Mechanical Engineer''''s Reference Book 2011 Part 15 ppsx

Vibration isolation and limits 15/137 2 4 68 2 4 68 2 4 68 50 30 1 10 Amplitude (p) Figure 15.172 Human sensitivity: Rezher-Meister scale (vertical vibration) Octave pass band L I 6.3 m/s2 4.0 0.63 0.4 0.25 , ~~l~l~~~~~il~~~~~ll~il,,,., 0.4 0.63 1.0 1.6 2.5 4.0 6.3 10 16 25 40 63 Hz 100 Qne-third octave band centre frequency Figure 15.173 IS0 vibration criteria for a person in a vertical position 15/138 Plant engineering 100 50 30 - N I 20 2 m 3 0- LL E 10 5 13 104 Amplitude (pm) Figure 15.174 Building damage criteria. Zone A - no damage, zone B - plaster cracking possible, zone C - damage to structure, zone D - total destruction At relatively low vibration levels cracks can occur in plaster (particularly around windows). At higher levels, structural members may crack and ultimately fail. These two types of damage may be easily attributed to mechanical vibration. Another type of damage may result from building settle- ment caused by ground-borne vibrations compacting the ground differentially beneath buildings. This type of damage is indistinguishable from settlement caused by other occur- rences. Specifications for maximum permissible vibrations may be found in DIN standards which are given in terms of maximum velocity (in mm s-l) which is allowable for different classes of buildings from ruins and historical buildings up to reinforced concrete structures. More accurate criteria may be found in the technical press and HMSO publications. One such type of drawing is shown in Figure 15.174. 15.9 Acoustic noise 15.9.1 Introduction - basic acoustics Sound can be defined as the sensation in the ear caused by pressure variations in the air. For a pressure variation to be known as sound it must occur much more rapidly than barometric pressure variations. The degree of variation is much less than atmospheric pressure. Audible sound has a frequency range of approximately 20 Hz to 20 kHz and the pressure ranges from 20 X N m-2 to 200 N m-’. A pure tone produces the simplest type of wave form, that of a sine wave (Figure 15.175). The average pressure fluctuation is zero. Measurements are thus made in I I I 1 I Figure 15.175 Sine wave terms of the root mean square of the pressure variation (abbreviated to RMS). For the sine wave the RMS is 0.707 times the peak value. Since RMS pressure variations have to be measured in the range 20 x N m-2 to 200 N m-* (a range of 10’) it can be seen that an inconveniently large scale would have to be used if linear measurements were adopted. Additionally, it has been found that the ear responds to the intensity of a sound (q2) in a logarithmic fashion. The unit that has been Acoustic noise %/I39 adopted takes these factors into account and relates the measured sound to a reference level. For convenience, this is taken as the minimum audible sound (Le. 20 X PO-6 N m-') at 1 kHz. The logarithm (to the base 10) of the ratio of the perceived pressure l(squared) to the reference pressure (squared) is known as the Bell, i.e. Since this would give an inconveniently small scale (it would range from approximately 0 to 14 for human response), the Bell is divided numerically by 10 to give the decibel. The equation Itherefore becomes: D2 D 15.9.2 Sound intensity Sound intensity, I, is a measure of energy and its units are watts per metre. Intensity is proportional to the square of pressure. Sound intensity Ievel (SIL) is defined in a similar manner to sound pressure level. In this case the equation is dB (sound intensity level) = 10 loglo, - I {ref 15.9.3 Sound power Similarly, the power of a source (measured in watts) can be expressed in terms of decibels (in this case called the sound power level (SWL)) dB (sound power level) = 10 loglo, - Wref is taken as IO-''. It can thus be seen that it is important not only to express the unit but also to state sound pressure level (SPL), sound intensity ]level (SIL) or sound power level (SWL). W Wref 15.9.4 Addition and subiraction of decibels For coherent sound waves addition of values is possible. It will be apparent that as the scale is logarithmic, values cannot merely be added to one another. Intensities can, however, be added anld thus the equation becomes 11 + I2 SIL (total) = 10 loglo - Iret Le. 70 dB -k 73 dB 11 70 = 10 loglo - Irer antilog 7 I 1- 4et 12 73 = 10 loglo - Iref antilog 7.3 p 2- Ire, SIL (total) = 10 log (antilog 7 + antilog 7.3) = 10 log (2.99526 x lo7) = 74.76 dB 5 10 15 L.2-L, (dB1 Figure 15.176 Noise-level addition graph The square of individual pressures must be added and thus the equation in this case must utilize P(tota1) = V(P: + P;) 15.9.5 Addition of decibels: graph method It is possible to use a graph to calculate the addition of decibels, even in the case of multiple additions (Figure 15.176). The graph is used in the following way: In the case of the addition of two levels - the difference between the higher and lower levels is plotted on the lower scale of the graph. The correction is then read from the vertical scale by projecting a horizontal line across to this scale from the point on the graph. The correction is added to the highest original level to give the total level. In the case of subtraction of levels - the difference be- tween the total sound level and the one to be subtracted is plotted onto the graph and the correction obtained as above. In this case the correction is subtracted from the total level to give the remaining sound level. In the case of multiple additions - if there are more levels to be added the first two levels are added using the graph and then the third is added to the resultant using the same method. 15.9.6 The relationship between SPL, SIL and SWL The total acoustic power of a source can be related to the sound pressure level at a distance r by the following equation (assuming spherical propagation): w = P2/(pc47Tr?) where p = density of the medium and c = velocity of sound in that medium. By substituting this back into the SPL equation we obtain SPL = SWL - 20 loglor - 11 (spherical propagation) situations, i. e. Point source on a hard reflecting plane Line source radiating into space Line source on a hard reflecting plane These equations are: SPL = SWL - 20 loglor - 8 (hemispherical propagation) SPL = SWL - 10 loglor - 8 (line source in space) It is also possible to derive equations for other common W140 Plant engineering SPL = SWL - 10 loglor - 5 (line source radiating on a These equations are useful for calculating distance attenuation effects. If the sound pressure level at distance ro is known it is possible to calculate the sound pressure level at position rl quite easily: SPLo - SPLl = 20 loglorl - 20 logloro plane) rl r0 SPLO - SPL, = 20 loglo - dB If rl is double ro it will be seen that the SPL will be approximately equal to 6 dB (2OlO log 2). This gives us the principle of a decrease in level by 6 dB per doubling of distance (inverse square law). For the line source the same calculation produces a difference of only 3 dB per doubling of distance. 15.9.7 Frequency weighting and the human response to sound In practice, noises are not composed of one single pure tone but are usually very complex in nature. It is essential that more than the overall noise level (in dB) is known in order to appreciate the loudness of a noise, as the ear does not respond uniformly to all frequencies. As previously stated, the ear can respond from 20 Hz to 20 kHz and the response can be demonstrated by equal- loudness contours (Figure 15.177). It can be seen in Figure 15.177 that there is a loss in sensitivity (compared to 2 kHz) of approximately 60 dB at the low-frequency end of the chart. It will also be seen that all the curves are approximately parallel, but there is a tendency to linearity at the higher noise levels. In order to produce meaningful readings it is therefore important to state the sound pressure level in dB and the frequency of the noise. A weighting can be imposed on noise readings which corresponds to the inverse of the equal- loudness contours. If this weighting is used all readings which are numerically equal will sound equally loud, regardless of frequency. 500 lk 2k Frequency Figure 15.177 Equal-loudness contours 20 50 100 200 500 Ik 2k Frequency (Hz) Figure 15.178 Weighting networks Originally, three networks were proposed (A, B and C) and it was suggested that these be used for low, medium and high noise levels, respectively. It was proved, in practice, that this introduced numerous difficulties particularly with rapidly changing noise levels when a change of filter network was necessary. It was also found that the ‘A’-weighting network corresponded very well to annoyance levels at all noise levels (Figure 15.178). It was therefore decided that the ‘A’ weight- ing would be used as the norm for noise readings concerning human response. There is another weighting network (the ‘D’ network) that is used for aircraft noise measurement. If it is necessary for engineering purposes to know the tonal make-up of a noise, several approaches are possible. The noise can be processed by a bandpass filter. The most common filters are octave band filters and the agreed centre frequen- cies are as follows 31 63 125 250 500 lk 2k 4k 8k 16k (Hz) If further resolution is necessary, one-third octave filters can be used, but the number of measurements that are required to be taken is most unwieldy. It may be necessary to record the noise onto tape loops for the repeated re-analysis that is necessary. One-third octave filters are commonly used for building acoustics. Narrow band real-time analysis can be employed. This is the fastest of the methods and most suitable for transient noise. Narrow band analysis uses a visual display screen to show the graphical results of the fast Fourier transform (FFT) and can also provide octave or one-third octave bar-graph displays. 15.9.8 Noise indices All the previous discussions have concerned steady-state noise. It will, however, be apparent that most noises change in level with time. It may therefore be necessary to derive indices which describe how noise changes with time. The commonest of these are the percentiles and equivalent continuous noise levels. Percentiles are expressed as the percentage of time (for the stated period) during which the stated noise level was ex- ceeded, i.e. a 5-minute L90 of 80 dB(A) means that for the Acoustic noise 15/141 5-minute period of measurement for 90% of the time the noise level exceeded 80 dB(A). Therefore LO is the maximum noise level during any period and Ll00 is the minimum. The variation of noise levels within a discrete period of time can best Ibe described by a set of Ln results (the more results available, the greater the representation of the noise event). Sound-level meters commonly measure Ln’s at seven points (commonly, L1, L2, L10, L50, L90, L95, L99). More so- phisticated modern machines are capable of being adaptea by the user ,and non-standard Ln’s are available. Leq (the equivalent continuous noise level) is defined as the continuoils steady noise level which would contain the same total acoustic energy as the actual fluctuating noise, measured over the same period of time. This concept may be understood by considering electrical power consumption. If a machine uses 4 kW for i hour 2 kW for 2 hours 1.5 kW flor 4 hours 1 kW fix 1 hour the total usage of power is 15 kW h-’. The equivalent power for the &hour period would be 1.875 kW. If two events are to be added together and the Leq derived we must first convert to intensity units. Addition may then take place directly using the equation: Y Zeq = - (Zltl + 12r2 + 13i3 . . .) where 7 = total time, I1 = intensity for the first event, tl = time for the first event, I2 = intensity for the second event; t2 = time for the second event, etc. The total intensity is then converted back to decibel units by T I dB(A) = 10 loglo- Zref where Iraf = reference intensity. However we usually know the levels in terms of dB(A) rather than intensities, therefore by substitution where Ll = level 1 in decibels, etc. As noise is often measured on the ‘A,’-weighted scale Leq is usually expressed in this way. In this case the nomenclature becomes LAeq. A further derivation of equivalent continuous level is the single-event level (SEL), also known as sound-exposure level or Lax. This a special type of Leq used for transient events such as the passage of aircraft, gunshots, etc. The SEL is a one-second Leq and can be defined as the steady level which over one second would contain the same ‘A’-weighted energy as the actual event (regardless of its duration). Thus 1 T Leq = IO log - (tl x 10~1’“ + r2 x 10~2”” + , . , etc.) SEL = 10 log (tl x 10~1’” + t2 x 1oL2’l0 + . . . etc.) where tl + t2 etc. are the durations of levels L1, Lz, etc. in seconds. 15.9.9 Noise-rating curves These are a set of graphs that are commonly used as a specification for noise from machinery. They are similar to Noise Criteria Curves (used in the USA to specify noise from ventilation systems). The rating of a noise under investigation 120 110 100 50 40 30 20 10 625 125 250 500 1000 2000 4000 8000 Mid-frequency of octave band (Hz) Figure 15.179 Noise-rating curves is the value of the highest noise-rating curve penetrated by the readings when plotted on the graphs (Figure 15.179). 15.9.10 Community noise units Noise has been defined as unwanted sound. To quantify noise is therefore much more complicated than to quantify sound itself (which is what we have previously considered). Units have to be derived from these purely acoustic measurements by assessment of experimental psycho-acoustic data. It has been found that the response to different types of aural stimulation cannot be described by one single measurement, and hence a number of different noise measures are used. We now have three distinct classes of measurement: 1. Noise Units - these are the basic physical measurements of sound (i.e. decibel). 2. Noise Scales - these are composed of a combination of physical measurements (usually sound level, time, etc.) (i.e. Ln’s, Leq, SEL). 3. Noise Indices - here other factors are used to modify the noise scales in order to more closely relate the noise scale to other factors (annoyance. for instance) A criterion is a noise index value which is used to describe the reaction of a given percentage of the population. 15.9.11 Road traffic Road traffic is assessed by an 18-hour L10. This is not the percentile for 18 hours but rather the arithmetic average of the 15/142 Plant engineering 18 one-hour LlO’s between 6 a.m. and midnight on a normal working day. 15.9.12 Air traffic It has been found that annoyance caused by airdraft flyovers is related to the average value of the maximum perceived noise levels and the number of events. The index is known as NNI (noise and number index) and is obtained from: NNI = Lpn(max) + 15 log(lON)-80 where Lpn(max) is the logarithmic average of the maxima of the flyovers and N is the number of flyovers. 15.9.13 Railway noise Railway noise is assessed in Leq units. 65 dB(A) Leq is the usual criterion at which double-glazing is fitted where new housing is built near to railway lines. 15.9.14 Noise from demolition and construction sites Hourly Leq is used as the index. 15.9.15 Noise from industrial premises British Standard 4142: 1990 is described in detail in Section 15.9.25 and is derived from the noise measured in Leq compared to a background level measured in Ln. 15.9.16 Measurement of noise The simplest sound-level meter consists of a microphone, an amplifier and a meter of some type. Sound-level meters are graded according to British and international standards. For most precision work a Type 1 (precision) sound-level meter is used. This has an accuracy of approximately *1 dB(A). Type 0 meters (laboratory) grade are rarely encountered. Type 2 (industrial) grade sound meters may be suitable for some initial survey work but may not be sufficiently accurate to comply with legislative requirements at all frequencies. In particular, the lower grade of instruments have poor perfor- mance above 10 kHz (the human ear responds to noise at least up to 16 kHz). 15.9.17 Microphones The microphone is a device for converting pressure fluctua- tions in the air into an electrical signal. For precision work two types may be chosen. The polarized condensor microphone consists of a very thin metal diaphragm stretched in close proximity to a back plate. This diaphragm is charged to a polarization voltage of 200 V (some are lower). The diaph- ragm thus forms a condensor with the back plate. Sound causes the diaphragm to move in relation to the back plate, thus changing the charge on the condensor. This can be sensed electrically and used to measure the sound. The pre-polarized (or electret) microphone is a develop- ment of the polarized microphone, the main difference being that the charge across the diaphragm is permanent (or almost) and no polarization is needed (which simplifies the electronics of the pre-amplifiers). The disadvantage of the polarized microphone is that it is very moisture sensitive. Condensation on the diaphragm may result in electrical breakdown which causes sparks. These damage the diaphragm, thus ruining the microphone. The pre-polarized microphone has the disadvant- age of slightly reduced long-term stability (although this has now been largely overcome). Other types of microphone have been used - notably the piezoelectric type - but these are not suitable for anything more than the most basic noise ‘survey’ meters. Microphones should be capable of measuring the pressure changes in the air without altering the pressure waves they are trying to measure. This may seem to be a fairly fundamental point but, unfortunately, this is not physically possible. The diagphragm must have sufficient frontal area in order to capture the pressure wave and hence produce a reasonably sensitive output. Some reflections,will occur at the diaphragm and hence produce addition and/or cancellation effects with incoming pressure waves. This effect will differ depending upon the angle of incidence of the sound on the diaphragm and the frequency of the pressure fluctuations. In the past it was necessary to have 25 mm diameter diaphragms in order to get a sensitive response and reflection errors were a significant problem. It is now common to employ 12 mm diameter microphones and these problems are now reduced. There are, however, still many specialized micro- phones produced but they fall broadly into three types: 1. Pressure microphones - used for measuring sound in ducts, etc.; 2. Free field - used for measuring sound (usually out of doors) in which the angle of incidence is at 0” to the centre line of the microphone; and 3. Random incidence - used for measuring sound (usually indoors) in a reverberant field where the angle of inci- dence is more random. Note that most precision sound-level meters are fitted with a switch which can change electronically the response between free field and random response. For infra-sound (sound below the normal audible range) measurement special microphones may have to be used. Although some ordinary microphones are capable of operat- ing at low frequency, great care has to be exercised in impedance matching if low-frequency cut-off is to be avoided. 15.9.18 The sound-level meter The precision sound-level meter incorporates the pre- amplifier in the nose of the meter (usually in the stem that the microphone fixes on to). The main amplifier is contained within the body of the meter and may either be auto ranging or may have one or more user-adjusted ranges. In older instru- ments the range had to be adjusted in 10 dB steps (which was very awkward to use with rapidly changing noise levels). Simple sound-level meters merely display the output of this amplifier onto an analogue meter (Figure 15.180). Modern sound-level meters are equipped with internal filters and intergrating circuits and can produce outputs in terms of percentiles, Leq and frequency spectra. Some sound- level meters have a computer-controlled circuitry that is addressable from a ROM cartridge which is inserted to load a program and then removed. These sound-level meters can then perform many functions as several cartridges are avail- able. The sound-level meter thereby becomes dedicated to one particular type of task. Memory power of sound-level meters is increasing daily and it is now common to hold many sets of data (for instance, percentiles) in the sound-level meter memory and download later (perhaps in a kinder environment) either to a printer directly or to a personal computer. If the PC option is chosen the data can be introduced to a graphics program and results displayed in a chosen graphics format which can produce elegant displays. Digital outputs are available on most sound-level me’ters which will enable connection to portable computers if much Acoustic noise W143 These units are now available in laptop computers. They are not, at present, being produced by the major instrumentation companies, who continue with their dedicated machinery. It has to be said, however, that the add-on units are not as fully developed as they might be. Current developments include the provision of amplifiers and power supplies to enable microphone connection directly, and if these prove successful the end of the dedicated sound- level meter may be in sight. Figure 15.1180 Schematic diagram of a sound-level meter greater memory is required (or if on-site processing is chosen). Sound-level meters are also equipped with a.c. or d.c. outputs which will enable the connection of tape recorders, etc. Ruggedized sound-level meters are available which are designed for leaving out of doors. These devices (often referred lo as environmental noise analysers) are fitted into steel wealhertight cases and have a large battery capacity (and the provision for external battery connection). They are fitted with their own printers. Battery and paper life is in the order of six days. Longer life may be obtained by the use of external batteries and minimizing the amount. of data being printed to the paper roll. 15.9.19 Digital signal analysis While analogue filtering of signals may be of some use, as previously described, if detailed information is needed inevi- tably digital processing is called for. The principle of frequen- cy analysis is known as Fourier Analysis. The Fourier series states that any complex signal can be represented as a series of sine waves of various frequencies, magnitudes and relative phase angles. An example of this is the square ‘wave. This signal may he represented by the series of sine waves composed of the fundamental frequency - a sine wave at three times the funda- mental and one-third of the amplitude, a sine wave at five tiirnes the frequency and one-fifth the amplitude, etc., with the progression carrying on to infinity. Electrically, this process is known as, FFT (Fast Fourier Transform) analysis. The narrow band FFT analyser displays this signal graphically (as a display with frequency on the x-axis and amplitude on the y-axis). Octave OF one-third octave analysers usually employ digital filters which are arranged such that real-time analysis is possible ((where the whole of a signal is analysed rather than merely a snapshot). The sophistication of the machine and the required upper frequency will determine whether real-time operation is possible or not. Bothi types of analyser have digital outputs which will enable downloading to larger com- puters for further manipulation or to allow long-term storage. It is now possible to obtain add-on hardware and software systems for existing personal computers which will enable them to be used both as statistical (Ln and Leq, etc.) and frequency analysers (both narrow and octave band, etc.). 15.9.20 Noise control Noise is capable of causing psychological, physiological and pathological reactions as well as physical damage to plant, machinery and building structures. The need for the control of noise is recognized in many statutes for the protection of both workers and members of the public in their homes. 15.9.21 Noise nuisance Section 80 of the Environmental Protection Act 1990 gives local authorities the power to serve a notice where certain classes of nuisance have occurred or may occur. The expres- sion ‘nuisance’ is not defined in the Act or indeed in any other. The use of the expression ‘nuisance’ can be traced back to legal action as far as the thirteenth century and its meaning is now well understood. Nuisance describes anti-social un-neighbourly behaviour, and has been taken to mean the interference with one’s neighbours in their day-to-day-activities. Noise nuisance can therefore be a statutory nuisance (by virtue of the Environ- mental Protection Act), a private nuisance (actionable at common law as a tort) or a public nuisance (a crime). FOF a noise to be a statutory nuisance it must also be a common law nuisance and hence a private or public nuisance. The concept of private nuisance is now well developed. Private nuisance is a land owner’s tort and is a complaint that the use or enjoyment of his or her land has been interfered with. The nuisance only applies to the occupier of the land and not his or her family or sub-tenants. There are two types of private nuisance. The first concerns rights attached to land (for instance, right of way) and the second to enjoyment of the land (which does have relevance to noise control). This class of nuisance is described as ‘where a person is unlawfully annoyed, prejudiced or disturbed in rhe enjoyment of land or with his health, comfort or convenience as an occupier’. The interference must be substantial and the duration, nature and level of the noise must be considered. A single event may not therefore constitute a nuisance. The area affected by the nuisance must therefore be consi- dered. One often-quoted remark is taken from the case of Sturgess v. Bridgam (1879), in which Theiseger, L. J., said ‘What would be a nuisance in Belgrave Square would not necessarily be so in Bermondsey’. However, care must be taken if it is to be assumed that because an area is already noisy extra noise will not constitute a nuisance. In one case another printing press in Fleet Street proved to be a nuisance (1907). Two other legal precedents should be considered at this stage. The first concerns sensitivity. In the case of Walrer v. Selfe (1851) the expression ‘ought this inconvenience to be considered - not merely according to elegant or dainty modes of habit or living, but according to plain and sober and simple notions amongst the English people’ was quoted. This forms a cornerstone of nuisance law and gives rise to the question of reasonableness of a nuisance. Special sensitivities are not therefore to be considered when the question of nuisance arises. This may have relevance to shift workers, for instance, 15/14 Plant engineering who while they might expect their daytime sleep to be protected by law, may be disappointed to find that the law will only protect their property against noise that would affect the enjoyment of the average person (i.e. one who is not sleeping during the day). The second precedent concerns the case of the aggrieved person who moves next to a noise source and hence suffers a nuisance. The law of prescription concerns private nuisances (but not public) and states that if things are done which affect your neighbour (with his or her knowledge) and continue for 20 years, you obtain the right to continue. However, this does not translate well to noise nuisance. If, for example, the noise has continued for more than 20 years but no one has been affected by it, there has been no noise nuisance and hence there can be no prescriptive right. This can be illustrated by the case of Sturgess v. Bridgrnan (1879). The plaintiff was a doctor who built a consulting room at the bottom of his garden against a neighbouring property and was affected by the noise of machinery from that proper- ty. The judge ruled that as the doctor had not known about the noise until he built his consulting room no prescriptive right accrued. Therefore in the common case of a complainant moving next door to a factory the normal rules of nuisance will apply, despite the factory occupier’s insistence that ‘they were there first’. 15.9.22 Health effects Exposure to noise has been shown (in clinical experiments) to cause nausea, headache, irritability, instability, argumenative- ness, reduction in sexual drive, anxiety, nervousness, insom- nia, abnormal somnolence, and loss of appetite, as well as the more well-known hearing loss. Generally these health effects were shown to occur at noise levels greater than 85 dB(A). In the case of hearing damage, numerous experiments have been conducted with the aim of arriving at a safe exposure to noise. It has been found that some individuals are much more susceptible to hearing damage than others. Some people may suffer permanent damage over a few months’ exposure while others may take years to develop the same damage (at the same noise levels). Physical injury occurs at sound pressure levels in excess of 140 dB (at this level there is a risk of rupture of the tympanic membrane) while levels greater than 130 dB result in acute pain. Statistical studies on workers exposed to noise levels between 75 dB(A) and 9.5 dB(A) lead to the following conclu- sions: 1. For a 40-year working life a daily Leq of less than 75 dB(A) will lead to negligible risk. 2. The experimental data would indicate that for higher noise levels, and corresponding shorter time periods, the risk to hearing damage is the same. For example, 78 dB(A) for an 8-hour period is the same as 81 dB(A) for a 4-hour period. 3. Above 7.5 dB(A) 8-hour Leq the risk of hearing damage increases proportionately with the rise in levels. 4. Most countries have legislation which restricts noise levels to 85 dB(A) k5 dB(A) with a tendency to reducing acceptable levels. It should be noted that at the UK’s limit of 90 dB(A) there is some risk of hearing damage. 5. Infra-sound (sound below the normal human audible range) is capable of causing health effects. More recent research indicates an effect similar to excess alcohol con- sumption and indeed a synergistic effect with alcohol has been noted. It may be that in certain cases infra-sound is capable of causing an increase in accident rates. High infra-sound levels are noted in the foundry industry and in drivers’ cabs in large vehicles. 15.9.23 Damage to plant/machinery/building structures Noise can lead to damage in two ways: 1. Directly - as a result of induced vibrations 2. Indirectly - as a result of interference with the operative’s normal function Direct damage includes vibration fractures of electrical com- ponents (particularly switch contacts), structural panels, etc. Damage to buildings occurs particularly around windows (infra-sound is particularly troublesome in this effect). Indirect damage is probably the greatest effect of noise levels. Operator performance is affected by fatigue and also the inability to hear potential problems with the machine (that might ordinarily be attended to with no significant damage resulting). In addition, the inability to hear shouted warnings may result in accidents and further plant damage. 15.9.24 Legislation concerning the control of noise 15.9.24.1 Environmental Protection Act 1990, Section 80 A notice may be served where a nuisance has occurred or the Local Authority think a nuisance may occur. Noise nuisance is not defined as such, but includes vibration. The notice may not be specific and may merely require the abatement of the nuisance. A notice may, however, require the carrying out of works or specify permissible noise levels. The time period for compliance is not specified in the Act, but must be reasonable. Appeals against a Section SO notice must be made to the magistrate’s court within 21 days of the serving of the notice. The grounds of appeal are given in the Statutory Nuisance (Appeals) Regulations 1990 and are as follows: 1. That the notice is not justified by the terms of Section 80. The most common reason for this defence is that the nuisance had not already occurred, and that the Local Authority did not have reasonable grounds to believe that the nuisance was likely to occur. 2. That there had been some informality, defect or error in, or in connection with, the notice. It may be that the notice was addressed to the wrong person or contained other faulty wording. 3. That the Authority have refused unreasonably to accept compliance with alternative requirements, or that the requirements of the notice are otherwise unreasonable in character or extent, or are unnecessary. This defence is self-explanatory. The Local Authority are only permitted to ask for works that will abate the noise nuisance. Other works (perhaps to comply with other legislation) should not be specified in the notice. They may, however, be contained in a letter separate from the notice. An example of this would be where food hygiene requirements were breached by the fitting of acoustic enclosures to food-manufacturing machines. Readily cleanable enclosures may be a require- ment of the Food Hygiene Regulations, but it should not be contained in a Section 80 Environmental Protection Act notice. 4. That the time (or, where more than one time is specified, any of the times) within which the requirements of the notice are to be complied with is not reasonably sufficient for the purpose. 5. Where the noise to which the notice relates is that caused by carrying out a trade or business, that the best practic- able means have been used for preventing or for counter- acting the effects of the noise. ‘Best practicable means’ Acoustic noise 15/14 incorporates both technical and financial possibility. The latter may be related to the turnover of a company. Theredxe a solution that may be the best practicable means for one company may not be so for another. 6. That the requirements imposed by the notice are more onerous than those for the time being in force in relation to the noise to which the notice relates of (a) Any notice under Sections 60 or 66 of the Control of (b) Any consent given under Sections 61 or 65, or (c) Any determination made under Section 67. Section 60 relates to a construction site notice. Section 61 is a consent for construction works. Sections 65-67 relate to noise-abatement zones (see below). 7. That the notice might lawfully have been served on some person instead of the appellant, being the person respon- sible for the noise. 8. That the notice might lawfully have been served on some person instead of, or in addition to, the appellant, being the owner or occupier of the premises from which the noise is emitted or would be emitted, and that it would have been equitable for it to have been so served. 9. That tlhe notice might lawfully have been served on some person. in addition to the appellant, being a person also responisible for the noise, and that it would have been equitable for it to have been so served. Pollution Act 1974, or 15.9.25 British Standard 4142: 1990 This British Standard is a revision of a standard first published in 1967 and was revised in 1975, 1980, 1982 and 1990. The standard purports to rate noises of an industrial nature affect- ing persons living in the vicinity. It gives a method of determining a noise level, together with procedures for assess- ing whether the noise in question is likely to give rise to complaints. It does make the point that while there is a correlation between the incidence of complaints and general community annoyance, quantitive assessment of the latter is beyond the scope of the document, as is the assessment of nuisance. The previous document has been used extensively as a guide to the assessment of nuisance in various circumstances (cer- tainly outside the scope of the document) and has gained a status that outweighs its original intention. Unfortunately, the early document was very flawed in its methodology (as is the current one) and resulted in numerous difficult legal decisions when it was produced in court as the definitive guide to noise nuisance. In particular, the old BS 4142 had a method for obtaining a ‘notional background level’ where the actual background level (i.e. that level which exists when the noise in question was suppressed) could not be measured, which was widely discredited as being grossly inaccurate. The new BS 4142 rates noise in terms of Leq over a measured time interval (one hour in the daytime and 5 minutes at night) and compares this level with a background measured in terms of the L90 of the ambient. If a noise has a duration shorter than the measurement period in question, an ‘on-time’ correction is applied by the use of the following equation: Ton LAeq Tr = LAeq T, -k 10 loglo - Tr where LAeq T, = Leq for reference period LAeq T,,, = measured Leq for the event Ton = time on T, = reference time period (5 or 60 minutes) Table 15.33 Corrections to noise level readings Noise level reading LAeq T minus background LA90, T Correction subtract from noise level reading (dB1 (dB) 6-9 4-5 3 <3 1 2 3 Make measurements closer to source and back-calculate theoretical noise ievel in isolation from background A further correction may need te be applied if the specific noise does not exceed the background by more than 10 dB. A simplified correction table is used (Table 15.33). Finally a correction is applied dependent upon the nature of the noise. If the noise contains a distinguishable, discrete. continuous note (whine, hiss, screech, hum, etc.) or if there are distinct impulses in the noise (bangs, clicks. clatters or thumps), or if the noise is irregular in character enough to attract attention, add 5 dB to the specific noise level to obtain the rating level. The assessment for complaint purposes is then made by comparing this rating level with the background noise level. If the rating level exceeds the background by 5 dB the standard states that the result is marginal, and if the rated level exceeds the background by 10 dB or more, complaints are ‘likely’. This background noise level is one of the main criticisms of the document as it is intended to include any existing noise sources in the area. The new noise source is therefore com- pared against the existing noise climate, even if most of this is produced by the same factory. The example given in the British Standard further reinforces this point by considering premises which produce 40 dB(A) at the nearest house when operating normally and yet the ambient fails to 29 dB(A) during a factory shutdown. Thus the existing contribution is already 11 dB. A new source is assessed which adds 4 dB to the existing (40 dB) ambient, and the result is determined to be marginal! If this situation were to continue the background sound level (as defined in the standard) would ‘creep’ upwards - obviously an undesirable situation and one that is addressed in a planning circular (Circular 10/73) that deals with planning and noise. This circular particularly addresses creeping am- bients and states that ‘the introduction of a new noise source into an area is liable to result in a creeping growth of ambient noise level, and consequent deterioration in the quality of the environment, even though each of the new sources, consi- dered separately, would not be liable to give rise to com- plaints’. This point alone is sufficient for the method to be discredited by Environmental Health Officers when investi- gating nuisances, and the standard is unlikely to be used by them as a definitive guide. Consequently, operators of indus- trial premises should not use the information given in this British Standard as evidence when arguing (in legal situations) that they are not causing a statutory nuisance. Further, this British Standard takes very poor account of the effects of discrete frequency components. It is quite possible for a narrow band component to cause a serious nuisance while being almost unmeasurable on an ‘A’-weighted scale. Consequently, more detailed narrow band analysis would be necessary and it is essential to compare the actual noise with the background noise within that narrow band (usually octave or one-third octave). The British Standard makes no mention of such a situation. [...]... Standards (1983) 14 BS 848, Fans for General Purposes Part 1: Methods of Testing Performance (1980): Part 2: Methods of Noise Testing (1985) 15 BS 2009, Code for Acceptance Tests for Turbotype Compressors and Exhausters (1953) 16 BS 157 1, Testing of Positive Displacement Compressors and Exhausters Methods for Acceptance Testing (1987): Part 2: References 15/ 151 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31... passes into the partition and the rest is reflected Energy that passes into a partition may be partially absorbed and transformed into heat This is likely to be very small in a plain partition The remainder of the energy will then pass through the partition by displacement of molecules and pass as sound in the same way that sound travels in air This can then pass to the edge of the partition and be... splitter silencer is employed (Figure 15. 184) These are capable of providing a high degree of attenuation, depending on the width between the elements The smaller the gap, the higher the attenuation Again, tables of performance are published by the major manufacturers It is necessary, in order to decide on the Sesign of the silencer to be 15/ 150 Plant engineering 15. 9.34.6 Rubber mounts Although these... - this is known as flanking transmission By far the greatest amount of energy, in a thin partition, will pass through the partition by actually causing the partition to vibrate in sympathy with the incident sound and, hence, re-radiating the sound on the opposite side The amount of sound transmitted through a partition is represented by the ratio of the incident energy to the transmitted energy This... Hygiene: 10 Asbestos Department of the Environment, Waste Management Papers: Number 18: Asbestos Wastes Number 23: Special Wastes (Chapter 4), HMSO ‘Material Health & Safety Data Sheets’, Asbestos Information Centre A Mechanical Seal Guide io API 610 Standard, 7th edn, John Crane Inc., New York (1990) Summers-Smith, J D (ed.), Mechanical Seal Practice for Improved Performance, Mechanical Engineering... 7346, Components for smoke and heat control systems Part 1: Specification for naturai smoke and heat exhaust ventilators BS 7346, Components for smoke and heat control systems Part 2: Specification for powered smoke and heat exhaust ventiilators 47 EH 40/89 Occupational Exposure Limits, Health and Safety Commission (1989) 48 Fuel Efficiency Booklets, Department of Energy (1984-1986): 1 Energy audits 2... found in references 68-70 References 1 Turton, R K., Principles of Turbomachinery,S & FN Spon, London (1984) 2 Horlock, J H., Axial Flow Compressors, Butterworths, London (1958) 3 Addison, H., Centrifugal and other Rotodynamic Pumps, 2nd edn, Chapman & Hall, London (1955) 4 Wislicenus, G., Fluid Mechanics of Turbomachinery,Vols 1 and 2, Dover Press, London (1965) 5 Karrasik, I et al., Pump Handbook, McGraw-Hill,... incidence (90”) At greater frequencies the partition will still be driven, but in this case by progressively lower angles of incidence The coincidence dip is not, therefore, a single dip but will result in a loss of soundreduction index at progressively higher frequencies The desirable insulation panel will, therefore, be massive but will not be stiff 15. 9.32 Absorbers 15. 9.32.1 Porous absorbers As sound... (Hz) IV Frequency (Hz) Figure 15. 181 Typical insulation characteristics of a partition mance of the panel at low freqencies until resonance occurs As the driving frequency increases, the resonance zone is passed and we enter the mass-controlled area The increase in sound-reduction index with frequency is approximately linear at this point and can be represented by Figure 15. 181 3 Coincidence - a panel... Handbook, BHRA, Cranfield (1984) Fern, A G and Nau, B S., Seals, Engineering Design Guide 15, Oxford University Press/Design Council, Oxford (1976) Merry, S L and Thew, M T., ‘Comparison between a hydra’dynamic disc seal and neckrings for a small process pump running in water and in mercury’, 9th B H R R International Conference on Fluid Sealing, 1981, Paper H2, p 333 Neale, M J (ed.), Tribology Handbook, . heat exhaust ventiilators References 15/ 151 47 EH 40/89 Occupational Exposure Limits, Health and Safety Commission (1989) 48 Fuel Efficiency Booklets, Department of Energy (1984-1986):. ary, in order to decide on the Sesign of the silencer to be 15/ 150 Plant engineering Figure 15. 183 Centre-pod silencer Figure 15. 184 Splitter silencer installed, to know the required attenuation. into the partition and the rest is reflected. Energy that passes into a partition may be partially ab- sorbed and transformed into heat. This is likely to be very small in a plain partition.