Educational Articles: Article nr. 3 Revision of CMM performance specifications Premise The working group (WG) Specifications is one of several working groups of ia.cmm, the International Association of Co-ordinate Measuring Machine Manufacturers. This working group is active in the area of CMM specifications. The performance specifications for CMMs are defined by the ISO 10360 series. Based on ISO 10360-1, the Specifications WG has generated a structure for a common data sheet within all associated Companies. The general goals of WG Specifications are:

The development of human skills by education and training. Giving users worldwide the assurance that machine performance is verified by CMM

Manufacturers in compliance with the regulations, methods and equipment stipulated by international standards.

Promotion of and support in the adoption of standards for metrology performance testing. Some of the Members of the Specifications WG also take part on national and international meetings of committees for standardisation of CMM performance specifications. One task of the Specifications WG is to inform the Members of ia.cmm and the Users about the upcoming new revision of ISO 10360 part 2 and of part 5, both dealing with acceptance and reverification tests for CMMs. These new revisions will contain more precise definitions, some new performance specifications and new procedures and new artefacts for application to a larger variety of CMMs. They are scheduled for release in 2007. Furthermore, additional new technical specifications related to CMM technology will be released in 2005. These include ISO/TS 23165 Guide to the evaluation of CMM test uncertainty and ISO/TS 15530-3 Uncertainty assessment using calibrated work pieces as one of several techniques for determining task specific uncertainties of measurements with CMMs. Makoto Abbe (Mitutoyo J) Michele Verdi (Hexagon Metrology SpA I) Josef Wanner (Carl Zeiss Industrielle Metechnik GmbH D) 1 Coming soon: ISO10360-2 CMMs used for measuring linear dimensions 1.1 Reason for revision Technologies applicable to coordinate measurement in industry have evolved rapidly in recent years. The current trend requires better clarified CMM specifications and enhancement of the corresponding ISO standards. The revised version of ISO10360-2, the acceptance test and reverification test for CMM, has been discussed by ISO/TC213/WG10. A brief overview of this upcoming revised standard is provided in the following. 1.2 Enhanced and clarified test procedures The new test procedure more clearly reveals the performance of CMMs by verifying their conformance or non-conformance to the specification. The test scheme of using a conventional E-test to verify MPEE in seven different positions and orientations in the CMM volume on five different sizes has largely been retained, although the

ia.cmm 1/12

conventional E-test is now called E0-test for clarification. On the other hand, in keeping with the concept of modularity, the test for determining the probing error (P-test), which verifies MPEP, has been moved from ISO10360-2 (Part 2) to ISO10360-5 (Part 5). An acceptance test performed according to the upcoming ISO10360-2 requires a tester to refer not only to Part 2 but also to Part 5 in an appropriate manner. The following topics are also dealt with in the new standards. 1.2.1 Thermal properties of test length Major test uncertainty results from thermal property of the test length used, regardless of whether the CMM is thermally compensated or not. Therefore, a CMM stating the specification conforming to the upcoming ISO10360-2 is required to disclose the necessary thermal properties of the test length used for acceptance, i.e. the upper limit of test length CTE (Coefficient of Thermal Expansion) and the upper limit of uncertainty in test length CTE. Furthermore, the default for the calibrated test length is specified as a normal CTE which is a material with a CTE between 8 x 10-6 /K and 13 x 10-6 /K. A specification statement requiring a test length with a non-normal CTE is obliged to indicate MPE*E0 instead. Especially in a case where use of a low CTE test length is stated by the CMM specification, an additional E0-test is performed on a normal CTE test length. 1.2.2 Improving test sensitivity to ram axis roll A new EL-test has been introduced with test sensitivity to ram axis roll proportional to the offset distance between the centreline of the ram axis and the centre of the stylus tip. Two EL-tests are performed with a 150mm offset distance in the default case. The user can choose the position and orientation for the test from eight possibilities along plane diagonals on planes parallel to the ram axis movement. 1.2.3 Repeatability test Repeatability R0 is calculated and tested using the results from an E0-test comprising three repetitive measurements performed on the CMM. Conformance or non-conformance to the specification is verified. 1.3 Newly introduced concept The application possibilities for the coming standard in industry are expected to expand widely. Major possibilities are mentioned below. 1.3.1 Various artefacts that represent a calibrated test length The upcoming standard allows us to adopt a variety of artefacts with a proper test procedure, as far as bidirectional measurement is performed directly or composed of a combination of size measurement on a short gauge block and a series of unidirectional measurements. Various unidirectional measurement possibilities such as laser interferometer, ball bar, ball plate, and so on, are listed in the annex. The typical difficulty of performing the acceptance test on a large CMM, for example, is

ia.cmm 2/12

overcome technically. The test result thus becomes much more practical and conforms fully to the upcoming standard. 1.3.2 Application possibilities for CMMs with particular functionalities ISO10360-2 is intended to apply to Cartesian CMMs with three orthogonally combined straight guideways equipped with contacting probing systems. Non Cartesian type CMM or optical CMM are of increasing importance. The scope of the upcoming standard mentions that the standard may be applied to these kinds of CMM by mutual agreement between the supplier and the CMM user. The test procedure for a dual ram CMM constituted of two independent arms and in a single coordinate system is specified as well. 1.4 Summary The benefit of the test according to ISO10360-2 is that the measured result has a direct traceability to the unit length. It therefore gives information on how the CMM will perform in connection with similar length measurements. The importance of this standard is increasing year by year, since ISO 10360-2 was published in 1994 and revised in 2001. The upcoming standard, ISO 10360-2, will surely be able to meet the expectations of industry regarding higher productivity. 2 ISO 10360-5: CMM Probing Performance with Contacting Probing System 2.1 Reasons for revision The revision of ISO 10360-5 provides some new definitions, clarifications and better modularity. The scope of ISO 10360-5 will be limited to contacting probing systems. Additional new parts of ISO 10360 to be released later will cover the performance tests for further probing systems. ISO 10360-7 for CMMs equipped with video probing systems is already being prepared. 2.2 Enhanced test procedures As a major change in the long term revision, the test of probing error P will be taken out of part 2 and will be integrated in part 5 as single stylus probing system test for modularity reasons. All tests related to a contacting probing system will be in part 5. Part 2 will contain the basic background description and will be focused on size tests only.

ia.cmm 3/12

The concentration of all contacting probing system tests in part 5 will also eliminate double measurements. Measurements taken for the single stylus probing system test can be used for the multiple stylus configuration test as well. The probing system performance is contaminated by the CMM performance. The tests in ISO 10360-5 are sensitive to many errors depending on both the CMM and the probing system, and are to be performed in addition to the length measuring tests given in ISO 10360-2. It is the CMM probing performance which is specified, due to the impracticality of isolating the performance of the probing system from that of the CMM. 2.3 Expanded scope The definition of specific parameters for additional types of contacting probing systems will allow the application of ISO 10360-5 for:

Multiple styli connected to the CMM probe (e.g. a star). Installations using an articulating probing system (motorized or manual) that

can be pre qualified in empirical mode at each utilised angular position or at a few angular positions in inferred mode with interpolation for any angular positions.

Installations using a repeatable probe-changing system. Installations using a repeatable stylus-changing system and multiple-probe

installations. 2.4 Renaming of the symbols For a better overview of the high number of definitions, it was necessary to rename the symbols. The symbols for errors and values are in the form Px_Ty with the indications: P for probing performance x F for a form parameter

L for a location parameter S for a size parameter

T for a contacting (touch) probing system y E for an articulating system using empirical qualification

I for an articulating system using inferred qualification M for a fixed multiple-stylus configuration N for a fixed multiple probe configuration U for a single (unique) stylus configuration

Examples: PF_TI probing system form error based on an articulating tactile probing system using inferred qualification.

ia.cmm 4/12

PF_TU single stylus probing system form error based on a tactile probing system (equal to probing error P as defined in ISO 10360-2:2001).

PS_TU single stylus probing system size error based on a tactile probing

system. PL_TM probing system location value based on a tactile fixed multiple-stylus

configuration. The single stylus probing system size error PS_TU is a new definition. For PS_TU it is not required a specification but it is a signed error. It is needed to get comparable results between bidirectional 2-point-distance measurements and other methods for size measurements according ISO 10360-2. 2.5 Clarified test procedures The new ISO 10360-5 will contain definitions which supersede similar definitions in ISO 10360-1:2001, because some symbols used have been revised and expanded for clarification.

Specification and test according ISO 10360-5 is necessary for one extension length only.

The effective stylus diameters are not identical within a stylus system used for

the test and even different nominal diameters could be specified.

The test sphere (10 - 50 mm) needs to be calibrated for form and for size.

Form errors of styli and test sphere have to be smaller than 20 % of the specified MPE. They must be covered by the specification and can not be subtracted from the test result.

The user is free to choose the location of the test sphere within the measuring

volume if not otherwise specified by the manufacturer. 3 ISO/TS 23165 Guide to evaluation of CMM test uncertainty 3.1 Purpose, motivation and status ISO/TS 23165 aims to define the proper criteria for evaluating the CMM test uncertainty. ISO 14253-1 (1) regulates conformance/non conformance decisions concerning acceptance of a work piece or a measuring equipment. In the case of a work piece, the specification is given by the tolerance: in the case of a measuring device (i.e. a CMM) the specification is given by MPE (Maximum Permissible Error), as prescribed

ia.cmm 5/12

by ISO 10360 standards (2). All ISO 10360 chapters dealing with CMM compliance to specifications prescribe that the CMM performance is to be verified if the error does not exceed the related MPE, taking into account the uncertainty of measurement as specified by ISO 14253-1. However, no guidance was provided to CMM users and CMM manufacturers for establishing which factors are the main uncertainty contributors and the related expanded uncertainty U by which the specification zone must be either reduced or expanded. Now, with ISO/TS 23165, this gap has been eliminated, and ISO 14253-1 can be uniformly applied in the CMM performance verification which is so important for regulating the related contractual aspects. The application field of ISO/TS 23165 is CMM performance verification, and not the general measurements made by a CMM. In this last case, the ISO 15530 series applies. The TS (Technical Specification) status of the document means that the document is officially released and therefore valid even though it does not yet has the full status of an international standard. 3.2 Key concepts The ISO/TS 23165 guide is based on some key concepts: #1 the concept of test uncertainty. #2 the concept of tester and tester counterpart. The test uncertainty is the expanded uncertainty U associated with the testing equipment and its use in the test, i.e. with the test conditions. A quality index can be in fact associated to both test conditions and CMM performance:

test uncertainty U is the quality index associated with the test conditions: o conventional errors like the error of indication, E, or probing error P, are

the quality index of CMM performance. ISO/TS 23165 defines the criteria of interaction between test conditions and CMM performance in the spirit of ISO 14253-1. Test conditions are determined by choices of the tester. When a poor material standard of size is used instead of a better one, this will immediately affect the test uncertainty. The tester is the main player in the test uncertainty scenario, and ISO/TS 23165 consequently identifies in the tester the main responsible for the test uncertainty. When the tester is required to provide relevant test inputs (i.e., the uncertainty of CTE of material standard of size), the related contribution to the test uncertainty budget must be considered. 3.3 Test uncertainty contributors ISO/TS 23165 provides the list of the test uncertainty contributors referring to:

Error of indication on size E. Probing error P.

as stated in ISO 10360-2.

ia.cmm 6/12

Concerning probing error P, the main contributors to test uncertainty are: Form error F as reported in the calibration certificate. Standard uncertainty u(F) as reported in the calibration certificate.

Concerning the error of indication for size, the main contributors are:

Uncertainty due to calibration of the material standard of size, u(cal): this uncertainty contributor quantifies the testers ability to choose a proper material standard of size for the test.

Uncertainty due to the CTE (Coefficient of Thermal Expansion) of the material standard of size u(): this uncertainty contributor quantifies the testers ability to choose a proper material standard of size and proper thermal test conditions. It has to be included in the uncertainty budget only when the CMM expects the tester to input a CTE value. In case of thermally uncompensated CMMs, the related inaccuracies are taken into account in the MPE specification.

Uncertainty due to the input temperature of the material standard of size u(t): this uncertainty contributor quantifies the tester ability to provide good temperature measurements used for compensating the measuring results. It has to be included in the uncertainty budget only if the temperature measurement depends on the testers own equipment. In cases involving temperature sensors embedded inside the CMM and thermally uncompensated CMMs, the related inaccuracies are taken into account in the MPE specification.

Uncertainty due to misalignment of material standard of size: this uncertainty contributor quantifies the testers ability to provide a good alignment of the material standard of size. It can be considered as negligible under specific conditions (i.e. if differences between alignment procedures in the calibration and testing phases, the effects of CMM tunnelling errors and improper sampling strategy are properly minimized)

Uncertainty of fixturing the material standard of size: this uncertainty contributor quantifies the testers ability to minimize deformations in the material standard of size. Statistical analysis or expert judgement can be used in determining it.

(1) ISO 14253-1 Inspection by measurement of work pieces and measuring

equipment Part 1: Decision rules for proving conformance or non-conformance with specifications.

(2) ISO 10360-2 CMMs used for measuring sizes. ISO 10360-3 CMMs with the axis of rotary table as fourth axis. ISO 10360-4 CMMs used in scanning measuring mode. ISO 10360-5 CMMs using multiple stylus probing systems. 4 Techniques for evaluation of the uncertainty of measurement 4.1 ISO 15530-1 Overview and metrological characteristics

ia.cmm 7/12

4.1.1 Purpose, motivation and status ISO 15530-1 aims to introduce the techniques for determining the uncertainty of measurement for a CMM, providing the proper terminology and a reference list of metrological characteristics, i.e. a list of factors that can potentially affect the measurements produced by a CMM. The document is still in a draft version. 4.1.2 Key concepts ISO 15530-1 is based on the following key concepts: #1 Concept of task specific measurement uncertainty. #2 Concept of metrological characteristic. Task specific measurement uncertainty is the expanded uncertainty U including the uncertainty contributors related to the overall process concerning a work piece measured by a CMM. The metrological characteristic of a measuring device is defined as the characteristic of a measuring device which may influence the results of measurements and which can determine an immediate uncertainty contribution. 4.1.3 List of ISO 15530 series standards General title: Geometric Product Specification (GPS) Coordinate Measuring Machines (CMM): techniques for determining the uncertainty of measurement

Part 1: Overview and metrological characteristics. Part 2: Use of multiple strategies in calibration of artefacts. Part 3: Use of calibrated work pieces or standards. Part 4: Use of computer simulation. Part 5: Use of expert judgement, sensitivity analysis and error budgeting.

4.2 ISO 15530-2 Use of multiple strategies in calibration of artefacts 4.2.1 Purpose, motivation and status ISO 15530-2 aims to introduce the technique of multiple measurement strategies, for determining the uncertainty of measurement related to a work piece measured by a CMM. The multiple measurement strategy aims to combine several orientations of the work piece and point distributions on its surface in order to get a better estimation for the conventional true value of the measurand. By applying a simple procedure, work pieces can be calibrated by a CMM. The document is still in a draft version.

ia.cmm 8/12

4.2.2 Key concepts ISO 15530-2 is based on the following key concept: #1 Concept of multiple strategies. Multiple strategies is a technique that, based on measurement of the work piece in different orientations, repeated measurements varying the point distribution on the work piece, and proper measurements of the material standard of size, permits the generation of work piece calibration values and the related uncertainty by means of a CMM. 4.2.3 Procedure The work piece must be measured in at least three (default four) different orientations corresponding to positions capable of guaranteeing good measurement conditions. In each orientation, the work piece must be measured with at least five different point distributions. In case distance/size measurements are required on the work piece, subsidiary measurements on a material standard of size similar in its length to the above distance/size, must be performed along the three CMM coordinate axes and repeated three times each. The calibration value, and related calibration uncertainty are determined by proper calculation based on the database generated by all of the measuring results obtained. 4.3 ISO 15530-3 Use of calibrated work pieces or standards 4.3.1 Purpose, motivation and status ISO 15530-3 aims to provide a technique for a simple uncertainty evaluation of measurements performed by a CMM. This technique applies to specific measuring tasks and to CMM results obtained from both uncorrected and corrected measurements. The standard was published on 2004-03-01. 4.3.2 Key concepts #1 Concept of non substitution measurement. #2 Concept of substitution measurement. #3 Concept of similarity condition. Non substitution measurements are results in which the CMM indication is not corrected by systematic errors: in substitution measurement, the CMM indication is corrected by systematic errors, where both the work piece and a proper material standard of size are measured.

ia.cmm 9/12

Similarity conditions define the constraints binding work piece and material standard of size in the uncertainty assessment (i.e. dimensions and materials) the procedure used for measuring them, the environmental conditions, the CMM stylus configuration. 4.3.3 Procedure In case of a nonsubstitution strategy, calibrated artefacts are measured on a CMM in at least ten measurement cycles, each composed of the handling of each artefact and its consequent measurement. A total of at least 20 measurement repetitions must be achieved in different conditions simulating the real measuring process, including position and orientation of the work piece. This means that, in case only one artefact is involved in the assessment, the above cycle must be repeated 20 times. The uncertainty contributors include the following:

The calibration uncertainty stated in the artefact certificate. The standard uncertainty assessed by the above procedure. The standard uncertainty resulting from the variations of form errors,

roughness, CTE, and other relevant parameters in different corresponding workpieces.

In case of a substitution strategy, the same procedure as stated above applies, i.e. the measuring cycle comprises the handling of the calibrated artefact, its consequent measurement, the handling of the work piece, and its consequent measurement. 4.4 ISO 15530-4 Estimating task-specific measurement uncertainty using simulation 4.4.1 Purpose, motivation and status ISO 15530-4 Estimating task-specific measurement uncertainty using simulation aims to define criteria for simulation techniques applied to task-specific uncertainty evaluations. The target of these techniques is to provide measurement results combined with the related measuring uncertainty. Their application will enable the user to take immediate decisions about the consistency of the measuring process and the conformance or non-conformance of the work piece to the specification as required by ISO 14253-1. ISO 15530-4 is the result of pluriennial research activities focused on simulation techniques applied to coordinate metrology (Virtual CMM). In Virtual CMM (VCMM), the VCMM model is assessed on the basis of a set of input parameters by means of proper CMM measurements of sphere/hole plates and of environmental parameters: using VCMM, it is possible to generate calibrated artefacts, i.e. to associate to the measurement result with the related uncertainty.

ia.cmm 10/12

ISO 15530 is still in the DTR (Draft Technical Report) phase, namely it is an unpublished document in the first phase of the TR status which does not have the full status of an international standard. 4.4.2 Key concepts ISO 15530-4 is based on the following key concepts: #1 Concept of UES (Uncertainty Evaluating Software). #2 Concept of UES model. #3 Concept of UES validation. The UES is a software used to provide uncertainty evaluation by simulating the overall CMM measuring process on a work piece: it can be resident inside the software equipping the CMM, or in another system working in conjunction with it. UES model can be based on mathematical and/or numerical procedures able to handle input quantities (like the temperature of the air) and to determine the related uncertainty contribution. UES can consider only a part of the metrological characteristics and related uncertainty contributions in its evaluating process: in this case, the uncertainty evaluation must be based on a combination of uncertainty evaluation techniques, some of which are derived from pure simulation, and others of which are based on an experimental approach (as described in ISO 15530-3). The UES manufacturer must provide:

A list of CMM metrological characteristics managed by the UES. Proper documentation of the techniques used for the uncertainty

evaluation. UES can be validated by following two different approaches:

Experimental approach consisting of proper testing with a CMM on a calibrated artefact and/or on a proper calibrated material standard of size. Iterative testing, combined with proper analytical tests focused on some specific factors like scale and probing errors, can consolidate the basis of the uncertainty analysis.

Computer-aided techniques (CVE, Computer aided Verification and Evaluation). The process can be based on error vector field analysis, comprising the vector applied to the expected (true) measured point and its direction versus the determined position of the measured one. By proper perturbation of a reference true condition, based on the input of different input quantities it is possible to achieve an uncertainty evaluation related to the metrological characteristic under examination. Uncertainties caused by perturbation of contact points derived from a set of reference input in the UES under examination can thus be compared with ones related to contact points given by a testing body by means of coherently simulated instances. Historical data referred to proper predefined tasks can be considered too.

ia.cmm 11/12

4.5 ISO 15530-5 Use of expert judgement, sensitivity analysis and error budgeting 4.5.1 Purpose, motivation and status ISO 15530-5 aims to define the criteria of the technique for uncertainty evaluation based on expert judgement. The use of expert judgement can be particularly important when type B standard uncertainty is considered in the uncertainty budget. Personnel aiming to provide expert judgements must be properly qualified. ISO 15530-5 is still in draft status. 4.5.2 Key concepts #1 concept of qualified personnel Qualified personnel, is here defined as personnel possessing the academic qualifications, experience and educational background in the field of accreditation and is required to provide an expert judgement. It is expected that any signatory representing a calibration laboratory accredited according to ISO/IEC 17025 by a national accreditation body is able to provide proper judgements.

ia.cmm 12/12