Specification standards for coordinate measuring machines

|° Specification standards for coordinate measuring machines Should there be a British Standard? In the past few issues brief descrip- the CMMA standard had been aimed I would propose, is a good start. It tions hae been carried of the Euro- at the acceptance test. has a reasonable degree of inter- pean (CMMA) 1 and USA (B89.1.12) 2 Short though his presentation national credence and with a lot of specification standards for coordi- was, Dr Jacoby mentioned many of work might be better than starting nate measuring machines (cmms). the arguments that must have con- again'. But it was pointed out, if The British have been biding their tinued for months during the pre- Ferranti themselves are already time, following the discussions, but paration of the standard - - the deft- meeting the ANSI interim Standard now look as though they are on the nitions (all length dependent so that in the USA why should the issue be move. A 'consultative meeting' accuracies are given as an absolute further complicated? held at NPL on 14 March 1984 pro- value plus a length dependent term), duced a consensus that there was a the choice of gauge blocks ('at the User survey need for a British Standard in this time they were the only measure- area and that: ment aids which were traceable to Further manufacturer and customer national standards'), and the mea- opinion was provided by Phil Smith • simplicity was the target surement approach (accuracies valid (National Engineering Laboratory) • the timescale for adoption within the whole volume, measure- who reported the findings of an NEL should be short ments must represent shop floor survey of cmm usage in the UK 3. • the 'software problem' should operation). It was conceded that The survey was undertaken to dis- be considered at all stages there was room for improvement: cover, and define, the potential mar- rotary tables needed to be con- ket for the mechanical space frame sidered, and there was a need for a that had been developed at NEL 4. Introducing the meeting, Dr Peter derating factor for temperature The principal objectives of the sur- Campion, Deputy Director of NPL, dependence. In response to a ques- vey were to establish: stated that engineering metrology tioner from Rank Taylor Hobson, Dr had been rescued from 'the dol- Jacoby also indicated that certain • the number of cmm used by the general body of mechanical drums often years ago' by the corn- questions concerning software-- eg, engineering companies in the bination of three factors: the rapid relating to alignment of gauge blocks UK advances in microelectronics and in in space-- had been passed over • the manufacturers, types and the speed of data acquisition and after a quick evaluation. A German sizes of cram in use processing; the increased sophisti- VDI standard, which agreed closely • the usage from one industrial cation of production methods, and with the CMMA document, had gone sector to another and the usage the demands that had placed on slightly further in considering by individual companies rnetrology; and an increased aware- software problems but this also • the verification practices ness of the importance of quality, omitted details. • the level of interest in a national Industry had created the demand for The more detailed approach of a standard of the type suggested, the American B89.1.12 draft standard standard to introduce a method of verification available to all No one at the meeting - - repre- was described by Professor Pat cram users sentatives of all the major UK cmm McKeown (CUPE). Most of those in manufacturers and several large attendance were familiar with the Findings relevant to verification and users - - disagreed with the opening proposals it contained and some of the need for a Standard are shown remarks; but what were the prob- the manufacturers were already on page 153. The survey also lems to be encountered, and how using it in the USA. Ball bar mea- showed that Ferranti dominates the had other attempts succeeded (or surements were the manufacturers UK market, followed some way failed)? main complaint, in that they were behind by LK Tools. DEA Soci~td The Coordinate Measuring said to be unrepresentative of 'the Genevoise and C.E. Johannson are Machines Manufacturers Associa- real situation' involving deflection of in a group below LK Tools. Motor tion (CMMA) was represented by its the probe tip. Despite this it soon vehicle manufacturing is estimated President, Dr Jacoby (Director for became clear that the prevailing to account for between one-quarter metrology equipment development, mood was one of avoiding the rein- and one-third of all cmms in use. Leitz). The CMMA comprises the vention of the wheel. If two good This is followed by aerospace, main European manufacturers plus documents were already on the pumps, valves and compressors, three Japanese companies. They table, and with the need for prompt metal working machine tools, and had set out with the 'brief' that the action having already been empha- scientific and industrial instru- user should know, and be able to sized, the course to follow had to be rnents - - each with about 10% of all verify, the accuracy of the measure- one of improvement and not of start- crams in the survey. ments made with a cram, compared ing from scratch. Although the majority of cram to the figures given by the cmm A British Standard based on an users rely on the cmm manufac- manufacturer. To this end the user improved CMMA standard was turers to verify the accuracy of their needed definitions and values which suggested by George Macpherson machines, a few of the users were could be used in (1) tests prior to (Ferranti): 'The CMMA standard ~ a dissatisfied with this situation for a purchase, (2) the acceptance test, slim document - - is the distillation variety of reasons, says the report. and (3) verification throughout the of the experience of a significant For instance, the certificate of verifi- lifetime of the machine. In the end number of cram manufacturers and, cation given by a cram manufacturer 152 JULY 1984 VOL 6 NO 3 | ' PItE¢IS is reluctantly accepted by the users' Similarly, other causes of concern accuracy appropriate to the users' customers who prefer to work to among cmm users would also be needs would be simplified. Further- national or equivalent standards. This solved - - a machine's accuracy more, the current practice of user- problem would be solved if a British would be traceable to the national verification could continue with the Standard were available that gave standards of length, and the task of increased confidence of complying advice and guidelines on verification, specifying a new machine with the with a standard method. The comments are diverse and Results of NEE survey of coordinate measuring they include the recognition of the machine usage in the UK complexity of verification; the poten- tial benefits to the user and his cus- Verification Practices three Soci~t~ Genevoise machines, tomers; the fact that any verification Some respondents used both a system should be backed by a stan- Frequency of verification laser and a mechanical means; in dard; the need for traceability to the In the NEL survey 64 respondents one case the laser was apparently national standards; and the realiza- gave information on the length of used to check Zeiss 650 while a step tion that the lack of a standard leaves time between each verification of standard was used for DEA Iota; in a a gap in the system. Some respon- their machines. A few stated that second case a laser was used for an dents deal with large organizations verification was carried out 'on LK Granite and a Talyvel for a DEA (eg MoD, CEGB, CAA, British Ley- demand'; this may mean an in- Beta. Two respondents used a corn- land, Ford Motor Company) that are house check using a standard work- bination of mechanical methods; accustomed to specifying to British piece or jig, and when a discrepancy one relied on length bars and a calm- and Defence Standards; part of this between recorded measurements is brated workpiece, the other used group of users stated that the certifi- noted the user refers to an outside slip gauges and a Talyvel. cation of a cmm by the manufacturer agency for a professional verifica- was not entirely acceptable to their tion. In comparison, another user Verifying organization customers. In the opinion of other users, a standard for verification Respondents were asked who ver- would allow the intercomparison of Interval No. of users ified their cram, whether it was their cram accuracy; this would be par- own company, the supplier, or ticularly useful when specifying a 12 months 37 another organization. Of the 93 new machine. 6 months 16 respondents 89 replied to this ques- Of the 23 respondents who 2-6 months 11 tion; the information is tabulated rejected the idea of a national stan- below. Eleven respondents sup- dard, some commented that the supported a regular three monthly ported the supplier's verification by present levels of accuracy were self-verification with an annual one an in-house verification at inter- by the supplier. The majority of mediate periods. Of the remaining acceptable for their type of work. users, however, had their machines four respondents, two used a cmm verified at regular intervals, by subcontract, and two had their Replies from Royal Ordnance own cmm verified by a sub- Factories contractor. Means of verification During the analysis of the replies In the survey 57 respondents gave Verifying organization No. of users from industry the survey was information on the means used to extended to include the Royal Ordnance Factories. The reply gave verify their cmm. The replies showed User 22 information on the methods used by that a laser interferometer was the Supplier or manufacturer 56 this organization as a group. single most common instrument Supplier and in-house 11 The methods of calibration vary used; however, nearly twice as many Other 4 greatly and in view of the numbers respondents used specified or of machines and the differing unspecified 'mechanical' means methods of calibration used for iden- than those that used lasers. Two Introduction of standards tical machines these have not been respondents refer to verification by tabulated but are the Johannson method; one of them The users of cram were asked for has a Johannson Cordimet but sur- their opinion on the need for a British basically: prisingly the other has a Ferranti and Standard that would introduce a • By contract with manufacturer common practice for verifying the (at six-monthly intervals) interim Means of verification No. of users accuracy of these machines. Of the check (at six-monthly intervals) 91 cram users who replied to this by factory quality staff using Laser 20 question 66 felt that there was a Sogenique machine calibrator need and 23 said that they had no • Calibrated blocks, straight edges Step standard 2 such need. Of the 66 users who and squares with and without Length bar 4 expressed a need for a standard lasers at 6 or 12-monthly Slip gauge 4 method, 40 added a comment based intervals Calibrated workpiece 8 on the experience that justified their • Ball frame with lasers and Unspecified 20 opinion. Two non-commital replies oscilloscope at 12-monthly were received, intervals PRECISION ENGINEERING 153 | ' Most respondents agreed that a produced very good hardware. Still, already prepared by other groups it standard verification method was however, investment in hardware is hoped that a draft standard will be required, said Smith, but opinions far outweighed investment in available by September 1985. differed within the same organiza- software development and proving, tion. Even within the same factory and this position had to be reversed. References separate methods are used to verify 1 CMMA. Accuracy specification for different machines. Accepting that coordinate measuring machines. different standards of accu racy a re After effects 1982, Coordinate Measuring required by different parts of indus- Machines Manufacturers Associa- try, the present diversity of methods Following the NPL meeting its tion, c/o BCM, 27a Old Gloucester promotes a situation in which agree- organizer, Dr K.F. Poulter 6, has held Street, London WC1N3XX, UK. See ment of accuracy between crams is further discussions with senior BSI also: Precision Engineering, April unlikely simply because there are no staff and formally requested the 1983, 5 (2),86-87 standard guidelines for verification. Institute to start new work in this 2 ASME B89.1.12 Working Group. field. BSI have agreed to form a new Method for performance evaluation Software committee in the MEE series entitled of coordinate measuring machines. March 1983 See also: Precision It was no great surprise to most pre- 'Computer-Aided Dimensional Engineering, July 1983, 5 (3), 125- sent at the meeting that algorithms Engineering Metrology'. This corn- 128 and software were the subject of mittee will be charged with produc- 3 Shelley M., Cook M., Smith P. and much discussion. Both users and ing three standards: Harvie A. A survey of coordinate measuring machine usage in the UK. manufacturers had their say; the (1) Methods for verifying the Publication EAU M39, March 1984, former calling for algorithms to meet mechanical performance of a Economic Assessment Unit, National specifications in the same way as coordinate measuring machine. Engineering Laboratory, East Kil- accuracy and for access to funda- (2) Formulating definitions of bride, Glasgow, UK mental definitions, the latter main- geometrical form suitable for 4 Harvie A. and Beattie J. S. NEL's taining their right to protect their use with computer aided coordinate measuring machine 'stock in trade', equipment, calibration service for industrial The question of definitions used (3) Methods for verifying the application. Production Engineer, June 1982 in the algorithm base was raised by software used with computer 5 Cox M. G. and Ja©kson K. Algo- several people. A first stage, said aided equipment, rithms and software for engineering Professor McKeown, was to investi- gate the algorithm base - - how do Trade Associations and Professional metrology: A statement of need. we define the displacement between Institutions are being asked to nomi- NPL Report MOM 65, June 1983 two holes, or planes, or flatness? Dr nate people to serve on the commit- 6 Dr.Metrology,K. F. Poulter,DivisionHeadof-MechanicalEngineering Maurice Cox, a member of the NPL tee. It is intended that the committee and Optical Metrology, National team investigating algorithms and will tackle tasks (1) and (2) in parallel Physical Laboratory, Queens Road, software in metrology s, suggested and that (3) will follow in due course. Teddington, Middlesex TWl I OLW, that considerable advantages would By making use of the documents UK accrue from devising, implementing and using good algorithms within the general principles and proce- dures of a 'complete measuring system'. It was necessary to develop mathematical conceptual methods, and in the first instance work would of necessity be confined to geometri- cal quantities - - straightness, flat- ness, cylindricity, conicity. George Macpherson agreed that the underlying mechanical definitions were not sufficiently Fixture makers often have to use levers to gain access to bores, which can specific and that this had led to prob- be expensive, difficult to manufacture and prone to error. The Mercer lems both of comparability and of miniature lever probe, however, can be implanted into ground slots in the bore mandrel or fixture to allow direct measurement of bores. Other appli- reliability of results. Several users indicated that this had now led them cations, such as the gauging of lengths and the measurement of narrow to write their own specifications and /ands, can also make use of the miniature probe. The probe consists of MoD QAD at Woolwich were even three main mechanical parts - - a steel body (about25 mm long), a prob- using their specifications at the ing head with a 2.00 mm carbide ball inserted, and a leaf spring which acceptance stage. Sixteen years connects the head to the body. Inside the probe is an armature and an experience have been gained at inductance coil which produces a gauging signal for use with standard or Woolwich and for the last six there microprocessor based electronic units. Tip force of the probe is 150 to has been general satisfaction with the 220 g at electrical zero and repeatability is to +_ 0.0001 ram. The measuring hardware. In the last two years atten- range is +_0.250 mm and the total travel of the tip is 0.650 mm with 0.050 tion has shifted to software and the mm pre-travel. Maximum linearity error is +_ 0.2% of range. instability of algorithms. Dr Cox Thomas Mercer Limited, Eywood Road, St A/bans, Hertfordshire agreed that the manufacturers now AL 1 2ND, UK 154 JULY 1984VOL 6 NO 3