[机械模具数控自动化专业毕业设计外文文献及翻译]【期刊】评估用于汽车尺寸检测的多传感器水平双臂坐标测量机-外文文献

Int J Adv Manuf Technol (2014) 72:16651675DOI 10.1007/s00170-014-5737-3ORIGINAL ARTICLEEvaluation of a multi-sensor horizontal dual arm CoordinateMeasuring Machine for automotive dimensional inspectionGlen A. Turley Ercihan Kiraci Alan Olifent Alex Attridge Manoj K. Tiwari Mark A. WilliamsReceived: 8 November 2013 / Accepted: 23 February 2014 / Published online: 30 March 2014 The Author(s) 2014. This article is published with open access at SAbstract Multi-sensor coordinate measuring machines(CMM) have a potential performance advantage over exist-ing CMM systems by offering the accuracy of a touchtrigger probe with the speed of a laser scanner. Before thesesystems can be used, it is important that both random andsystematic errors are evaluated within the context of itsintended application. At present, the performance of a multi-sensor CMM, particularly of the laser scanner, has not beenevaluated within an automotive environment. This studyused a full-scale CNC machined physical representation ofa sheet metal vehicle body to evaluate the measurementagreement and repeatability of critical surface points usinga multi-sensor horizontal dual arm CMM. It was found thatthere were errors between CMM arms and with regard topart coordinate frame construction when using the differ-ent probing systems. However, the most significant effectupon measurement error was the spatial location of the sur-face feature. Therefore, for each feature on an automotiveassembly, measurement agreement and repeatability has tobe individually determined to access its acceptability formeasurement with a laser scanner to improve CMM utili-sation, or whether the accuracy of a touch trigger probe isrequired.G. A. Turley (envelopeback) E. Kiraci A. Attridge M. A. WilliamsWMG, The University of Warwick, Coventry, CV4 7AL, UKe-mail: glen.turleywarwick.ac.ukA. OlifentJaguar Land Rover Limited, Abbey Road, Whitley, Coventry CV34LF, UKM. K. TiwariIndian Institute of Technology Kharagpur, Kharagpur, 721302,IndiaKeywords CMM Laser scanner Measurement systemsassessment1 IntroductionAccurate measurements are important in being able to mon-itor production processes and ensure conformity to designspecifications 1. In the automotive industry, inspectionof key dimensional and geometrical tolerances is typicallydone using a coordinate measuring machine (CMM) 2, 3.Traditionally, a touch trigger probe has been employed asthe CMM sensing mechanism because of well-establishedcalibration processes and knowledge of measurement uncer-tainties 4, 5. More recently, due to the increase in accuracyof non-contact laser triangulation sensors (LTS), CMMmachines with both contact and non-contact sensing sys-tems are becoming common 1, 5. LTS digitising, morecommonly referred to as laser scanning, offer advantagesover the touch trigger method with regards to faster mea-surement speed, higher resolution and non-contact measure-ment to prevent local part deformation during inspection68. All three are important for automotive manufac-turers offering improvements in CMM utilisation, betterestimation of feature characteristics as well as measurementvalidity by preventing sheet metal and plastic componentdisplacement during inspection 2, 9. CMM utilisation is anarea of particular interest, as greater measurement through-put will provide a better estimation of process capabilityand prevent the need for further CMM investment. How-ever, while laser scanning is well-established in the reverseengineering field and in the inspection of freeform sur-faces, lack of knowledge about measurement uncertaintieshas meant their use in the dimensional control of mechanicalparts has been more limited 4, 10, 11. This is particularly1666 Int J Adv Manuf Technol (2014) 72:16651675true with regard to large volume components, such as auto-motive assemblies, where measurement accuracy dependsupon both the surface material and the geometrical ele-ments which have to be measured 11. Both of these factorsaffect laser scanner measurement uncertainty as the technol-ogy is sensitive to issues such as colour, surface roughnessand reflectivity which do not influence tactile measurements4, 12. However, it has been recognised that the greaterpoint density provided by laser scanners can lead to bet-ter conditioned fitting algorithms for characteristics suchas the diameter or centre point of a sphere or circle 7,13. Multi-sensor data fusion, which combines data fromtwo or more sensors in a common spatial representationalformat, has been proposed as a way to provide greater mea-surement information while maintaining or improving themeasurement uncertainty 11. Consequently, the combina-tion of using a touch trigger probe in conjunction with alaser scanner has the potential to provide accuracy, speedas well as detailed surface information about the measuredartefact. However, because the measurement uncertainty ofa laser scanner is affected by both the artefact being scannedand the particular surface feature, an evaluation of a multi-sensor CMM needs to be relevant to its intended purpose.Therefore, the main objective of this study was to use anautomotive artefact to assess the suitability of laser scanningin the dimensional inspection of an automotive vehicle bodyas part of a multi-sensor solution.2 BackgroundA CMM is a computer numerically controlled (CNC) mea-surement system used to detect the spatial coordinates of anartefact surface with the aid of a probing device. It typicallyhas 5 degrees of freedom with three X,Y,Z translations anda further two A and B rotations at the probe head. TheX,Y,Z translations provide controlled displacements withinthe work envelope of the CMM while the A and B rota-tions ideally orientate the measurement probe with respectto the artefact surface 7, 12. The touch trigger probe isa tactile sensing element which uses nominal (theoreticallyperfect) part geometry to approach normal to the artefactsurface and triggers once contact has been made. The CMMtreatment system corrects this measured point according toknown error variables derived from the CMM calibration toprovide a comparison with the nominal part geometry 14.In contrast, the laser scanning probe consists of a transmit-ter that emits a focussed laser line onto the artefact surfacewhich diffuses the laser light. This diffuse reflection is fil-tered by the receiving lens of the laser scanner and focussedto form a 2D laser image on the LTS photo detector 6, 15.The CMM calibration then provides an error-compensatedtransformation which converts the coordinates of the photodetector 2D image into the 3D spatial coordinates of theCMM 15. Therefore, as the laser scanner moves acrossthe artefact surface, detailed 3D point cloud information isgenerated which can be compared with the nominal partgeometry. The different data acquisition methods of thetouch trigger and laser scanning probes cause systematicerrors which affect the agreement between the measurementresults and also random errors which affect individual proberepeatability.2.1 Laser scanner measurement uncertaintiesThere have been a number of published tests which haveevaluated the measurement agreement and repeatability ofCMM-mounted laser scanning probes compared to touchtrigger probes. These experiments used a variety of differentartefacts which included stand-alone reference plates 12,1619, reference plates combined with reference spheres6, 7, reference cylinders 4 and truncated pyramids 7.There have also been artefacts containing a range of dif-ferent geometries such as holes, slots, fillets and chamfers2, 5. The findings of these studies have uncovered manyextrinsic parameters which can affect the measurementuncertainty of laser scanners which have been summarisedbelow, although recent advances have sought to minimisethese effects.Scan depth Each laser scanning probe has a specified fieldof view, as shown in Fig. 1. This field of view representsthe window in which the laser scanner photo detector canacquire measurement points from the scanned surface 12.When the laser scanner is positioned such that the artefactsurface is located at the beginning of this field of view, theacquired point cloud will have a higher resolution and lowerrandom error compared to if it was located at the end of thefield of view. Studies have also found a systematic error of15 m between scans taken at the beginning of the fieldof view compared to the end when the laser scanner is fur-ther away from the artefact surface 16, 17. When scanningoutside of the field of view, the performance of the laserscanning probe deteriorates significantly 20.Incident angle The in-plane () and out-of-plane () scan-ning angles, shown in Fig. 1, can affect the quality of theacquired point cloud 7. Laser intensity is at its maximumwhen the in-plane and out-of-plane angles are perpendicu-lar to the surface; this can cause point saturation resultingin positioning errors 12. A further benefit of not scan-ning perpendicular to the surface is that random error isdistributed more evenly between the individual x-, y-andz-axes. This is due to measurement noise principally beingaffected by the scan depth which acts uni-directionally 17.However, there is a limit with regard to how far away fromInt J Adv Manuf Technol (2014) 72:16651675 1667Fig. 1 Illustration of laser scanning parameters: Stand-off, field ofview (scan depth and width), in-plane angle (), and out-of-plane angle()perpendicular the laser scanner can be compared to theartefact surface. Once the incident angle goes beyond 60,the intensity of the reflected laser light is too small to bedetected 16.Probe head orientation The laser scanning probe can beideally orientated to the artefact surface by changing theA and B CMM rotational axes. However, using multi-ple probe angles increases the time it takes to measure theartefact as well as the time to qualify a greater amountof probe angles, without improving accuracy 2, 5. Otherinvestigations have also shown that there is a lack of agree-ment between point clouds obtained using different probehead orientations 4, 10. However, multiple probe anglescannot be avoided as they are necessary when scanningsmall objects in order to measure enough of the artefact sur-face to be able to have a well-conditioned fitting algorithm5, 17.Surface properties Both colour and reflectivity of thescanned surface influence the measurement result. Dark sur-faces or colours reflect the minimal intensity of light makingpoint acquisition difficult 2, 21. On the other hand, smoothmetallic materials are not reflective within the range of thelaser light spectrum which again makes point acquisitiondifficult 12, 16. Ideally, diffuse reflection is preferred asit reflects equally in all directions maintaining scan qualitywhen measuring from different angles 16.These findings have led to improved measurement prac-tices and with the laser scanning technology itself, par-ticularly with regard to how it handles different surfaceproperties 5, 17. However, they reveal that measurementuncertainties are dependent upon the intended application.Therefore, a measurement study needs to be designed thatis relevant to the automotive industry.2.2 Study objectivesThere are a number of characteristics which distinguish therequirements for an automotive measurement task from theprevious evaluation studies. Firstly, measurement of a vehi-cle body shell and completed vehicles is generally done witha horizontal dual arm CMM, as shown in Fig. 2. Secondly,vehicle geometry is generally complex with a number ofocclusions. Therefore, different probe angles are required tomeasure surface features, sometimes with a less than idealincident angle. The features that are required to be evalu-ated are discrete surface points, holes and slots rather thanfree form geometry. Finally, the global coordinate frameof the CMM is required to be transformed to the localpart coordinate frame of the vehicle through the measure-ment of defined datum features. All these factors have thepotential to affect measurement uncertainty. The currentmethod for the verification of a CMM is controlled by ISO10360-2 whereby its measurement uncertainty and repeata-bility is determined through the measurement of a calibratedtest length which has direct traceability to the metre unit22. Therefore, the purpose of this study was to evaluatea multi-sensor CMM whose performance had been verifiedto international standards for its suitability for a complexautomotive inspection task. The specific study objectiveswere (1) to evaluate the random and systematic errors ofa horizontal dual arm CMM using a multi-sensor probingsystem and (2) to assess the impact of using a multi-sensorapproach upon the dimensional inspection of automotiveassemblies.3 Materials and methodsTo evaluate the capability of a multi-sensor probing system,a LK H Horizontal Dual Arm CMM was utilised (NikonMetrology, UK)pictured in Fig. 2. The configuration usedis seen regularly in the automotive industry and providesFig. 2 Horizontal dual arm CMM with Land Rover Environmen-tal Cube (E-Cube) CNC machined vehicle body simulator located onmeasurement bed1668 Int J Adv Manuf Technol (2014) 72:16651675measurement access to the exterior, interior and underbodyof the vehicle. Mounted on each of the CMM horizontalmeasuring arms was a PH10MQ indexing probe head (Ren-ishaw, UK) which was used in conjunction with an ACR3probe changing rack (Renishaw, UK) to allow automaticchanging between touch trigger and laser scanning probeswithin the measurement program. For contact measurement,a TP20 5-way kinematic standard force touch trigger probewas used with 140-mm extension and 20-mm long by 2-mmdiameter stylus (Renishaw, UK). For non-contact measure-ment, a Metris XC65Dx (Nikon Metrology, Belgium) laserscanner was used, shown in Fig. 1. This laser scanner emitsthree laser lines in a cross hatch formation each having ascan depth of 65 mm, scan width of 65 mm and a stand-off distance of 75 mm allowing in total 75,000 points persecond to be captured. Prior to the measurement experi-ment, both CMM arms when fitted with a touch triggerprobe were verified in accordance with ISO10360-2 and hadan expanded measurement uncertainty (k = 2) of within1.0 m+1.0m/m. The Metris XC65Dx laser scannerswere verified to the same standard achieving a measurementuncertainty of within 12 m when measuring a ceramicsphere 22. To evaluate the performance of the certifiedmulti-sensor horizontal dual arm CMM for a complex auto-motive measurement task, an aluminium CNC machinedfull-scale physical representation of the manufactured sheetmetal body of a vehicle (Jaguar Land Rover Limited, UK)was selected. This artefact is known as an environmentalcube (E-Cube) and is shown in Fig. 2 23. The E-Cube hasall the interior and exterior surface features of a vehicle bodyto allow fitment of trim components for quality matura-tion purposes to assess their fit, finish and alignment duringproduct development 24, 25. The following sections detailthe artefact experimental set-up, the measurement study andstatistical analysis procedures for this study.3.1 Experimental set-upThe E-Cube had machined flat faces on its base to enableit to locate on the CMM measurement bed without the needof a separate measurement fixture. The E-Cube remainedon the measurement bed for the duration of the study andthe environment was maintained at a standard 20C 1Cthroughout 26. To transform the global coordinate frameof the CMM into the local part coordinate frame of the vehi-cle, four datum plates located at the front and rear lowercorners of the E-Cube were measuredtwo with the LHhorizontal CMM arm and two with the RH CMM arm.Figure 3 provides a schematic of the local coordinate frame.To construct the local x-axis, the mid-point between the twofront datum plates and between the two rear datum plateswere determined and a line constructed between them; theunit vector of this line formed the x-axis according to Eq. 1.Z-AxisY-AxisX-AxisFront Right Datum(FR)Front Left Datum(FL)Rear Right Datum(RR)Rear Left Datum(RL)Primary PlaneFig. 3 Schematic of the E-Cube local part coordinate frame using thefront and rear datum platesThe local z-axis was constructed normal to a plane formedbetween the mid-point of the two front datum plates andboth of the rear datum plates. To do this, the vector cross-product was calculated using the x-axis vector and a vectorformed by the line between the two rear datum plates, (2).The vector cross-product between the x-andz-axes wasthen used to form the y-axis, (3). The x-, y-andz-axesunit vectors were then combined to form the transforma-tion matrix R,(4). This matrix R was calculated using boththe CMM measured values (RCMM) as well as using thenominal values of the E-Cube datum plates (RCube). Anymeasured value in the CMM global coordinate frame couldthen be transformed into the local part coordinate frameusing Eq. 5. Finally, the rotated local coordinate framewas translated so that its origin coincided with the nomi-nal origin of the local part coordinate frame, completing thereference point system alignment (RPS).xaxis=12(FL+ FR) 12(FL+ FR)bardbl12(FL+ FR) 12(FL+ FR)bardbl(1)zaxis= xaxisRL RRbardblRL RRbardbl(2)yaxis= xaxis zaxis(3)R =x1x2x3y1y2y3z1z2z3(4)Rcarline= RTCube.RCMM(5)Int J Adv Manuf Technol (2014) 72:16651675 16693.2 Experimental procedureTo evaluate the systematic error between the two horizontalCMM arms, a ceramic reference sphere (Kolb + Baumann,Germany) of 29.9912 mm diameter was measured by thetouch trigger probe mounted on each of the horizontal CMMarms. This sphere was measured within the CMM globalcoordinate frame. This procedure was repeated for the laserscanning probe later in the measurement program. Follow-ing measurement of the reference sphere, the E-Cube datumplates were m