Documente Academic
Documente Profesional
Documente Cultură
PATH STRUCTURE
FITRIAN IMADUDDIN
Replace this page with form PSZ 19:16 (Pind. 1/07), which can be
obtained from SPS or your faculty.
Signature
Name
Date
:
:
:
Signature
Name
Date
:
:
:
Replace this page with the Cooperation Declaration form, which can be
obtained from SPS or your faculty.
FITRIAN IMADUDDIN
JUNE 2015
ii
I declare that this thesis entitled A Novel Magnetorheological Valve with Meandering
Flow Path Structure is the result of my own research except as cited in the references.
The thesis has not been accepted for any degree and is not concurrently submitted in
candidature of any other degree.
Signature
Name
Date
:
:
:
Fitrian Imaduddin
iii
iv
ACKNOWLEDGEMENT
I wish to praise the Almighty Allah the Most Gracious for the blessing and
strength that have been given to my life. My deepest gratitude goes firstly to my main
supervisor, Associate Professor, Ir. Dr. Hj. Saiful Amri bin Mazlan for his tremendous
support during my entire study. His intensive encouragement, enthusiasm and guidance
have made me able to pass through this process easier. I also would like to thank
my co-supervisor Dr. Hairi bin Zamzuri for his advice, interest and support to my
work. I must also express my gratitude to the Malaysia-Japan International Institute
of Technology (MJIIT) for the financial support provided during my study though the
MJIIT scholarship. I also would like to thank the Lembaga Pengelola Dana Pendidikan
(LPDP) for providing the additional incentive during the completion of my thesis
writing.
Appreciation is also given to the faculty members and colleagues in the Vehicle
System Engineering (VSE) research laboratory, especially Jamal, Yasser and Izyan
for helping me a lot during my earlier time in the university. Thanks also to Mr.
Hairullail, Madam Aishah and the remaining MJIIT staffs that have been supportive to
me during my study. Particular credit is also given to my Indonesian friends in UTM
Kuala Lumpur especially Ubaidillah and Burhanuddin for being my family abroad. I
would also like to thank Aizzat and the developers of the utmthesis LATEX project for
making the thesis writing process a lot easier for me. Special acknowledgement goes to
my previous supervisors Dr. Khisbullah Hudha and Dr. Gunawan Nugroho. They have
influenced me with the passion and love to the scientific research. I regret that I cannot
mention all the valuable names here, but I believe and pray that Allah will reward all
the good deeds that have been given to me.
ABSTRACT
vi
ABSTRAK
vii
TABLE OF CONTENTS
CHAPTER
TITLE
DECLARATION
DEDICATION
ACKNOWLEDGEMENT
ABSTRACT
ABSTRAK
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES
LIST OF APPENDIX
PAGE
ii
iii
iv
v
vi
vii
x
xi
xiv
INTRODUCTION
1.1
Introduction
1.2
Motivation of Study
1.3
Research Objectives
1.4
Research Scope
1.5
Significance of Research
1.6
Outline of Thesis
1
1
2
6
6
7
7
LITERATURE REVIEW
2.1
Introduction
2.2
Magnetorheological Fluid
2.2.1
Composition of Magnetorheological
Fluid
2.2.2
Operational Modes of Magnetorheological Fluid
2.3
Magnetorheological Valve
2.3.1
Annular Magnetorheological Valve
2.3.2
Radial Magnetorheological Valve
9
9
9
10
11
14
15
18
viii
2.3.3
2.4
2.5
2.6
2.7
20
20
22
23
27
30
31
35
39
41
42
42
42
43
45
47
EXPERIMENTAL ASSESSMENT
4.1
Introduction
4.2
Experimental Apparatus
4.2.1
Magnetorheological Fluid
4.2.2
Magnetorheological Valve
4.2.3
Testing cell
4.3
Experimental Set-up
4.4
Experimental Results
4.4.1
Off-state and On-state Pressure Drop
Characteristics
4.4.2
Effect of Gap Size
4.4.3
Effect of Current Input Variation
62
62
62
62
64
67
68
71
50
54
61
71
75
76
ix
4.5
REFERENCES
Appendix A
4.4.4
Effect of Excitation Frequency Variation
Summary of Chapter 4
79
80
82
82
82
87
90
95
96
96
96
97
97
98
99
100
100
103
105
120 129
LIST OF TABLES
TABLE NO.
3.1
3.2
3.3
4.1
4.2
5.1
5.2
5.3
5.4
5.5
TITLE
Materials selection of valve component for valve routing
List of MR valve parameter
Performance benchmarking between the proposed MR valve
concept and the counterparts
Typical properties and material compatibility of MRF-132DG
The variable arrangement of experimental test using
Shimadzu Fatigue Dynamic Test Machine
Correlation test results between the model coefficient a and
current input I
List of coefficients for the polynomial-based parametric MR
valve model
List of approximated function for different parameters
Comparison of relative error at 0.75 Hz frequency excitation
Comparison of relative error at 0.50 Hz and 1.00 Hz
frequency excitations
PAGE
46
53
60
63
71
85
87
90
91
94
xi
LIST OF FIGURES
FIGURE NO.
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
2.11
2.12
2.13
2.14
2.15
2.16
2.17
2.18
2.19
2.20
2.21
2.22
2.23
2.24
2.25
2.26
2.27
2.28
2.29
TITLE
Movement of magnetic particles in the MR fluid with and
without magnetic field
Shear mode
Valve mode
Squeeze mode
Magnetic Gradient Pinch mode
MR throttle valve
C-shaped pressure control valve
Three port MR valve
Double-coil annular MR valve
Basic structure of single stage radial MR valve
Two-way controllable radial MR valve
Annular-Radial MR Valve
Typical arrangement of constant flow assessment method
Typical arrangement of variable flow assessment method
Illustration of significant variables in an MR valve
Bouc-Wen model
Parametric hysteretic polynomial model
Artificial Neural Network model
Valve mode MR damper
Shear mode MR damper
External coil MR damper
MR damper with bifold valves
Bifold MR damper for high impulsive loads
Bifold MR damper for shock vibration mitigation
Basic structure of Bypass MR damper
Bypass MR damper for large scale seismic application
Vane type MR damper with arc valve
Vane type MR damper with outer coil valve
Vane type MR damper with inner coil valve
PAGE
10
12
13
13
14
16
16
17
18
19
19
20
21
22
25
29
30
31
32
32
33
33
34
34
35
35
36
37
38
xii
2.30
2.31
2.32
2.33
2.34
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
39
39
40
40
41
44
45
48
48
49
50
52
54
55
55
57
59
63
64
65
65
66
67
69
70
71
xiii
4.10
4.11
4.12
4.13
4.14
4.15
4.16
5.1
5.2
5.3
5.4
5.5
5.6
6.1
6.2
72
73
76
77
78
79
80
83
85
86
90
92
93
100
102
xiv
LIST OF APPENDIX
APPENDIX
A
TITLE
CAD Drawings
PAGE
120
CHAPTER 1
INTRODUCTION
1.1
Introduction
One of the most popular devices that utilized the unique characteristics of
MR fluid is MR damper [13], which has been commercially available for highend passenger vehicles as a semi-active suspension or adjustable suspension [14].
The working principle of an MR damper is basically similar to a conventional
viscous damper which employs flow restriction concept to generate damping. The
flow restriction in a conventional viscous damper is normally generated by an orifice
channel which act as the valve. Since the gap of the orifice channel is fixed, the flow
restriction that can be generated by the valve of the conventional viscous damper is
also fixed. The MR dampers use different approach by employing MR fluids as its
working fluid and an MR valve in its flow restriction mechanism. Although the gap
size of the channel in an MR valve also can be fixed, the magnetic field strength in the
flow channel of the MR valve can be regulated [15]. Therefore, the flow of MR fluid
2
that pass through the MR valve can be controlled without having to modify the gap
size of the channel. On the other hand, it can be said that the performance of the MR
valve to generate flow restriction highly determines the overall performance of the MR
damper.
1.2
Motivation of Study
MR damper for semi-active vehicle suspension systems are among the most
popular and commercially successful MR fluid devices [2834]. In general, vehicle
suspension system can be divided into three categories; passive suspension system,
semi-active suspension system, and active suspension system [35]. Passive suspension
3
system is the common suspension system installed in most vehicle nowadays
which typically consists of spring and damper in parallel configuration. Semi-active
suspension system is similar with passive suspension system but the stiffness of the
component (spring and/or damper) can be controlled to suit the desired ride or handling
performance [36, 37]. Active suspension system, on the other hand, is the suspension
system with the involvement of active actuators such as hydraulic [38], pneumatic [39]
or electro-mechanic [40,41], which could provide external force to the suspension. MR
dampers are usually implemented as a semi-active device to retrofit hydraulic dampers
to enhance passive suspension performance. Enhancement of suspension performance
is feasible since the performance limitations of passive suspension system occurred due
to a fixed stiffness value of the spring and damper. In this case, MR damper, in contrast
to conventional linear hydraulic damper, has the capability to change its damping
stiffness by varying the magnetic field strength inside the damper. Together with
embedded control system, MR dampers have gained popularity and proved its potential
to enhance the performance of suspension systems. Aside of dampers, other types of
MR devices have been developed to meet other automotive application demands such
as engine vibration suppressors [4245], seat suspensions [4649], brakes [5053] and
clutches [5457].
According to the location of the valve, the MR damper can be divided into two
groups, the MR damper with internal valve and the MR damper with bypass valve.
The MR damper with internal valve typically has MR valve embedded in the piston
of the damper, similarly with the configuration of the valve in a conventional viscous
damper. This configuration is the most common valve installation in an MR damper
since it is neat and compact. However, the internal valve configuration is not without
setback. The disadvantages of internal valve configurations are mainly in the space
limitation of MR valve installation, the complexity of wiring and in the risk of thermal
build-up from the immersed valve. The MR valve integration to the piston is the main
reason why the construction of the MR damper with internal valve can be really neat
and compact. However, since the available space inside the cylinder is very limited
and the MR valve requires sufficient space for electromagnetic coil and magnetization
channel, the performance range of the damper is very narrow. Moreover, since the
coil is embedded with the piston, the common way of wiring installation is normally
made through the conduit along the rod, which made it prone to leakages and tends
to be costly for fabrication. On the other hand, the heat dissipation, as a result of
kinetic energy conversion into heat, can be more severe in an MR damper than in a
conventional viscous damper because the magnetically altered damping stiffness will
definitely increase the heat dissipation. In the case where the MR valve is immersed in
the MR fluid, the heat dissipation from the MR valve will have to disperse to the MR
4
fluid first, which responsible in the increase of fluid temperature, before eventually
released to the environment. The experimental observation conducted by [58] reported
that the temperature rise of MR fluid in an MR damper after 400 s of operation at
current input of 2 A and frequency excitation of 6 Hz have caused the damping force to
degrade in about 38%. However, the same experiment observed that less degradation
can be achieved if the MR damper is properly finned, whereas increase the thermal
release to the environment.
5
in the conventional annular MR valve is very low because not all areas of the annular
channel can be utilized as the effective area. Therefore, any improvement effort on
the annular valve performance will typically tend to increase the valve size either in
length, by enlarging the effective area, or in diameter, by enlarging the electromagnetic
coil. Thus, in a constrained space device such as in the MR damper with internal valve
configuration, the desired performance improvements are sometimes difficult to be
achieved.
Due to the limitation of the annular MR valve, another type of valve, known as
the radial MR valve, was introduced by [21]. The radial MR valve, as a distinction from
annular MR valve, has radial flow channel inside the valve and utilize it as the effective
area. The utilization of radial channel as the effective area offers several benefits than
the effective area of the annular channel, especially in terms of area efficiency since
the radial channel can be made in multi-stage configuration. Therefore, with multistage capability of the radial MR valve, the performance improvement of radial valve
typically has lower implication to the valve size than the one in the annular MR valve.
As a result, the radial valve concept has been installed to serve several concepts of
large scale MR dampers [61, 62, 70, 71]. Recently, another concept of MR valve also
has been developed by combining both annular and radial valve concept in a single
valve [23, 24]. The combination of both annular and radial channel in an MR valve has
been proven effective to improve the performance of MR valve. It has been reported
by [72] that the MR valve with combination of annular and radial channel has higher
achievable pressure drop than annular valve with lower power consumption although
at the cost of lower valve ratio. The MR valve with combination of annular and radial
channel also has been implemented in MR mount design [42] and MR damper design
[73].
6
hysteresis behavior is a disadvantage since the controller will face difficulties to track
the damper behavior. For example, according to Wang and Liao [74], tracking ability of
damping force is one of the highly important issues that should be considered in order
to get an accurate MR damper controller. However, a controller with such capability
will tend to be more complex, require more computational resources, be costly and
less robust. Therefore, innovation in the control design is also vital to support the
final implementation of MR devices. Innovation of the control algorithm will be more
difficult if the model that is used in the controller design phase is not able to simulate
the hysteresis phenomenon accurately. A simple and accurate model of an MR valve, in
particular, is needed in order to design an appropriate controller with good robustness,
stability and reliability. Therefore, the advancement of modeling technique that have
the ability to accommodate the hysteretic behavior of MR valve is as important as the
advancement of the MR valve concept.
1.3
Research Objectives
(a)
(b)
To analyze the effect of gap size selection to the achievable pressure drop of MR
valve.
(c)
(d)
1.4
Research Scope
7
package. The performance of MR valve, in this study, is only evaluated in terms of
the achievable pressure drop as a function of gap size of the flow channel, magnitude
of current input charged to the coil, and fluid flow rate. This research is also covering
the experimental evaluation of the MR valve using an MR valve testing cell in variable
flow rates, to reveal the hysteretic behavior, with constant current inputs. The measured
performance of the MR valve is also used to model the hysteretic behavior of the MR
valve, which is not covered in the steady-state model. However, the optimization of the
concept is not discussed in this research and the demonstration of control application
is only performed.
1.5
Significance of Research
(a)
(b)
This research provides knowledge of the effect of gap size selection to the
achievable pressure drop of the valve which will be further useful for valve sizing
process.
(c)
1.6
Outline of Thesis
8
established in the chapter. The outline of this thesis is organized as shown:
Chapter 3 explains the new concept of the MR valve with meandering flow
path, the design consideration for the performance assessment, the steady-state model
derivation, the magnetic simulation as well as the performance prediction of the new
MR valve with respect to various dependent variables.
Chapter 6 concludes the work and presents the achieved contribution of the
research as well as recommends open problems for future work.
CHAPTER 2
LITERATURE REVIEW
2.1
Introduction
2.2
Magnetorheological Fluid
The MR fluid was first discovered in the Rabinows magnetic fluid clutch in
1948 [11] and has been classified as one of the class of smart materials [9]. MR fluid is
classified as smart material because of its adaptive behavior and their inherent ability to
provide a simple, fast and robust interface between electronic controls and mechanical
components. When the fluid is subjected to magnetic field, the iron particles start
aligning along the magnetic fluxes, as shown in Figure 2.1. Hence, the movement
of the fluid is restricted by particle chains thus increasing its apparent viscosity.
Consequently, the fluid changes its state from liquid to a viscoelastic solid depending
on the strength of the magnetic field. The advantages of MR fluid have created great
interest in MR based device development in a wide range of applications such as in
civil applications [60, 61, 63, 7580] medical prosthetic applications [8184], ground
vehicle applications [28,30,31,34,55,85], aerospace applications [8688] and military
10
applications [33, 8991]. Since the rheological properties of MR fluid can be modified
magnetically, an MR fluid based device requires no moving parts and therefore less
vulnerable to wear and tear.
Figure 2.1 Movement of magnetic particles in the MR fluid with and without magnetic
field
2.2.1
11
approach of MR fluid synthesis which are mixing non-magnetizable particles dispersed
in magnetizable liquid (ferrofluid) that is known as the inverse ferrofluid [98] and the
Carbonnanotube (CNT)-magnetite based MR fluid [99].
MR fluid are often distincted from ferrofluid from its particle size. Ferrofluid,
similarly with MR fluid, is also a class of field sensitive fluid. However, the particle
size of MR fluid are often bigger (few microns) than the nano-sized particle ferrofluid
because it seek larger yield stress due to stonger dipole moment between the particles.
The yield stress of MR fluid represents the peak value of the stress versus strain curve,
since the gel-like structure will break when the stress reached this peak value [1].
However, since the particle size of the MR fluid is relatively large, it is susceptible to
settling and often caused unwanted abrasion [100]. The ferrofluid is less prone settling
because the nano-sized particles exhibit Brownian motion that keeps them dispersed
in the carrier fluid. The larger the size of the particle, the lesser the effect of Brownian
motion and the faster the tendency of settling. According to Bell et al. [101], the MR
fluids with particle size above 10 microns tend to settle in less than 20 minutes. That
is why, in order to inhibit sedimentation and particle aggregation, some approaches
such as the addition of thixotropic agents and surfactants [102, 103] or the attempt of
non-spherical particle shape implementations [104,105] were often given during to the
fluid synthesis process.
2.2.2
There are three known basic operational modes in any employment of MR fluid
in a device, shear mode, flow mode and squeeze mode [15, 92, 106, 107]. Recently,
another working mode was proposed by Goncalves and Carlson in [108], called the
magnetic gradient pinch mode. However, the working modes of MR fluid are not
strictly limited to these four since the MR devices can be operated in a combination of
these modes [109114].
12
Shear mode, also known as the clutch mode, occurs when MR fluid is exposed
to magnetic field between two parallel magnetic surfaces. One of the surfaces is
moving whilst the other is fixed, as shown in Figure 2.2. The fluid shear area where
MR fluid is exposed to magnetic field is called the effective area of shear mode. Shear
mode is commonly used in brakes and clutches but also appears in some particular
design of dampers, called the shear mode damper.
Flow mode, also known as the valve mode occurs, when MR fluid is exposed
to a magnetic field whilst the fluid flows between two fixed parallel magnetic surfaces.
The concept of flow mode is depicted in Figure 2.3. Similarly with the shear mode,
flow mode has the effective area, which is defined as the area where the flowing MR
fluid is exposed to magnetic field. Flow mode is commonly used in dampers and in
other applications in which the devices requires a valve to control fluid flow.
Despite the concept of squeeze mode has been used in several early rotor
vibration damper designs, the fundamental principle of this mode in MR fluid was
not discussed until later in several publications [115118]. Squeeze mode occurs when
MR fluid is exposed to magnetic field while at the same time being compressed or
decompressed, as shown in Figure 2.4. Since it involves very small displacement most
of its application appears in vibration suppression devices.
The latest working mode is called the magnetic gradient pinch mode. The
basic idea of this mode is similar to flow mode but with a different configuration of
13
magnetic circuit design. In the magnetic gradient pinch mode, as shown in Figure
2.5, the magnetic poles are arranged axially along the flow path and separated by a
non-magnetic material. This kind of poles arrangement will create elliptical magnetic
fibrils, which will block the flow of MR fluid in the valve gap. One of the unique
characteristics of this mode is the slope between pressure and velocity relationship in
magnetic gradient pinch mode will be significantly increased when the magnetic field
is increased [108]. This is unique, since in the conventional flow mode the slope tends
to remain constant under any magnetic field strength modifications. Another advantage
14
is the possibility to use MR fluid with coarser particles up to 100 microns, since a larger
orifice is feasible to be used with the magnetic gradient pinch mode [108].
2.3
Magnetorheological Valve
Valve is the key components in almost any flow control mechanism. The
common practice in hydraulic system usually combines solenoid actuation valve
electronic devices in order to regulate fluid flow. However, conventional hydraulic
valve consists of several moving parts inside the valve, which makes it less responsive
while at the same time more vulnerable to wear and tear. Therefore, an MR fluid based
valve was introduced in order to improve the performance of valve. Moreover, since
working principles of most MR devices are based on the manipulation of fluid flow
rate, the key performance of MR device is determined by the performance of MR valve.
A successful improvement of MR valve performance can give a significant impact to
the development of other MR devices.
15
Considering the importance of MR valve, many designs of MR valve have
been proposed. According to the structure of its flow path, MR valve can be divided
into three categories, the annular type MR valve, the radial type MR valve and the
combination of annular and radial MR valve.
2.3.1
The annular MR valve can be easily distincted as the MR valve with effective
area in the annular channel. Since the effective area of the valve mode is defined as
the area where the fluxes cross the fluid in perpendicular to the flow direction, the
fluxes in annular MR valve is always directed in radial direction. The simplicity of
mechanical structure of the annular MR valve has made it the most common MR valve
type employed in various MR devices. However, the length of annular channel that can
be exploited as the effective area is limited and consequently the magnetorheological
effect that can be generated is relatively narrow.
Another concept of annular MR valve was proposed by Yokota et al. [18]. The
design employed the very basic principles of valve mode using a C-shaped magnetic
16
core attached to the flow channel as shown in Figure 2.7. The C-shaped magnetic
core act as the main flux guiders which conducted the magnetic flux loops to induce
the MR fluid flows in perpendicular direction. Although the experimental results were
only reported around 0.7 MPa of achievable pressure drop at similar flow rate and field
intensity with the previous design, the design supposed to be very efficient since flux
loss can be prevented to a very minimum value. However, the bulkiness of the design
would quite inconvenience for application and compared with the previous design,
this design is only equipped two flow port which made it only suitable for serial flow
connection.
The work of Yokota et al. [18], was improved by Yoshida et al. [19] through the
17
introduction of a three-port annular MR valve as depicted in Figure 2.8. The three-port
annular MR valve design comprises of serially connected two-port MR valves similarly
with the one in Yokota et al. [18] concept. However, one supplementary port is added
in between these two-port to ease the application that requires parallel flow connection.
A NdFeB permanent magnet is added in between both coil to create a magnetic offset
as well as to regulate the direction of the fluid.
18
2.3.2
19
magnetic flux to the radial gaps so that the off-state condition already has pre-yield
value. The two-way condition of the valve defined the ability of the valve to reduce the
pressure drop as well as to enhance the pressure drop depending on the resultant of the
flux directions. The presented design utilized the manipulation of flux similarly with
the one presented by Yoshida et al. [19]. The combination of permanent magnet and
electromagnet, which knows as the hybrid magnetic circuit, also provide additional
advantages such as the possibility to enhance the on-state performance, to reverse the
magnetorheological effect as well as to provide fail-safe features during power failure
to the electromagnet [120].
20
2.3.3
2.4
21
design. The characteristics of MR valve pressure drop are typically evaluated in terms
of response to the variation of flow rate and response to the variation of current input.
In general, the experimental assessment method of MR valve can be divided into two
types. The first type is the constant flow assessment while the second type is the
variable flow assessment.
22
simultaneously generates pressure difference between cylinder chambers. The pressure
difference between hydraulic chambers induces alternating flow between those
chambers which the flow line passes through the MR valve. Since the flow is
alternating, the variable flow assessment is very ideal to observe the hysteresis behavior
of the MR valve. The result of variable flow assessment of an MR valve is also
very useful to further predict the dynamic behavior of a damper that utilizes the MR
valve. The measurement arrangement is also relatively simpler than the constant flow
assessment given that the dynamic test machine or fatigue test machine can be used
as the source of movement to the hydraulic cylinder such as shown in Figure 2.14.
However, the method cannot be used to measure the time constant of the MR valve
since the valve response to rapid change of current input in a constant flow rate cannot
be executed.
2.5
23
the other hand, is defined as the behavior when the influenced variables are changing
over time. According to Snyder et al. [125], the linear steady state behavior model is
only sufficient to predict the energy dissipation, however, it cannot accurately portrays
the force response of the MR damper. The force response of the MR damper is known
as a highly nonlinear variable that is dependent to amplitude and frequency of motion
and it can only be modeled using nonlinear dynamic model [126129].
2.5.1
Steady-state Model
(2.1)
where is the MR fluid viscosity when no magnetic field applied, is the shear-strain
rate and sgn ()
is the signum function for shear-strain rate.
(H) + ,
y
po
=
,
pre
f or | | > ys
(2.2)
f or | | < ys
The expression of the biviscous model in Equation 2.2 has two states. For
absolute shear stress | | above static yield stress, ys , the expression is similar to
the Bingham plastic model with post-yield viscosity, po as the viscosity component
of the fluid. For shear stress below static yield stress, the expression is similar to a
conventional Newtonian fluid with pre-yield viscosity pre as the viscosity component
of the fluid. The relationship between dynamic yield stress y (H) and static yield stress
24
ys can be expressed as [6]:
po
y (H) = ys 1
pre
(2.3)
Wang and Gordaninejad [131] are among the first who generalized the use
of the Herschel-Bulkley model in MR fluid. The Herschel-Bulkley model can be
considered as the improvement of the Bingham plastic model since it has parameter
to anticipate shear thinning and shear thickening behavior, which only occurs at high
shear rates. The Herschel-Bulkley model can be expressed as [6]:
1
= y (H) + K ||
m sgn ()
(2.4)
In the case of MR valve, most of its steady-state behavior are also modeled
based on Bingham plastic model equation, although in some conditions the more
sophisticated models, such as the Herschel-Bulkley model, are more preferred [117].
In particular, the MR valve behavior normally declared by the following equation:
P = Pviscous + Pyield
(2.5)
25
fluid. The annular based MR valve model can be described as follows [23]:
Pviscous =
Pyield =
6QL
d3 R
(2.6)
c (H) L
d
(2.7)
where the pressure drop from viscous properties in Equation 2.6 is proportionally
related to the fluid base viscosity (), the volumetric flow rate (Q), and the annular
channel length of the valve (L) but inverse-cubical to the valve gaps (d) and inverse
proportional to the channel radius (R). Meanwhile, the field dependent pressure drop
in Equation 2.7 is proportionally associated with the field dependent yield stress value
( (H)) of MR fluid, annular channel length (L), and flow-velocity profile coefficient
(c) but inverse-proportionally related to the gap size (d). The coefficient c is normally
defined as the ratio between field dependent pressure drop and viscous pressure drop
which its value is in the range of 1 to 3.
Note should be taken that the model shown in Equation 2.6, described the
viscous pressure drop using the constant of 6. In some other literature [22, 27], the
viscous pressure drop was described using the constant of 12. The reason behind the
difference is in the understanding of the annular channel length (L). In a common
annular valve design, there will always be two annular channel length (L1 and L2) that
26
are separated by the coil winding as shown in Figure 2.15. Some researchers, prefer
to define both length separately and therefore the pressure drop expression of each
annular channel length is defined using the constant of 6. Some others, assume both
channel length identical (L1 = L2 = L) and therefore since the length of L defined as
2L, the constant of 6 became 12.
The other form of annular MR valve model was also presented using an
as shown in the following
analytical terms called the nondimensional plug thickness, ,
equation [20, 133]:
P =
bd3
12QL
2
1
1+
(2.8)
The other steady-state model that has been developed is the radial MR valve
model. Wang et al. [21] proposed the radial valve model as described in the following
equation:
Pviscous
Pyield
Do
Di
i
(H)
2 Do D
2
=
d
6Q
ln
=
d3
(2.9)
(2.10)
where the valve gap size (d) in Equations 2.9 and 2.10 refers to the size of radial gaps
while the Do and Di refer to the outer radius of the radial gaps and the inner radius
of the radial gaps respectively. The field dependent yield stress value ( (H)) of MR
fluid in Equations 2.9 and 2.10 is also specifically refers to the yield stress value of in
the radial gaps. Similarly with the annular valve model, the gap size is also inversecubically related with the viscous pressure drop and inverse proportionally related with
the field dependent pressure drop.
27
In the same form with the Equation 2.9 and 2.10, the radial valve model was
also presented in the following form [24]:
Pviscous
6Q
=
ln
d3
Pyield =
Do
Di
c (H)
(Ro Ri )
d
(2.11)
(2.12)
where the main differences between these models are only in preference of using
diameter (D) and radius (R) of the radial gaps in the equation and also in the constant
selection of the field dependent pressure drop where in equation 2.10 the moderate
constant of 2 is used while in equation 2.12, the flow-velocity profile coefficient, c, is
chosen.
2.5.2
Dynamic Model
While the steady state modeling approaches for MR valve has been established,
the dynamic modeling technique of MR valve as a single component is rarely
disseminated. This can be understood, since most of the literatures were only focused
their modeling in the prediction of the valve peak performance. However, to fulfill
other purpose such as the control design process, the steady-state model only will
not adequately represent the device behavior. The necessity of dynamic modeling
of MR valve also can be considered higher than MR damper because the MR valve
28
and obviously the model can be used in the wider range of applications aside of
MR damper. For example, the knowledge of MR valve behavior and its virtual
representation can be used as a reference to design and predict the performance of
an MR damper or to help further development of new concept of actuators that require
MR fluids flow control [25, 128, 134].
There are various kinds of parametric model that have been developed. For
example, Sakai et al. [137], proposed a simple parametric model based on the LuGre
friction operator as an inverse model of an MR damper as an actuator for vibration
control. The LuGre friction model also explored and modified by Jimenez and AvarezIcaza [138] to model an MR damper as a part of a semi-active control system. The other
type of parametric model, which also apparently the most popular parametric model,
known as the Bouc-Wen model has been extensively explored as the parametric model
of MR damper [136,139143]. The Bouc-Wen model approximate the device behavior
through a set of hypothetic spring and damper as shown in Figure 2.16.
29
Gavin et al. [144]. The method was then utilized by Kwok et al. [135] to model
an MR damper incorporating particle swarm optimization (PSO) as the parameter
identification method. Jiang and Christenson [145] also use similar model in a realtime hybrid simulation environment for a large scale MR damper. A performance
comparison study of various parametric modeling approach that has been conducted
by ahin et al. [146] reported that the class of algebraic model is able to provide better
performance fit than differential model. Although, according to comprehensive review
by Wang and Liao [74], the accuracy of the parametric model is not just rely on the
basic form of the model itself but also highly determined by the identification method
of the parameters. In general, Wang and Liao [74] inferred that model with higher
number of parameters and equipped with an appropriate identification method will
have better accuracy than the one with less parameters.
The other parametric method that is also often explored in the MR damper
model is the polynomial model. Choi et al. [147] introduced the approximation of the
force-velocity curve of an MR damper using polynomial equations. The method is
simple yet unique because there were two separated 6th order polynomial equations
for each positive and negative directions as shown in Figure 2.17. to compensate
the inequality of the force-velocity curve in positive and negative direction due to
hysteresis. The method was then applied by Ubaidillah et al. [148] in the modeling
approach of an MR damper for automotive semi-active suspension system. The
polynomial model is favored because it is convenient to use and easy to be adapted in
an open-loop system [149]. However, according to Sahin et al. [146], the polynomial
model has weakness in characterizing the device behavior in low velocity region due
to the absence of variables that characterize the pre-yield force.
30
2.6
Despite the lack of specific literature discussed about MR valve as a standalone device, in fact, it is the most vastly used MR devices. The MR valve has become
the most influential component in the MR damper, which is commercially the most
successful MR device. Recently, the development progress of new MR actuators also
stimulated the rapid development of MR valve. In this section, the utilization of MR
31
valve is elaborated in three different MR devices, namely the linear MR damper, the
rotary MR damper and the new MR actuator.
2.6.1
Linear MR damper is one of the most discussed MR devices so far. The range
of application of MR damper is widely ranged from automotive to civil applications.
According to the working modes, there are two basic contruction of MR damper. The
most common design is the valve mode MR damper with valve located in the piston as
shown in Figure 2.19. The other MR damper equipped with shear mode as depicted in
Figure 2.20.
Since both types of damper regulated flow using the narrow gaps in the piston,
these narrow gaps are functioned as the valve that is embedded in the damper piston.
Although these structures are the most common structure of MR damper due to its
compactness the fact that the coil should be wound in the piston has increased the
complexity of wire assembly as well as increasing the weight and thickness of the
piston head [14]. The internal coil placement also made the heat dissipation from the
coil more difficult and potentially increase the temperature of the MR fluid [58].
The alternatives of coil placement has been proposed by Chen et al. [155]
using external coil structure (see Figure 2.21). The construction of the damper not
just allow the heat dissipation to be directly released to the atmosphere but also made
it possible to add the self-powered and self-sensing features [157]. The coil also can
32
33
be easily accessed for maintenance without having to interfere with the internal part of
the cylinder nor the fluid. However the basic construction of the external coil design
is quite complicated and the performance of the damper is apparently less than the
internal coil design due to difficulties in optimally routing the magnetic flux.
The other approach of coil placement are using the bifold valve design shown
in Figure 2.22 [28] where the coils are statically installed in both ends of the cylinder.
The bifold valve damper gain the advantage in design compactness similarly with the
damper with internal valve. Good damping performance was also reported especially
in high shock velocity. However, the design is also quite complicated and the valve
34
effective area is very limited and thus the performance expansion cannot be executed.
Since the design can provide advantage in high shock velocity, the design has been
explored for vibration mitigation application by Facey et al. [156] and Mao et al. [158]
as illustrated in Figure 2.23 and 2.24.
If the design simplicity and the ease of capacity expansion are the priority,
then the bypass structure can be a better option. The bypass structure where the valve
35
are separated from the piston and physically located outside the cylinder as depicted
in Figure 2.25 [8]. The separation between the cylinder and the valve has made the
assembly process of the bypass damper is far simpler than the other damper since
variety of commercially available hydraulic cylinder can be used. In the other words,
the customization requirement of damper components is very minimum. Though the
assembly process is simpler, the mounted valve outside the cylinder typically made the
size of bypass damper larger than the damper with internal valve. Nevertheless, most
of applications of bypass MR damper is on a large scale damping component such as
the seismic dampers shown in Figure 2.26 [60].
Figure 2.26 Bypass MR damper for large scale seismic application [60]
2.6.2
36
rotary damper also more suitable in a hostile working environment since the damper
shaft in a rotary damper is sealed better than the damper rod in a linear damper.
However, the rotary MR damper that equipped with valve is so far is limited to the
vane type MR damper only. The other devices that can be classified as rotary MR
damper, e.g. MR brake, are typically not equipped with valve and operate mainly in
shear mode.
37
on damper performance were presented.
Figure 2.28 Vane type MR damper with outer coil valve [32]
Generally, the vane type MR damper design by Giorgetti et al. in [32] has
shown simplicity and compactness. The experimental results showed that the damper
could provide up to 200 Nm damping torque at 3 A current input in a feasible geometry
to be applied in passenger vehicle suspension. However, the incorporated internal valve
could have the same complexity in terms of coil maintenance as that described in [159].
Yang et al. [160] discussed a vane type MR damper design with flow type
internal valve. The design has shown similarities with the work done by Giorgetti et
al. in [32] in the context of annular flow mode valve design, where the effective areas
are located in line with the direction of fluid flow. The main difference was in the
placement of the coil, as shown in Figure 2.29. In Giorgetti et al. design [32], the coil
was located in the outer radial side of the fluid path, while in Yang et al. design [160],
the coil was located in the inner radial side of the fluid path.
38
Figure 2.29 Vane type MR damper with inner coil valve [160]
Since Yang et al. [160] put the coil in the inner radius, a lower resistance value
of the coil could be obtained leading to lower power consumption. The valves in the
designs of Yang et al. [160] and Giorgetti et al. [32] have similar dimension and number
of coil turns as well as the wire gauge, the coils in the Giorgetti et al. design [32] would
simply have higher resistance than the coils in the Yang et al. design [160]. Higher
resistance in coil means higher power consumption for the same current input, since
power relates proportionally to resistance and quadratically to current. Therefore, if it
is assumed that both valves have the same effective area and gap width, then both valve
would have the same performance but the Yang et al. design [160] would consume less
power.
The disclosed experimental results show that the achievable damping force
that can be generated by Yang et al. [160] could be up to 3 kN at 4A current input.
However, since the damping performance was evaluated in different measurement units
the results are difficult to be compare with those of Giorgetti et al. [32]. Giorgetti et
al. [32] demonstrated the damper performance in torque unit, while Yang et al. [160]
showed them in force unit. Although it is known that the achievable damping force
can be converted to torque by multiplying them with the length of rotating hub, the
length of rotating hub was not specifically mentioned in any part of their publications.
Nevertheless, logically, both designs should be in the same class in terms of achievable
damping torque, since both of them were designed for passenger vehicle suspension.
Both of them were also capable of providing almost the same ratio between on-state
and off-state damping at around 6.
39
2.6.3
Aside from the device component, the MR valve also has been developed as a
stand-alone device. The stand-alone MR valve is an independent component that can
be used to retrofit the hydraulic valve or to complete a newly developed actuators.
There are several discussions about a new MR based actuator that can be functioned as
an active actuator similarly with a hydraulic or pneumatic actuator such as the bellowdriven motion controller shown in Figure 2.30. [19], the MR hydraulic power actuation
system in Figure 2.31 [25], the MR actuation system with embedded terfenol-D pump
in Figure 2.32 [26], and the MR based link manipulator in Figure 2.33 [134].
40
One of the unique working application of the MR valve is the concept of 4/3
way directional MR valve proposed by Saloom and Samad [27] as depicted in Figure
2.34. The 4/3 way directional valve is one of the common functional hydraulic valve
that serve the hydraulic actuator and determine the direction of force actuation. The
4/3 way directional valve normally implemented in servo hydraulic actuation but in
order to operate in higher performance, a sophisticated valve actuation is required. The
design of 4/3 way directional MR valve was one of the proposed solution that utilizing
the rapid response of MR fluid to magnetic field. The proposed 4/3 way MR valve
concept was basically constructed of four single MR valves that each of the valve
element was arranged in between the valve ports. There are five ports in this valve
design, one inlet pressure port, two actuator ports, and two return ports. By activating
and adjusting the strength of magnetic field in different combination of annular valve,
the direction and rate of MR fluid flow can be regulated so that the actuation control of
force and direction of hydraulic can be achieved.
41
2.7
Summary of Chapter 2
CHAPTER 3
3.1
Introduction
3.2
43
3.2.1
Conceptual Design
There are two disclosed MR valve concepts that provided less than 50 mm
diameter MR valve with achievable pressure drop more than 1.5 MPa at 40 ml/s
flow rate. The first concept was proposed by Yoo and Wereley [20] using double coil
annular configuration while the second concept was introduced by Ai et al. [23] using
combination of annular and radial configuration which experimentally elaborated [24].
The double coil annular concept, as depicted previously in Figure 2.9, was proposed as
a high performance MR valve that can generate pressure drop up to 2 MPa at fluid flow
rate of 40 ml/s with outer diameter size of 25.4 mm and uses annular flow path channel
with internal coil attachment. The valve size still can be reduced up to 40% without any
performance degradation if high permeability material such as Hiperco alloy is used.
Meanwhile, the combination of annular and radial concept, shown in Figure 2.12, was
introduced as an effort to improve the performance of conventional MR valve design.
The first concept of MR valve with annular and radial flow path has outer diameter size
of 45 mm and capable to perform in similar pressure drop range with the double coil
annular MR valve.
When the valve size matters, any improvement efforts which involve
enlargement of valve dimension are not preferable. For that reason, extending the
flow path of MR fluid is the feasible option. The combination of annular and radial
concept by Ai et al. [23] is a good example of effective area expansion by extending
the flow path. Another study by Nguyen et al. [72] has also confirmed that the MR
valve with combination of annular and radial gaps could provide higher achievable
pressure drop than other types of annular MR valve with the same outer radius and
44
power consumption. However, since the current design uses only a pair of annular
and radial gaps, the total effective area improvement is still very narrow; thus, the
enhancement of pressure drop that can be achieved is also limited. Therefore, in order
to expand the total effective area more, the length of flow path should be increased
using multiple combination of annular and radial gaps forming a meandering flow path.
The conducted study in this research will explore the effectiveness of the meandering
flow path concept to improve the pressure drop performance of the MR valve in the
sequence shown in Figure 3.1.
The detailed structure of MR valve with meandering flow path that is explored
in this study is shown in Figure 3.2. The MR valve structure can be divided into
three components, the casing, the coil and the valve core. The casing resembles the
outer shell of the valve with two main functions. The first function is to secure the
whole structure of the valve together as well as to connect the valve with other devices
through its embedded fittings. The second function is to properly guide the magnetic
flux from the coil so that the loss of flux to the air, which reduce the efficiency, can be
minimized. The coil is the main component that generates and determines the magnetic
field strength of the MR valve. In this proposed MR valve structure, the coil is turned
in a coil bobbin that is also functioned as the wall of the flow channel. The core is the
component that formed the annular and radial flow channel, which can be divided into
four parts, the side core, the center core, the orifice core and the spacer.
45
3.2.2
Design Consideration
In Table 3.1, the combination summary of all possible material arrangement for
the three main components of the valve is shown. The main goal of these combinations
is to find the best material selection that capable to guide the fluxes passing through
46
No.
1
2
3
4
5
6
7
the fluid gaps with sufficient magnetic field strength. From these eight possible
combinations, only two combinations show the possibility of flux to pass through the
fluid flow gaps. However, only the combination of magnetic casing and core with nonmagnetic coil bobbin is capable to provide a high magnetic field strength in the fluid
gaps. From the remarks in Table 3.1, it can be concluded that in order to successfully
guide the fluxes to pass through the fluid gaps, the permeability of the core component
should be higher than the permeability of the coil bobbin. However, to get a higher
magnetic field strength in the fluid gaps, the permeability of the casing should be at
least similar to the permeability of the core.
In this study, the selected material of the casing is the mild steel with properties
that compatible with the AISI 1010. The design of the casing consists of two identical
parts, which each part has four threaded holes as the holder for the locking bolts and
one female BSPP (British Standard Pipe Parallel) 1/4 fluid channel port. Since the
coil bobbin has to be made from the low permeability material, aluminum is chosen
for the material of the coil bobbin. The coil bobbin is also designed with a pair of
grooves for O-rings installation and has the threaded holes in pair to the holes in the
casing. Meanwhile, the valve core is separated into two side cores, two orifice cores, a
center core and six spacers. All parts in the valve core are also made from mild steel
47
that comply with AISI 1010 material except for the spacers, which are made from
aluminum.
The outer diameter size of the valve is determined of 50 mm with overall length
of 77 mm. On the other side, the inner diameter size, that is the inner diameter of
the coil bobbin that becomes the bore size of the flow channel is determined of 15
mm. The core parts are then arranged to form the annular and radial gaps alternately
in such a way it meanders the flow channel. As one of the subject of the study, the
size of annular and radial gaps are both varied from 0.5 to 2.0 mm. Since the valve
is designed to generate high pressure drop, the casing should be strong enough to
withstand such pressure and for that purpose the thickness of the casing is determined
of 5.0 mm. As a consequence, the space available for the coil winding that is made from
23 SWG copper wire is sufficient for around 545 turns with total resistance of around
3.03 . In this study, the current input is limited to 1.0 A therefore the maximum power
consumption of the valve is also limited to around 3.03 W.
3.2.3
Magnetic Simulation
(3.1)
where B is the magnetic flux density in Tesla which is highly determined by current
input, coil turns and dimension, and total reluctance of the medium. The permeability
curve of the MRF-132DG, together with the other permeability value of the materials
used in the MR valve can be used in the calculation to determine the magnetic flux
density, B, at the effective area. Since the number of turns of the coil is already
determined and the permeability of magnetic material was assumed to follow the B-H
curves of AISI 1010, the magnitude of magnetic flux density can be determined by
the magnitude of current input supplied to the coil. However, due to the complexity
of magnetic structure and non-linear permeability of the materials, the magnitude of
48
magnetic flux density for each zone is difficult to be calculated analytically [111].
Thus, in order to determine the value of B for each corresponding current input, a
finite element method based software for magnetic simulation called FEMM is used.
The results from FEMM are shown in Figure 3.4 and 3.5.
Figure 3.4 shows two dimensional axisymmetric meshed model in the FEMM
using triangular element with total elements number of 103771 and total nodes number
of 52281. Figure 3.5 shows the contour and flux lines of magnetic flux density for the
49
Figure 3.5 Flux lines and contour of magnetic field of the MR valve in FEMM
whole magnetic circuit. Notes should be taken that the flux that can influence yield
stress of MR fluid is only the fluxes that passed through MR fluid gaps. The MR fluid
gaps area that crossed by the magnetic flux is known as the effective area. According
to the result in Figure 3.5, it can be observed that the flux lines have passed across the
outer annular gap, outer radial gap and inner radial gap. Therefore, only MR fluid that
passed through these three gaps can be rheologically influenced by the changing of the
current input. Meanwhile, the inner annular and orifice gaps are not intensively exposed
to magnetic field and therefore the influence of magnetic field to yield stress of MR
fluid in these areas can be neglected. Yet, although the inner annular and orifice gaps
do not provide field-dependent pressure drop, these gaps still have the contribution to
the viscous pressure drop. Nevertheless, more focus should be taken at the flux lines
in the outer annular gap and outer radial gap since the structural configuration of outer
annular and outer radial gaps have made these gaps formed as a parallel channel for the
flux lines. Thus if flux-loss is neglected, the total number of magnetic flux that passed
through the outer annular and outer radial gaps are equal to the number of magnetic
flux that passed through the inner radial gap. This is why the inner radial gaps have
higher magnetic flux density than the outer annular gaps or the outer radial gaps. The
difference of magnetic flux density between the outer radial gaps and the inner radial
gaps is also the reason why the radial gaps are divided into the outer and inner zones.
The distribution of magnetic flux density value along the flow path channel for
0.5 mm gap size is shown in Figure 3.6 where the flux density is seen to be increased
to the augmentation of current input to the coil. In this study, the maximum current
50
Figure 3.6 Magnetic flux density along MR fluid flow path for 0.5 mm gap size with
respect to various current input
input is limited to 1.0 A while the maximum flux density that can be reached along the
flow path channel is around 0.7 Tesla at the inner radial gaps. As a result of parallel
magnetic channel, the second highest flux density is reached at the outer radial gaps
of around 0.5 Tesla while the flux density of the outer annular gaps is only around 0.2
Tesla. The magnetic flux density values for each corresponding current input is used to
predict the yield stress of MR fluid using Equation 3.1. Therefore, from these results,
it can be predicted that the pressure drop of the inner radial zone will be the most
significant contributor to the total MR valve pressure.
3.3
51
viscous properties of the fluid and pressure drop from the field dependent yield stress
of the fluid. The basic expression of pressure drop in an MR valve can be declared by
following quasi-steady equations [24]:
P = Pviscous + Pyield
Pviscous =
Pyield =
(3.2)
6QL
d3 R
(3.3)
c (B) L
d
(3.4)
where expressed that the pressure drop from viscous properties in Equation 3.3 is
proportionally related to the fluid base viscosity (), flow rate (Q), and annular channel
length of the valve (L) but inverse-cubical to the valve gaps (d) and inverse proportional
to the channel radius (R). Meanwhile, the field dependent pressure drop in Equation
3.4 is proportionally associated with the field dependent yield stress value ( ) of MR
fluid, annular channel length (L), and flow-velocity profile coefficient (c) but inverseproportionally related to the gap size (d). The coefficient c is obtained by calculating
the ratio between field dependent pressure drop and viscous pressure drop using the
approximation function as defined by [72] in the following equation:
c = 2.07 +
12Q
12Q + 0.8Rd2 (B)
(3.5)
However, the expressions in Equations 3.3 and 3.4 are only valid for MR valve
with annular gaps. For radial gaps, the viscous pressure drop and the yield pressure
drop can be expressed as [24]:
Pviscous
6Q
=
ln
d3
Pyield =
Ro
Ri
c (B)
(Ro Ri )
d
(3.6)
(3.7)
where the valve gap size (d) in Equations 3.6 and 3.7 refers to radial gaps while the
and refer to the outer radius of the radial gaps and the inner radius of the radial gaps
respectively.
The mathematical expression for the orifice gaps is slightly different, because
no field dependent yield stress is incorporated into the equation since the fluxes did not
pass through the orifice gaps. Therefore, the pressure drop equation for the orifice gaps
52
is only expressed by the equation of viscous resistance of the fluid [22].
P =
8QL
R 4
(3.8)
Since the proposed MR valve design has both annular and radial gaps, the
mathematical expressions of pressure drop should include all the expressions in
Equations 3.3 to 3.8. However, to make the derivation simpler, the valve gaps is
separated into five different zones, the outer annular gaps zone, the outer radial gaps
zone, the inner annular gaps zone, the inner radial gaps zone, and the orifice gaps zone
as shown in Figure 3.7. The valve needs to be clustered into five zones because each
zone was presumed to have different magnetic flux density. These five zones also could
also be divided into two categories, the zones that categorized as effective area and the
zones that have only viscous resistance. The zones that categorized as effective area
are the outer annular gaps zone, the outer radial gaps zone, and the inner radial gaps
zone while the zones that have only viscous resistance are the inner annular zone and
the orifice zone.
Figure 3.7 Gaps zone in MR valve with multiple annular and radial gaps
From Figure 3.7, it can be seen that there are two outer annular gaps, two
outer radial gaps, three inner annular gaps, four inner radial gaps and two orifice gaps.
Using the expressions of annular and radial pressure drop in Equation 3.3 to 3.8, the
quasi-steady pressure drop of MR valve with multiple annular and radial gaps can be
53
described in the following equations:
Pvalve = Pannular outer + Pradial outer + Pannular inner
+ Pradial inner + Porif ice
Pannular outer
ca a (B) Lao
6QLao
+
=2
3
dao Rao
dao
Pradial outer
6Q
=2
ln
d3ro
Pannular inner = 3
6Q
=4
ln
d3ri
Porif ice = 2
R1
R0 outer
cro ro (B)
+
(R1 R0 outer )
dro
6QLai
d3ai Rai
Pradial inner
(3.9)
(3.10)
(3.11)
(3.12)
R1
R0 inner
cri ri (B)
+
(R1 R0 inner )
dri
8QLo
R0 inner 4
(3.13)
(3.14)
The complete parameters for the MR valve with multiple annular and radial
gaps are shown in Table 3.2 while the dimensions and the specific location of each
corresponding variable of the proposed MR valve are shown in Figure 3.8.
54
3.4
Figure 3.9 illustrates the results of maximum achievable pressure drop of the
multiple annular and radial MR valve at 40 ml/s flow rate with gap size of 0.5 mm
and current input 1.0 A is around 5.3 MPa. As a comparison, the achievable pressure
drop of the previous design of compact MR valve at 40 ml/s flow rate is only around 2
MPa [20,24]. In other words, the proposed MR valve design has shown the contribution
to improve the achievable pressure drop of a compact MR valve. Furthermore, it is
shown that the pressure drop from the inner radial gap zone is the highest contributor to
the total achievable pressure drop of the MR valve. Surprisingly, although the magnetic
flux density of the outer annular zone was already known to be the lowest from the
other effective area, the maximum pressure drop of the outer annular zone is higher
than the outer radial zone. Meanwhile, the total viscous pressure drop of the MR valve
is shown to be nearly 10 MPa which mostly contributed from the outer annular zone,
followed by the inner radial zone and the outer radial zone. Since the contribution of
viscous pressure drop from the inner annular zone and the orifice zone are relatively
small, these zones are not included in the graph in Figure 3.9. The percentage of
contribution from each zone to the MR valve pressure drop is illustrated in Figure
3.10.
55
Figure 3.9 Estimation of achievable pressure drop of MR valve with 0.5 mm gap size
The illustrations of pressure drop percentages from each zone in Figure 3.10
can be used to explain why the pressure drop at the outer annular zone can outperform
the outer radial zone although the magnetic flux density in the outer annular zone is
lower than the outer radial zone. The viscous pressure drop of outer annular zone has
been shown to be the highest contributor of viscous pressure drop of the MR valve. The
magnitude of viscous pressure drop at the outer annular zone has contributed more
than half of the total viscous pressure drop of MR valve while the outer radial and
the inner radial zone have only contributed around 11% and 28% respectively. The
channel length, channel radius, gap size and fluid flow rate, as shown in Equation 3.10,
(a)
(b)
Figure 3.10 Percentage of pressure drop contribution from each zone (a) viscous (b)
field-dependent at 1 A current input
56
are the variables that responsible in determining the magnitude of the viscous pressure
drop at the outer annular zone. However, among them, gap size is known as the most
significant variable since it inverse-cubically influences the viscous pressure drop.
As the most significant variable that determines the viscous pressure drop, the
effect of gap size variations on the pressure drop of MR valve should be evaluated.
Since the flow path has been divided into five zones, there are five different gaps to be
varied. However, to simplify the combinations, the gap size for both outer radial and
inner radial zones are determined to be equal (dro = dri ) and varied from 0.5 mm to
2.0 mm. The same variations and interval are also used at the outer annular gap size.
Meanwhile, for zones that have only viscous resistance such as the inner annular gap
and the orifice gap, the gap size is kept constant. The effect of each combination on the
viscous and field dependent pressure drop of MR valve are shown in Figure 3.11:
Figure 3.11a and 3.11b show that the gap size has a very significant effect
to the magnitude of viscous pressure drop as well as the magnitude of field-dependent
pressure drop. However, the most significant influence of the annular gap to the viscous
pressure drop of the MR valve is seen in the interval between 0.50 to 0.75 mm where
the viscous pressure drop of the MR valve could be reduced in nearly 0.4 MPa.
Obviously, the change of annular gap also has some effect to the magnitude of fielddependent pressure drop. Figure 3.11b shows that the increase of annular gap size
from 0.50 to 0.75 mm reduces the field-dependent pressure drop of around 0.8 MPa.
Meanwhile, the enlargement of radial gap seems to have more influential effect to the
reduction of field-dependent pressure drop magnitude than to the viscous pressure drop
magnitude. From Figure 3.11b, it can be observed that increasing radial gap size from
0.50 to 0.75 mm reduces the viscous pressure drop of only 0.25 MPa but responsible
to the decrease of field-dependent pressure drop for almost 2.5 MPa.
57
(a)
(b)
Figure 3.11 Effect of gap size on the pressure drop (a) viscous (b) field-dependent at
1 A current input
58
only reduces viscous pressure drop about 27%. The reason of this can be explained
by considering the effect of gap size changing to the magnetic circuit performance.
Since the magnetic resistance of MR fluid in the gaps is a lot higher than the magnetic
resistance of magnetic material in casings and cores, small change in the gap size will
significantly change the magnetic circuit and influence the magnetic flux density [24].
Therefore, although in Equations 3.10, 3.11 and 3.13 the gap size seems to only
inverse-proportionally related to the field-dependent pressure drop, in actual, the gap
size also influences the field-dependent pressure drop by simultaneously changing the
magnitude of field-dependent yield-stress of MR fluid.
However, in terms of overall MR valve performance, high achievable fielddependent pressure drop only is not sufficient to show that the MR valve is working
in a decent performance. Since, MR valve with high field-dependent pressure drop
but also with high viscous pressure drop value means that the operational range of MR
valve is low. The operational range of MR valve can be determined using the following
expression.
=
Pyield
Pviscous
(3.15)
From Figure 3.12, it can be observed that the operational range of the valve
59
consistently increased when the gap size at the outer annular zone is increased. The
most significant operational range improvement occurred when the outer annular gap
size is increased in the interval between 0.50 to 0.75 mm. Since the increase of
annular gap size provides more significant impact to the viscous pressure drop than
the field-dependent pressure drop, consequently, the occurrence of operational range
augmentation in this interval can be considered as a result of significant viscous
pressure drop reduction. In the contrary, it is clear that the increase of radial gap
size will have inverse relationship with the operational range since the increase of
radial gap size will cause significant degradation to the field-dependent pressure drop.
However, some exceptions are found for the combinations of 1.0, 1.5 and 2.0 mm
annular gap sizes with radial gap size in the interval between 0.50 to 0.75 mm. The
anomaly of trend in these intervals appears because the performance degradation of
the viscous pressure drop in these gap combinations are higher than the performance
degradation of the field-dependent pressure drop. The higher viscous pressure drop
degradation is also shown to be increased in the combination with higher annular gap
size. However, as the size of radial gap increased, the dominance of field-dependent
pressure drop degradation is started to overlap the degradation of viscous pressure
drop. Nevertheless, it still can be concluded that the increase of annular gap size
generally has positive effect to the operational range while the increase of radial gap
size tend to reduce the operational range. In this case, due to trend anomaly and
significant impact to the change of pressure drop value, the gap size interval in between
0.5 to 1.0 mm require more attention and will be further explored in the experimental
60
assessment.
Table 3.3: Performance benchmarking between the proposed MR valve concept and
the counterparts
200 kN
400 kN
250 kN
AnnularProposed
Specification MR damper MR damper MR damper
Radial MR
MR valve
[165]
[166]
[21]
valve [24]
Valve type
Annular
Annular
Radial
Annular-Radial Meandering
Maximum
pressure
3.8
16.1
9.6
2.5
5.3
drop (MPa)
Maximum
Power (W)
262
1920
900
-NA3
Pressure
drop per
Power input
14.5
8.39
10.7
-NA1766
(kPa/W)
Gross
Volume
1
31
19
0.1
0.15
(dm3 )
Pressure
drop per
Volume
3.8
0.52
0.48
24.87
35.08
3
(kPa/cm )
61
3.5
Summary of Chapter 3
A new concept for a MR valve with meandering flow path was presented in
this chapter. The new MR valve is presented by elaborating the conceptual design
of the valve followed by the explanation about the consideration of the design
and magnetic field simulation. The derivation of the MR valve steady-state model
and the performance prediction of the MR valve were also presented in separate
section. The novelty of the new MR valve concept is mainly justified by the unique
arrangement of the effective area which makes it capable to deliver higher pressure
drop capacity than its counterparts. The steady-state modeling and simulation work of
the achievable pressure drop in accordance to the viscous and field-dependent pressure
drop contributions have been conducted to predict the achievable performance of the
MR valve. The effect of gap size selection in the annular and radial channel to the
magnitude of viscous and field-dependent pressure drop also has been analyzed to
evaluate the significance of each channel to the overall performance of the MR valve.
According to the assessment results, the outer annular channel was shown as the most
influential region to the characteristics of viscous pressure drop. Meanwhile, the inner
radial channel was more influential to determine the characteristics of field-dependent
pressure drop. The assessments of the gap size effect in each regions were also
analyzed in terms of dynamic range of MR valve. The gap size of annular channel has
been reported to have positive correlation with the dynamic range of MR valve while
the gap size of radial channel were reported to have inverse effect. The knowledge of
these characteristics is important as a reference to perform the experimental testing
that will be discussed in the next chapter.
CHAPTER 4
EXPERIMENTAL ASSESSMENT
4.1
Introduction
The advantage of the meandering flow path concept has been showed in
Chapter 3 through the comparison between the achievable pressure drop of the
meandering valve concept with its predecessors, which typically has maximum
pressure drop up to 2.5 MPa. In this chapter, the concept will be experimentally
explored to prove and demonstrate its performance capability in improving pressure
drop. This chapter is arranged in three sections, which is started with the explanation
about the experimental apparatus, followed by the elaboration of the experimental setup, which explains the measurement methodology, and finally the discussion about
the experimental results. The chapter is discussed in four different subsections, the
off-state characteristics, the on-state characteristics, the analysis about the frequency
excitations effect, and the analysis about the current input effect. This chapter is ended
by the last section that concludes the chapter.
4.2
Experimental Apparatus
4.2.1
Magnetorheological Fluid
63
are listed in Table 4.1 while its magnetic properties, presented in the B-H curve, and
the curve are shown in Figure 4.1 and 4.2 respectively.
Table 4.1: Typical properties and material compatibility of MRF-132DG [167, 168]
FLUID PROPERTIES
Base Fluid
Hydrocarbon
Solid Content by Weight, %
80.98
3
Density, g/cm (lb/gal)
2.95-3.15 (24.6-26.3)
Operating Temperature, C ( F)
-40 to 130 (-40 to 266)
Flash Point, C ( F)
>150 (>302)
64
4.2.2
Magnetorheological Valve
65
radial and 0.5 mm annular gaps, the 0.5 mm radial and 1 mm annular gaps, the 1
mm radial and 0.5 annular gaps, and the 1 mm radial and 1 mm annular gaps. These
gap configurations, according to the simulation results in Section 3.4, were reported to
potentially have the highest achievable pressure drop.
66
However, after some initial tests, it was found that the design has weakness
in the strength of bolt-locking mechanism between the casing and the coil bobbin.
As shown in Figure 4.4, the M2 female thread in the aluminum bobbin is apparently
not strong enough to hold the casings together under high internal fluid pressure which
caused failure. Therefore, the modification is made to the design, especially to improve
the strength of casings assembly joint. The modifications is made by changing the boltlocking mechanism from the casings to the aluminum bobbin to a directly-coupled
thread between casings as shown in Figure 4.5. With the modified design, better
strength of casing assembly can be achieved and less component is required in the
valve structure.
(a)
(b)
Figure 4.5 Modification and comparison of the MR valve prototype(a) Exploded view
of the MR valve design (b) Fabricated prototype of MR valve
67
4.2.3
Testing cell
For instance, the arrangement of the MR valve in the testing cell has made the
device looks similar with a bypass linear damper. The main difference, in this case, is
that the dynamic force-velocity behavior of the testing cell is not the point of interests
since the behavior of the valve is mainly related to the valve pressure drop and the fluid
68
volumetric flow rate. In order to measure the valve pressure drop, two pressure sensors
are installed in the nearest point to both MR valve ports to ensure that the pressure
drop effect from the flow conduits does not alter the pressure measurements. The
volumetric flow rate is measured indirectly by multiplying the measured velocity of
the piston and the effective piston area. The velocity data is obtained by differentiating
the displacement data of the piston that is measured by the native displacement sensor
of the dynamic test machine.
4.3
Experimental Set-up
The platform is used to provide the oscillatory movement to the testing cell
which will result in pressure changing inside the cylinder and will periodically pump
the MR fluid to flow through the MR valve. The pressure sensor measures the realtime changing of the fluid pressure in each port of the valve. The pressure difference
between the inlet port pressure and the outlet port pressure represents the MR valve
pressure drop. In order to form the relationship between pressure drop and flow
69
rate as well as the current input, the measurement was conducted in three different
inverter-controlled frequencies namely 5.0 Hz, 7.5 Hz, and 10 Hz, which, according
to the measurement, are approximately equal to peak flow rate of 37 ml/s, 55 ml/s
and 73 ml/s. The peak flow rate and its associated pressure drop will be plotted to
form the relation. The approach is similar with the common method in any damper
characterization process to form the force-velocity curve [154].
Since the force induced to the testing cell is also proportional to the pressure
drop of the valve, the larger the pressure drop is, the higher the force required to
move the testing cell. Unfortunately, the in-house dynamic test machine, which use
the inverter-controlled AC motor as the prime-mover of the platform, is not able to
provide sufficient force to move the testing cell in low frequency excitation. The
speed reduction of the motor with inverter control is proportional to the reduction
of motor power, which obviously decrease the torque of the motor. Therefore, even
when the platform is moving, it is difficult to keep the excited frequency in constant
value, especially during the increased pressure drop of the MR valve. As a result, the
measurement range of the in-house dynamic test machine is very limited.
Due to these limitations, the results of the in-house test machine will only
be used to evaluate the peak off-state pressure drop of the valve and the peak onstate pressure drop. The dynamic pressure drop, on the other hand, is evaluated with
70
During experimental testing with the Fatigue Dynamic Test Machine, the
testing cell is subjected to a sinusoidal wave, where one end of the testing cell is
fixed and the other end is actuated. Since the sinusoidal wave can be expressed in
u = A sin (2f t)), the amplitude, A, is fixed to +/- 25 mm, while the excitation
frequency, f , is varied from 0.50 to 1.00 Hz. In order to demonstrate the MR effect,
the magnetic field strength is also varied by the changing the magnitude of the current
input. The summary of variable arrangement for the experimental test is shown in Table
4.2.
71
Figure 4.9 Testing cell installation in the Shimadzu Fatigue Dynamic Test Machine
Table 4.2: The variable arrangement of experimental test using Shimadzu Fatigue
Dynamic Test Machine
Excitation Frequency Excitation Amplitude Current Input Cycles
(Hz)
(mm)
(Ampere)
0.50
25
0-1
25
0.75
(interval 0.1)
1.00
4.4
Experimental Results
4.4.1
The off-state pressure drop can be defined as the pressure drop generated by the
MR valve when the electromagnet is charged with 0.0 A current. In the other words,
the off-state pressure drop is the pressure drop with only viscous contribution without
any magnetorheological effect from the MR fluid. The peak value of the measured off-
72
state pressure drop across the MR valve with meandering flow path as a function of
flow rate is shown in Figure 4.10.
Figure 4.10 Comparison of measured and theoretical off-state peak pressure drop at
various flow rates for 0.5-0.5 mm (annular-radial) gaps configuration
Figure 4.10 shows that generally the trend of the theoretical off-state pressure
drop value is in accordance with the measured one, in the range of flow rate
measurement between 37-73 ml/s, the pressure drop increased proportionally to the
flow rate. In specific, the exact value of the measured off-state pressure drop is slightly
lower than the theoretical off-state pressure drop from the model if the assumed
viscosity value is equal to 0.092 Pa s, which is the specified viscosity value of the
MRF-132DG from the manufacturer. However, since according to the documentation
from Lord Corp. [168], the exact viscosity value also can be varied up to +/- 0.015
Pa s from the specified value, thus the theoretical off-state pressure drop value also
can be dropped lower than the measured value if lower viscosity value is selected
in the calculation. If the range of viscosity value is taken into consideration, then the
theoretical off-state pressure drop as a function of flow rate will not form into a straightline curve, but will be more like a band that representing the range of predicted offstate pressure drop values. Figure 4.10, in this case, shows the agreement of the range
of predicted off-state pressure drop values with the measured off-state pressure drop
values in the measurement range between 37-73 ml/s.
73
The on-state pressure drop is the pressure drop that is generated when the
valve coil is charged with a current. The performance of the valve during the on-state
condition is measured in the range of current input between 0.2 to 1.0 A with 0.2 A
interval. The relationship between the measured on-state peak pressure drop and the
current input for three different flow rate variations, along with its comparison to the
predicted value from the model are shown in Figure 4.11.
Figure 4.11 Comparison of measured and theoretical on-state peak pressure drop at
various current inputs and flow rates for 0.5-0.5 mm (annular-radial) gaps configuration
In Figure 4.11, the trend line agreements of the pressure drop to the current
input curve between theoretical and measurement are shown. Due to the power
limitation of the test rig prime mover, the measurement data for 37 ml/s flow rate
are only able to be recorded up to 0.6 A current input. From Figure 4.11, the measured
pressure drop values are shown smaller than the theoretical value when the current
input is lower than 0.8 A. However, during and above 0.8 A, the measurement results
from the experiment show higher pressure drop than the predicted one. The reason
can be related to the explanation of the off-state pressure drop behavior where the
viscosity value plays an important role in the accuracy of the predicted value. Because
although the graph in Figure 4.11 is showing the on-state performance of the valve,
the off-state pressure drop is actually still exists as an initial pressure drop. However,
the percentage of off-state pressure drop to the total on-state pressure drop will be
74
varied to the magnitude of the current input. Typically, the percentage of the off-state
pressure drop is the highest at the lowest current input but will be getting lower as the
current input is increased. For example, at the 73 ml/s theoretical curve, the percentage
of the off-state pressure drop is around 62 % for the total on-state pressure drop at
0.2 A current input but dropped to 23 % at 1.0 A current input. Therefore, it can be
considered that the effect of the base viscosity value in the total pressure drop is still
quite significant in the measured pressure drop below 0.8 A.
However, the superiority of the measurement results over the predicted results
in the current input higher than 0.8 A is quite surprising. Normally, the measured onstate pressure drop will be lower than the estimated pressure drop due to unpredicted
flux loss and prediction error in the field strength value of the FEM software. The
reason is not clear now, but several causes can be the source of the phenomenon. The
first cause can be assumed as a result of the thickening effect of the fluid, which in
several literatures were mentioned to be occurred due to spalling of the particles inside
the MR fluid [12, 22]. The thickening effect of the fluid is currently not considered in
the model since the MR valve model is based on the Bingham fluid model. However,
the thickening effect can be modeled if the Herschel-Bulkley fluid model is used as
the basis of the MR valve modeling [74]. The second cause is assumed due to the
enhanced yield stress phenomenon due to aggregation process [4], which occurred
when several single-chain structure of magnetic particle joined together and formed
a thick column. As discussed by Tao [5], the yield stress enhancement of MR fluid
can be done by changing the microstructure arrangement of magnetic particle during
on-state condition. As already shown in Figure 4.2, The MRF-132DG that is used in
this study has saturated yield stress value in around 48 kPa. However, the relationship
curve between yield stress and the field strength is derived based on the assumption
that the particles are arranged in a single chain structure [94]. If the microstructure
arrangement of magnetic particle inside the MR fluid is no longer arranged into a
single chain structure but rather a multiple chain structure, the yield stress value of the
MR fluid can be potentially enhanced. However, the validity of this assumption need
further confirmation since the occurrence of multiple chain structure that was discussed
in [4] and [5] is, so far, only reported in compression mode, which is possible due to
fluid-particle separation phenomena [171]. The yield enhancement in valve mode, to
the knowledge of the author, has not been reported yet. Nevertheless, the MR valve
concept with meandering flow path is already proving an improvement by delivering
higher achievable pressure drop than the previous concept of MR valve.
75
4.4.2
The testing results for all gap combinations at flow rate of 37 ml/s are shown
in Figure 4.12. It is shown that the pressure drop of all gap size combinations is
consistently increasing with the increase of current input. The slope, however, is
slightly gradual in the condition where the current input is lower than 0.2 A. The
difference in the slope is actually expected considering the nonlinearity of the fluid
yield stress curve provided by the fluid manufacturer as shown in Figure 4.2. From the
curve, it is shown that the yield stress is only appeared when the flux density value is
above 0.08 Tesla. Meanwhile, from the previous Section 3.4, it has been shown that
the flux density of the MR valve at the effective area with the excitation current to the
coil below 0.2 A are below 0.1 Tesla. Due to that reason, the slope of the pressure
drop curve for the MR valve with all gap size combinations below 0.2 A of current
excitation is not as steep as the slope of the curve above 0.2 A.
Furthermore, it can be seen that the smaller the gap size, the higher the pressure
drops of the MR valve. This effect is obvious, because the smaller the gap, the higher
the effect of wall friction to the fluid, which increased the viscous pressure drop and
field-dependent pressure drop simultaneously. However, a specific note should be taken
that the change of the gap size of the radial channel provide more significant effect to
the pressure drop than the gap size of the annular channel. This is the proof of the
simulated performances that have been presented Section 3.4, which can be explained
in two reasons. The first reason is regarding the number of the radial channel, which
is twice the number of the annular channel, while the second reason is regarding the
higher flux density at the radial channel than the flux density at the annular channel.
Another important aspect that can be evaluated from the data in Figure 4.13 is
the dynamic range of the MR valve. The dynamic range of MR valve is defined as the
ratio between the on-state and the off-state pressure drops of the valve as described in
Equation 3.15. Using the expression in Equation 3.15, the dynamic ranges of the MR
valve for each gap size combination are shown in Figure 4.13. The dynamic range is
an important variable as an indication that the valve can be controlled in a wide range
of performance. Ideally, a good MR device is expected to provide the highest on-state
performance while capable to suppress the off-state to the lowest value.
Based on the dynamic range values shown in the Figure 4.13, it can be
concluded that among four types of gap size combinations, the MR valve with 1.0
mm annular gap size will always have a higher dynamic range than the MR valve with
76
7
0
0
0.1
0.2
0.3
0.4
0.5
0.6
Current Input (A)
0.7
0.8
0.9
Figure 4.12 Comparison of measured peak pressure drop in various gap size
combinations
0.5 mm annular gap size. On the contrary, the increase of radial gap size, although not
as significant as the change in annular gap, has a weakening effect to the dynamic range
of the MR valve. Both of these effects are in agreement with the prediction shown in
Section 3.4. The knowledge of both factors, the achievable pressure drop value and the
dynamic range, will be useful to decide which gap size combination to be used when
designing an MR valve for a particular application.
4.4.3
Figures 4.14a to 4.14c detail the characteristics of valve with current input
variations at 0.50 Hz, 0.75 Hz and 1.00 Hz of frequency excitations. As predicted, the
MR valve with meandering flow path is consistently showing pressure drop adjustment
with the variation of the current input. As seen in Figure 4.14a, the peak pressure drop
is around 0.65 MPa at the current of 0.0 A with 0.50 Hz excitation frequency. When
the current increases in the interval of 0.1 A, the peak pressure drop are also steadily
increased with average increments of 0.58 MPa up to the maximum pressure drop of
around 6.42 MPa at the current of 1.0 A. Similarly, in Figure 4.14b, at the current of 0.0
77
Figure 4.13 Comparison of the MR valve dynamic range for each gap size
combinations
A with 0.75 Hz excitation frequency the maximum pressure drop that can be achieved
is around 0.95 MPa. However, with 0.1 A of current increment, the pressure drop are
normally increased in around 0.56 MPa until the maximum observed pressure drop
of 6.53 MPa is achieved at 1.0 A current. Lastly, at 1.00 Hz excitation frequency as
shown in Figure 4.14c, the maximum off-state pressure drop is around 1.35 MPa with
the average increments of pressure drop are seen around 0.55 MPa with each 0.1 A
increment of current so that the peak pressure drop at 1.0 A is measured at 6.86 MPa.
The relationships between peak pressure drop in each current increment for 0.50 Hz,
0.75 Hz and 1.00 Hz excitation frequency are depicted in Figure 4.15.
As shown in Figure 4.15, the rise of peak pressure drop is almost linear to
the increment of the current input. However, there is an interesting phenomenon that
can be observed from the curves where the slopes between 0.0 to 0.2 A and 0.8 to
1.0 A are slightly lower than the other regions. The phenomenon can be explained
using the curve of yield stress versus magnetic field strength of the MRF-132DG in
Figure 4.2 where the slope is leaner at the lower magnetic field strength value and also
leaner when it almost reached saturation in the higher magnetic field strength value.
78
8
1.0 A
0.9 A
0.8 A
0.7 A
0.6 A
0.5 A
0.4 A
0.3 A
0.2 A
0.1 A
0.0 A
0
2
4
6
8
40
30
20
10
0
Flow Rate (ml/s)
10
20
30
40
(a)
8
1.0 A
0.9 A
0.8 A
0.7 A
0.6 A
0.5 A
0.4 A
0.3 A
0.2 A
0.1 A
0.0 A
0
2
4
6
8
60
40
20
0
Flow Rate (ml/s)
20
40
60
(b)
8
1.0 A
0.9 A
0.8 A
0.7 A
0.6 A
0.5 A
0.4 A
0.3 A
0.2 A
0.1 A
0.0 A
2
0
2
4
6
8
80
60
40
20
0
Flow Rate (ml/s)
20
40
60
80
(c)
Figure 4.14 The pressure dynamics of MR valve at various current input for 0.5-0.5
mm (annular-radial) gaps configuration, (a) 0.50 Hz (b) 0.75 Hz (c) 1.00 Hz
79
8
6
4
2
0
2
4
6
8
0
0.50 Hz
0.75 Hz
1.00 Hz
0.1
0.2
0.3
0.4
0.5
Current Input (A)
0.6
0.7
0.8
0.9
Figure 4.15 The trend of peak pressure drop at various current input for 0.5-0.5 mm
(annular-radial) gaps configuration
Although, the declaration that the 1.0 A current input is already reached saturation
region would be too premature, since the magnetic simulation results in Section 3.2.3
showed that the peak magnetic flux density at 1.0 A current input is only around 0.85
Tesla. The saturation region of the yield stress, according to Figure 3.3 is started to
appear above 0.9 Tesla. However, the slope of the yield stress curve is already seen
to gradually decrease after the magnetic flux density reached 0.7 Tesla. In the other
words, it is valid to predict that the attempt to increase the current input above 1.0 A
will not give any significant improvement to the pressure drop performance of the MR
valve, due to the yield stress saturation of the MR fluid.
4.4.4
The relationship between valve pressure drop against flow rate of the MR fluids
at the coil current of 1.0 A at various frequencies is shown in Figure 4.16. The results
are the typical characteristics from the fifth cycle onwards since the characteristics of
the first until the fourth cycle are usually unstable and inconsistent as similarly reported
by Li et al. [172]. From Figure 4.16, it can be concluded that the peaks of the valve
pressure drop are increased with the increase of maximum flow rate. Similarly, the
shifting of peak pressure drop also shifts the width of the hysteretic region. In other
words, the width of the hysteretic region of pressure drop is highly dependent to the
value of peak flow rate as typically found in the characteristics of MR damper in its
analogous force-velocity pattern.
80
8
8
80
0.50 Hz
0.75 Hz
1.00 Hz
60
40
20
0
Flow Rate (ml/s)
20
40
60
80
There is an interesting note regarding the bend of the curve shape in each time it
passes through the 0 MPa of pressure drop. The bends of the curve were interpreted as
the pressure lag effect similarly with the ones reported in the behavior of MR dampers
by Yang [173] and Zhang et al [174] which caused by the uncompressed air pocket
that inevitably exists in the testing cell. The existence of air pocket inside the cylinder
creates force lag during the transition between the positive and negative direction of
the piston. In fact, these effect were not a unique case in the MR fluids device since its
appearance was also reported in the characteristics of a passive hydraulic damper [175].
In order to minimize the air pocket effect, a vacuum technique is often used in the
fluid filling process as well as the utilization of the accumulator to pre-compress the
air pocket in the damper assembly [123, 173]. However, since no report has strongly
mentioned about side effects of the air pocket to the overall damping characteristics
other than just the force lag, in this study, the pressure lag effect will not be taken into
consideration.
4.5
Summary of Chapter 4
81
type Fatigue Dynamic Test Machine. The experimental results are assessed in terms
of off-state pressure drop characteristics, on-state characteristics, the effect of current
variations and the effect of frequency excitation to the MR valve performance. The
measurement results of the off-state characteristics concluded that the prototype could
deliver off-state pressure drop in the range that is already predicted by the steadystate model, while the on-state characteristics of the prototype with both 0.5 mm
annular and radial gaps proved that the concept successfully increased the achievable
pressure drop of MR valve more than 6 MPa. The effect of flow rate and current
variations to the dynamic characteristics of the MR valve were also demonstrated
through the experimental assessment where the increment of flow rate and current
result in consistent increment of the pressure drop. In addition the hysteretic behavior,
which is also captured in the measurement, can be used as a source to develop the
dynamic model of the MR valve which will be covered in the next chapter.
CHAPTER 5
5.1
Introduction
5.2
n
X
i=0
ai x i , n = order of polynomial
(5.1)
83
where ai is the coefficient that should be empirically determined using the curve fitting
of the experimental data. In both [147] and [148],the changing trend of each coefficient
as a variable to the magnitude of current input is then approximated using linear
regression so that
ai = bi + ci I , i = 0, 1, ..., n
(5.2)
where the b and c are the tracking coefficient for the a as a function of current input, I.
By substituting the equation 5.2 to equation 5.1, the damping force can be represented
as:
F =
n
X
(5.3)
i=0
.
x
MR Damper model
MR Valve model
DP
Figure 5.1 The difference between MR damper model and MR valve model excitation
Adopting the method to the case of MR valve requires some changes in the
input and output variables, because there are some differences between the modeling
requirement of MR damper and MR valve as shown in Figure 5.1. Therefore, the
84
generalized form of the polynomial-based parametric model for the MR valve is as
follows:
P =
n
X
ai Qi , n = 6
(5.4)
i=0
where the P is the pressure drop of MR valve, similarly to the force F of MR damper
and Q is the volumetric flow rate of the MR fluid across the MR valve, similarly to the
piston velocity x of the MR damper. Sixth order of polynomial is also chosen to model
the MR valve with reference to [147] and [148].
In order to find the set of coefficient a, the experimental data from the frequency
excitation of 0.75 Hz is selected as a reference. There are 11 sets of curve that
representing the pressure drop response for current inputs of 0.0 to 1.0 A with interval
of 0.1 A. There are 1000 points of measurement for each curve which can be evenly
divided into two regions, the region with positive flow acceleration (lower loop) and
the region with negative flow acceleration (lower loop). Each region is then fitted with
the sixth order polynomial equation to find the specific sets of coefficient ai for each
curve. The results of curve fittings for both lower loop and upper loop are shown in
figure 5.2 and 5.3 respectively.
(5.5)
85
25
9
a *2*10
6
20
a *108
15
a4*6*10
Coefficient Value
10
5
0
5
10
15
0
0.1
0.2
0.3
0.4
0.5
0.6
Current Input (A)
0.7
0.8
0.9
0.7
0.8
0.9
(a)
35
a *1*104
30
a *2*102
2
Coefficient Value
25
a *10
1
a *1
0
20
15
10
5
0
5
0
0.1
0.2
0.3
0.4
0.5
0.6
Current Input (A)
(b)
Figure 5.2 Trend of the normalized coefficient values at the positive flow acceleration
(lower loop) to the variations of current input (a) a6 , a5 and a4 (b) a3 , a2 , a1 and a0
to the current input variability are shown in table 5.1 It can be seen that most of the
coefficient has non-linear correlation with the current input and therefore cannot be
represented by linear function.
Table 5.1: Correlation test results between the model coefficient a and current input I
Correlation coefficient
Model coefficient
Lower loop Upper loop
a0
-0.9863
0.9906
a1
0.8396
0.8511
a2
0.5486
-0.8251
a3
0.0068
-0.1176
a4
-0.1181
0.5095
a5
-0.2821
-0.1492
a6
-0.0722
-0.3090
86
25
9
a *2*10
6
20
a *108
15
a4*6*10
Coefficient Value
10
5
0
5
10
15
0
0.1
0.2
0.3
0.4
0.5
0.6
Current Input (A)
0.7
0.8
0.9
0.7
0.8
0.9
(a)
45
Coefficient Value
40
a3*1*10
a *2*10
35
a1*10
30
25
20
15
10
5
0
5
0
0.1
0.2
0.3
0.4
0.5
0.6
Current Input (A)
(b)
Figure 5.3 Trend of the normalized coefficient values at the negative flow acceleration
(upper loop) to the variations of current input (a) a6 , a5 and a4 (b) a3 , a2 , a1 and a0
(5.6)
n
X
i=0
bi I 4 + ci I 3 + di I 2 + +ei I + fi Qi , n = 6
(5.7)
87
Using the same curve fitting process, the fourth order polynomial coefficient
determination is conducted for a6 , a5 , a4 , a3 , a2 , a1 and a0 with respect to current input
variations. The specific values of bi , ci , di , ei and fi are listed in table 5.2.
Table 5.2: List of coefficients for the polynomial-based parametric MR valve model
Curve
Order of
Coefficient Values
Region Polynomial
b
c
d
e
f
6
-3.337e-08 1.156e-08 1.641e-08 8.011e-10 6.406e-10
5
-1.149e-06 8.699e-07 2.292e-07 -1.139e-07 8.383e-09
Lower
4
1.209e-04 4.760e-06 -1.053e-04 -7.317e-06 -3.218e-06
Loop
3
5.091e-03 -3.820e-03 -9.721e-04 1.661e-04 -3.785e-05
2
-0.0197
-0.224
0.239
0.0151
4.698e-03
1
-3.2792
-1.2161
4.9471
0.5715
0.2141
0
-170.0049
335.074
-193.1929
-4.9548
-1.6389
6
4.634e-08 -4.068e-08 -3.928e-09 -3.063e-10 -4.239e-10
5
-2.286e-06 2.938e-06 -8.588e-07 5.549e-08 1.565e-09
Upper
4
-1.715e-04 1.138e-04 5.632e-05 3.617e-06 2.291e-06
Loop
3
9.737e-03
-0.0123
3.502e-03 -5.047e-04 -9.416e-06
2
0.0514
0.1260
-0.2001
-8.136e-03 -3.707e-03
1
-7.7430
7.0455
0.6173
1.1776
0.1862
0
199.877
-373.947
211.433
1.279
1.9134
5.3
(5.8)
where the 0 , 1 , and 2 are the parameters of the model and z is the variable that can
be expressed as:
z = x
|x|
z
g (x)
(5.9)
88
where g (x)
is the additional function that depends on various factors such as material
properties and temperature.
(5.10)
z = x 0 a0 |x|
z
(5.11)
where a , 0 , 1 , 2 , b , and a0 are the parameters of the model while v is the voltage
input to the coil and the z is interpreted as the variable that expressed the average MR
fluids transient deformation generated when the direction of force is changing or also
known as the evolutionary variable.
The other modification to the LuGre model was also proposed by Jimenez and
Avarez-Icaza [138] in the following form:
F (t) = 0b zv + 1 z + 2a x
(5.12)
z = x 0b a0 |x|
z (1 + a1 v)
(5.13)
It can be seen that both forms use the voltage, v, as the input variable that is
related to the coil magnetization strength. In another model, the magnetization strength
is expressed with the magnitude of the current, i, given to the coil [127]. Physically
both can be considered as equal term given that the change of coil resistance during
magnetization is neglected. However, both current and voltage are just the indirect
variables to express the units of magnetization strength. In other words, the value of
voltage or current that is expressed in the model will not be universally applicable since
the strength of magnetic field that influences the MR effect is also highly subjected to
coil turns and dimensions.
89
of valve model. The modifications of the base model are also needed to reduce the
parameter and simplify the model. Therefore, the generalized form of the modified
LuGre based MR valve model proposed in this paper is as follows:
P = Az + Bz + CQ
(5.14)
z = Q a0 |Q| z
(5.15)
thus, there are five independent parameters that need to be identified, and these set of
parameters are defined as follows:
= [A, B, C, , a0 ]
90
14
A*20
B/5
C*2.5
*5
a0*5
12
10
8
6
4
2
0
0
0.2
0.4
0.6
Current Input (A)
0.8
5.4
91
excitation of 0.75 Hz at current inputs of 0.3 A, 0.6 A and 0.9 A are shown in Figure
5.5a to 5.5c respectively.
Generally, in these figures, the results from both models are in a good
agreement with the experimental data, especially in terms of the peak pressure drop
value. Though there are some compromises when the curve is reaching the pressure
drop of 0 MPa due to the air pocket effect as mentioned in section 4.4.4. These air
pocket effects, unfortunately, failed to be represented in both models and apparently
have been tampering the parameter values during the parameter identification process.
Although in general both model are able to represent the hysteretic behavior of
the MR valve, in order to compare the model accuracy, a specific measurement need to
be taken. In this study, the relative error, that is adopted from [127], is used to measure
the level of accuracy of each model. The relative error is expressed in the following
equation:
Pn
P exp Pimodel
i
RE = i=1 Pn
exp
|
i=1 |Pi
(5.16)
According to table 5.4, most of the results from both model are showing relative
92
3
-1
-2
Experimental Data
Polynomial-based Hysteresis Model
LuGre based Hysteresis Model
-3
-60
-40
-20
0
Flow Rate (ml/s)
20
40
60
(a)
6
-2
-4
Experimental Data
Polynomial-based Hysteresis Model
LuGre based Hysteresis Model
-6
-60
-40
-20
0
Flow Rate (ml/s)
20
40
60
(b)
8
-2
-4
-6
-8
-60
Experimental Data
Polynomial-based Hysteresis Model
LuGre based Hysteresis Model
-40
-20
0
Flow Rate (ml/s)
20
40
60
(c)
Figure 5.5 Comparison between the test data and the model results for various current
input, (a) 0.3 A (b) 0.6 A (c) 0.9 A
93
error less than 10%, which can be considered good since the results from the model
developed by [127] were reported to have even higher relative error in around 15%.
In average, the relative error of the polynomial-based hysteresis model is also lower
than the relative error of the LuGre based hysteresis model, which is demonstrating
the advantage of the polynomial model in terms of accuracy as already explained in
chapter 2.
-2
-4
-6
-8
-40
Experimental Data
Polynomial-based Hysteresis Model
LuGre based Hysteresis Model
-30
-20
-10
0
Flow Rate (ml/s)
10
20
30
40
(a)
8
-2
-4
-6
-8
-80
Experimental Data
Polynomial-based Hysteresis Model
LuGre based Hysteresis Model
-60
-40
-20
0
Flow Rate (ml/s)
20
40
60
80
(b)
Figure 5.6 Comparison between the test data and the model results for current input
of 1.0 A at various frequency excitations, (a) 0.50 Hz (b) 1.00 Hz
However, while it is obvious that the models, that have been specifically
identified using the 0.75 Hz experimental data, can mimic the 0.75 Hz measurement
data well, the real challenge is actually to perform at the frequency excitations other
than 0.75 Hz. As an example. the performance comparisons of each model with the
experimental data at 0.50 Hz and 1.00 Hz frequency excitations for current input
94
of 1.0 A are depicted in Figure 5.6. The results in these figures have shown that
the deviations are apparent for both models in the 0.50 Hz and 1.00 Hz frequency
excitations. However, the deviations of the polynomial-based hysteresis model is seen
much larger than the deviations of the LuGre based hysteresis model, which are also
shown by the comparison of relative error in table 5.5. These large deviations can be
explained by considering that the polynomial equations, at some values, will reach its
extreme points. In this case the extreme points of the pressure drop are visible at the
1.00 Hz frequency excitations, which occurred at the flow rate slightly below +/- 55
ml/s.
Table 5.5: Comparison of relative error at 0.50 Hz and 1.00 Hz frequency excitations
Current
Relative Error
Input
Polynomial-based model LuGre based model
(A)
0.50 Hz
1.00 Hz
0.50 Hz 1.00 Hz
0.0
0.1268
0.5871
0.0767
0.0674
0.1
0.1240
0.5132
0.1053
0.0950
0.2
0.1582
0.5029
0.0685
0.0544
0.3
0.1670
0.6500
0.0894
0.0507
0.4
0.1891
0.8125
0.1274
0.0803
0.5
0.1550
0.9269
0.0603
0.0682
0.6
0.1503
1.0091
0.0669
0.1066
0.7
0.1405
0.9001
0.0614
0.0757
0.8
0.1794
0.6162
0.1098
0.0423
0.9
0.2073
0.3196
0.1424
0.0821
1.0
0.3033
0.8669
0.1427
0.1369
Average 0.1728
0.7004
0.0955
0.0781
95
MR valve including actuator design and damper design as well as the development of
control system for suspension and vibration isolation devices that involves MR fluids
flow control.
5.5
Summary of Chapter 5
The hysteretic modeling of the MR valve has been discussed in this chapter.
The hysteretic modeling is presented in two types of parametric modeling approaches,
namely the polynomial-based hysteresis modeling approach and the modified LuGrebased hysteresis modeling approach. The performance of these two approaches in
modeling the hysteretic behavior of the MR valve is compared with the experimental
data. According to the performance comparison results, it can be concluded that
the polynomial-based hysteresis model has better accuracy than the LuGre-based
hysteresis model but only at the range of inputs that is recognized during the
identification process. In the wider range of inputs, the LuGre-based model show better
capability to consistently mimic the behavior of MR valve than the polynomial-based
model. Moreover, the LuGre based model has fewer parameters to identify than the
polynomial based model so that the identification process is relatively simpler.
CHAPTER 6
6.1
Conclusions
6.1.1
In this study, the new concept of MR valve has been introduced using the
combination of multiple annular and radial gaps to form a meandering flow path. The
meandering flow path approach is used to increase the path length of MR fluid flow
inside the valve so that the opportunity of the MR fluid to be exposed with magnetic
field is higher. In order to predict the performance of the new MR valve concept, the
97
steady-state model of the MR valve with multiple annular and radial valve is derived.
According to the simulation results of the steady-state model, the new MR valve with
meandering flow path concept is able to reach pressure drop of around 6 MPa in a
comparable valve size and power consumption of a conventional 2.5 MPa annular MR
valve. The performance of the MR valve was also analyzed by assessing the effect of
current input variations, flow rate variations and gap size variations to the generated
pressure drop of the MR valve.
6.1.2
6.1.3
The predicted results of the steady-state model are well confirmed by the
experimental work with small deviations in the off-state pressure drop value. The
experimental work is conducted using a double rod hydraulic cylinder, as the testing
cell, actuated by the dynamic test machine. The experimental test was conducted in
three different frequency excitations with constant amplitude of 25 mm to emulate the
variations of peak fluid flow rate. In each frequency excitations the current input is
varied from 0.0 A to 1.0 A with 0.1 A increment. Using the reciprocating sinusoidal
movement from the dynamic test machine, the fluid flow rate across the MR valve
98
is also varied in sinusoidal value, which allow the observation of hysteretic pressure
drop behavior. According to the experimental results, the hysteretic pressure behavior
occurs when there are different measured pressure drop at the same flow rate with
non-zero flow acceleration at different flow direction.
6.1.4
Since the verified hystetic model is a very good representation of the MR valve,
the model can be used to design the control system for further applications of MR
valve. In this study, the process of designing a pressure tracking control system using
PI controller is presented as a demonstration on how the MR valve model can be used
to serve various applications. The assessment of performance of the pressure tracking
control is evaluated by introducing three different types of reference input, which
representing the class of continuous and discontionuous functions, to the controller.
The simulation results have shown that the controller is able to track these three signals
well. Further, the pressure tracking control system can be used as the embedded inner
loop controller of the MR valve that act to ensure that the generated pressure drop of
an MR valve can be accurately delivered as commanded by the outer loop controller to
serve the application demands.
99
6.2
In this research, the new way to improve the achievable pressure drop of MR
valve has been demonstrated with the meandering flow path concept. The new concept
has been studied thoroughly by evaluating the MR valve behavior in steady-state and
dynamic environment. The results that were obtained from the study showed that the
new concept can be potentially used in various applications that requires flow control
as well as enhancing the performance of MR devices that utilize valve as one of its
main component. Nevertheless, the contributions of this research are listed as follows:
(a)
(b)
The second contribution of this research was related to the knowledge on the
effect of gap size selection to the achievable pressure drop of the valve. The
specific selections of the gap size have a very significant effect to the pressure
drop characteristics of the valve where the annular gaps have shown to be
responsible in the determination of off-state pressure drop while the radial gaps
were highly influencing the achievable on-state performance. In terms of valve
dynamic range, it has been confirmed in both simulation and experimental work
that the reduction of annular gap size is highly affecting the degradation of valve
dynamic range while in contrary occured with the reduction of radial gap size,
although the effect is not significant. Further, the knowledge on how to select
the size of both gaps are very important in the sizing process for particular
applications.
(c)
The third contribution was regarding the development of new parametric model
based hysteretic MR valve model. The hysteresis was identified during the
experimental assessment of MR valve. The model is the first parametric model
derived specifically to model the hysteretic behavior of MR valve.
100
6.3
6.3.1
There are wide ranges of applications that can utilize the advance features of
MR valve, since most of actuators that rely on flow control mechanism will require
valve as their main control element. However, in order for the MR valve to be
well functioned as the final control element of an actuator, the pressure drop that
is generated by the valve has to be controllable accurately. Therefore, the pressure
tracking control system is needed as the valve inner-loop controller and integrated
module. As a case study for future reference, the explanation of the basic structure
of the pressure tracking controller and its simulated performance using the MR valve
hysteresis model in the case of tracking various types of input references are presented.
Desired
pressure
drop
+
Error
signal
PI
Controller
Control
signal
Current
Driver
Current
input
MR
Valve
Flow
rate
Actual
pressure
drop
The basic structure of the pressure tracking control of the MR valve are shown
in figure 6.1, which depicts a closed-loop control system with PI controller. The PI
controller is chosen because it is simple, easy and widely recognized as the most
common controller in industrial application [178]. The input of the controller is the
error signal, which is acquired from the subtraction between the reference signal and
the feedback signal. The reference signal in the case of MR valve pressure tracking
control is the desired pressure drop and the feedback signal is the actual pressure drop.
Therefore, the formulation of the PI controller for the pressure tracking control can be
expressed as follows
Z
u (t) = Kp e (t) + Ki
e (t)
(6.1)
(6.2)
101
where the u(t) is the control signal which drives the current driver to charge the
electromagnetic coil of the MR valve with current, the Kp and Ki are the proportional
and integral gain respectively as well as the Pd and Pa as the desired pressure drop
and actual pressure drop.
The current driver is needed to adjust the value of the output signal from the
controller with the signal input that can drive the MR valve. Since the control signal
will not always satisfy the signal requirement of the MR valve. For example, the value
of Pa can be higher or lower than the value of the Pd . Therefore, the value of e(t)
can be positive or negative which concurrently cause the control signal u(t) to be in
positive or negative value. However, the positive of negative sign of the current input
to the electromagnetic coil of the MR valve in this study practically has no effect to the
magnitude of the pressure drop. Therefore, it is the task of the current driver to adjust
the sign, which simply can be done using absolute function.
In order to assess the performance of the pressure tracking control, the LuGrebased hysteresis MR valve model is used with several types of reference signal, namely
the sinusoidal signal, the pulse signal and the saw-tooth signal. The selection of these
signals as the reference input are based on the assumption that if the system can
track these signal well, then the system is considered robust to track various types
of continuous and discontinuous functions [148]. The simulation results of pressure
tracking control under various reference functions are shown in figure 6.2. The results
are the outcome of pressure tracking control with controller parameters of Kp = 0.08
and Ki = 4, which acquired from MATLAB auto-tuning tools, and Kp = 0.05 and
Ki = 2, which acquired from trial and error method. From figure 6.2, it can be
concluded that the pressure tracking control of the hysteresis model of MR valve is
able to track various types of continuous and discontinuous functions well.
Although the explanation of the pressure tracking control in this section only
involved a closed-loop PI controller, the further application of inner-loop controller
of the MR valve is not just limited to the PI controller. Various control strategy can
be developed and integrated with the MR valve. The demonstrated pressure tracking
control system using PI controller shown in this section is merely an illustration of
the kind of applications that the MR valve can serve. With the involvement of the
inner-loop controller, the MR valve can serve both semi-active and active application
accurately. Considering the necessity of the study, the further exploration of the
pressure tracking control system of the meandering type MR valve is recommended
for future works.
102
5.5
Desired Pressure Drop
Controlled Pressure Drop (PI Autotuning)
Controlled Pressure Drop (Trial and Error)
4.5
3.5
2.5
2
6
Time (s)
10
(a)
5.5
Desired Pressure Drop
Controlled Pressure Drop (PI Autotuning)
Controlled Pressure Drop (Trial and Error)
4.5
3.5
2.5
2
6
Time (s)
10
(b)
5.5
Desired Pressure Drop
Controlled Pressure Drop (PI Autotuning)
Controlled Pressure Drop (Trial and Error)
4.5
3.5
2.5
2
6
Time (s)
10
(c)
Figure 6.2 Simulation results of pressure tracking control under various functions as
reference, (a) Sinusioidal (b) Pulse (c) Saw-tooth
103
6.3.2
There are also several problems encountered that were not able to be discussed
in this study and can be recommended for future works as a continuation of this study.
These problems are as the following:
(a)
The conceptual design of the MR valve with meandering flow path was not
considering any optimization on valve design. The gap sizes of annular and radial
flow channel, the coil properties, the MR fluid types and the overall valve size in
this study were selected merely to serve the observation purposes and therefore
not considering any optimum value. In the future, it will be necessary to put the
consideration on the optimization on these variables such as to conduct the study
on the critical thickness of the valve casing with respect to magnetic reluctance
and mechanical strength to withstand the internal pressure of the valve.
(b)
(c)
(d)
The meandering flow path structure that consists of the combination of annular
and radial flow channel inevitably formed many sharp corner similarly with the
effect of elbow. The fluid flow in elbow, in general, has been widely known as a
complex phenomenon due to frictional and separation effects [179,180]. Though
104
for the flow case of non-Newtonian fluid the friction loss coefficient has been
shown lower than in the case of Newtonian fluid [181], the losses in a small flow
channel such as in an MR valve may become significantly large [124]. Therefore,
it may be necessary in the future to consider the elbow effect in the meandering
flow path structure in the pressure drop characteristics of MR valve to further
examine the concept.
(e)
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
106
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
107
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
108
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
109
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
110
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
[73]
[74]
[75]
Liao, C. R., Zhao, D. X., Xie, L. and Liu, Q. A design methodology for a
magnetorheological fluid damper based on a multi-stage radial flow mode.
Smart Materials and Structures, 2012. 21(8): 085005.
Dyke, S., Spencer, B., Sain, M. and Carlson, J. Modeling and control
of magnetorheological dampers for seismic response reduction. Smart
Materials and Structures, 1996. 5(5): 565575.
Milecki, A. Investigation and control of magnetorheological fluid dampers.
International Journal of Machine Tools and Manufacture, 2001. 41(3): 379
391.
Li, W. H., Du, H. and Guo, N. Q. Design and testing of an MR
steering damper for motorcycles. The International Journal of Advanced
Manufacturing Technology, 2003. 22(3-4): 288294.
Milecki, Andrzej and Sedziak, Dariusz and Ortmann, Jaroslaw and Hauke,
M. Controllability of MR shock absorber for vehicles. International Journal
of Vehicle Design, 2005. 38(2/3): 222233.
Yu, M., Liao, C. R., Chen, W. M. and Huang, S. L. Study on MR semiactive suspension system and its road testing. Journal of Intelligent Material
Systems and Structures, 2006. 17(8-9): 801806.
Batterbee, D. C., Sims, N. D., Stanway, R. and Rennison, M.
Magnetorheological landing gear: 2. Validation using experimental data.
Smart Materials and Structures, 2007. 16(6): 24412452.
Huang, J., He, J. and Zhang, J. Viscoplastic flow of the MR fluid in a
cylindrical valve. Key Engineering Materials, 2004. 274-276: 969974.
Sahin, H., Liu, Y., Wang, X., Gordaninejad, F., Evrensel, C. and Fuchs,
A. Full-scale magnetorheological fluid dampers for heavy vehicle rollover.
Journal of Intelligent Material Systems and Structures, 2007. 18(12): 1161
1167.
Guo, P., Guan, X. and Ou, J. Physical modeling and design method of the
hysteretic behavior of magnetorheological dampers. Journal of Intelligent
Material Systems and Structures, 2014. 25(6): 680696.
Nguyen, Q., Choi, S. and Wereley, N. Optimal design of magnetorheological
valves via a finite element method considering control energy and a time
constant. Smart Materials and Structures, 2008. 17(2): 025024.
Bai, X.-X., Wang, D.-H. and Fu, H. Principle, modeling, and testing of an
annular-radial-duct magnetorheological damper. Sensors and Actuators A:
Physical, 2013. 201: 302309.
Wang, D. H. and Liao, W. H. Magnetorheological fluid dampers: a review of
parametric modelling. Smart Materials and Structures, 2011. 20(2): 023001.
Yang, G., Spencer, B. F., Carlson, J. D. and Sain, M. K. Large-scale MR fluid
111
[76]
[77]
[78]
[79]
[80]
[81]
[82]
[83]
[84]
[85]
[86]
[87]
[88]
112
[89]
[90]
[91]
[92]
[93]
[94]
[95]
[96]
[97]
[98]
[99]
[100]
[101]
[102]
113
[103]
[104]
[105]
[106]
[107]
[108]
[109]
[110]
[111]
[112]
[113]
[114]
114
[115]
[116]
[117]
[118]
[119]
[120]
[121]
[122]
[123]
[124]
[125]
[126]
[127]
115
[128]
[129]
[130]
[131]
[132]
[133]
[134]
[135]
[136]
[137]
[138]
116
[139]
[140]
[141]
[142]
[143]
[144]
[145]
[146]
[147]
[148]
[149]
[150]
[151]
[152]
117
[153]
[154]
[155]
[156]
[157]
[158]
[159]
[160]
[161]
[162]
[163]
[164]
[165]
118
[166]
[167]
[168]
[169]
[170]
[171]
[172]
[173]
[174]
[175]
[176]
[177]
[178]
119
[179]
[180]
[181]
APPENDIX A
CAD DRAWINGS
ISOMETRIC VIEW
LEFT VIEW
4X M4X0.7
1.09
Q.A
MFG
APPV'D
CHK'D
DRAWN
FI
NAME
38
FINISH:
DATE
WEIGHT:
MATERIAL:
SECTION B-B
SIGNATURE
16 + 0.02
22.64
16.98
3.5
O-ring Groove
DEBUR AND
BREAK SHARP
EDGES
10.0
15
SCALE:1:1
DWG NO.
TITLE:
REVISION
Coil Bobbin
SHEET 2 OF 2
A2
A4
3.5
40 - 0.02
4X M4X0.7
RIGHT VIEW
Alumunium
10
121
12
TOP VIEW
ISOMETRIC VIEW
3/8 inch
15
12
Q.A
MFG
APPV'D
CHK'D
DRAWN
FI
NAME
DATE
WEIGHT:
MATERIAL:
SECTION A-A
FINISH:
SIGNATURE
15
50
DEBUR AND
BREAK SHARP
EDGES
24
5 - 0.02
19 - 0.02
SCALE:1:1
DWG NO.
TITLE:
16 - 0.02
REVISION
SHEET 1 OF 2
A1
50
40 + 0.02
Valve Casing
BOTTOM VIEW
A4
122
49
39
50
123
22
A
50
15
6+
12.5
READ
11.5
0.0
EMA
LE TH
24
1/4 F
SPP
11 (B
10
44
40 + 0.1
14 + 0.02
FI
DEBUR AND
BREAK SHARP
EDGES
FINISH:
SIGNATURE
16 - 0.02
CUT-FACE A-A
6 + 0.02
11.8
5 - 0.02
DATE
REVISION
TITLE:
29/4/13
CHK'D
APPV'D
MFG
Q.A
MATERIAL:
WEIGHT:
Casing2_rev_ver2
DWG NO.
SCALE:1:1
SHEET 1 OF 1
A4
49
124
50
22
A
46
50
50
.02
6+0
15
23
E THR
EAD
12
EMA
L
11.8
DEBUR AND
BREAK SHARP
EDGES
FINISH:
SIGNATURE
16 - 0.02
CUT-FACE A-A
6 + 0.02
5 - 0.02
DRAWN
1/4 F
11 (B
10
SPP
40 + 0.1
25 + 0.02
DATE
REVISION
TITLE:
29/4/13
CHK'D
APPV'D
MFG
Q.A
MATERIAL:
DWG NO.
Casing2_rev_ver2_2
SCALE:1:1
SHEET 1 OF 1
A4
20
TOP VIEW
ISOMETRIC VIEW
70
Q.A
MFG
APPV'D
CHK'D
DRAWN
FI
NAME
FINISH:
DATE
WEIGHT:
MATERIAL:
SECTION A-A
SIGNATURE
1 + 0.03
2.5 - 0.02
3 - 0.02
R6.5 + 0.03
R5.5
DEBUR AND
BREAK SHARP
EDGES
Alumunium
SCALE:3:1
DWG NO.
TITLE:
B1
REVISION
SHEET 1 OF 2
0.5
16 - 0.02
Bush 0.5mm
BOTTOM VIEW
A4
125
TOP VIEW
ISOMETRIC VIEW
70
20
Q.A
MFG
APPV'D
CHK'D
DRAWN
FI
NAME
FINISH:
DATE
WEIGHT:
MATERIAL:
SECTION B-B
SIGNATURE
1 + 0.03
2 - 0.02
3 - 0.02
R6.5 + 0.03
R5.5
DEBUR AND
BREAK SHARP
EDGES
Alumunium
SCALE:3:1
DWG NO.
TITLE:
0.5
REVISION
SHEET 2 OF 2
B2
16 - 0.02
Bush 1mm
BOTTOM VIEW
A4
126
ISOMETRIC VIEW
Q.A
MFG
APPV'D
CHK'D
DRAWN
FI
NAME
FINISH:
DATE
15.5
SIGNATURE
5 - 0.02
WEIGHT:
MATERIAL:
DEBUR AND
BREAK SHARP
EDGES
10.5
SCALE:3:1
DWG NO.
TITLE:
REVISION
SHEET 1 OF 4
C1
5 - 0.02
13 - 0.02
Side Core 1
TOP VIEW
A4
127
Q.A
MFG
APPV'D
CHK'D
DRAWN
FI
NAME
FINISH:
SIGNATURE
5 - 0.02
DATE
WEIGHT:
MATERIAL:
DEBUR AND
BREAK SHARP
EDGES
SCALE:3:1
DWG NO.
TITLE:
REVISION
SHEET 1 OF 2
D1
5 THRU ALL
16 - 0.02
Orifice Core
A4
128
Q.A
MFG
APPV'D
CHK'D
DRAWN
FI
NAME
FINISH:
SIGNATURE
4 - 0.02
5 - 0.02
DATE
WEIGHT:
MATERIAL:
DEBUR AND
BREAK SHARP
EDGES
SCALE:3:1
DWG NO.
TITLE:
REVISION
13 - 0.02
13 - 0.02
SHEET 2 OF 2
D2
Middle Core
A4
129