車輛工程外文翻譯-測試工業制動器襯片摩擦特性【中文3700字】【PDF+中文WORD】
車輛工程外文翻譯-測試工業制動器襯片摩擦特性【中文3700字】【PDF+中文WORD】,中文3700字,PDF+中文WORD,車輛,工程,外文,翻譯,測試,工業,制動器,摩擦,特性,中文,3700,PDF,WORD
TECHNICAL A RTICLE
Testing the Friction Characteristics of Industrial Drum Brake Linings
J. Van Wittenberghe, W. Ost, and P. De Baets
Department of Mechanical Construction and Production at Ghent University, Ghent, Belgium
49
Experimental Techniques 36 (2012) 43– 49 ? 2010, Society for Experimental Mechanics
Keywords
Drum Brake, Friction, Testing, Coef?cient of Friction, Temperature
Correspondence
J. Van Wittenberghe,
Department of Mechanical Construction and Production at Ghent University,
Ghent, Belgium
Email: Jeroen.VanWittenberghe@UGent.be
Received: December 7, 2009; accepted:
August 30, 2010
doi:10.1111/j.1747-1567.2010.00675.x
Abstract
In the present study a new brake setup was developed to test drum brake linings on an industrial brake with drum diameter of 3011 . During the tests performed on the setup, the brake undergoes a series of cycles in which the drum is slowed down from service speed to standstill. In each cycle the same amount of energy is dissipated as during a realistic safety stop. This was obtained by adding a flywheel in the setup so that the system’s kinetic energy at service speed matches the energy of the hoisting system dissipated during an emergency stop. Two different brake lining materials were characterized. Both materials were subjected to two test series to study the changes in coefficient of friction over a number of cycles. It was observed that the coefficient of friction of both linings was dependent on the drum temperature. The coefficient of friction of the first material decreased with increasing drum temperature, while the latter had the opposite behaviour.
Introduction
Spring applied, electrically released drum brakes are used in industrial environments, such as steel mills, to control the movement of travelling cranes as well as the hoisting apparatus of the crane. Such cranes are typically powered by an electromotor, but although the hoist motors are normally geared to produce greater torque and reduce the output speeds to an acceptable level for lifting and lowering heavy objects, it remains nevertheless possible for the motor to be driven by a heavy object in case of an electrical failure during lifting. This dangerous situation is referred to as ‘‘block drop.’’ To stop the motor in case of block drop, spring applied, electrically released drum brakes are used. These brakes contain heavy springs which push the brake shoes against a drum that rotates with the motor or the transmission output shaft. To retract the springs, a built-in electric solenoid has to be powered. The solenoid is generally wired in the motor’s electrical circuit, so when power is lost to the motor, the solenoid also loses power allowing the springs to thrust the brake shoes against the drum and hence preventing the motor to turn freely. When block drop appears, the drum brake is closed, stopping
the lifted load to fall down and keeping it at its height. But before trying to solve the failure of the electrical power circuit of the crane, it is important to put the load safely on the ground. Normal procedure is then to use a backup circuit with manual control to open the brake for a moment. To prevent a too fast rate of descent, the brake is closed after a moment, stopping the load again. These actions are repeated several times until the load is lowered completely. During this procedure, the drum brake material is severally put to the test because total load has to be slowed down repeatedly without the help of hoisting apparatus traction.
The drum brake’s braking power depends not only on the force applied by the springs, but is also determined by the frictional properties between the material used in the braking shoes and the drum of the brake. The behaviour of this friction material during its service life has to be known because a lack of friction can cause the brake to slip due to heavy loads. Nevertheless, a coefficient of friction (COF) that is too high can overload the drum axle and can cause high drum temperatures and high dynamic loads on the drum which can lead to cracks at the drum surface
J. Van Wittenberghe, W. Ost, and P. De Baets
Friction of Drum Brake Linings
or even drum fracture. Nowadays, a wide range of friction materials is available, but as is known from Zhang and Wang1 the behaviour of those material is highly dependent on their composition and service conditions. Through a series of tests on small-scale samples, they found the friction performances and wear resistance of the same material to be changing with load, sliding speed, and temperature. In another study2 they showed that also the drum material can have an impact on the tribological behaviour of the brake because of changes in specific heat capacity and thermal conductivity. Hence when new brake materials are developed, it is still necessary to perform experimental tests to characterize the lining material in combination with the drum material. In addition to this it is known that the pressure distribution is not evenly spread across the surface of brakes due to both geometrical deviations of drum and brake shoes and dynamic effects. This means that extrapolations of results of small scale tests on friction material cannot always be used to make reliable predictions on the behaviour of the full-scale brake. Hence in most cases full-scale tests are the only option to get accurate information about the performance of the brake.
Full Scale Test Setup
Principles of the drum brake setup
During previous studies, setups were developed mainly to quantify the frictional behaviour during continuous braking.3 In that case, the local fric- tion intensity of an imaginary friction lining segment changes during braking. This process is called ther- moelastic instability (TEI) and causes, over a critical speed, a steady-state regime with harmonic changes in friction. The TEI can be predicted accurately by finite element analyses.4 However, in the case of block drop and the procedure of safely lowering the load afterwards, the transient regime is the region of interest because the steady-state regime is not reached. For this purpose, a new setup was designed to simulate the block drop situation in a better way.
In the new setup the brake undergoes a series of cycles in which the drum is slowed down from service speed to rest. Of course to have a realistic situation, there should be an equal amount of energy dissipated during one cycle as in a real safety stop. To obtain this, the system’s mass moment of inertia was chosen in such a way that the kinetic energy of the system at service speed would match the maximum energy to be dissipated during an emergency stop.
In the following paragraphs, firstly, the test setup details are presented together with a calculating
method to obtain the COF. Later the test data of the two different brake lining materials will be discussed.
Test setup description
Schematic drawings of both the frontal and the section view of the setup are shown in Figs. 1 and 2. A view of the total setup is given in Fig. 3. The setup consists of a spring applied and electrically released Igranic safety brake type M 3011 , whose drum (1) is driven by an electrical DC compound 100 kW (at 5000 rpm) motor (17). The braking force is applied by the spring
(4) that pushes the brake shoes (2) against the drum. Different friction lining materials (3) can be mounted in the brake shoes to test their behaviour during brak- ing. The braking pressure can be set by adjusting the spring compression with the bolt (5) and can be varied between 0 and 16.6 N/cm2. The latter corresponds to a maximum braking torque of approximately 10 kNm for a COF between the drum and the friction lining of 0.6. To open the brake the solenoid (6) is powered pulling part (7) to the left and compressing the spring. To obtain a system that contains enough kinetic energy to simulate a realistic block drop situation, a drive wheel (8) is added to increase the inertia of the system. Drum (1) and drive wheel (8) are carried by the main axle (10). The drive wheel is connected to the main axle using two locking assemblies (9). The main axle is supported by two self-aligning ball bearings (11) and is connected to the DC motor by a
flexible jaw coupling (12).
Drum and drive wheel have the same diameter of 3011 or 760 mm. The moments of inertia of the differ- ent rotating parts of the setup are given in Table 1. Drum, drive wheel, and main axle are the parts that contribute the most to the moment of inertia of the system. Since a jaw coupling is used, the rotor of the DC motor rotates with the other rotating parts of the setup and its inertia of 6 kg·m2 has to be taken into account. This gives the setup a total moment of iner- tia of 95.1 kg·m2 resulting in a total kinetic energy of 422 kJ at the service speed of 900 rpm. Because the total brake shoe area is 0.28 m2, the mean energy density during each braking cycle is approximately 1500 kJ/m2. In a previous study by Severin5 a brake with a 2511 drum dissipating 168 kJ during each brak- ing cycle starting from a service speed of 900 rpm was used, giving an energy density of approximately 1100 kJ/m2. Hence the setup of this study is able to apply a much higher energy density into the material starting from the same service speed.
During a braking cycle, the drum and the drive wheel are brought up to service speed, while the brake is open. Once the speed of 900 rpm is reached,
0.500m
5
4
6
7
2
3
e
FG
MB
a
b
1
15
FL
13
14
Figure 1 Schematic front view of the drum brake setup
1
8
12
11
9
to the motor
10
11
16
Figure 2 Schematic section view of the drum brake setup
17 1 18
8
13
from happening and will apply a force FL (N). Because the loadcell is rigid, the actual rotation is very small and the position of the brake on the inclined surfaces
(13) will not change significantly. Hence the reaction forces in the supports stay aligned with the connection lines a and b in Fig. 1. This means the vector of the reaction forces goes through the centre of rotation of the drum and the reaction forces do not contribute to the torque equilibrium around this point.
Figure 3 Drum brake setup
Table 1 Properties of the rotating parts of the setup
Inertia (kg
Calculating the coefficient of friction
The COF can be calculated from the applied braking torque MB, which can be calculated from the force measured by the loadcell FL by expressing the torque equilibrium around the centre of the drum (Fig. 1):
MB = FL·0.500 ? FG·e (Nm) (1)
with FG (N) the gravitational force of the brake and e
Part Mass (kg)
m2) Material
(m) the eccentricity of the centre of mass to the centre
Drum 320 28.8 Cast iron
Drive wheel 700 56.7 Structural steel
Main axle 60 3.6 42CrMo4 alloy steel
Coupling 9 0.01 Steel + elastomer spider Two locking assemblies 5 0.02 Steel
the power of the motor is switched off and the brake is closed. When finally the drum has come to rest, the brake is opened again and the cycle repeated.
During the tests, the rotational speed was measured using a tachometer mounted on the motor and the surface temperature of the drum was continuously measured using an SP i-tec 2005D infrared sensor (see (18) in Fig. 3). The control of the system and measuring of all signals are carried out by a computer with a Texas Instruments BNC-2110 data acquisition card and a Labview programme. Speed, surface temperature and force in the loadcell were recorded at a frequency of five samples/second.
In order to measure the brake torque, the brake is mounted on two inclined surfaces (13) and (14), as can be seen in Fig. 1. These two supports are manufactured in the way that the supporting surfaces are perpendicular to the two construction lines a
and b. As the drum rotates in the counter clockwise
of rotation of the drum. The gravitational force of the brake is constant and because the actual rotation of the brake is very small, the eccentricity can also be considered constant. When the brake is open, no braking moment is applied, but due to the eccentric centre of mass of the brake, there is still a force applied on the loadcell. For this case (MB = 0) Eq. 1 becomes
FL·0.500 = FG·e (Nm) (2)
where FL is a measured value. By this way a value for FG·e of 3136 Nm was found. With the mass of the brake of approximately 1 tonne, an estimated eccentricity of 0.31 m was obtained. In the calculations only the product FG·e is used. The estimated value of the eccentricity is only mentioned as an illustration.
From the braking torque MB, calculated from Eq. 1, the COF μ can be calculated as explained in the following section.
As is schematically shown in Fig. 4, the braking pressure p (N/m2) multiplied by the COF, integrated over the surface of the brake shoes equals the braking torque MB:
r α
direction, the reaction force on the support (14) can become negative. To counter this force the part (15)
MB = 2·b·
r·μp·rdθ (Nm) (3)
?α
is present, whose contact surface is parallel to the contact surface of (14). A loadcell (16) with a capacity of 20 kN is mounted 500 mm below the centre of rotation of the drum. During braking the brake will try to rotate with the drum. The loadcell will prevent this
The factor 2 in Eq. 3 results from the two brake
shoes that are present. Equation 3 can be simplified to
MB = 4·μ·b·r2·p·α(Nm) (4)
mp
p
a
-a
r
braking
Speed
Torque Temperature
5000 100
4000 80
Speed [rpm] Torque [Nm]
Temperature [°C]
3000 60
2000
1000
0
40
20
0
-1 0 1 2 3 4
Time [s]
Figure 5 Measured signals during one braking cycle
Figure 4 Schematic view of the pressure in the brake shoe
Hence
μ MB [—] (5)
= 4·b·r2·p·α
with b the width of the brake shoes (0.300 m), r the radius of the brake drum (0.380 m), p the mean braking pressure during the tests (8.1 N/cm2 = 8.1·104 N/m2) and α the half angle of one brake shoe (35? or 0.611 rad).
With the above values Eq. 5 becomes
MB(Nm)
μ = 8574(Nm) [—] (6)
Course of a braking cycle
During each braking cycle, the drum and the drive wheel were brought up to 900 rpm. This took about 90 s. Once the drum was at the required speed, data acquisition started and 2 s later the brake was closed. Two seconds after the drum stopped, data acquisition was interrupted and the break opened again, after which the cycle restarted. In order to control the dataflow and avoid recording excess data, data logging was interrupted when the drum was brought up to service speed.
In Fig. 5, the course of a braking cycle is shown. For this cycle the braking time is 2.2 s, in which the braking speed is brought from 900 rpm to rest. The course of the braking torque is somehow different from what one could expect from small- scale material tests. Common frictional behaviour of braking materials includes a difference in static and dynamic COF, from which we could expect the braking torque to have a peak when the brake is
closed and remain constant until the drum is brought to a halt. In Fig. 5, however, it can be observed that the braking torque increases linearly for about
1.4 s after which the torque reaches a more or less stable value. This linear increase is caused by electromagnetic effects in the solenoid ((6) in Fig. 1) of the brake. When the current over the solenoid is removed, the force of the spring ((4) in Fig. 1) is not immediately applied on the braking shoes. Due to the solenoid’s self-induction, the original magnetic field only decreases gradually and hence, the braking torque is applied over a certain period of time instead of instantaneously. In this cycle the maximum braking torque is 4045 Nm, from which a COF of
μ = 0.47 can be calculated according to Eq. 6. The drum temperature increases here from 27?C before the braking to a maximum of 47?C during the braking.
Experimental Tests
In following sections the results of the test series performed on two different composite brake linings with a different composition is presented. Both materials were subjected to two test series on the new setup. First, a short test series was conducted, where the objective was to test until the mean surface temperature of the drum saturated. The short test series was stopped after 50 cycles. Second, a long test series was conducted, consisting of 250 successive cycles to study the integrity of the lining material when subjected to a high number of braking cycles.
The conducted tests are summarized in Table 2. The noted numbers for the materials and tests will be used according to this table in the rest of this article. Test series 1 and 3 are the short test series, 2 and 4 are the long series.
Temperature [°C]
Coefficient of Friction [-]
Material 1 Material 2 100
Test Test Test Test
80
Series 1 Series 2 Series 3 Series 4
0.60
0.50
0.40
Number of cycles
50
250
50
250
60
0.30
Final speed (rpm)
900
900
900
900
Environment 22.5 20.4 21.0 20.8 40
temperature at
start (?C) 20
Drum temperature 31.2 22.8 27.2 21.1
Min. Temperature
Max Temperature
Coefficient of Friction Mean Temperature
0.20
0.10
Table 2 Summary of the tests short test series
120
at start (?C)
Mean drum temperature at
end (?C)
63.6 64.8 69.9 64.9
0
0 10 20 30 40 50
Number of Cycles
0.00
Coef?cient of friction last cycle
0.47 0.49 0.35 0.31
Figure 6 Coef?cient of friction and temperatures during test series 1
(short) on material 1
Short test series
The results for the short test series of materials 1 and 2 are shown in Figs. 6 and 7. For both materials, the COF together with the minimum, maximum, and mean temperatures are plotted as a function of the cycle number. For both materials it can be seen that the mean temperature saturates at about 65?C after approximately 30 cycles. At this point the minimum and maximum temperatures are also saturated, with a minimum drum temperature of about 50?C for both materials. The maximum drum temperatures are different for both materials, as can be seen in
120
100
Temperature [°C]
80
60
40
20
0
Min. Temperature
Max. Temperature
Coefficient of Friction Mean Temperature
0 10 20 30 40 50
Number of Cycles
0.60
Coefficient of Friction [-]
0.50
0.40
0.30
0.20
0.10
0.00
Fig. 6, the maximum drum temperature with lining material 1 can reach peak values of about 118?C, while only 104?C for lining material 2 (Fig. 7). This difference is caused by the difference in COF between the two materials. The COF of material 1 is higher than that of material 2, which means the braking time will be shorter for material 1. Consequently, the same amount of kinetic energy has to be transferred from the drum to the friction lining in a shorter time, resulting in higher peak temperatures. However, because the actual braking time (about 2.5 s) is short in comparison to the total cycle time of about 96 s, the minimum and mean drum temperatures for both friction linings are practically the same.
Additionally, it can be observed from Fig. 6 that the COF shows a slight increase with increasing temperature: the COF started at a value of 0.44 for a mean drum temperature of 36.0?C and increased to a value of about 0.47. The opposite behaviour was observed for material 2 (Fig. 7). Here the COF decreased with increasing drum temperature: at start COF = 0.47 and the mean drum temperature was 27.2?C, while the
收藏