Hand-arm Vibration

VIBRATION TO HAND-ARM SYSTEM by Enrico Marchetti Researcher of the Italian Institution for Occupational Health and Safety Rome, 25th April 2007 Introduction Vibration transmitted to the hand-arm system are one of the main hygienistical problem for workers, besides noise exposure. Effects of hand-arm vibration (HAV) exposure interests different tissues and organs. Some effects are transients, other are permanent. In following pages we will give a sketchy description of effects. Moreover we will give an overlook of measurements, measurements apparatus, physical quantities used to describe vibration exposure and limits coming from the European Directive 2002/44/CE on this specific risk factor. As a matter of fact we will use this directive as a pathway to explore the matter, explaining different subject as they are mentioned in the directive that is attached to this Chapter as Attachment 1. Workers that are employed in automotive industry are strongly involved in this risk and should take care to contribute in risk assessment with their cooperation. To do so imply knowledge of the subject and this is the reason for giving this kind of information to young apprentices. Effects The article 2 says that the main effects of hand-arm vibration exposure are: “in particular vascular, bone or joint, neurological or muscular disorders”. The vascular effect is said “Raynaud’s phenomenon” and it consists of a vasospasthic crisis triggered by cold exposure. This looks like the fingers are blanching without any apparent reason. This effect is a permanent one and it is considered invalidating to the point of insurance economic compensation in most country of European Community.

Figure 1 – a finger blanching due to protracted vibration exposure. The Raynaud’s phenomenon needs long time exposure (we can talk about years) and high level of accelerations (see fig. 2). This syndrome has been first seen in mine and quarry workers, whose exposure is, even today, the maximum of all known. This exposure is characterized by the high level of acceleration and the low frequency of vibration.

Since the triggering is the cold, the subject should be kept always in warm condition with gloves, during work in cold environment.

Duration of exposure (Dy, years)

20

Dy = 31.8 A(8)-1.06

10

5

3 2

1 2

3

4

5

6

7

8 9 10

20

30

A(8) (ms-2) Figure 2 – probability of insurgence of Raynaud’s syndrome in time and acceleration exposure. Ostheoarticular (bones and joints) effects are permanent and frequently need surgical operation to resume part of the previous functionality of the arm. This effect is a modification of bones and articulation due to vibration exposure to low frequency vibrations.

Figure 3 – radiography of bone anomalous accretion due to vibration exposure. Mainly these are pseudocysthis, arthritis and bone accretion like the one in figure 3.

Neuromuscular effects are mainly tendonitis, nerve compression syndromes (i.e.: carpal tunnel nerve), other pathologies like Dupuytren. These includes both permanent and transient effects and are caused by medium frequency vibration. All transient effect allows return to normal condition after adequate rest: for example after the night rest without addictive exposure. The skin desensibilization is one of the more evident transient effects. This is due to the epithelial neuroterminations that are sensory saturated while exposed to high frequency vibration. This effect is typical of the transient ones, because after few hours of rest it is completely eliminated. All the effects of exposure to hand-arm vibration are affected by subjectivity, so they will not occur in all workers in the same way and at the same time, but they are more probable to occur if the exposition is severe. So we need a way to define the severity of the exposition. This is the topic of the next paragraph. Measurement and physical definition The article 3 says that vibration exposure measurement and assessment should be done following the technical annex, part A – point 1. This point refers directly to international standard ISO 5349-1 (2001), chapters 4 and 5. This standard describes in detail what should be done to measure and to assess personal vibration exposure level. The exposure quantity is indicated as A(8) and is directly related to the triaxial acceleration measured inside the palm of the hand that holds the vibrating tool. So we must start from the acceleration that goes from the vibrating tool to the hand.

Figure 4 – strong periodic component in a typical vibratory acceleration signal. This acceleration has a strong periodic component (as can be seen in figure 4) that gives the sensation of vibration. This component has to be characterized to assess the effects that are frequency dependent. So we need to measure acceleration not only in value but also in frequency. This means to measure acceleration in time and then to Fourier transform it to determinate the energy that is present in each frequency band (see table 1 for the division of frequency bands). To understand what means to Fourier transform we must start from a single sinusoidal signal. This is characterized by an oscillating value between two extremes that are the amplitude of the signal (see fig. 5 for reference on what we are saying). Besides that we can notice that there is a time in which the signal goes from one extreme to the other. This time is called period. Its inverse is called frequency. The amplitude is measured in the same variable quantity as the signal so, if it is an acceleration, this will be meter per square second (m/s2). The period is measured in seconds in the International System (IS) of measure, that is compulsory in Europe. The frequency is measured in hertz, after Gustav Hertz (symbol: Hz). One hertz is one complete oscillation in one second: 1 Hz = 1/1 s-1 (and is indicated as a symbol as time elevated to minus one). If we have a multisinusoidal signal, instead of the more simple one we have just seen, we can say that it will be composed from many pure sinusoidal signals linearly summed, having different frequencies (see fig. 6). To know this signal we will have to know all the frequencies that contribute to it.

Sinusoidal signal 1,5

1

750

730

710

690

670

650

630

610

590

570

550

530

510

490

470

450

430

410

390

370

350

330

310

290

270

250

230

210

190

170

150

90

130

70

110

50

f1

30

0 10

Amplitude of the signal

0,5

-0,5

-1

-1,5 Time (s)

Figure 5 – pure sinusoidal signal Every signal can be described mathematically as the sum of many pure sinusoidal signals of known frequency. If the signal is very strange this sum can be infinite. Usually acceleration signals are composed of a big but finite number of frequencies. To study this signal we must know all the frequency values. It is not necessary, anyway, to find every each frequency, but it is enough to know the energy contained in some range of frequency that are called frequency band. There are some way of defining frequency bands, but the more generally used is that of octave bands, or its sub multiple: one third of octave bands (1/3 octave bands). Multisinusoidal signal 3

2

Amplitude

1

0 1

3

5

7

9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75

-1

-2

-3 Time (s)

Figure 6 – Sum of three signals with different frequencies giving, as a result, the black line signal. Usually octave bands between 8 and 1000 Hz are admitted to assess vibration exposure. This means a frequency range from 5,6 (the lowest frequency value of the 8 Hz band) to 1400 Hz (the highest value of the 100 Hz band – not seen in tab. 1). More frequently, one third of octave bands between 8 and 1000 Hz are those used to assess vibration exposure. This means a frequency range from 6,3 (the lowest frequency value of the one-third 8 Hz band) to 1250 Hz (the highest one third of octave value of the 1000 Hz band – not seen in tab. 1). This second range is slightly narrower than the octave one, but this is of no concern, given the fact that in such areas weighting is very small. The real difference is in the fact that if we have a peak it is more likely to see it in one third octave band than in simple octave band.

Octave 1/2 octave 1 1 1,4 2

2 2,8

4

4 5,6

8

8 11,2

16

16 22,4

Table 1 – generally adopted frequency bands 1/3 octave 1/6 octave octave 1/2 octave 1 1 31,5 31,5 1,12 1,25 1,25 1,4 45 1,6 1,6 1,8 2 2 63 63 2,24 2,5 2,5 2,8 90 3,15 3,15 3,55 4 4 125 125 4,5 5 5 5,6 180 6,3 6,3 7,1 250 250 8 8 9 10 10 11,2 355 12,5 12,5 14 16 500 500 16 18 20 20 22,4 710 25 25 28 1000 1000

1/3 octave 31,5 40 50 63 80 100 125 160 200 250 315 400 500 63 800 1000

1/6 octave 31,5 35,5 40 45 50 56 63 71 80 90 100 112 125 140 160 180 200 224 250 280 315 355 400 450 500 560 630 710 800 900 1000

The most general case gives a signal that is the summation of different signals with different frequencies and amplitudes, but this doesn’t add any understanding in the concepts we introduced and complicates significantly the processing of data. The Fourier transformation process of the acceleration time signal to the frequency power spectrum (i.e. the calculation of frequencies and amplitudes) can be done analytically with the following formulas: ∞ a nπx nπx ⎞ ⎛ + bn sin f ( x) = 0 + ∑ ⎜ an cos ⎟ (1) 2 n=1 ⎝ L L ⎠ Where the amplitudes are multiples of the fundamental one: 2L ⎧ 1 nπx = a f ( x) cos dx ⎪ n ∫ L 0 L ⎪ (2) ⎨ 2L 1 n x π ⎪b = ⎪ n L ∫ f ( x) sin L dx 0 ⎩

The Cartesian system of axis that will be used to measure acceleration are those reported in figure 7. Actually it is possible to choose from two set of axis: biodynamic and basicentric. The first is oriented in function of the direction of transmission of vibration in the forearm, so the z axis has

been chosen along this direction and other axis are arranged to be orthogonal to this one. The basicentric system is centred on the emission of vibration by the tool, so its y axis is parallel to the tool handle direction, the other axis being arranged in order to be orthogonal to that one.

Figure 7 – Axis systems for acceleration measurement. The acceleration signal should be processed in order to extract the r.m.s. (root mean square) value, to cope with its strong periodical component that would put to zero the simple average. As a matter of fact the average of a sinusoidal signal is well known be zero, as easily demonstrated by the following figure 8: Siusoidal value and its root square 1,5

1,0

0,5

0,0 1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

-0,5

-1,0

-1,5

Figure 8 – difference in averaging a sinusoidal signal and its square. The difference arises because all sinusoidal values have an equal value of different sign, so it gives zero when summed up. For the square this is no longer true, because the squaring operation gives only positive values. Hence when summed up they give always a positive value. Of course this value is higher than the original one, so there must follow the extraction of the square root. So we have the root mean square. This is done with the following formula: T

1 2 a hi = a (t )dt T ∫0

(3)

Different frequencies have different probability of inducing effect so, to allow confrontation between frequency band effects and to admit summation of probability of causing those effects, we

need a weighting factor to multiply for each frequency band measured acceleration. This is done by the set of factors reported in figure 9 as a curve and in table 2.

Figure 9 – Weighting factors curve for all three axis. Table 2 – weighting factors for one third frequency bands.

The single axis acceleration r.m.s. signal ahki in the frequency band i should be weighted and summed with the signals of the other frequency bands. This can be done with the procedure described in the following paragraph.

The ahki signal will be multiplied for the appropriate weighting factor, then, being comparable for the effects induced, they can be summed up. In order to preserve their energy content from the problem of zero averaging that we mentioned before, this will be done with square root: a hwk =

∑ (W

i

⋅ a hki )

2

(k = x, y, z )

(4)

i

Then different axis should be composed to give a single number that is representative of the acceleration exposure of the worker: 2 2 2 a hw = a hwx + a hwy + a hwz

(5)

Usually measurement instrumentation does the whole job of recording, transforming, weighting and summing acceleration, freeing the user from the task of doing complicated mathematical processing. The exposure has a duration time T that has to be evaluated. This is the time that the workers use effectively the vibrating tool. Since the daily working time is generally 8 hours, the exposure will be referred to this period of time, as stated by article 3. To do so we use a equal energy principle; this says that if an acceleration a has a duration T, then it will have the same effect of an acceleration a’ that has a duration given by the following proportion that relates energy of both situations: a 2 ⋅ T = a ' 2 ⋅T ' (6) This gives the exposure indicator A(8): T (7) A(8) = a hw ⋅ T0 where T0 = 8 h as in the European directive (see art. 3).

Different tools exposure of the same worker will be composed following the same equal energy principle, previously exposed. So if a worker is exposed to three tools having acceleration respectively of 2, 3.5 and 10 ms-2 and a duration of the exposure respectively of 1, 3 and 0.5 h, then the daily exposure will be: 1 (8) A(8) = ⋅ (2 2 ⋅ 1) + (3.5 2 ⋅ 3) + (10 2 ⋅ 0.5) = 3.4 ⋅ ms − 2 8 This is a practical example of the general formula that gives the way of summing contribution of different exposure to different tools in the same working day:

[

A (8 ) = tool

]

1 8

n

∑

i =1

2 a hwi Ti

(9)

Where ahwi is the measured acceleration for each i tool and Ti is the exposure time for each i

Measurement instrumentation The technical annex, part A – point 1 defines characteristics of the measurement instrumentation that is included in the same standard 5349-1 that we mentioned previously. In this standard is indicated the instrumentation and the methodology of measurement, that we will summarize here. The measurement chain is composed by a transducer, a cable connection, a preamplification, an analyzer and a recorder. Some of the parts can be integrated in a single component (generally preamplification, analyzer and recorder). The transducer is an object that transform an acceleration signal in an electric one. This is performed, in principle, by a piezoelectric crystal: a crystal that has a defined electric field due to its structure (see fig. 10). Any mechanical deformation modifies the electric field and so induce an

electric signal that is directly proportional to the mechanical deformation. Since the acceleration is proportional to the force (F = ma) and the force is proportional to the electric field multiplied by the electric charge (Eq = F), this leads to the relation of direct proportionality between acceleration and electric field and charge (a = Eq/m). So we have two choices: to fix the electric field and measure variation of the charge; otherwise we can fix the electric charge and measure the variation of the field. Technically the difference between both solutions consists in placing a charge amplifier between the transducer and the analyzer or a field amplifier inside the transducer.

Figure 10 – The electric field of a piezoelectric crystal. The accelerometer (or the transducer) is a small object like those represented in figure 11 and 12. The dimension is about 1 centimetre, the weight is about 20 mg. The cable connect the accelerometer with the measurement instrument. If the accelerometer is triaxial (like the one in fig. 11) this cable will be a bundle of three cables, otherwise it will be a single cable.

Figure 11 – Tension (electric field) accelerometer.

Figure 12 – Charge accelerometer.

The charge amplifier is bulky and delicate, while the field amplifier is a normal tension amplifier, that is a transistor. This can be as small as we want and resistant to anything but the nuclear explosion. Conversely the charge accelerometer is very resistant while the tension one is delicate and tend to suffer from hysteresis. The measurement instrumentation can be an electrometer. In order to simplify computations and to record results, besides the weighting operation, there are composite instruments that do all the thing we said. An example is that of figure 13, that is a four channel portable analyzer with weighting curves both for whole body and for hand-arm. Other more sophisticated instruments can have more channels, or the real time processing and visualization of signals. This kind of instruments offer immediate frequency band distribution of energy and allow to adjust the measure to the effective situation selecting, for instance, the better choice of band for highlighting the physical phenomenon, or showing the time history to allow exclusion of spurious peaks. An example of such an instrument is reported in figure 14.

The most complete instrument is in the head of God (that means it will never be thought or build). It will be composed of an infinite number of channels and the best real time processing, allowing to show single, infinitesimal, fraction of hertz. However, waiting for better times, the best instrument actually in use is the OROS 38, with 32 channels and a good real time processing software. It is controlled by a portable PC that provides post elaboration. It has an hard disk built-in to allow stand alone measurement and it is damped to avoid dangerous effects of vibration on the instrument (see fig. 15). Of course any channel has a cost. This is valid for vibration measurements as for anything else. So if you follow the upper description you will also follow an increasing price list. The first being the cheapest.

Figure 13 – analyzer Svantek with hand-arm accelerometer and adapter to insert it between hand and tool.

Figure 14 – Real time analyzer Harmony, four channels, power spectrum and time history.

Figure 15 – OROS 38 with LAN cable to the PC and connections to accelerometers. Besides the analyzer there is a need of a field calibrator, to check for environmental or traumatical measurement error. Due to the fact that the measure is a derivate one, we must check for imprecision arising from non-linearity of the transducer or the analyzer. Field calibration is done with an electric vibrator (see fig. 16) that emits a pure sinusoidal signal at a specified frequency and at a given intensity. Usually it is approximately 160 Hz and 10 m/s2. This allows to check every time the measurement chain before and after the set of measures. If there is a mismatch between the calibration run before and after, all the measure are to be discarded. This suggest as a safe routine to repeat frequently calibration during long and/or important measurement session.

Figure 16 – Field calibrator for accelerometers and measurement chain. Risk evaluation and limits The article 4, paragraph 1, says that the: “the employer shall assess and, if necessary, measure the levels of mechanical vibration to which workers are exposed”. This means to have a couple of data that defines exposure: acceleration of the tool and exposure time. The evaluation of this set of data complete the assessment for the single tool exposure. We have now all the supporting instrument and knowledge to run this evaluation. We just need to refine our measurement protocol and put everything together. We have, as stated in the directive annex part A, first paragraph, two ways of performing acceleration assessment: first making use of declared emission vales and, second, running autonomous measurement. The second path is preferable to the first because it makes a “photography” of the vibrational exposure of workers in specific conditions (tool use, tool’s point or blade consumption, material worked, etc.), but some national implementation of the directive (for instance the Italian and the Swedish) have started the national making of a database of acceleration values to allow for the first assessment path. The Italian database contains both kind of values: field data measured by national experts and emission values declared by the tool producer. Of course this two set of data are not comparable. This is so because the sensibility of the measure to the specific condition of work. Many aspects are influencing the measure output: the condition of the tool, the physical constitution of the worker, the object and/or the surface on which the tool is used, the personal way of utilising the tool of the workers, the grip and feed force on the tool, etc. Generic emission data are not representative of such a variability and must be used with great care. It is far better to use on spot measured data than generic emission ones. Unfortunately there are not so many measured values and so we need to have also declared ones. Those data should be used with some care. Declared data are measured by producers following specific emission standard and are not intended to be used for prevention activity. So often they are far from realistic situation. Such a case is, for example, the grinders or the sanders that are measured on a smooth surface to evaluate the emission of their own electrical engine. These data should be amplified of a factor of about 1.5 to be used for preventionistic purpose. Otherwise chainsaw declared data are measured in more realistic situation and are immediately fit for prevention. In Italian database (that will soon be translated in English to be usable in most European country) there is an instruction page that help people to look for the correct factor for the usage of declared data. In this page there are three tables picked from the tobe-published standard CEN 15350, where this procedure is first proposed.

Besides databases the best way of running vibration exposure assessment is to measure exposition as described in the standard listed in annex A: EN ISO 5349-1 (2004). This standard ask to measure three times the acceleration between hand and tool and perform the average. Measurement set up should be thought to approximate at the most the real working condition. To comply with the measurement reference axis system the point of attachment of accelerometers should be the nearest to the centre of hand’s palm. The use of an adaptator like the one represented in figure 17 is strongly encouraged.

Figure 17: palm adaptator for lodging a triaxial accelerometer and positioning it between palm and tool. Having evaluated the exposure time of the worker to that tool it will be assessed his exposure with the formula n° 7. Limits are stated, in Directive, in article 3 and are: − the daily exposure limit value standardised to an eight-hour reference period shall be 5 m/s2; − the daily exposure action value standardised to an eight-hour reference period shall be 2,5 m/s2. The limit, of course, is intended not to be exceeded. The action value is a value above which the employer should take some cares to reduce exposure of workers and put them under health surveillance. In the next chapter we will see some exposure reduction procedures. Health prevention and exposure reduction The article 5 it is stated that the employer should tend to reduce exposure is action value is overcome. In the same article there are some provision aimed to that purpose: (a) other working methods that require less exposure to mechanical vibration; (b) the choice of appropriate work equipment of appropriate ergonomic design and, taking account of the work to be done, producing the least possible vibration; (c) the provision of auxiliary equipment that reduces the risk of injuries caused by vibration, such as (…) handles which reduce the vibration transmitted to the hand-arm system; (d) appropriate maintenance programmes for work equipment, the workplace and workplace systems; (e) the design and layout of workplaces and work stations; (f) adequate information and training to instruct workers to use work equipment correctly and safely in order to reduce their exposure to mechanical vibration to a minimum; (g) limitation of the duration and intensity of the exposure; (h) appropriate work schedules with adequate rest periods; (i) the provision of clothing to protect exposed workers from cold and damp.

These provisions, but for g), h) and i), are generic and should be technically implemented with solutions that take into account technical progress. This is the field of safety and work health prevention researchers. To describe all the activity in this field will take too long and lead us too far. We have an international conference on the subject that takes place every four years and always introduce new entries. This year it will be held in Italy in June. The logical progress from high exposure to a safe one is to measure exposure in order to know which frequencies are more dangerous. To substitute the tool, if practicable. In alternative (or in addition) to insert an handle, between tool and hand, that smoothes down the more dangerous frequencies. To provide protecting gloves, that are to be tested following the EN ISO 10819 (1998). At this point another measurement should be done to assess the actual level of exposure and, if still higher than the limit, the reduction of effective exposure time should be provided. Gloves have the extra effect of protecting the hand from cold and damp that are vasospasthic and so synergic with vibration inn inducing finger blanching. Besides these reduction provisions the employer shall establish a suitable programme of maintenance of tools that prevent the condition of maximal exposure due to tool deterioration. This will be done knowing the frequencies that are more dangerous, and which is the part of the tool that produces them, and acting with maintenance on them previously of deterioration. The tool manual can be of use in doing such a practice. This is why the manual should be provided in the national language by the tool producers. The employer should make a programme of information and training (art. 6) that will be run the first time at the beginning of activity by the worker and repeated every time there is a significant change in the working practice. I wish to thank Ing. Federica Morgia for her valuable contribution in helping me and giving me suggestion to improve this very synthetic work. Further readings in internet 1) “Vibrations and Waves”, by Benjamin Crowell, Book 3 in the Light and Matter series of introductory physics textbooks, www.lightandmatter.com 2) “Guide to good practice on Hand-Arm Vibration” - Non-binding guide to good practice with a view to implementation of Directive 2002/44/EC on the minimum health and safety requirements regarding the exposure of workers to the risks arising from physical agents (vibrations). Found on www.humanvibration.com. Or on the European Community site.