On Forced Excitation of Electromagnetic Clutches

M. F. Filippov
Associate Professor, Candidate of Technical Sciences

Introduction

Electromagnetic friction clutches make it possible to use, for the rotation of periodically started or reversible drives, the cheapest, simplest, and most reliable asynchronous motor in operation.

Electromagnetic clutches find their greatest application in drives with very frequent starting and reversing, where the duration of the cycle is measured in several seconds. With long cycle durations, transient regimes—starting, braking, reversing—are of no appreciable importance; but in drives with short cycle durations, transient regimes have a major influence on the operation of the drive and take up a relatively significant amount of time, essentially determining the total duration of the cycle. Further shortening of the cycle duration in this case can be achieved only by reducing the duration of the transient processes.

Electromagnetic clutches are also used for controlling auxiliary mechanisms, where in many cases accuracy of control in time is required. Accuracy of control can be attained only when starting, stopping, or reversing occurs almost instantaneously. Thus, in this case as well, the requirement arises for the maximum reduction of the duration of transient processes.

The study of methods for intensifying transient processes in a drive with an electromagnetic clutch is at present of great importance also because high-speed electromagnetic clutches are included among the elements of electric drives “which must be mastered and produced in the third five-year plan” (Electricity, 1939, No. 4, p. 10).

The Transient Process During Starting

The total time required for starting a drive with an electromagnetic clutch consists of three parts: 1) \(t_a\)—the time of rise of the excitation current and, consequently, of the magnetic flux from the moment the core coil is switched on to the moment the rubbing surfaces of the clutch come into contact; 2) \(t_1\)—the time of acceleration of the disk, with slip present between the rubbing surfaces, up to the moment of their engagement; and 3) \(t\)—the time of acceleration to the steady-state speed with the core and disk engaged, if the drive motor has a soft characteristic and, during the process of clutch engagement, appreciably reduces its rotational speed.

To reduce the duration of the transient process, one should seek the maximum reduction of all three components.

Forced Excitation of a Clutch

When the excitation winding of the clutch core is connected to a constant voltage \(U_n\), the excitation current does not reach its steady-state value \(I_m\) immediately, but gradually; consequently the attraction of the disk to the core will occur not at the moment the core coil is switched on, but with a certain delay by the time \(t_a\), since the magnetic flux, only upon reaching a certain magnitude, is capable of attracting the disk to the clutch core, overcoming the friction of the disk during axial displacement, the tension of the release springs, and the inertia of the disk.

The law of increase of the current in the core coil may be expressed by the equation

\[ I = I_m \left( 1 - e^{-\frac{r_m}{L_m}} \right) = I_m \left( 1 - e^{-\frac{t}{T_m}} \right), \tag{1} \]

where:

\[ r_m \]
— ohmic resistance of the clutch core winding;

\[ I_m = \frac{U_n}{r_m} \]
— steady-state excitation current of the clutch;

\[ L_m \]
— self-inductance coefficient of the core winding;

\[ T_m = \frac{L_m}{r_m} \]
— time constant of the electromagnetic inertia of the clutch.

If the influence of Foucault currents in the core and disk is neglected, then, for a weakly saturated magnetic circuit of the clutch, proportionality may be assumed between the excitation current and the magnetic flux of the clutch, i.e.,

\[ \Phi = \Phi_m \left( 1 - e^{-\frac{t}{T_m}} \right), \tag{2} \]

where \(\Phi_m\) is the magnetic flux corresponding to the steady-state excitation current of the clutch \(I_m\).

It follows from equation (2) that the duration of the flux increase process will be the shorter, the smaller the value of \(T_m\).

One of the methods of forced excitation is based on this; it consists in inserting an additional resistance \(R\) into the excitation circuit (Fig. 1).

In this case the voltage \(U_c\) is chosen so that, despite the additional resistance \(R\), the excitation current settles at a value equal to the nominal one. In this case the value

\[ T'_m = \frac{L_m}{r_m + R} \]

will be smaller than

\[ T_m = \frac{L_m}{r_m}, \]

and the process of current increase will be accelerated.

In addition to the method of forced excitation indicated above, a method is used in which, at the instant of switching on, a voltage \(U_c\) higher than the nominal one is applied to the excitation winding.

For this purpose, a resistance \(R\) is connected in series with the winding (Fig. 2), which is initially shunted by contactor \(K\).

When the excitation current reaches a value close to the nominal value \((I_m)\), the contacts \(K\) open, the additional resistance \(R\) is inserted, and forced excitation ceases.

In Fig. 3, curve 1 corresponds to the process without forced excitation; curves 2 and \(2'\) refer to forced excitation by the first method (for two different values of \(R\)); and curve 3 is obtained with forced excitation by the second method.

Despite the use of the above methods of forced excitation, the transient processes of an electromagnetic clutch require further reduction.

Fig. 1

Fig. 2

Fig. 3

Forcing the engagement process

As was indicated above, contact of the rubbing surfaces of the clutch occurs with some delay after the coil is switched on. Fig. 4 shows the rise curve of the excitation current, and point a corresponds

corresponds to the instant of contact between the disk and the core. Consequently, the beginning of acceleration or reversal should be counted after a time interval \(t_a\) from the moment the core coil is switched on.

In addition, the curve in Fig. 4 shows that the excitation current, and hence also the torque of the clutch, does not reach its greatest (normal) value immediately after the beginning of friction between the disk and the core.

Fig. 4

This circumstance also leads to an increase in the duration of the starting process. For example, the duration of acceleration of a drive with an electromagnetic clutch due to friction between the disk and the core is determined by the expression:

\[ t_1=\frac{J_2\omega_s}{M_T-M_2} \left(1-\frac{M_1+M_2}{A}\right), \tag{3} \]

where \(M_T\)—the torque during friction between the disk and the core of the clutch—is taken as constant.

Consequently, the greater the value of \(M_T\) in the process of engagement, the faster the drive accelerates.

From all that has been set forth above it follows that, in order to shorten the duration of the transient processes of a drive with an electromagnetic clutch, a method of forcing is needed which would accelerate the build-up of the magnetic flux and, in addition, would provide a substantial torque during the process of clutch engagement.

In a drive with an electromagnetic clutch, where the non-reversing part of the drive has a large flywheel moment, it is permissible for the torque \(M_T\) (during slip between the disk and the core) to be considerably greater than the steady-state clutch torque \(M_m\), corresponding to the steady-state excitation current \(I_m\).

When an induction motor is used, as the author’s experiments have shown, the torque \(M_T\) in the process of clutch engagement may be greater than the pull-out torque of the motor \((M_{\mathrm{opr.}})\), since, owing to the non-

1) Filippov M. F., “Transient conditions of an induction motor in a drive with an electromagnetic clutch,” Izvestiya GII, 1940, No. 59, issue 11.

because, owing to the duration of the clutch-engagement process, the motor, possessing considerable kinetic energy, does not have time to reduce its speed to the slip corresponding to the overturning torque. The author proposes, especially for electromagnetic clutches, a new method of forcing that satisfies the conditions set forth above. For forcing by the proposed method, a resistance is connected in series with the clutch coil, the magnitude of which rises sharply as the temperature increases. In the author’s experiments, incandescent lamps with a metallic filament were used as such a resistance (Fig. 5).

Fig. 5

Fig. 6

At the first moment after switching on, with the lamps cold, the excitation-coil circuit has a reduced resistance and the coil is, as it were, connected to an increased voltage.

The excitation current rapidly rises to a value greater than the steady-state value; but, owing to the rise in temperature of the lamp filaments caused by the larger current, the resistance of the lamps increases and the current rapidly falls, assuming the steady-state value.

The oscillograms in Fig. 6 show, for comparison, the curves of the change in the excitation current for different methods of forcing, but for one and the same value of the steady-state current (\(I_m = 6\) A).

Fig. 7

Corresponding to curves 1, 3, and 4 in Figs. 7, 8, and 9 are oscillograms of the acceleration of the clutch disk. To make comparison possible, all the curves are superimposed in Fig. 10.

Fig. 8

From a comparison of the oscillograms obtained, it may be concluded that the forcing method proposed by the author gives a considerable reduction in the transient process.

If, without forcing, acceleration during the process of clutch engagement lasts 0.22 sec., while with forcing by the first of the methods described above it takes 0.16 sec., then with forcing by variable resistance only 0.085 sec. is spent on acceleration.

Fig. 9

Fig. 10

Curve 4—without forcing: \(r_M = 4.3\) ohms; \(U_c = 26\) volts.

Consequently, forcing by variable resistance can be successfully applied for high-speed electromagnetic clutches.