Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Ahmad Proj

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 11

1-INTRODUCTION

The theory of electrical circuits represents one of the most important parts of any
electrical engineering education. The aim of this paper is analysis circuit and to
experience the actual behavior of a D.C. shunt motor, this requires a powerful software
mathematical tool [1]. MATLAB is a good software package for high performance
numerical combination of analysis and visualization. It makes the combination of
analysis capabilities more flexibility, more reliability, and powerful graphics [2, 3]. The
modeling and simulation of this paper helped to generate expected outcomes of the
project design, simulation software MATLAB is used to provide simulation design and
results for evaluation the speed, torque and the armature current for a D.C. shunt
motor. Simulink which is a sub program of matlab was used to complete the modeling
and simulation [4]. Simulations are interactive, so user can change parameter on the spot
and immediately see what happens [5].
These can be achieved by changing the setting in Matlab/ Simulink to investigate a
D.C. motor responds to these changes.
The Simulink program will help the students without returning to the laboratory to use
the actual D.C. motor.
Many researches dealt with this subject. Saffet Ayasun and Gultekin Karbeyaz
(2007)," in their paper describes the matlab simulink realization of the D.C. motor speed
control methods",. Karung Berkunci´s (2006) work," presents shunt connected direct
current motor analysis using matlab laboratory",. Tan Kiong Howe´s (2003), evaluate the
transient response

130
of a D.C. motor by variation of terminal voltage, armature resistance, and field
resistance using matlab/ simulink.

2- PRINCIPLES OF OPERATION OF D.C MOTOR


An electric motor operation is based on simple electromagnetism. Motor is a
machine which converts electrical energy into mechanical energy that when a current-
carrying conductors is placed in an external magnetic field, it experience a mechanical
force proportional to the current in the conductor and to the strength of the external
magnetic field and between them generate rotational motion [6].
The effect of flux distribution is very important, because the limits of successful
commutation are directly influenced by the flux, also both the generated voltage and
torque of armature current are influenced thereby [7].
The brushes, commutator contacts and rotor windings are such that when power is
applied, the polarities of the energized winding and the stator magnets are misaligned,
and the rotor will rotate until it is almost aligned with the stator‫ۥ‬s field magnets. As the
rotor reaches alignment, the brushes move to the next commutator contacts and
energize the next windings [8].
In the field resistance control, a series resistance is inserted in the shunt-field circuit
of the motor in order to change the flux by controlling the field current. It is theoretically
expected that an increase in the field resistance will result in an increase in the no-load
speed of the motor and in the slope of the torque- speed curve [9].
The field current may be varied by placing a variable resistance in series with the
field windings. Since the current in the field circuit is low, a low-wattage rheostat may be
used to vary the speed of the motor due to the variation in field resistance. As f.eld
resistance increase, field current will decrease. A decrease in field current reduces the
strength of the electromagnetic field. When the field flux is decreased, the armature will
be rotate faster, due to reduced magnetic field interaction. Thus the speed of a D.C.
shunt motor may be easily varied by using a field rheostat [10].

3 -MOTOR MODELING SIMULATION AND MATHEMATICAL REPRESENTATIONS


The system contains a D.C. shunt motor, a model based on the motor specifications needs
to be obtained, as shown in Fig(1).
The basic motor equations are:-
Eb = Vt – Ia Ra ………… (1)
Eb = K = K f If Wm, …….. (2)
Eb =Km Wm
T = K Ia = K f I f Ia …….. (3)
T =Km Ia , Km =Kf If ,
:. Km = Eb / Wm at separately excited
Ka = Kf = Eb / Wm If, at shunt excited
Applying Kirchhoff's Voltage Law:
Vt = Eb + Ia Ra +La dia /dt ….. (4)
Vt = If Rf + Lf dif / dt ………. (5)
From equation (2)
Vt = K a If Wm + Ia Ra +La dia /dt …. (6)
The Laplace transform of equations (5, 6)
Vt (s) = Ka If Wm (s) + Ia Ra +LaIa (s)
Vt(s) – Eb = Ia(s) [La(s) + Ra]
:. Ia(s) =Vt(s) – Eb / [La(s) +R ……. (7)
Vt = If Rf + Lf If (s)
:. If (s) = Vt(s) / [Lf(s) + Rf ] ………. (8) Or
equation (7): -
Vt (s) =Ka If Wm(s) + Ia Ra (1 +s )
Where: = La / Ra is the electrical time constant of the armature.
The dynamic equation for the mechanical system:-
T = Ka If Ia = J dWm / dt + BWm +TL ….... (9) BWm
= is the rotational loss torque of the System.
The Laplace transform of equations:-
T(s) = Ka If Ia (s)
= J.s.Wm(s) + BWm(s) + TL(s)
T(s) –TL(s) = Wm(s) × [J.s. +B]
:. Wm(s) = T(s) –TL(s) / [J.s. + B] ……. (`10)
From (3) and (10)
T(s) = J.s.Wm(s) +BWm(s) +TL(s)
= Wm(s) / [J.s. +B] + TL(s)
: .Wm(s) = T(s)-TL(s) / B [1+ m.s] ..…. (11)
Where m = J / B is the mechanical time constant of the system.
From (2) and (8)
Vt (s) = Eb(s) + Ia(s) R (1+s a)
:. Ia(s) = Vt(s) – Eb(s) / Ra (1+s a) …… (12)
The values of Pin, Pout, Plosses, and (η) can be calculated by:
Output power =Pout = Wm × T [KW]
Pout = 2 NT/60 …….. (13)
Input power = Vt × Iin [KW]
Pin =Vt × (Ia +If) ……. (14)
Losses power = Pin - Pout ……… (15)
Efficiency (η) = Pout / Pin × 100 % …. (16)
A block diagram which represents the equations (7), (8) and (10) is shown in fig. (2).

4- MODELING RESULT AND DISCUSSIONS


In this paper, the specifications of D.C. shunt motor were obtained from the
engraving on the metal tag attached onto the motor, shown in table (1).
To produce a good model design, there needs to be some amount of simulations, to
avoid aimless trial and error techniques with the actual equipment of a D.C. shunt
motor.

From the specification of the D.C. motor and the equations; Calculate the constant
torque (Kf ) or (Ka), at no-load;
Vt = 220 volt, If = 0.75 ampere, from the equation (2); then the torque constant:
(Kf) = Eb / If Wm,
As the speed in terms of (N) (r.p.m.), N = 1800 r.p.m., to convert to (rad / s), then
the speed Wm =2 N/60, Wm = 2× 3.14 ×1800/60 = 188.4 rad/s,
And
Kf = 220/ 0.71 ×188.4 = 1.644 =Ka
The value of B obtains from calculates the mechanical equation as follows:-
From equation (9) and when TL = 0 (no-load), at steady state, Ia and Wm,
stabilized
then; dWm/dt
= 0.
Kf If Ia = J *dWm / dt + BWm +TL.
:. dWm/dt = Kf If Ia - BWm = 0, TL = 0.
:. Kf If Ia = BWm,
:. B = Kf If Ia / Wm,
B = 1.644× 0.71 ×1.8 /188.4,
B = 0.01115.

Ia = 1.8 ampere from the measurements of experiment at no-load.


J = 0.117 Kg*m2 the value of the armature (rotor) inertia.
After the measuring of the values resistance and inductance (Ra, La) of the
armature windings and the shunt field windings (Rf, Lf )for the D.C. shunt motor in
electrical machines laboratory(D.C. Machines laboratory), their values were:-
Ra = 2 Ω. La =16.2mh. Rf = 210 Ω. Lf = 5.47 H.
And varies resistance connected in series with shunt field windings to obtain speed
about (1800 r.p.m.) at no load, the varies resistance shown in table (2). The average
of the constant torque, K f =Ka=1.607. And IL= 0, 2, 3, 5, 6, 8, 9, 10, 12, 13, 14, 15.
Ampere.
Where the field current (If) is constant If= 0.71 ampere. Take K a at no load when the
speed is (1800 r.p.m.) and Eb =Vt, IaRa is neglected because the voltage drop (IaRa) is very
small.
Ka=Km= Eb/ If Wm, Ka=220/0.71×188.4 =1.644.
Comparing the results of table (3) and table (4), the values of (If) is kept constant of
a (0.71 amp.) while this value was (0.698 amp.) at simulation technique. This slight
difference may be of recording the direct data obtained from the devices.
The same observations were recorded for the Ia (amp.) T (N.M.) and Wm
(rad/s). This difference was calculated and it was found to be about (1-2) %
The D.C. shunt motor was modeled using characteristics transfer function of
electrical and mechanical of the motor as shown in fig.(3).

5- SIMULATION RESULTS AND DISCUSSIONS


D.C. motors are used to drive mechanical loads, some applications require that the
speed remain constant as the mechanical load is applied to the motor changes.
A: Simulation Results At no-load:
At speed (Wo) on no-load, the produced no-load current is small and not enough to
carry the load so the motor starts to slow down. And the e.m.f. becomes smaller,
resulting in a higher current and higher torque.
At no-load to obtain the speed (1800r.p.m.) connect variable resistance in series
with the shunt field circuit, its shown in table (2), increase the resistance, the speed
increase until to obtain (N= 1800r.p.m.), the relation between them linearity, the results
of simulations for the D.C. shunt motor model at no load is shown in Fig. (4).
A series resistance is inserted in the shunt-field circuit of the motor in order to
change the flux by controlling the field current. It is theoretically expected that an
increase in the field resistance will result in an increase in the no-load speed of the
motor; this method is one of the most common speed control methods for dc shunt
motor.
Fig.(5), shows the output waveform of the starting torque which has a high value in
the beginning, after that it drops to an approximate torque equal to the torque imposed
(2.1 N.M.), the output waveform armature current shows a large increasing, at starting,
after that it drops to reach the approximate armature current (Ia) (1.8 amp.), and the
output waveform of the speed in (rad/s), it is very clear that there is a large increase in
the speed at the beginning but at steady state the speed remains constant at (188.4
rad/s)(1800r.p.m.)at no load.

B: Simulation Results At load:


At load (change in the mechanical load), when the mechanical load applied to the
shaft vary, the armature current rises and the speed drops. The mechanical load torque
reaches (21 N.M.).
From the simulation results shown in Fig. (6) the armature current rises, from no-
load at
(1.8 amp.) to the full load current (19.8amp.) at steady state, shows the variation of
torque from (2.1N.M.) at no-load to the value (21 N.M.) at full load at steady state and
Te =22.7 N.M., and the reduction of speed in (rad/s) of the shunt motor from no-load at
(188.4 rad/s)(1800r.p.m.) to ( 157 rad/s)(1500r.p.m.) at full load, this cause to diminish
e.m.f., resulting in higher current and a corresponding higher torque, leads to slow
down the speed.
The simulation results for calculating Pout, Pin, Plosses, and η, from the equations
(13, 14, 15, and 16) are shown as; fig. (7) and fig. (8).
Fig (7) Shows the output power versus time at full load, P out = 3.564KW, the input
power versus time at full load, Pin =4.5 KW. And shows the curve of the losses in power
versus time at full load, P losses = 0.9365 KW.
Since the power losses =0.9365KW, then power is dissipated in:-
1- Power losses converted directly to heat in the resistances of the current paths.
2- Mechanical energy developed within the device is absorbed in friction and windage
converted to heat.
3- The energy absorbed by coupling field is converted to heat in magnetic core loss for
magnetic coupling or loss for electric coupling.
Fig. (8) Shows the efficiency versus time at full load, the efficiency of the motor, η =
80%.

8- CONCLUSION
** In this paper, the block diagram of a D.C. shunt motor was developed by using
matlab / Simulink, the exact simulated with expected waveform output were obtained,
for example the armature current, torque, speed and output power characteristic of the
d.c. shunt motor.
** High protection, increases complexity in operation of the protection equipments
by increasing the supply voltage drop and transient torque which can damage the
mechanical drive, one of the advantages of having simulink is, increasing the simplicity
of the operation of protection equipments and low cost.
** From the obtained results, it was very clear that simulation can be very helpful tool
to study the dynamic behavior of D.C. shunt motor and its interaction, with reading
experiment. ** Simulation model of D.C. shunt motor and feedback control system
for D.C. motor drives have been developed using MATLAB/SIMULINK and it has been
shown that proposed simulation model correctly predict the effect of the field resistance
on the torque-speed characteristic of the D.C. shunt motor.

9- REFERENCES
1. Chee-Mun Ong, 1998, " Dynamic Simulation of Electric Machinery", Prentice Hall
DTR,.
2. User´s Guide, 2002, "SimPower Systems for use with Simulink", MathWork Inc.
Natick, MA, .
3. S. Li. And R. Challoo, 2006, " Restructuring an Electric Machinery Course with an
Integrative approach and Computer- Assisted teaching Methodology", IEEE Trans
Educ. 49,.
4. Adrian Biran, Moshe Breiner, 1999,"Matlab 5 for Engineers", 2nd edition
published ISBN, 0-201-36043-8,.
5. The Math Work, Printing, January 2001,"Matlab Student Version Learning Matlab
6" (Release 12), 2nd.
6. Theodore Wildi, 2000,"Electrical Machines, Drives, and Power Systemes", Fourth
Edition, Prentice Hall International, Inc.,.
7. M. S. Sarma, 1994,"Electric Machines, Steady- State Theory and Dynamic
Performance", 2nd edition, West, ST. Paul, MN,.
8. Peter F. Ryff, David Platnick and A. Karnas, 1987, "Electrical Machines and
Transformers Principles and Applications", Prentice Hall, Inc.,.
9. S. J. Chapman, 1998,"Electric Machinery Fundamentals", 3rd edition, WCB/Mc
Graw –Hill, New York,.
10. M.-Y. Chow and Y. Tipsuwan, 2003,"Gain Adaptation of Networked DC Motor
Controllers Based on QOS Variations", IEEE Trans Ind. Electron 50,.

Table(1): Specifications of a D.C. Shunt Motor

D.C. motor specification


Tipo 160L
Model Number 61.5.75F
Power 2.94KW
Horse Power 4
Voltage 220 V
Current 15.4 A
Field Voltage 220 V
Field Current 1.06 A
Speed 1500 r.p.m.
Wd Compound

Table (2): Simulation results at no load condition.

Eb = Vt- Wm N T Ia Km=Ka
IaRa rad/s) .p.m.) .M.) Amp.) =Eb/IfWm
N
216.4 188.4 1800 1 .22 1.8 1.618
213.2 185.78 1775 3 .58 3.4 1.616
211.2 183.17 1750 4.8 4.4 1.624
206.6 178.98 1710 7 .63 6.7 1.625
204.4 177.93 1700 9.3 7.8 1.618
198.8 174.27 1665 12.4 10.6 1.607
196.4 172.7 1650 14.1 11.8 1.602
194 170.61 1630 15.5 13 1.602
188 167.47 1600 19.6 16 1.581
185.6 164.85 1575 21.5 17.2 1.586
183 162.23 1550 22.9 18.5 1.588
180 157 1500 23.8 20 1.615
Table (3): The experimental results at load.

Rf +Rser. Wm N If Ia T
(Ω) (rad/s) (r.p.m.) (Amp.) (Amp.) (N.M.)
210 126.8 1211.47 1.048 0.795 1.369
220 132.7 1267.83 1.0 0.887 1.459
230 138.7 1325.16 0.957 0.980 1.542
240 144.6 1381.53 0.917 1.023 1.549
250 150.5 1437.90 0.88 1.125 1.627
260 156.4 1494.27 0.846 1.243 1.729
270 162.2 1549.68 0.815 1.315 1.762
280 168.1 1606.05 0.786 1.443 1.864
290 173.9 1661.47 0.759 1.55 1.935
300 179.7 1716.88 0.733 1.629 1.965
315 188.4 1800 0.698 1.797 2.063
Table (4): The simulation results at load.

N Wm TL Te Ia If
(r.p.m.) (rad/s) N.M.) N.M.) Amp.) (Amp.)
1800 188.4 0 2.06 1.787 0.698
1771.3 185.4 2 4.06 3.54 0.698
1743.6 182.5 4 5.97 5.2 0.698
1715.0 179.5 6 8.0 6.96 0.698
1686.3 176.5 8 9.95 8.66 0.698
1657.6 173.5 10 11.9 10.4 0.698
1629.0 170.5 12 13.9 12.1 0.698
1600.3 167.5 14 15.9 13.8 0.698
1571.7 164.5 16 17.8 15.5 0.698
1544.0 161.6 18 19.8 17.23 0.698
1515.3 158.6 20 21.7 18.9 0.698
1500 157 21 22.68 19.8 0.698

SW.
La Ra

Ia TL

VtLf
Vt Ar.
If Eb
Vt
Vt
Rf

Fig.(1): Schematic Diagram of D.C. Shunt


Motor.
Fig.(2): Simulink Model of D.C. Shunt Motor.
Fig.(3): Simulink Model for D.C. Shunt Motor with the Constant Values.

Fig.(4): Rotor Speed (Wm) Versus Field resistance (Rf +Rser.)(Ω) At no load.

Fig.(5): Simulated Output Ia, Te, and Wm Versus Time (Sec.) at no load.

Fig.(6): Simulated Output Ia, Te, and Wm Versus Time (Sec.) at full load.
Fig.(7):Simulated Pout, Pin, and losses Versus Time (Sec.).

.)

You might also like