Click here to login if you're an NAE Member
Recover Your Account Information
Author: William W. Lang and George C. Maling Jr.
Noise can be controlled at the source, along its path, or at the location of the receiver.
The Source-Path-Receiver Model
For a systems approach to the technological issues of noise reduction, the source-path-receiver model, which has been in use for many years, can provide a helpful starting point (Bolt and Ingard, 1957). This model enables engineers to treat each part of a system separately, thus facilitating identification of the technical challenges for that part.
Noise can be controlled at any stage of the model—by reducing it at the source, interfering with its transmission paths, and/or changing the characteristics of the receiver. The descriptors most frequently used in a noise control system are the levels of sound emission from the source, attenuation level of noise along the path(s), and the sound-pressure level at the location of the receiver (i.e., imission). A source is usually characterized by the total acoustic power it radiates into the environment and, less frequently, by the directivity of the sound radiation. Because a wide range of acoustic powers can be radiated by sources (and sound pressures at receiver locations), a logarithmic scale is used for each descriptor.
The sound power level, the logarithmic ratio of the acoustic power of the source and a reference power (10-12 W), can be conveniently expressed in bels. The acoustic power is often A-frequency-weighted to bring the level more in line with human perception (IEC, 2002). For practical reasons, most industries prefer to use the decibel (dB) for this ratio (1 bel equals 10 dB).
An A-weighted level frequently used to describe cumulative noise exposure is the day-night sound level, Ldn (the time-weighted average level over a 24-hour period, but with 10 dB added for nighttime hours [10:00 p.m. to 07:00 a.m.]). The abbreviations dBA or dB(A) are often used as the unit of an A-weighted level, even though the level, and not the decibel, is A-weighted.
At the receiver’s location, the sound-pressure level, also expressed in dB, is 20 times the logarithmic ratio of the root-mean-square sound pressure and a reference pressure (2x10-5 N/m2). Sound-pressure levels are also often A-frequency weighted.
The acoustic power radiated by a source is generally a very small fraction of the total electrical or mechanical power of the source. Shaw (1975) estimated this fraction to be 10-5 with a range of 10-3 to 10-7 for a wide variety of sources—from dishwashers to the airplanes of that era. A more recent illustration can be derived from a current European Union (EU) Directive on noise emissions from outdoor equipment (EU, 2005). The examples below are Stage II values, which became effective on January 3, 2006.
Example 1. A welding and power generator with an electrical rating of 10 kW must radiate no more than 100 dB of acoustic power. The ratio of the two powers is 10-6. This ratio is not very different from those of a wide range of electrical ratings.
Example 2. A compressor with a net installed power of 15 kW must radiate no more than 97 dB of acoustic power. The ratio of the two powers is 0.33?10-6. The ratio is smaller for net installed powers < 15kW, but not very different for powers > 15 kW.
Aerodynamic noise generated by turbulent flow or fluctuating lift forces is a common problem for aircraft, air-moving devices, and air-cooled machinery. Many sources generate aerodynamic noise, and, although once again the acoustic power radiated is a very small fraction of the power of the flow, aerodynamic noise creates significant problems in homes, offices, and communities.
Because the acoustic power radiated is small, even if the mechanical motion of a noise source is known, it is often very difficult to predict the level of sound radiation. The propagation of sound through a structure and subsequent radiation from the structure further complicate the situation.
Theoretically, sound waves outdoors spread geometrically at a rate of 6 dB per doubling of distance. However, this seldom occurs in the real world because of sound absorption at the ground surface, ground geometry, air turbulence, and wind and temperature gradients. Although propagation models have been developed, conditions are constantly changing, and predicting noise levels at a receiver outdoors is a complex undertaking.
Silencers and mufflers are common ways of attenuating noise along a path, indoors or outdoors, and their performance can be predicted or measured. However, these devices are frequently expensive add-ons that can raise the cost of a product.
For equipment used indoors, attenuation depends on the sound-absorptive properties of room surfaces, room geometry, and the scattering of sound from objects in the room. Thus, designing machines to ensure that a given sound-pressure level reaches the receiver can be challenging indoors as well.
Measuring sound-pressure levels, usually A-weighted, at receiver locations is a straightforward process. However, difficulties arise in correlating the physical magnitudes of the sound to the subjective and physiological reactions of the receiver. The human ear is, perhaps, the most variable receiver in its response to physical signals, because the principal effects of noise on people are both psychological and physiological.
Prolonged exposure to
high noise levels may have
The psychological effects of noise include annoyance, sleep disturbance, interference with communication, and adverse effects on learning, social behavior, job performance, and safety—all of which affect quality of life. However, people can adapt to many sounds. For example, they may become accustomed to sounds that were extremely disturbing when they were first heard. After hearing a sound over a long period of time, some people may find it tolerable, or even acceptable.
An example of this is someone who purchases a residential property adjacent to a busy highway. One of the motivating factors for the purchase is the reduced price of the property because of its exposure to highway noise; an equivalent property in a quiet neighborhood would be significantly more expensive. At first, the buyer may find the traffic noise annoying. But after a period of time, the negative psychological effect may fade as he or she becomes accustomed to the noise.
The ability of humans to adapt to noise, particularly when there are financial advantages in doing so, presents a major difficulty in determining metrics for community noise. It is difficult to determine physical parameters for annoyance, loudness, sleep disturbance, and other effects of noise that are subject to adaptation.
Prolonged exposure to high noise levels may have physiological effects, such as significant hearing loss caused by the atrophy of hair cells in the inner ear. Noise-induced hearing loss (NIHL) is well understood and is subject to international standardization (ISO, 1990). NIHL is frequently accompanied by tinnitus, a persistent ringing in the ears, which is sometimes severe, even for people with mild hearing impairments. Even relatively low levels of noise that are well below the threshold for causing hearing loss can have adverse effects on the auditory system—such as masking signals and alarms and interfering with speech.
Noise may also increase stress and damage the cardiovascular system. Such non-auditory effects are under investigation but have not been subject to international standardization.
Excessive noise in the workplace, which has been a problem since the industrial revolution, remains a problem despite years of effort. In a study of noise and military service in 2005, the Institute of Medicine provided an assessment of costs to the federal government for compensation for hearing loss incurred during military service. “At the end of 2004, the monthly compensation payments to veterans with hearing loss as their major form of disability represented an annualized cost of some $660 million . . . [W]ith tinnitus as the major disability . . . [the cost is] close to $190 million on an annualized basis” (IOM, 2005). Corresponding costs for civilian federal employees in 2001 were reported by the U.S. Army Center for Health Promotion and Preventive Medicine to be about $43.8 million (CHPPM, 2003). Thus the total cost for compensation to veterans and other federal employees is almost $1 billion per year. Similar figures for non-federal employees in American industry are not available.
The technological challenges to controlling noise in the workplace include (1) using available technology to solve as many of the problems of excessive noise as possible and (2) developing new technology for addressing the remaining problems. According to Bruce and Wood (2003), “The lack of clear direction at the national level is the reason this very solvable problem remains a challenge to our society . . . we lack the will to resolve this problem.”
The principal tactic for reducing surface-transportation noise is the noise barrier, a palliative measure that does nothing to reduce the level of noise radiated by vehicles in motion (i.e., the sources). The Federal Highway Administration (FHWA) has reported that, by the end of 2004, more than 3,500 km of noise barriers had been constructed by 45 state departments of transportation and the Commonwealth of Puerto Rico at a cost of more than $2.6 billion ($3.4 billion in 2004 dollars) (FHWA, 2006). Some roads have barriers on both sides, others on only one side.
The cost per square meter of a noise barrier depends on its height, construction materials, and terrain (Polcak, 2003). However, the effectiveness of roadside noise barriers is marginal. They provide a modest degree of shielding from traffic noise for residences immediately behind the barrier, but their effectiveness diminishes rapidly with distance from the barrier (Daigle, 1999).
At the present time, a low level of interior noise, particularly in a passenger vehicle, is an important sales feature. Thus market forces are driving the automotive industry to invest large sums in research and development to reduce interior vehicle noise.
No comparable efforts are being made to reduce exterior vehicle noise. At highway speeds, most exterior noise is produced by the interaction between tires and the road surface. Table 1 shows crossover speeds—the speeds at which road/tire interaction noise is equal to the power train noise—for different types and vintages of vehicles. At higher speeds, tire/road-surface interaction noise dominates.
A body of evidence now indicates that significant noise reductions can be achieved by reducing tire/road-surface noise (Donavan, 2005). The technological challenge is to engineer road surfaces and tires to reduce noise levels.
TABLE 1 Crossover Speeds for Various
Types of Vehicles