QMT Features: March 2013
Right arm verification?
The NPL offers some guidelines on choosing the right verification method for your articulated arm co-ordinate measurement machine.

Articulated Arm Co-ordinate Measurement Machines (AACMMs) are becoming increasingly popular for measurement and inspection tasks in many industries. An Articulated Arm CMM is checked against specification before being delivered, but how sure can you be that it is performing to that same standard today? Articulated Arms are, by their very nature, portable but easily subjected to knocks during normal usage and transportation that could affect their performance. The only way to be sure is to have your arm regularly verified.

Although these instruments are often thought of as portable CMMs, there is a much less well developed after market support capability for these instruments than for traditional Cartesian or bridge CMMs. This leads to some confusion amongst users about the right way to verify their instruments.

One of the complications facing users is the variety of performance verification standards that are often purported as being relevant to AACMMs, for example:

- ASME B89 4.22 (Methods for Performance Evaluation of Articulated Arm Co-ordinate Measurement Machines) is a US standard method specifically written for the assessment of AACMMs.
- VDI/VDE 2617-9 (Acceptance and Reverification for Articulated Arm Co-ordinate Measurement Machines) a German standard extending the use of the well-known general CMM verification standard (ISO 10360-2) to such instruments.
-  ISO 10360-2 specifies the acceptance tests for verifying the performance of a CMM used for measuring linear dimensions. The tests apply only to contact probe Cartesian CMMs operating in a discrete-point probing mode. However, the standard allows that the method may be used to verify to non-Cartesian CMMs by mutual agreement.

These standards mandate different tests and different approaches to the performance test.  It is important to understand the differences between the various standards in order to make an informed choice of performance verification method.
The popular Faro, Nikon, Cim Core and Romer brands of arms quote specifications according to standard ASME B89.4.22 (Faro and Romer also claim specifications meet VDI/VDE 2617-9). Faro and Romer have ISO 17025 accreditations for in-house verifications of their products based on ASME B89.4.22. Third party suppliers have started to offer verification services, most are based on ISO 10360-2.

Types of Performance Verification tests
The Effective Diameter Test is required as a single test in ASME B89 4.22 and equivalent results are evaluated in the Testing of Probing Error Test in VDI/VDE 2617-9. This evaluates the error in measuring the diameter of a calibrated sphere.
The Single Point Articulation Test (SPAT) required in ASME B89 4.22 establishes the capability of the instrument to measure the same point in space as the arm is articulated through a wide range of movements. Three measurement positions, with respect to radial distance from the centre of the arm) are used – near, middle and far – with the arm base rotated so that the direction of the measurement position relative to the front of the instrument is different for each position.

Testing of Probing Error Test required in VDI/VDE 2617-9 is a reference test to determine the measurement arm repeatability using the ball probe and a calibrated reference sphere. Five points on the surface of the sphere are measured from five different approach directions. The deviation of the point defined from the average sphere centre and the deviation of each point to the average centre are calculated.

Three measurement positions (near, middle and far) are required with the arm base rotated differently for each measurement. This test combines the SPAT and Effective Diameter Test from ASME B89 4.22.

Experience suggests that the single point accuracy test is a vital component of the assessment of an arm. Firstly, failure to meet specification at the mid position is always accompanied by poor performance in other tests and is a good indication of an arm that needs servicing or adjustment. The near and far positions provide fairly stringent tests of an arm’s performance and give good indications of the limits of where the arm can be used with confidence. The closer or further the measurements are made, the greater the likelihood of the arm failing to meet specification.

All of the methods require a length measurement accuracy assessment, known as the Volumetric Accuracy Test (ASME B89 4.22) or Error of Length Measurement Test (VDI/VDE 2617-9 or ISO 10360-2). These tests are designed to provide reasonable expectations for machine performance in practical measuring applications. The tests involve measuring a certified length standard many times in several locations and orientations and compare the resultant measurements to the actual length. However, the requirements of ASME B89 4.22 differ from VDI/VDE 2617-9 or ISO 10360-2.
ASME B89 4.22requires 20 different measurement lines on calibrated length standards (ball bars or similar recommended) covering different combinations of:

•    Lengths – short (25% to 37.5% of arm range) or long (60% to 75 %) of arm range.
•    Direction –horizontal, vertical, ±45? diagonals (tangential movement with the measured points in different octants), radial (with the measurement points in the same octant).
•    Distance between base and artefact – near (< 50% of arm length to the centre of the artefact) or far (> 50%).
•    Rotations of the arm – measurements are required to start and finish in different octants of the sphere volume defined by the reach of the arm.

VDI/VDE 2617-9 or ISO 10360-2 require measurement along 7 different measurement lines (in Cartesian CMMs these are X, Y, Z, XY, XZ, YZ and XYZ but these are not the best lines for AACMMs) using calibrated length standards (length bars or step gauge recommended).

•    Lengths – at least 5 different lengths per line.
•    The longest length must exceed 66% of the arm range (132% of radius).
•    Stitching of results is allowed if a sufficiently long standard is not available but the artefact must be at least 40% of the arm range.

VDI/VDE 2617-9 applies some specific requirements to AACMMs not contained in ISO 10360-2

•    Rotations – 3 different rotations required
o    In the first rotation 1 horizontal line and two oppositely inclined diagonals are measured measured
o    In the second and third rotations 1 horizontal line and one diagonal are measured
o    Different positions of starting rotation of the first rotary encoder are required

Other considerations
AACMMs vary in size considerably and the size of calibration artefact can become an issue in carrying out measurements over a sufficient range of lengths and encompassing a large enough proportion of the measurement volume of the arm to meet the requirements of the standards.

Length bars and step gauges used for ISO 10360 verifications of Cartesian CMMs are relatively common up to ca. 1 m in length. NPL receives very few gauges above ca. 1 m for calibration, 1.5 m gauges being the largest we calibrate. Specially constructed and calibrated artefacts are required to properly assess the largest arms.

A further consideration is whether the artefacts used for verification support the type of measurement that the arm is used to make. Step gauges and length bars provide length measurements between parallel faces. Measurement of ball gauge type artefacts involves probing around the surfaces of spherical targets in order to achieve centre to centre distances. The latter type of measurement can involve more articulations of the arm and is probably closer to the way arms are used in practice.

There are several apparent choices available when choosing a method for verifying the performance of an AACMM. However, these methods vary in rigour and the extent to which the functionality is tested. NPL’s view is that currently ASME B89 4.22 is the most appropriate method for verifying an arm as the method tests a greater range of positions and movement than the other standards. Additionally, ASME B89 4.22 is the basis for the original manufacturer’s certification and only through employing tests based on this standard is it possible to determine whether an arm still meets original specification. If using another standard as the basis of performance re-verification tests it is advisable to understand the exact nature of the tests being performed and make an informed decision whether these tests are appropriate for the way you use your arm. l
You can now view all QMT Magazine issues on your favourite tablet or smart phone.
Download the free Quality Manufacturing Today App from the Apple iTunes App Store or from QMT Magazine on Google Play.

Rob Tremain Photographer
Click above to see full page display and links to QMT articles.
Hexagon Logo 2
Nikon logo
Bowers logo