( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( )k23,sk23,LSPk,24LSP k1,-isk1,-iLSPk,iLSP k,3sk3,LSPk,4LSP k3,SPLk,3LSP +¢¢=¢¢
•
• +¢¢=¢¢
•
• +¢¢=¢¢ =¢¢
a. Explanation
Steps 4 through 7 specify methods to smooth one-third-octave-band spectra to obtain pseudobroadband levels that would have been present in the absence of tones.
b. Supplemental Information
None
c. Procedures
1. Data Adjustments for Background Noise: When the procedure presented in Appendix 3 of this Advisory Circular is used for correction for the effects of background noise, Steps 4 and 5 of this tone-correction procedure should be modified as follows: Step 4(3) - The “Last Good Band” (LGB) should be used in place of the highest-frequency band (i=24). Step 5 - A new slope, s’(25,k), should be calculated for the band beyond LGB as described for an imaginary 25-th band. This slope should be used in place of the slope derived from the actual level of the band beyond LGB. (See Supplementary Information under section A36.3.10.2(b)(1) for information on background noise correction procedures.)
(h) Step 8: Calculate the differences, F (i,k), between the original sound pressure level and the final background sound pressure level as follows:
( ) ( ) ( )F i k i k i k, ,= − ′′SPL SPL ,
and note only values equal to or greater than 1.5.
a. Explanation
Step 8 specifies a requirement to compare the differences between a derived sound pressure level and the original sound pressure level.
b. Supplemental Information
None
c. Procedures
None
(i) Step 9: For each of the relevant one-third octave bands (3 through 24), determine tone correction factors from the sound pressure level differences F (i,k) and Table A36-2.
500 Hz ≤ f ≤ 5 000 Hz Tone correction C. 5 4 3 2 1 < 500 Rz 1 > 5 000 Hz Level difference F, dB Level Tone Frequency difference correction f, Hz F, dB C, dB 50 ≤ 5 < 500 14½* ≤ F<3 F/3 - 4½ 3≤ F < 20 F/6 20 ≤ F 3% 500 ≤ $ ≤ 5 000 1½* ≤ F < 3 2 F/3 - 1 3≤ F < 20 F/3 20 ≤ F 6⅔
5.000 < 5 ≤ 10.000 1½* ≤ F<3 F/3 - ½ 3≤F<20 F/6 20 ≤ F 3⅓ * See Step 8, 4.3.1.
Table A36-2: Tone Correction Factor
a. Explanation Step 9 specifies a method of determining a tone correction factor for each tone.
b. Supplemental Information None
c. Procedures None
(i) Step 10: Designate the largest of the tone correction factors, determined in Step 9, as C(k).
(An example of the tone correction procedure is given in the current advisory circular of this part.
Tone-corrected perceived noise levels PNLT(k) must be determined by adding the C(k) values to corresponding PNL(k) values, that is:
( ) ( ) ( )PNLT PNLk k C k= +
For any i-th one-third octave band, at any k-th increment of time, for which the tone correction factor is suspected to result from something other than (or in addition to) an actual tone (or any spectral irregularity other than airplane noise), an additional analysis may be made using a filter with a bandwidth narrower than one-third of an octave. If the narrow band analysis corroborates these suspicions, then a revised value for the background sound pressure level SPL"(i,k), may be determined from the narrow band analysis and used to compute a revised tone correction factor for that particular one-third octave band. Other methods of rejecting spurious tone corrections may be approved.
a. Explanation
Step 10 specifies the method for identifying the largest tone correction factor associated with the spectrum and then adding this correction to PNL(k) to obtain PNLT(k). An example of tone correction calculation for a turbofan engine is presented in Table 4-2 of the appended TM. The additional text following Step 10 describes an allowance to eliminate tone-corrections that are attributable to sources or causes other than the test aircraft, including pseudotones.
Note: The term “background” in this case, refers the broadband noise level that would be present in the one-third octave band in the absence of the tone, and not to the ambient noise described in section A36.3.10.
b. Supplemental Information
None
c. Procedures
(1) Data Precision (after Calculation of Tone Correction Factor): At this point, the original sound pressure level resolution of .01 dB must be restored. Although the required precision of reported EPNL is 0.1 dB, all other intermediate calculations external to the tone correction process must maintain a precision of at least
0.01 dB.
(2) Identification of Pseudotones: Appendix 2 of the appended ICAO TM presents guidance material on methods for identifying pseudo-tones. Note that the use of ground plane or 10-meter microphones is supplemental to the required 4-foot (1.2-meter) microphones, and is allowed only for identification of frequency bands within which pseudotones occur, not for the determination of aircraft noise certification levels.
(3) Tone Correction Factor Adjustment: When tone-correction factors result from false or fictitious tones, recalculation is allowed using revised sound pressure level values (based on narrow-band analysis) of the smoothed spectral levels obtained in Step 7. Once the levels have been revised, the tone-correction factor must be recomputed for the revised one-third octave band spectrum. This recomputed maximum tone-correction factor must be applied, even if it occurs at or near the band associated with an artificial tone, and FAA approval must be obtained for the methodology used.
197. Section A36.4.3.2
The tone correction procedure will underestimate EPNL if an important tone is of a frequency such that it is recorded in two adjacent one-third octave bands. An applicant must demonstrate that either:
(a) No important tones are recorded in two adjacent one-third octave bands; or
(b) That if an important tone has occurred the tone correction has been adjusted to the value it would have had if the tone had been recorded fully in a single one-third octave band.
a. Explanation
None
b. Supplemental Information
(1) Band sharing is discussed under section A36.4.4.2.
c. Procedures
None
198. Section A36.4.4 Maximum tone-corrected perceived noise level
199. Section A36.4.4.1
The maximum tone-corrected perceived noise level, PNLTM, must be the maximum calculated value of the tone-corrected perceived noise level PNLT(k). It must be calculated using the procedure of section A36.4.3. To obtain a satisfactory noise time history, measurements must be made at 0.5 second time intervals.
Note 1: Figure A36-2 is an example of a flyover noise time history where the maximum value is clearly indicated.
Note 2: In the absence of a tone correction factor, PNLTM would equal PNLM.
Figure A36-2: Example of Perceived Noise Level Corrected for Tones as a Function of Aircraft Flyover Time
200. Section A36.4.4.2
After the value of PNLTM is obtained, the frequency band for the largest tone correction factor is identified for the two preceding and two succeeding 500 ms data samples. This is performed in order to identify the possibility of tone suppression at PNLTM by one-third octave band sharing of that tone. If the value of the tone correction factor C(k) for PNLTM is less than the average value of C(k) for the five consecutive time intervals, the average value of C(k) must be used to compute a new value for PNLTM.
a. Explanation
This section specifies the method of accounting for the effects of the suppression of tones occurring at or near the band edges of one-third-octave-band filters when determining a tone correction factor at the time of PNLTM.
b. Supplemental Information
(1) Band Sharing Adjustment Concept: The one-third-octave-band filtering process specified for analysis of airplane noise certification data may allow the tone-correction procedure presented in section A36.4.3 to under-predict a tone correction factor when the frequency of a tone is located at or near the edge of one or more one-third-octave bands. To account for this phenomenon, a band sharing adjustment is computed that takes advantage of the fact that, as a result of the Doppler Effect, a tone that is suppressed at PNLTM will probably appear normally in the spectra that occur before or after PNLTM. By averaging the tone correction factors calculated for the spectra within a 2-second period around PNLTM, the tone correction factor that would have occurred at PNLTM if it were not suppressed can be reasonably estimated.
c. Procedures
(1) Computation of Band Sharing Adjustment: Although part 36 refers to identification of the frequency bands in which maximum tone-corrections occur for the records near PNLTM, the presence or absence of band-sharing cannot be established merely by observing these frequencies. Even though the maximum tone that occurs in a one-third-octave band spectrum may not be related to the band of maximum tone-correction in the PNLTM spectrum, a related tone may still be present. Therefore, the average of the tone-corrections of all spectra within 1-second (five one-half second data records) of PNLTM must be used regardless of the bands in which maximum tones are found. If the band sharing adjustment is believed to result from effects other than band sharing, the applicant must demonstrate its absence for each event.
(2) Adjustment of PNLTM for Band Sharing: The band sharing adjustment must be computed before the determination of the 10 dB-down period and must be included in the reported PNLTM and EPNL values for the test condition data.
(3) Application of Band Sharing Adjustment for Simplified Procedure: When the simplified procedure is used to adjust data to reference conditions, the bandsharing adjustment, must be applied to the PNLTr at time of PNLTM before “delta 1" and EPNLr is calculated.
(4) Application of Band Sharing Adjustment for Integrated Procedure: When the integrated procedure is used to adjust data to reference conditions, a new band sharing adjustment must be calculated from the five data records for the one-third-octave-band spectra (adjusted to reference conditions) encompassing the PNLTr at time of PNLTM (this includes two data records prior to PNLTr at time of PNLTM, the PNLTr at time of PNLTM record, and two data records after PNLTr at time of PNLTM).
201. Section A36.4.5 Duration correction
202. Section A36.4.5.1
The duration correction factor D determined by the integration technique is defined by the expression:
PNLTMdt 2 t 1 t 10 PNLTantilogT1log10D -œœœßøŒŒŒºØ ò÷łöçŁæ= ÷łöçŁæ ÷łöçŁæ
where T is a normalizing time constant, PNLTM is the maximum value of PNLT, t(1) is the first point of time after which PNLT becomes greater than PNLTM-10, and t(2) is the point of time after which PNLT remains constantly less than PNLTM-10.
203. Section A36.4.5.2
Since PNLT is calculated from measured values of sound pressure level (SPL), there is no obvious equation for PNLT as a function of time. Consequently, the equation is to be rewritten with a summation sign instead of an integral sign as follows:
( ) PNLTM10 kPNLTlogt.antiT1log10D td 0k -œßøŒºØ ÷łöçŁæ= å=D D
where Dt is the length of the equal increments of time for which PNLT(k) is calculated and d is the time interval to the nearest 0.5s during which PNLT(k)remains greater or equal to PNLTM-10.
204. Section A36.4.5.3
To obtain a satisfactory history of the perceived noise level use one of the following:
(a) Half-second time intervals for Dt; or
(b) A shorter time interval with approved limits and constants.
205. Section A36.4.5.4
The following values for T and Dt must be used in calculating D in the equation given in section A36.4.5.2:
T = 10 s, and
Dt = 0.5s (or the approved sampling time interval).
Using these values, the equation for D becomes:
( ) 13PNLTM10 kPNLTlogantilog10D d2 0k --œßøŒºØ= å=
where d is the duration time defined by the points corresponding to the values PNLTM-10.
a. Explanation
None
b. Supplemental Information
(1) EPNL Equations: The equation for the duration correction, D, as shown in A36.4.5.4, is valid only for records of one-half second in length. The constant value 13 is used to normalize the one-half second values to the 10-second standard duration (10*log10 (10 seconds of ½ -second data records = 20 records) =13.01).
c. Procedures
None
206. Section A36.4.5.5
If in using the procedures given in section A36.4.5.2, the limits of PNLTM-10 fall between the calculated PNLT(k) values (the usual case), the PNLT(k) values defining the limits of the duration interval must be chosen from the PNLT(k) values closest to PNLTM-10. For those cases with more than one peak value of PNLT(k), the applicable limits must be chosen to yield the largest possible value for the duration time.
a. Explanation
This section specifies a requirement to calculate a duration correction, D, which is determined by means of integrating (summing) the PNLT values within 10 dB of PNLTM.
b. Supplemental Information
None
c. Procedures
(1) Identification of the First and Last Records Within the 10 dB-Down Period: When identifying the records that define the limits of the 10 dB-down period, those records having PNLT values closest to the actual value of PNLTM-10 dB must be used. As a result, the PNLT values for the PNLTM-10 dB points may not always be greater than or equal to PNLTM-10 dB.
(2) Calculation of the Duration Correction, D: When calculating the value of D, the full PNLT value and duration of each record within the 10 dB-down period must be included. Interpolation of discrete PNLT values to the exact point of PNLTM-10 dB is not allowed.
(3) Recorded Data: If recorded data do not encompass the entire 10 dB-down period, an EPNL cannot be calculated from those data, and the event must not be used for aircraft noise certification purposes.
207. Section A36.4.6 Effective perceived noise level
208. Section A36.4.6
The total subjective effect of an airplane noise event, designated effective perceived noise level, EPNL, is equal to the algebraic sum of the maximum value of the tone-corrected perceived noise level, PNLTM, and the duration correction D. That is:
EPNL = PNLTM + D
where PNLTM and D are calculated using the procedures given in sections A36.4.2, A36.4.3, A36.4.4 and A36.4.5.
209. Section A36.4.7 Mathematical formulation of noy tables
210. Section A36.4.7.1
The relationship between sound pressure level (SPL) and the logarithm of perceived noisiness is illustrated in Figure A36-3 and Table A36-3.
211. Section A36.4.7.2
The bases of the mathematical formulation are:
(a) The slopes (M(b), M(c), M(d) and M(e)) of the straight lines;
(b) The intercepts (SPL(b) and SPL(c)) of the lines on the SPL axis; and
(c) The coordinates of the discontinuities, SPL(a) and log n(a); SPL(d) and log n = -1.0; and SPL(e) and log n = log (0.3).
212. Section A36.4.7.3 Calculate noy values using the following equations:
(a) ( ) ( ) ( ) [ ] { }
SPL SPL antilog SPL SPL ≥ = − a n M c c
(b) ( ) ( ) ( ) ( ) [ ] { }
SPL SPL SPL antilog SPL SPL b a n M b b £ < = −
(c) ( ) ( ) ( ) ( ) [ ] { }
SPL SPL SPL antilog SPL SPL e b n M e e £ < = − 0 3 .
(d) ( ) ( ) ( ) ( ) [ ] { }
SPL SPL SPL antilog SPL SPL d e n M d d £ < = − 0 1 .
213. Section A36.4.7.4
Table A36-3 lists the values of the constants necessary to calculate perceived noisiness as a function of sound pressure level.
a. Explanation
This section specifies a requirement for using table A36-3 constants to calculate perceived noy values related to sound-pressure-level values.
b. Supplemental Information
None
c. Procedures
None
Siope = M(c) SPL(a), log n (a) Logarithm perceived noisiness, log n SPL(d) Slope - - M(b) SPL(C) SPL(b) Sound pressure level SPL М(e) Slope = SPL(e), log(0.3) M(d) Slope
- 1.0 Figure A36-3: Perceived noisiness as a function of sound pressure level Page 115
BAND SPL SPL SPL SPL SPL HZ (a) (b) (c) (d) (e) M(b) M(c) M(d) М(e) 1 50 91.0 64 52 49 55 0.043478 0.030103 0.079520 0.058098 63 85.9 60 51 44 51 0.040570 0.068160 80 87.3 56 49 39 46 0.036831 0.052288 4 100 79.9 53 47 34 42 0.059640 0.047534 5 125 79.8 51 46 30 39 0.035336 0.053013 0.043573 6 160 76.0 48 45 27 36 0.033333 7 200 74.0 46 43 24 33 0.040221 250 74.9 44 42 21 30 0.032051 0.037349 9 315 94.6 42 41 18 27 0.030675 0.030103 0.034859 10 400 40 40 16 25 0.030103 500 40 40 16 25 12 630 40 40 16 25 13 14 1 000 800 40 40 40 40 16 16 25 25 NOT APPLICABLE 0.053013 15 1 250 38 38 15 23 0.030103 0.059640 0.034859 16 1 600 34 34 12 21 0.029960 0.053013 0.040221 17 2 000 32 32 9 18 0.037349 18 2 500 30 30 15 0.047712 0.034859 19 3 150 29 29 4 14 20 4 000 29 29 5 14 0.053013 21 5 000 30 30 6 15 0.034859 22 6 300 00 31 31 10 17 0.029960 0.068160 0.037349 23 8 000 44.3 37 34 17 23 0.042285 0.029960 0.079520 24 10 000 50.7 41 37 21 29 0.059640 0.043573 Table A36-3: Constants for mathematically formulated noy values.
214. Section A36.5. Reporting of Data to the FAA
215. Section A36.5.1 General
216. Section A36.5.1.1
Data representing physical measurements and data used to make corrections to physical measurements must be recorded in an approved permanent form and appended to the record.
217. Section A36.5.1.2
All corrections must be reported to and approved by the FAA, including corrections to measurements for equipment response deviations.
218. Section A36.5.1.3
Applicants may be required to submit estimates of the individual errors inherent in each of the operations employed in obtaining the final data.
a. Explanation
None.
b. Supplemental Information
(1) Compliance Records: For compliance with section A36.5.1.1 through A36.5.1.3, all data measured during noise certification testing, including all time histories of physical measurements, noise recordings, instrument calibrations, etc., are to be recorded in permanent form and made available to the FAA for review, inspection, and approval. A common procedure is for the applicant to submit representative samples of test data for each noise measurement point and adjustments to measured data to permit the FAA to determine compliance with the regulation. The applicant may either: 1.) submit the complete test records along with the required corrections, or 2.) when approved by the FAA, the applicant may instead submit samples of test data along with the required data adjustments.
c. Procedures
(1) Test Records: All test records are to be submitted to the FAA for compliance review and approval. Adjustments to the measured data are also to be submitted to the FAA for review and approval. The FAA may require an applicant to submit records showing estimates of the magnitude of potential errors at each step in the measurement, analysis, and adjustment of data. Specific data-reporting requirements are identified in section A36.5.2.
219. Section A36.5.2 Data reporting
An applicant is required to submit a noise certification compliance report that includes the following.
a. Explanation
None.
b. Supplemental Information
(See section 36.1501(a))