SIST EN 15433-3:2008

2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.Transportation loads - Measurement and analysis of dynamic-mechanical loads - Part 3: Data validity check and data editing for evaluationNLKRYUHGQRWHQMHCharges de transport - Mesurage et analyse des charges mécaniques dynamiques - Partie 3 : Contrôle de validité des données et édition des données pour évaluationTransportbelastungen - Messen und Auswerten von mechanisch-dynamischen Belastungen - Teil 3: Datengültigkeitsüberprüfung und Datenaufbereitung für die AuswertungTa slovenski standard je istoveten z:EN 15433-3:2007SIST EN 15433-3:2008en,de55.180.01ICS:SLOVENSKI
STANDARDSIST EN 15433-3:200801-februar-2008







EUROPEAN STANDARDNORME EUROPÉENNEEUROPÄISCHE NORMEN 15433-3December 2007ICS 55.180.01 English VersionTransportation loads - Measurement and evaluation of dynamicmechanical loads - Part 3: Data validity check and data editingfor evaluationCharges de transport - Mesurage et analyse des chargesmécaniques dynamiques - Partie 3 : Contrôle de validitédes données et édition des données pour évaluationTransportbelastungen - Messen und Auswerten vonmechanisch-dynamischen Belastungen - Teil 3:Datengültigkeitsüberprüfung und Datenaufbereitung für dieAuswertungThis European Standard was approved by CEN on 28 October 2007.CEN members are bound to comply with the CEN/CENELEC Internal Regulations which stipulate the conditions for giving this EuropeanStandard the status of a national standard without any alteration. Up-to-date lists and bibliographical references concerning such nationalstandards may be obtained on application to the CEN Management Centre or to any CEN member.This European Standard exists in three official versions (English, French, German). A version in any other language made by translationunder the responsibility of a CEN member into its own language and notified to the CEN Management Centre has the same status as theofficial versions.CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland,France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal,Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland and United Kingdom.EUROPEAN COMMITTEE FOR STANDARDIZATIONCOMITÉ EUROPÉEN DE NORMALISATIONEUROPÄISCHES KOMITEE FÜR NORMUNGManagement Centre: rue de Stassart, 36
B-1050 Brussels© 2007 CENAll rights of exploitation in any form and by any means reservedworldwide for CEN national Members.Ref. No. EN 15433-3:2007: E



EN 15433-3:2007 (E) 2 Contents Page Foreword.3 Introduction.4 1 Scope.5 2 Normative references.6 3 Identification of physical events.6 3.1 General.6 3.2 Identification of periodic signal components.7 3.3 Identification of time-varying signals.7 4 Visual inspection of analogue time histories.10 4.1 Signal clipping.10 4.2 Excessive instrumentation noise.11 4.3 Intermittent noise.13 4.4 Power line pickup.14 4.5 Spurious trends.15 4.6 Signal dropouts.17 5 Visual inspection of digital time histories.18 5.1 General.18 5.2 Signal clipping.18 5.3 Excessive digital noise.19 5.4 Wild points.19 5.5 Spurious trends.20 6 Visual inspection of analysed data.20 6.1 General.20 6.2 Probability density plots.20 6.2.1 General.20 6.2.2 Signal clipping.21 6.2.3 Intermittent noise.21 6.2.4 Wild points.22 6.2.5 Power line pickup.22 6.2.6 Signal dropouts.23 6.3 Narrow band spectral analysis.24 6.3.1 General.24 6.3.2 Excessive instrumentation noise.25 6.3.3 Power line pickup.25 7 Corrective editing of time histories.26 7.1 General.26 7.2 Corrections of excessive instrumentation noise.27 7.3 Removal of intermittent noise spikes and wild points.27 7.4 Removal of spurious trends.28 7.5 Removal of temporary signal dropouts.28 8 Identification of periodic components.30 8.1 General.30 8.2 Narrow-band spectral analysis.30 8.3 Band-limited probability density analysis.31 9 Identification of stationary and non-stationary trends.31 Bibliography.33



EN 15433-3:2007 (E) 3 Foreword This document (EN 15433-3:2007) has been prepared by Technical Committee CEN/TC 261 “Packaging”, the secretariat of which is held by AFNOR. This European Standard shall be given the status of a national standard, either by publication of an identical text or by endorsement, at the latest by June 2008, and conflicting national standards shall be withdrawn at the latest by June 2008. Attention is drawn to the possibility that some of the elements of this document may be the subject of patent rights. CEN shall not be held responsible for identifying any or all such patent rights. According to the CEN/CENELEC Internal Regulations, the national standards organizations of the following countries are bound to implement this European Standard: Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland and the United Kingdom.



EN 15433-3:2007 (E) 4 Introduction This standard was originally prepared by working group NAVp-1.4, Requirements and Testing, of the German Standardization Institute (DIN).
It is part of a complete normative concept to acquire and describe the loads acting on goods and influencing them during transport, handling and storage. This standard becomes significant when related to the realisation of the European Directive on Packaging and Packaging Waste (Directive 94/62 EC, 20 December 1994). This directive specifies requirements on the avoidance or reduction of packaging waste, and requires that the amount of packaging material is adjusted to the expected transportation load, in order to protect the transportation item adequately. However, this presumes some knowledge of the transportation loads occurring during shipment. At present, basic standards, based on scientifically confirmed values, which can adequately describe and characterize the magnitudes of transportation loads, especially in the domain of dynamic mechanical loads do not exist nationally or internationally.
Reasons for this are mainly the absence of published data and insufficient description of the measurements or restrictions on the dissemination of this information. This standard will enable measurement and evaluation of dynamic mechanical transportation loads, thus enabling the achievement of standardized and adequately documented load values. This series of standards consists of the following parts:  Part 1: General requirements  Part 2: Data acquisition and general requirements for measuring equipment  Part 3: Data validity check and data editing for evaluation  Part 4: Data evaluation  Part 5: Derivation of test specifications  Part 6: Automatic recording systems for measuring randomly occurring shock during monitoring of transports.



EN 15433-3:2007 (E) 5 1 Scope This standard defines procedures for assessing the validity of results acquired in accordance with EN 15433-2, and for evaluating these results.
NOTE When measuring and analysing dynamic processes, quite often unnoticed or difficult to recognize disturbances or erroneous measurements occur, which impair the application of these values. These procedures are necessary in order to detect possible errors before any actual analysis occurs. Figure 1 provides an overview of the data validation and editing processes in this standard.



EN 15433-3:2007 (E) 6
Figure 1 — Outline of data validation and editing procedures 2 Normative references Not applicable. 3 Identification of physical events 3.1 General A measured time signal shall be associated with the physical events that happen during a measurement.
If the data are produced by a printer or plotter or with an analogue recorder, then the frequency response of these devices shall be equal to or greater than the frequency range of interest in the data.



EN 15433-3:2007 (E) 7 NOTE The first step in data validation and editing is identifying each signal at all relevant physical events associated with the measurement.
Identification should preferably be achieved by inspecting the analogue or digital signals visually, either on paper copies or on the monitor. It is assumed that the measured signal is of a periodic, random or transient nature [see Figure 2 a) and b)].
In practice, these signals are most commonly of a combined nature [see Figure 2c)]. a) YX b) YX c) YX Key X-axis Time t Y-axis Instantaneous value x(t) Figure 2 — Periodic (a), random (b), and mixed signals (c) 3.2 Identification of periodic signal components Periodic components in measured signals shall be identified, e.g. by visual inspection of paper recordings, in order to treat them correctly during the analysis. 3.3 Identification of time-varying signals Transient or non-stationary physical events shall be identified by measured time signals (see Figure 3), in order to separate them at a later time, and to perform a separate analysis. NOTE 1 Transient signals are broadly defined as those that have a definite beginning and end [see Figure 3a)].



EN 15433-3:2007 (E) 8 a) YX b) YX Key X-axis Time t Y-axis Instantaneous value x(t) Figure 3 — Transient (a) and non-stationary random signal (b) NOTE 2 Non-stationary occurrences are due to long-lasting events with continuous varying characteristics. Figure 4 shows the main transients and superimposed occurrences during a road transport. The identification of transients and non-stationary events is not only needed to assist the data validation, but is essential also for the selection of appropriate analysis procedures. Based upon physical considerations, situations may arise where a measured time history reveals an apparent non-stationary trend, which is not anticipated. This trend can be wrong. On the other hand, it might be indicative of an unexpected time-varying property of the measured phenomenon, in which case the presence of a trend could have important physical implications.



EN 15433-3:2007 (E) 9
YX0,02,04,06,08,010,012,014,016,018,020,022,024,026,0-15,0-10,0-5,00,05,010,015,01234 Key 1 Deceleration 2 Branching off; changing road surface condition 3 Acceleration 4 Pothole Figure 4 — Identification of physical events in a measured signal



EN 15433-3:2007 (E) 10 4 Visual inspection of analogue time histories 4.1 Signal clipping Measured time signals of a periodic, random or transient nature shall be checked for signal clipping (see Figure 5). If signal clipping is detected during data acquisition, then the recorded data are useless.
No attempt shall be made to introduce non-linear corrections to signals that have been clipped. NOTE 1 One of the most common errors in data acquisition is too high a setting of the sensitivity of any one of the data acquisition instruments. The result is signal limitation or signal clipping. An insufficient (high) sensitivity setting can also result in signal limitation, because the signal disappears within the noise. Such problems are easily corrected, but the corrections shall be performed immediately after the first measurements, and checks shall be repeated. Contrary to the two-sided clipping shown in Figure 5, a clipping can appear one-sided as well. Low-pass filtering of clipped signals obscures the results shown in Figure 5. After a filtering operation, it is difficult to detect a limited signal. Signal saturation in certain instruments of the measuring chain may also produce more complicated results than the ideal amplitude limiting shown in Figure 5, and shall therefore not be used. Specifically, there may be a zero shift in the signal level followed by a slow recovery, which appears as a time-varying trend in the mean value of the signal. The probability density analysis of a signal (in particular, a stationary random signal), provides a powerful tool to detect clipping. As Figure 5c) shows, the detection of signal clipping by visual inspection is most difficult for a transient signal, particularly if it is a single pulse transient. To assist the detection of possible clipping in transient signals, it is recommended that the peak output voltage of each instrument within the measuring chain be determined and compared to the peak voltage represented by the measured transient. If the peak voltage of the signal is equal to or greater than 95 % of the peak voltage of the instrumentation, this suggests that clipping might have occurred.



EN 15433-3:2007 (E) 11
a) YX b) YX c) YX Key X-axis Time t Y-axis Instantaneous value x(t) Figure 5 — Clipped periodic (a), random (b), and transient signals (c) In the case of random signals, it is recommended that a measurement be rejected if the clipping occurs within ± 1,5 standard deviations of the mean value of the clipped signal. If a measurement cannot be repeated, then extrapolation of singular events, e.g. drop test of a container may be performed, should the physical causes leading to the exceeding of the measuring range and the boundary conditions be known, thereby permitting a reconstruction of the event. NOTE 2 For periodic and transient signals, clipping may dramatically reduce the indicated peak amplitude of the signal. Clipping also erroneously increases the high frequency content of the signal. 4.2 Excessive instrumentation noise Excessive noise introduced by external loads shall be detected and corrected during the system calibration. To detect excessive instrumentation noise, the output signal from the data acquisition system shall be measured prior to and after the dynamic activity of interest. Acquiring the output signal shall be performed using the same sensitivity setting as for the measurements. The influence of excessive instrumentation noise on the measured time histories of periodic, random and transient data signals is shown in Figure 6.



EN 15433-3:2007 (E) 12 Using telemetering systems with voltage-controlled oscillators and multiplexers, it is recommended that the noise floor of the system be measured and analysed to identify potential noise problems prior to the data acquisition. NOTE 1 Too low a sensitivity leads to an inadequate signal-to-noise ratio. Excluded are those cases where the system is vulnerable to noise induced by external loads (triboelectric noise); here instrumentation noise can be assumed to be an additive and statistically independent of the signal.
Instrumentation noise is usually obvious in transient signal measurements as well, since the noise is present before and after the transient event.
Instrumentation noise in stationary random signals is usually not obvious through visual inspection of the time history (see Figure 6b), because the noise itself is generally a stationary random signal. a) YX b) YX c) YX Key X-axis Time t Y-axis Instantaneous value x(t) Figure 6 — Excessive instrumentation noise in periodic (a), random (b) and transient signals (c) The synchronous averaging procedure shall be applied to extract a periodic signal from excessive instrumentation noise. Only the calculation of the spectrum of a stationary signal with instrumentation noise can decide whether that measurement has to be rejected. To extract a transient out of excessive instrumentation noise, it is necessary that in addition to a shock response spectrum computed from the transient, a shock response spectrum of the signal prior to or after the transient is computed. The signal duration required to compute the shock response spectrum of the random signal shall be the same as that used to compute the shock response spectrum of the transient.



EN 15433-3:2007 (E) 13 NOTE 2 In the case of periodic signals, if only one spectrum is required from the data analysis, the narrow band filtering inherent in the computation of a spectrum can suppress the influence of excessive instrumentation noise dramatically.
Excessive instrumentation noise is not a reason to reject periodic signal measurements for spectral analysis purposes if the procedure of synchronous averaging is used to gain signals from noisy measurements. 4.3 Intermittent noise Intermittent noise shall be detected and corrected during the calibration of the data acquisition system.
To identify noise spikes, the following procedures shall be applied. a) In those cases where multiple-channel measurements are made, and the source of noise spikes may be common to all channels, directly compare the time histories of all simultaneous measurements to determine if suspected noise spikes appear at exactly the same time in all measurements.
The possibility that an actual physical event might have produced an extreme value in all the measurements shall also be considered. b) If the upper frequency limit of the data acquisition system substantially exceeds the upper frequency limit of the dynamic signals being measured, noise spikes will generally have a much shorter duration than actual data.
Specifically, the duration of noise spikes is characteristically of the same order as the reciprocal of the upper cut-off frequency of the data acquisition system. NOTE 1 Noise spikes may appear in FM telemetered data due to "click" (or FM) noise caused by too low a signal level at the demodulator.
"Click" noise generally will not be detected during the calibration of the data acquisition system. In the case of periodic data, the presence of intermittent noise spikes is usually obvious from a visual inspection of the time history [see Figure 7a)]. For random and transient signals, the problem may become more obscure [see Figure 7b) and 7c)].
The detection of noise spikes in stationary random signals can be augmented by a probability density analysis. It is recommended that all measurements with identified noise spikes in their analogue time histories be rejected, unless the noise spikes can be removed by corrective editing, detailed in 7.3. NOTE 2 Intermittent noise spikes appear in a spectral analysis as additive broadband noise, which can severely distort the spectral values of the actual signal. Because noise spikes in time signals with broadband stationary random signals can be mistaken with extreme signal values from probability density analyses, the time signal of broadband random signals exceeding ± four standard deviations should be considered erroneous. These limits may be increased when the signal is produced by an accelerometer mounted on a structure with a hardening spring non-linear response characteristic, or when the signal has an unusually long duration.



EN 15433-3:2007 (E) 14
a) YX b) YX c) YX Key X-axis Time t Y-axis Instantaneous value x(t) Figure 7 — Intermittent noise in periodic (a), random (b) and transient signals (c) 4.4 Power line pickup As improper shielding or grounding of a data acquisition system can cause strong power line contamination and thereby distort the measured signal, it is essential to detect and correct this contamination during data acquisition system calibration. A break in the shielding of a signal transmission line, or the grounding of the data acquisition system at two or more points, are the most common causes of excessive power line pickup. NOTE 1 Power line pickup may occur occasionally in the data playback equipment. In this case, the data signals are not contaminated and exchanging the data playback equipment easily eliminates the problem. Power line contamination can properly be detected with a spectral analysis of the signal. NOTE 2 Power line pickup produces a periodic oscillation, often with multiple harmonics, superimposing itself on the signal time history of interest, A visual inspection of an analogue time history shows a strong pickup easily, as shown in Figure 8.



EN 15433-3:2007 (E) 15 a) YX b) YX c) YX Key X-axis Time t Y-axis Instantaneous value x(t) Figure 8 — Power line pickup in periodic (a), random (b) and transient signals (c) Severe power line contamination may occur early in the data acquisition system due to a ground loop involving the transducers, and thereby saturate the instruments that follow in the data acquisition system, or force sufficiently low sensitivity gain settings on the instruments to cause the signal to be buried in the instrumentation noise.
In such cases, all meaningful information is lost and the measurement shall be rejected. NOTE 3 Unless not too severe, power line contamination is no reason to reject measurements which are intended for spectral analysis purposes, although it may somewhat limit time domain interpretations of such data. If a spectrum is computed, it is recommended that all spectral peaks indicative of a sine wave at the power line frequency and all multiples thereof, are ignored. The only exception is when there is a conclusive, independent reason to believe that the data include a periodic component at the power line frequency, e.g. a measurement that represents the vibration response of a rotating machine operating at line frequency. 4.5 Spurious trends If a spurious trend or "zero shift" due to saturation of a piezoelectric transducer or amplifier is identified, the measurement shall be rejected and no corrective editing or trend removal shall be attempted.



EN 15433-3:2007 (E) 16 NOTE 1 If time history signals reveal a relatively slow variation (trend) in the mean value as a function of time, often with a period that is longer than the measurement duration, then such trends may be physically meaningful if the signal being measured has a time-varying mean value, and the frequency response for the data acquisition system permits the recording of statistical values. In the case of extreme shocks, the spurious trends commonly occur due to a severe saturation of a signal conditioner or the sensing element of a piezoelectric transducer, and generally render the data worthless. In other cases, spurious trends may occur when a signal is integrated, e.g. when a vibration response measured by an accelerometer is converted to a velocity signal. Spurious trends in otherwise valid measurements are usually obvious in the signal time history, as long as the lower frequency limit (f1) of the data acquisition system and/or the data signal is substantially higher than the reciprocal of the duration (Tr), of the measurement; i.e. f1 » 1/Tr. Figure 9 shows spurious trends in periodic, random and transient signals. a) YX b) YX c) YX Key X-axis Time t Y-axis Instantaneous value x(t) Figure 9 — Spurious trends in periodic (a), random (b) and transient signals (c) NOTE 2 Where a spurious trend is not due to saturation of an amplifier or piezoelectric transducer element, the presence of the trend in the data signal will in most cases only distort the low frequency content in a computed spectrum. Assuming the lowest frequency f1 of interest in the signal is substantially higher than the reciprocal of the measurement duration Tr (say f1 > 10/Tr), a valid de-trended signal can usually be recovered by high pass filtering operations.



EN 15433-3:2007 (E) 17 4.6 Signal dropouts Measured signals will diminish rapidly into the instrumentation noise floor for no apparent reason, and may or may not appear again at a later time.
The reason for such behaviour shall be detected, as it can cause non-valid results in the data analysis. In those cases where multiple-channel measurements are made, and the source of dropouts is common to all channels, a direct comparison of the time histories of two or more simultaneous measurements shall be performed to identify dropouts. NOTE 1 A temporary dropout, where the signal level ultimately recovers, usually indicates a malfunction due to a momentary saturation of a transducer or signal conditioner, or a tape dropout. Temporary dropouts can represent a malfunction in a transducer, an instrument in the data acquisition system, or a connector connection. A permanent signal dropout usually indicates a catastrophic malfunction, e.g. an accelerometer detaching from the structure or a signal conditioner losing power. Temporary or permanent dropouts may also occur in telemetered data when the transmitter antenna moves to a position where it is physically shielded, or when it is excessively remote from the receiver antenna. Figure 10 illustrates permanent and temporary dropouts in random signals. NOTE 2 With the exception of transient occurrences, dropouts can easily be detected by visual inspection of the time history signal.
In the case of transient occurrences, a signal dropout can result that
...