Enhanced Modeling of Aircraft Taxiway Noise, Volume 2: Aircraft Taxi Noise Database and Development Process (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

4-1 CHAPTER 4. Hybrid Methods for Estimating Turbofan (Jet) Aircraft Taxi Noise Levels There are three sources of aircraft noise level information that can provide a basis for estimating aircraft taxi noise levels, as follows: 1. INM NPD data – provides various flight noise levels as a function of power setting and distance for several classes of certified aircraft; 2. ANOPP Predictions – A NASA code which can predict flight noise levels based on engine performance predictions (NPSS code) and aircraft flight performance (NASA FLOPS code) and definitions of aircraft and engine geometry, cycle parameters, etc.; and 3. Empirical data – e.g., Madrid, IAD, etc., from airport survey measurements, corrected to some reference condition using the Wyle AAM code. Sufficient data was made available by Georgia Tech, to run predictions of taxi noise for five different aircraft configurations using the NASA ANOPP procedure. Taxi noise predictions were carried out over a range of engine thrust settings and distances from aircraft taxi path to microphone/observer for these configurations. The aircraft taxi predictions were made using ANOPP and EDS representations of the following aircraft:  Boeing B737-800 with CFM56-7B engines;  Boeing B777-200 with both GE90-90B and PW4098 engines;  Boeing B747-400 with PW4062 engines;  Boeing B767-300 with CF6-80C2 engines; and  Bombardier CRJ900 with CF34-8C5 engines. 4.1. Attempts to Utilize INM Flight NPD Data The first approach attempted was to utilize the INM flight NPD data to generate a relationship between noise and power setting (typically net corrected thrust per engine), and then extrapolate the results to typical taxi thrust levels, while applying a combination of ANOPP predictions and measured data to adjust the INM results. The INM NPD data were adjusted from the standard 160 Kts flight speed to the recommended “standard” taxi speed of 16 kts using the duration correction of 10 log (160/16) = 10 dB. Both departure and arrival INM NPD’s were created, and plots of SEL (dBA) vs. net corrected thrust (Fn/delta) developed for various distances, from 200 ft. to 25,000 ft. The INM NPDs were developed for the aircraft listed above for which ANOPP calculations had been performed. ANOPP was used to determine the individual noise contributions from the airframe and engine components. ANOPP results were added to the plots, as well as measured taxi noise data. Examples of these comparisons are shown in Figures 4-1, 4-2 for the B737-800 (200 ft and 1000 ft) and Figures 4 3, 4-4 for the B747-400 (200 ft and 1000 ft), respectively. The INM NPD data for approach and departure is depicted as red squares and green triangles, respectively. The empirical data (from Madrid or Wyle measurements) is indicated at the assumed thrust by a yellow triangle. The two techniques for subtracting airframe noise from the INM NPDs are indicated by cyan squares (Logarithmic subtraction) or orange squares (dB difference). Polynomial fits with extrapolations are also indicated in the graphics in Figures 4-1 through 4-4 and in Appendix E1. 1 ANOPP computations (engine only, under the taxi condition) are plotted in combination with the other data presented in Figures 4-1 and 4-2 and may be found in Appendix E. ANOPP computations are indicated by blue diamonds.

4-2 One additional point to note is that the INM data includes all aircraft noise sources, and so it was deemed necessary to attempt to extract airframe noise from the data using information obtained from the ANOPP predictions. The ANOPP data were computed both with and without airframe noise included, and the results were anti-logarithmically subtracted to estimate the airframe noise itself. The resulting airframe noise levels were then used to extract airframe noise from the INM projections. Two methods were used, logarithmic subtraction and differencing, i.e., taking the difference in decibels obtained from the ANOPP information and subtracting from the INM information. All these variants are shown to some extent on the example plots of Figures 4-1 through 4-4. It should be noted that plots like these were generated for all the ANOPP aircraft listed above, and for distances of 200 ft., 400 ft., 630 ft., 1000 ft. and 10,000 ft., and may be found in Appendix E. FIGURE 4-1 Taxi noise vs. thrust comparison, B737-800 for 16 knots at 200 ft. distance. 85 90 95 100 105 110 115 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 SE L, d B A Corrected Net Thrust Taxi Noise vs. Thrust Comparison - B737-800 16 Kts - 200 ft. INM NPD - Approach INM-NPD - Dep Madrid-AAM-MeasData INM App - no AFN (Log) INM Dep - no AFN (Log) INM App - no AFN (Diff) INM Dep - no AFN (Diff) Poly. (INM-NPD - Dep) Poly. (INM App - no AFN (Log)) Poly. (INM Dep - no AFN (Log)) Note: INM-NPD data corrected from 160 Kts to 16 Kts by adding 10 dBA

4-3   FIGURE 4-2 Taxi noise vs. thrust comparison, B737-800 for 16 knots at 1000 ft. distance. FIGURE 4-3 Taxi noise vs. thrust comparison, B747-400 for 16 knots at 200 ft. distance. 80 85 90 95 100 105 110 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 SE L, d B A Corrected Net Thrust Taxi Noise vs. Thrust Comparison - B737-800 16 Kts - 1000 ft. INM NPD - Approach INM-NPD - Dep Madrid-AAM-MeasData INM App - no AFN (Diff) INM App - no AFN (Log) INM Dep - no AFN (Log) INM Dep - no AFN (Diff) Poly. (INM-NPD - Dep) Poly. (INM Dep - no AFN (Log)) Note: INM-NPD data corrected from 160 Kts to 16 Kts by adding 10 dBA 95 100 105 110 115 120 125 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 SE L, d B A Corrected Net Thrust Taxi Noise vs. Thrust Comparison - B747-400 16 Kts - 200 ft. INM NPD - Approach INM-NPD - Dep Madrid-AAM-MeasData INM App - no AFN (Log) INM Dep - no AFN (Log) INM App - no AFN (Diff) INM Dep - no AFN (Diff) Poly. (INM-NPD - Dep) Note: INM-NPD data corrected from 160 Kts to 16 Kts by adding 10 dBA

4-4 FIGURE 4-4 Taxi noise vs. thrust comparison, B747-400 for 16 knots at 1000 ft. distance. It was concluded, having studied the results of all of these comparisons, that extrapolation of INM results, whether from the departure data or arrival data, and whether airframe noise was extracted by either method or not, does not in general provide a reliable method for estimating taxi noise, when compared to the measured data points. In addition, the trend or curve shapes obtained from the INM results do not agree with the ANOPP curve shapes of Taxi noise (SEL) vs. Taxi thrust. Based on the outcomes of the analysis presented in Figures 4-1 through 4-4 and Appendix E, it was decided that extrapolation of INM NPD data would not result in a satisfactory prediction model, and so an alternate technique which relies exclusively on Empirical data and ANOPP trends was developed. Three approaches were identified for investigation, as discussed in the Introduction: Method I. Empirical Taxi Noise Data and ANOPP data; Method II. Empirical Taxi Noise Data Only; and Method III. No Empirical Taxi Noise Data. The logic behind selecting these three approaches was that method I utilizes all the available information, whereas Method II utilizes empirical data for the nominal taxi condition, and assumes that insufficient information is available to generate thrust-noise sensitivity data using ANOPP. Method III corresponds to the case where one has neither ANOPP nor measured data. Attempts were made to utilize existing INM flight NPD data in conjunction with the Empirical and ANOPP taxi data but were unsuccessful. Section 5 of this report documents which Method (I, II or III) is to be used for each aircraft type in the INM / AEDT database. 85 90 95 100 105 110 115 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 SE L, d B A Corrected Net Thrust Taxi Noise vs. Thrust Comparison - B747-400 16 Kts - 1000 ft. INM NPD - Approach INM-NPD - Dep Madrid-AAM-MeasData INM App - no AFN (Log) INM App - no AFN (Diff) INM Dep - no AFN (Log) INM Dep - no AFN (Diff) Poly. (INM-NPD - Dep) Note: INM-NPD data corrected from 160 Kts to 16 Kts by adding 10 dBA

4-5 4.2. Correlation of ANOPP Data and Empirical Data The ANOPP generated taxi noise data for the five aircraft listed in the previous section were correlated vs. thrust at each distance to see if the resulting curves could be collapsed or normalized into a “universal” trend line of taxi noise vs. thrust. Such a correlation would be needed for those aircraft in the INM database for which there is no empirical taxi data. Further, measured data for several aircraft were plotted on the ANOPP trend curves to see whether adjustments or calibrations might be needed to match existing taxi noise data. In order to compare with the ANOPP data, taxi thrust for the measured data points was needed, and not all the acoustic taxi data available had associated with it nominal taxi thrusts. Therefore an attempt was first made to develop a taxi thrust correlation from existing information. A set of ICAO Best Practices Data Base (BPDB) tables (ICAO, 2008) was queried and a table of aircraft maximum takeoff weight (MTOW) and takeoff thrust was generated. By multiplying the engine takeoff thrust listed in the BPDB by the number of engines on the aircraft, and assuming that the taxi thrust was nominally about 5% of takeoff thrust , a correlation of taxi thrust vs. MTOW was obtained, as shown in Figure 4-5. An assumed value of 5% thrust is consistent with recent ICAO guidance (ICAO, 2006) for taxi power settings in cases where the taxi operating state is not known by other means. Also shown are taxi thrusts estimated for the ANOPP case aircraft. With the exception of the CRJ900 case, the ANOPP thrust cases tend to straddle the correlation line curve fit, suggesting that this curve may be used, with a reasonable amount of uncertainty, to estimate taxi thrust for any of the aircraft for which taxi thrust is not known but noise measurements are available. This correlation was used to estimate taxi thrusts for the taxi noise data used in the correlation model that follows. For the purposes of developing a Taxi Noise NPD dataset aircraft were grouped into “small” and “large” categories with small aircraft defined as those with a maximum certificated TOGW less than 300,000 lbs. and large aircraft with a MTOGW at or above 300,000 lbs. This division corresponds to the division between the INM/AEDT Large and Heavy Weight Class (S = Small, L = Large, H = Heavy). For the purposes of using the Taxi Noise regressions, AEDT Aircraft in Weight Class Heavy are considered Taxi Large and Weight Class Small and Large are considered Taxi Small. The Aircraft fleet dataset from INM / AEDT is illustrated graphically with the Taxi Noise division in Figure 4-6. Alternate means of categorizing the aircraft were explored – such as by max rated takeoff thrust, bypass ratio and single/twin aisle – but they did not let the ANOPP and measured data collapse as nicely.

4-6 FIGURE 4-5 Taxi thrust and weight correlation. FIGURE 4-6 Taxi thrust and weight correlation. The ANOPP taxi noise vs. thrust trend curves, for all five aircraft, were plotted at several distances. Note that the ANOPP taxi noise estimates were for engine-only, no airframe noise, and with inflow distortion (ground vortex ingestion) turned on. Figures 4-7 through 4-9 show examples of these y = -2.1852E-06x2 + 1.5114E-02x R² = 9.7640E-01 0 2 4 6 8 10 12 14 0 100 200 300 400 500 600 700 800 900 1000 To ta l T ax i T hr us t/1 00 0, lb f MTOW/1000, lbm Taxi Thrust Correlation 5% of ICAO BPDB T/O Thrust ANOPP Aircraft Cases Poly. (5% of ICAO BPDB T/O Thrust) 100 1000 10000 100000 1000000 1000 10000 100000 1000000 T H R _S T A T IC ( po un ds ) MTOW (pounds) Taxi Noise Class Categorization: Small < 300 klbs Max TOGW Taxi: MTOW Division FAA MTOW limit AEDT Heavy AEDT Large AEDT Small All INM/AEDT AC Single Aisle Madrid Jets Double Aisle Madrid Jets Taxi Small Taxi Large F A A M T O W L im it 30 0, 00 0 lb s M T O G W AEDT Weight Class: Small AEDT Weight Class: Large AEDT Weight Class: Heavy

4-7 comparisons for 400 ft. 1000 ft., and 4000 ft. Plots for the additional NPD distances and other metrics may be found in Appendix C. FIGURE 4-7 ANOPP predicted Taxi noise SEL trend lines – 400 ft. distance. Figure 4-8. ANOPP predicted Taxi noise SEL trend lines – 1000 ft. distance. y = 10.361ln(x) + 15.647 R² = 0.9902 y = 9.0214ln(x) + 21.821 R² = 0.9828 75 80 85 90 95 100 105 110 115 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 SE L,  dB A Corrected Net Thrust/2, Fn/delta, Lbs ANOPP Predicted SEL (dBA) Regression Trend Lines ‐ 400 Ft Distance Small AC (B737, CRJ) Large AC (B747,B767,B777PW,B777GE) B737‐800 CRJ900 B747‐400 B767‐300 B777‐300,GE90 B777‐300,PW4098 Log. (Small AC (B737, CRJ)) Log. (Large AC (B747,B767,B777PW,B777GE)) y = 9.8063ln(x) + 14.433 R² = 0.9866 y = 8.4121ln(x) + 22.094 R² = 0.9716 70 75 80 85 90 95 100 105 110 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 SE L,  dB A Corrected Net Thrust/2, Fn/delta, Lbs ANOPP Predicted SEL (dBA) Regression Trend Lines ‐ 1000 Ft Distance Small AC (B737, CRJ) Large AC (B747,B767,B777PW,B777GE) B737‐800 CRJ900 B747‐400 B767‐300 B777‐300,GE90 B777‐300,PW4098 Log. (Small AC (B737, CRJ)) Log. (Large AC (B747,B767,B777PW,B777GE))

4-8 FIGURE 4-9 ANOPP predicted Taxi noise SEL trend lines – 4000 ft. distance. The first observation made is that the small aircraft (B737-800 and CRJ900) fall on top of each other, while the larger aircraft (B767, B747 and B777) group together forming more-or-less a common curve, with very little scatter. The large aircraft grouping is about 4 to 5 dB lower than that for the two small aircraft. As distance increases, these two curve sets tend to come closer together, and the spread or variation within the groupings tends to increase, but there is still a reasonably close agreement. The second observation is that the measured data seem to agree very well with the curve shapes and even absolute levels, and the small aircraft data tends to cluster around the ANOPP small aircraft grouping, while the large aircraft data tends to cluster around the large aircraft ANOPP grouping. From the above results, it is concluded that a reasonable prediction model for taxi noise can be obtained using the above discussed curve shapes defined by the ANOPP data. Two sets of curves were utilized, one for small aircraft and one for large aircraft. Further, reasonable estimates of taxi thrust can be obtained using the correlation described above which assumes 5% of takeoff thrust correlated against aircraft MTOW. Curve fits for each metric (SEL, EPNL, PNLTmax, LmaxA) have been generated for each individual ANOPP aircraft type for each of the standard INM NPD distances. Logarithmic (natural) regressions have also been generated for the single aisle and twin aisle aircraft types. These regressions are of the form in Equation 4-1. Ltaxi = m* ln(Fn/delta) + b Eq. 4-1 where Ltaxi is the resultant Taxi noise level for the noise metric and distance for the corresponding m and b as provided in Table 4-1 or Appendix A4 at a Corrected Net Taxi Thrust/2 (Fn/delta/2, in lbs). The Fn/delta is specified as half of the taxi thrust. For 2 engine aircraft this represents single engine thrust, and for a 4 engine aircraft this is thrust from 2 engines. Coefficients for Equation 4-1 are provided in Tables 4-1 for single and twin aisle aircraft and in Appendix C, Table C.4 for individual aircraft types for which ANOPP models were available. y = 9.0296ln(x) + 8.9046 R² = 0.9724 y = 7.316ln(x) + 21.04 R² = 0.9353 60 65 70 75 80 85 90 95 100 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 SE L,  dB A Corrected Net Thrust/2, Fn/delta, Lbs ANOPP Predicted SEL (dBA) Regression Trend Lines ‐ 4000 Ft Distance Small AC (B737, CRJ) Large AC (B747,B767,B777PW,B777GE) B737‐800 CRJ900 B747‐400 B767‐300 B777‐300,GE90 B777‐300,PW4098 Log. (Small AC (B737, CRJ)) Log. (Large AC (B747,B767,B777PW,B777GE))

4-9 Table 4-1. Taxi Noise Regressions Coefficients for Turbofan Aircraft Taxi L max SEL EPNL PNLMX Aircraft Metric (dBA) (dBA) (dB) (dB) Size Dist (Ft) m b m b m b m b Small 200 10.3200 7.9640 10.7198 16.2886 12.3369 6.2596 11.9024 8.5163 Small 400 10.0152 4.2812 10.3613 15.6472 12.0901 4.2351 11.7303 3.3375 Small 630 9.7806 1.8226 10.0867 15.2065 11.8601 3.0685 11.5176 0.3356 Small 1000 9.5152 -0.8341 9.8063 14.4328 11.5525 2.1026 11.2950 -3.0609 Small 2000 9.1453 -5.7928 9.3922 12.4806 10.9702 0.6455 10.6570 -6.7214 Small 4000 8.7779 -12.0089 9.0296 8.9046 10.5707 -2.8424 9.8925 -9.0451 Small 6300 8.5778 -17.2758 8.8286 5.4338 10.7297 -8.9944 9.8592 -16.4073 Small 10000 8.4241 -23.8710 8.7901 -0.2324 11.4021 -19.6520 10.5515 -27.2874 Small 16000 8.4740 -33.1212 9.0338 -9.0055 12.9697 -40.0949 12.0674 -50.1154 Small 25000 8.8536 -45.2275 9.7704 -21.8142 15.4121 -71.7522 14.5871 -83.8764 Large 200 8.6658 15.2803 9.4283 21.4469 11.0498 9.1065 10.7826 8.5383 Large 400 8.3909 12.0880 9.0214 21.8208 10.6717 8.4541 10.3440 6.1362 Large 630 8.1262 10.4477 8.7461 21.8783 10.3093 8.7098 9.9457 5.1270 Large 1000 7.7369 9.4940 8.4121 22.0943 9.8061 9.9059 9.3669 5.2708 Large 2000 7.3104 5.9932 7.9140 21.6949 8.9051 12.6168 8.4123 5.8678 Large 4000 6.9264 0.9634 7.3160 21.0404 7.8294 16.6097 7.3173 8.7751 Large 6300 6.6231 -2.7306 6.9022 19.9089 7.3639 16.8470 7.0291 4.6488 Large 10000 6.2946 -7.2202 6.4337 18.2300 7.2036 13.7611 6.9945 0.5930 Large 16000 5.9846 -13.1308 6.0489 14.6151 7.4439 4.4268 7.2952 -11.6531 Large 25000 5.8408 -21.1360 6.0714 6.8268 8.4492 -15.0634 7.9305 -28.9701 4.3. Methods for Taxi NPD Database Creation As described above, aircraft in the INM database fall into one of three categories. Taxi noise NPDs may be generated for each category using the following procedures. Method I. Empirical Taxi Noise Data and ANOPP data Step 1. Process the Empirical Acoustic Taxi data (as described in Section 3.1 and Appendix A) and determine the taxi NPD noise metrics (SEL, EPNL, Lmax) as a function of distance for the reference atmospheric conditions for the measured “nominal” thrust condition. Step 2. Assign a “nominal” taxi thrust (Fn/delta) to the acoustic data based on either (in priority order): • Known engine operating state at the time of the acoustic data acquisition; • Generalized nominal taxi thrust based on historical FDR data; or • Assumed taxi thrust. Step 3. Evaluate the NPD levels at other Power settings using ANOPP predictions (Appendix C4) to create a noise sphere and AAM to propagate the source to the required distances. Method II. Empirical Taxi Noise Data Only Step 1. Process the Empirical Acoustic Taxi data (as described in Section 3.1 and Appendix A) and determine the taxi NPD noise metrics (SEL, EPNL, Lmax) as a function of distance for the reference atmospheric conditions for the measured “nominal” thrust condition.

4-10 Step 2. Assign a “nominal” taxi thrust (Fn/delta) to the acoustic data based on either (in priority order): • Known engine operating state at the time of the acoustic data acquisition; or • Generalized nominal taxi thrust based on historical FDR data; or • Assumed taxi thrust. Step 3. Evaluate the NPD levels at other Power settings above and below the “nominal” taxi thrust using the generalized thrust-noise relationship provided in Figures 4-6 through 4-8 and Appendix C4. Method III. No Empirical Taxi Noise or ANOPP Data Step 1. Assign a “nominal” taxi thrust (Fn/delta) to the acoustic data based on: • Generalized nominal taxi thrust based on historical FDR data; • Assumed taxi thrust based on certification max rated thrust; or • Assumed taxi thrust based on certification maximum takeoff gross weight. Step 2. Evaluate the NPD levels at other Power settings above and below the “nominal” taxi thrust using the generalized thrust-noise relationship provided in Figures 4-6 through 4-8 and Appendix C4. 4.4. Uncertainty of the NPD Creation Process The correlations of Method III have with them a spread or scatter, from which can be extracted uncertainties, in terms of standard deviations of noise predictions. The best empirical data was used in the creation of the final INM/AEDT NPD Dataset using Method I (Step 1) or Method II (Step 1), but for the purposes of assessing the uncertainty, it will be compared directly with Method III predictions, which do not use empirical data in the determination of Taxi Noise. The aircraft TOGW was selected from the INM/AEDT Database for the aircraft represented in the noise measurements. In the case of the Madrid data where several variants were grouped together (i.e. 737-300, -400, -500) the weights were averaged. The assumed taxi thrust was computed based on the TOGW-Taxi Thrust relationship described in Figure 4-5. For aircraft with a MTOGW below 300,000 lbs the Taxi-Small noise coefficients (Table 4-1) were used to generate the SEL, Lmax, EPNL, and PNLTmax for the standard NPD distances. Taxi Noise Method III predictions were compared with the empirical taxi data. A full itemization of the empirical taxi data is contained in Appendix C.5. The mean and standard deviation of the predicted minus measured differences were computed for each distance and each metric. The median of the mean difference and the Average of the Standard Deviations were calculated for all distances combined for a given metric, and are provided in Table 4-2. The Standard Deviation for all distances and all aircraft within a category (Small and Large) and the 90%, 95% and 99% Confidence Levels were then computed (Table 4-3). Slight changes in the assumed thrust will drive large changes in the noise predictions due to the steep slope of the noise – thrust curve at low power settings. Figures 4-10 through 4-13 illustrate the comparison with empirical data for the 1000 Ft distance for the metrics SEL, Lmax, EPNL and PNLTmax respectively. The final INM/AEDT dataset for the specific aircraft2 included in this uncertainty assessment has by virtue of the methodology, 0. Median of the Means and 0. Average Standard Deviation, as outlined in Method I and Method II. 2 Specific aircraft include those itemized in Appendix C.5, for which empirical acoustic taxi noise data exists and comparisons with Method III could be performed.

4-11 TABLE 4-2 Taxi Noise Uncertainty: Method III Compared with Empirical Data Metric Taxi-Small Taxi-Large Median of the Means Average of the St.Devs Median of the Means Average of the St.Devs SEL -3.1 3.8 1.7 2.5 Lmax -0.9 3.8 1.7 2.0 EPNL -5.8 4.8 -1.8 2.7 PNLTm -2.8 4.8 -1.8 2.5 * Note: The final INM/AEDT dataset did not use Method III for these aircraft. TABLE 4-3 Taxi Noise Confidence Levels: Method III Compared with Empirical Data Taxi- Small Taxi- Large Count 140 40 St.Dev 4.4 3.1 90% Confid 0.6 0.8 95% Confid 0.7 0.9 99% Confid 1.0 1.2 FIGURE 4-10 Comparison of Method III with empirical data, 1000 Ft – SEL (dBA). y = 9.8063ln(x) + 14.433 R² = 0.9866 y = 8.4121ln(x) + 22.094 R² = 0.9716 70 75 80 85 90 95 100 105 110 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 SE L,  dB A Corrected Net Thrust/2, Fn/delta, Lbs ANOPP Predicted SEL (dBA) Regression Trend Lines ‐ 1000 Ft Distance Small AC (B737, CRJ) Large AC (B747,B767,B777PW,B777GE) B737‐800 CRJ900 B747‐400 B767‐300 B777‐300,GE90 B777‐300,PW4098 Taxi Data ‐Small AC Taxi Data ‐ Large AC Log. (Small AC (B737, CRJ)) Log. (Large AC (B747,B767,B777PW,B777GE))

4-12 FIGURE 4-11 Comparison of Method III with empirical data, 1000 Ft – Lmax (dBA). FIGURE 4-12 Comparison of Method III with empirical data, 1000 Ft – EPNL (dB). y = 9.5152ln(x) ‐ 0.8341 R² = 0.9839 y = 7.7369ln(x) + 9.494 R² = 0.9422 55 60 65 70 75 80 85 90 95 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 Lm ax , d BA Corrected Net Thrust/2, Fn/delta, Lbs ANOPP Predicted Lmax (dBA) Regression Trend Lines ‐ 1000 Ft Distance Small AC (B737, CRJ) Large AC (B747,B767,B777PW,B777GE) B737‐800 CRJ900 B747‐400 B767‐300 B777‐300,GE90 B777‐300,PW4098 Taxi Data ‐Small AC Taxi Data ‐ Large AC Log. (Small AC (B737, CRJ)) Log. (Large AC (B747,B767,B777PW,B777GE)) y = 11.552ln(x) + 2.1026 R² = 0.9944 y = 9.8061ln(x) + 9.9059 R² = 0.9802 70 75 80 85 90 95 100 105 110 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 EP N L,  dB Corrected Net Thrust/2, Fn/delta, Lbs ANOPP Predicted EPNL (dB) Regression Trend Lines ‐ 1000 Ft Distance Small AC (B737, CRJ) Large AC (B747,B767,B777PW,B777GE) B737‐800 CRJ900 B747‐400 B767‐300 B777‐300,GE90 B777‐300,PW4098 Taxi Data ‐Small AC Taxi Data ‐ Large AC Log. (Small AC (B737, CRJ)) Log. (Large AC (B747,B767,B777PW,B777GE))

4-13 FIGURE 4-13 Comparison of Method III with empirical data, 1000 Ft – PNLTmax (dB). y = 11.295ln(x) ‐ 3.0609 R² = 0.9916 y = 9.3669ln(x) + 5.2708 R² = 0.9564 65 70 75 80 85 90 95 100 105 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 PN LT m ax , d B Corrected Net Thrust/2, Fn/delta, Lbs ANOPP Predicted PNLTmx (dB) Regression Trend Lines ‐ 1000 Ft Distance Small AC (B737, CRJ) Large AC (B747,B767,B777PW,B777GE) B737‐800 CRJ900 B747‐400 B767‐300 B777‐300,GE90 B777‐300,PW4098 Taxi Data ‐Small AC Taxi Data ‐ Large AC Log. (Small AC (B737, CRJ)) Log. (Large AC (B747,B767,B777PW,B777GE))