The purposes of this work are to plan, theoretical account and imitate a magnetorheological brake system that has benefits over the conventional braking system. Basically, the drum-type MRB consists of revolving disc immersed in a MR fluid and enclosed in an electromagnet spiral. The magnetic spiral causes hardening of the MR fluid so that the shear stres between the traveling portion and inactive portion increases ensuing in the decrement velocity of the traveling portion. The shear emphasis can be varied by using assorted electric current to the twist spiral. In this paper, the survey began from the portion design utilizing 3D mold package, following with the finite component simulations affecting magnetostatic analysis. The mathematical theoretical account is derived based on drum-type MR brake. The simulation and experimental survey on the fillet clip and braking torsion were besides performed and compared each other. Several public presentations in footings of torsion versus clip and velocity decrease versus clip were investigated by using assorted electric current.
Vehicle public presentation, safety and cost have been going a major focal point in automotive industry for many old ages due its possible betterment. In footings of cost and safety, automotive applied scientists were struggled to develop high safety vehicle with low production cost. A common attempt to accomplish this purpose is by minimising the figure of parts used in vehicle. X-by-wire is one of subjects as a solution for this recent automotive issue.
In a vehicle, x-by-wire has been employed in several sections such as maneuvering and braking. It means replacing conventional mechanical constituents by electrical 1s ( Park et al. , 2006 ) . This treatment relates with the braking system as one of x-by-wire execution that the so called brake-by-wire system. The chief end of this survey is the development of an actuator for brake-by-wire system that employs magnetorheological ( MR ) fluid.
Brake-by-wire system can be realized by replacing the mechanical constituents which connect the brake units on each wheel and the brake pedal with electrical constituents. Harmonizing to Park et Al. ( 2006 ) , the first coevals of brake-by-wire system still use the conventional hydraulic brake system for fail safety demands. All the brake control maps are implemented in one chief electronic control unit. A hydraulic system is needed for safety grounds to guarantee braking in the instance of an electrical failure.
There are several benefits of implementing brake-by-wire system. The hold between the clip the brake pedal pressed by the driver and the corresponding brake response of a conventional hydraulic brake exhibits 200-300 MS ( Falcao da Luz, 2004 ) . It is due to the force per unit area build-up within the hydraulic lines. The usage of electric brake system has the possible to cut down this clip hold drastically, resulted a decrease in braking clip and distance. Besides that, the belongingss and behaviors of the brake will be easy to accommodate by merely altering package parametric quantities and electrical end products alternatively of seting mechanical constituents ( Park et al. , 2007 ; Karakoc et al. , 2008 ) .
In the initial phase of this paper, the rule work of MR fluid will be briefly explained. MR fluid is a sort of suspension contains fine Fe atoms which have diameter of 20-50 micrometers. The fluid stiffens in the presence of magnetic Fieldss in a fraction of msecs. When it is released from the magnetic field, the fluid shows like a Newtonian fluid behavior. This irreversibility is frequently used for semi-active devices that need hydraulical and rheological behaviors such as semi-active quiver absorbers ( Carlson, 2001 ) .
It is of import to observe that this work utilizes MR fluid as a portion of the brake-by-wire actuator. In fact, there is another type of fluids viz. Electrorheological ( ER ) fluid which has the same rule work with the MR fluid. ER fluid is besides a additive syrupy liquid whose rheological behavior alterations under the influence of an applied electric field, alternatively of a magnetic field. However, there are many drawbacks to ER fluid, including comparatively little rheological alterations and utmost belongings alterations in temperature. Due to the drawbacks of ER fluids that require high control electromotive forces, incapableness in bring forthing high shear forces and are susceptible to contaminations, they are non ideally suited for automotive applications ( Craft et al. , 2003 ; Jolly et al. , 1998 ) .
The application of a MR fluid in braking system is comparatively a recent subject. Some old plants on MR brake can be found in literatures. Park et Al. ( 2006 ) presented a design optimisation process utilizing fake tempering combined with finite component simulations affecting magnetostatic, unstable flow and heat transportation analysis. In 2007, Park et Al. continued their surveies on MR fluid choice for MR brake application, magnetic circuit design and torsion demands for automotive application. Followed by Karakoc et Al. ( 2008 ) , in which the work was focus on probe of practical design standards for MR brake such as stuff choice, sealing, working surface country, syrupy torsion coevals and MR fluid choice for basic automotive braking system.
However, the kineticss of the MR brake is non clearly explored peculiarly the MR braking speed which relates with halting clip response and besides its braking torsion in assorted applied inactive electric current. These standards are of import for velocity and torsion controls application. This paper discusses about the MR brake behavior in footings of its speed and braking torque responses in assorted electric currents. This will be performed by revolving an inertial organic structure with a certain mass at a coveted speed, and so the coupled MR brake is energized to cut down angular speed of this organic structure until zero velocity quickly by utilizing the MR braking torsion.
The lineation of this paper is as follows. In subdivision 2, the description on rule work and proposed design of drum-type MR brake are presented. Section 3 explains the trial rig installation and measuring of the MR brake parametric quantities. Section 4 is the mathematical modeling of the MR braking speed and torque relationships. Section 5 contains the simulation of magnetic field utilizing finite component package. Section 6 discuss about the speed responses of the MR brake under assorted input parametric quantity. Section 7 explains the braking torsion generated under assorted changeless electric currents. Finally, this paper is enclosed with decision and recommendation.
Description of the MR brake
This subdivision describes the constellation of MR brake and its rule work. First, it will be briefly explained the rule work of MR brake. As it was introduced in Section 1, ER and MR fluids possess the belongings of altering their viscousness as an electric or magnetic field is applied. This altering viscousness is in bend relative to the clash exerted on anybody traveling within the fluid. Hence, the application to brakes consists of holding a disc immersed in MR fluid. The braking torsion exerted on the disc is so controlled the magnetic field applied to the fluid. In automotive application, the magnetic field is varied harmonizing to the driver ‘s force per unit area on the brake pedal. Normally, an electromagnet will be used to bring forth the magnetic field ( Wang and Gordaninejad, 1999 ; Falcao district attorney Luz, 2004 ) .
The conventional diagram of the cross-section MR brake is presented in Figure 1. Following, the assembly of the MR brake is described as follows. This MR brake consists of three chief parts viz. ; a traveling rotor ( A ) , a inactive organic structure or lodging ( B ) and a twist spiral ( C ) . Parts A and B are chiefly made of mild-steel, while the spiral C is normally a bronze wire as used for DC motor spiral. The rotor is stiffly fitted to the brake lodging via two deep channel ball bearings. The membranophone active diameter is 100 millimeter, while for the brake lodging ; its internal diameter is 103 millimeter. The annulate spread between inactive organic structure and membranophone is 3 millimeter. For the propulsion of burden inactiveness, a mild steel round rod and a mild steel shaft from AC motor are coupled by a jaw yoke. In the other terminal of MR brake rotor is besides coupled with the inertial burden of 10 kilograms by a piece of jaw yoke.
C
Bacillus
A
Figure 1 Schematic of the proposed MR brake
The inactive organic structure has the chief map of enveloping the MR fluid inside the brake paradigm. It has to forestall the fluid from leaking. To avoid unstable leaking from the traveling portion, 4 viton lip seals a located besides these ball bearings. For measuring demands, the inactive organic structure is designed to be freely singing relations to the rotor. In fact, the inactive organic structure is fixed to the detector during operations. The inactive organic structure is attached into two roller bearings which are mounted in the rig frame. On the other manus, the lodging is used for attaching the electromagnetic spiral every bit good as its base. The bronze wire has the diameter of 0.3 millimeters and the twist of 400 bends is proposed for the electromagnetic spiral.
The samples of tried MR fluids in this survey are hydro-carbon based MRF-132AD fluid manufactured by Lord Corporation. This fluid has a about additive experimental emphasis rate curve that is good approximated by the Bingham theoretical account. This fluid type was chosen chiefly due to its higher temperature opposition features ( Falcao da Luz et al.,2004 ) .
Description of The Test Facility
Figure 2 illustrates the MR brake trial rig installation developed by Smart Material and Automotive Control research group and available at the Autotronic Lab. , UTeM. In Figure 2, an AC motor is coupled to the input shaft or rotor of the MR brake via an A-type V-belt. The belt is tensioned to drive the rotating system until making the desired speed and is released when the electric current is applied.
The lodging of this MR brake is coupled to a burden cell via an arm which has the length of 238 millimeter. In this equipment, the burden cell is employed for braking torque measuring. A velocity detector that normally used as ABS velocity detector is utilized for mensurating drum rotational velocity. The MR brake trial rig is equipped with an I/O device for informations processing. The Integrated Measurement and Control ( IMC ) device provides signal processing of the centripetal system. These signals are digitally processed and stored in a personal computing machine utilizing FAMOS control package ( IMC, 2002 ) . IMC device is connected to the personal computing machine utilizing NetBEUI protocol ( Scholz, 2000 ) . A DC power supply manufactured by GW-INSTEK is used for providing inactive electric currents to the MR brake electromagnetic spiral.
LOAD SENSOR
LOAD SENSOR
MR BRAKE
INERTIAL LOAD
Figure 2 Mechanical assembly of the MR brake trial rig
Mathematical Model of MR Brake
The indispensable magnetic field dependent unstable features of MR fluids can be described by a simple Bingham plastic theoretical account ( Philips, 1969 ) . By utilizing the constituent equation for a Bingham plastic fluid, the entire shear emphasis ( ? ) is stated as follows.
( 1 )
where is the output emphasis due to the applied magnetic field H, is the changeless plastic viscousness which is considered equal to the no-field viscousness of the fluid, and is the shear strain rate. Here, the plastic viscousness is defined as the incline between the shear emphasis and shear emphasis rate, which is the traditional relationship for Newtonian fluids.
Based on Eq. ( 1 ) and the given geometrical constellation shown in Figure 1, the braking torsion which is caused by the clash on the interfaces between the MR fluid and the solid surface within the MR brake can be written as ( Park, 2006 ) ;
( 2 )
Where N is the figure of surfaces of the brake disc in contact with the MR fluid, and are the inner and outer radii of the brake disc, severally. The output emphasis and shear strain rate are defined as ;
and
where ? is the angular speed of the revolving disc, H is the thickness of the MR fluid spread, H is the magnetic field strength, and K and ? are changeless parametric quantities that approximate the relationship between the magnetic field strength and the output emphasis for the MR fluid. Then, Eq. ( 2 ) can be rewritten as ;
( 3 )
The applied magnetic field H can be generated inside the MR brake when current I is presented to the electromagnet spiral as,
( 4 )
where is a relative addition. By executing the integrating in Eq. ( 3 ) and replacing Eq. ( 1 ) , it can be understood that the ensuing braking torsion is contributed by two torque component. First is torque due to the output emphasis induced by the applied magnetic field ( TH ) and another is torque due to the clash and viscousness of the MR fluid ( T? ) . Both torque elements are expressed as follows ( Li and Du, 2003 ; Park et al. , 2006 ) ) .
( 5 )
( 6 )
where is the rotational velocity of the disc. In other manner, the entire braking torsion outputted by the MR brake can be written as follows,
( 7 )
Next, the net or effectual MR braking torque response is derived by utilizing the free organic structure diagram as shown in Figure 3.
During rotary motion, the inertial burden will bring forth a changeless burden or falling burden as stated by Tan et Al. ( 2007 ) . Mention to Figure 3, the TL term is the lading torsion that acts on the end product rotor of the MR brake. This burden torsion is generated due to the weight of the burden mass ( m ) about the radius ( Rhode Island ) of the burden membranophone. The TL term is mathematically expressed in Eq. ( 8 ) . Physically, the entire end product MR braking torsion ( Tb ) is used to get the better of the torsion generated by the falling burden ( TL ) . This will ensue a net torsion which is used to slow the kineticss of all inactiveness ( J ) coupled stiffly to the membranophone shaft of the MR brake. The falling burden ( TL ) can be written as,
( 8 )
By detecting the free organic structure diagram, the MR effectual braking torsion is shown in Eq. ( 9 ) as follows.
( 9 )
?MR
Jpulley
Jsprocket
Jbearings
Jbearings
Thallium
Terbium
Figure 3 The free organic structure diagram of the MR brake shaft with its inertial burden
In Eq. ( 9 ) , the entire minute of inactiveness of the MR brake ( J ) consists of rotor membranophone shaft, four bearing inner parts, a sprocket, a block and the inertial burden. Therefore, the entire end product inactiveness ( J ) of the MR brake is expressed in the Eq. ( 10 ) .
( 10 )
The entire minute inactiveness is about about 0.932 kilograms M2.
By replacing Eqs. ( 7 ) and ( 8 ) into Eq. ( 9 ) , the Eq. ( 11 ) can be written as follows.
( 11 )
Magnetic Field Simulation
To accurately qualify the MR brake behaviour, a simulation based finite component theoretical account ( FEM ) is developed utilizing ANSYS MAXWELL. This theoretical account is a multi-physics theoretical account that accounted for magneto atmospherics within the MR brake. The finite component analysis in this survey consists of a magnetostatics simulation which gives the form of magnetic flux denseness and magnetic flux lines. This is utile for the get downing measure of design in which the interior decorator can gauge the magnetic field distribution within the MR brake.
The first measure in the finite component mold is to specify the brake geometry. Since the job is axisymetric, intending that the geometry, stuff belongingss and all tonss are all consistent along the digressive way, merely the cross-sectional is modeled ( Park, 2006 ) . This manner, the solution becomes that a planar job, leting the usage of ANSYS MAXWELL plane elements. This method reduces the computational cost of each simulation. Figures 4, 5 and 6 present the preliminary consequence of the finite component simulation.
Figure 4 shows the magnetic flux lines distribution in drum-type constellation. The lines represent the distribution of magnetic flux covered country within the MR brake. From the figure, the spread around the membranophone perimeter can be influenced good by the electromagnet spiral. The consequence is simulated based on 1.5 Ampere of applied electric current to the spiral. From the figure, it can be seen that the flux lines become weaker when it is far from the magnet beginning. The colour line shows the different value of the magnetic flux in [ Wb/m ] . The highest value of the magnet flux line is the ruddy line that has the magnetic flux of 4.6909 mWb/m. In that country, the Relatively Movable Poles ( direct-shear manners ) of magnetorheological fluid are utilised because of the high concentration of the flux line.
Figure 4 Flux magnetic lines distribution
Figure 5 shows the magnetic flux denseness within the MR brake by using 1.5 Ampere of electric current to the spiral. Equal to the flux line tendency, the magnetic field strength becomes weaker when it is far from the magnet beginning. The magnetic field strength is measured in [ A/m ] and indicated by different colourss. The highest value of the magnetic strength is 7.4046kA/m which is near the beginning or at the spread between membranophone and spiral wall, while the lowest magnetic field strength is 0.4293kA/m.
Finally, the finite component simulation presents the magnetic flux denseness which is shown in Figure 6. The magnetic flux denseness is measured in Tesla ( T ) . This parametric quantity will act upon the magnetic shear emphasis within the magnetic country. The more current applied to the spiral, the bigger magnetic flux denseness produced. By using 1.5 electric current, the highest magnetic flux denseness field is 9.6565e-1 Tesla, while the lowest value is 1.0593e-2 Tesla.
Figure 5 The form of the magnetic flux intensity/strength
Figure 6 The form of the magnetic flux denseness
Velocity Response
In this subdivision, the simulation and experimental consequences of the speed responses of the MR brake are presented. The simulation survey is performed in SIMULINK-MATLAB based on the regulating equations. All parametric quantities from the derived equations have been assigned in the MATLAB as followed:
J = 0.932 kgm2 ( entire minute of inactiveness )
N = 2 ( figure of surface )
? = 0.09Pa s ( MR fluid viscousness ? )
rz = 0.0415m ( outer radius of phonograph record )
rw = 0.04m ( interior radius of phonograph record )
r1 = 0.1m ( inertial burden radius )
K = 0.269Pa m/A ( electric invariable )
n = 0.00525m ( MR fluid spread )
? = 12500m-1 ( relative addition )
Figure 7 shows the sample of simulation consequences. The x-axis is clip variable and the y-axis is revolving velocity variable. This consequence is obtained by using 0.5 Ampere electric current to the spiral. It can be seen that the clip needed to halt an inertial burden of 10 kilograms from 2.5 rad/s until to the full halt.
Figure 7 Velocity versus clip
Using the developed MR brake, the experimental probe was conducted as follows. Initially, the AC-motor thrust the block attached on the input shaft of MR brake stationary. When the shaft rotates in a changeless velocity ( indicated from velocity detector ) , a certain current is supplied to the MR brake together with let go ofing the belt tensioner. The applied current will give the magnetic field across the spiral and therefore the MR consequence is generated. The burden cell so measures the produced braking torsion, while the velocity detector records the revolving speed response. All these measured informations are so displayed in a Personal computer for farther analysis.
Different current values viz. : 0.25, 0.5, 1, 1.5, 2 and 2.5 Ampere were applied to the MR brake spiral during experimental work. All of trials are performed in room temperature of 25 – 30 oC. Figure 8 shows the revolving speed responses under assorted currents. From the Figure, it can be seen that the magnitude of the clip needed for halting the burden decreases proportionately with the addition of the current applied to the MR brake spirals.
Figure 8 Revolving velocity versus clip
The simulation consequences obtained from the regulating equation of gesture demand to be validated with the experimental consequences. This measure is of import for the future work. When the theoretical account is valid, it is ready to be integrated with a control system that will be proposed for ABS utilizing MR brake. In this paper, the proof is provided in several currents applied viz. : 0.5 Amp, 1 Amp, 1.5 Amp and 2 Amp. The consequences are shown in Figures 9 ( a ) , ( B ) , ( degree Celsius ) and ( vitamin D ) severally. In Figure 9, the intimacy between the measured and modeled consequences indicates that the dynamic theoretical account of MR brake is valid.
( a ) Applied current 0.5 Ampere
( B ) Applied current 1 Ampere
( degree Celsius ) Applied current 1.5 Ampere
( vitamin D ) Applied current 2 Ampere
Figure 9 Validation consequences of MR brake dynamic theoretical account
Braking Torque Response
In this subdivision, the behaviour of the MR braking torsion responses are studied in order to cognize capableness of the MR brake in bring forthing the torsion. The torsion recorded by the burden cell is the entire torsion response generated by the MR brake. The experimental consequences of retarding torsion in clip sphere are displayed in Figure 10. From the figure, it can be noted that to acquire the shorter halting clip in changeless inertial burden which has initial revolving velocity, the higher torsion should be applied. On the other manner, the increasing current applied to the MR brake spirals, the bigger braking torsion obtained. This fact agrees with the speed response, in which the higher electric current allows the MR brake to bring forth larger MR effectual braking torsions to slow the tonss more quickly.
Figure 10 Braking torsion in clip sphere
The mean braking torsion of each current supplied is besides calculated. The relation between the mean braking torsion and current is shown in Figure 11. From the figure, it can be seen that the torsion increases exponentially in the increasing current.
Figure 11 Average braking torque versus current
Decision
A MR brake trial rig which was instrumented with several detectors has been developed. The mathematical theoretical account of the braking system has besides been derived based on the free organic structure diagram. To analyze the influence of electric current to the magnetic flux, the finite component theoretical account simulation was performed utilizing ANSYS MAXWELL SV. From this mold, the behaviours of magnetic flux lines, strength and denseness have been obtained.
The MR brake behaviour was studied in this paper through simulation and experiment. By utilizing the energized MR brake to get the better of the burden torsions and inactiveness, the speed and torque responses of the MR brake were measured by the velocity and burden detectors severally. In this work, the simulation consequences of speed responses have been compared with the experimental consequences. The intimacy of the consequences indicated that the mathematical theoretical accounts of braking system were validated. By and large, it can be concluded that the increasing current applied to the MR brake spirals will shorten the fillet clip as the consequence of the increasing braking torsion.
Acknoledgement
This undertaking is conducted under the short term grant funded by Universiti Teknikal Malaysia Melaka. Writers thank to whom involves in this undertaking.