Deformation analysis of aircraft wheels using a speckle ... · PDF fileDeformation analysis of aircraft wheels using a speckle shearing interferometer ... all wheel units are 100 per

Loughborough UniversityInstitutional Repository

Deformation analysis ofaircraft wheels using a

speckle shearinginterferometer

This item was submitted to Loughborough University's Institutional Repositoryby the/an author.

Citation: IBRAHIM, J.S., PETZING, J.N. and TRYER, J.N., 2004. Deforma-tion analysis of aircraft wheels using a speckle shearing interferometer. Proceed-ings of the Institution of Mechanical Engineers, Part G: Journal of AerospaceEngineering, 218 (4), pp. 287-295 [DOI:10.1243/0954410041872807]

Additional Information:

• This article was published in the journal, Proceedings of the In-stitution of Mechanical Engineers, Part G: Journal Journal ofAerospace Engineering [ c© IMechE] and is also available at:http://journals.pepublishing.com/content/119782

Metadata Record: https://dspace.lboro.ac.uk/2134/3779

Publisher: c© IMechE / Professional Engineering Publishing

Please cite the published version.

https://dspace.lboro.ac.uk/2134/3779

This item was submitted to Loughborough’s Institutional Repository (https://dspace.lboro.ac.uk/) by the author and is made available under the

following Creative Commons Licence conditions.

For the full text of this licence, please go to: http://creativecommons.org/licenses/by-nc-nd/2.5/

Deformation analysis of aircraft wheels using a speckleshearing interferometer

J S Ibrahim, J N Petzing* and J R Tyrer

Wolfson School of Mechanical and Manufacturing Engineering, Loughborough University, Loughborough, UK

Abstract: Speckle shearing interferometry has developed over the last three decades and is used as anon-destructive analysis tool for aerospace engineering. The primary role of the technique hastraditionally been for the examination of defects in fibre composite structural materials, but morerecent attention has considered strain and displacement measurement. Furthermore, aerospaceregulatory bodies such as the UK Civil Aviation Authority are demanding better understanding andtraceability of the technique.

The novel work reported here has investigated the use of a Michelson-based speckle shearinginterferometer for the analysis of compression-loaded aircraft wheels and tyres. Whole-fielddeformation data have been obtained for repeated compression loading of a BAe146 main wheel,providing displacement derivative and displacement data. The quality of the measurement techniquehas been ascertained through the analysis of the repeatability and reproducibility data, withcorrelating data being produced using contact displacement transducers.

Keywords: aircraft, wheels, tyres, non-destructive testing, metrology, optical method, speckleshearing interferometry, deformation, displacement

NOTATION

IA image plane light intensity beforedeformation

IB image plane light intensity after deformationI1 wavefront intensity from first mirrorI2 wavefront intensity from second mirrorn fringe order numberqu=qx in-plane first-order partial displacement

derivativeqw=qx out-of-plane first-order partial displacement

derivativeq2w=qx2 out-of-plane second-order partial

displacement derivativeqw=qy out-of-plane first-order partial displacement

derivativex Cartesian coordinate point referenceX Cartesian coordinate system axisy Cartesian coordinate point referenceY Cartesian coordinate system axisz Cartesian coordinate point referenceZ Cartesian coordinate system axis

dx small lateral shear in the X axisdy small lateral shear in the Y axisda small angular tilty laser illumination anglel laser wavelengths standard deviationf optical phase change caused by deformationc initial optical phase

1 INTRODUCTION

The manufacture of aircraft wheels requires qualitycontrol procedures and mechanisms to identify poten-tially defective units, during manufacturing and testingstages and operational service life. Defects can beformed due to random or systematic errors within themanufacturing processes and may cause problems suchas fatigue fractures within the wheel structure. Use ofsuch defective wheel and tyre units by an aircraftoperator leads to growth of the defects, which in turncan potentially generate sudden unexpected catastrophicfailure during flight-landing operations.

Qualification and approval of all wheel units ismandatory under UK Civil Aviation Authority

The MS was received on 5 November 2003 and was accepted afterrevision for publication on 17 June 2004.* Corresponding author: Department of Mechanical Engineering,Loughborough University, Loughborough, Leicestershire, LE11 3TU,UK. e-mail:[email protected]

1

G05003 # IMechE 2004 Proc. Instn Mech. Engrs Vol. 218 Part G: J. Aerospace Engineering

(UKCAA) and Federal Aviation Administration(FAA), USA, regulations [1, 2], as well as aircraftmanufacturer internal standards. Hence, all wheel unitsare 100 per cent defect tested before customer delivery.The regulatory authorities and manufacturers have beenrequesting improvement and traceability of analysismethods, during manufacture and during routine in-service maintenance checks, linked through to thecertification of testing systems and the operatingpersonnel [3].

One particular analysis method which has seengrowing acceptance within the aerospace industries asa useful analysis tool is speckle shearing interferometry(or shearography) [4], this technique being recognizedunder an FAA Advisory Circular [5]. The technique is alaser-based non-contact optical method, originallyproposed in the early 1970s [6] and developed by anumber of different optical engineering research groupsduring the 1970s [7] and 1980s [8]. Speckle shearinginterferometry has now found some significant nicheapplications, specifically that of fault detection in fibrecomposites, and quality control of aircraft tyres.

The first application area has been the subject ofrecent UK [9] and European research contracts [10, 11],while other researchers are currently investigating theapplication of speckle shearing interferometry to rotat-ing aircraft components [12]. The exploitation of muchof this applied research has been in the form ofcommercial speckle shearing instrumentation. Furtherexamples of application analysis of speckle shearinginterferometry has been completed by The NationalAeronautics and Space Administration (NASA) withreports available on the NASA Technical ReportsServer; however, specific examples of the applicationof speckle shearing interferometry are available in theliterature [13, 14].

In the case of the latter example of aircraft tyreanalysis, equipment used to be based on photographicfilm, such as the former Grant Engineering Inc. L-RAYsystem. More recently, with the development of cameraand image-processing technology, digital speckle shear-ing interferometry equipment such as the SteinbichlerOptotechnik GmbH Intact2 and the LTI Inc. AST-series shearography tyre casing analyser now demon-strate the availability of the analysis technique. This isregarded somewhat as a mature technology with opticalsystems identifying and locating a number of differenttypes of tyre defect during partial pressure tests.

A common aspect of much of this technology is thegrowing demand for quantitative data. Using image-processing algorithms and techniques [15] it is routinefor a speckle interferometer to produce quantifiedwhole-field meshes of deformation data. However, thereis a increasing issue of the reliability of the datagenerated from speckle shearing interferometers. Thisconcern is highlighted by a mounting body of evidencefrom a number of independent sources [16–20] over the

last few years which have analysed aspects of the opticalphysics of the instrument, in order to determine sourcesof error which contribute to the final data values. Whatis clear from this work is that there are some specificoptical design features (some of which have onlyrecently come to light) which must be included withinthe instrumentation, in order to maximize the repeat-ability, reproducibility and consequently the reliabilityof the measurement method and the consequential data.

This paper examines the novel use of speckle shearinginterferometry for aircraft component analysis, andspecifically structural loading of aircraft wheels. Thework considers the speckle shearing interferometer as acomplete instrument and uniquely identifies the engi-neering reality of the repeatability and reliability of thetechnique, both in the controlled environment of themetrology laboratory, and within the industrial work-place. This work provides further evidence concerningthe quality and strictures of the optical technique andmay be of use when considering the complex issues oftechnique certification for aerospace structural analysis,and training issues for operatives.

2 SPECKLE SHEARING THEORY

Speckle shearing interferometers can be divided intogroups, depending upon the optical method by whichthe spatial derivative data is produced. The Michelsonshearing interferometer (shown schematically in Fig. 1)is one of several designs which use optical elements tosplit an incoming laser wavefront into two components,subsequently focusing these on to the image plane of therecording system [4, 6–8].

Mirrors are placed parallel to two adjacent sides ofthe 50–50 non-polarizing amplitude beam-splitting cube,while a camera system views the laser illuminated objectvia the cube and mirrors. Spatial differentiation iscontrolled by the amount of lateral displacement orshear (dx or dy) created between the two wavefronts.This optical set-up is one of two common designs incommercial equipment, due to the ability to change theinstrumentation sensitivity by altering the magnitude ofthe lateral shear. The alternative common commercialdesign uses an optical wedge, causing a fixed lateralshear determined by the wedge angle.

In the case of the Michelson interferometer, manip-ulation of the wavefront components is achieved usingthe mirrors (or prisms) positioned around the beam-splitting cube. Generally one mirror will be parallel tothe respective cube face, while the other will have a smalltilt da introduced with respect to the second cube face. Itis considered important to limit applied lateral shear toonly one axis at a time when recording measurements.Shear in both the X and the Y axes is required tocharacterize fully the slope function of an object, but

J S IBRAHIM, J N PETZING AND J R TYRER2

Proc. Instn Mech. Engrs Vol. 218 Part G: J. Aerospace Engineering G05003 # IMechE 2004

analysis of the data is difficult if multiple shears are usedsimultaneously.

The light intensity at the image plane can be describedin a simplified form which is synonymous with specklemetrology:

IA ¼ I1 þ I2 þ 2ffiffiffiffiffiffiffiffiI1I2

pcosc ð1Þ

where the optical phase term c ¼ yðxþ dx, yÞ � yðx, yÞ,and I1 and I2 are the general intensity terms of the twowavefronts. Note that the descriptor IA is used in thisinstance as a means of identifying the intensity distribu-tion, attributable to the object in its initial state.

If the object receives a static deformation, the opticalpath alters slightly, represented as a phase change offðx, yÞ and fðxþ dx, yÞ respectively for the two imagewavefronts created by the shearing optics. This creates asecond set of equations which describe the object in itsnew state after the deformation function has beenapplied, which include the change in phase, and can besummarized as

IB ¼ I1 þ I2 þ 2ffiffiffiffiffiffiffiffiI1I2

pcosðcþ fÞ ð2Þ

The change in intensity at the image plane can bedescribed as DI ¼ IA � IB and is typically achieved bystoring an initial object state reference image andsubtracting subsequent images from this reference,such that via trigonometric identities

DI ¼ � 4ffiffiffiffiffiffiffiffiI1I2

psin cþ f

2

� �sin

f2

� �ð3Þ

Examination of the relationship described in equation(3) leads to the conclusion that subtraction correlationfringe intensity (the raw data of the instrument) is at amaximum when f ¼ ð2nþ 1Þp, whereas minimum

correlation fringe intensity occurs when f ¼ 2np, where,in both cases, n is an integer number. It should be notedthat equation (3) represents a single subtractioncorrelation interferogram, and phase-stepping techni-ques [15] are typically used to extract the optical phasein a quantified form.

Analysis of the optical path length changes within theinterferometer can lead to a relationship that links theoptical phase change and the spatial derivatives, in aform that is much used and referenced for commonanalysis of speckle pattern shearing correlation fringepatterns:

f ¼ 2plð1þ cos yÞ qw

qxþ sin y

quqx

� �dx ð4Þ

This relationship relates the optical phase to the partialdisplacement derivatives qw=qx and qu=qx, where y isthe laser illumination angle with respect to the camera–image plane axis. Equation (4) is specific to theillumination axis being in the XZ plane and a horizontallateral shear dx with respect to the X axis. Three furtherequations of a similar form may be developed whenconsidering different orthogonal illumination vectorsand lateral shear directions.

When examining an object for deformations typifiedby out-of-plane (OOP) signatures, then ideally thisrequires the illumination angle y to be minimized, thusdesensitizing the instrument with respect to in-planeterms qu=qx and maximizing the OOP componentsensitivity qw=qx. While the OOP slope component isof specific interest, it should be recognized that this termis not particularly significant with respect to engineeringanalysis and requires numerical integration to displace-ment, or further numerical differentiation before flex-ural strain q2w=qx2 can be derived. It is also important

Fig. 1 Basic optical design of a speckle shearing interferometer (Nd: YAG, neodymium doped yttriumaluminium garnet; ESPSI, e$$$ speckle shearing interferometry)

DEFORMATION OF AIRCRAFT WHEELS 3


that the magnitude of the lateral shear dx is small,because this has bearing upon the sensitivity of theshearing interferometer, the quality of the transducersignal and the inclusion of additional error terms [17–19].

Using normal collimated illumination of the object,equation (4) can be reduced to the form

f ¼ 4pdxl

qwqx

ð5Þ

Assuming that data are produced in the form ofsubtraction correlation fringes, the OOP partial dis-placement derivative can be finally expressed as

qwqx¼ nl

2

1

dxð6Þ

Likewise, a similar relationship can be developed for thevertical OOP derivative component when using a smallvertical shear dy and illumination in the YZ plane:

qwqy¼ nl

2

1

dyð7Þ

Care must be taken when using these relationships as thebasis of data processing, because they only hold trueunder certain conditions, these being collimated illumi-nation of the object, viewing along the camera opticalaxis of the instrument and small shear values. Otherwork [19, 20] has already identified that equation (4) hasvery limited applicability (noted above) if quality andtraceability of output are the priorities.

3 EXPERIMENTAL DETAIL

The interferometer design was based on the Michelsonoptics and is shown in Fig. 1. This type of speckleshearing interferometer was chosen because it providedsensitivity adjustment which was required in order tooptimize the experimentation. However, it should bementioned that the Michelson optics contain more air/glass interfaces than a simple wedge-based system,resulting in a poorer system efficiency. The specificoptical design of the instrumentation was constrained bydominating issues: firstly, the layout of the wheel testingfacilities at Dunlop Aerospace Braking Systems Limitedand, secondly, the requirement for collimated illumina-tion, although every effort was made to consider andinclude instrument design features which acknowledgethe existing body of knowledge concerning error sourcesand instrument performance.

The wheel units were illuminated using a continuous-wave diode pumped neodymium-doped yttrium alumi-nium garnet, laser ðl ¼ 532 nmÞ. The laser wasexpanded and collimated to 98 mm diameter, with theillumination angle b being minimized to between 58 and

78 depending on object distance but could not bereduced further due to the optical arrangement. Theinstrument used a Pulnix TM9701D digital camera,linked through to a personal computer mounted CorecoViper Digital frame-grabbing card. Image-grabbing andimage-processing algorithms were developed and oper-ated under the Coreco WIT and Sapera softwarepackages. Two-, three- and four-step optical phasecalculations [15] were implemented, used and evaluated.Final data processing was in Matlab, with optical phaseextraction completed using PhaseVision Limitedunwrapping software.

Whole-field maps of spatial derivatives were gener-ated for all test conditions. Where appropriate, datawere numerically integrated to produce specific displa-cement information. Sylvac P10 linear variable-displace-ment transducer (LVDT) displacement probes were usedto provide pointwise correlating information, whileTekscan flexistrip pressure sensing units were used tomonitor wheel/tyre interface pressures.

Wheel testing involved a BAe146 main wheel, with aDunlop Aircraft Tyres Limited tyre, inflated to 160 lbf/in2 using dry nitrogen. Wheel loading was completed ina load-cell-calibrated Denison compression-testingmachine, specifically designed and configured for wheelloading, with Fig. 2 showing the loading system in a verysimplified diagrammatic form. Compressive loading wascompleted in two parts: firstly, a preload of 10 000 lbf(45.3 kN), 15 000 lbf (67.95 kN) or 20 000 lbf (90.6 kN)was applied, which significantly reduced movementwithin the stub axle bearings; secondly, an incrementalload of 500–2500 lbf [in 500 lbf (2.26 kN) steps] wasapplied in addition to the preload. Note that, due to theage of the equipment, all scales were marked in BritishImperial units. Where appropriate, equivalent SIderived/related units were used and are provided forreference.

Experiments were designed and conducted to provideinformation to answer several issues:

1. Data were required to identify the repeatability/reproducibility of the speckle shearing interfero-meter. This was achieved through an extensive seriesof repetitive loading experiments, by applying onepreload and then repeatedly loading and unloadingwith small incremental loads.

2. Data were required to demonstrate the capability ofthe speckle shearing interferometer to measurewhole-field wheel deformations while under largestructural loads. This was achieved by measuring thedeformation of the wheel while applying a variety ofpreloads and subsequent incremental loads, simulta-neously generating correlating data with LVDTprobes. It should be noted that applications of thisnature (very large loading and analysis using specklemetrology) have received little if any attention withinthe engineering research literature.



3. Finally, data were required by other design elementsof the aircraft wheel manufacturing company for thewheel design cycle, which was automatically afunction of the two previous stages. However, owingto commercial confidentiality issues, wheel redesign isnot discussed within this reported work.

4 DISCUSSION OF EXPERIMENTAL RESULTS

The experimental results are discussed in two parts.Firstly, it is demonstrated that the speckle shearinginterferometer can produce meaningful engineering-based numerical data from an aircraft wheel undersimulated operational loads. Secondly, the quality andconfidence of the data are discussed.

4.1 Demonstrating the output of the instrumentation

The primary output from a speckle shearing interfe-rometer is a subtraction correlation fringe pattern,which depicts lines of isodisplacement derivative orisoslope. In reality, these are difficult to interpretmanually or visually, and it requires automated extrac-tion of the optical phase terms from the fringe patterns,in order to generate useful engineering information. Thisis a routine process within speckle metrology [15]. Theseautomated stages are exemplified by Fig. 3 which showsthe wrapped and unwrapped optical phase maps for theilluminated section of the BAe146 wheel of 98 mmdiameter. The two small semicircular cut-outs at thebottom of each image in Fig. 3 are the positions of twoconsecutive wheel nuts, out of the total of 16 wheel nuts.The larger curved cut-outs at the upper sides of each

image in Fig. 3 are the position of wheel vent holes.These features have been blanked out during theanalysis because data were not present in these regions.

The wrapped-phase map highlights an issue of dataquality. In an ideal experimental situation, the grey-scalegraduations between the black and white extremities ofeach wrapped optical phase fringe (left-hand image inFig. 3) should be gradual, even and smooth [15]. Incontrast, this data set contains discontinuities and non-smooth data and is perhaps more typical of the qualityof data obtainable in a non-controlled industrialcontext. Only moderate low-pass Fourier filtering wasapplied during the use of the speckle shearing inter-ferometer. The reasons for this type of noise in the datais a function of different sources: structural vibration,local variations in object surface reflectivity, localvariations in air refractive index, and instrument errorssuch as fringe contrast, speckle size and phase-steppingnon-linearities. Removal of these noise terms is often

Fig. 3 Wrapped and unwrapped optical phase maps gener-ated from the speckle shearing interferometer whiletesting the wheel/tyre under load

Fig. 2 Simplified schematic diagram of the compression loading test machine for aircraft wheels



desirable but may be problematic if using image-processing techniques, because the spatial frequenciesof the noise terms overlap those of the specklecorrelation data itself.

Depending on the type of phase-unwrapping algo-rithm used, then noise within the wrapped optical phasemap may translate into the unwrapped map andconsequently into three-dimensional data meshes. Thisis particularly the case when simple linear unwrappingroutines, such as those supplied automatically inMatlab, are used. In this case, the research workincluded the prototyping of new unwrapping software(PhaseVision Limited) which automatically recognizesphase discontinuities caused by noise elements andsignificantly reduces noise transfer [21].

Calculating numerical data from the wrapped-phasemap is a consequence of the application of equations (5)and (6), when considering the case of the qw=qx terms.An example of the displacement derivative output isshown in Fig. 4, the data mesh being generated from theoptical phase data shown in Fig. 3. Three issues ofimportance should be mentioned: firstly, that thedisplacement derivative data is dimensionless in termsof SI units, secondly, the instrumentation has been usedto analyse the vertical OOP derivative componentsqw=qy and, thirdly, for the purposes of clarity andpresentation, the data mesh has been inverted. Hencethe wheel nut positions are now represented by the twosemicircular cut-outs at the top of the mesh.

Displacement data can be produced via numericalintegration and are demonstrated in Fig. 5. The datashow the surface of the wheel deformed away from theinstrument as a function of load. Typically, loads of 500

to 2500 lbf applied to the wheel (on top of the appliedpreload) caused surface displacement components ofbetween 2 and 15 mm. Clearly, the deformation of thewheel as a function of the preloading would besignificantly larger, but this was not measured due tomovements of the wheel bearings. Furthermore, it wasfound that incremental loads larger than 4000 lbfoverloaded the sensitivity of the instrumentation.Sensitivity adjustment could be made to allow largerload analysis, but desensitizing speckle shearing inter-ferometers tend to lead to poorer data contrast [16] andmore noise in the numerical data.

Care must be taken when interpreting the data shownin Fig. 5, because it only represents the numericalintegration of one of the two OOP displacementderivative components, namely qw=qy. The analysis forthe other OOP component qw=qx has not beenpresented here because the magnitude of the qw=qxdata was negligible in comparison with the qw=qy data.However, when analysing other structural objects withdifferent deformation characteristics, the qw=qx termmay be more significant. If so, then it would benecessary to measure and process both sets of deforma-tion derivatives qw=qx and qw=qy, numerically tointegrate and then to combine, to obtain the correctrepresentation of the OOP plane object surface dis-placement.

4.2 Issues of data quality

Defining the accuracy of a measurement instrumentshould normally involve a direct or indirect (via

Fig. 4 Displacement derivative data qw=qy calculated fromthe wrapped optical phase data

Fig. 5 Displacement data calculated from the displacementderivative data



working/transfer standards) comparison with the rele-vant primary standard. This is often achieved throughcomparison measurements, or the use of an artefact ofknown properties. Unfortunately, speckle shearinginterferometers and other similar whole-field techniquesdo not currently have a route of traceability to theprimary length standard, the metre, although a recentlyinitiated European Union funded project [22] is con-sidering these issues as its core topic area.

It is, however, possible to define the quality andconfidence of the instrument and measurements throughan understanding of the repeatability and reproduci-bility of the data. This has been completed by thepresent authors and has been demonstrated for oneparticular design of speckle shearing interferometer,ideal laboratory-based repeatability versus the reality ofindustrial measurement repeatability. This work isfurther extended by the results presented here, whichdemonstrates the repeatability of the speckle shearinginterferometer for the displacement analysis. Figure 6shows data obtained for a preload of 20 000 lbf andincremental load of 2500 lbf on the BAe146 wheel. Thesedata have been processed and developed to providedisplacement component line profiles vertically throughthe analysis area of 98 mm diameter, from the qw=qypartial derivative data.

The experimentation represented in the data withinFig. 6 was repeated on seven occasions, leading to seven

line profiles of displacement component. It is noticeablethat there is a variation with time, this possibly beingcaused by changing equipment usage (and consequentlyenvironmental noise sources) within the test environ-ment, other than the equipment reported here. Thevariability of the changing noise sources introducesmore uncertainty into the comparability of the data, butthis is the reality of operating interferometric measuringequipment in industrial environments. In many cases, itwill not be possible to control the environment, only thelocal test conditions and variables.

The operational set-up of the speckle shearinginterferometer and the wheel testing equipmentremained identical throughout the testing period,although it is recognized that issues of backlash,hysteresis, creep/relaxation in the rubber matrix of thetyre, and other machine errors may also have con-tributed to the variation. The data in Fig. 6 may besummarized by Fig. 7, which shows the mean data line,the 2s standard deviation bounds and data obtainedfrom the LVDT.

The purpose of showing the 2s bounds is that theyrepresent the 95 per cent confidence limits of the data.This leads to a maximum 2s value for the speckleshearing derived displacement data of 3.7 mm for thisparticular work. The pointwise LVDT data tend toagree with the optical data, although the extent of the1 mm error bars on the LVDT data is significant. The

Fig. 6 Repeated line profiles of displacement taken through a vertical section of the wheel profile



stated accuracy of the Sylvac P10 displacement trans-ducer (10 mm range) is 1 mm with a 0.1 mm repeatability.However, the transducer was used to measure smalldisplacements (0–12 mms), which is very close to its noisefloor. Any transducer errors may therefore become moresignificant and will influence the correlation with theoptical data. A simple example is if the transducer is notplaced normal to the measured surface, thus causing acosine error and an incorrect reading.

It is important to realize that this statement ofdisplacement component measurement quality, in termsof repeatability (or reproducibility [23]), is specific to thedesign of the speckle shearing interferometer, thesoftware and hardware specifications of the imageprocessing, and the operational characteristics of theaircraft wheel compression-testing machine. If themeasurement instrument is transferred to anotherapplications area, then the experimental variables (anderror sources) will be changed and potentially therepeatability may also change. An additional issue thatshould also be considered is the form of the object beinganalysed, because the relationships discussed in section 2are particularly relevant to the analysis of planar (flat)surfaces. If the object being analysed is not planar, thenthe change in geometry will cause a change ininstrument sensitivity, which if not taken into accountwill create further errors in the data processing andinterpretation.

5 CONCLUSIONS

This paper has described the novel and successfulapplication of speckle shearing interferometry to theanalysis of large-scale, structurally loaded aircraftwheels. Specifically, a BAe146 main wheel has under-gone a series of compression load tests, using a preloadof between 10 000 and 20 000 lbf, with incremental loadsof 500–2500 lbf. Whole-field non-contact analysis hasidentified the real-time deformation of the wheelstructure. These data have been processed to producedisplacement component data showing the surfacedeformations and have been obtained on a routinebasis, in what can be described as an active industrialenvironment, with no form of vibration isolation.

Another important novel aspect of this work has beento consider the quality and confidence of the measure-ments. The shearing interferometer has measured sur-face displacement components in the range 0–15 mm,which correlates with pointwise data collected using anLVDT. Through repeated experimentation, the 2sdisplacement component repeatability of the instrumentfor this specific object/load case has been shown to be3.7 mms, although it is important to note that this valuewill change if the experimental conditions or the designof the optical transducer are changed.

As part of the discussion, the work has identifiedcritical aspects of speckle shearing interferometer design

Fig. 7 2s repeatability bands and LVDT data, compared with the mean displacement profile derived fromthe speckle shearing interferometer



and application, all of which will influence the qualityand integrity of data produced. With respect to currentUK CAA and FAA regulations, this work has generatedevidence of data confidence, instrumentation designcriteria and instrumentation procedure, which will bevaluable within the aerospace (and related) sector.Finally, the work has demonstrated that the instrumen-tation can produce deformation data in standardengineering measurands which is suitable for designanalysis verification.

ACKNOWLEDGEMENTS

The work reported has been completed under theauspices of the DTI/LINK Engineering and PhysicalSciences Research Council sensors and sensor systems forindustrial applications contract; ‘Aircraft tyre examina-tion using optical strain testing (AIRTEST)’ (GR/M78960), in association with Dunlop Aerospace BrakingSystems Limited and Laser Optical Engineering Limited.

REFERENCES

1 Aircraft wheels and wheel-brake assemblies. Technical

Standard Order TSO-C26c, Federal Aviation Administra-

tion, Washington, DC, 1984.

2 Tyres and wheels fitted to aircraft certificated in the

transport category. Airworthiness Notice 93(5), UK Civil

Aviation Authority, March 2002.

3 Personnel certification for non-destructive testing of air-

craft, engines, components and materials. Airworthiness

Notice 94(7), UK Civil Aviation Authority, March 1999.

4 Steinchen, W. and Yang, L. Digital Shearography: Theory

and Application of Digital Speckle Pattern Shearing

Interferometry, 2003 (Society of Photo-optical Instrumen-

tation Engineers, Bellingham, Washington).

5 Acceptable methods, techniques and practices—aircraft

inspection and repair. Advisory Circular AC 43.13-IB,

Federal Aviation Administration, Washington, DC, 2001.

6 Leendertz, J. A. and Butters, J. N. An image shearing

speckle pattern interferometer for measuring bending

moments. J. Physics E, 1973, 6, 1107–1110.

7 Hung, Y. Y. and Taylor, C. E. Measurement of slopes of

structural deflections by speckle shearing interferometry.

Expl Mechanics, 1974, 14, 281–285.

8 Sirohi, R. S. Speckle Metrology, 1993 (Marcel Dekker,

New York).

9 Huntley, J. M., Richardson, M. O. W. and Nurse, A. D. In-

situ quantitative flow detection using advanced optical

inspection techniques. Report (GR/M57835), Engineering

and Physical Sciences Research Council, 2003.

10 Pulsed digital holography and shearography. BRITE/

EURAM 3 Report (BRPR 960152), EU Fourth Frame-

work Programme, 1999.

11 Multi-wavelength shearography. BRITE/EURAM 3

Report (BRPR 960272), EU Fourth Framework Pro-

gramme, 1999.

12 Irving, P. E., Tatam, R. P. and Thompson, C. P. Damage

detection and damage tolerance in rotating mechanical

systems. Report (GR/N11872), Engineering and Physical

Sciences Research Council, 2003.

13 Parker, S. C. J. and Salter, P. L. A novel shearography

system for aerospace non-destructive testing. Proc. Instn

Mech. Engrs, Part G: J. Aerospace Engineering, 1999,

213(G1) 23–33.

14 Steinchen, W., Yang, L. and Mackel, P. Non-destructive

testing of aerospace composite materials using digital

shearography. Proc. Instn Mech. Engrs, Part G: J. Aero-

space Engineering, 1999, 212(G1) 21–30.

15 Robinson, D. W. and Reid, G. T. Interferogram Analysis:

Digital Fringe Pattern Measurement Techniques, 1993

(Institute of Physics, Bristol).

16 Ng, T. W. and Chau, F. S. Statistical properties of phase

variation in shearing speckle interferometry and its effect

on fringe visibility. J. Mod. Optics, 1992, 39(12), 2463–

2469.

17 Krishna Mohan, N., Masalkar, P. J., Murukeshan, V. M.

and Sirohi, R. S. Separation of the influence of in-plane

displacement in multiaperture speckle shear interferometry.

Opt. Engng, 1994, 33(6), 1973–1982.

18 Steinchen, W., Yang, L., Kupfer, G., Mackel, P. and

Vossing, F. Strain analysis by means of digital shearo-

graphy: potential, limitations and demonstration. J. Strain

Analysis, 1998, 33(2), 171–182.

19 Wan Abdullah, W. S., Petzing, J. N. and Tyrer, J. R. Wave-

front divergence: a source of error in quantified speckle

shearing data. J. Mod. Optics, 2001, 48(5), 757–772.

20 Farrant, D. I. and Petzing, J. N. Sensitivity errors in

interferometric deformation metrology. Appl. Optics, 2003,

42(28), 5634–5641.

21 Huntley, J. M. Automated fringe pattern analysis in

experimental mechanics: a review. J. Strain Analysis,

1998, 33(2), 105–125.

22 Standardisation project for optical techniques of strain

measurement. Report G6RD-CT-2002-00856 (SPOTS),

EU Sixth Framework Programme, 2002.

23 PD 6461: Part 2 Vocabulary of Metrology: Vocabulary of

Legal Metrology—Fundamental Terms, 1980 (British Stan-

dards Institution, London).



Documents

Deformation analysis of aircraft wheels using a speckle ... · PDF fileDeformation analysis of aircraft wheels using a speckle shearing interferometer ... all wheel units are 100 per