GB2543535A - Test methods and apparatus - Google Patents

Abstract

A method and apparatus of testing, e.g. fatigue testing, a component 8, e.g. tubular such as a gas turbine engine rotary component, comprises enclosing the component to define first 2 and second 3 pressure chambers on opposite sides thereof, the component providing a fluid-tight wall between the pressure chambers; and adjusting the respective pressures of fluid in the first and second pressure chambers to apply stress, e.g. hoop stress, to the component. The component may be subjected to a predetermined pattern, e.g. a repeating cycle, of variation of differential pressure between the first and second pressure chambers and/ or variation of temperature. The component may be monitored by sensors 41 for detecting stress, strain, displacement, expansion, contraction, deformation, temperature, surface state, crack development, joint integrity and/ or structural failure of the component during test. The stresses of high-speed rotation conditions can be simulated in a static rig.

Description

TEST METHODS AND APPARATUS

This invention relates to methods and apparatus for testing mechanical components under load. Particularly but not exclusively it is relevant for fatigue testing, such as high frequency or low frequency cycling testing. We particularly envisage application with engine components, such as gas turbine components or other aeronautical engineering components, but the skilled person will understand from what follows that the methods and apparatus can be useful in other areas.

BACKGROUND

The components of gas turbine engines, and particularly the rotating components, operate under extremes of mechanical stress, temperature and temperature gradient. For obvious reasons of integrity and safety in such a large and critical piece of equipment, components are rigorously tested to check design and to confirm their adequacy for an intended lifetime.

Where possible, testing simulates operational conditions with actual components at full scale. Thus, for example a segment of a rotor shaft may be mounted in a test rig adapted to drive high speed rotation at a controlled raised temperature for an extended period. The test may be run until the part fails, or until a prescribed number of cycles corresponding to an intended service life is completed. Such testing is difficult because of the challenge of monitoring and instrumentation at high rotation speeds. Also, it is seldom possible to simulate actual load cycles with their complex combination of thermal and stress conditions. Some kinds of operational load are difficult to simulate in a test rig. This leads to caution, i.e. reduction in the service life prescribed.

Another known method simulates engine loading on small (sub-element) test pieces in a fatigue test machine. Residual stress and surface finish factors in the small specimens make them represent only poorly the corresponding element of the actual component. Machining of specimens tends to relieve residual stresses which would be present in the full component. There is a risk of missing properties associated with structural features if these features are divided between tested subelements. Because the specimens are small many tests are needed to represent the overall component, making this method expensive.

It is also known to test the fatigue strength of hollow aerospace components by filling them with pressurised gas to apply tensile load.

THE INVENTION

The aim herein is to provide new and useful testing methods and corresponding apparatus, enabling testing of large-scale or full-scale components more conveniently and offering the possibility of different modes of loading.

We propose a method of fatigue testing a component, comprising enclosing the component to define first and second separate pressure chambers on opposite (first and second) sides thereof, and adjusting the respective pressures of fluid in the first and second pressure chambers to apply stress to the component.

The component provides a fluid-tight wall which defines at least part of the pressure chamber(s). Thus, the term “component” may refer to a complete component, or to a part of a larger component being a part that provides a fluid-tight wall. If necessary, a component may be modified for test purposes to make it fluid-tight e.g. by closing off or not forming one or more openings that would be present in the actual operational component.

The method is particularly suitable when the component is hollow or tubular, the first and second pressure chambers then being inner and outer pressure chambers respectively. In particular the procedure is suitable for testing of rotary components which are hollow or tubular in form, such as shaft components of gas turbine engines. However, the method is applicable to a wide range of component forms where an operational mode can be represented or simulated by differential pressure between opposed first and second faces of the component. This may be relevant for a range of flat or curved layer-form components, as well as tubular or hollow components. Use with metal components is particularly envisaged.

With tubular components such as shaft segments a centrifugal operational load due to rotation can be simulated by raised relative pressure in the first/inner pressure chamber. Because it is not necessary to rotate or even move the component under test, monitoring of behaviour or properties, such as stress, strain, temperature, surface state, cracking etc. becomes much more practical and effective than with a fast-moving test piece. The positioning and operation of monitors and sensors, positioned on or directed at the component under test, can be straightforward.

As will be understood, the method is particularly suited for testing components (or component portions) in the form of a wall or closed layer. Testing is likely to be applicable to components in which this wall has one or more localised structural nonuniformities or discontinuities, such as a join (e.g. weld or other bond) or joint between adjacent parts, a change of material or composition, or change of dimension/structure such as a change of thickness, diameter or angle relative to an operational movement axis: in short any structural feature that may raise a question about the ability to withstand operational loads.

The test procedure may comprise maintaining a fluid pressure difference between the first and second pressure chambers over a test period. The pressure difference may be maintained at a constant level, or varied e.g. in a cycling or repeated manner, or according to any other predetermined pattern corresponding to an operational load pattern or simulation thereof.

Importantly, by providing pressure chambers to either side of a component it becomes possible to subject that component to excess pressure from one side or the other, and to do both at different stages of a test procedure. Conveniently this can be done using only pressures at or above atmospheric pressure, which simplifies pumping.

Importantly, the temperature of the test component may be controlled by controlling the temperature of fluid supplied to the first and/or second pressure chamber. The fluid temperatures in the chambers may be different as part of the test regime.

Desirably the temperature is varied or is variable over a range of at least 300°C, more preferably at least 400°C, still more preferably at least 500°C.

To define the separate pressure chambers, edge or end portions of the component are desirably sealingly engaged by corresponding or complementary walls of a pressure housing comprised in the test apparatus. Such a pressure housing may have opposed end walls which are brought into sealed engagement against opposite end of edge portions of the component to be tested. Inner surfaces of these housing end walls may then define wholly or partly the first or second pressure chambers. The pressure housing may incorporate feed openings for a supply of fluid under controlled and adjustable pressure to each of the first and second pressure chambers. It will be understood that while we refer here to first and second pressure chambers, there may be more than two pressure chambers in some cases e.g. where the component to be tested is in some way segmented or compartmented, or has different portions which would be subject to different operational conditions.

In a preferred embodiment a component to be tested has a tubular form, e.g. as mentioned above, and has opposite ends one or both ends having a respective end face or end edge. The test housing has opposed end walls which are brought into sealed engagement against the respective ends, especially end faces/edges, of the component, so that the component separates the inner pressure chamber from an outer pressure chamber. A peripheral wall of the test housing may be provided to enclose the outer pressure chamber; depending on the circumstances the peripheral housing wall may be integral with one or both of the end walls thereof, or may be a separate component, provided that the necessary enclosed chamber can be formed thereby.

In this arrangement a stress loading corresponding to rotation in use, such as tensile hoop stress, can be simulated by raised pressure in the inner chamber relative to the outer chamber. Compressive wall stress, such as a compressive hoop stress in an annular component, can be simulated by raised pressure in the outer chamber relative to the inner chamber. Further, by adjusting the temperature of the fluid in the chambers an actual thermo-mechanical loading condition can be represented, e g. corresponding to that of a rotary component in a gas turbine engine. Typically a pressure difference of at least 1 MPa, preferably at least 2 MPa, more preferably at least 5 MPa and perhaps 10 MPa or more is used in the method, and the pressurised fluid supply is able to pump at corresponding pressures.

As with the differential pressure loading, temperature conditions may be maintained, modified or cycled according to a predetermined pattern throughout the test period.

The pressure fluid used in the first and second pressure chambers may be gas or liquid according to the intended test conditions, component materials and available facilities. As the skilled person will appreciate, liquid is safer under high pressure whereas gas is superior if significant heat transfer is needed. The chemical nature of the fluid may also be part of the test regime, for example where possible corrosion is to be monitored. Sealing between a pressure housing and the test component may use auxiliary (deformable) seal components if necessary, or may be by direct interface contact between component and housing, with fastening or clamping if appropriate. This may be determined by the skilled person taking into account the materials being tested, the pressure differences involved and the fitting force available or desired between housing and component.

As well as applying stress to the test component by differential fluid pressures, and optionally also temperature load, the method and apparatus may apply mechanical test load to the test component. For example the component may be subject to compression and/or torsion between opposed walls of a pressure housing. So that such loads may be effectively transferred to the component under test, the apparatus may provide for some relative movement between apparatus portions engaging the test component while maintaining the pressure differential between the pressure chambers. For example, mechanical loading may be by a loading apparatus or device separate from but movable relatively to (e.g. housed within) the pressure housing, or part of the pressure housing may be deformable, slidable or otherwise allow for relative movement to accommodate controlled application of mechanical stress to the test component. This may be simply by providing that one or more pressure housing wall portions extending in the direction of the intended stress are sufficiently deformable, e.g. allowing sufficient elastic flexing in a peripheral wall portion around a tubular component as described above.

Typical tests to be implemented using the invention include thermomechanical cycles, tension-compression cycles, and testing of annular components in hoop tension and/or compression, optionally with axial and/or torsional load. Testing of turbine shaft elements is particularly envisaged. Generally the test method includes the measurement and/or monitoring of characteristics of the tested component during the test. At the simplest, a component may be tested until it fails. Joint integrity e.g. of welds may be tested. Other properties and events which may be monitored and measured, and the means for monitoring/measuring them, include the following. Crack development: this may be monitored by means of any of pressure drop, acoustics (listening for crack formation, or sending ultrasonic pulses through the structure and monitoring for any attenuation), voltage drop measured across a potential crack site, eddy current detection, visual detection either by eye e.g. through a viewing port, or by fibre optic inspection. Strain: this can be monitored using a strain gauge or extensometer by contact, or by image monitoring. Bodily displacement of the component or part thereof, e.g. accompanying expansion, contraction or deformation such as bending: this can be monitored by a suitable position/displacement transducer such as an LVDT. Temperature: this can be monitored directly by a thermocouple or remotely by a pyrometer.

The test rig of the invention and the operation of the corresponding method can be static, or quasi-static (because only fluid need move), so the mounting and operation of monitors and sensors is greatly facilitated. Some instrument types can be housed inside the apparatus, while any necessary connections between the inside and outside of the apparatus can pass through sealed couplings in the wall.

Control and monitoring can be more detailed and accurate (as compared with a rapidly moving test piece), so the present proposals offer the possibility to achieve a better understanding of response and potential failure modes under the test conditions, leading to more efficient component design relative to design lifetime.

The proposals are now illustrated by way of example with reference to the accompanying figures illustrating schematically an apparatus and method for testing a shaft segment (specifically, an inter-stage drive arm) of a gas turbine engine, wherein

Fig. 1 is a schematic axial cross-section through part of a test apparatus with the drive arm in place for testing;

Fig. 2 is a schematic graph showing the ability to apply a range of load conditions, and

Fig. 3 is a graph showing how an engine thermo-mechanical fatigue test cycle can be simulated.

Fig. 1 shows schematically the set-up for fatigue testing a test component 8 -specifically, a shaft drive arm of a gas turbine - in a test rig 1. The drive arm 8 is a component of the gas turbine shaft extending between two rotors; it is a generally cylindrical component usually made from a nickel-based superalloy with a cylindrical wall 81, and in this case having a circumferential weld zone 83 with a weld 84 (typically an inertia weld, i.e. a friction weld) joining two tubular segments to form the drive arm segment of the shaft. On its outer surface the drive arm 8 carries two integrally-formed seal fins 85, which when in situ inhibit gas flow along between the shaft and an inwardly-projecting stator component. At each circular end edge the drive arm 8 is formed for test purposes with a radially-projecting securing flange 82.

The dimensions of the drive arm 8 might be for example 200 - 400 mm diameter and 1 - 10 kg in weight and its operational conditions include rates of revolution up to 14,000 rpm and temperatures up to 750°C. Direct simulation of such conditions in a test rig is very challenging and it would be highly desirable to provide meaningful testing under more convenient conditions. Features particularly raising performance/reliability questions are the non-uniformities in the structure, including the weld 84 where there may be high residual stresses, the surrounding weld zone 83 which is also stressed by the weld and where the wall thickness changes, and the seal fins 85 which have complex stress patterns and surface properties associated with their forming processes.

Fig. 1 shows schematically and in axial/radial cross-section (from the centre line “C”) a test rig 1 for subjecting the drive arm 8 to test loads for fatigue testing.

The test rig 1 essentially consists of a pressure housing with means for adjusting fluid pressures inside it in different zones or chambers. It comprises first and second flat opposed end plates 11 and a generally cylindrical tubular outer wall 12 whose annular end faces abut sealingly around corresponding peripheries of the end plates 11, so that they enclose a cylindrical volume between them. The end plates 11 are secured onto the peripheral wall by bolts 99.

The peripheral wall is made the same length as the segment of turbine shaft drive arm 8 to be tested, and the latter is placed coaxially in the housing with its end flanges 82 abutting against the respective inwardly-directed faces of the end plates 11. This abutment is fluid-tight, by means of bolts (schematically indicated at 98) clamping the flanges 82 to the end plates 11 and/or by means of deformable seal elements (not shown) positioned between them. The tubular wall 81 of the drive arm 8 is a closed wall, and divides the interior of the rig housing 11,12 into a central or inner pressure chamber 2 and a peripheral or outer annular pressure chamber 3. Respective communication openings for pressurised fluid are provided between the pressure chambers 2,3 and the exterior, shown here schematically as an inner pressure feed 62, an outer pressure feed 63, an inner vent 72 and an outer vent 73. By connection of the feeds to appropriate supplies of pressurised gas/liquid and control of the vents, the chambers 2,3 can be adjusted to any desired pressure. Suitable gases may be selected in line with the technical situation and the skilled person’s knowledge, and include argon, nitrogen and air. An oil is usually suitable if liquid is to be used. The skilled person is familiar with suitable high pressure delivery pumps, typically of the displacement type. The fluid supplies may also be heated by external heaters (not shown). Accordingly, it is possible to control the pressures on the tubular wall 81 from the inner and outer zones 2,3 to load the wall in hoop tension (to the extent that the inner pressure exceeds the outer) or hoop compression (to the extent that the outer exceeds the inner). At the same time, the fluid temperature can be controlled to adjust the temperature conditions experienced by the component 8.

Optionally, and as indicated by arrows A, the component 8 may be placed under axial compression by outside loading on the end plates 11 provided that the housing peripheral wall 12 is moderated in stiffness to allow for this. Alternatively it may be placed under axial tension if the flanges 82 are bolted on and the housing walls allow for some expansion movement while maintaining sealing under these conditions. Or, torsion may be applied.

The skilled person will appreciate that by these means the inter-stage shaft component 8 can be subjected to thermo-mechanical stress cycles in a manner designed to correspond to operational loading. Because the rig 1 is essentially static, sensors and monitors can easily be arranged to operate in the rig to determine its response. Fig. 1 shows schematically a generic sensor arrangement 4 including a sensor/transducer 41 on the test piece, a lead 42 communicating out through a gas-tight fitting of the wall 12, and a read-out or recorder device 43 for the corresponding parameter.

Fig. 2 shows schematically how, because of the ability to control pressures Pi and P2 independently in the first and second pressure chambers 2,3, the component may be placed under a cycle of either compressive or tensile hoop stress over time.

By way of example, Fig. 3 shows how an apparatus and method embodying the invention, such as shown in Fig. 1, can be used to simulate the operating conditions for the drive arm component 8 shown in Fig. 1. In Fig. 3 the hoop stress in the component is shown on the y axis (negative values being compressive, positive tensile). The x axis represents temperature. The curved line X shows a typical operational stress/temperature TMF cycle. Initially, such as on take-off, the outside of the shaft heats much more quickly than the bore side: it wants to expand but cannot, because the bore holds it in, and it becomes highly compressively stressed. This compressive stress relaxes as the bore side heats up and engine speed reduces after take-off. The cycle reaches a maximum of tensile stress at a lower temperature and lower absolute stress value. These conditions are not practically reproducible in a spin rig. Flowever they can be simulated by the present invention, as illustrated in simplified form by the triangular test profile shown in Fig. 3 with Stages A, B, C. In Stage A the outer chamber 3 is pressurised progressively, using heated gas, to a pressure excess ΔΡαβ over the inner chamber which is above 10 MPa or more. The exact ΔΡ to be applied can readily be calculated based on the diameter and thickness of the test component. In Stage B pressure in the outer chamber 3 is progressively reduced and that in the inner chamber 2 progressively increased, at the same time the fluid temperatures being moderated, to reach a condition in which there is excess pressure APBc in the inner chamber 2 corresponding to the anticipated operational tensile hoop stress in the component. Again this pressure difference is readily calculated, and might be e.g. between 5 and 10 MPa. In a third Stage C the pressures of the inner and outer chambers 2,3 are progressively equalised and the temperature reduced.

By this means the component is subjected to temperature and hoop stress conditions corresponding to those arising during take-off and shut-down but without needing to be moved. Monitoring of the component over a large number of cycles, e.g. 10,000 cycles or more, then becomes practical. A particular difficulty for the corresponding spin rig test is that the peak tensile stress of the operational cycle does not occur at the peak temperature; attempts to run a spin rig at both the peak stress and peak temperature result in non-representative extreme loads leading to a non-representative failure of the test piece. By contrast, the easy variability and control of hoop stress in the present method enables this aspect of the operational conditions to be accommodated readily.

Claims (21)

1. A method of testing a component, comprising enclosing the component to define first and second pressure chambers on opposite sides thereof, the component providing a fluid-tight wall between the first and second pressure chambers, and adjusting the respective pressures of fluid in the first and second pressure chambers to apply stress to the component.

2. A method according to claim 1 in which the component is enclosed in a test housing having opposed end walls, the component has opposite ends each with an end face or end edge, and the opposed end walls of the test housing are brought into sealed engagement with the end face or end edge of the respective component end.

3. A method according to claim 1 or claim 2 in which the component is tested by applying varying differential pressure between the first and second pressure chambers.

4. A method according to claim 3 in which the predetermined pattern of varying differential pressure includes at least one stage when the first pressure chamber is at higher pressure than the second pressure chamber and at least one stage when the second pressure chamber is at higher pressure than the first pressure chamber.

5. A method according to any one of the preceding claims in which the difference between the pressures in the first and second pressure chambers reaches at least 1 MPa.

6. A method according to claim 5 in which the difference between the pressures in the first and second pressure chambers reaches at least 5 MPa.

7. A method according to any one of the preceding claims in which the pressures in the first and second pressure chambers are at or above atmospheric pressure throughout the test.

8. A method according to any one of the preceding claims comprising controlling the temperature of the component by controlling the temperature of fluid supplied to the first and/or second pressure chamber.

9. A method according to claim 8 in which the temperature of a said fluid is varied over a range of at least 300°C.

10. A method according to any one of the preceding claims comprising subjecting the component to a predetermined pattern of variation of differential pressure between the first and second pressure chambers and/or variation of temperature.

11. A method according to claim 10 in which the predetermined pattern is a repeating cycle.

12. A method according to any one of the preceding claims in which the component is hollow or tubular in form, the first and second pressure chambers being respectively inner and outer chambers relative to the component.

13. A method according to claim 12 in which the component is a gas turbine engine shaft component.

14. A method according to any one of the preceding claims comprising monitoring any one or more of stress, strain, displacement, expansion, contraction, deformation, temperature, surface state, crack development, joint integrity or structural failure of the component during the test.

15. A method of fatigue testing a tubular component in the form of a gas turbine engine rotary component, the tubular component having opposite ends each having a respective end edge, comprising enclosing the component in a test housing having a peripheral wall and opposed end walls, the end walls being brought into sealed engagement against the edges of the respective ends of the component to define an inner pressure chamber inside the component and an outer pressure chamber between the component and peripheral wall, and applying hoop stress to the component by adjusting the respective pressures of fluid in the inner and outer chambers to in a predetermined pattern of pressure variation.

16. A method of claim 15 in which the predetermined pattern includes at least one stage when the inner pressure chamber is at higher pressure than the outer pressure chamber applying tensile hoop stress and at least one stage when the outer pressure chamber is at higher pressure than the inner pressure chamber applying compressive hoop stress.

17. A method of claim 15 or 16 in which a said fluid is heated to heat the component.

18. A method according to any one of claims 15 to 17 comprising monitoring any one or more of stress, strain, displacement, expansion, contraction, deformation, temperature, surface state, crack development, joint integrity or structural failure of the component during the test.

19. Test apparatus for implementing a test method according to any one of claims 1 to 18, comprising a test housing with opposed end walls and a peripheral wall shaped and dimensioned to surround the component and define the first and second or inner and outer pressure chambers in relation thereto, and means for supplying pressurised fluid to the pressure chambers whereby either chamber can be put under excess pressure relative to the other at a pressure difference of 1 MPa or more.

20. Test apparatus according to claim 19 comprising a heater for heating a said fluid to 300°C or more.

21. Test apparatus according to claim 19 comprising one or more instruments for monitoring any one or more of stress, strain, displacement, expansion, contraction, deformation, temperature, surface state, crack development, joint integrity or structural failure of a said component in the test housing during testing.