M. Ivantysynova, "Displacement Controlled Linear Actuator with
Differential Cylinder-a Way to Save Primary Energy in mobile machines," in
Proceedings of 5th International Conference on Fluid Power Transmission and Control
(ICFP 2001), Hangzhou, China, 2001.
[27] R. Rahmfeld, "Development and Control of Energy Saving Hydraulic Servo Drives for
Mobile Machimes," PhD Thesis, TUHH, Germany, 2002.
[28] C. Williamson and M. Ivantysynova, "Pump Mode Prediction for Fourquadrant Velocity
Control of Valveless Hydraulic Actuators," in Proceedings of 7th JFPS International
Symposium on Fluid Power, Toyama, Japan, 2008.
[29] S. Yuming, A. Imam, C. Wu and N. Sepehri, "Stability Study of a Pump-Controlled
Circuit for Single Rod Cylinders Via The Concept of Lyapunov Exponents," in
ASME/BATH 2017 Symposium on Fluid Power and Motion Control, Sarasota, Florida,
USA, 2017.

160
[30] G. Altare and A. Vacca, "Design Solution for Efficient and Compact Electro-hydraulic
Actuators," in Proceedings od Dynamics and Vibroacoustics of Machines (DVM2014),
Samara, Russia, 2014.
[31] G. Altare, A. Vacca and C. Richter, "A Novel Pump Design for an Efficient and Compact
Electro-Hydraulic Actuator," in Proceedings of IEEE Aerospace Conference, Montana,
USA, 2014.
[32] MathWorks, "Cylinder Friction," [Online]. Available:
https://www.mathworks.com/help/physmod/hydro/ref/cylinderfriction.html. [Accessed
1 May 2020].
[33] C. Williamson, "Power Management for Multi-Actuator Mobile Machines with
Displacement Controlled Hydraulic," PhD Thesis, University of Purdue, USA, 2010.
[34] J. D. Zimmerman, "Toward Optimal Multi-actuator Displacement Controlled Mobile
Hydraulic Systems," PhD Thesis, University of Purdue , USA, 2012.
[35] A. Imam, M. Rafiq, E. Jalayeri and N. Sepehri, "Design, implementation and evaluation
of a pump-controlled circuit for single rod actuators," Actuators, vol. 6, no. 10, pp. 1-15,
2017.
[36] A. Imam, M. Rafiq, T. Zeljko and N. Sepehri, "Improving the Performance of Pump-
Controlled Circuits for Single-Rod Actuator," Actuators, vol. 8, no. 2, pp. 1-18, 2019.
[37] L. Wang and W. Book, "Using Leakage to Stabilize Hydraulic Circuits for Pump
Controlled Actuators," Journal of Dynamic Systems, Measurement, and Control, vol.
123, no. 1, pp. 061007-1-12, 2013.
[38] E. Engineers, "Overcenter Valves," [Online]. Available:
http://www.eaton.com/ecm/groups/public/@pub/@eaton/@hyd/documents/content/pct
_273380.pdf. [Accessed 1 May 2020].
[39] A. Imam, M. Rafiq, E. Jalayeri and N. Sepehri, "A Pump-Controlled Circuit for Single-
Rod Cylinders that Incorporates Limited Throttling Compensating Valves," Actuators,
vol. 7, no. 13, pp. 1-21, 2018.

161