784 

14 CFR Ch. I (1–1–19 Edition) 

Pt. 36, App. A 

noise generator must be determined by a 
method traceable to the U.S. National Insti-
tute of Standards and Technology or to an 
equivalent national standards laboratory as 
determined by the FAA. Changes in the rel-
ative output from the previous calibration at 
each one-third octave band may not exceed 
0.2 dB. The correction for frequency response 
must be reported to the FAA and applied to 
the measured one-third octave sound pres-
sure levels determined from the output of 
the analyzer. 

A36.3.9.6 The performance of switched at-

tenuators in the equipment used during 
noise certification measurements and cali-
bration must be checked within six months 
of each test series to ensure that the max-
imum error does not exceed 0.1 dB. 

A36.3.9.7 The sound pressure level produced 

in the cavity of the coupler of the sound cali-
brator must be calculated for the test envi-
ronmental conditions using the manufactur-
er’s supplied information on the influence of 
atmospheric air pressure and temperature. 
This sound pressure level is used to establish 
the acoustical sensitivity of the measure-
ment system. Within six months of each test 
series the output of the sound calibrator 
must be determined by a method traceable 
to the U.S. National Institute of Standards 
and Technology or to an equivalent national 
standards laboratory as determined by the 
FAA. Changes in output from the previous 
calibration must not exceed 0.2 dB. 

A36.3.9.8 Sufficient sound pressure level 

calibrations must be made during each test 
day to ensure that the acoustical sensitivity 
of the measurement system is known at the 
prevailing environmental conditions cor-
responding with each test series. The dif-
ference between the acoustical sensitivity 
levels recorded immediately before and im-
mediately after each test series on each day 
may not exceed 0.5 dB. The 0.5 dB limit ap-
plies after any atmospheric pressure correc-
tions have been determined for the cali-
brator output level. The arithmetic mean of 
the before and after measurements must be 
used to represent the acoustical sensitivity 
level of the measurement system for that 
test series. The calibration corrections must 
be reported to the FAA and applied to the 
measured one-third octave band sound pres-
sure levels determined from the output of 
the analyzer. 

A36.3.9.9 Each recording medium, such as a 

reel, cartridge, cassette, or diskette, must 
carry a sound pressure level calibration of at 
least 10 seconds duration at its beginning 
and end. 

A36.3.9.10 The free-field insertion loss of 

the windscreen for each one-third octave 
nominal midband frequency from 50 Hz to 10 
kHz inclusive must be determined with si-
nusoidal sound signals at the incidence an-
gles determined to be applicable for correc-
tion of directional response effects per sec-

tion A36.3.9.3. The interval between angles 
tested must not exceed 30 degrees. For a 
windscreen that is undamaged and 
uncontaminated, the insertion loss may be 
taken from manufacturer’s data. Alter-
natively, within six months of each test se-
ries the insertion loss of the windscreen may 
be determined by a method traceable to the 
U.S. National Institute of Standards and 
Technology or an equivalent national stand-
ards laboratory as determined by the FAA. 
Changes in the insertion loss from the pre-
vious calibration at each one-third-octave 
frequency band must not exceed 0.4 dB. The 
correction for the free-field insertion loss of 
the windscreen must be reported to the FAA 
and applied to the measured one-third octave 
sound pressure levels determined from the 
output of the analyzer. 

A36.3.10 Adjustments for ambient noise. 

A36.3.10.1 Ambient noise, including both an 

acoustical background and electrical noise of 
the measurement system, must be recorded 
for at least 10 seconds at the measurement 
points with the system gain set at the levels 
used for the aircraft noise measurements. 
Ambient noise must be representative of the 
acoustical background that exists during the 
flyover test run. The recorded aircraft noise 
data is acceptable only if the ambient noise 
levels, when analyzed in the same way, and 
quoted in PNL (see A36.4.1.3 (a)), are at least 
20 dB below the maximum PNL of the air-
craft. 

A36.3.10.2 Aircraft sound pressure levels 

within the 10 dB-down points (see A36.4.5.1) 
must exceed the mean ambient noise levels 
determined in section A36.3.10.1 by at least 3 
dB in each one-third octave band, or must be 
adjusted using a method approved by the 
FAA; one method is described in the current 
advisory circular for this part. 

Section A36.4 Calculation of Effective Perceived 

Noise Level From Measured Data 

A36.4.1 

General. 

A36.4.1.1 The basic element for noise cer-

tification criteria is the noise evaluation 
measure known as effective perceived noise 
level, EPNL, in units of EPNdB, which is a 
single number evaluator of the subjective ef-
fects of airplane noise on human beings. 
EPNL consists of instantaneous perceived 
noise level, PNL, corrected for spectral 
irregularities, and for duration. The spectral 
irregularity correction, called ‘‘tone correc-
tion factor’’, is made at each time increment 
for only the maximum tone. 

A36.4.1.2 Three basic physical properties of 

sound pressure must be measured: level, fre-
quency distribution, and time variation. To 
determine EPNL, the instantaneous sound 
pressure level in each of the 24 one-third oc-
tave bands is required for each 0.5 second in-
crement of time during the airplane noise 
measurement. 

VerDate Sep<11>2014 

12:50 Apr 30, 2019

Jkt 247046

PO 00000

Frm 00794

Fmt 8010

Sfmt 8002

Y:\SGML\247046.XXX

247046

spaschal on DSK3GDR082PROD with CFR