Verification of large CMMs - Artefacts and Methods - Eugen Trapet, Spain (Trapet Precision and ISM3D) eugen@trapet.de, www.trapet.de
Contents - Quick review of ISO 10360-2 - Types of artefacts specifically useful for testing large CMMs - Methods to extend the measurement range of artefacts - Verifying corrections and maintaining corrections valid - Conclusions
ISO 10360-2 - general rules Measure 5 lengths in each of the specified lines of measurement Measure these 5 lengths in each line 3 times The maximum artefact length measured in each line must be >66% of the largest measurable length in the respective line All lengths must be measured in a bi-directional way *) The errors of all measured lengths must be smaller than the respective MPE The repeatability must be smaller than the specified limit *) if artefacts which allow bidirectional probing are not available, it is allowed to measure the errors due to bidirectional probing on separate artefacts and add them to the length measurement error (examples of non bidirectional standards of length: laser interferometers, ball beams)
ISO 10360-2 procedure in short Measure 3 axis-parallel lines with zero-ram offset styli Measure 4 corner-to-corner space diagonals with approximately zero ram axis offset styli Measure 2 edge-to-edge plane diagonals in a plane parallel to the ram axis with ram axis offset styli (per default 150 mm offset) On (large) CMMs where the longest axis is much longer than the other axes, it is recommended to measure additionally 2 cornerto-corner plane diagonals in a plane parallel to the ram axis and orthogonal to the longest axis, using styli which have no or little ram axis offset
Rules to deal with the uncertainty of the test Use decision rules as stated in ISO 14253-1 Take into account artefact calibration uncertainty and drift Take into account temperature influences as far as they are not attributed to the machine performance *) Take into account alignment uncertainty and overlapping / joining uncertainty where needed *) In case of a not-temperature-compensating CMM and when using steel artefacts no such influences need to be accounted for; on a CMM with object temperature compensation, normally the uncertainty of the CTE must be accounted for; on a CMM with no temperature compensation and when using low-cte-artefacts account for: - thermometer uncertainty - temperature assessment uncertainty in situ - uncertainty of the CTE of the artefact
Suited artefacts for large CMMs Gage blocks (normally up to 1 m length) Step gages (normally up to 2 m length) Ball beams and ball bars (normally up to 3 m, in some cases up to 5 m) *) Disassemblable ball bars (up to 12 m) *) Laser interferometer *) If CTE of artefacts is < 0.000002 /K, then measure a 0.5 m gage block additionally (errors must be within specifications). On non-temperature-compensated CMMs: use Virtual Steel concept *) additionally to be taken into account: bi-directional probing error, determined by the measurement of a sphere diameter or a gage block length in direction of the respective line of measurement
Step gages for the verification of CMMs according to ISO 10360-2 (2009) Courtesy: KOBA
Ball beams for the verification of CMMs according to ISO 10360-2 (2009) Here shown: ESCALON 2 m with precision joint on high accuracy CMM with 3 m maximum axis travel Courtesy: BIMAQ
3 different ways to large artefacts 1. Segmented ball beam: Calibrated bars between spheres; the sum of the length measurement errors of the segments is very similar to length error of direct/total length: the difference is only the cosine-error of the measured machine error, and not of the entire measured length! The joining uncertainty per segment is <0.7 m. The maximum length is >12 m, more than reached with any other artefact. The individual elements can be calibrated on smaller reference measuring systems. error length 2. Overlapping by displacement of beam or by placing 2 or more beams in series without precisely joining them and by adding the length measuring error at the end of the prior position to the length measuring error at the beginning of the following position. 3. Overlapping by precision joints between longer segments of several spheres Each segment is calibrated individually. Segments can be used individually. Typical segment lengths are 1.5 m to 2 m.
Ball beams for the verification of CMMs according to ISO 10360-2 (2009) Here shown: ESCALON ball beam 2.7m on CMM with 10 m x 2 m x 1 m axis travel, 4 times stitching in X-axis by simple shifting Courtesy: LOMG
Ball beams for the verification of CMMs according to ISO 10360-2 (2009) Here shown: ESCALON ball beam 2.7m on CMM with 10 m x 2 m x 1 m axis travel, Diagonal placement and measurement with offset stylus Courtesy: LOMG
Courtesy: TRIMEK Large solid carbon fibre ball beam ( 3D-ball beam ), up to 5 m
Large multi-sensor ball beam
Ball beams for the verification of CMMs according to ISO 10360-2 (2009) 3 x 2 m displacing and stitching of 1 bar on a 5 m holder = 6 m 2 x 1.4 m with precision joint = 2.9 m
Disassemblable ball beams for the verification of large CMMs Here shown: Disassemblable ball bar of 5 m length Courtesy: HEXAGON
Disassemblable ball beams for the verification of large CMMs Courtesy: ZEISS, Daimler Benz
Disassemblable ball beams for the verification of large CMMs
Disassemblable ball beams for the verification of large CMMs Here shown: disassemblable ball bar of 8 m length Courtesy: Metronom
Laser ball bar for the verification of CMMs according to ISO 10360-2 (2009) Courtesy: Caterpillar
Renishaw plc 2002-2004 Not to be reproduced without written permission from Renishaw Laser diagonal tests - laser set-up Laser & Calibration Products Division 22-Nov-2004 Slide 20 ML10 laser, and linear optics aligned via a swivel mirror to a machine body diagonal.
The Laser Tracer Laser Tracer hardware on the fly measurement Courtesy: ETALON
Special features of the Laser Tracer concept Machine performance verification according to ISO 10360-2 and ISO 230-2 ff: axial and space-diagonal displacement measurements Courtesy: ETALON
Basic test measurement lines according to ISO 10360-2
Stylus-offset length measurement test lines according to ISO 10360-2
Specific case Long CMM according to ISO 10360-2
Stylus-offset length measurement test lines according to ISO 10360-2 in case of HAMs
Duplex specific error Twin-Coupling is checked by ISO 10360-2 procedure Z 2 Z 1 X 2 X 1 Y 2 Y 1 Duplex Alignment translation and rotation; desirable: scale factor, pitch, yaw and roll harmonization
Specific errors of large HAMs, cross-table machines and gantries where the coverage by the ISO standards is less good Y Y Z Y Z X Z Y Z Z X Y Y X Y X Z Z
Specific choices of the test lengths for HAMs operated in duplex mode
Overlapping or stitching concept to test CMMs larger than the standard
Overlapping or stitching concept to test CMMs larger than the standard, specific considerations for HAMs measure these spheres with arm No 1; measure overlap of positions 1-2 completely with arm No 1 (both ball beam positions) L5 L4 L3 L2 masure these spheres with arm No 2 in CSY created with arm No 1; measure overlap of pos 2-3 completely with arm No 2 (both ball beam positions) L1 measure these spheres with arm No 1; measure overlap of positions 1-2 completely with arm No 1 (both ball beam positions) stitching 2 bar positions L5 without stitching L5 L4 stitching 3 bar positions (... and so on...) L3 L2 L1 masure these spheres with arm No 2 in CSY created with arm No 1; measure overlap of pos 2-3 completely with arm No 2 (both ball beam positions) due to the fact that the common measuring volume of both arms is in the centre of the machine and is very limited, the short lengths will always be in the centre of the machine too L4 L3 L2 L1 ISO 10360-2 prescribes for HAMs that eac h test length consists of one side of the standard being measured by the first arm and the other side by the second arm
Requirement of standards with bi-directional length representation ( standards of size ) Probing Direction Probing Direction Position 1 Position 2
Rules for converting one-directional or centre-to-centre artefact measurements into bi-directional measurements Measure the 5 lengths in in a given line of measurement in a uni-directional or centre-tocentre way, each length 3 times Measure the length of a gage block or equivalent standard of size 3 times in direction of the measurement line concerned Add length measurement errors obtained on gage block to length measurement errors obtained for the 3 repetitions of each of the 5 lengths
Measured Error [µm] Example: Results from verification of large high accuracy CMM with ESCALON ball beam (without stitching) 8,0 6,0 4,0 2,0 0,0-2,0-4,0-6,0-8,0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Length [mm]
Measured Error [µm] Example: Results from verification of very large CMM with ESCALON ball beam with 4 position stitching Metrological Properties "Length Measurement Errors" in defined Lines of Measurement" and "Probing Errors" Page 4 of (Verification according to ISO 10360-2) 10 0 01.10.07 01/00/00 Measurement line 1 (parallel to the X-axis) The following deviations from the calibrated lengths were measured: Calibrated Lengths [mm] 149,909 2699,589 5249,251 7798,913 10348,576 Measured Lengths Uncert. max.perm. deviations measured Meas 1 Meas 2 Meas 3 (k=2) MPE E E [mm] [mm] [mm] [µm] [µm] [µm] [µm] [µm] 149,912 149,914 149,915 2,34 11,05 3,14 5,14 6,14 2699,614 2699,617 2699,618 4,74 28,90 25,68 26,68 28,68 5249,275 5249,273 5249,272 7,26 46,74 23,83 21,83 20,83 7798,913 7798,913 7798,913 9,11 64,59-0,35-0,35-0,35 10348,580 10348,580 10348,580 10,64 82,44 4,46 4,46 4,46 100,0 80,0 60,0 40,0 20,0 0,0-20,0-40,0-60,0-80,0-100,0 0 2000 4000 6000 8000 10000 12000 Length (mm)
Measured Error [µm]. Example: Results from verification of very large CMM with disassemblable ball bar Metrological Properties "Length Measurement Errors" Page 0 in defined Lines of Measurement" and "Probing Errors" 4 of (Verification according to ISO 10360-2) 10 01.10.07 01/00/00 Measurement line 2 (parallel to the Y-axis) The following deviations from the calibrated lengths were measured: Calibrated Lengths [mm] 499,5245 998,9792 1997,7631 3496,5564 4995,2820 Measured Lengths Uncert. max.perm. deviations measured Meas 1 Meas 2 Meas 3 (k=2) MPE E E [mm] [mm] [mm] [µm] [µm] [µm] [µm] [µm] 499,5028 499,5028 499,5028 1,00 44,99-21,70-21,70-21,70 998,9676 998,9676 998,9676 1,41 54,98-11,63-11,63-11,63 1997,7789 1997,7789 1997,7789 2,00 74,96 15,84 15,84 15,84 3496,5888 3496,5888 3496,5888 2,65 100,00 32,38 32,38 32,38 4995,3157 4995,3157 4995,3157 3,16 100,00 33,68 33,68 33,68 150,0 100,0 50,0 0,0-50,0-100,0-150,0 0 1000 2000 3000 4000 5000 6000 Length [mm]
Laser tracker verification with overlapping ball beam
Calibration certificat of disassemblable ball bar
Interim checking large CMMs Periodic checking large CMM in ISM3D Courtesy photograph on right: UNIMETRIK
Interim checking large CMMs, Here: disassemblable tetrahedron 2m Courtesy: METRONOM/AIMESS
Courtesy: BalTec Interim checking of large CMMs, Here: Virtual Tetrahedron
Conclusions The ISO 10360-2 (2009) provides rules for the verification of large CMMs, including duplex machines. One important rule is that the artefact (reference system) must have at least 66% of the length of the maximum machine travel. Artefacts suited for the standardized procedures are commercially available as well are accredited laboratories to calibrate them. One may choose between different concepts of reaching the required length: - stitching smaller artefacts mathematically to one bigger artefact (with and without auxiliary devices to place the artefact in the subsequent positions) - using precision-connections to assemble several artefacts to create one bigger artefract - using segmented artefacts where each individual length is separately calibrated and joined on a stable holder