STUDY OF PARAMETERS OF CENTRIFUGAL FANS AND VENTILATION SYSTEMS OF DC MACHINES

R. Ya. Klyain, D. I. Sannikov

(Presented by the scientific seminar of the Departments of General Electrical Engineering and Electrical Machines)

To intensify the cooling of DC machines, various methods are used: increasing the cooling surface; compounding the windings to create good thermal contact with structural parts that conduct heat away; increasing fan output, etc. The latter method is quite effective and less labor-consuming.

The design of the ventilation system used in modern DC machines of the unified P series, because of its imperfection, cannot provide a sufficient cooling intensity in highly utilized machines of newly designed series. The low pressure of the fan of the existing design is due to large losses in the annular channel between the fan and the bearing end shield. In particular, the deterioration of the aerodynamic characteristics of this section is associated with the one-sided arrangement of the outlet openings in the bearing shield and with the underutilization of their area as a result of the formation of aerodynamic shadows and vortex zones. The obstructed nature of the air outlet increases the leakage flows of the fan, especially through the gap between the front disk and the frame.

Measures aimed at increasing the cooling intensity must consist, first of all, in eliminating the indicated shortcomings.

The fan design shown in Fig. 1 makes it possible to largely eliminate leakage flows and to equalize the air-velocity field at the outlet from the fan chamber.

The special feature of the fan under consideration, 2, is that it is made without a front disk; instead, a stationary diaphragm, 1, of annular shape is installed. The diameter of the rear disk, 3, is increased to the outside diameter of the fan. Four outlet windows are arranged uniformly around the circumference of the bearing shield; the width of the spaces between them is chosen from the condition of sufficient rigidity of the structure. For brevity, we shall call such a shield “open.”

The study of the fan parameters was carried out on machines of the P series of two frame sizes, P42 and P63, at a rotational speed of 1500 rpm. The tests were carried out both with the open shield and with the shield of the standard design. The area of the outlet windows of the open shield was 3.5 times greater for the P42 machine, and 2 times greater for the P63 machine, than that of the bearing shield of the existing design.

Ventilation tests were carried out using an aerodynamic test stand made in the form of an air-permeable chamber, in the outlet opening of which the machine under test was placed. The inlet to the chamber was equipped with a flow-measuring device operating on the principle of measuring the vacuum in the smooth inlet collector. The air flow rate was varied by means of an auxiliary fan built into the aerodynamic test stand.

Fig. 1

The influence of the outer diameter \(D_2\) (see Fig. 1, b) on the fan output is shown in Fig. 2 by curves 1—4. The limits of variation of the outer diameter relative to the internal bore of the bearing shield were \(0.8—0.97\). Curves 1 and 3 correspond to the shield of the existing design, and 2 and 4 to an open shield. A reduction in the aerodynamic resistance of the outlet leads to a sharper dependence of the air flow rate on the outer diameter of the fan. As can be seen from curves 1—4, as the diameter \(D_2\) is increased and brought closer to the internal bore of the shield, the air flow rate increases continuously.

Fig. 2

However, this phenomenon is accompanied by a large increase in the power consumed for ventilation and by an increase in fan noise. Obviously, in this case the criterion of optimality cannot be taken to be only the magnitude of the air flow rate. It is necessary to proceed from the ratio of the effect of in-

cooling intensity due to increasing the fan diameter and the power expenditure on ventilation, while taking into account the permissible limits of fan noise.

The internal diameter of the fan \(D_1\) has a very insignificant effect on the flow rate and pressure (curves 5 and 6), regardless of whether the internal diameter changes by changing \(D'_1\) (Fig. 1, b) with \(D''_1 = \mathrm{const}\), or by changing \(D''_1 = \mathrm{var}\) with \(D'_1 = \mathrm{const}\). The ventilation losses change only slightly. The greatest air flow is provided by the fan for which the ratio \(D_1/D_a\) is 1.1–1.12. Here \(D_a\) is the armature diameter. Curves 5 and 6 were obtained with a shield of the standard design.

The annular diaphragm has a large influence on fan performance. The dependence of air flow on the diaphragm diameter is shown in Fig. 2 in the form of curves 7–11. Curves 7 and 10 were obtained with a shield of the existing design, and curves 8 and 11 with an open shield; curve 9 corresponds to a fan with twice the number of blades as compared with the other relationships. Increasing the diaphragm diameter is accompanied by a significant decrease in fan pressure, owing to the occurrence of vortexing at the leading edge of the blades and increased leakage in the gap between the diaphragm and the fan. The dependence of pressure on the diaphragm diameter is shown in Fig. 2 by curves 12 and 13, respectively for machines P42 and P63. However, with a small diaphragm diameter the aerodynamic resistance of the machine air duct increases. Ventilation power losses, as experience shows, do not depend on the diaphragm diameter. The maximum flow rate of cooling air can be obtained when the ratio of the diaphragm diameter to the armature diameter lies within 1.2–1.5, which is optimal. Moreover, the exact value of this quantity must be determined individually for each machine size. On the basis of investigations by a number of authors, the clearance between the fan and the diaphragm is recommended to be taken as 2–4 mm.

Experiments show that increasing the width of the fan blade leads to an increase in ventilation losses, although the air flow rate increases only slightly. Therefore the blade width adopted in modern machines of the P series should be considered sufficient; specifically, the ratio of blade width to armature diameter should be taken within the limits 0.3–0.35.

The dependence of air flow on the number of fan blades is shown in Fig. 3 by curves 1–4. Here dependences 1 and 3 were taken with a shield of the existing design, and curves 2 and 4 with an open shield. It is evident from Fig. 3 that increasing the number of fan blades above 15 not only does not increase the air flow, but even leads to a certain decrease in it, while the power expenditure on ventilation increases considerably (curve 5). Therefore the number of blades 11–15 should be considered optimal. Reducing the aerodynamic resistance of the air outlet from the fan chamber by increasing the area of the outlet openings in the bearing shield is a substantial reserve for increasing fan performance. As can be seen from Figs. 2 and 3, the increase in flow rate solely by reducing the aerodynamic resistance of the outlet, all other conditions being equal, reaches 35% for the machines investigated. Figure 4 shows the dependence of air flow on the area of the outlet openings in the bearing shield. Here \(S_{\max}\) is the area of the ring on the side surface of the bearing shield.

The criterion for selecting the optimal opening area of the shield, providing a sufficiently small value of the aerodynamic resistance of the outlet, can only be the condition of splash protection and the strength condition of the shield structure.

The recommendations obtained as a result of the study of the fan parameters were checked in the thermal regime on a P42 motor, whose rated data are \(P_n = 3.8\ \text{kW}\), \(U_n = 110\ \text{V}\), \(n_a = 1500\ \text{rpm}\).

Fig. 3.

The tests were carried out at the rated armature current \(I_a = 42\ \text{A}\) in the following regimes:

I — reference regime with the standard fan and end shield;

Fig. 4

II — improved fan (Fig. 1) with optimal dimensions and the standard end shield;

III — improved fan and open end shield;

IV — the same, with a casing installed on the open end shield, ensuring splash protection. The temperature rises of the windings were determined by the resistance method, and the temperature rise of the frame by thermocouples built into the surface of the frame. The results of the experiment are presented in the table.

Here \(V\) is the working air flow rate through the machine;

\(\Theta_a\) — temperature rise of the armature winding above the ambient air;

\(\Theta_{\sh}\) — temperature rise of the field winding;

\(\Theta_d\) — temperature rise of the interpole winding;

\(\Theta_{st}\) — temperature rise of the frame;

\(t_{\text{amb}}\) — ambient temperature;

\(\Delta \Theta_f\) — heating of the cooling air in the machine.

As experience shows, the use of a fan without a front disk, but with a stationary diaphragm, provides a 1.6-fold increase in air flow rate, owing to which the overheating of the armature winding is reduced by 11% and that of the field winding by 25%. With an open shield, the flow rate

increases by a factor of 2.5, which provides a reduction in winding overheating by an average of 20%. The installation of a protective casing does not worsen the results achieved.