HIGH-SPEED TEST BENCH FOR TESTING GEAR TRANSMISSIONS

A. E. BELYAEV

(Presented by the scientific seminar of the Department of Applied Mechanics)

Test benches for testing gear transmissions are designed either with an open or with a closed power flow. Until recently, high-speed gear transmissions were tested on benches of the first type, despite their very substantial drawback: the power transmitted by the gear pair was determined by the power of the braking system. However, the increase in the power of motors for high-speed transmissions (marine and aviation drives) and the desire of researchers to carry out experiments with full-scale transmissions (or transmissions close to them) make it difficult to use benches with an open power flow under laboratory conditions, since they require powerful energy absorbers, which in a number of cases practically cannot be manufactured. Experience in the study of gear transmissions has shown that the results of experiments do not depend on the type of installation. This circumstance allowed us to adopt a scheme with a closed power flow¹ in designing a test bench for testing transmissions at high speeds.

The bench is intended for testing gear wheels for contact strength and bending strength under pulsating and alternating loads with different asymmetry coefficients. In addition, the design of the bench makes it possible to study transmissions under loads that change sharply in magnitude.

Brief technical characteristics of the bench

1. Rated power in the closed loop \((N_c)\) — 1000 hp.
2. Electric-motor power \((N)\) — 40 kW.
3. Electric-motor speed \((n)\) — 965 rpm.
4. Maximum torque on the low-speed shaft of the closed loop \((M_k)\) — 12000 kgf·cm.
5. Minimum distance between the centers of the shafts of the closed loop \((A_{\min})\) — 120 mm.
6. Maximum distance between the centers of the shafts of the closed loop \((A_{\max})\) — 300 mm.

1. Module of the wheels being tested . . . . . $(m)$ up to 6 mm
2. Maximum diameter of the wheels being tested $(D)$ 300 mm
3. Maximum face width of the wheels being tested $(B)$ 60 mm
4. Maximum gear ratio of the wheels being tested . . . . . $(i)$ 6
5. Maximum speed . . . . . . . . . . . . . $(v)$ 90 m/sec.

Figure 1 shows the general view of the installation², consisting of the following units and auxiliary systems: platforms—gearboxes 5 and 9 with experimental reducers, the gears of which are connected by shafts 7, 15, 17; an AK-82-6 type phase-rotor electric motor; multiplier 3, as well as lubrication and cooling systems, loading systems, and electrical equipment.

Fig. 1. Diagram of the high-speed test bench: 1—main electric motor; 2, 4—connecting couplings; 3—multiplier (gearbox); 5—fixed platform; 6, 18—gear couplings; 7, 15—low-speed shafts; 8, 14—loading friction couplings; 9—movable platform; 10, 22—fixed rear housings of the wheels being tested; 11, 21—movable middle housings of the wheels being tested; 12, 20—movable front housings of the wheels being tested; 13, 19—connecting blind bushings; 16—high-speed shaft (torsion shaft); 17—support of the high-speed shaft; 23—auxiliary electric motor for running-in pairs.

All units, except for the oil tank and the hydraulic loading system, are mounted on one common frame fixed to the foundation.

1. Closed loop of the test bench and its drive

The right-hand gearbox (Fig. 1) is mounted on the fixed platform 5 and is connected through the safety pin coupling 4 to the multiplier 3. The input shaft of the multiplier is connected to the main electric motor through the pin coupling 2.

The left-hand gearbox is mounted on the rocking platform 9, whose rocking axis is horizontal and lies in the same vertical plane as the axis of the middle shaft of the gearbox. The axis of the platform rests on two roller radial-thrust bearings, the housings of which are fixed to the frame. The left-hand platform (depending on the method of loading)

can remain stationary. This is achieved by means of pins that fix the platform in a strictly horizontal position.

The gearboxes have a split in the horizontal plane. Each gearbox consists of a fixed rear housing 10 and 22, a movable middle housing 11 and 21, and a movable front housing 12 and 20. The fixed housings of both gearboxes are pinned and secured on the platform. The front and middle housings can move in transverse slots and are fixed with pins at the set center distance. The middle housings are fastened with special bolts. The low-speed shafts of the gearboxes are connected to each other by means of shafts 7 and 15, loading friction (8 and 14) and gear (6 and 18) couplings. The high-speed shafts of the gearboxes are connected by torsion shaft 16 by means of blind splined bushings 13 and 19 (Fig. 1).

All shafts of the experimental gear wheels are mounted on rolling bearings of AB or A quality class. With this test-bench layout, the supports of the high-speed shafts are relieved of axial and radial loads, which makes it possible to mount them on radial bearings. The wheel shafts are mounted on radial-thrust bearings. High-speed shaft 16 is supported by steady rest 19, consisting of a housing, a bearing (rolling or sliding), and a cover.

The test-bench multiplier is single-stage, with replaceable pairs of gear wheels, and makes it possible to obtain from 150 to 6380 rpm at the input shaft of the circuit. An auxiliary motor 23 (AOL 41–6), used for running-in experimental gear pairs, is also mounted on the common bed. The large pulley of the V-belt drive \((i = 10)\) is mounted on coupling 4.

The number of loading cycles is recorded by a revolution counter.

In coupling 2 there are two replaceable safety pins, sheared off when the main motor is overloaded twofold. When centering the gearboxes, the permissible nonparallelism and crossing of the axes of the bores were established within 0.015 mm/m.

2. Lubrication and Cooling System

For lubrication of the friction pairs and for removal of the heat generated during transmission testing, the installation is equipped with a circulating lubrication system, the diagram of which is shown in Fig. 2.

Gear pump 7 supplies oil to main line 8 from tank 1 by means of intake branch pipe 6, at the end of which a mesh filter 3 is installed. An additional plate-and-slot filter 11 and flowmeter 9 are included in main line 8. After the filter (through pipeline 13), the oil is supplied to distribution manifold 15, from which it is fed to the bearings and gear pairs through special drillings in the gearbox housings.

To regulate the temperature of the oil, electric heater 2 and coil 4, through which water circulates, are installed in the tank.

When testing heavily loaded gear pairs, additional cooling of the lubricant can be introduced by connecting coils 18 and 28, located in the crankcases of gearboxes 19 and 29, to the system.

To measure the pressures after the oil pump and at the manifold, pressure gauges 10 and 16 are installed. Sensor 5 of the TK-8 remote thermal annunciator, monitoring the oil temperature, is placed in the tank near the delivery pipeline 6. To prevent possible test-bench failures caused by a drop in oil pressure in the system or by interruption of the lubricant supply, in the system of distribution pipelines

pressure line 13, in which pressure relay 12 is installed, whose contacts are included in the power-supply circuit of the main electric motor.

The design of the test bench provides for the possibility of carrying out tests both with jet lubrication and with dip lubrication.

In the first case, a jet of lubricating oil is supplied to the meshing zone through pipelines 14 and 17, respectively. From the housing, the used oil drains directly into the tank through openings in the bottoms of the gearbox crankcases 19 and 29.

Fig. 2. Lubrication and cooling system: 1—oil tank; 2—electric heater; 3—mesh filter; 4, 18, 28—coils; 5—sensor of the remote temperature alarm; 6—intake branch pipe; 7—gear pump; 8, 13, 14, 17, 23, 22, 25, 23, 27—pipelines; 9—flowmeter; 10, 16—pressure gauges; 11—plate-and-slot filter; 12—pressure relay; 15—distribution manifold; 19, 29—gearbox crankcases; 21, 24—equalizing tanks; 26—valve.

In the case of dip lubrication (when testing low-speed transmissions on the bench), a flow-through oil bath is provided (with mains 14 and 17 shut off), which is due to the need to remove a significant amount of generated heat.

To maintain the required oil level in the bath, 2 equalizing tanks 21 and 24 are provided (one for each test gearbox). The tanks are mounted on special brackets directly—

directly on the frame of the installation. In this case hoses 20 and 25 are disconnected from the tank and are connected from the side to the equalizing reservoirs through the corresponding fittings, into which oil from the test gearboxes enters. In turn, the lower branch pipe of the equalizing reservoirs is connected directly to the tank, as required, by means of Durit hoses (23 and 22), and drainage is carried out there. Draining oil from gearboxes 29 and 19 during splash lubrication is impossible, since line 27 is shut off by valve 26. The system for maintaining the oil level and draining it into the tank is shown schematically by dashed lines in Fig. 2.

The oil level in the crankcases of the wheels under test is regulated by placing the reservoirs at different heights directly on the stand frame. In doing so, the maximum diameter and module of the gear pair under test are taken into account.

3. Methods of Loading Transmissions

The stand can be loaded as follows:

1. By twisting the torsion shafts (with the stand not operating, using a socket wrench and weights).

2. By twisting the torsion shafts by rocking the platform (in this case the loading of the circuit can be carried out with the stand operating), effected either by special hydraulic systems or by mechanical systems.

In our tests the stands were loaded by means of weights. In this case the lugs of the special wrench are placed in the corresponding slots of the right-hand parts of the loading friction clutches 8 and 14 (Fig. 1), the left half of which is locked. Under the action of the applied moment, the shafts of the circuit are twisted. The friction forces of the disks arising at the calculated tightening provide transmission of the torque \(M_k = 120\text{--}130\ \text{kgm}\), which considerably exceeds the requirements of the tests. The corresponding axial tightening force \(P = 12\text{--}15\ \text{kg}\) is produced by 4 threaded plugs uniformly arranged along the diameter on the end face of the clutches.

All tests were carried out with two power flows closed through a common torsion shaft³. The circulation of power for different directions of the loading moments on the outer shafts is shown in Fig. 3a, and the direction of the power flows when the rotation is changed is shown in Fig. 3b.

Fig. 3. Power circulation for different directions of loading moments (a) and direction of power flows when the direction of rotation is changed (b): \(\omega\)—angular velocity; \(M_z\)—loading moment; \(N\)—circulating power.

When the stand operates according to a two-flow scheme with a central shaft, the type of cycle according to which both circuits operate also depends on the direction of loading of both friction clutches. Two cases are possible.

1. The loading moments of the wheel shafts are directed identically and are equal to one another (Fig. 4b). In this case the circumferential forces are

\[ P_1 = P_2 = \frac{2M_k}{d_{dk}}, \]

where \(d_{dk}\) is the diameter of the pitch circle of the wheel;

\(M_k\) is the moment on the loaded wheel.

The moment on the gear is

\[ M_{\text{sh}} = P_1 \cdot d_{d\text{sh}} = \]

\[ = \frac{2M_k}{d_{dk}} \cdot d_{d\text{sh}} = \frac{2M_k}{i}. \]

Here \(d_{d\text{sh}}\) is the diameter of the pitch circle of the gear;

\(i\) is the gear ratio of the pair being tested.

The moment on the torsion shaft is

\[ M_{\text{v sr}} = \frac{M_{k_1}}{i} + \frac{M_{k_2}}{i} = \frac{2M_k}{i}. \]

Fig. 4

In this case, the same profile will be the working profile on the gear, and the gear tooth works in bending according to a pulsating cycle (Fig. 4a) with a frequency equal to \(2n_k \cdot i\), where \(n_k\) is the number of revolutions of the wheel meshing with the gear. The load on the gear supports is \(R_n = |P_1| - |P_2|\), i.e., the supports are unloaded \((P_1 - P_2)\).

In the general case, if the loading moments of the first and second circuits are not identical (i.e., \(M_{k1} \ne M_{k2}\)), then the same profile of the gear tooth will operate according to a pulsating cycle with variable amplitude (Fig. 4c).

In this case the supports will carry a load equal to

\[ R_n = |P_1| - |P_2|. \]

1. The loading moments of the wheel shafts are directed oppositely (Fig. 4). In the general case, assuming that the loading moments of the wheels are not equal \((M_{k1} \ne M_{k2})\), the moment on the central shaft will be

\[ M_{\text{v sr}} = \frac{M_{k_1}}{i} - \frac{M_{k_2}}{i}. \]

In this case the gear teeth will operate with different profiles in bending, with cycles having any asymmetry coefficient \((r = 0 \div 1)\). A load \(R_n = P_1 + P_2\) will act on the gear supports. In particular, for equal values of the loading moments \((M_{k_1} = M_{k_2})\), \(M_{\text{v sr}} = 0\), and the load acting on the support of the intermediate shaft is \(R_n = 2P_1\).

1. The closed loop is currently also used in testing large marine transmissions. According to recent work in the USA and England, full-scale drives with circulating power up to 30,000 hp have begun to be tested on the closed benches of “Pametrada” and AVGRA. ↩

2. The design of this installation was developed by Doctor of Technical Sciences Professor V. N. Kudryavtsev; the drawings of the installation were issued under the supervision of engineer A. K. Suslov. ↩

3. The stand can also operate according to a single-flow scheme without the middle housings and intermediate shaft, and also with a “parasitic” middle gear (the stand scheme and the loading cycle in the latter case are shown in Fig. 4). ↩