GB2343515A - Method of and test rig for checking the fit of radial tooth couplings - Google Patents

Abstract

A method of checking the fit of a radial tooth coupling comprises locating a first part 9 of the coupling over a test shaft upstanding from a reference surface 14, locating the second part 8 of the coupling on the first part, applying a predetermined tensile force to the shaft to compress the two parts of the coupling together, and increasing the tensile force by a predetermined amount and measuring, 16, the change in positions of a plurality of locations around the outer surface of the second part resulting from the increase.

Description

METHOD OF AND TEST RIG FOR CHECKING THE FIT OF RADIAL TOOTH COUPLINGS Field of the Invention This invention relates to a method of checking the fit of a shaft or disc coupling device of the interlocking radial gear tooth type. Such couplings are typically used to join together parts of a gas turbine engine rotor and to transmit drive between the parts.

Background to the Invention In a gas turbine engine rotor having a bladed compressor section and a bladed turbine section connected in axial alignment, the aerodynamic forces of gases from the engine combustion section act on blades of the turbine section to turn the rotor so that in consequence blades of the compressor section also turn and air is drawn into the engine. Such engine rotors may include a number of separate disc elements which are stacked axially in concentric alignment and clamped together to make a rigid assembly capable of transmitting the full engine power. One method of simultaneously achieving disc alignment and power transmission capability in a rotor is to employ a coupling of the radial gear tooth type between the disc elements. Two typical couplings of the type are known as Hirth and Curvic couplings.

To operate efficiently and reliably it is imperative that an engine rotor has sufficient stiffness to counter a tendency to"whip"when rotating at speed; it must also be capable of withstanding and transmitting the engine output torque without movement between rotor discs and further, when completely assembled and in use it must be within strict dimensional limits in order that balance will not be adversely affected. It follows that any coupling between disc elements or sections of a rotor must therefore satisfy criteria of stiffness, torque, dimensional accuracy and balance.

For use in transmitting loads at high speed, as in gas turbine duty, radial gear coupling types must be machined to fine tolerances in order to achieve the necessary high quality of fit. Tooth profiles must be dimensionally accurate and the surface finish of the tooth flanks must be of high quality to achieve a good fit so that the coupling runs true. As an illustration of the consequences of slight errors in machining, in a gas turbine of 5MW output and a rotor speed of 17,000 rpm an error in machining resulting in a shift of the centre of gravity of the rotor of just 1 pm would generate a centrifugal force of 1200N.

Whilst radial gear tooth couplings are excellent in many respects, they are not problem-free when applied to high speed gas turbine duty. In particular, slight imperfections of tooth form may cause an assemble coupling to adopt a tilted or eccentric position and thus give rise to out-of-balance problems. While it is possible to measure coupling tilt and concentricity, and thus deduce out-of-balance forces for a static condition, just what the extent of such tilt and/or eccentricity might be in a working engine has hitherto been especially difficult to predict. This is because the imperfections give rise to an instability in the tooth fit so that, as the coupling becomes affected by forces in a working engine (heat expansion for example), the coupling position may shift slightly and consequently change the centre of gravity, which may give out-of-balance problems.

As part of the engine manufacturing process, rotors are separately assemble and individually tested on a high speed test rig to measure vibration levels. Those found to have unsatisfactory vibration levels are rejected, dismantled, reworked, re-assembled and then retested (perhaps several times over) before allowing them to be built into engines offered for sale. Clearly this additional work is both time-consuming and expensive, and anything that can be done to reduce the amount of such work will be of benefit. For gas turbine duty, traditional methods of gear tooth inspection to assess di mensional tolerances and surface finish are inadequate to assess reliably the quality of gear tooth couplings under actual service conditions, and the use of such methods causes much unnecessary work.

For example, a bail gauge inspection (in which a steel ball of set diameter is rested on the flanks of adjacent teeth and a measurement taken from the top of the ball to a datum point in order to measure tooth profile depth) is reliant on a number of geometric features such as tooth flank angles. Also, such a method may be affected by the surface finish of the tooth flank, so that if the surface has a high or low spot local to the ball position, the measurement will be adversely affected. A further traditional method to indicate the quality of contact fit between mating teeth is to use"engineers blue"on the teeth flanks, but unfortunately this method is highly dependent on operator skill, and thus gives inconsistent results in practice.

Summary of the Invention According to the invention, there is provided a method of checking the fit of a radial tooth coupling, comprising: locating a first part of the coupling over a test shaft upstanding from a reference surface; locating the second part of the coupling on the first part; applying a predetermined tensile force to the shaft to compress the two parts of the coupling together; and increasing the tensile force by a predetermined amount and measuring the change in positions of a plurality of locations around the outer surface of the second part resulting from the increase.

Preferably, the method comprises comparing the change at each location with a predetermined change value and rejecting the coupling if any change exceeds said predetermined change value. Alternatively, or additionally, the method comprises comparing the change at each location with that at each other location and rejecting the coupling if the difference between the change at any two locations exceeds a predetermined second value.

The invention also provides a test rig for testing the fit of a radial tooth coupling, the rig comprising: a reference surface having a test shaft upstanding therefrom for receiving the coupling over the shaft; means locatable on the upper part of the coupling for applying a tensile force to the shaft, thereby compressing the coupling ; and means for measuring the change in position of the shaft end and of each of a plurality of locations on the periphery of the upper part of the coupling when the tensile force applied to the shaft is increased from a first predetermined value to a second, higher, predetermined value.

The method of the invention results in lower production costs and improved throughput in the number of serviceable rotors shipped.

Brief Description of the Drawings In the drawings, which illustrate a typical coupling and one exemplary embodiment of the invention: Figure 1 is a side elevation of one of the coupling elements ; Figure 1 a is a front elevation of the element shown in Figure 1; Figure 2 is an enlarged detail view showing the meshing of the teeth of the two coupling elements together ; and Figure 3 is a part-sectional side elevation of a test rig in accordance with the invention.

Detailed Description of the Illustrated Embodiment Figures 1,1 a and 2 illustrate a typical coupling of the type with which the invention is concerned. Two matching coupling elements 1 and 6 each have a number of radially spaced matching teeth2,4 and 5 on adjacent mating side faces. The pitch lines of the teeth extend radially from the centre axis 3 of the coupling so that when the teeth are engaged the two coupling elements 1 and 6 automatically align on the common centre axis 3, the drive being transmitted between them through the mating tooth flanks.

As may be seen from Figure 2, a tooth 2 of one element 1 engages with teeth 4 and 5 of the mating coupling element 6.

In use, the two coupling elements 1 and 6 are held together under a compressive force by a bolt (not shown) through the centre axis. In practice, the teeth may be cut into side faces of rotor discs so that adjacent discs become an actual coupling pair.

Referring to Figure 3, a coupling consisting of bladed disc members 8 and 9 is positioned with its central axis vertical on a supporting surface 14. A centre bolt 11 passes through the supporting surface 14 and the disc members 8 and 9, on the centre axis 10 of the members. A hydraulic tensioning means 12 is located on the upper member 8 and engages the bolt 11 to apply a tensile force thereto, thereby compressing the members 8 and 9 between the tensioning means 12 and the surface 14. Hy- draulic fluid is supplied to the tensioning means 12 via hydraulic pipes 13. A probe 15 is positioned to measure changes in the bolt length, while four probes 16 (of which only two may be seen in Figure 3) are positioned equi-spaced around the periphery of the upper disc member 8 to measure changes in the positions of different points on the sur face of the disc member as compressive forces are applied thereto. The body of each of the probes 15 and 16 is fixed in relation to the surface 14.

The term"predetermined"is used throughout to indicate various measurement values, previously established by calculation or determined through experimentation in each case. The actual values will vary from one coupling type to another and may depend on such factors as coupling size and the number of couplings in series. The method of measuring tilt in the coupling comprises the steps set out hereinafter: 1. With a coupling assembled as shown in Figure 3, the five measuring probes are positioned with one (15) in contact with an end face of the centre bolt 11 (the centre probe) and the other four (16) in contact with the side face of the disc (the tilt measuring probes) and equi-spaced on a common pitch circle around the axis 10; 2. The centre bolt 11 is tensioned by the hydraulic tensioning means 12 to apply an initial loading to the coupling; 3. All probe readings are zeroed; 4. The hydraulic pressure In the tensioning means 12 is increased suffi ciently to increase the length of the bolt 11 by a predetermined amount, as measured by the centre probe 15, and then each individual tilt probe reading is recorded; 5. The tilt probe readings are compared with known acceptable values for the coupling type. For an acceptable coupling, the change in any individ ual tilt probe reading should not exceed a predetermined change value ; 6. Each individual tilt probe reading is then compared with each other. If the difference between any two is greater than a predetermined differ ential value, the coupling is again rejected; Where a coupling is rejected in either of tests 5 and 6, it is dismantled, and each coupling element is separately measured using the above method with a master coupling element of proved quality. In this way, faulty coupling elements can be quickly isolated for corrective action without the need to rework the complete coupling.

It will be appreciated that, while the above method is described with reference only to a pair of coupling elements constituting a single coupling, the method of the in vention may be applied to the measurement of a series of couplings in concentric alignment, for example representing several discs of a rotor. In this case, the probe measurements and the force applied would depend on the number of couplings, their sizes and constituent materials.

While the use of four probes will normally be sufficient to indicate alignment problems, the invention is not limited to the use of a particular number, and more or fewer probes may be used, for example for larger or smaller diameter couplings. In addition, the invention has been described with the reference surface horizontal, but it will be appreciated that the method of the invention is applicable regardless of the orientation of the coupling during testing.

Claims (6)

1. A method of checking the fit of a radial tooth coupling, comprising: locating a first part of the coupling over a test shaft upstanding from a reference surface ; locating a second part of the coupling on the first part; applying a predetermined tensile force to the shaft to compress the two parts of the coupling together; and increasing the tensile force by a predetermined amount and measuring the change in positions of a plurality of locations around the outer surface of the second part resulting from the increase.

2. A method according to Claim 1, comprising comparing the change at each location with a predetermined change value and rejecting the coupling if any change exceeds said predetermined change value.

3. A method according to Claim 1 or 2, comprising comparing the change at each location with that at each other location and rejecting the coupling if the difference between the change at any two locations exceeds a predetermined second value.

4. A method of checking the fit of a radial tooth coupling, substantially as described with reference to the drawings.

5. A test rig for testing the fit of a radial tooth coupling, the rig comprising: a reference surface having a test shaft upstanding therefrom for receiving the coupling over the shaft ; means locatable on the upper part of the coupling for applying a tensile force to the shaft, thereby compressing the coupling ; and means for measuring the change in position of the shaft end and of each of a plurality of locations on the periphery of the upper part of the coupling when the tensile force applied to the shaft is increased from a first predetermined value to a second, higher, predetermined value.

6. A test rig for testing the fit of a radial tooth coupling, substantially as described with reference to, or as shown in, Figure 3 of the drawings.