NOISE METHODS OF ANALYSIS
1.0 NOISE DESCRIPTORS AND EFFECTS
Noise is generally described as unwanted sound. Unwanted sound can be
based on objective effects (hearing loss, damage to structures, etc.) or
subjective judgments (community annoyance). Noise analysis thus requires
a combination of physical measurement of sound with psycho- and socioacoustic
Launch vehicles generate two types of sound. One is engine noise, which
is continuous sound. The other is sonic booms, which are transient impulsive
sounds. These are quantified in different ways.
1.1 DESCRIPTORS OF CONTINUOUS SOUNDS
Measurement and perception of sound involves two basic physical characteristics:
amplitude and frequency. Amplitude is a measure of the strength of the sound
and is directly measured in terms of the pressure of a sound wave. Because
sound pressure varies in time, various types of pressure averages are usually
used. Frequency, commonly perceived as pitch, is the number of times per
second the sound causes air molecules to oscillate. Frequency is measured
in units of cycles per second, or Hertz (Hz).
Amplitude. The loudest sounds the human ear can comfortably hear
have acoustic energy one trillion times the acoustic energy of sounds the
ear can barely detect. Because of this vast range, attempts to represent
sound amplitude by pressure are generally unwieldy. Sound is therefore usually
represented on a logarithmic scale with a unit called the decibel (dB).
Sound on the decibel scale is referred to as a sound level. The threshold
of human hearing is approximately 0 dB, and the threshold of discomfort
or pain is around 120 dB.
The difference in dB between two sounds represents the ratio of those
two sounds. Because human senses tend to be proportional (i.e., detect whether
one sound is twice as big as another) rather than absolute (i.e., detect
whether one sound is a given number of pressure units bigger than another),
the decibel scale tends to correlate linearly with human response.
Frequency. The normal human ear can hear frequencies from about
20 Hz to about 15,000 or 20,000 Hz. It is most sensitive to sounds in the
1,000 to 4,000 Hz range. When measuring community response to noise, it
is common to adjust the frequency content of the measured sound to correspond
to the frequency sensitivity of the human ear. This adjustment is called
A-weighting (American National Standards Institute, 1988). Sound levels
that have been so adjusted are referred to as A-weighted sound levels. The
amplitude of A-weighted sound levels is measured in dB. It is common for
some noise analysts to denote the unit of A-weighted sounds by dBA or dB(A).
As long as the use of A-weighting is understood, there is no difference
between dB, dBA or dB(A). It is only important that the use of A-weighting
be made clear. It is common to use the term A-weighted sound pressure level
(AWSPL) to refer to A-weighted sounds.
For analysis of damage to structures by sound, it is common not to apply
any frequency weighting. Such overall sound levels are measured in dB and
are often referred to as overall sound pressure levels (OASPL or OSPL).
C-weighting (American National Standards Institute, 1988) is sometimes
applied to sound. This is a frequency weighting that is flat over the range
of human hearing (about 20 Hz to 20,000 Hz) and rolls off above and below
that range. C-weighted sound levels are often used for analysis of high-amplitude
impulsive noise, where adverse impact is influenced by rattle of buildings.
Time Averaging. Sound pressure of a continuous sound varies greatly
with time, so it is customary to deal with sound levels that represent averages
over time. Levels presented as instantaneous (i.e., as might be read from
the dial of a sound level meter), are based on averages of sound energy
over either 1/8 second (fast) or one second (slow). The formal definitions
of fast and slow levels are somewhat complex, with details that are important
to the makers and users of instrumentation. They may, however, be thought
of as levels corresponding to the root-mean-square sound pressure measured
over the 1/8-second or 1-second periods.
The most common uses of the fast or slow sound level in environmental
analysis is in the discussion of the maximum sound level that occurs from
the action, and in discussions of typical sound levels. Figure A-1 is a
chart of sound levels from typical sounds.
Assessment of cumulative noise impact requires average levels over periods
longer than just the fast or slow times. The sound exposure level (SEL)
sums the total sound energy over a noise event. Mathematically, the mean
square sound pressure is computed over the duration of the event, then multiplied
by the duration in seconds, and the resultant product is turned into a sound
level. SEL is sometimes described as the level which, occurring for one
second, would have the same sound energy as the actual event.
Note that SEL is a composite metric that combines both the amplitude
of a sound and its duration. It is a better measure of noise impact than
the maximum sound level alone, since it accounts for duration. Long sounds
are more intrusive than short sounds of equal level, and it has been well
established that SEL provides a good measure of this effect.
SEL can be computed for A- or C-weighted levels, and the results denoted
ASEL or CSEL. It can also be computed for unweighted (overall) sound levels,
with a corresponding designation.
For longer periods of time, total sound is represented by the equivalent
continuous sound pressure level (Leq). Leq is the average sound level over
some time period (often an hour or a day, but any explicit time span can
be specified), with the averaging being done on the same energy basis as
used for SEL. SEL and Leq are closely related, differing by (a) whether
they are applied over a specific time period or over an event, and (b) whether
the duration of the event is included or divided out.
Just as SEL has proven to be a good measure of the noise impact of a single event, Leq has been established to be a good measure of the impact of a series of events during a given time period. Also, while Leq is defined as an average, it is effectively a sum over that time period and is thus a measure of the cumulative impact of noise.
Noise tends to be more intrusive at night than during the day. This effect
is accounted for by applying a 10-dB penalty to events that occur after
10 PM and before 7 AM. If Leq is computed over a 24-hour period with this
nighttime penalty applied, the result is the day-night average sound level
(Ldn or DNL). Ldn is the community noise metric recommended by the U.S.
Environmental Protection Agency (U.S. Environmental Protection Agency, 1972)
and has been adopted by most federal agencies (Federal Interagency Committee
on Noise, 1992). It has been well established that Ldn correlates well with
community response to noise (Schultz, 1978; Finegold et al., 1994).
The state of California quantifies noise by Community Noise Exposure
Level (CNEL). This metric is similar to Ldn except that a penalty of 5 dB
is applied to sounds in the evening, after 7:00 p.m. and before 10:00 p.m.
It was noted earlier that, for impulsive sounds, C-weighting is more
appropriate than A-weighting. The day-night average sound level can be computed
for C-weighted noise, and is denoted LCdn or CDNL. This procedure has been
standardized, and impact interpretive criteria similar to those for Ldn
have been developed (CHABA, 1981).
1.2 DESCRIPTORS OF SONIC BOOMS
Figure A-2 shows time histories (pressure versus time) for the two types
of sonic boom signatures generated by launch vehicles: N-wave carpet booms
and U-wave focus booms. Each consists of a pair of shock waves connected
by a linear expansion (N-wave) or a U-shaped curve (U-wave). Each type of
boom is well described by its peak overpressure in pounds per square foot
(psf), and its duration in milliseconds (msec). Duration tends to have a
minor effect on impact, so the peak pressure is all that is normally required.
For assessment of impact via LCdn as discussed in Section 1.0, the peak
pressure is related in a simple way to CSEL, from which LCdn can be constructed.
The peak pressure P (psf) is converted to the peak level (Lpk) dB by the
Lpk = 127.6 + 20 log10 P (A-1)
CSEL is then given by Plotkin (1993):
CSEL = Lpk - 26 (N-wave) (A-2)
CSEL = Lpk - 29 (U-wave) (A-3)
2.0 NOISE EFFECTS
Studies of community annoyance to numerous types of environmental noise
show that Ldn is the best measure of impact. Schultz (1978) showed a consistent
relationship between Ldn and annoyance. This relationship, referred to as
the ìSchultz curveî, has been reaffirmed and updated over the
years (Fidell et al., 1991; Finegold et al., 1994). Figure A-3 shows the
current version of the Schultz curve.
Some time ago Ldn of 55 dB or less had been identified as a threshold
below which adverse impacts to noise are not expected (U.S. Environmental
Protection Agency, 1972). It can be seen from Figure A-3 that this is a
region where a small percentage of people are highly annoyed. Ldn of 65
dB is widely accepted as a level above which some adverse impact should
be expected (Federal Interagency Committee on Noise, 1992), and it is seen
from Figure A-3 that about 15 percent of people are highly annoyed at that
A limitation of the Schultz curve is that it is based on long-term exposure
to noise. The proposed action is for a single launch. Therefore, analysis
in the current study examines this on a single-event basis.
2.2 SPEECH INTERFERENCE
Conversational speech is in the 60 to 65 dB range, and interference with
this can occur when noise enters or exceeds this range. Speech interference
is one of the primary causes of annoyance. The Schultz curve incorporates
the aggregate effect of speech interference on noise impact.
Because only one launch is planned, and noise would last for only a few
minutes, speech interference is not expected to be a significant impact.
2.3 SLEEP INTERFERENCE
Sleep interference is commonly believed to be a significant noise impact.
The 10-dB nighttime penalty in Ldn is based primarily on sleep interference.
Recent studies, however, show that sleep interference is much less than
had been previously believed (Pearsons et al., 1989; Ollerhead, 1992).
Traditional studies of sleep disturbance indicate that interference can
occur at levels as low as 45 dB. Data indicates that at indoor SEL of 70
dB, about 20 percent of people will awaken (Federal Interagency Committee
on Noise, 1992). Assuming a nominal outdoor-to-indoor noise reduction of
20 dB, these correspond to outdoor sound exposure levels of 65 dB and 90
dB, respectively. Note that the awakening threshold is comparable to the
threshold of outdoor speech interference.
2.4 TASK INTERFERENCE
Due to startle effects, some task interference may occur to sonic booms.
High levels of rocket noise may cause some task interference close to the
launch sites. It is difficult to estimate degrees of task interference,
since this is highly dependent on specific tasks. Startle from sonic booms
is often stated as a concern, but there are no credible reported incidents
of harm from sonic boom startle. Task interference from rocket noise is
expected to occur at higher levels than speech interference.
2.5 HEARING LOSS
Federal OSHA guidelines (Title 29 CFR 1910.95) specify maximum noise
levels to which workers may be exposed on a regular basis without hearing
protection. Pertinent limits are a maximum of 115 dBA for up to 15 minutes
per day, and unweighted impulsive noise of up to 140 dB. Exceeding these
levels on a daily basis over a working career is likely to lead to hearing
impairment. These levels are conservative for evaluating potential adverse
effects from occasional noise events.
Nonauditory effects of long-term noise exposure, where noise may act
as a risk factor, have never been found at levels below federal guidelines
to protect against hearing loss. Most studies attempting to clarify such
health effects found that noise exposure levels established for hearing
protection will also protect against nonauditory health effects (von Gierke,
1990). There are some studies in the literature that claim adverse effects
at lower levels, but these results have generally not been reproducible.
2.7.1 Launch Noise
Damage to buildings and structures from noise is generally caused by
low-frequency sounds. The probability of structural damage claims has been
found to be proportional to the intensity of the low-frequency sound. Damage
claim experience (Guest and Sloane, 1972) suggests one claim in 10,000 households
is expected at a level of 103 dB, one in 1,000 households at 111 dB, and
one in 100 households at 119 dB.
Figure A-4 shows criteria for damage to residential structures (Sutherland,
1968), and compares them to launch noise spectra that could occur a few
kilometers from the launch point of a medium (300,000 to 500,000 pound thrust)
rocket. These data show that noise-induced damage to off-base property would
typically be very minimal.
2.7.2 Sonic Boom
Sonic booms are commonly associated with structural damage. Most damage
claims are for brittle objects, such as glass and plaster. Table A-1 summarizes
the threshold of damage that might be expected at various overpressures.
There is a large degree of variability in damage experience, and much damage
depends on the pre-existing condition of a structure. Breakage data for
glass, for example, spans a range of two to three orders of magnitude at
a given overpressure. While glass can suffer damage at low overpressures,
as shown in Table A-1, laboratory tests glass (White, 1972) have shown that-properly
installed window glass will not break at overpressure below 10 psf, even
when subjected to repeated booms.
The maximum sonic boom overpressures for the proposed launch will be
2.7 psf during launch (maximum focus boom) and 3.2 psf during entry, near
the water impact point. These are well below the threshold where structural
damage would be expected, were there structures in the vicinity.
3.0 NOISE MODELING
3.1 LAUNCH NOISE
On-pad and in-flight rocket noise was computed using the RNOISE model
(Plotkin et al., 1997). Rocket noise prediction via this model consists
of the following elements:
1. The total sound power output, spectral content and directivity is based on the in-flight noise model of Sutherland (1993). Noise emission is a function of thrust, nozzle exit gas velocity, nozzle exit diameter, and exhaust gas properties.
2. Propagation from the vehicle to the ground accounts for Doppler shift, absorption of sound by the atmosphere (American National Standards Institute, 1978), inverse square law spreading, and attenuation of sound by the ground (Chien and Soroka, 1980). A semi-hard ground surface (1,000 mks rayls) was assumed.
3. One-third spectral levels were computed at the ground, for every flight trajectory point, on a grid of 3721 points. ASEL and maximum A-weighted and overall sound levels were then derived from the results at each grid point.
The computed noise levels were then depicted as contours of equal level.
3.2 SONIC BOOM
Sonic boom was computed using the U.S. Air Forceís PCBoom3 software
(Plotkin, 1996). This is a full ray tracing model. Details of sonic boom
theory are presented by Plotkin (1989) and Maglieri and Plotkin (1991).
The specific approach to sonic boom modeling included the following elements:
1. Trajectories provided by the vehicle manufacturers were converted into PCBoom3 TRJ format using PCBoom3ís TRAJ2TRJ utility. This utility generated required higher derivatives, as well as converting file formats.
2. Vehicle F-functions were calculated using the method of Carlson (1978). Area distributions were obtained from vehicle drawings. The shape factors computed were used to obtain nominal N-wave F-functions.
3. The F-function associated with the plume was obtained by a combination of the Universal Plume Model (Jarvinen and Hill, 1970) and Tiegermanís (1975) hypersonic boom theory.
4. Ray tracing and signature evolution were computed by integration of the eiconal and Thomasís (1972) wave parameter method.
5. Focal zones were detected from the ray geometry, and focus signatures computed by applying Gill and Seebassís (1975) numerical solution.
The resultant sonic boom calculations were depicted as contours of constant overpressure (psf).
American National Standards Institute, 1988. Quantities and Procedures
for Description and Measurement of Environmental Sound, Part 1, ANSI
Carlson, H.W., 1978. Simplified Sonic Boom Prediction, NASA TP
CHABA, 1981. Assessment of Community Noise Response to High-Energy Impulsive
Sounds, Report of Working Group 84, Committee on Hearing, Bioacoustics and
Biomechanics, Assembly of Behavioral and Social Sciences, National Research
Council, National Academy of Sciences, Washington, DC.
Chien and Soroka, 1980. ìA Note on the Calculation of Sound Propagation
Along an Impedance Boundary,î J. Sound Vib. 69, pp. 340-343.
Federal Interagency Committee on Noise, 1992. Federal Agency Review
of Selected Airport Noise Analysis Issues, Federal Interagency Committee
on Noiseî, August 1992
Fidell, S., D.S. Barger, and T.J. Schultz, 1991. ìUpdating a Dosage-Effect
Relationship for the prevalence of Annoyance Due to General Transportation
Noiseî, Journal of the Acoustical Society of America, 89, pp.
221-223, January 1991
Finegold, L.S., C.S. Harris, and H.E. von Gierke, 1994. ìCommunity
Annoyance and Sleep Disturbance: Updated Criteria for Assessing the Impacts
of General Transportation Noise on Peopleî, Noise Control Engineering
Journal, Volume 42, Number 1, January-February 1994, pp. 25-30.
Gill, P.M., and A.R. Seebass, 1975. ìNonlinear Acoustic Behavior
at a Caustic: An Approximate Solutionî, AIAA Progress in Aeronautics
and Astronautics, H.J.T. Nagamatsu, Ed., MIT Press.
Guest, S. and R.M. Sloane, Jr. 1972. ìStructural Damage Claims
Resulting from Acoustic Environments Developed During Static Firing of Rocket
Engines,î presented at NASA Space Shuttle Technology Conference, San
Antonio, Texas, April. Published as NASA Technical Memo NASA TM X-2570,
Haber, J. and D. Nakaki, 1989. Sonic Boom Damage to Conventional Structures,
HSD-TR-89-001, April 1989
Jarvinen, P.O. and J.A.F. Hill, 1970. Universal Model for Underexpanded
Rocket Plumes in Hypersonic Flow, Proceedings of the 12th
JANNAF Liquid Meeting, November 1970.
Maglieri D.J. and K.J. Plotkin, 1991. ìSonic Boom,î Chapter
10, Aeroacoustics of Flight Vehicles, edited by H.H. Hubbard,
NASA RP 1258 Vol. 1, pp. 519-561.
Ollerhead, J.B., et al., 1992. Report of a Field Study of Aircraft
Noise and Sleep Disturbance. The Department of Transport, Department
of Safety Environment and Engineering. Civil Aviation Authority, London,
Pearsons, K.S., D.S. Barber, and B.G. Tabachick, 1989. Analysis of
the Predictability of Noise-Induced Sleep Disturbance, HSD-TR-89-029,
Plotkin, K.J., 1989. ìReview of Sonic Boom Theory,î AIAA
Plotkin, K.J., 1993. ìSonic Boom Focal Zones from Tactical Aircraft
Maneuvers, Journal of Aircraft, Volume 30, Number 1, January-February
Plotkin, K.J., 1996. PCBoom3 Sonic Boom Prediction Model: Version
1.0c, Wyle Research Report WR 95-22C, May 1996
Plotkin, K.J., L.C. Sutherland, and M. Moudou, 1997. Prediction of
Rocket Noise Footprints During Boost Phase, AIAA 97-1660, May 1997
Schultz, T.J., 1978. ìSynthesis of Social Surveys on Noise Annoyanceî,
Journal of the Acoustical Society of America, 64, pp. 377-405,
Sutherland, L.C., 1968. Sonic and Vibration Environments for Ground
Facilities - A Design Manual. Wyle Laboratories Research Report WR68-2,
Sutherland, L.C., 1993. Progress and Problems in Rocket Noise Prediction
for Ground Facilities, AIAA 93-4383.
Tiegerman, B., 1975. ìSonic Booms of Drag-Dominated Hypersonic
Vehicles,î Ph.D. Thesis, Cornell University, August 1975
Thomas, C.L., 1972. Extrapolation of Sonic Boom Pressure Signatures
by the Waveform Parameter Method. NASA TN D-6832, June 1972
U.S. Environmental Protection Agency, 1972. Information on Levels
of Environmental Noise Requisite to Protect the Public Health and Welfare
With an Adequate Margin of Safety, U.S. Environmental Protection Agency
Report 550/9-74-004, March 1972.
von Gierke, H.R., 1990. The Noise-Induced Hearing Loss Problem,
National Institute of Health Consensus Development Conference on Noise and
Hearing Loss, Washington, DC, 22-24 January 1990
White, R., 1972. Effects of Repetitive Sonic Booms on Glass Breakage,
FAA Report FAA-RD-72-43, April 1972
Table A-1. Possible Damage to Structures From Sonic Booms
Sonic Boom Overpressure Nominal (psf)
0.5 - 2
|Cracks in plaster||Fine; extension of existing; more in ceilings; over door frames; between some plaster boards.|
|Cracks in glass||Rarely shattered; either partial or extension of existing.|
|Damage to roof||Slippage of existing loose tiles/slates; sometimes new cracking of old slates at nail hole.|
|Damage to outside walls||Existing cracks in stucco extended.|
|Bric-a-brac||Those carefully balanced or on edges can fall; fine glass, e.g., large goblets, can fall and break.|
|Other||Dust falls in chimneys.|
2 - 4
|Glass, plaster, roofs, ceilings||Failures show that would have been difficult to forecast in terms of their existing localized condition. Nominally in good condition.|
|Glass||Regular failures within a population of well-installed glass; industrial as well as domestic greenhouses.|
|Plaster||Partial ceiling collapse of good plaster; complete collapse of very new, incompletely cured, or very old plaster.|
|Roofs||High probability rate of failure in nominally good state, slurry-wash; some chance of failures in tiles on modern roofs; light roofs (bungalow) or large area can move bodily.|
|Walls (out)||Old, free standing, in fairly good condition can collapse.|
|Walls (in)||Inside ("Party") walls known to move at 10 psf.|
Greater Than 10
|Glass||Some good glass will fail regularly to sonic booms from the same direction. Glass with existing faults could shatter and fly. Large window frames move.|
|Plaster||Most plaster affected.|
|Ceilings||Plaster boards displaced by nail popping.|
|Roofs||Most slate/slurry roofs affected, some badly; large roofs having good tile can be affected; some roofs bodily displaced causing gale-end and will-plate cracks; domestic chimneys dislodged if not in good condition.|
|Walls||Internal party walls can move even if carrying fittings such as hand basins or taps; secondary damage due to water leakage.|
|Bric-a-brac||Some nominally secure items can fall; e.g., large pictures, especially if fixed to party walls.|
Source: Haber and Nakaki, 1989
|Join the GlobalSecurity.org mailing list|