Requirements governing measurements in product standards

Measurements of physical parameters repeatedly differ in their results. For this reason, a study commissioned by KAN was conducted in order to determine the extent to which the harmonized standards and draft standards pursuant to the Machinery and PPE Directives contain safety-related requirements that are described in terms of measurement variables, and whether suitable measurement methods are also set out where this is necessary.

Where measurements are performed by different laboratories, deviations of 40% to 50% may easily occur, for example during measurement of the parameters of exposure to vibration from hand-held machines in accordance with the EU Machinery Directive (cf. Schenk, Th.; Gillmeister, F. (1999): Ermittlung der Meßunsicherheit bei der Anwendung von Prüfverfahren für die Schwingungsemission von handgehaltenen vibrierenden Maschinen). The reason for the high deviation in this case is not particularly the influence of the measurement apparatus, but rather the selection of the form and location of coupling for the vibration transducer. The operating personnel and their particular ways of working, and the operating forces which they exert, are also major influencing factors, as are the experience and theoretical knowledge of the personnel conducting the measurements. For measured data to be in any way comparable, it is therefore absolutely essential that the standards contain precise descriptions of the measurement methods.

Methodology of the study

At the outset of the study, the data contained in the standards which were to be excluded from the analysis were defined. These include definitions and explanations of terminology, calculated variables, and data describing items of test apparatus. In addition, the study defined which variables are trivial to measure, i.e. for which a description of the measurement methods is superfluous. For example, where variables such as room temperature (in °C) or length (in mm or m) must be measured, the task may be presumed to be straightforward provided an excessively high accuracy of measurement is not specified.

The results in brief

• In total, 941 standards and draft standards from the 54 CEN and CENELEC technical committees (TCs) responsible for machinery and PPE were analysed.
• Measurement of a spatial dimension (clearance, length, width, etc.) was a requirement at almost 2,100 points in the standards studied. The measurement variables of sound-pressure level (sound power level, emitted sound-pressure level), force and time (such as a duration) each occurred as a requirement in almost 1,000 provisions.
• Measurement of spatial dimensions and time variables may however be regarded as trivial in most cases. By contrast, measurement of forces and in particular of sound pressure and vibration require a detailed description of the method to be employed.
• The analysis of the standards revealed that in many cases, the required measurement methods for variables to be measured were not stated, nor were references to relevant standards provided. This is the case, for example, for the variables of load-carrying capacity, wind speed and energy. This deficit is particularly serious for the measured variables of force and velocity (45% and 32% of cases respectively), owing to the frequency with which they must be measured, and for the measured variable of vibration, since the Machinery Directive requires statement not only of the emission value, but also of the uncertainty of measurement for this variable.

The affected standards committees should therefore use the study to review all non-trivial measurement provisions for which neither a suitable measurement method is described, nor reference made to such a method. It would be desirable in future for suitable measurement methods to be added or reference made to such methods, where necessary in such cases.

With the support of the project partner, KAN also intends to draw up recommendations for constructive ways of formulating measurement provisions in product standards. This is to be achieved with consideration for the results from the first phase of the project and based upon existing standards provisions where available, for example those found in the ISO/IEC Directives, the CEN/CENELEC Internal Regulations, or the DIN 820 series of standards. Answers are to be found in particular to the following questions:

1. In what cases must the measurement method be described?
2. What form should this description take?
3. How can proper consideration be given to the uncertainty of measurement?

Stephan Riedel, Frank Gillmeister
info@ergonomieexperten.de