Integrated Noise Model Accuracy for General Aviation Aircraft (2014)

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.

2 State of the Practice: How are GA Aircraft Being Modeled? This state of the practice section documents the results of two efforts to determine current noise modeling practice for GA aircraft: Section 2.1a literature search; Section 2.2a web-based survey of Integrated Noise Model (INM) users. Harris Miller Miller & Hanson Inc. (HMMH) completed the literature review, and HMMH and Montgomery Consulting Group (MCG) worked together to produce and distribute the survey and summarize the results. Literature Review 2.1 The literature review was a broad search of published information to identify additional data on the noise levels produced by GA aircraft or on how GA aircraft noise is modeled. The search yielded eight published articles or reports, which provided limited information useful for the analysis of the modeling of noise generated by GA aircraft. The following paragraphs summarize these publications and their potential utility for this project. The bold italic text at the end of each summary gives the usefulness of the information to this project. Documents reviewed are listed in Appendix A. 2.1.1 “Review of Integrated Noise Model (INM) Equations and Processes” Forsyth, D.W., NASA/CR-2003-212414, May 2003 The FAA's Integrated Noise Model (INM) relies on the methods of the SAE AIR-1845 "Procedure for the Calculation of Airplane Noise in the Vicinity of Airports" issued in 1986. Simplifying assumptions for aerodynamics and noise calculation were made in the SAE standard and the INM. One the objectives of this study of interest to modeling GA aircraft was the test of some of the simplified assumptions against Boeing source data, but the testing was related to variations in only weight, atmospheric conditions and flap retraction schedule. Generally, the INM equation A1 assumptions appear to predict the corrected net thrust to ±2% of Boeing determined performance. It must be said, however, that the report assumes the reader has an in-depth understanding of engine performance and provides more information than can easily be summarized here. For the current study, when we examine modifying aircraft coefficients, the results suggest only that the form of equation A1 of SAE AIR-1845 is reasonable and that aircraft weight and flap retraction should be considered for alteration (if physically reasonable). 2.1.2 “Assessment of Tools for Modeling Aircraft Noise in the National Parks” Fleming, G.G., K.J. Plotkin, C.J. Roof, B.J. Ikelheimer, D.A. Senzig, FICAN, March 18, 2005 Because the main aircraft types of concern in the National Parks context are tour aircraft, this study might have suggested useful modeling issues to include. The first objective of the study was to evaluate the performance of two aircraft noise models, the FAA’s INM and NMSim (a model evolved from a study for the NATO Committee on the Challenges of Modern Society) for modeling the noise produced by tour aircraft over national parks. A second objective was to examine the usability of each model, i.e., ease of operation, runtime, data input / output availability, etc. Because high altitude jet aircraft also produce noise audible in parks, field measurements of some of these overflights were also conducted and are reported in appendices. In general, the study focuses on comparing calculations of audibility over long distances for propeller and rotary wing aircraft of the type used for air tours. None of these comparisons are judged applicable to the present assessment of INM accuracy in computing standard noise metrics for GA aircraft in the vicinity of airports. 5

2.1.3 “Fitchburg Municipal Airport Noise Measurement Study: Summary of Measurements, Data and Analysis” Reherman, C.N., C.L. Roof, G.G. Fleming, D.A. Senzig, D.R. Read, C.S.Y. Lee, DOT-VNTSC-FAA-03- 09, November 2005 This document reports on the measurement and analysis of the noise produced by six aircraft: Maule M-7- 235C, Piper Twin Comanche PA-30, Piper Navajo Chieftain PA-31-350, Piper Warrior PA-28-161, Beech 1900D, Eurocopter EC-130 Helicopter, and the Robinson R-22 Helicopter. It describes the planning and execution of the measurements made at Fitchburg Municipal Airport in Fitchburg, Massachusetts. Additionally, the data reduction procedures and data adjusted to standard conditions are presented. These aircraft, with the possible exception of the Beech 1900D are unlikely to be in the list of GA aircraft proposed for evaluation – they are unlikely to be significant contributors to GA aircraft produced noise exposure because we expect the GA jets to dominate noise exposure wherever they operate. Appendix A of the report includes a clear description of the calculation of INM coefficients for the 1900D using the measured and the manufacturer supplied data. The information contained in this reference for the Beech 1900D aircraft, should there be sufficient noise data from the monitoring systems / airports used in this INM GA aircraft accuracy study could provide a check on the potential errors in our determination of discrepancies. 2.1.4 “Integrated Noise Model (INM) Noise Contour Comparison: Version 7.0 vs. 6.2a” He, B., R. Cointin, E. Boeker, E. Dinges, C. Roof, FAA-AEE-07-01, October 2007 This report was intended to provide a general sense of noise level changes expected by using INM version 7.0 rather than previous version 6.2a. Four different airport studies were run, modifying, one at a time, terrain effects, lateral attenuation and bank angle as implemented in 7.0. One of the four was a “Large GA” airport, likely with the noise dominated by jets. Each was run with variations in headwind component and temperature. Results were examined in terms of DNL contours and their areas and LAMAX. Interestingly, the changes in contour area were, for most of the cases, greatest for the Large GA airport. No rigorous analysis was conducted, however, to isolate specific reasons for any changes, but the report suggests that the change in lateral attenuation is the dominant factor affecting the increase in DNL contours. We note that lateral attenuation is built into INM and is generic by aircraft type – e.g., for under-wing or for fuselage mounted engines. It is our intention to explore only the aircraft specific variables, and not the general means by which the INM uses those specifics in the computation of noise exposure. The information provided in this reference is judged non-applicable to the present assessment of INM accuracy in computing standard noise metrics for GA aircraft in the vicinity of airports. 2.1.5 “Sensitivity of the FAA integrated noise model to input parameters” Gaja, E., G. Clemente, A. Reig, Applied Acoustics, Volume 66 Issue 3, March 2005 The study conducted a sensitivity analysis of INM computations as they might be affected by errors in input information. It analyzed the effects of variations in aircraft weight, speed, flaps ID, climb rate and altitude on INM computed SEL. Not surprisingly, weight is the primary factor affecting SEL during takeoff since it affects climb rate and altitude. On approach, flaps ID is also important. This study was reviewed in hopes that it would shed light on engine and aerodynamic coefficients, which we have proposed to alter in our study of INM accuracy for GA aircraft. It did not. However, the importance of weight confirms our approach in the Modified Work Plan, page 9, to determine the effects on INM accuracy by modifying an aircraft’s weights. The information on the effects of weight confirms the 6

value of our approach, but the study provides no information directly applicable to our assessment of INM accuracy in computing standard noise metrics for GA aircraft. 2.1.6 “Error Sensitivity Analysis of the Integrated Noise Model” Restrick, K., EEC Note: EEC/ENV/2002/006, 28/06/2002 This report was reviewed because, under the expanded work plan, we proposed under “Identify causes of discrepancies: modification of INM input” to “Modify Coefficients,” page 9 of the plan. This is one of the few studies that reports on input error sensitivity. Though there is sufficient information to believe the effort was thorough, the text explanations leave much to be desired and it is not possible to easily connect all the analysis steps and the results. Thirteen aircraft types were run through the analysis for six flight segments. The study presents a method for analyzing and calculating the effect of errors in the inputs to each of the equations used within INM. The most influential variables are found to be aircraft weight, local pressure, and a number of aircraft flap and engine coefficients. The coefficients examined are: Aerodynamic – B – Ground roll flap coefficient (multiplied by weight, thrust to get ground roll) C – Takeoff flap coefficient (multiplied by sq. root of weight gives climb CAS, Calibrated Airspeed) D – Landing flap coefficient (multiplied by sq. root of weight gives approach CAS) R – Drag coefficient to lift coefficient for a given flap setting (landing gear retracted) (for a given configuration and airspeed, used with weight and segment thrust and to give climb angle for the segment) Engine corrected net thrust equation – E – Constant F – Constant multiplying calibrated airspeed G – Constant multiplying pressure altitude, h (height above MSL at ISA for performance); in some cases there are two G constants, one multiplying h, one multiplying h squared H – Constant multiplying ambient air temperature at plane Interpreting the average percent errors in report Table 3, it appears that in descending order of importance, errors in aircraft weight and in coefficients R, E and F, are likely to have the greatest effect on aircraft position. Presumably, since they all relate to power required / delivered by the aircraft, they may also have the greatest influence on computed noise level. The information contained in this report will be useful to our analysis in that we will consider the results in prioritizing the modification of variables to determine their effect on resulting INM computed GA aircraft noise levels. 2.1.7 “The Mosquito Effect: Community Reaction to Noise from a General Aviation Airport” Lloyd, D., Inter-noise 2000, Nice, France, 2730 August 2000 This study determined the reaction of residents to aircraft noise from Jandakot, a GA airport having about 1100 daily aircraft operations. The results of the study were compared with the expected responses calculated using the Australian Noise Exposure Forecast (ANEF) system. The response of 330 residents was determined by conducting a social survey in the areas within the INM computed ANEF contours out to 10 ANEI (approximately 45 DNL). The results of the survey indicated that reaction to light aircraft 7

noise was approximately 7 ANEF units higher than predicted by a 1982 National Acoustics Laboratory survey of residents living around five major Australian airports. Validation of the INM ANEF contours showed that actual noise levels from aircraft overflights were 3 ANEF units higher than the computed ANEF levels. When taking into consideration the actual noise, it was concluded that the reaction to aircraft noise from this airport was 4 ANEF units higher than would be expected for major airports. This would suggest that a 4 ANEF adjustment should be considered when planning for residential land use around GA airports. However, it should be noted that the difference of 4 ANEF units which has been found at Jandakot, while statistically significant, is not outside the bounds of variation between airports which would have been expected from results at major airports from the 1982 NAL study. It appears from examination of the tarmac in Google Earth and from reading a description of the airport that it is primarily, if not exclusively, an airport of propeller aircraft operations. It is interesting as a data point for the hypothesis that annoyance is higher for low noise, large numbers of aircraft than for loud less frequent jet aircraft, but has probable little relevance for the present study. 2.1.8 “Simplified Computation Procedure for Assessment of the Noise Load in the Vicinity of Sport Airfields” Sandor, H., 2006 Annual Report, KTI Institute for Transport Sciences Nonprofit Limited, Hungary ISSN: 1789-4042, pp. 5157 Hungary is apparently complying with the European Noise Directive, 2002/49/EC, that requires EU members to create noise maps, provide the information to the public and establish Noise Action Plans to reduce numbers of people exposed to excessive noise. The method for airports has been judged as too complicated and too precise for “sport airfields,” which we take to mean low use GA airports. This study describes a method to determine “noise loads” (noise exposure) as a function of distance from the runway / flight track and of the number of operations at a sport airfield. It provides graphs that can be used knowing any two of the three variables of number of operations per 16 hours (maybe they only operate during daylight), distance and noise load, to derive the third. The method has no direct application to our study. 2.2 INM User Survey The INM user survey was completed to determine how GA aircraft noise modelers currently use the INM, collect input data, validate the model and determine noise exposure from GA aircraft at a variety of airports, from strictly GA to large hub commercial airports. The survey was conducted using a web-based provider.4 The survey process had three steps that generated the responses: 1. Development of the database of INM users 2. Design of the survey questionnaire 3. Execution of the questionnaire A brief summary of the method for the respective steps is described in the following three sections. The survey results are presented and discussed in Section 2.2.4. The questionnaire is provided in Appendix B. 2.2.1 Development of database INM users Potential INM users including airport, airline industry, and state and federal agency users were identified and compiled from the following sources: • HMMH proprietary lists of HMMH-hosted INM Training Workshop attendees from 2001-2011 4 SurveyMonkey™ 8

• LinkedIn5 INM User Group • Airport Consultants Council (ACC)6 database of firms identified as providing noise compatibility consulting • Known INM modelers HMMH and MCG contact lists • Internet research conducted by MCG – determined that most potential INM users were already included in the above lists Combined, the initial potential U.S. and international INM users lists contained 483 contacts. Since the user database contained e-mails from 2001 – 2011, it was expected that many of the e-mails would bounce (discontinued e-mail addresses). If the e-mail address was non-existent, MCG attempted to resolve the user contact information by locating the contact at another organization or contact address (i.e., firm had been acquired, contact had moved to another firm, etc.). MCG received 123 “bounced” e- mails upon execution of the survey. The final database of potential INM user names consisted of 256 potential U.S. users and 104 potential international users. Appendix C summarizes the types of users surveyed and responding. 2.2.2 Questionnaire development HMMH in combination with MCG constructed the survey. Two separate survey questionnaires were developed, a survey of the potential U.S. users and a survey for potential international users. The U.S. and international user surveys were slightly different (“FAA” as an organization applies to only the U.S.), and are provided separately to permit analysis of differences in GA aircraft modeling approach, if any. An online software application, SurveyMonkey™, was used to execute the questionnaire. The questionnaire included twelve questions with multiple choice responses and/or text fillable user input. The survey began requesting the type of user based on their organization and asked whether they modeled GA aircraft using the INM (those who did not model GA aircraft exited the survey). The survey then focused on model input data sources, how they modeled GA aircraft and whether they have compared and/or used noise measurement data to validate or modify the model. The survey ended by asking the participant whether they could be contacted for further questioning. The twelve survey questions and summary of responses are provided in Section 2.2.4.2. 2.2.3 Questionnaire execution The survey commenced on December 6, 2011 with the original invitation to complete the online survey, and was followed-up with reminder e-mails on December 13 and December 20, 2011. A final reminder e-mail advised recipients that survey data collection would conclude on December 23, 2011. The following summarizes the INM user participation: • 121 U.S. users responded to the 2617 queries ( 46.4% response rate) • 35 international users responded to the 998 queries (35.4% response rate) 2.2.4 Summary of responses The following two sections summarize the responses, first as generalizations, then more specifically in tables. 5 LinkedIn is a business oriented social network, http://www.linkedin.com/ (accessed 2 February 2012) 6 ACC is an international trade association that represents private businesses involved in the development and operations of airports and related facilities, http://www.acconline.org/ (accessed 2 February 2012) 7 Note that 256 emails were sent, but five were forwarded by recipients to others who were the INM users. 8 Of the 104 originally identified emails, in subsequent emails, 5 addresses thought valid ultimately bounced. 9

2.2.4.1 Summary Generalizations The following summary percentages are based on the number of respondents who answered each specific question. Except for question 1, which was answered by 97 from the U.S. and by 38 international participants, each question was answered by 61 to 71 U.S. respondents and by 21 to 29 international respondents. In this section, percentages have been rounded. For both U.S. and international respondents, their division among organizations was similar with about 55% consultant, 23% government and 12% airport personnel.9 For respondents who model GA aircraft more often than rarely, about 1 ½ times as many U.S. respondents model GA aircraft as compared with international respondents, or about 69% of U.S. compared with 45% of international respondents. U.S. respondents report modeling all aircraft types about equally with a slight tendency toward piston props, while the international respondents report modeling turbojets slightly more often than modeling the other two types. For determining how to model existing fleet mixes, a wide range of resources are used, particularly in the U.S. There is a clear preference for interviews, while use of permanent noise and operations monitoring systems and radar data combined represent a close second for international respondents. For determining future fleet mixes, all respondents report primary use of airport and consultant forecasts, while U.S. respondents also make frequent use of the current fleet adjusted by the Terminal Area Forecast or other federal forecasts. The weather data that are used are predominantly annual averages, though international respondents make heavy use of seasonal averages and long term (decades of) historical data. Though half or fewer of the respondents modify GA aircraft data for modeling, of those who do, modifications are more often to the profiles of altitude, speed or thrust. Respondents report that modifications to weights or noise-power-distance curves are made about half as often as they are made to profiles. Occasionally spectral classes are altered. Well over half the respondents find they need to model GA aircraft types not in the INM database, and generally do so by substituting other INM aircraft types. A few also develop user defined. While most respondents never or rarely compare modeled GA aircraft noise levels with measured levels, about 40% report doing so. Finally, though most respondents report never adjusting computed GA aircraft noise levels using measured data, 27% to 28% report doing so. 2.2.4.2 Specific Responses by Question The following tables provide more detail on respondent answers to the twelve questions. They compare U.S. and international responses. The percentages shown were used to develop the previous section’s generalizations. Percentages are based on answers of “rarely” which is for these tables the sum of “never” and “rarely” in the answers. “Mostly” is the sum of the responses of “occasionally” or more frequently. The intent is to emphasize those respondents who fairly frequently take the action in question versus those who almost never do. In this way, the “mostly” responses can be considered to be from 9 For these generalizations and the following section, respondents who marked “other,” and who could clearly be categorized as government, consultant or airport were included in the totals for those types of organizations. 10

those who are very familiar with modeling GA aircraft and who best represent the state of the practice for using the INM to model GA aircraft. Question #1 What best describes your organization? Answer Summary U.S. Respondents Int'l Respondents Airport 13.1% 11.9% FAA - U.S./CAA - Int. 14.0% 4.8% Military 1.9% 2.4% Other Government 6.5% 16.7% Consultant 55.1% 54.8% Other 9.3% 9.5% Other responses: Noise Committee Member, Manufacturer, University, Research & Development Center Research Center, Air Navigation Services Provider, University, Airline Question #2 Do you model GA aircraft with the INM? Answer Summary U.S. Respondents Int'l Respondents Rarely (once/year or less) 31.0% 55.2% Mostly (>twice/year) 69.0% 44.8% Question #3 For airports where you use the INM to model GA aircraft, what have been the predominant GA aircraft types? Answer Summary U.S. Respondents Int'l Respondents Rarely Mostly Rarely Mostly Turbojets 45% 31% 29% 38% Turboprops 34% 34% 34% 33% Piston-engine props 21% 36% 37% 29% 11

Question #4 What data sources have you used for estimating types of GA aircraft to model EXISTING noise? Answer Summary U.S. Respondents Int'l Respondents Rarely Mostly Rarely Mostly Enhanced Traffic Management System Count 24% 14% 19% 13% Radar data 19% 18% 19% 13% Perm. Noise & Ops. Monitoring Systems 22% 16% 15% 22% Interviews FBO, Tower, Based Operator(s) 8% 28% 13% 38% Commercial Available Sources 18% 18% 23% 6% Other 9% 8% 10% 9% Other responses: Master Plans, Observation, INM database, Internal data, Airport sponsor, Fictional ops. data (for education), FAA databases, Demand forecasts, NextGen data Consultant data, ANP Database, Internet search Question #5 What data sources have you used for estimating types of GA aircraft to model FUTURE noise? Answer Summary U.S. Respondents Int'l Respondents Rarely Mostly Rarely Mostly Fleet TAF Adjusted or federal sources 36% 44% 65% 17% Airport/Consult. forecast 21% 48% 6% 74% Other 43% 7% 29% 9% Other responses: Socioeconomic data, Internal forecasts, Fictional ops. data (for education), Aviation forecast trends, Primary users input, Airport Managers, Letters of Interest Internet search, Other consultants, Interviews with major operators and FBO's 12

Question #6 What data do you use to specify weather conditions? Answer Summary U.S. Respondents Int'l Respondents Rarely Mostly Rarely Mostly Annual Average 6% 67% 0% 48% Seasonal Average 71% 19% 67% 41% Other (specify below) 23% 14% 33% 11% Other responses: Actual data, Consultant Wind Study, Design day, Daily, Single-event conditions, Specific day/time periods, Fictional ops. data (for education purposes), Metric-specific time periods Actual weather data of the last 30 years, Actual weather cases, 20 year averages Question #7 Have you ever modified INM provided GA aircraft data? Answer Summary U.S. Respondents Int'l Respondents Rarely Mostly Rarely Mostly Takeoff weights 35% 20% 33% 24% Departure/Arrival alt- speed-thrust 26% 44% 23% 48% Noise-Power-Dist curve 31% 24% 35% 19% Other 8% 11% 9% 10% Other responses: Substitute aircraft types within INM database, Spectral class for user- defined aircraft, Spectral class based on internal noise estimates, Investigate performance coefficients Situation changes and factoring of movement numbers Question #8 Have you needed to model a GA aircraft that is not in the INM database? Answer Summary U.S. Respondents Int'l Respondents Rarely 37.7% 28.6% Mostly 62.3% 71.4% 13

Question #9 When you have wanted to model a GA aircraft not in the standard INM database, have you: Answer Summary U.S. Respondents Int'l Respondents Rarely Mostly Rarely Mostly Excluded it - modeling 50% 6% 45% 3% Used INM standard sub 5% 43% 14% 38% Assigned standard type 9% 36% 14% 41% Developed user defined 36% 15% 27% 19% Question #10 Have you compared INM modeled GA aircraft noise with measured noise? Answer Summary U.S. Respondents Int'l Respondents Rarely 61.0% 61.9% Mostly 39.0% 38.1% Question #11 Have you adjusted INM modeled GA aircraft noise using measured noise? Answer Summary U.S. Respondents Int'l Respondents Rarely 88.1% 85.7% Mostly 11.9% 14.3% Question #12 May we contact you directly if we have additional questions? Answer Summary U.S. Respondents Int'l Respondents Yes, provided name 41.2% 47.4% Recommendations 2.3 The literature review provides useful information about how modifications of coefficients used within the INM might be made to the GA aircraft database, should that be a useful approach to improving computed results. However, the literature provided no information on comparisons of GA aircraft modeled results using the INM with measured values or any information about current modeling procedures or methods. The user survey, however, suggests that current INM modeling of GA aircraft noise would benefit, probably significantly in reductions of modeling effort, cost of modeling, and in uniformity of results if: 14

1. More GA aircraft types were provided as standard types in the INM database 2. Clear methods were provided / described for: a. Selecting substitution standard INM GA aircraft types for those types not specifically included in the INM database b. Modifying departure altitude and thrust profiles, as described in Section 8 and as recommended in Section 9. 15