Everything you need to learn about controlling road traffic noise.
Contents:
- Introduction to Road Traffic Noise
- Noise from Individual Vehicles
- Subjective Rating of Automotive Vehicle Noise
- Characteristics of Vehicle Noise
- Vehicle Noise Legislation
- Relation between Noise, Engine Design, and Operating Parameters
- Relation between Noise and Engine Combustion System
- Low Noise Engine Design
- Traffic Noise Reduction as a Function of Reducing Noise Levels of Vehicles
- Vehicular Component Hardware Development for Control of Engine-Related Truck Noise
- Relating Light-Vehicle Sound Level Ratings to Actual Operation Levels
- Noise Control Hardware Development of Automobiles and Light Trucks
ADVERTISEMENTS:
Introduction to Road Traffic Noise:
Various noise surveys show conclusively that road traffic has been at the present time the predominant source of annoyance; no other single noise has been of comparable importance. Such a finding has been not surprising due to the large number of automotive vehicles in comparison with other machines. The total horse power which is ‘built in’ in automotive vehicles exceeds 20 times the horsepower of all other prime movers combined (aircraft, ships, power stations, etc.).
Having such a wide use (not only by industry, but by private persons), road transport has been generally very cost conscious. Economy has been therefore one of the prime factors which has so far dictated the development of vehicle design and operational methods.
For these reasons, in the commercial field, the more efficient diesel engine has replaced the petrol engine with a reduction of fuel consumption by a factor of two. Diesel engines operate at considerably higher peak combustion pressures and higher rates of pressure rise and thus result in greater noise and vibration.
The higher noise and vibration, however, has been in no way detrimental to the life expectancy of a diesel engine; its life in general exceeds that of a petrol engine. Improvements in economy have been also obtained by reducing the vehicle weight for the same load carrying capacity, and also engine are made smaller and lighter by running at higher speeds to produce the same or greater power. Due to all these reasons, the prime mover—the engine and the vehicle were gradually becoming noisier; therefore it becomes essential to introduce noise legislation in many countries.
ADVERTISEMENTS:
Apart from noise, there have been now numerous legislations in road transport in many countries which have been introduced to improve man’s standard of safety and comfort; the laws now cover many aspects of construction and operation of the vehicle.
These laws have already had a marked success in dictating basic principles of car and commercial vehicle design. For example, legislation covering exhaust emission has resulted in great advances and improvements in the combustion system design of a petrol engine. There are also indications that in the near future radical changes in engine and vehicle design would result from noise legislation.
The motorcyclists with silencers removed or cut go zooming past in residential areas enjoying a thrill by making all that racket, oblivious to the annoyance they are causing to other citizens.
In Delhi, the rapid growth of vehicles from 2.5 lakh to 6.8 lakh in last 10 years had added to noise pollution by a much higher level. The multiplicity of high types of vehicles and the failure to segregate fast and slow moving traffic results in high noise emissions as vehicles break and then accelerate frequently. Traffic noises have been of two types— noises generated by individual vehicles, and noises generated by a continuous flow of vehicles of all types.
ADVERTISEMENTS:
Noise from Individual Vehicles:
Noises from individual vehicles could be summarized as follows:
(a) Noise from Engine and Transmission:
It depends very markedly on the design of the car and particularly upon the method of support used for its moving parts. More expensive cars employ a more elaborate damping system so that noises are not transmitted to the body shall and thence to the outside world. Considerable improvements are now made to improve engine mounting systems, even in cheaper cars.
ADVERTISEMENTS:
(b) Exhaust Noise:
The reduction of exhaust noise has been a simple acoustic problem which has been completely solved. Unfortunately, the more effective and elaborate the exhaust system of a motor car, the more expensive, heavier and bulkier it has been likely to be.
An efficient silencing system is able to reduce the power output to any engine and, in consequence, the design of the exhaust system is a matter of compromise. Systems have been usually designed to keep within the law without significantly affecting cost and performance. A special case has been that of so-called ‘sports’ cars where there has been often a deliberate attempt to retain as much exhaust noise as possible because a ‘fruity’ exhaust noise constitutes a powerful; sale appeal.
Prosecution in different countries for excessive noise from exhaust systems do occur but seem somewhat illogical. It is usually the subjective judgement of a policeman which determines whether a car is deemed to be excessively noisy, although noise meters are available to be used.
(c) Slamming of Car-Doors:
This aspect of noise from motor vehicles causes most annoyance late at night, because it has been an intense yet intermittent noise of a type most likely to disturb sleep. There is of course, no law to control it. Some car firms have effectively solved the problem of noisy door closure, but most motor manufacturers are still turning out vehicles which make plenty of noise when doors are closed. It is a problem which can only be solved at the design stage of a new car model —and will be done only when there is legislation forcing motor manufacturers to produce noiseless door shutting devices.
(d) Brake Squeal:
Brake squeal has been particularly noticeable with modern dies brakes although drum brakes also exhibit this phenomenon. Its cause has been that the vibration produced during the application of brake resonates within the brakes structure, and is then further magnified by the body of the car. It appears that little could be done about brake squeal, except by introduction of some damping devices into the vehicle.
(e) Use of Horn:
All motor cars must be fitted with horns by law. It has been found that a visual stimulus like the flashing of headlights is more effective than an audible one in attracting attention.
Relative Noise of Vehicle Types:
Not all vehicles make the same amount of noise.
Tests carried out by using a test procedure gave the following results:
Luxury limousine- 77 dB
Small passenger car- 79 dB
Miniature passenger car- 84 dB
Sports car- 91 dB
Motor-cycle (2-cylinder 4-stroke)- 94 dB
Motor scooter (1-cylinder 2 stroke)- 80 dB
The difference between the noise level of a standard small passenger car and a sports car has been no less than 12 dB, which means the sports car is roughly 15 times noisier than the saloon car. Motor cycles, with their exposed engines and inadequate silencing arrangements, are notorious noise producers with a sound level roughly 30 times higher than that of a saloon car.
Motor scooters, on the other hand, only produce the same noise as a motor car. It is debatable whether there is any longer room on modern roads for fast motor-cycles and sports cars, which not only contribute excessive noise to the environment, but also figure prominently in road accidents. It can be argued with objective justification that these vehicles should be used only on enclosed race tracks, and not on the congested public highway.
Noise from a Continuous Stream of Vehicles:
To reduce traffic noises as a whole, steps should first be taken to reduce the noise from each individual vehicle, but road re-planning can be of immense help. It has been found that vehicles produce the maximum noise when accelerating in low gear- thus reduction of the number of stops and starts will lower overall traffic noise.
Bypasses, ring-roads and urban motorways all make useful contributions to the reduction of urban noise. It has been, however, necessary to plan these so that they do not themselves constitute a threat to the environment. Cuttings, shading by artificial hills and planting bushes and trees near busy roads all play their part in cutting down traffic noises.
Subjective Rating of Automotive Vehicle Noise:
A comprehensive study of the relation between subjective ratings is noise emitted by motor vehicles and the objective measurements with a sound-level meter has been carried out by Mills and Robinson. These tests were done using ‘live’ automotive vehicle noise in the open air on one of the test tracks.
The subjects were asked to rate the noises which were presented to them according to a six-point rating scale by verbal description, as shown in Table 1.
No descriptions were ascribed to the first and last categories, which the subjects were instructed to regard as extremes. For convenience in expressing results, the verbal categories of the rating scale were first expressed numerically, so that quiet became 2, acceptable 4, and so on.
The average subjective rating for each vehicle has been plotted against the recorded sound level in decibels (dB(A)). The scatter has been generally small and to a reasonable approximation, a straight-line relation exists between the subjective rating and the measured dB(A) level. The most significant point on the rating scale has been the numerical value 5 which corresponds to the demarcation line between ‘acceptable’ and ‘noisy’.
Only a small number of commercial vehicles have been able to comply with the 80 dB(A) criterion and over the past ten years a large number of commercial vehicles have now reached the level of ‘excessively noisy’; that is, 92 dB(A).
Rough values of traffic noise can be obtained by spot reading a meter that reads only in dB(A). To do this, you simply note down at regular intervals (say, one every 10 seconds, although a shorter interval, 4-5 seconds is preferable), the sound level as shown on the meter. When a hundred readings are obtained, the top ten readings, that is the highest ten levels in dB(A), are deleted.
The highest of the remaining ninety levels would be then roughly equivalent to the L10 value. A more accurate method of obtaining L10 is to plot the distribution of levels (level versus number of registrations at that level) and to calculate the standard deviation about the mean of the distribution.
Assuming that the distribution is Gaussian-normal (i.e., fits a bell-shaped curve which is symmetrical about the mean) then the 10 per cent level is found by adding 1.28 x the standard deviation to the mean. Similarly, the 90 per cent level (L90) is found by subtracting 1.28 standard deviations from the mean.
The formula for standard deviation is-
where x denotes the arithmetic mean of the total of N readings.
Needless to say, this is a very tedious method of calculating L10 and needs considerable dedication, especially when it has to be done eighteen times. An equally accurate out substantially less time consuming method of calculating L10 from spot readings has been again based on the assumption of a normal distribution.
It is necessary to subdivide the total range of noise levels into consecutive 2.5 or 5 dB(A) ranges and to compute the fractional cumulative probabilities in each 2.5 dB(A) level range. These can either be plotted on cumulative probability graph paper or the probability function values corresponding to each fractional cumulative probability, can be read from statistical tables and plotted on ordinary graph paper.
More sophisticated apparatus may be used to get the instantaneous registrations of level then the sharp-eye with a sound-level-meter. A combination of a graphic level recorder and a statistical analyser can be used, but still needs a cumulative probability plot to yield L10. Recent developments in instrumentation enable the whole procedure to be carried out automatically, yielding a digital read-out of the 18-hour L10 in dB(A).
Having obtained our L10 value, we have a measure that approximates the peak noise that can be expected from the road. A measure of the steady noise contributed by the general (remote) traffic-stream is the L90 value—that is, the value that is exceeded for 90 per cent of the time.
This value must be lower than that of L10 and approximates the steady background noise level present during the readings due to the local distribution of noise sources.
In many countries there is a composite unit called the Traffic Noise Index (TNI) which is given by the formula:
TNI = L90 + 4(L10 – L90) – 30
where all values are measured in dB(A)
This formula indicates that two variables were regarded important for the Traffic Noise Index- one is the straight forward L90 value and the other is the difference between L10 and L10 which represents the variability of the sound reckoned that a sound which varies over a wide range of decibels is more annoying than one which is constant with time. The Traffic Noise Index has never been widely used in noise standards and for the purposes of the noise insulation regulations. The 18-hour L10 value is used exclusively; L90 is still used.
The basic equipment necessary to perform measurements of L10 and L90 is a precision sound level meter and a windshield. It has been important to recalibrate the meters before and after each set of readings with a pistonphone or other calibration device provided by the manufacturer.
Adding a further piece of equipment will cut labour costs. The signals from the sound level meter get passed to a sampling chart data recorder. The signals are recorded continuously for a period not exceeding 30 hours, if powered by a mains supply. When batteries are used they must get charged every eight hours.
One noise level reading is taken by the sampling recorder every 3¾ A seconds. The range gets limited to 20 dB and the precise peaks and troughs of the noise level are not recorded. Once it is set up, particularly with the insertion of a mains control switch which will automatically start sampling at 6600 hrs and cut out at 2400 hrs, and by a sampling device (such as used in air pollution sampling) which enables a 15-minute sample to be taken every hour throughout the measuring period, the system is completely automatic and readout is accurate to ± 2 dB(A). This is sufficiently accurate measurement of the L10 index.
However, a more widely used system passes the signals from the external microphone through a precision sound level meter and into a tape recorder. The tape is then processed subsequently by passing the signals back through the precision sound level meter (now used as an attenuator) into a pen recorder and then to a statistical distribution analyser.
From the statistical distribution analyser it becomes necessary to plot the 12 channel readings onto probability paper before deducing the required value from the graph. This requires experience. Furthermore, the accuracy of the tape recorders as a means of recording audible sound for later analysis is in question.
As such, the tape recorder used in this way has been a system weakness. If the weather is very cold or very humid, the accuracy of the tape recorder can be affected. Nevertheless, this system enables measurements of L10, L50 L90 or LEQ with reasonable accuracy given fair weather conditions.
More recently developed alternative systems could be able to record levels from an external microphone onto a cassette tape recorder via a data logger in digital form for subsequent or real-time processing by a mini-computer. These systems have been good and accurate without necessarily being more expensive.
Characteristics of Vehicle Noise:
The mechanism of radiation of noise to outside from a vehicle has been basically different from the generation of noise inside the vehicle. None of the noise-producing systems of the vehicles have been fully enclosed; if anything, they have been only partially screened.
Thus, the noise emitted depends on the relative levels, characteristic, and the interaction of the directly radiated noises from these systems. For current production vehicles the principal noise source is the power unit and its auxiliaries. Other important generators have been the transmission system, tyres, and braking system.
Vehicles could be classified from the emitted noise point of view as heavy commercials, light commercials, public service vehicles, small cars, large cars, and high performance cars.
The ISO drive past test involves accelerating the vehicle at maximum rate in a low gear over a distance of 20 m from the re-equarters of the engine maximum power speed past a microphone positioned 10 m from the start of the acceleration and 7.5 m from the centre line of the vehicle. The use of low gear (2nd or 3rd depending on the vehicle) ensures the maximum engine speed is reached at or shortly after the end of the test section.
This test procedure forms the basis of vehicle construction regulations in several countries, and for this reason most investigations of vehicle noise are limited to this test. The drive past test is frequently supplemented by a stationary test at maximum governed engine speed. It is to be noted that this test is only justifiable when the total vehicle noise is engine controlled.
The spectra of number of heavy commercial vehicles have been obtained. Most of these are powered by four stroke cycle diesel engines of cubic capacity from 4 to 16 litres with power output from 80 to 300 bhp. The spectra from most of the vehicles have been almost constant at about 80 dB up to 3000 Hz. A broad peak in the frequency range from 800 to 3000 Hz constitutes the characteristic ‘diesel knock’ emitted by the engine surfaces.
Above 3000 Hz the noise gets decreased by about 6 dB per octave. The scatter in the low-frequency noise between various vehicles has been attributed to variations of the exhaust and inlet silencing systems used.
The spectra of light commercial vehicles (less than 3.5 ton gross vehicle weight or with less than 12 seats) fitted with petrol and smaller capacity diesel engines have been recorded. These show the spread of the noise in this class of vehicle is considerably larger, being of the order of 15 to 20 dB. It is generally found that reduction of noise obtained by body shielding has been appreciable in this type of vehicle, which largely explains for greater differences measured.
The low-frequency noise which has been comparable with heavy goods vehicles arises not only from the exhaust and inlet, but also from the noise radiated by the body structure. High frequency noise from 800 to 3000 Hz has been 5 to 8 dB lower than from heavy commercial vehicles.
The spectra from small and large cars fitted with petrol engines are reasonably flat, the average spectrum level of the small cars being above 70 dB and for larger cars being somewhat higher. The high-frequency content of the spectra has been in one way less significant than that of the diesel engine vehicles.
The high performance and sports cars generate very high levels of low-frequency noise (100-200 Hz) which is intentionally allowed by inadequate silencing of the exhaust. The high frequency noise, however, has been almost the same as from other types of car.
Vehicle Noise Legislation:
During the 1950’s there occurred great advances made in automotive diesel engine development. It was found that automotive diesel engines could be satisfactorily run at higher speeds and higher specific loads, thus considerably improving the specific power output.
Although Daimler Benz successfully introduced a high speed diesel engine in a car before the war, it was not until the 1955’s that a widespread development of the high-speed diesel engine for vans and light industrial applications also took place. The noise of the automotive truck diesel engines produced before 1950, running at moderate speeds, was around 98 to 100 dB(A) measured at a distance of 1 metre.
By 1960 the engines were running much faster with often large bore-to- stroke ratios and the noise levels reached 103 to 107 dB(A). The result was an appreciable increase of vehicle noise. Similar trends could be noted also in petrol car engine developments.
It is for these reasons that in the mid-sixties the first noise legislation was introduced which since then have remained virtually the same. The legislation is usually based on practically observed data; trucks have been permitted higher levels of noise.
The detailed investigations on the origins of road vehicle noise have shown that noise radiated by the engine surfaces in various vehicles has been the main contributor. As there are significant differences in the legislated levels the question is whether the individual vehicle in a particular group is quieter because of a quieter engine, or whether it is due to basic principles of vehicle design.
The test-bed overall noise at 1 metre distance from the side of the engine, for a large number of engines running at rated conditions (maximum speed and load) has been determined and classified in the various vehicle categories. The data presented are from the engines measured at ISVR and AVL laboratories.
The mean values of the noise have been as follows:
Group 1-Truck diesel engines above 200 hp- 102.4 dB(A)
Group 2-Truck diesel engines below 200 hp- 101.8 dB(A)
Group 3-Van high-speed diesel engines- 101 dB(A)
Group 4-Car petrol engines- 99 dB(A)
It can be seen the mean levels between various engine groups only differ by 1.4 dB(A). Also the mean noise level of the petrol engines is only 2 dB(A) quieter than the diesel engines.
It may be concluded that differences of the noise vehicle groups have been mainly due to principles of vehicle designs.
As can be seen, trucks are having a high ground clearance, the engine gets more exposed, and the vehicle also provides little attenuation of noise. In buses, vans, and cars, the vehicle body provides considerable shielding of engine noise and the ground clearance is considerably smaller.
Investigation on the ISVR open-air pad have shown that attenuation with distance from 1 meter to 7.5 meters also depends on engine size. For large engines attenuation of noise is about 12 dB(A); for medium-size engines 13 dB(A); while for small high-speed diesel and petrol engines about 14 to 15 dB(A).
The general attenuation of engine noise measured as the noise on the test bed at 1 meter distance less the noise from the vehicle with no specific acoustic treatment, at 7.5 meters has been as follows:
Large truck engines above 200 hp- 12 dB(A) (approx)
Medium truck engines below 200 hp- 13 dB(A) (approx)
Large truck engine in a bus or coach- 15 dB(A) (approx)
Medium size truck engine in a bus or coach- 16 dB(A) (approx)
Vans- 16 dB(A) (approx)
Cars- 19 dB(A) (approx)
These values of attenuation correspond to values which are imposed by present noise legislation.
It is found that in vehicles where nosier engines are installed, the vehicle manufacturer is to incorporate in the vehicle design some noise reducing features, such as partial shielding.
Relation between Noise, Engine Design, and Operating Parameters:
Despite the numerous exciting forces which almost simultaneously excite engine structure there has been some justification to look at the problem in a simpler way. As the gas force resulting from combustion tends to be the predominant force in most of the engines, the relationships between the gas force characteristics and emitted noise can be used to establish a basic model to identify the effects of fundamental engine design and operating parameters.
The three basic parameters of an engine have been as follows:
(a) Engine Speed;
(b) Engine Size;
(c) Engine Load.
(a) Engine Speed:
The engine structure characteristics can be defined by use of electrodynamic vibration generators, and the broad response readily established. It will be seen that when the engine structure gets subjected to a constant magnitude sinusoidal force it gives rise to maximum response in the high-frequency range from 800-2000 Hz.
Electronic analyses show in some detail the existence of numerous natural frequencies at which the structure can vibrate. It also shows that it is reasonably heavily damped and thus any better resolution of various modes of vibration by any instrument is impossible.
The existence of the high density of various modes confirmed by finite element computer calculations of a simple model of a crankcase which confirms that in a very narrow frequency range of one third octave there can be at least some 20 natural frequencies of the crankcase walls.
The gas force which again in itself has been very complex, can be subjected to frequency analyses to quantify its exciting propensities. Analysis of the gas force reveals that in the low frequency range the magnitude of the harmonics has been a maximum, gradually decreasing with increasing frequency at higher orders. Comparing this force spectrum with the response of the structure one can visualise that only the high order harmonics (frequency range 800- 2000 Hz) are responsible for the predominant noise of the engine.
If the engine speed gets doubled, the pressure diagram remains the same shape and thus the amplitude of all the harmonics should remain the same, i.e., the same force spectrum. Since the actual event occurs in one-half the time (see inset of the pressure diagrams at 2000 rev/min) it implies that the actual force spectrum gets shifted sideways by a factor of two, namely from 8.34 to 16.7 Hz in this example.
We can see that the engine structure, in which the response frequencies remain the same, now gets excited with lower order harmonics which have higher amplitudes. Since the general slope of the force spectrum is about 30 dB/decade an increase of excitation by 9 dB will be obtained. With further speed increases the same pattern is followed.
It may be concluded that the characteristics of force determine the rate of increase of noise with engine speed which in this example, for a naturally aspirated diesel engine, is 30 dB per ten-fold increase of speed.
Effect of Combustion System on Noise:
The amplitudes of the high-frequency harmonics in the gas force are produced at the onset of combustion in the diesel engine; that is the explosion. It requires only slight modification of the diagram in this region (indicated by a circle) to reduce the noise by as much as 20 dB. This part of the diagram is having no effect on power developed by the engine. Only a few low frequency harmonics are needed to develop the necessary mechanical power.
It will be seen that the combustion system determines the rate of increase of noise with engine speed. The exponent for a petrol engine is 5, while for the most vicious diesel engine about 2.5. It is seen that at low speeds there is a possibility of reducing noise by as much as 25 dB(A) by smoothing the development of the gas force. As speeds are increased the lines tend to converge, clearly indicating that in high-speed engines the form of the gas force is irrelevant.
(b) Engine Size:
Measurements done on a large number of engines show that increase of noise with engine size is considerably less. An increase of size to ten times results in an increase of noise of 17.5 dB(A). The detailed investigations now reveal that vibration levels of the engine surfaces have been about the same irrespective of their size, thus the increase of noise with size has been simply due to a larger radiating surface area.
(c) Engine Load:
Engine load does not have effect on noise, which is in agreement with the findings that noise is simply due to the initial ignition of the fuel. This takes place at the same intensity whether the engine is running at no load at all or full load.
It can be concluded that:
(i) The form of the exciting gas force ascertains the rate of increase of noise with engine speed.
(ii) At high engine speeds the form of the gas force is having a less significant effect on noise.
(iii) Engine noise is independent of the horsepower produced.
Summing up Effects of Engine Design and Operating Parameters:
In the investigation of the variations in noise level produced by engines of similar cylinder capacity it becomes possible to find relevant examples which have provided further evidence on the parameters which control engine noise. There exist significant differences in the spectra because the natural frequencies of each of the structures are markedly different. The general overall levels of the noise spectra, however, are all within 1 dB.
Therefore, we conclude that one can add to the engine any number of cylinders and thus increase the power of the engine without increasing the noise. In this case the power increased by a factor of two.
Another example is illustrated for two engines with the same cylinder capacity and the same number of cylinders. Both engines have been of V8 form. The only differences between the two engines is the stroke-to-bore ratio. The fundamental differences of the two engines are illustrated in somewhat exaggerated form.
The over-square engine has been very attractive to the vehicle manufacturer because of its reduced engine bulk volume and weight for the same power output. Furthermore, it can run considerably faster for the same limiting piston speed. The result as can be seen, is a marked increase of noise of the over-square engine despite its smaller size.
The increase of noise is from 103 to 108 dB (A). These observations led to the conclusion that the basic parameters which determine the noise of an engine are only its speed and cylinder diameter, and the following relation can be derived.
I ~ Nn x B5 (n is combustion index)
By taking into account the level of gas force and engine structure characteristics the engine noise intensity becomes as follows:
I ~ Cf C5 (Nn x B5)
where Cf defines the level of the gas force
C5 defines the structure characteristic
B is bore diameter
N is engine speed
n is 1/10 slope of cylinder pressure spectrum in dB decade (combustion index)
Based on these, and other investigations the ISVR formula for predicting Direct Injection (DI) engine noise was derived which gave noise predictions for automotive engines over wide size and type ranges. The formula was also applied to Indirect Injection (IDI) and turbocharged engines but results were less satisfactory.
Relation between Noise and Engine Combustion System:
In order to establish the relevant differences in noise characteristics of engines of different combustion systems a more detailed study has been made of the results of all the different engines tested at ISVR laboratories. On each engine gas force diagrams have been taken, and cylinder pressure, noise and vibration analyses have been carried out.
The noise of all the engines show straight-line relationships with speed, except for some small high-speed IDI and petrol engines which show two slopes, a low rate of increase in the low-speed range and a high rate in the high-speed range.
If the results have been compared on the basis of constant speed the increase of noise with engine size becomes apparent. At the rated speeds, all engines (except the opposed piston two-strokes) reach about the same level of noise within a band of some 10 dB (A).
An attempt has been also made to classify the engines in various groups according to combustion system and fundamental design principles.
In each of these groups listed below, there have been about seven engines:
(a) Turbocharged in-line DI
(b) Naturally aspirated in-line DI
(c) Naturally aspirated vee-from DI
(d) Two-stroke DI
(e) Naturally aspirated IDI
(f) Spark ignition petrol
In order to find some correlation with a simple combustion model the overall noise levels are plotted against engine bore diameter, at a constant speed of 2000 rev/min which represents a realistic speed for all the engines in this wide range.
It can be seen that the engines do fall into specific groups:
(a) All normally aspirated DI engines fit within a 3 dB band of slope (bore). It is clear that there have been no differences between the overall noise of vee form and in line engines. Some of the IDI engines also fall within this same band, but these generally have abrupt or ‘advance’ pressure diagrams.
(b) Turbocharged engines occupy a band just below.
(c) The remaining IDI engines all within a band some 8 dB(A) below the DI engines. These engines generally have smooth or ‘retarded’ type pressure diagrams.
(d) Two-stroke cycle engines fall within a band some 4 dB(A) higher than the DI engines.
(e) Opposed piston two-stroke cycle engines fall in a band 12 dB(A) higher.
(f) Petrol engines show considerable scatter but are about 15 dB(A) the DI engines.
It is possible to explain many of the differences between these various groups by the salient features of cylinder pressure development. Both ‘advance’ and ‘retarded’ type IDI engines, and a petrol engine. The diagram of the turbocharged engine has been extremely smooth but with a high peak pressure of 100-135 bar.
The DI engine pressure diagram has been abrupt but the peak pressure is considerably lower at 65-80 bar. The advanced IDI diagram has been similar to that of the DI, while the ‘retarded’ diagram has a flat peak of smooth double hump peaking at 65-75 bar.
The petrol engine diagram has been smooth and the peak pressure very low at 35-50 bar. The different forms of the cylinder pressure development can be full described by their spectral analyses. The rate of increase of noise of the cylinder pressure form as cylinder pressure levels increase with increasing speed according to the slope of the spectrum.
Relation between Combustion-Induced Noise and the Overall Noise of the Engines with Various Combustion Systems:
In order to compare various combustion systems it becomes necessary to reduce them to a standard form which has been independent of speed (simplified normalized spectra). It is the levels of the harmonic components between 800 and 3000 Hz (the predominant range of engine structure natural frequencies) that, to a first approximation, decide the importance of the cylinder pressure spectrum.
Measured cylinder pressure levels for a DI engine taken at various speeds plotted on a base of normalized frequency. It can be seen that all spectra are coincident in the important frequency range and can be approximated by a single straight line. Such spectra have been derived for each engine tested and results have been grouped according to their various combustion systems.
It has been previously stated that noise from the normally aspirated DI engine has been mainly combustion- controlled, therefore the noise of engines of other combustion system should be judged on the basis of the DI engine results. A combustion model, developed at the ISVR enables the prediction of combustion-induced noise of the engine based on fundamental principles of thermodynamics and structural attenuation.
Using this model, overall noise of all engines could be predicted. The model clearly reveals that the combustion-induced noise of turbo-charged engines should be considerably lower than DI engines. As the overall noise of the turbocharged engine has been relatively higher than would be expected from combustion alone, it suggests relatively high levels of mechanical exciting forces in turbocharged engines.
Relation between Mechanically-Induced Noise and the Overall Noise of the Engines with Various Combustion Systems:
Prediction of the mechanically-induced noise has been more complex because the excitation occurs in many different parts of the engine, like piston slap, bearing impacts at big end and main crankshaft bearings, etc. The effects produced by the oil firm have been also at present difficult to predict. It can be regarded, however, that the main parameter which determines the intensity of the mechanically-induced noise has been the rate of the reversible force, that is dF/dt.
In the case of the piston slap it can be shown that intensity of noise, assuming only the kinetic energy release on impact, is-
I~ [Fd/dt δ4MK3]1/3.B
where
δ = clearance
M = piston mass
K = liner stiffness
B = cylinder bore.
The various parameters in the equation have been related to bore in the following manner:
δ ~ B
M ~ B
K ~ B
Thus, the intensity of noise is approximately given by-
I ~ dF/dt. B4
As for a given speed dF/dt is proportional to engine bore, the intensity of piston slap noise will be proportional to be (bore)5 –
That is, I ~ B5.
These considerations are able to explain why overall noise of the engines where combustion included noise is low i.e., turbocharged and IDI engines, the predominant mechanically-induced noise also increases with (bore).
The values of dF/dt for samples from each engine group are plotted against engine speed. It is apparent that dF/dt for each group has distinct values. Turbocharged engines, mainly on account of very high values of peak pressure, show rates of change of reversible forces higher by a factor of two than on naturally aspirated DI engines. IDI engines with smooth pressure diagrams and petrol engines have values lower than DI engines by a factor of two.
As the combustion-induced noise of both IDI and turbocharged engines has been of similar level, the higher mechanical noise (higher value of dF/dt) must be the cause of the higher overall noise level of the turbocharged engine.
Low Noise Engine Design:
From the point of view of noise legislation it has been the maximum level of noise which is important; that is, at its rated conditions of load and speed. There has been considerable latitude to select engine design parameters for low noise performance; that is, operating cycle, combustion system, and bore size.
There are, however, good economic reactions—production costs, gross vehicle weight limitations, and specific fuel consumption requirements—which tend to reduce this latitude in practice and to dictate the type of engine design needed for a specific duty. Thus choice is limited.
Another factor influencing engine noise has been the design of the internal parts themselves. That the natural frequencies of the crank system and bearing supports influence the noise radiated by the outer engine structure is clear, but serious research must be carried out in this area before any positive design recommendations could be made.
Some optimism must be derived from the fact that the noise of engines in a specific group covers a certain range at least 3 dB(A), in most cases more, between the highest and the lowest levels.
The factors influencing these variations could be summarized as follows:
(a) Pistons:
Pistons slap noise can vary widely, either from chance or design. Prediction procedures have been now well advanced, and optimum pin offsets can be readily calculated.
(b) Timing Systems:
Controlled experiments have revealed that chain derives are quieter than gears by 2-4 dB(A).
(c) Gear Location:
Timing systems at the rear of the engine have been preferable, where not only they are shielded by the transmission but are located near an antinode of the crankshaft. 2.5 dB(A) has resulted from this one modification.
(d) Engine Layout:
The petrol engine is having a smaller frontal area than the diesel because the camshaft and distribution derives are usually located on the same side. Diesel engines which have pump and camshaft on the same side have some advantage.
(e) Non-Stressed Covers:
Experiments have proved that on most engines a 3-6 dB(A) noise reduction would be possible if all cover noise could be completely eliminated. Selective use of damping, isolation, and stiffness has been the criterion of quiet cover design.
(f) Engine Structure:
Minor design differences in casting details can provide small advantages. It has been particularly true where attention has been paid to the parts of the casting to which covers have been attached.
It is stressed that where a major noise reduction is needed, the optimum design have been selected together with a radical structure redesign, that is, every possible detail must be attended to.
An essential prerequisite to the design of diesel engines for low noise has been to know how much of the noise gets radiated by individual parts of the engine. The largest noise procedures can then be examined, first ensuring that a maximum noise reduction is achieved for the minimum effort.
A simple covering technique using lead sheet lined with sound-absorbing material makes the measurement of the noise radiating characteristics by an individual surface or component of the engine to be evaluated. All the outer surfaces of engine have been first shield with 1.5 mm thick lead sheet lined with 25 mm thick Fiberglass wool.
In order to assess the contribution from a particular surface the lead shielding from that surface have been then removed and the resultant noise spectrum compared with both the spectra of the normal and fully shielded engine.
As an addition to the lead covering techniques the average surface vibration technique can be used. It has been particularly useful where interference to surface vibration (mass loading of thin covers by lead shielding) or lower frequencies have been of interest as well as a check on the lead shielding used.
The fundamental relationship between the direct radiated acoustic power of a surface Srad of area average mean square velocity <u2 (f)> over the surface has been as follows:
Wrad(f) = Pc Srad δrad(f) <u2> (f)
p = density of air
c = velocity of sound in air
δrad(f) = radiation ratio at f Hz.
The sound pressure measured on a spherical surface Strav around the engine would be-
P(f) = (pc)2<u2> Srad/Strav) δrad (f).
Thus, a knowledge of the surface area of the radiating component, the surface area of microphone transverse or position, the area average mean square vibration velocity of the surface, and the radiation ratio of the surface makes the sound power or pressure level contribution of the surface to be obtained as a function of frequency or overall level.
In this way a noise balance equivalent to and complementing the lead-shielding technique could be made. At present, the radiation ratio values for the main load carrying surfaces have been well-understood, but the situation for small, lightweight covers has been not so well documented.
The actual block structure of an engine could be represented by a three-dimensional network of essential plates and rods which by finite element analysis technique enables the overall distortion pattern to be determined when the structure is loaded by typical combustion forces. Modifications can then be made to the various elements of this representation until a more desirable structural deformation is achieved.
At present, this exercise can only be performed in an existing engine structure as certain relevant testing is necessary so as to establish the base-line parameters of the model.
These tests are:
(a) Modal Analysis:
Due to damping at sliding surfaces in a running engine, mode shapes have been difficult to ascertain, and therefore a non-running procedure is adopted. A bare engine block has been suitably suspended and excited electro-dynamically. Vibration acceleration measurements are done over a grid on the internal and external surfaces of the block, and essential information of vibration amplitude and phase derived.
(b) Static Deformation of the Engine Block:
High pressure oil gets pumped into the cylinder of a suitably suspended block, and the resultant distortion of the block measured.
In order to interpret the dynamic relevance of these results this static definition pattern has been considered as a vibration phenomenon at zero frequency. In practice, the static deflected shape has been composed essentially of the first few block mode patterns determined from the modal excitation.
The modifications determined by this analytical method usually take the form of altering the load paths to the structure, and thus, it may only become necessary to modify certain parts of the engine such as bulkhead and crankcase ribbing layout, and block wall thickness and curvature.
Over a number of years some ten different experimental low-noise engine structures have been built embracing all the groups namely five IDI engines, two normally aspirated DI engines, one vee-form engine, one opposed piston two- stroke engine, and two high-output turbocharged DI engines. Many different principles of design have been explored.
(1) Skeleton frame, both cast and fabricated, clad with damped materials.
(2) Increased stiffness for the same weight using magnesium alloy.
(3) Reduction of area of cast outer surfaces and replacing with large non-stiff covers.
(4) Increased stiffness using bedplate, ladder frame, or split crankcase principles.
(5) Combinations in increased damping and stiffness.
(6) Partial enclosure of cast surfaces.
General conclusions have been hard to draw, but have been based on economic, production, and durability considerations a universal structure design with the following features is suggested:
(1) An engine having walls of normal thickness cast as flat as practicable with raised flanges at the periphery (top deck, sump flange, and block side edges) to which damped shields can be attached (by screwing, gluing or riveting).
(2) Rear gear location with heavy gear cover to which all ancillary equipment can get attached (fuel pump, compressor, power take-off etc.).
(3) Crankcase split at crankshaft axis; the lower portion comprising stiff integral bearing deadplate providing a rigid support to which the side shields and oil sump can be attached.
(4) Damped or isolated rocker cover and sump.
Although this combination of design principles could be applied to all engines it has been found to be particularly suited to those of larger size where small increased costs and weight are more acceptable. For the smaller engines, particularly on the IDI range, completely new thinking is needed.
There is experimental evidence that lighter weight structures of reduced stiffness could well offer a solution, and consequently studies are being made at the ISVR on these lines to produce a light-weight the passenger can diesel.
Traffic Noise Reduction as a Function of Reducing Noise Levels of Vehicles:
Traffic noise is related to characteristics of the roadway. Of course, other factors in community noise control have been the capability and the feasibility of reducing vehicular noise by designing noise control features into newly produced vehicles and by prohibition of deliberate vehicular modifications that produce increased noise.
If these measures have been taken, what quantitative improvement may we expect along the streets and highways in a community? General Motors engineers have developed a simplified mathematical modelling technique for predicting roadside noise from freely flowing traffic in order to assess the effectiveness of various vehicle-noise reduction strategies. By entering sound level data into the model, which represent various design possibilities relative to vehicular noise control hardware, changes in urban-street Ldn can be projected.
An example, represents the reduction in Ldn that can be expected if:
1. A statistically derived operational sound pressure level of 70.5 dBA at 50 ft (15.2 m) for automobiles and light trucks has been selected as a future low- speed value. This regards the elimination of higher vehicular sound levels resulting from exhaust system modification by owners and from poorly maintained exhaust systems.
2. Medium and heavy-truck low-speed sound levels do not exceed a standard of 83 dBA.
3. Tyre noise from cars and trucks has been idealized at the level consistent with best-known present technology.
4. The existing fleet of trucks has been modified as needed to meet a low-speed pass by standard of 86 dBA.
5. The density of vehicles on the national street and highway system increases in proportion to the slope the top line on the chart, as an extreme.
Of course, the reduction of Ldn could be further based downward by choosing lower operational vehicular noise standards, but the estimate of incremental cost must then be developed for the attainment of each characteristics new vehicular sound level proposed.
The results of successive Ldn reductions and their associated costs may then be compared to the population-exposed goals discussed in relation to the values. Studies in the field of cost for increasing these benefits have depicted a sharp rise in the progression, beyond which the costs rise exponentially. It is not surprising that a point of diminishing returns exists in this relationship; social planners should not attempt to derive the benefit factors beyond that point.
Vehicular Component Hardware Development for Control of Engine-Related Truck Noise:
It immediately becomes obvious to anyone who makes a traffic-noise survey that aside from modified automobiles and motorcycles, the major source of noise on the streets is the truck. It has been largely because truck engines are ordinarily used at near-full power output for a much higher proportion of operating time than are automobiles or light trucks with gross vehicle weight ratings (GVWR) of less than 10,000 lbs (4536 kg).
This higher power usage from the engine arises out of the fact that medium and heavy trucks have been relatively low power to weight-type vehicles as compared to cars and light trucks. Thus, to start the load of the truck and to accelerate it, the engine is usually operated near its governed speed and power rating regardless of whether the truck is gasoline-powered or diesel-powered.
Ordinarily, the sound one hears from truck operating at speeds of 35 mph (56 km/hr) or less is engine related, and not tyre-related. The main contributors to engine-related noise have been the engine cooling system fan, the engine exhaust, the internal mechanical reciprocating and rotating engine parts, and the ringing of the engine block as a product of combustion gas pressure pulses within the cylinders of the engine. The latter phenomenon is more apparent in the case of the diesel engine than in the gasoline engine.
If the truck engine noise has been controlled to levels consistent with Interstate Motor Carrier Noise Regulation, then at cruise speeds of 50.55 mph (80.88 km/hr), tyre noise becomes a most significant contributor to total pass by noise. Note that the truck manufacturers’ responsibility has been involved with low-speed high-power-output, engine-related noise.
To qualify the matter of who bears responsibility for truck tyre noise, it must be understood that truck manufacturers furnish only original equipment tyres, and do not control the aftermarket in tyres. Of all truck tyres being used in service at any point of time, it has been estimated that 75 per cent of them have been the product of after-market purchases.
Of the remaining 25 per cent of tyres on the road that are original equipment truck tyres, truck purchasers either specify or supply their own choice of tyre for the truck in 25 to 35 per cent of the cases. Finally, in the case of semi-trailer rigs, the truck tractor manufacturer is having on control at all over the tyres supplied for trailers. For all these reasons, the truck tyre manufacturer is responsible for the noise performance of tyres, not the truck manufacturer.
The truck engine-cooling-system engineer is having a pair of option in terms of the methodology that may be used to control fan noise. The first approach has been to attempt adjust the various cooling-system design parameters so that the fan tip speed will be minimized, in as much as fan noise is a function of tip speed.
The best combination has been ordinarily the use of a large surface area liquid coolant radiator with a large diameter low-rotational speed fan. If there happens to be insufficient space in the engine compartment for that combination, there may be room for a close-fitting fan shroud with sufficient axial length to improve flow conditions for the fan.
If cooling system optimization has been not possible in the degree needed to sufficiently reduce fan noise, the second option is to drive the fan through a thermally actuated clutch. The truck accommodates this concept by the fact that ram airflow has been of sufficient volume to satisfy medium to high vehicle- speed cooling requirements, even in quite warm air temperatures.
The fan would be automatically cycled to the “on” condition when critical coolant temperatures get exceeded, and the fan will get automatically declutched when the coolant temperatures falls below the critical level. In the likely event that the fan would engage during an engine idle period is the result of “hot soak” after a high temperature run, it has been of little sequence from the standpoint of noise because of the relatively low rotational rate of the fan at engine idle.
Thus, not counting engagement at engine idle, tests reveal that statistical expectancy of “fan on” time could be as low as 1 per cent and as high as 10 per cent. This of course, implies that fan noise would be radiated less than 10 per cent of the time.
Exhaust noise attenuation must take into account three aspects of noise generation resulting from the release of high-temperature gases under the pressure conditions existing within the engine cylinders due to combustion of the fuel:
1. Gas-pressure pulse propagation through the exhaust-valve exist route from the interior of the cylinders.
2. High velocity viscous flow of expanding gas from the exhaust-pipe outlet.
3. Mechanical ringing of the exhaust piping and the muffler body, sometimes known as “shell noise”.
Exhaust pulse noise has been primarily attenuated through the use of chambered mufflers having reactive and resistive elements that decrease the amplitude of the transmitted gas pressure pulses. The ability of mufflers to accomplish this has been quite largely a function of correctly proportioned chamber volumes, taking into consideration that excessive back pressure on the engine has been intolerable.
The exhaust-system designer is not having great freedom to use high-volume devices but rather is having quite specific space restraints placed upon the system design. If the exhaust noise situation remains critical after the best muffler has been applied, the usual alternative has been to split the exhaust gasflow and to install a second muffler. Of course, dual exhaust systems add weight and cost.
Viscous flow noise could be usually controlled by reducing gas velocity through the use of large-diameter exhaust pipes. Shell noise radiation could be controlled through the use of double-wall exhaust-system pipes constructed so that the two layers of steel attenuate the transmitted sound by interfacial friction.
It may be tend to dampen exhaust-pipe ringing. The use of large diameter double wall pipes adds weight and cost to the truck. In addition, the mufflers must be wrapped with some type of absorption material to dampen muffler-body ringing, and a protective layer of sheet must be used over the absorption material.
During the combustion process in diesel-cycle engines and otto-cycle engines, a high rate of gas-pressure has been rise followed by a rapid pressure-decay rate. These pulses excite the many natural frequencies of the engine structure and cause it to ring. This is termed as combustion noise, and, as it turns out, this source is more difficult to control than the other major sources. Much work has been done particularly to understand and to control diesel-engine noise, and research is continuing as more sophisticated analytical instrumentation becomes available.
Diesel-engine manufacturers have studied the noise control possibilities of acoustic treatment of sheep-metal engine panels like rocker-box covers. Commercial noise control packages attempting to exploit such treatment have been available, but they are far from a total answer to engine noise.
When it is needed for engine noise to be further reduced, engineers have resorted to designing engine and transmission enclosures, which must be built into truck cab and frame subassembly. These enclosures must be lined with acoustic absorption material and must be designed so that mating surfaces fit perfectly to provide a high integrity absorbing barrier.
Also, the enclosure should be designed for quick removal for engine and transmission accessibility when normal maintenance has been. Such enclosures may be envisioned as the counterpart of an aircraft engine nacelle and, in fact, quasi-aircraft design practice will be required if very low engine-noise levels are needed.
Development engineers have reported serious disabilities with the use of such enclosures, in addition to the expensive design inherent in their application. The enclosures have been enclosing a higher level of engine cooling difficulty because truck engine compartments, as we known them, have been quite open to cooling airflow axially as well as downward toward the roadway. In order to significantly increase cooling capacity, disproportionately large engine compartments are required to accommodate large radiators, or else disproportionately small engines must be used with existing compartment envelopes.
Full-scale model building and testing show that, to attain satisfactory engine cooling, either cargo space will be lost due to engine compartments-space claims, or prime-mover power capability will be lost due to use of smaller engines. It is contrary to best design practice for trucks in a period of petroleum shortages.
Furthermore, because of the fire hazard that arises from accumulations of engine oil and fuel in the acoustic absorption material, particularly around the bottom of the enclosure, this engine enclosure approach to noise control must be considered marginal at best.
Taken together with the maintenance difficulties inherent due to inaccessibility of the engine and transmission, the conclusion is that engine enclosures will change the functional character of the transport truck in a manner that is unacceptable.
Development work has indicated that, expensive as it may be, a somewhat less severe application of engine enclosures may reduce the overall low-speed, full power noise of a heavy diesel truck to the 80 dBA level, measured at 50 ft (152 m). A significant reduction of noise is attainable in the national community with the use of trucks, related at 83 dBA.
The impact of trucks controlled to that level has been found to be more significant to the spectator when one realizes that, in the recent past, uncontrolled trucks in commercial use could generate sound level in excess of 93 dBA at 50 ft (15.2 m) under low-speed, full power conditions.
Noise Control of Existing Truck Fleet:
It is recognized that one of the most effective measures that can be taken to reduce urban highway and street noise has been to control the existing fleet to past model trucks, considering that new trucks will be manufactured with appropriate noise control features.
A level of 90 dBA measured at 50 ft (15.2 m) has been established as a maximum sound level for these vehicles when operated at speeds in excess of 35 mph (56 km/hr). The regulation is written to control the operational characteristics of trucks in service and not the manufacture of new trucks. The truckers must institute periodic maintenance of noise control systems to assure effective community noise reduction.
Relating Light-Vehicle Sound Level Ratings to Actual Operation Levels:
Engine-related noise of passenger cars and light trucks must be understood in an entirely different context than truck-engine noise. These light trucks and cars in general much lighter in proportion to their engine-power ratings than are medium and heavy trucks. The light vehicles have been capable of being accelerated in traffic with the use of proportionately less power and thus they generate much less noise.
Notice in Table 1 that during non-highway driving (that is, along urban streets and arterials), light vehicles are accelerating about 15% of the time. In Table 2, we can see that the time spent accelerating light vehicles in excess of 0.25g is extremely small, certainly less than 0.5% of the time.
In a separate investigation, GM engineers have established that the first row of light vehicles accelerating from stoplight situations in fast-moving traffic consistently move away so that they travel in the first 100 ft (30.4 m) in 5 seconds. While the actual acceleration rate in that situation is not uniform, it is equivalent to a calculated average rate of 0.25 g. From these data, we may predict noise levels of accelerating vehicles in traffic by measuring the sound level of light vehicles during the performance of such controlled “urban accelerations” on the proving ground strip.
The sound levels given for wide-open-throttle acceleration are conditioned by the term average. The reader must not construe the values shown as absolute sound levels for a given car or class because considerable variability exists, even between automobiles manufactured to a set of identical specifications.
This is true of light, medium, and heavy trucks as well. If the single highest value is thrown out (defective muffler), the range of variability is 5.5 dB. The stand deviation is 0.81 dB. This means that the designer/ manufacturer must target the sound level design point some 2 to 3 dB below the sound level specification if it is required that production vehicles must not exceed the specification.
This explains why, the averages of the wide-open-throttle sound levels fall significantly below the upper limit specification of 80 dBA. Any lowering of such a specification implies that a redesign of the vehicle must take place to maintain the mean sound level some to 3 dB below the new specification.
Noise Control Hardware Development of Automobiles and Light Trucks:
The components that have been main contributors to the low-speed full-power noise level of light vehicles are the engine-cooling fan, engine exhaust, engine structure vibration, and the combustion air-taken system. High speed noise has been largely a function of tyre noise. Cooling system fan noise has been controlled by refinement of the aerodynamic design of the cooling-system component for trucks.
If the vehicle is air conditioned, recent practice is to equip the car or truck with a viscous drive fan clutch. The units act as a fluid coupling under high-temperature conditions, driving the fan to cool the engine. At reduced ambient temperatures, the fluid has been internally diverted, and the coupling “declutches”, no longer driving the fan.
Furthermore, this type of viscous drive is usually designed as a torque limited coupling. This implies that even when the fan is thermally engaged, its driven speed gets limited to a value less than the input speed of the driven pulley. Since low-vehicle speed cooling of the engine requires greater fan-generated airflow volume than high-speed cooling, then considering the availability of supplemental ram air, the speed limiting feature of the viscous fan drive is attractive in preventing excessive fan speed, and thus noise.
Other devices such as “flex” fans have been developed, which are molded from synthetic material that allows the blades to flatten out at high speeds under air load. This provides partial control of fan noise at lower cost but is not as effective from the standpoint of noise control as the viscous drive.
The same general technology used in truck exhaust systems is applicable, to light vehicles, except that space is even less available.
Under the full-throttle-acceleration newer model automobiles and light trucks, and exterior noise specification 80 dBA is not uncommon.
To reduce the engine-related sound level of automobiles and light trucks below those levels would appear to require the following:
1. Some provision for shielding exhaust systems to reduce radiated noise for most vehicles.
2. Fan noise reduction by means of the clutch controlled fan.
3. Additional air injection reaction pump silencing on vehicles with air pumps.
Also, some vehicles will require air-intake noise silencing. A number of other vehicles may require considerably more control for engine noise or other mechanical noise sources/Of course, this is even more apparent at speeds greater 35 mph (56 km/hr), since tyre noise becomes increasingly dominant compared to engine-related noise.