388.42 Sp34r REPORT NO. D339-10019 RAIL TRANSIT NOISE ABATEMENT FOR MINIMUM COST ASSESSMENT OF URBAN RAIL NOISE CLIMATES AND ABATEMENT OPTIONS R. H. Spencer S. L. Wolfe .< OF Tito*. MARCH 1976 Final Report Document is available to the public through the National Technical Information Service, Springfield, Virginia 22151 f Prepared for DEPARTMENT OF TRANSPORTATION TRANSPORTATION SYSTEMS CENTER CAMBRIDGE, MASS. 02142 DOT-TSC-850-3 March 1976 NOTICE THIS DOCUMENT IS DISSEMINATED UNDER THE SPONSORSHIP OF THE DEPARTMENT OF TRANS¬ PORTATION IN THE INTEREST OF INFORMATION EXCHANGE. THE UNITED STATES GOVERNMENT ASSUMES NO LIABILITY FOR ITS CONTENTS OR USE THEREOF. THE UNITED STATES GOVERNMENT DOES NOT ENDORSE PRODUCTS OR MANUFACTURERS. TRADE OR MANUFACTURERS' NAMES APPEAR HEREIN SOLELY BECAUSE THEY ARE CONSIDERED ESSENTIAL TO THE OBJECT OF THIS REPORT. Technico! Report Documentgtion Page 1. Report No. 2. Government Accession No. 3. Recipient's Catolog No. 4. Title ond Subtitle Rail Transit Noise Abatement for Minimum Cost — Assessment of Urban Rail Noise Climates and Abatement Options 5. Report Dote March 1976 6. Performing Orgonizotion Code 8-2791 8. Performing Orgonizotion Report No. D339-10019 7. Aurho r 1 s) Spencer, R. H.; Wolfe, S. L.* 9. Performing Organization Nome ond Address The Boeing Vertol Company P.O. Box 16858 Philadelphia, Pa. 19142 10. Work Unit No. (TRAIS) 11. Controct or Grcnt No. DOT-TSC-850 13. Type of Report ond Period Covered Final Report March 1975 - March 1976 12. Sponsoring Agency Nome and Address Department of Transportation Transportation Systems Center Kendall Square Cambridge, Mass. 02142 14. Sponsoring Agency Code 612 15. Supplementary Notes *Wilson- Ihrig and Associates, Inc. 16. Abstrccr This report contains data on the effectiveness and costs of noise abatement techniques appropriate to the rapid rail transit lines of the San Francisco Bay Area Rapid Transit, Cleveland Transit System, Port Authority Transit Cor¬ poration (Lindenwold Line), and Southeastern Pennsylvania Transportation Authority (Broad Street Subway and Market-Frankford Line). Data is tabulated by categories of a) rapid rail vehicle applications, and b) track, roadbed and associated structures. Non-acoustic constraints bearing on the applicability of each abatement technique are cited. A noise abatement methodology, developed during the MBTA Pilot Study, is applied to select the optimum combination of noise abatement techniques. The cost and applicability data developed here is combined with the noise environment data base previously reported separately for each of the above transit systems. The abatement methodology stresses achieving a specified level of noise reduction for the minimum cost. The report concludes with a discussion of the applicability of the study results and recommendations for their use by the transit systems. UNIVERSITY OF ILLINOIS LIBRARY AT URBANA - CHAMPAIGN 17. Key Words 18. Distribution Stotemenf Noise, Cost, Rapid Transit, Transportation Noise, Abatement Techniques, Abatement Methodology, Abatement Effectiveness, Sources, Paths, Receivers Document is available to the public through the National Technical Information Service, Springfield, Virginia 22151. 19. Security Clossif. (of this report) 20. Security Clossif. (of this page) 21* No. of P ages 22. Price Unclassified Unclassified 146 Form DO I F 1700.7 (8-72) Reproduction of completed page authorized PREFACE This report has been prepared by The Boeing Vertoi Company and Wilson-lhrig and Associates under Contract DOT-TSC-850 as part of the Urban Rail Noise Program. Authors of the report were R. H. Spencer of Boeing Vertoi, and S. L. Wolfe of Wilson-lhrig and Associates. Techni¬ cal Monitor for the program was Dr. R. Lotz of the Transportation Systems Center. in Digitized by the Internet Archive in 2019 with funding from University of Illinois Urbana-Champaign Alternates https://archive.org/details/railtransitnoiseOOspen TABLE OF CONTENTS Page 1. Summary.. 1 — 1 2. Introduction. 2—1 3. Noise Abatement Techniques. 3—1 A. Discussion of Abatement Effectiveness .............. 3—1 1. Rail Vehicle. 3-1 Vehicle Component Noise .. 3—1 Ducted-Forced Ventilation of Propulsion System Isolation of Floor from Vehicle Superstructure Isolation and Balancing of Auxiliary Equipment Components Non-Skid Braking System Air Brake Vent Mufflers Redesign, Repair and Maintenance of Doors Wheel/Rail Noise.3—7 Resilient Wheels Damped Wheels Wheel Truing Soft Journal Sleeves Sound Absorption Treatment on Bottom of Car Floor and Inside Face of Car Skirt 2. Track, Roadbed and Associated Structures . .......... 3—10 Wheel/Rail Noise.3-10 Welded Rail and Rail Grinding Maintenance of Rail Joints Rail Lubrication General Way Structures .3—11 Absorptive Treatment in Tunnels with Concrete Trackbed Ballast and Tie Trackbed in Tunnels Absorptive Treatment in Subway Stations Ballast and Tie Trackbed in Subway Stations Sound Barrier Walls Trackbed in Retained Cut B. Application to Rail Transit Systems . ... . 3—16 1. Bay Area Rapid Transit...3—16 Description of Transit System Vehicle Noise Reduction v TABLE OF CONTENTS (Continued) Page Wheel/Rail Noise Trackbed and Way Structure Noise Reduction Bay Area Rapid Transit District Line Summary 2. Cleveland Transit System.3—27 Description of Transit System Vehicle Noise Reduction Wheel/Rail Noise Trackbed and Way Structure Noise Reduction CTS Rapid Transit Noise Summary 3. Port Authority Transit Corporation.3—33 Description of Transit System Vehicle Noise Reduction Wheel/Rail Noise Trackbed and Way Structure Noise Reduction General Way Structures PATCO Noise Summary 4. Southeastern Pennsylvania Transportation Authority.3—40 Broad Street Subway Description of Transit System Vehicle Noise Reduction Wheel/Rail Noise Trackbed and Way Structure Noise Reduction Broad Street Subway Noise Summary Market-Frankford Line Description of Tranist System Vehicle Noise Reduction Wheel/Rail Noise Trackbed and Way Structure Noise Reduction General Way Structure Market-Frankford Noise Summary 4. Transit Systems Minimum Cost Noise Reduction.4—1 A. Methodology Applied to Investigation.4—1 1. Urban Rail Noise Sources, Paths and Receivers.4—1 2. Formulation of Noise Control Scenarios.4—4 3. Determination of Minimum-Cost Noise Control Options.4—4 VI TABLE OF CONTENTS (Continued) Page B. System Noise Reduction.4—4 1. Bay Area Rapid Transit.4—8 Wayside Noise Reduction Station Noise Reduction In-Car Noise Reduction 2. Cleveland Transit System.. 4—21 Wayside Noise Reduction Station Noise Reduction In-Car Noise Reduction 3. Port Authority Transit Corporation.4—30 Wayside Noise Reduction Station Noise Reduction In-Car Noise Reduction 4. Southeastern Pennsylvania Transportation Authority ....... 4—41 Broad Street Subway Wayside Noise Reduction In-Car Noise Reduction Market-Frankford Subway Wayside Noise Reduction Station Noise Reduction In-Car Noise Reduction 5. Program Summary.5—1 6. Conclusions and Recommendations.6—1 7. References. 7—1 vu TABLE OF CONTENTS (Continued) Page B. System Noise Reduction.4—4 1. Bay Area Rapid Transit.4—8 Wayside and Community Station Vehicle Interior 2. Cleveland Transit System.4—21 Wayside and Community Station Vehicle Interior 3. Port Authority Transit Corporation.4—30 Wayside and Community Station Vehicle Interior 4. Southeastern Pennsylvania Transportation Authority.4—41 Broad Street Subway Wayside and Community Station Vehicle Interior Market-Frankford Line Wayside and Community Station Vehicle Interior 5. Program Summary.5—1 Background and Purpose 6. Conclusions and Recommendations.6—1 7. References.7—1 viii LIST OF ILLUSTRATIONS Figure Page 3—1 Summary of BART Noise Environment. 3—22 3—2 Summary of CTS Noise Environment.3—32 3—3 PATCO Lindenwold Line — Roadbed Summary.3—34 3—4 Summary of SEPTA Broad Street Subway Noise Environment.3—44 3—5 Summary of SEPTA Market-Frankford Line Noise Environment .... 3—49 3— 6 SEPTA Market-Frankford Subway-Elevated Roadbed Schematic .... 3—50 4— 1 Illustration of In-Car Noise Path Combination Used in Control Scenarios.4—3 4—2 Sample Noise Control Scenario Definition — PATCO — In-Car Noise Double Cars.4—5 4—3 Minimum Costs to Reduce Wayside Noise — Bay Area Rapid Transit.... 4—11 4—4 Minimum Costs to Reduce Station Noise — Bay Area Rapid Transit . . . 4—15 4—5 Minimum Costs to Reduce In-Car Noise — Bay Area Rapid Transit. . . . 4—22 4—6 Minimum Costs to Reduce Wayside Noise — CTS.4—26 4—7 Minimum Costs to Reduce Station Noise — CTS.4—31 4—8 Minimum Cost to Reduce In-Car Noise — CTS.4—35 4—9 Minimum Costs to Reduce Wayside Noise — PATCO.4—39 4—10 Minimum Costs to Reduce Station Noise — PATCO.4—45 4—11 Minimum Cost to Reduce In-Car Noise — PATCO . .. . 4—50 4—12 Minimum Costs to Reduce Wayside Noise — SEPTA Broad Street Subway.....4—55 4—13 Minimum Costs to Reduce Station Noise — SEPTA Broad Street Subway ....4—60 ix LIST OF I LLUSTRATIONS (Continued) Figure Page 4—14 Minimum Cost to Reduce In-Car Noise — SEPTA Broad Street Subway.4—64 4—15 Minimum Costs to Reduce Wayside Noise — SEPTA Market- Frankford Line.4—69 4—16 Minimum Costs to Reduce Station Noise — SEPTA Market- Frankford Line.4—74 4—17 Minimum Costs to Reduce In-Car Noise — SEPTA Market Frankford Line.4—78 x LIST OF TABLES Table Page 3—1 Summary of Noise Abatement Techniques.3—2 3—2 Summary of Wayside Noise Measurements — BART .3—23 3—3 Summary of Station Platform Measurements — BART .3—25 3—4 Summary of In-Car Noise Measurements — BART .3—26 3—5 CTS Railcar Inventory-Airport Rapid Transit.3—28 3—6 Specification of CTS St. Louis Car.3—29 3— 7 Distances Between Rapid Transit Stations — PATCO ......... 3—35 4— 1 Reference Code List for Noise Control Scenario Sources and Paths. ..... 4—2 4—2 Sample Scenario Breakdown — PATCO In-Car Noise Levels — Double Cars.4—6 4—3 Reference Code List for Noise Control Scenario System Fixes.4—7 4—4 Potential Wayside Noise Reduction Techniques — BART .4—9 4—5 Wayside Noise Abatement Techniques — BART .4—10 4—6 Computer Selected Techniques for Wayside Noise Reduction — BART . . 4—12 4—7 Potential Station Noise Reduction Techniques — BART .4—13 4—8 Station Noise Abatement Techniques — BART .4—14 4—9 Computer Selected Techniques for Station Noise Reduction — BART . . 4—16 4—9a Potential In-Car Noise Reduction Techniques — BART .4—17 4—10 In-Car Noise Abatement Techniques — BART . ..4—20 4—11 Computer Selected Techniques for In-Car Noise Reduction — BART . . . 4—23 4—12 Potential Wayside Noise Reduction Techniques — CTS ........ 4—24 4—13 Wayside Noise Abatement Techniques — CTS . ........... 4—25 4—14 Computer Selected Techniques for Wayside Noise Reduction — CTS Pullman Cars...4—27 4—15 Potential Station Noise Reduction Techniques — CTS ..4—28 4—16 Station Noise Abatement Technqieus — CTS. 4—29 4—17 Computer Selected Techniques for Station Noise Reduction — CTS . . . 4—32 4—18 Potential In-Car Noise Reduction Techniques — CTS.4—33 4—19 In-Car Noise Abatement Techniques — CTS.4—34 4—20 Computer Selected Techniques for In-Car Noise Reduction — CTS . . . . 4—36 4—21 Potential Wayside Noise Reduction Techniques — PATCO.4—37 4—22 Wayside Noise Abatement Techniques — PATCO.4—38 4-23 Computer Selected Techniques for Wayside Noise Reduction — PATCO . . 4—40 4—24 Potential Station Noise Reduction Techniques — PATCO.4—42 4—25 Station Noise Abatement Techniques — PATCO . ..4—44 4—26 Computer Selected Techniques for Station Noise Reduction — PATCO . . 4—46 4—27 Potential In-Car Noise Reduction Techniques — PATCO.4—47 4—28 In-Car Noise Abatement Techniques — PATCO ..4—49 4—29 Computer Selected Techniques for In-Car Noise Reduction — PATCO . . 4—51 4—30 Potential Wayside Noise Reduction Techniques — SEPTA Broad Street Subway...4—53 4—31 Wayside Noise Abatement Techniques — SEPTA Broad Street Subway . . 4—54 xi LIST OF TABLES (Continued) Table Page 4—32 Computer Selected Techniques for Wayside Noise Reduction — SEPTA Broad Street Subway.4—56 4—33 Potential Station Noise Reduction Techniques — SEPTA Broad Street Subway.4—57 4—34 Station Noise Abatement Techniques — SEPTA Broad Street Subway . . 4—59 4—35 Computer Selected Techniques for Station Noise Reduction — SEPTA Broad Street Subway. 4—61 4—36 Potential In-Car Noise Reduction Techniques — SEPTA Broad Street Subway.4—62 4—37 In-Car Noise Abatement Techniques — SEPTA Broad Street Subway . . . 4—63 4—38 Computer Selected Techniques for In-Car Noise Reduction — SEPTA Broad Street Subway.4—65 4—39 Potential Wayside Noise Reduction Techniques — SEPTA Market-Frankford Line.4—67 4—40 Wayside Noise Abatement Techniques — SEPTA Market-Frankford Line . 4—68 4—41 Computer Selected T echniques for Wayside Noise Reduction — SEPT A Market-Frankford Line.4—70 4—42 Potential Station Noise Reduction Techniques — SEPTA Market- Frankford Line. 4—72 4—43 Station Noise Abatement Techniques — SEPTA Market- Frankford Line.4—73 4—44 Computer Selected Techniques for Station Noise Reduction — SEPTA Market-Frankford Line.4—75 4—45 Potential In-Car Noise Reduction Techniques — SEPTA Market-Frankford Line .. 4—76 4—46 In-Car Noise Abatement Techniques — SEPTA Market- Frankford Line.4—77 4—47 Computer Selected Techniques for In-Car Noise Reduction — SEPTA Market-Frankford Line.4—79 • • Xll 1. SUMMARY This report is the second of two studies relating to noise assessment and abatement of five U.S. transit lines. These systems are the Bay Area Rapid Transit District in San Francisco, the Cleveland Transit System, the Delaware River Port Authority's Port Authority Transit Corpora¬ tion and the Southeastern Pennsylvania Transportation Authority's Market-Frankford Line and Broad Street Subway in Philadelphia. The first report of this series, issued in four volumes, described the noise climate on each of the lines (Reference 1). The current report describes techniques for abating noise on rail transit systems with direct application of specific techniques to each of the noted lines, and gives estimates of cost to reduce noise levels on each system to some prescribed goals. The intent of the report is to provide a realistic overview of what can be done to control noise on each of the identified systems. Each described abatement technique has been evaluated previously in other studies on one or more transit systems with data obtained on the effective¬ ness it provides in reducing noise. Some thought has been given to the non-acoustic constraints inherent in these techniques, as well. Recommendations for future work are given. 1-1 2. INTRODUCTION 2.1 Background Transit system noise, most often arises from a combination of sources, not all of which are generated by the car itself. The elimination or reduction of any one source, may not always result in a noticeable reduction in the total observed noise to a passenger within the car, or to residents residing near the system. For example, consider five radio re¬ ceivers tuned to different radio stations. When one receiver is turned off, there is very slight reduction in the observed noise level. This analogy provides an insight into the complexity of transit system noise, where wheel/rail, propulsion system, blowers, com¬ pressors, and brakes all contribute to the acoustic signature of the vehicle. Any approach to the control of noise would not be realistic without the addition of the cost involved. The cost of an improved noise environment can only be minimized by careful selection of the treatment used to achieve it. 2.2 Scope The Department of Transportation's Transportation Systems Center has reported on an initial program at the Massachusetts Bay Transportation Authority (MBTA), which for¬ mulated the approach to rail rapid transit noise control. This study (Reference 2) was conducted on the Red, Orange, and Blue transit lines of Boston. This report is the second of a two part series prepared by Boeing Vertol, which performs a similar function for four other U.S. transit systems, BART in San Francisco, CTS in Cleveland, PATCO and SEPTA in Philadelphia. The first report of this series, issued in four volumes (by system) was an assessment of noise levels on each of the noted operating properties (Reference 1). The subject report discusses the noise abatement techniques available for reducing system noise, and the approach to specified noise levels for minimum cost. The listing and des¬ cription of noise abatement techniques and their degree of noise reduction is applicable to both existing and future rail rapid transit systems. The techniques in all cases are practical, i.e., they have been tested on operational transit lines and can be implemented on most of the existing systems. The noise reduction effectiveness of each technique is based upon field data obtained at one or more transit systems. In some cases, the field data may have been taken for limited conditions, ie., a change in interior noise level having been evaluated for at-grade operation resulting from only one particular abatement technique. However, the reduction effectiveness observed for a particular condition has been generalized to other conditions where the abatement technique would also be useful in noise reduction. Presently TSC is beginning a large project sponsored by UMTA, which will systematically characterize the abatement effectiveness of resilient wheels, damped wheels, wheel truing and grinding, or smoothing the running rail. As a result, the noise reduction ef¬ fectiveness generated from these abatement techniques will require updating, as data from these tests become available. 2-1 The discussion of each technique contains a description of its applicability as well as additional background information, indicating the basis for the given reduction effective¬ ness of Table 3-1, in Section 3 of this report. More detailed information on the back¬ ground and tests used to determine the values of Table 3-1 are given as references where applicable. Costs to implement the noise reduction techniques are quite preliminary. Although the costs are valid only for a very short period at the time the report is published, the esti¬ mate does allow for a comparison of relative costs among the many varied approaches available from line fixes, as well as car fixes. The methodology employed (Reference 3) is uniform with the MBTA Pilot Study in that the minimum cost algorithm is the same one developed under, and applied to, that program. It will also be utilized for similar application to concurrent studies at Chicago (CTA), and New York (NYCTA) which are being conducted by other TSC contractors. 2-2 3. NOISE ABATEMENT TECHNIQUES A. DISCUSSION OF ABATEMENT EFFECTIVENESS 1. RAIL VEHICLE 1.1 VEHICLE COMPONENT NOISE (a) Ducted-Forced Ventilation of Propulsion System This technique, applicable mainly to new transit vehicles, can reduce the propul¬ sion system noise considerably, in the range of 10 to 14 dBA at the source. Changing from self-ventilated propulsion motors (used for many conventional transit vehicles), to force-ventilated propulsion motors would probably require a major overhaul of an existing transit vehicle. Since the noise from the force-ventilated propulsion system can be controlled to a greater extent than with the self-ventilated propulsion system, the noise can be controlled to be lower than the wheel/rail noise. However, in order to realize an overall reduction in noise, the wheels and rails must be in very smooth con¬ dition. The reduction effectiveness shown on Table 3-1 assumes that a transit vehicle with the forced-ventilated propulsion system will have resilient wheels, and that it operates on ground welded rail. (b) Isolation of Floor from Vehicle Superstructure Isolation of the floor from the vehicle superstructure reduces the undercar auxil¬ iary equipment and propulsion system noise and is effective for all operating modes. This isolation has been employed in the new SOAC vehicle described as "carpet underlayment: heavy vinyl laminated on foam." (Reference 4) Comparison of interior noise levels at different speeds between SOAC and BART (utilizing noise data obtained as part of this contract), while also considering the SOAC vehicle has a force-ventilated propulsion system and uses resilient wheels, indicates that a 4 to 6 dBA interior noise reduction can be achieved when the floor is isolated. Although the BART transit vehicle contains a floor designed to reduce the noise transmitted from the undercar equipment and propulsion sys¬ tem, the isolating medium used to improve the noise reduction is laid directly on the floor, rather than on top of a type of resilient or foam pad. The authors believe this differences contributes to the decrease of the interior noise level ob¬ served in the SOAC car. (c) Isolation and Balancing of Auxiliary Equipment Components Most transit vehicles with auxiliary equipment mounted undercar have some iso¬ lation between the units and vehicle. Optimizing the stiffness of the isolators to minimize vibration, changing the methods of mounting to provide for increased isolation and rebalancing of different rotational elements can improve the noise transmitted through the floor and/or decrease the noise radiated from floor vibration. 3-1 TABLE 3-1. SUMMARY OF NOISE ABATEMENT TECHNIQUES E >1 m e •c- J= to ro c TO O <-> r- t— 1/1 C <0 • 4 ) O > 1 — 0> -r- a> O J= c *0 «— ? O) co ■0 O r— 4-> C -C U 0 4-> O 4 ) >, .c in 0 •<- cn 0) -t- +J <1) ro cn m •— C 5 -m ♦*— •r- U c § > 4J C X) c c *3 «3 £ *-> 3 0 •*- <0 0J c— m *r- m 0 CL l/) r— V. C tO TJ M— U) *) Cf 4) 4-> 4-> 4-J to a) 4- ♦-> 0 U LJ CC. U — r- 5 0 0 0 qj 1 a <♦- O to <0 UJ X3f a* u qj in CL 3 r— JZ -*-> *— I_ s- g W)TJ v»- a» U- *r— -r- -O f— 4-> 03 O OJ 4 - O t«J O O > 1 -M 0 0 UJ 1 1 0 — E u c t- C S- C 0 t_J 1— - Q- =-> O -r- E .— CL ■r“ *r— ■ E 3 a; Cl +-> — E 1— CL CO _J O to Z X 3 O in X 3 t— 1 z 0 u 3 cr U >> 3 cr 0 z h— < Cl Lt z CL Cl in <£ < 1 ) UJ 1 £= O) oor UJ -0 j-> QJ M- (O U O >t 1— Ul 0 u to Z Z3 •—« 0 C UJ cr nc <+- 0 = S — 1- 0 Ul Z UJ I +j -r- h-“ Z > (O to < O ■0 1— <— £3 Ul _1 a. r— CL 4 - a -r- u "O 3 4- 0) 3 c ^ cr 0 — 4-J 03 O QJ u O c ■*- 3 C CT> !>^ O JC U 0 c il QJ 4-> •r- fO *-* > in 4-* U V- CO u 03 C f— — E QJ r~ ro -r- 0 0 Cl O *— X to U 3 m 03 3 >— 4- in ►—i -O 03 ■— c -j£ tO u JQ *0 •r— •x e to CL) I J~> C to O >, 2 : 1.1 UJ c l/) 0 ►H -r— Q in -L> «T3 CO CJ c s- 3 QJ a> —1 cr > CL m QJ O 0) r— -X i. to ai ftJ O -c- 0 S- O O —1 -C CG O Z UJ 3 UJ Z Z u to •r- co O z > C4 c >—« a> zc J-> lk: to c c to UJ c to s- cn-r- u 0J u ai •r- (O O *T— J3 r- tO e O -J r~— 4- a> 13 KH 'r~ U 4- X5 T3 _^*- r to -r- 3 O) C »4- OJ < £ a: 10 0 (X CC 3-2 01 I 0> JZ QJ j- f— r— 5 QJ r— OJ •*— »r> to JZ •»— CL,— to ro 4- qj E c s o T9 2 • X O) L- L- •r- to o o L. L- V 3 1 -a QJ O H- -M C 4- 1 » 3 § c x to C 1 c > C O QJ • 1 C O OJ 3 1 3 •r— to a> to JZ 1 t 3 UJ r— 4J O O 4J *-> 4- QJ CO O oc ■o CL o> o> •*— CO to C O 4— cl to a. to +-. +-> o V) 1 to to r—i C QJ L-> L» QJ L. UI r—» to O JZ rO 3 *<- a> cz cr> •r— JJ U _J a OJ to to 4-» o c +-> to to H- UJ o -C -C t— r— cn cn — 4- L- °r- CL C 4- C Z ZD >—« 3: 3: QJ a c e .3 O > S- L- O O °r- L- ui cr T* QJ —< * ■ 4-> 4-> SZ -C 3 3 L_ t—1 O JZ to to J-J 0) 3 U UJ z LU c C IS L. L. 3 U JZ CZ 3 C t— X > cu a 1— 4-> O to >> >> rO QJ c i— rO «c o •r* •r— •o “O •f~> QJ L- !_ s- E tO O CO UJ - I r —> r— QJ QJ r—~ i— > ro CJ ro 3 J-> L- L. «£ 1— •r> Ql CL QJ QJ 4-» QJ E JZ E tZ ro QJ O QJ to to E E OJ QJ 4- di -r- JZ »r* 3 QJ 3 0 0 a> uuur io • •a o c 2 o 3 CJ C 0 13 1 1 E -M 1 ■C > C t • i cc 1 1 CZ 1 CD *t— re I i i 1 1 CZ CD 1 lc j c 1 i i O E U 1 -r- r— LU •»— O DU CD CC ZD I_ IT) O <♦- *-» C a> a> 1 - 1 - re c c 1/1 Q. O JZ JZ 03 33 re X zu io EU co a 0 • L- O © o CVJ o in r* in Lfi J- r— #— LO CM i— o» r— r— OJ 1 1 i 1 i i 1 1 1 4 -> CO ID oo ID 00 O'. 00 o CM X r— r— r— >- LU «c cc (- ZD O o o CO OO “r— CO LO CM 1 O^ 00 r— r— L. 1 1 CO 1 1 1 • 1 1 03 VO o OO tn in • 4 U r —1 c c L—L CO •o l—l i- oo o oo •r— o LU S_ cr> LO CVJ OO f 1 i i i z a i i 1 1 1 t i i i LU +-> ID VO CO ID LU > LU X r— D£ L—L CD LU ZD f— < J— <_> CC LU o $- O U. 1 o LU ID 23 Ll_ »— • 1 — OO ID ID r— CC 1 1 i i i LU c S- L—L 1 1 CO 1 1 — 1 1 i i i Z r a> • 4 -> o CO ID OO o C z L—L 1 —L >“ 1 - < (_J i L 3 £ ZD CD LU L- O LU CO ID CO ID z _ 1 X LU zc 1 LU «£ uC o L—L m CC S_ 1 I LU o o c J- OO 00 1 ID 1 CO CVJ 1 OO 1 1 1 1 i i i i « i Of r— Q£ C ZD ID ' ►—L 1 — (— O ZD oc o £ § <-> o £ UJ. 0uJ £ I § 5 Cl. 00 Q LU o oo 00 < S- 4U r- 4-» i— «P r— *T“» -f— c o o O > U u re > o r3 > O 2 T z cr LU S. OJ r— 0 ) ^-> CD r— U> Qj r— 4-> u aj to oi c > *r- +-> re oj o 3 s- DC C O +■> jQ -a •<- 1— 0J •*— i/I 0) 01 00 C > -r- 4 -> ro 0 0 3 1 - CC CZ O 4 -> -O re i- a» -t-> aj i/i to > -r- -I— o o c CC CZ -T- Q c o Q LU 4 -J ra IA 4 D re VA LU •*- CJ r— 03 »> c CD cn CD 4 -> CO u a» < 1 » c ■LU •r» ■LU 2 •r» *r» Z ZD «£ r— -L- c JD 4 -> c « 4 -> lu cr o •r* -o U +-> 1 — o 3 L. •f* 0 > ZD •r* ra s: — w l—l 1 ^ c c c u > 4 -> U •a > CO 4 -> LU Z Qu c: -L— LO *r— JO lX •r— c -a “O •r— to T 3 *T 0 o a) -r—j •J X CL •r— u -U VA -Q r— CL •I- o IA -Q >> CO LU u JZ u O c o •LJ +J r— U £ r—• •r— •r® *r* •r” cc C ra r— to c: LO z re r - ra X 5 C£ L^L L. 00 L. 13 X> OJ -LU ra t- 3 H 3 : QC s: u DC t— < E 2 *-> CO -M •*-> c E (A 3-4 TABLES—1. (Continued) i- > , LU O 4-J O *J •r- G> ■r- CO *r— c +-» +J >, C •*-» P-> >, C 1 y iO nj 3 U ID Ifl 3 <1) •<- 5 F= < o .O C co uj o T3 tO P>£ •«- *=C J— *C C r— U to /V 33 i— to 4 ~> u> O fO U 0/ h- go 3 : k— W / 3-5 Examination of the floor vibration in the BART car (Reference 5) indicates that decreasing the vibration of the floor generated by operation of the undercar aux¬ iliary equipment would result in a 3 to 5 dBA decrease in the interior noise level when the car is stationary. When the car is operating, particularly at high speed, propulsion system noise and wheel/rail noise predominate, thus the isolation and balancing of auxiliary equipment provide little benefit when the car is operating. It is expected that similar noise reduction could be obtained on other existing systems and with new transit vehicles if attention is paid to the mounting and balancing methods of the undercar equipment in the design stage. (d) Non-Skid Braking System A non-skid braking system is a braking system with a slip-spin or skid control system, utilizing maximum friction or adhesion between the wheels and rails. During deceleration, if slower than normal wheel speed is detected, braking ef¬ fort is decreased at that wheel until normal wheel rotation is restored. Such a system is used on the BART car. The main advantage of the non-skid braking system is in the reduction in number and severity of wheel flats and rail burns which can add considerably to the overall noise, both within and outside of the transit vehicle. The reduction effectiveness attributed to the utilization of a non-skid braking system is that which would be produced by a transit system having no wheel or rail flats produced from the skidding or slipping of the wheels on the rail. (e) Air Brake Vent Mufflers Installing mufflers on the brake vents of transit vehicles utilizing air brakes can provide a substantial decrease in the noise produced by the escaping air as the transit train stops at a station. The mufflers would provide a large decrease in the overall noise due to the predominance of this vent noise when the air escapes, experienced by commuters on the station platform near the approaching trains. The mufflers would not provide much of a decrease in noise to those inside the vehicle, since the noise is generally not intrusive inside the vehicle, due to other noises associated with older transit vehicles. Some benefit could be realized inside the vehicle if windows or doors are open at the time of the air exhaust. (f) Redesign, Repair and Maintenance of Doors An annoying and intrusive interior noise can be caused by the operation of the transit vehicle doors. The noise level produced from the operation of the doors can usually be decreased with repair and maintenance of older doors. Results obtained on other transit cars indicate that a 5 to 8 dBA decrease in door opera¬ tion noise can be achieved. The interior noise in the BART car is approximately 15 dBA greater near the door during door operation than when the car is stationary with auxiliary equip¬ ment energized. 3-6 1.2 WHEEL/RAIL NOISE (a) Resilient Wheels Resilient wheels are wheels which provide some isolation between the wheel rim and its center hub or web section. Two types have been tested on rapid transit systems, the Penn Cushion or Bochum wheel, and the Acousta Flex wheel. In the Penn Cushion wheel, the isolating section is made up of a number of closely spaced rubber blocks, while the Acousta Flex isolating section consists of a con¬ tinuous rubber layer with a saw-tooth cross section. For the purposes of this dis¬ cussion on resilient wheels, based principally on the similarity of data obtained at BART (Reference 6), the reduction effectiveness shown on Table 3-1 is ap¬ plicable to either type of resilient wheel. Resilient wheels have been shown to be extremely effective in suppressing or eliminating wheel squeal on short radius curves. Results obtained with a BART car equipped with Bochum wheels showed an overall noise decrease in subway of up to 18 dBA inside the car, and 16 dBA outside the car on the safety walk, for a 540 ft radius curve. Measurements made at the CTA loop, with very tight radius curves on elevated structure, indicate that when traversing these short radius curves, resilient wheels reduce the interior noise level on the order of 10 to 15 dBA (Reference 7). This noise level reduction is applicable primarily to wheel squeal. Wayside measurements made at-grade for a small Personal Rapid Transit (PRT) vehicle indicate that resilient wheels decrease the wheel squeal by 10 to 15 dBA (Reference 8). Although some of the wayside noise measurements for both the BART and PRT tests were made within 4 or 5 feet of the transit vehicle near the wheel/rail interface where the wheel squeal is extremely severe, the decrease in squeal correlates well with the decrease in overall noise due to the predominance of the wheel squeal over the other noise sources when it occurs. The overall reduc¬ tion effectiveness of resilient wheels in reducing wheel squeal is shown in Table 3-1. The least amount of noise reduction occurring on low squeal, relatively gradual curves, and the maximum reduction in noise occurring for high squeal, relatively tight curves. Although no large-scale tests have yet been performed to assess the effect of resilient wheels on the reduction of impact noise at rail joints and/or switch gaps, experimental "drop tests" with a resilient wheel (Reference 8) indicate that resilient wheels on rapid transit vehicles should give the reduction indicated on Table 3-1 for impact noise. As part of the BART tests with resilient wheels (Reference 6), both Bochum and Acousta Flex wheels were tested for their reduction of interior and wayside noise on ballast and tie tangent track. The decrease in noise that is attributable to resilient wheels in this case is a decrease in roar noise. With each wheel on ground welded rail, a decrease of up to 2 dBA for both interior and exterior noise over a number of different conditions was observed. On quiet propulsion 3-7 systems, this decrease could conceivably be somewhat greater. Although no measurements were made on either aerial structure or in tangent subway, the de¬ crease could conceivably be somewhat greater due to the reflecting properties of the concrete trackbed which is used (on BART) for these transit structures. (b) Damped Wheels Transit car wheels have been damped using both free-layer and constrained-layer damping treatments, and a combination of the two. The different treatments were found to be effective in reducing noise, especially wheel squeal on curves (as resilient wheels). A particular damping treatment is usually designed within the specifications and restrictions applicable to a particular transit system. Measurements with damped wheels were also made at BART under the same con¬ ditions as those for interior and wayside measurements with resilient wheels. Results obtained indicated an overall noise decrease in subway of up to 11 dBA inside the car, and up to 6 dBA outside the car on the safety walk for a 540 ft radius curve (Reference 6). Measurements made at TTC indicated a 12 to 15 dBA decrease in the wayside noise at 5 ft from the train on a curve of 200 ft radius (Reference 9). Measurements made with the same PRT vehicle on the same curve (discussed for resilient wheels) also show a 10 to 15 dBA reduction in wheel squeal as with resilient wheels. Thus, damped wheels appear to have an equal effectiveness with resilient wheels in reducing both the interior and exterior noise due to wheel squeal. Further noise measurements at BART with damped wheels indicates, when com¬ pared to resilient wheels, they are equally effective in reducing roar noise. Al¬ though little information is available, the indication is that damped wheels are approximately half as effective as resilient wheels in reducing impact noise (i.e., provide half the dB reduction that resilient wheels provide). (c) Wheel Truing The process of wheel truing is similar to rail grinding. Truing wheels smooths the wheel surface and removes flat spots and imperfections which may develop during the transit vehicles operation. Truing the wheel results in a decrease in roar noise which arises from the wheel and rail roughness. Assuming that the wheels have no flats and that the trains are operating on welded ground rail, the decrease in the noise level will be on the same order as the decrease in roar noise using either resilient or damped wheels, up to a 2 dBA reduction for either interior or ex¬ terior noise at-grade. Similar noise reductions are also expected for aerial struc¬ ture and subway operations. If the wheel has a gross imperfection such a a wheel flat, wheel truing can give a significant decrease in the impact noise which results each time the wheel flat contacts the rail. Depending on the severity of the imperfection, wheel truing can give up to a 10 dBA reduction in both interior and exterior noise. Elimina¬ tion of a wheel flat is analogous to replacing relatively poorly maintained jointed rail with welded rail except that in the case of a flat wheel, the impact noise would occur at each wheel rotation. 3-8 (d) Soft Journal Sleeves Tests were made at CTA on a pair of 2000 series cars with modified bogies. The rubber journal sleeves on the bogies were replaced with softer journals to create the effect of primary springing on a bogie which does not incorporate primary springs. The reduction in stiffness of the journal sleeves was 30 to 1. Table 3-1 gives the noise reduction obtained for tests of the interior and wayside noise with the soft journal sleeves (Reference 7). The softer journals apparently increase the total damping in the system and decrease the vibration transmitted from the wheels and axles to the truck frames and car body. The result was an improvement in the noise levels in the cars. The softer journals also apparently decrease the impact forces which occur at the wheel and rail interface, thus de¬ creasing the vibration level of the wheels and rails resulting in less noise radiated from the vibration of the wheels and rails. It should be noted that if soft primary springs are used on other systems with concrete aerial structures, rather than the old steel aerial structures used at CTA, the noise reduction on aerial structure would be greater than shown on Table 3-1. (e) Sound Absorption Treatment on Bottom o f Car Floor and Inside Face of Car Skirts " Absorption material applied on the underside of the car floor, especially above each bogie, can be effective for reducing the interior noise level. When used in conjunction with the sound barrier wall, an improvement in the exterior noise level also results. Tests on a BART car with approximately 140 square feet of absorption material located over each bogie showed a noise reduction in the car of 3 to 4 dBA at-grade, and 2 to 3 dBA in subway. The wayside noise re¬ duction of 1 dBA was achieved at-grade without the use of a sound barrier wall (Reference 6). With the use of a sound barrier wall, an additional 3 to 4 dBA of noise reduction can be achieved over what would normally be achieved with only a sound barrier wall and no undercar absorption treatment. Addition of this undercar absorption material is relatively inexpensive both in terms of materials and added weight to the car, and can give significant benefit in terms of reducing the overall noise exposure for both patron and wayside community adjacent to the transit system alignment when a sound barrier wall is used. A moderate skirt extension used in conjunction with a sound barrier wall could prove to be an extremely effective method of reducing wayside noise. The use of car skirts could decrease the required height of sound barrier walls when needed at some noise critical areas. Although no cars have been designed with deep skirts for noise control purposes, there was at one time a consideration of ex¬ tending the car skirts on the BART car as a retrofit for noise control purposes. 3-9 2. TRACK ROADBED AND ASSOCIATED STRUCTURES 2.1 WHEEL/RAIL NOISE (a) Welded Rail and Rail Grinding The use of welded running rail eliminates the impact noise produced at rail joints as the transit vehicle wheels pass over the rail joints. Mechanical grinding or smoothing of the continuous welded running rail has been found to give signifi¬ cant noise reduction for transit train operations on tangent track. Wayside noise measurements made at CTA at-grade facilities indicate that a 5 to 7 dBA reduction in wayside noise was achieved with standard trains operating on new unground continuous rail, compared with noise produced from standard trains operating on joined rail. Further measurements made at CTA indicate that rail grinding further decreases both wayside and in-car noise levels. Wayside measurements made at both 25 feet and 50 feet from ballast and tie at-grade track after rail smoothing, indicate a 6 to 10 dBA reduction in wayside noise with smooth continuous welded rail, over noise with worn or corrugated jointed track. A wayside noise reduction of up to 14 dBA was obtained with CTA transit trains that had soft journal sleeves (as previously described) when operated on smooth continuous welded rail, compared to identifical trains operating on jointed rail (Reference 7). Measurements made at BART indicate that grinding the welded running rail resulted in a 6 to 9 dBA reduction of passby noise for operations on ballast and tie welded tangent track with standard wheels (Refer¬ ence 6). Interior noise measurements made at CTA indicate that trains traveling on either unsmooth continuous welded rail or smooth continuous welded rail produce on the order of 3 to 5 dBA less interior noise than for operation on ballast and tie jointed track. BART interior noise measurements indicate a 4 to 5 dBA reduc¬ tion of car interior noise with ground rail over unground rail for operation on ballast and tie tangent track with standard wheels. For subway operations, BART train operations on ground rail produce similar noise reductions: 4 to 5 dBA reduction of car interior noise, and 5 to 12 dBA reduction of in-tunnel or sta¬ tion noise with standard wheels. Based on the measured noise data for both at-grade and subway operations, the reduction effectiveness of welded rail and ground rail for concrete aerial struc¬ ture operations is shown in Table 3-1. (b) Maintenance of Rail Joints For existing transit systems with jointed running rail, maintenance of the rail joints to a particular tolerance should provide some wayside and interior noise reduction. The actual amount of noise reduction would depend on both the severity of misalignment and the degree of alignment possible and practical at each joint. The wayside noise reduction expected would be somewhat less than that obtained on unsmooth continuous welded rail, thus a 3 to 6 dBA noise reduc¬ tion could be expected if the rail joints are maintained to some particular tolerance. 3-10 Recent experimental scale model tests made with rail joints having height differ¬ ences (Reference 8), show that for travel in the step-up direction, the peak sound pressure level increases monotonically with increasing train speed. For travel in the step-down direction, the peak sound pressure level at low speeds coincides with the speed obtained for the step-up direction. At high speeds the sound pres¬ sure level versus train speed levels off, indicating the existence of a critical train speed, above which the wheel separates from the rail. This indicates that in main¬ taining rail joints, if a level joint cannot be maintained it would be desirable to align the joints so that transit trains would travel in the step-down direction under normal operating conditions. (c) Rail Lubrication Tests on short radius curves show that lubrication of the running rails reduce the wheel squeal noise levels on the order of 15 to 20 dBA. Measurements made on the platform of the PATH Old Hudson Terminal near the 90 foot radius subway curve indicated a 19 dBA reduction in squeal noise on the platform, with a 33 dB maximum reduction in the 2500 Hz one-third octave band when using the water spray lub¬ rication system for this curve (Reference 10). Although no car interior noise measurements were made during this series of platform noise measurements, a 19 dBA reduction in the squeal noise on the station platform indicates that lub¬ rication of the running rail does provide a significant noise reduction inside the car. Some problems have been associated with lubricating the rails with regards to traction and braking, however, at the PATH installation no traction or braking problems were reported. 2.2 GENERAL WAY STRUCTURES (a) Absorptive Treatment in Tunnels with Concrete Trackbed Transit trains operating in subway tunnels can produce a high noise level creating reverberant sound in the tunnel space which is transmitted into the interior of the train. When concrete invert with direct rail fixation is used, the increase in noise for tunnel operations is substantial because the interior of the tunnel surfaces are all hard reflecting surfaces. The enclosed space is highly reverberant, and the re¬ sult is a buildup of sound level because of the low rate of sound energy absorp¬ tion at the surfaces. The application of sound absorbing materials to the interior of subway tunnels reduces the reverberant noise level outside the transit car as it passes through the tunnel, reducing the noise inside the transit train, and hence the noise heard by commuters. This same sound absorbing material can also provide reduction of noise caused by transit vehicles transmitted along the tunnel to station platform areas, and can also reduce the noise transmitted from transit vehicles to vent shafts. Absorption treatment applied to subway tunnels has a beneficial effect for the commuter while riding the transit train, waiting for the trains on station platforms, and in community areas near vent shafts. 3-11 The WMATA subway system has made extensive use of the spray-on material called Pyrok. The nominal thickness of the treatment used is 1 inch, with the material covering approximately one-quarter of the tunnel surface area on the lower portions of the side wall surfaces. This same spray-on acoustical material has also been used at vent shafts, with the material covering up to 100 percent of the available surface area in the vent shaft. It should be noted that there are several spray-on mineral fiber materials which are effective as sound absorbing materials and meet the requirements of tunnel installation for reasonable mechani¬ cal durability, fire resistance and the ability to withstand cleaning. Tests made in the WMATA Metro subways indicate that the Pyrok spray-on acoustical treatment reduces the reverberant noise in the tunnel by 7 to 9 dBA (Reference 12). By reducing this reverberant noise impinging on the walls and roof of the car, a net reduction of car interior noise of 5 to 7 dBA can be realized with the WMATA Metro transit vehicle. For cars with equipment and wheel/rail noise more easily transmitted through the floor, characteristic of many older transit cars, the higher noise levels due to the greater amount of energy transmitted through the floor prevents achieving a similar degree of car interior noise level reduction. It should be noted that the sound absorption treatment ap¬ plies to the walls of the subway gives on the order of only 1 or 2 dBA reduction of the undercar noise level, so that most of the reduction in interior noise is due to reduced noise through the car walls and roof. A noise reduction of 3 to 7 dBA inside the cars of subways is quite favorable for commuters riding in the cars. However, the effective noise reduction of absorp¬ tion treatment is even greater when the effect for commuters standing on the platform areas is considered. The noise reduction achieved at the end of the platform as the train approaches the platform area is a combination of the noise reduction at the source, and the noise reduction with distance along the subway, so that the amount of noise reduction differs with location of the train and the arrangement of the sound absorption treatment. The noise reduction at the platform is greatest for trains operating in a treated section of the subway be¬ cause a substantial part of the total noise reduction is the source noise reduction in the treated section. Further tests at the WMATA Metro subway indicate that substantial noise reduc¬ tion is achieved through vent shafts when they are lined with the spray-on acoustical treatment. Comparison of two vent shafts, of somewhat different con¬ figuration, one with and one without the spray-on acoustical treatment indicates that a 13 to 15 dBA reduction in train noise can be achieved at the surface of the vent shaft when both the tunnel and shaft are treated with the 1 inch thick Pyrok spray-on material. Additional beneficial noise reduction can be achieved with 2 inch thick treatment instead of the 1 inch thick treatment which was used in the shafts where the tests were made. (b) Ballast and Tie Trackbed in Tunnels The use of ballast and tie trackbed instead of concrete invert with direct rail fixation in tunnels gives the same or greater noise reduction as a concrete trackbed 3-12 tunnel which has been treated with sound absorbing acoustical material. A ballast layer is a sound absorber located in the optimum location for reducing noise in a transit tunnel when ballast and ties are used, thus the effect of reflec¬ tion and reverberation in the tunnel is considerably diminished. Measurements made at the CTA transit facilities (Reference 7) indicate that transit trains operating at 30 to 40 mph produce between 8 and 11 dBA less exterior noise for operation on ballast and tie trackbed when compared with those operating on the concrete trackbed. This translates into a 4 to 8 dBA re¬ duction in interior noise, with the degree of reduction dependent upon the trans¬ mission loss characteristics of transit vehicle as discussed in the section on absorptive treatment in tunnels. (c) Absorptive Treatment in Subway Stations The application of acoustical treatment to the interior surfaces of transit stations and to the underplatform areas adjacent to the transit cars makes it possible to substantially reduce the noise due to all sources in the transit stations and par¬ ticularly to reduce the noise due to transit train operations in underground stations. Basically, the inclusion of acoustic treatment in a transit system station accom¬ plishes four major objectives: (1) Control and reduction of noise from transit train operations, (2) provision for good intelligibility of announcements from the public address system, (3) control of general crowd noise generated by commuter talking and walking, and, (4) assistance in the control of noise from the station's mechanical equipment. The treatment applied in the station reduces the noise due to a combination of absorption at the source (through reduction of the sound energy at the first re¬ flection) combined with reduction of the reverberation. The absorption of train noise at the source is most effectively accomplished by the underplatform absorp¬ tive material. The amount of control of reverberation and the consequent reduc¬ tion of noise is dependent upon the area of the acoustical treatment, the absorp¬ tion coefficient and the placement of the treatment. Tests were made at the WMATA Metro system to determine the effectiveness of the station acoustical treatment (Reference 11). The effectiveness was determined by measuring the reverberation times in one finished station with complete acoustical treatment, and one unfinished station complete structurally, but with¬ out acoustical treatment. The sound absorbing material installed in the treated station consists of panels made up of 2 inch thick glass fiber boards encased in plastic sheeting with a perforated sheet metal facing. These absorbing panels are located in fourteen rows of the vault coffers which are located above the platform 3-13 and track areas, the underplatform areas, and on the underside of the mezzanines. The tests, made with cannon shots and corrected for a typical transit train spec¬ trum in the station, indicate that the treated station has 6 to 8 dBA less noise due to the reduction of reverberation time only. Measurements made at TTC and CTA indicate that the actual noise reduction achieved in a transit station is actually somewhat greater than the expected or calculated reduction, considering only the reduction of reverberation time. If the effect of a reduction of the noise due to the absorption at the source is also considered, the predicted noise reduction agrees favorably with the observed con¬ ditions on other systems. Due to underplatform absorption, noise is not reflected from the underplatform area directly to the upper surfaces or ceiling, but absorb¬ ed, and the reverberant sound in the platform area tends to be composed of less - direct and more diffuse reflected sound. Thus, the net noise reduction due to treatment applied to the WMATA subway station is 10 to 12 dBA. BART subway stations all have extensive applications of sound absorbing ma¬ terial on the ceilings and underplatform spaces. The result is a more acceptable noise level from the transit train operations than found in older systems which have completely untreated highly reverberant stations. The Toronto Transit Commission has had good success with acoustic treatment. At one station where a suspended acoustical ceiling was installed over the tracks, and sound absorption placed on the underplatform area, the noise reduction obtained was about 13 dBA average. The absorptive treatment in subway stations can also decrease the noise experi¬ enced by commuters inside the train while it is in the station. The actual noise reduction experienced inside the train depends on several factors, but in general, can range from 5 to 10 dBA. No reduction in noise is apparent inside the transit train when all doors and windows are closed, and another train is in operation on an opposite track. When the doors of the stopped train are open, noise reduc¬ tion can be realized, depending on the noise level of the stationary train's aux¬ iliary equipment. In the case where the stationary train's auxiliary equipment noise level is quite low, the maximum noise reduction (up to 10 dBA) could be experienced by those commuters in the transit train near the door opening. However, if the stationary train has relatively noisy auxiliary equipment, then the benefits of station acoustical treatment would be minimized because of the dominance of the noise from the auxiliary equipment as perceived by the com¬ muters near the transit train's open door. In this case very little, if any, noise reduction would occur for the benefit of the commuters on the train. The optimum location for the sound absorption treatment for reducing train noise is on the underplatform edges, opposite the vehicle bogies, and on the side walls opposite the platform. Sound absorption material can also be applied to the ceilings and walls of mezzanine areas. In general, treatment, such as glass fiber boards or blankets, which are both acoustically efficient and sufficiently durable is acceptable for use as a sound absorbing material in reducing the noise in sub¬ way stations. 3-14 (d) Ballast and Tie Trackbed in Subway Stations Subway stations with ballast and tie trackbed like ballast, and tie tunnels have an important noise control feature in the absorptive properties of the ballast. Noise measurements were made at the CTA facilities at similar stations with some having concrete trackbed and others having ballast and tie trackbed (Reference 7). The results of the measurements indicate that the net effect of the sound absorption contributed by the ballast in the ballasted track section is to reduce the noise level from the transit trains on the platforms by 12 to 15 dBA. This is a very signifi¬ cant reduction of the noise and indicates the benefit of providing absorption at track level, or absorption at the source to reduce the noise exposure of commu¬ ters in the stations. The benefit of the ballasted trackbed in reducing the interior noise inside the transit train follows the same discussion as previously presented in the section on absorptive treatment in subway stations. The range of noise reduction to the commuter inside the transit vehicle is in the range of 5 to 13 dBA, with the maxi¬ mum benefit occurring when the auxiliary equipment of the transit train has very low noise levels. (e) Sound Barrier Walls The use of a low sound barrier wall at the side of both surface and aerial way struc¬ tures has been found to be an effective means for reducing the wayside noise exposure due to transit train operations. The data obtained from wayside noise tests for operation with sound barrier walls, for measurements made at both the BART Test Track (Reference 12), and with different configurations at the BART revenue aerial structure (Reference 13), both with and without absorptive ma¬ terial added to the track side of the sound barrier wall, indicate a noise reduction effectiveness at 50 feet of 5 to 7 dBA for the non-absorptive sound barrier wall on aerial structure, and 8 to 10 dBA for the absorptive sound barrier wall on aerial structure. For at-grade operations where the trackbed is ballast and tie, the ballast acts as an absorptive medium and thus the use of absorbing material on the track side of the sound barrier wall is less effective. The noise reduction at 50 feet for a non-absorptive sound barrier wall for at-grade trackbed is 8 to 10 dBA. It should be noted that in order to achieve the indicated noise reductions, a number of parameters should be optimized (within the limitations of the particular transit system). Among the important parameters are: effective height of the wall, mass and material of the wall, spacing between the car and the wall, thick¬ ness of the sound absorption treatment, car skirt depth and the shape of the car side wall. Inside the transit vehicle, the sound barrier wall will generally have no effect on the interior noise level assuming a transit car with closed windows and adequate transmission loss between the exterior and interior areas. Thus, the sound barrier wall is an effective means of reducing wayside noise only, and has no effect on the noise experienced by the commuters inside the transit vehicle. 3-15 (f) Trackbed in Retained Cut If an "at-grade" line can be put in a depressed cut, the effect of the depression of the trackbed can be equal to or more effective than a sound barrier wall adjacent to the alignment. The direct radiation of the noise from the transit train opera¬ tions is blocked by the banks or edges of the cut. Although most transit lines do not allow for long sections of the line which could be put in a cut, in areas where this alternative is available. Putting the alignment in a retained cut can decrease the wayside noise by 8 to 14 dBA depending on the depth of the cut and slope of adjoining walls. This is an improvement over the use of a sound barrier wall, if the line was left either at-grade or placed on aerial structure. As with the sound barrier wall, when the alignment is in a retained cut, the noise inside the transit train is unaffected because the retained cut's noise control properties benefits the wayside community, and not the commuters riding inside the transit vehicle. B. APPLICATION TO RAIL TRANSIT SYSTEMS The noise abatement techniques presented in Section 3-A, Discussion of Abatement Effectiveness, have been reviewed for applicability to each of the systems under study. Each of the techniques are listed by their effectiveness regarding vehicle noise reduc¬ tion and trackbed and way structure noise reduction for each transit line. 1. BAY AREA RAPID TRANSIT 1.1 DESCRIPTION OF TRANSIT SYSTEM Routes and Service — The BART system consists of four branches in the San Francisco Bay Area — Fremont, Concord, Richmond and Daly City. The system is 75 miles long and has 34 stations (33 presently in operation). Revenue operation began in 1972 with the opening of service between Fremont and MacArthur Stations. Service was extended to the Richmond Station in early 1973. Later that same year, service was initiated between the MacArthur Station and the Concord Station and between Daly City Station and Montgomery Street Station. The final link through the Trans- Bay Tube was opened in 1974. Engineering Features — The design of facilities and equipment for new transit systems, such as the BART system, include many features intended to produce lower noise and vibration levels than those traditionally expected for rail transit systems. The BART system contains many of the general and special design features which are intended for and do result in lower noise. The results from the BART system indicate considerable success in achieving lower noise operations with noise levels experienced at BART facilities being typically in the range of 20 to 30 decibels (dB) less than have been or are experienced at older facilities or older systems where noise and vibration were not considered as important or limiting design parameters. In the planning and design of the BART system facilities and equipment, data obtained from various operational and experimental transit vehicle structures and systems were used to determine the noise characteristics to be expected. Using the known noise characteristics of the best state-of-the-art components, equipment or facilities, and 3-16 estimating the improvements in performance which could be made by changes in the design, through the use of noise limit specifications on the equipment and as deter¬ mined from experimental vehicles, projections were made of the expected performance for the system and facilities. The results obtained with the BART vehicles and facilities are very close to the performance expected. Roadbed — There are several varieties of way structures, track designs and station de¬ signs included in the BART system. The following table indicates the main categories of the various types of facilities: (a) Above-Ground Track Structures (1) Ballast and tie at-grade tracks (2) Concrete aerial structure (3) Composite steel concrete aerial structure with trapezoidal girders (4) Composite steel concrete aerial structure with I-beam girders (5) Ballasted deck bridge (b) Underground Structures (1) Concrete double box section (2) Single box section (3) Concrete round tunnel (4) Steel-lined round tunnel (5) Concrete double box section with ballast and tie track (c) Additional Features (1) Continuous welded rail (2) Concrete aerial structures with resilient rail fasteners (3) Resilient direct fixation rail fasteners in subways (4) Use of a rail grinder for smoothing the rails before commencing revenue operations and for maintaining the rails in smooth condition With concrete trackbed and resilient rail fasteners Rail Vehicles — There are two configurations of the BART car. The "A" car contains the operator's cab and the automatic train operation equipment. The "B" car is identical to the "A" car except it does not contain the cab or automatic train opera¬ tion equipment. A standard revenue train consists of anywhere from two "A" cars to two "A" cars and eight "B" cars. The system schedule speed is approximately 45 mph with a maximum speed of 80 mph. The trains run on wide gauge track (5'-6"). Each axle of each car is powered, receiving the power from the 1,000 volt DC third rail. Each car is air conditioned by a 12-ton capacity refrigeration evaporator with automatic 3-17 heating and cooling control. With respect to noise and vibration, the BART car was designed with the following considerations: (1) Car equipment noise and vibration level performance limits (2) Car body sound insulation performance requirements (3) Lightweight trucks with minimized unspring weight, with rubber mounts and inserts for vibration isolation and with a low noise braking system (4) Use of wheel grinders and lathes for maintaining the wheels in smooth condition. Stations — Stations on the BART system can be grouped as follows: (a) Above Ground Stations (1) At-grade center platform in residential/commercial area. (2) At-grade side platform in: Industrial/commercial area Residential area. (3) Aerial center platform in: Freeway median Commercial area Residential/commercial area. (4) Aerial side platform in: Commercial area Residential/commercial area Residential area. (b) Underground Stations (1) Center platform, two track single level. (2) Single track multi-level. Sound absorption treatment has been included for subway stations. The Bay Area Rapid Transit system (BART) began revenue operation in 1972. A number of noise control features discussed in the Noise Abatement Techniques sec¬ tion have been used by BART for reducing the noise experienced by commuters and the wayside community. The design features or characteristics intended to reduce noise include: Sound absorption treatment in subway stations Use of a rail grinder for smoothing the rails Limited use of a “partial” sound barrier wall Car equipment noise and vibration level performance limits Car body sound insulation performance requirements Non-skid braking system Truing of wheels by wheel grinders and lathes 3-18 With the incorporation of these noise control features, the car interior noise levels com¬ pare favorably with those in private automobiles, and the wayside noise levels are lower than current trucks and buses. The actual noise levels of BART for wayside, interior and station platform locations are presented as part of this project in "Noise Assessment Report — Volume 1 — Bay Area Rapid Transit District" (Reference 1). Although BART has incorporated a significant number of beneficial noise control features to achieve the present noise levels, several additional noise abatement techniques are applicable to BART and would further reduce the noise levels from BART operations. 1.2 VEHICLE NOISE REDUCTION As previously mentioned, the BART car was designed with specific noise and vibration level performance limits and with car body sound insulation performance requirements. Although the car met most of the noise and vibration requirements of the specifica¬ tions, redesign of certain elements would decrease the interior noise, especially when the car is stationary. To decrease the interior noise radiated from the floor during the operation and vibra¬ tion of the undercar auxiliary equipment, redesign of the auxiliary equipment suspen¬ sion would be beneficial. A basic review encompassing the following would provide information leading to an optimum design: (1) Examination of the stiffness and elastomer material type for isolators used in mounting the undercar equipment, (2) Examination of alternate methods for mounting the undercar equipment to provide for increased isolation, (3) Examination of the balancing of the different rotational elements, (4) Examination of bearing condition. An additional decrease in the interior noise would be realized for stationary and low speed operation with a redesign of the air conditioning system air ducts, and a balanc¬ ing of the air conditioning system air flow. This would decrease the air flow noise, par¬ ticularly the air flow whistle encountered in some cars. Tests with BART Car 107 (Reference 6) indicated a significant decrease of interior noise (3 to 4 dBA at-grade), when 140 square feet of acoustical absorbing material was placed on the underside of the car floor over the bogies. Installation of this treat¬ ment on BART's fleet of 450 cars would be relatively easy and inexpensive and would provide a significant reduction of noise, especially inside the car. 1.3 WHEEL/RAIL NOISE The present fleet of BART cars is equipped with standard wheels with an aluminum center web. Tests also performed with Car 107 indicate up to an 18 dBA decrease in wheel squeal inside the car on a 540 ft radius curve. This curve is within 10 feet (in radius) of the tightest curve on the BART system. The maximum level of squeal would 3-19 tend to occur here. There are approximately 3 or 4 curves on BART where wheel squeal is encountered. The wheel squeal at these locations could be effectively eliminated if the standard wheels were replaced with resilient wheels. Since BART has ground welded rail, the roar noise produced at the wheel/rail interface, especially for subway operations, would also decrease. Impact noise on BART occurs at the rail gaps of switches at turnouts and crossovers. Although the impact noise would occur more often if BART had jointed rail, the effect at the switches would be significant. Since the BART cars are relatively new, with significant life remaining on the existing wheels, a program of phasing out the standard wheels after a certain amount of wear, and replacing them with resilient wheels could be implemented. TRACKBED AND WAY STRUCTURE NOISE REDUCTION Presently there are two locations on the BART Richmond Line where a "partial" sound barrier wall has been erected. The sound barrier wall has been installed along the transition structure from subway to surface, and on the aerial structure for a short dis¬ tance at the two subway to aerial transitions in Berkeley. The installed sound barrier wall is a vertical concrete wall without absorption treatment, spaced out from the aerial structure in a manner that creates a slot nominally about 4 inches in width between the aerial structure and the sound barrier wall at the bottom of the wall. The wall has narrow openings between the different concrete slabs. These sound barrier walls as installed are not of an optimum design or configuration and do not give the maximum wayside noise reduction which would be achieved with an improved config¬ uration. The non-optimum configuration is the reason this sound barrier wall is referred to as "partial". With an "optimum" sound barrier wall configuration, the 8 to 10 dBA reduction can be achieved for an absorptive sound barrier wall on aerial structure and a non-absorptive sound barrier wall adjacent to at-grade tracks. Considering typical maximum noise levels adjacent to the BART line at 50 feet for 4- to 6-car trains traveling at the maxi¬ mum speed of 80 mph, the maximum noise level for at-grade operations would decrease from 87 dBA to 77 dBA. The maximum noise level for aerial structure operations would decrease from 92 dBA to 85 dBA with the non-absorptive sound barrier wall, and to 82 dBA with the absorptive sound barrier wall. These are substantial decreases in the wayside noise, and would provide an improved sound environment in the com¬ munities adjacent to the BART alignment. If the sound barrier walls were installed along the entire length of aerial and at-grade tracks, excluding sections where the alignment is in freeway median, it would result in sound barrier walls on approximately 21 miles of aerial structure, and approximately 15 to 16 miles of at-grade tracks. Installing sound barrier walls over so much distance would be a large-scale undertaking. A more economical approach would be to install sound barrier walls in the most noise sensitive areas first, and over a period of time install sound barrier walls throughout the system where a decrease in the BART passby noise would benefit the wayside communities. Installing sound barrier walls in a free¬ way median or in a strictly industrial area would provide little benefit to the wayside community. Another line treatment applicable to BART is the absorptive treatment in tunnels. Most of the BART subway tunnels have a concrete trackbed with steel tunnel liners, while the Trans Bay Tube has a concrete trackbed with concrete lining. For all con¬ figurations of tunnels, the application of sound absorbing material to the interior of the tunnels would decrease both the noise inside the BART car, and the noise transmitted through the tunnels to the station platform. A spray-on mineral fiber with effective acoustical properties while meeting BART's requirements for maintenance and safety could be applied to the tunnel walls during the nights and weekends when BART (presently) does not operate revenue service. This would require treatment for approximately 20 miles of tunnel structure. As treatment applied in all of the tunnels would be an extensive undertaking, treatment applied in long sections of tunnel where maximum speed and hence maximum noise is attained, such as the Trans Bay Tube or Berkeley Hills Tunnel, would provide the greatest benefit to the commuter inside the BART car. 1.5 BAY AREA RAPID TRANSIT DISTRICT LINE SUMMARY General The following section summarizes the data obtained at selected community, station platform and in-car locations. Tables summarizing the results obtained from these measurements have been included. The wayside noise reported is an average of the maximum noise levels (L^max) ob¬ tained for each near train passby at a distance of 50 feet from the near track center- line. The station platform noise reported is an average of the maximum level (L^max) obtained for each train's arrival and departure at a position in the center of the station platform 2 meters from the platform edge. The interior noise data reported represents the maximum plateau noise level observed at the car center for operation over a par¬ ticular type of track structure. For the wayside community noise, no attempt was made to calculate L^p since BART presently (1975) does not operate at night. Thus noise measurements were made only during the daytime, rush hour and evening. As BART presently has plans to commence nighttime operations in 1976 (after the conclusion of this project), it is felt that the calculation of an L^p without the nighttime data would not be valid in light of future operation. Community Noise Wayside community noise measurements were made at 13 representative scenarios as shown in Table 3-2. The scenarios were chosen to be representative of residential and commercial areas which contain the different types of track structures on which the BART trains operate. Measurements were made along different lines to assess the effects of operation frequency and train length. Table 3-2 also summarizes the re¬ sults obtained at each wayside location. The following general observations can be drawn from the noise data: 3-21 RIGHT OF WAY (1,000'S OF FEET) NUMBER OF STATIONS FEET OF TRACK (1,000'S OF FEET) AERIAL STATION PLATFORM NOISE LEVELS - AVERAGE OF ENTERING AND DEPARTING PEAKS (dBA) OF NEAR TRACK TRAINS MAXIMUM PASS-BY LEVEL AT 50 FT (dBA) TABLE 3-1. SUMMARY OF BART NOISE ENVIRONMENT 3-22 TABLE 3-2. SUMMARY OF WAYSIDE NOISE MEASUREMENTS - BART Location & Line 1 Richmond 2 Richmond 3 Fremont 4 Fremont 5 Richmond 6 Concord 7 Fremont 8 Fremont Community Type Track Structure Train Speed [MPH] Near Far Average Maximum @ 50 Near Track Residential Aerial 80 80 93 Commercial Aerial 80 80 93 Residential Aerial 80 80 89' Commercial Aerial 80 80 88 Commercial At-Grade 80 80 87 Residential At-Grade 80 80 84 Residential At-Grade 80 80 85 Commercial At-Grade 70 80 87 Richmond Residential Vent Shaft 50 50 10 Richmond Commercial Vent Shaft 60 60 11 Concord Residential Walnut Creek Bridge 80 80 87 12 Fremont Residential/ Commercial Aerial Crossover 60 80 87 13 Fremont Residential At-Grade Crossover 80 80 78* * Noise level at 100 ft Levels - dBA Ft Far Track 80 81 79 79 85 76 77 79 78 76 73* 3-23 1. The passby noise levels at each measurement location are consistent. 2. The maximum passby noise levels for trains on the BART concrete aerial structure are 3 to 6 dBA higher than for trains on ballast and tie at-grade tracks. 3. The noise radiated from the two vent shafts was not a significant contributor to the noise climate during the samples. Station Noise Station platform noise measurements were made at 12 stations representative of the six different types of stations in use on the BART system. The six different types of stations are indicated on Table 3-3. Table 3-3 also presents the average of the maximum noise levels of both the entering and departing trains. The noise data ob¬ tained lead to the following general observations: 1. The average maximum noise levels at the aerial and subway stations are generally higher than those at the at-grade stations by 4 to 8 dBA. 2. The station platform configuration contributes only a small observable difference in the average maximum train noise levels on the platform from the near track. 3. The subway station absorption material contributes to the reduction of the subway station platform noise to comparable or less than that on the aerial station platforms. Interior Noise Five commute trips, simulating a typical commuter's trip on BART were made with noise instrumentation to assess the noise exposure during a typical trip. Table 3-4 summarizes the results of these trips. End-to-end interior noise measurements were also made between Concord and Daly City, and Richmond and Fremont. These tests were made with the microphone located at the center of the car. Both noise and train speed data were obtained for correlation between the noise level and train speed. A summary of the results also is shown in Table 3-4. From the interior noise data ob¬ tained, the following general observations can be made. 1. The average maximum interior noise levels in subway are from 4 to 7 dBA higher than those on aerial structure. 2. The average maximum interior noise levels on aerial structure are from 3 to 5 dBA higher than those for at-grade ballast and tie track. 3. The interior noise levels between cars on identical track structures are rela¬ tively consistent for measurements made at the same speed with the same microphone position. 3-24 TABLE 3-3. SUMMARY OF STATION PLATFORM MEASUREMENTS - BART Average Maximum Levels ~ dBA for Near Track at Center of Platform Platform 2 m From Edge Station Type Configuration Train Entering Train Departing Rockridge [Daytime] Center Aerial - Concrete Trackbed 80 82 Rockridge [Rush Hour] Center Aerial - Concrete Trackbed 85 86 Coliseum Center Aerial - Concrete Trackbed 80 85 Bay Fair Center Aerial - Concrete Trackbed 80 85 Walnut Creek Side Aerial - Concrete Trackbed 75 84 El Cerrito Del Norte Side Aerial - Concrete Trackbed 84 82 Pleasant Hill Side Aerial - Concrete Trackbed 79 83 Richmond Center At-Grade - Ballast & Tie 75 80 Union City Side At-Grade - Ballast & Tie 78 79 South Hayward Side At-Grade - Ballast & Tie 74 77 Lake Merritt Center Subway - Concrete Trackbed 77 86 19th Street [Upper Level] Side [Multi- Level] Subway - Concrete Trackbed 83 85 19th Street [Lower Level] Side [Multi- Level] Subway - Concrete Trackbed 84 87 Civic Center Center Subway - Concrete Trackbed 81 82 3-25 TABLE 3-4. SUMMARY OF IN-CAR NOISE MEASUREMENTS - BART COMMUTE TRIPS: Maximum Noise Level [dBA] [Excluding Transients] Trip Platform Subway At-Grade Aerial Concord Station to Montgomery Street Station 65 85 77 81 Fremont Station to Berkeley Station 65 82 75 79 Rockridge Station to Glen Park Station 1 85 84 73 78 El Cerrito Plaza Station to Lafayette Station 57 [El Cerrito] 86 74 76 80 [Mac Arthur] Powell Street Station to North Berkeley Station 65 [Powell St] 83 71 77 80 [Mac Arthur] END-TO-END TESTS: Maximum Noise Level [dBA] [Excluding Transients] Trip Subway At-Grade Aerial Concord Station to Daly City Station and Return 83 74 79 Richmond Station to Fremont 83 75 76 Station and Return Average of three trips 3-26 Bart Noise Summary A graphic summary of interior, station, and wayside noise is presented in Figure 3.1. The levels have been grouped into 5 dBA ranges centered at 72.5, 77.5, 82.5, 87.5 and 97.5 dBA. The levels have been further broken down in terms of aerial, at-grade, and subway structure, as the noise levels are highly dependent upon the type of structure on which the trains operate. 2. CLEVELAND TRANSIT SYSTEM 2.1 DESCRIPTION OF TRANSIT SYSTEM Routes and Service — The Cleveland Transit System rapid transit line (Airport Line) is 19 miles (30.6 Km) in length with 18 stations. The eastern portion of the line from the Cleveland Union Terminal to Windermere Station was opened in 1955. Five months later, the section from Union Terminal to West 117th-Madison was opened. Addition of the Triskett and West Park Stations at the western end of the line was completed in 1958. In 1968 the four-mile extension to Cleveland Hopkins International Airport was opened, including the Puritas and Brookpark Stations. The average running time from Windermere to Airport is 36 minutes. Roadbed — The roadbed consists of wood tie, rock ballast, and AREA 100 welded rail. Most of the track runs on-grade over right-of-way formerly utilized by the New York Central. Along the western portion of the line, the system parallels the Penn Central through mixed industrial, business, and residential communities. It parallels the Norfolk and Western tracks at the eastern end of the line. There are two underground track sections on the system — one near the Airport Station, 0.3 mile (.48 Km) in length, and the other at the downtown station. Public Square located in the Cleveland Union Terminal, 0.5 mile in length. Between W117-Madison and W25-Lorain, and between Campus and E-105-Quincy, the system is located in a cut. From Superior to Windermere, the roadbed is on elevated embankment. Where the roadbed is in a cut, a vertical concrete retaining wall is occasionally used on one side of the line. Reflections of undercar noise from this wall increases in-car noise by 1 dBA. The underground Airport station has one ventilation fan rated at 20,000 cfm which was installed to satisfy a 60 dB(A) noise criterion. The passenger tunnel is heated during the winter by a 50 KW, 2000 cfm forced air system. Short curve radii which produce wheel squeal are located at the Windermere yard ap¬ proach tracks, and entering and exiting the Public Square station. Intermittent moderate squeal noise or flange “sing" can be heard on most curves and on tangent track as well. Track gauge at CTS if 4 feet 8% inches, 14-inch tighter than standard railroad gauge, with wheel gauge set for standard track gauge. The resulting decrease 3-27 between wheel and track gauge appears to excite flange modes of the wheel which re¬ sult in flange "singing”. Inspection of the wheel reveals that the fillet between the flange and tire on cars at CTS is a smaller radius than on other systems, confirming that the flanges receive more excitation at CTS than was observed on other systems. Most residential areas are located between 200 and 300 feet (60 and 90 meters) from the centerline of the track with the exception of a section east of Triskett Station where some lower income homes adjoin the right-of-way within about 25-30 feet (7.5-9 meters). Rail Vehicles — The following table lists the railcars in use at CTS. TABLE 3-5. CTS RAILCAR INVENTORY - AIRPORT RAPID TRANSIT Single Cars Series Make Year Number 101-112 St. Louis 1955 12 113-118 St. Louis 1958 6 151-170 Pullman 1967 20 Total Single 38 Double Cars 201-256 St. Louis 1955 56 257-270 St. Louis 1953 14 Total Double 70 The cars are operated with approximately six minutes of headway on weekdays at peak travel periods, with 10 to 15 minutes on Saturdays and Sundays. There is no acoustical absorption in the car interior with the exception of three experimental Pullman cars whose floors and sidewall kick panels are covered with carpeting. The latter cars also have cloth-upholstered seats rather than the vinyl seats used on the standard cars. Although these experimental cars appear to be quieter to the casual rider, no significant change in noise level was recorded when compared with the "standard” Pullman cars. Stations — The two-track system serves patrons through 18 stations with an average station spacing of 1.13 miles (1.82 Km). The majority of stations are center platform style. East 55th is the only side platform station and it shares tracks with the Shaker Heights Transit, a light rail system. East 55th is an interchange station between systems. The Shaker Heights line uses the lower level center platform with the Airport Rapid line using the side platforms. The only elevated embankment stations are at Superior and Windermere. Underground stations are located at Public Square and at the Airport. 3-28 TABLE 3-6. SPECIFICATION OF CTS ST. LOUIS CAR Length of Car 48'6" single car 97'6" double unit Height, Rail to Roof 11'9" Width, at Floor Level 10' Width, at Window Level 10' Weight, Empty 56,000 pounds Total Weight, Pounds Per Foot 1,155 pounds Weight, Loaded 72,500 pounds Seating Capacity 54 passengers = -|09 55 passengers Motors (4 per car) 55 HP each Free Running Speed 47 MPH Windows Solex Safety Glass; laminated; stationary Heating and Ventilating Thermostatically controlled heat and air Some center platform stations have vertical dividers in sections which shield patrons on one side of the barriers from the direct radiating train noise on the opposite side. The underground terminal at Public Square station, which has three tracks divided by two island platforms about 20 feet wide, is primarily a reinforced concrete structure. Some side pillars and beams are covered by an approximate 2-foot wide sheet of corrugated and perforated steel facing. 2.2 VEHICLE NOISE REDUCTION St. Louis Cars — The predominant noise sources on the St. Louis cars arise from the propulsion system and the wheel/rail interface since undercar equipment is limited to the motor alternator set and the brake air compressor. As a result of the low equip¬ ment noise, static interior levels are quieter than the newer Pullman cars. When running, noise levels in the St. Louis cars are 1 to 5 dBA higher than the Pullman cars. A noise reduction program would require improved door seals and sealing of the ventilator system to increase the car body transmission loss. The sealing of the venti¬ lator, warrants a requirement for an air conditioning system. An investment in these cars of this magnitude would have to consider the remaining service life of the cars. Pullman Cars — A predominant source of noise on the newer Pullman cars is the air comfort blower. Silencers in this system would have to be designed with a low pres¬ sure drop in order to maintain a systematic air flow. Quieting the condenser blower for the air conditioning system would also be required. This noise source propagates to the interior from its undercar location through acoustic leaks in the car body. An in¬ crease in car body transmission loss, particularly the sidewalls and floor, and the ad- 3-29 dition of carpeting, such as installed on the experimental cars, would be required to achieve a measurable reduction in interior noise. 2.3 WHEEL/RAIL NOISE Wheel noise on the Airport Rapid cars contains a flange tone not normally heard on other rapid transit systems. This is thought to arise from a track gage of 4 feet 8% inches with a standard wheel gage. Inspection of the fillet on the wheel between the tread and the flange reveals that this area is worn on the Airport Rapid cars to a substantially greater degree than on other systems. The resulting sound of what is thought to be the flange scraping the rail is audible on tangent as well as curved trade. In addition, wheel flats are quite noticeable on many rapid cars, both interior and exterior to the car. Open¬ ing of the track gage of a companion reduction in wheel gage would probably elimi¬ nate the flange noise, while a regular program for truing wheels would substantially reduce slid flats as a source of noise on the cars. Wheel squeal is generally not a sub¬ stantial contributor to system noise. Where it does occur (underground near the air¬ port, just west of West 98th/Madison, on either approach to Cleveland Union terminal, and on the approach tracks to the Windermere shops beyond the station platform) squeal is generally of short duration and moderate in amplitude. Resilient, viscous-damped or friction-damped wheels would likely eliminate this noise source on the system, but benefits gained would have to be weighed against costs involved. In no case is the screech-generating track section near a noise sensitive community. All of the undercar equipment acoustical treatment recommended for interior noise reduction would be effective for community and station noise reduction as well. A sub¬ stantial noise source, particularly on underground station platorms was the air con¬ ditioner condenser fan. 2.4 TRACKBED AND WAY STRUCTURE NOISE REDUCTION Although the track is welded on the Airport Rapid system, the rail surface contributes to both interior and wayside noise when it is not smooth. System noise levels could be reduced by 4-5 dBA, if rails were ground smooth and wheels were trued in a regular maintenance program. In addition, filling insulated rail joints (and other rail gaps) with a nonconducting material would improve noise levels in localized areas, as well as in-car noise. In noise sensitive areas or where the system is located close to residential homes or apartments, a reduction in wayside noise could be achieved with wayside sound barrier walls. Since the system is located in commercial areas where residents are already conditioned to railroad noise (the rapid tracks are adjacent to Penn Central and Norfolk and Western tracks), the length of line where barrier walls could be used effectively is relatively short (estimated to be less than 10 percent of the system). To be effective, a wall should be somewhat higher than the top of the wheels, and be of sufficient mass to provide sound transmission loss at frequencies of wheel/rail and undercar equipment noise. Unfortunately, any walls installed to reduce noise of the 3-30 transit system would not be effective for the railroad system which masks all local noise sources when a highspeed train with several locomotives passes. Addition of an absorptive treatment on the tunnel walls where the system is located underground would reduce the wheel/rail and undercar noise in that reflective en¬ vironment. 2.5 CTS RAPID TRANSIT NOISE SUMMARY A graphic summary of community, station and in-car noise on the rapid transit line is presented in Figure 3.2. Levels have been grouped into 5 dBA ranges: 75-80, 80-85, 85-90, 90-95 and 95-100 dBA. Wayside measurements were made at a distance of 15M from the near track, station noise recorded at the center of a stopped multi-car train, and in-car data was taken in the second car of a multi-car train for one round trip. In-car noise presented represents steady-state plateau levels between stations. For all the cars surveyed, these levels are established by wheel/rail noise primarily, with pro¬ pulsion system noise contributing at a second-order level. The cars surveyed in each case did not have any audible wheel flats, but these cars were judged to be generally unrepresentative of CTS wheels which typically have audible slid flats. Another source of wheel rail noise on the rapid transit line is rail/flange "sing". Wheel excitation at the flange mode is different from wheel squeal generated on curved track and the noise is heard on tangent track as well, particularly at higher speeds. Inspection of the wheel contour reveals that the wheels wear to a sharp radius at the flange-tread fillet. Indi¬ cations are that the difference between rail gauge and wheel gauge may be less at CTS than on other systems, resulting in fillet wear as well as acoustic excitation of the flange. Passengers riding the system end-to-end in a St. Louis car would experience plateau noise levels in the 80-85 dBA group 75 percent of the distance traveled and from 85-90 for 25 percent of the distance. Noise levels in the Pullman cars, however, never exceed the 80-85 dBA range, and the modified Pullman cars have plateau levels in the 75-80 dBA grouping for 5 percent of route traveled. The CTS rapid transit has many stations which are similar in construction and layout, and the resulting noise exposure to patrons on the station platforms is rather uniform throughout the system. For example, 89 percent of the station noise environments due to train arrival and departure are in the 80-85 dBA interval with only two stations on the system displaying other noise levels. One of these is Public Square (85-90) and the other is Windermere (75-80). The microphone at Windermere Terminal was placed at the center of a multicar train and this position was held constant for all measurement periods. During the day, only single cars are operated, and cars are positioned nearer 3-31 0 20 1 5 10 5 0 50 40 30 20 1 0 0 r COMMUNITY X X X - X X X X X X X X x K X X X 1 X X X X X X X X X X X X X i V 70,75 75,80 80,85 85,90 90,95 95,100 GROUPED SOUND LEVEL, L Am ax - dBA NOTE: Grouped sound level interval includes lower, but not upper end point. FIGURE 3-2. SUMMARY OF CTS NOISE ENVIRONMENT 3-32 the end of the station platform and thus do not pass by the microphone at any time. If the microphone position had been adjacent to the single car location, noise levels would have been in the 85-90 dBA interval. Noise in the community has been ranked as follows: Approximately 49 percent in the interval 90-95 and 51 percent in the 95-100 dBA interval. Where the system operates in a cut, levels are in the 90-95 dBA interval and where it is on-grade or at elevated embankment, levels are in the 95-100 dBA interval. Near the terminals, or other locations for low speed operation, noise levels are in the 80-85 dBA grouping. Wheel squeal is audible in the community near the Windermere Yard (there are few residence nearby) and near W98-Detroit Station where the system is in a cut and "S" turns prior to the eastbound approach to the station. For the remainder of the system, noise levels are primarily established by wheel/rail sources. 3. PORT AUTHORITY TRANSIT CORPORATION 3.1 DESCRIPTION OF TRANSIT SYSTEM Routes and Service — The Delaware River Port Authority's Port Authority Transit Corporation Lindenwold High Speed Line has a route structure and operates rapid service between Philadelphia and Lindenwold, New Jersey. The line is 14.2 miles (22.9 km) long and has 12 stations. It went into operation in January 1969 between Camden and Lindenwold, with service extended to Philadelphia one month later on a section of track formerly used by the SEPTA bridge cars. The entire distance is covered in less than 23 minutes, for an average speed (including 10 intermediate stops) of 40 mph (64 kph). West of Camden, speeds are held to 40 mph (64 kph) maximum in the four miles (6.4 km) of subway which has several sharp curves requiring 30, 20 and 15 mph limits (48, 32, 24 kph). Southeast of Camden, on the new section of track, normal running speed is 75 mph (121 kph). Roadbed — Fully welded track is used, except in the subway between 8th and Race Streets (near the Ridge Avenue connection) and 16th Street in Philadelphia, and on the Benjamin Franklin Bridge and its approaches where the rail is jointed. The roadbed in these latter sections consists of short wood ties set in concrete, with every fifth tie a long tie. East of the Broadway station in Camden, the roadbed is new with continuously welded rail seated on double-shouldered tie-plates, and anchored using compression clips. Ten miles of roadbed in New Jersey is above ground, with 45 percent on an em¬ bankment, 5 percent on concrete viaduct. Figure 3—3, 40 percent at grade, and about 10 percent in a cut. The remaining section, some of which is in Camden and about 2.5 miles (4 km) on the Pennsylvania side, is underground. Near Lindenwold there is a short section of roadbed which runs at-grade, parallel to the Pennsylvania-Reading Seashore Line (PRSL). The Lindenwold station itself is on ele¬ vated embankment, with the PRSL track at-grade parallel to it. 3-33 p LOMuapu l i pue LMsV p L9Lj.uoppeH luoun S9M pOOMSDU L L L°3 0nu3A\/ X"eMp eoug L LbH MBW 8 asnoog o L-6 isnoog Zl^Zl ^ snoog 9L’SL <_> GO o cc < LU O GO LU CO O CC o u_ a LU CO o cc I— GO Z o o C_3 < ac o_l LU LU (_} CC=) OQ O*— <->> Qlu LUZ I— ^ >< LU CD LULU LU CO Q CC CQ >- < CQ ZD GO > CC < ID CO Q LU CD Q < O CC LU 3-34 The two stations on the system are located on concrete viaducts. These are the West¬ mont and Collingswood stations. At the Collingswood station, residential dwellings are closer to the roadbed 75 ft (23m), than anywhere else on the system. Wheel squeal was noted at the Lindenwold yardloop, and at six underground locations in both Philadelphia and Camden. Impact noise occurs at insulated track joints. Rail Vehicles — PATCO uses 75 Budd Company electric cars with a third rail shoe col¬ lecting power for the stainless steel cars which, according to type, are in two weight classes: double cars and single cars. There are 25 single-unit double-end cars seating 72, and 50 cars arranged as 25 married pairs, each car seating 80. Operation of the cars is under automatic control (ATO) and the most frequent head¬ way intervals are two minutes. These intervals lengthen to T /2 to 10 minutes during non¬ rush hour periods and on weekends, and to one hour during 1:30 to 5:30 a.m. “owl" hours. Service is continuous, 24 hours a day, 365 days a year. The car interiors are fully climate controlled and employ transverse, high-backed suburban seating. The seats provide some measure of acoustical absorption by their upholstered covers. The floor construction is plymetal with thermal/acoustical in¬ sulation on the underside covered with stainless steel sheets. The walls and ceiling also contain acoustical/thermal insulation to increase the car body transmission loss. Stations — The PATCO system has 12 stations with an average station spacing of 1.29 miles (2.08 km). Distances between the stations are shown in the following table: TABLE 3-7. DISTANCES BETWEEN RAPID TRANSIT STATIONS - PATCO Miles Km Lindenwold to Ashland 1.79 2.88 Ashland to Haddonfield 3.19 5.13 Haddonfield to Westmont 0.87 1.40 Westmont to Collingswood 1.05 1.69 Collingswood to Ferry Avenue 1.61 2.59 Ferry Avenue to Broadway 2.16 3.48 Broadway to City Hall 0.25 0.40 City Hall to 8th-Market 2.28 3.67 8th-Market to 9-10/Locust 0.43 0.69 9-10 Locust to 12-13/Locust 0.29 0.47 12-13/Locust to 15-16 Locust 0.28 0.45 Total 14.2 22.85 All of the stations are of the center platform type on short concrete pillars. Ferry Avenue station in Camden is somewhat unique in that its platform is split by a tail track for local trains. 3-35 An acoustical feature noted on the PATCO line is the use of thin metal perforated ceiling throughout its six New Jersey stations. 3.2 VEHICLE NOISE REDUCTION Plateau noise levels in the PATCO cars range from 72 dBA on elevated embankment to 84 dBA in some sections of the subway. While the lower levels are quite acceptable, within the subway, riders would benefit from reduced noise inside the car. Recognizing that reducing noise of a transit vehicle, once the car is designed and constructed, is a complex and costly procedure, some general statements can be made about those items on the car which would require acoustic treatment in order for a reduction in noise to be achieved. Detail measurements of system noise to identify individual components by amplitude and frequency were not within the scope of the program, but some ob¬ servations can be made. Interior noise levels at speed are established primarily by wheel/rail and propulsion system sources. The air comfort system may contribute a small amount near the end of each car where the blowers are located, but it is undercar noise which predominates inside the car. This is noted from time histories of 'A' weighted noise which is approximately 65 dBA when the car is at rest. A tone can be heard in the propulsion system noise and this is thought to be the gear mesh funda¬ mental frequency. Damping of the gears would be a logical first step in reducing gear¬ box noise. A second item requiring treatment is traction motor cooling. The self ventilated motors can be quieted by installation of a forced air blower with ducting to the motors. This treatment would be required along with all others noted to achieve a measurable reduction in noise. In addition, since interior noise increases by 12 dBA upon entering some portions of the subway (72-84) an increase in the car body acoustic transmission loss would be required. This would require improved door seals and additional insulation in the car body walls, ceiling and floor. The cooling blowers for the air conditioning compressor are audible in the car when this system is in use. It is noted more when the car doors are open while standing at a station. At very low speeds (0-5 mph) during both acceleration and deceleration, brake squeal is audible. Although these levels were not read as part of the vehicle interior noise, they are pure tones which are audible for short durations. It was also noted that brake squeal did not always occur, an indication that it might be possible to eliminate this source of noise through investigation of the units which do not squeal. 3.3 WHEEL/RAIL NOISE Wheel screech is generated on several short radius curves in the subway at PATCO. These levels are generally below 90 dBA within the car interior but could be con¬ sidered annoying by some commuters. Resilient or damped wheels would be bene¬ ficial in reducing this noise. PATCO has independently evaluated one type of resilient wheel on an in-service car, although this wheel type was in service for a short 3-36 duration as a result of an overheated wheel caused by a dragging hand brake. One car of a married pair was fitted with the resilient wheels, the second having steel wheels. Noise measurements were made in the car with resilient wheels while negotiating the short radius curves. Screech from the steel wheels on the adjacent car prevented a true evaluation of the resilient wheel effectiveness regarding screech, however. PATCO is currently evaluating a ring-damped steel wheel configuration for screech noise reduction in a joint program with Boeing Vertol. Measurement of in-car noise in the subway showed up to a 10 dBA noise reduction in wheel squeal. This type wheel, while not completely eliminating squeal on short radius curves, appears to be a cost effective ap¬ proach to reducing wheel squeal. Installation of rings on a complete train (e.g., 6 cars) would be a logical next step in this program. Wheel truing is already being accomplished on a routine basis at PATCO, and no men¬ tion has been made of this as a noise reduction technique. Slid flats are rarely noted on PATCO cars since wheels are frequently trued. Most of the acoustical treatment recommended for interior noise reduction would benefit the wayside or exterior noise as well. The one treatment for interior noise reduction, which would not reduce wayside noise however, is the addition of acoustical insulation within the car body. Since wheel squeal occurs only within the subway (Lindenwold shop area excepted), the installation of damped or resilient wheels would not reduce community noise by any real amount, although some subway station commuters would benefit by reduced squeal levels. Noise reduction of the air conditioning compressor cooling blowers would improve the exterior signature of the car at all stations and particularly within the subway, where decay times are relatively long due to low absorption of acoustic energy in the station area. Noise levels of cars standing in subway stations are 73-75 dBA. Levels on the station platform without cars drop 10 dBA to 63-65 dBA. A reduction in undercar equip¬ ment noise by this amount would result in cars which would sound only one half as loud as the present vehicles. To achieve this, blowers and fans probably would require installation of shrouds lined with acoustically absorptive material. Frequently, the undercar space to accomplish this is not available. On an existing car, the amount of noise reduction which can be achieved by modifications of this nature is dependent on the space available and the ability to design an installation which allows unhampered maintenance procedures. These criteria are not always compatible. 3.4 TRACKBED AND WAY STRUCTURE NOISE REDUCTION All track is welded on the PATCO system except in the subway between 8th and Race Streets (near the Ridge Avenue connection), 16th Street in Philadelphia, the Benjamin 3-37 Franklin Bridge and its approaches, and for occasional insulated joints and special trackwork, no impact noise is heard. Filling the insulated joints with a nonconducting material would improve noise levels in localized areas, however, particularly on steel structure overpasses. It is estimated that this would reduce impact noise at the way- side by 5 to 10 dBA. In the subway, a small reduction of in-car noise, 1 to 2 dBA, could be obtained by adding ballast between the rails to act as an acoustical absorber. Noise from wheel/ rail and undercar equipment sources would be scattered from this surface rather than reflected from the smooth concrete invert which is used throughout the subway. Ad¬ ditional absorption on the tunnel sidewalls and ceiling would be beneficial as well, and this is discussed in the following section. 3.5 GENERAL WAY STRUCTURES In noise sensitive areas or where the system is located close to residential homes or apartments, a reduction in wayside noise could be achieved with wayside sound bar¬ rier walls. To be effective, a wall should be somewhat higher than the top of the wheels and be of sufficient mass to provide sound transmission loss at frequencies associated with undercar noise. Walls would clearly have to satisfy many non- acoustical constraints as well. Some of these are safety, maintenance, snow removal, cost, etc. Subway tunnels were constructed in the 1930's and are acoustically reverberant. Ad¬ dition of an absorptive treatment on the walls of the tunnel such as sprayed-on materials would be an alternate to installing more insulation in the car body. To de¬ termine effectiveness, a section of the tunnel extending between two adjacent sta¬ tions could be evaluated in a preliminary test covering less than one-third mile. Complete coverage of the tunnel on the entire system would require treatment for 2.5 miles of subway. The reduction on each noise generating component on the vehicle should be pursued as the primary method of achieving lower system noise levels. The application of acoustically absorptive material at the Haddonfield station (located in a cut lined with vertical concrete walls) and in all the subway stations, could reduce noise during train approach and departure of reflected sound by 5-10 dBA, depending on the specific treatment, and the area covered. Direct noise, that which propagates to a station patron without benefit of reflecting from a treated surface, would not be reduced. However, most acoustic energy radi¬ ating from a transit car in the subway is located under the car. Only door noise and warning signals would propagate directly to a patron. 3-38 3.6 PATCO NOISE SUMMARY Patrons in underground stations are exposed to higher noise levels than patrons at sta¬ tions located on elevated embankment. Noise levels at Haddonfield (located in a cut) lie between those measured for underground and elevated embankment stations. It should be noted that while no attempt was made to measure vehicle speeds during passby in high speed territory, at each wayside location the trains were operating at equivalent speed somewhat below the maximum of 75 mph (120 kph). At the Benjamin Franklin bridge plaza train speeds were substantially lower, approximately 30 mph (48 kph) when eastbound. Lowest wayside noise levels were measured when the system operates in a cut, due to the reflection of acoustic energy vertically. The concentration of acoustic energy in the cut causes Haddonfield station patrons to be aware of an approaching or departing train for 20-30 seconds longer than patrons on other station platforms. The 15m com¬ munity noise with the trains operating in a cut was 15 dBA lower than measured adja¬ cent to concrete viaduct. Data recorded adjacent to elevated embankment was 10 dBA lower than for concrete viaduct but 5 dBA higher than for operation in a cut. Although the system operates at grade at certain locations between Haddonfield and Lindenwold, the total percentage is small compared with the other designated roadbed types, and for simplicity, these sections have been included in the elevated embankment category. It is estimated that the levels adjacent to the at-grade sections would not differ by more than 3 dBA from those measured at elevated embankment sites. Noise at the wayside (15m) can be characterized predominantly by the elevated embank¬ ment site. Approximately 86 percent of the community noise lies in the grouped data from 80-85 dBA with remaining wayside characterized nearly equally between concrete viaduct (90-95) and in-cut levels (75-80). 3-39 4. SOUTHEASTERN PENNSYLVANIA TRANSPORTATION AUTHORITY BROAD STREET SUBWAY 4.1 DESCRIPTION OF TRANSIT SYSTEM Routes and Service — The Broad Street system is a cut-and-cover subway located directly under Broad Street with the southern terminal at Pattison Avenue and the northern terminal at Fern Rock Station, a distance of 9.8 miles. The Fern Rock terminal is the only station on the system which is on-grade. The subway was con¬ structed in 1928 and operates a total of 300 cars. Ridge Avenue spur joins the system at the Fairmount station and terminates at the Market-Frankford line at Eighth Street, all underground. This spur, opened in 1936, originally extended across the Benjamin Franklin Bridge to the Broadway terminal in Camden. In 1969, the section between 8th and Market Streets in Philadelphia and Broadway in Camden became an integral part of the Delaware River Port Authority's Port Authority Transit Corporation (Linden- wold Line) and the Broad Street-Ridge Avenue spur was terminated at Eighth Street. Roadbed — The entire Broad Street subway system is underground except for the Fern Rock station, which is on-grade. The roadbed in the underground section from Pattison Avenue to Snyder consists of jointed rail set on tie plates located on resilient pads on concrete roadbed. A drainage ditch covered by steel grating is located between the rails. Between Snyder Avenue and the on-grade section at Fern Rock, the rail is set on short wood ties set in concrete. Every fifth tie is a cross tie. The on-grade section of track at Fern Rock is ballast and tie construction. The rail has bolted joints for the entire route. The rail surface is smooth with no corrugations ob¬ served, and only slight pitting was noted. Except for the tunnel ceiling at the City Hall station, where perforated steel sheets are used, no noise control treatment is employed. The Broad Street line consists almost entirely of tangent track except at City Hall station, where curved track routes the system around the City Hall Plaza, and also near Fern Rock, where the track turns east under Grange Avenue, leading to the terminal. In both instances, wheel screech occurs. The underground track sections are ventilated directly to the atmosphere through vent shafts located along the sidewalk at street level. Except for periods of very low street noise, or when an observer stands directly on the grating at street level, subway system noise goes unnoticed in the community. Rail Vehicles — 150 cars were built in 1928 and additional 50 constructed in 1938. The Bridge cars operate only on the Ridge Avenue spur. In the summer, the cars operate with windows as well as the doors between cars open for ventilation. Noise levels while running between stations for these conditions are generally between 90 dBAand lOOdBA. Power pickup for the car is through a third-rail system entirely. 3-40 Stations — Station configurations, as listed in Table 5.1, consist of one center plat¬ form between two tracks (7), two center platforms between four tracks (8), and two side platforms outside four tracks (7). All station surfaces visible from the station platforms at underground locations are of hard masonry, or ceramic tile construction. The four-track, two-side platform stations are skip-stop stations for express trains which operate during peak hours. Cashier's booths are generally located one level above the station platform on a mezzanine level, although a limited number of sta¬ tions do have the booth located at platform level. The Pattison Avenue terminal is a two level, two track center platform station. Plat¬ form width is 40 feet compared with 20 feet (12,6M) for on-line center platform sta¬ tions. The lower platform level is used only during periods required by the sports arena area which it serves. Walnut-Locust is an interchange with the Port Authority Transit Corporation's Linden- wold Line which crosses above the Broad Street line. Passengers exit overhead to mezzanine level and the PATCO line, and one level above that to Broad Street. The station at City Hall is an interchange with the Market-Frankford Line and the Subway-Surface Line (PCC cars). The Broad Street subway passes below both of the other systems. Wheel screech is very evident at this station as the tracks curve around City Hall plaza. Car operators also sound the horn in this location and the combined noise level may reach 100 dBA. A cashier's booth located one level above the tracks is exposed to this noise although it is occupied only during peak hours. Fairmount is an interchange station with the Ridge Avenue spur which terminates at Eighth and Market Streets. The construction of the spur is generally similar to the Broad Street line, but it is a two-track system except at Eighth Street where it single tracks into the terminal station. At Eighth Street, passengers can transfer to the Market-Frankford Line. VEHICLE NOISE REDUCTION Noise levels in the Broad Street cars range from 90 to 100 dBA while operating between stations. Standing at stations when the brake air compressor is not operating, the cars are very quiet with levels determined by sources other than the car itself (commuters in the car, or on the station platform). Any noise reduction program on these cars would have to treat several sources of noise and the paths by which they radiate to the interior. The predominant noise source on the car is the propulsion system, with the gearbox the main contributor. Gearbox noise propagates by a direct structure-borne path into the 3-inch thick concrete flooring re-radiating from that surface as airborne noise. A second path for propulsion system noise is airborne from the undercar location through the car body by way of the doors, windows and venti¬ lators. During summer months, windows and doors between cars are open allowing large areas for noise to enter the car body. To attenuate structure-borne gearbox noise, the truck requires isolation using a resilient mount between truck bolster and kingpin. Airborne noise should be at¬ tenuated by improved door seals and completely sealed ventilator system. This re¬ quirement implies that an air conditioning system would be required for summer months. Insulation would have to be added to the car body walls and ceiling. Treat¬ ment of the brake air compressor would require isolation by resilient mounting and some added insulation to reduce airborne noise radiation. 4.3 WHEEL/RAIL NOISE Rail surface on the Broad Street Subway is generally in excellent condition in that it is smooth with no obvious corrugation, pitting, etc. Rail grinding would offer little improvement to the noise environment (2 dBA) even in combination with wheel truing, since slid flats were not observed at any occasion in the subway. Squeal is generated at a few locations and where it was noted, noise levels were high. The squeal sites were north of City Hall station, where the tracks negotiate a curve around City Hall plaza, in the tunnel approaching Fern Rock and just outside the tunnel portal at Fern Rock on the tracks leading to the terminal. Resilient or damped wheels would reduce or eliminate the major wheel modes excited by this trackage, but as with the State-of-the-Art Car which had resilient wheels and was demonstrated on this line, certain wheel flange tones remained audible. Reduction of undercar noise levels, particularly the propulsion system, could result in station platform noise reduction of 4-7 dBA. A reduction in gearbox noise to achieve this would most likely require a redesigned gear train configuration or tooth profile or both. Installation of resilient or damped wheels would eliminate curve squeal at City Hall and Fern Rock stations, but would be ineffective in reducing station platform noise elsewhere on the system. Acoustic treatment of the brake air compressor would improve the car noise signature only at those stations where the air compressor was cycling. However, this is the only source of noise while the car is standing at a station, except for door operation. 4.4 TRACKBED AND WAY STRUCTURE NOISE REDUCTION Welding of rail joints would eliminate this source of noise in the car, although joints are generally well aligned and do not generate large magnitude impact noise. Quieting the propulsion system would have to be accomplished before this source became audible in the car. Cross overs do generate substantial levels of impact noise, however. A reduction of in-car noise to 2-3 dBA could be obtained by the addition of ballast between the rails to act as an acoustical absorber. Noise from wheel/rail and undercar equipment would be absorbed or scattered from this type of surface. Additional ab¬ sorption on the tunnel sidewalls and ceiling would also be beneficial as discussed in the following section. 3-42 4.5 GENERAL WAY STRUCTURES Subway tunnels are acoustically reverberant all along the system, but are particularly reflective between Pattison and Snyder Avenues, where sidewalls are smooth straight concrete, rather than scalloped, as is the case north of Snyder. Addition of an absorp¬ tive spray-on material would reduce in-car noise between stations by 3-7 dBA, and panels of absorptive materials on the station sidewalls and ceiling would reduce re¬ flected noise on the station platform by 10 dBA. Panels used for noise control should also be vandal resistant, durable, washable, etc. 4.6 BROAD STREET SUBWAY NOISE SUMMARY A graphic summary of community, station and in-car noise on the Broad Street line is presented in Figure 3—4. The levels have been grouped into eight 5 dBA ranges of noise from 70-75, 75-80, 80-85, 85-90, 90-95, 95-100, 100-105, and 105-110 dBA. Wayside measurements were made at a distance of 15M from the near track, station noise was obtained at the center of a stopped train and in-car data was taken in the second car of a multicar train. In-car noise levels are established primarily from the propulsion system, with wheel/ rail noise audible when trains are not accelerating or maintaining speed. In-car data was taken with all windows closed, however, much of the year cars are operated with doors between cars and ventilators open, and undercar noise is little attenuated. At the higher speeds, communication in the car is not possible at normal voice levels. Wheel squeal is generated just north of City Hall station as the trains negotiate curved track around City Hall Plaza and also near the tunnel portal of Fern Rock as the system tracks align with Nedro Avenue. A patron riding the system from terminal to ter¬ minal would experience noise plateau levels in the 95-100 dBA group 19 percent of the time, levels in the 100-105 dBA group 69 percent of the time and levels in the group 95-100 dBA 12 percent of the time. A patron's actual exposure is dependent on his specific commuting route of course, but if that route is between Snyder Avenue and Logan stations the mean level of exposure for each station-to-station plateau is 91 dBA with a standard deviation of only 1.25 dBA. Due to vehicle propulsion system noise and the reverberant characteristics of the sub¬ way stations, noise levels with trains entering and departing range between 88 and 99 dBA. The one station located at grade. Fern Rock, displays the lowest levels on the line, 86 dBA. The majority of the stations, 46 percent, have noise levels in the 90-95 dBA range with 23 percent in the group from 85-90, and 31 percent in the group from 95-100 dBA. Community noise exposure is restricted to the vicinity of Fern Rock station. In the immediate location of the station, community noise is in the 75-80 dBA group when the 15M noise is estimated. It is in this locale that most of the residential buildings are located. A few hundred meters from this site, community noise levels are also in the group from 75-80 dBA resulting from the generation of wheel squeal, but the distance to the nearest housing increases from 15M to more than twice this, tending 3-43 RIGHT-OF-WAY NUMBER OF TRACK LENGTH (lOOO's of feet) STATIONS (lOOO's of feet 1 5 1 0 5 . 0 STATION 70—75 75—80 80—85 85—90 90—95 95—100 100—105 IC5—110 0 T T COMMUNITY 15M T T 70— 75 75—80 80—85 85—90 on_95 95—100 . . . 7 —— — loo—ms 105-no GROUPED SOUND LEVEL, LAmax - dBA NOTE: GROUPED SOUND LEVEL INTERVAL INCLUDES LOWER, BUT NOT UPPER ENDPOINT. FIGURE 3-4. SUMMARY OF SEPTA BROAD STREET SUBWAY NOISE ENVIRONMENT 3-44 to forget the increased level due to squeal. Elsewhere on the line, the noise of subway train operation is not a measurable quantity in the community. MARKET-FRANKFORD LINE 4.7 DESCRIPTION OF TRANSIT SYSTEM Routes and Service — The Market-Frankford subway elevated line operates over all three basic types of roadbed (elevated, subway and at-grade) for a total distance of 12.8 miles, extending from 69th Street in Upper Darby, Pa., just west of the Phila¬ delphia county line, to the Frankford area in North Philadelphia. Its route is along Market Street in West Philadelphia, through center city, to Front Street where it turns north. It then follows Front Street to Kensington Avenue (York-Dauphin Station) running in a northeast direction until it joins with Frankford Avenue. From there, it continues to Bridge Street terminal. Maintenance facilities for the transit vehicles are located at 69th Street. The 69th Street area is a major business district near the Philadelphia line. Connections to all the major western suburban areas are available at the 69th Street Terminal by bus, light rail, or other high-speed rapid transit lines. The region which the route serves in West Philadelphia is comprised of residential homes, as well as commercial and business establishments which front directly on Market Street. Center city, where the system is located underground, is the hub of commercial and business enterprise. The North Philadelphia segment is similar to West Philadelphia in that it serves a residential and commercial area, although there tends to be more housing and less commerce along this northern segment than is the case for West Philadelphia. At 33rd Street, a route of the subway-surface system (light rail) parallels the Market- Frankford Line in the subway and continues to the 15th Street station with inter¬ mediate stops at 30th, 22nd and 19th streets. The Market-Frankford system interchanges with the subway surface line only at 15th and 30th Street stations, where passengers exit overhead and descend to grade level platforms on separate outside tracks. Noise levels of the PCC cars contribute sub¬ stantially to the acoustical environment of the Market-Frankford patrons. Roadbed — Inbound from the 69th Street Terminal, where the system is on-grade, the roadbed is ballast and tie construction for a distance of just over one-half mile. At this point, it becomes elevated above Market Street on steel structure where it continues to 46th Street station. The structure over this segment of the route region supports a concrete sub-base on which ballast and tie roadbed is laid. In some station areas, the roadbed consists of short wood ties set in concrete. Just east of 46th Street, the line descends into the subway, below Market Street, continuing under center city to Front Street where it makes a sharp turn northbound around a 200-ft radius curve. The roadbed in the tunnel consists of short wood ties in concrete. 3-45 There are no tunnel ventilating shafts and patrons on subway station platforms feel an onrush of air due to train motion through the tunnel just prior to train arrival at the station. At Front Street immediately north of Market, the line becomes elevated on steel struc¬ ture and remains so to the Bridge-Pratt terminal, which is also elevated on steel struc¬ ture. The line goes on-grade into a reversing loop although most trains remain in the station area and are operated in the reverse direction for the return run to 69th Street. There are numerous locations on the line where wheel screech is generated: the 69th Street reversing loop, entering and leaving the terminal, between the Milbourne and 63rd Street stations, entering the subway east of 46th Street, north of 2nd Street, just prior to leaving the subway, north of York-Dauphin station where the line joins Kensington Avenue and south of Church Street station. The rail is jointed along the entire route except for a section on the westbound tracks between 46th and 52nd Street in West Philadelphia. As rail is replaced, new welded rail is being installed. Many joints in the existing rail are not aligned, or have large gaps and produce large amplitude impact noise. Rail Vehicles — The transit cars in use on the Market-Frankford Line were built by Budd in 1960. There are 273 cars of this type on the system. The cars are ventilated by overhead fans. Speeds between stations are typically 45 mph, although trains accelerate to 55 mph descending under the Schuylkill River. Between 2nd and Bridge Streets, roadbed and elevated structure maintenance has reduced speeds between stations to 20 mph. This speed restriction will be in force until the extensive structural maintenance is com¬ pleted. It has currently been in effect for a period of about one year. Power pickup is by outside third rail. Noise levels in the car typically are in the range of 80 to 90 dBA. There is no specific acoustical treatment of the cars other than the thermal in¬ sulation normally applied to the car body. One car (No. 635) had Acousta Flex wheels for evaluation purposes. After about 40,000 miles, these were removed for re¬ furbishing by the manufacturer and have been reinstalled. These wheels virtually eliminate the wheel screech normally generated by all steel wheels. Stations — At 69th Street Terminal, passengers exit on one side platform overhead to other connections within the terminal. Entering patrons board trains by way of the center platform. Passengers at the only other on-grade station, Milbourne, enter and exit the station area from the cashier's booth on the eastbound platform. An overhead walkway is used for crossing to and from the westbound platform. Elevated station platforms are entered from street level through a cashier's booth fre¬ quently located under the station on a crossover walkway but above street level. Pass¬ enger crossovers between platforms are either below track level, or in some instances 3-46 north of 2nd Street above the tracks. Patrons are exposed to general community noise on these platforms in addition to the noise of the transit vehicles. The underground stations are virtually all masonry in construction, consisting of large surface areas of concrete and ceramic tile. Fifteenth Street station is an interchange with the Broad Street Subway and also the subway-surface system. Thirtieth Street also has a subway-surface station located on outboard tracks. Patrons on the 2nd Street station platform are exposed to wheel squeal as trains negotiate the 200-foot radius curve just north of the station. North of this curve the roadbed becomes elevated and, in most instances, the stations are similar in layout to the elevated stations in West Philadelphia. 4.8 VEHICLE NOISE REDUCTION Plateau noise levels in the Market-Frankford cars range from 75-90 dBA. In-car noise levels on the elevated structure in West Philadelphia are generally between 80-85 dBA. Operation on the elevated structure in North Philadelphia, where speeds are reduced due to structure maintenance, produces plateau levels from 75-80 dBA. In the subway, levels range from 85-90 dBA. The propulsion system and wheel/rail noise contribute primarily to in-car levels. A tone is noted in propulsion system noise as the cars accelerate and decelerate. This could be due to traction motor cooling fans or gearboxes. Any noise reduction pro¬ gram should identify the source of this tone more specifically. The brake air com¬ pressor is a predominant source of noise and would require acoustic treatment, al¬ though its operation is cyclic and probably less annoying than a continuous noise of the same level. In the summer the air conditioner compressors and blowers are a source of noise that would require treatment although noise data was not taken on this system during that season. An increase in car body acoustic transmission loss would be required to reduce in-car noise levels in the subway. Improved door seals and additional insulation in the car body would provide the means for achieving this. 4.9 WHEEL/RAIL NOISE Although some flats were noted on the Market-Frankford cars, wheel grinding is pro¬ vided on a routine basis. All cars have wheels trued on a STAN RAY wheel grinder approximately once a year. Maintenance of smooth rails through grinding would re¬ duce noise levels by 2-3 dBA. A Speno rail grinder is available at SEPTA and is shared with the Broad Street subway and PATCO line. Rail joints produce large amplitude impact noise on the elevated segments of the Market-Frankford line. Alignment of the rails and elimination or reduction in gaps would reduce or eliminate impact noise. As rail is replaced on the Market-Frankford line, welded rail is installed. Wheel squeal is produced on a number of curves on the Market-Frankford line, and could be reduced or eliminated with damped or resilient wheels. A car set of Acousta 3-47 Flex resilient wheels was installed on one in-service car and noise measurements were made inside the car, in stations and in the community during passby of the vehicle. Wheel squeal was virtually eliminated with this resilient wheel, and the test demon¬ strated noise reductions on some curves of 15 dBA. Other resilient or damped wheels also would be effective in reducing wheel squeal. Reduction of propulsion system noise, brake air compressor and wheel/rail noise, par¬ ticularly the elimination of wheel flats and rail joints would be required for station and community noise reduction. 4.10 TRACKBED AND WAY STRUCTURE NOISE REDUCTION Impact noise at rail joints and special trackwork propagates through the steel structure and re-radiates into the community and stations. Isolation of the rail and track or damping of the steel structure should be investigated to reduce community and station noise. Reports of trackbed isolation by a resilient pad under the ballast and tie road¬ bed on elevated steel structure by the Japanese National Railway have been encourag¬ ing. In the subway, a small reduction of in-car noise, 1-2 dBA, could be obtained by adding ballast between the rails to act as an acoustical absorber. Noise from wheel/rail and undercar equipment would be reduced by reflection or scattering from the surface of the ballast. 4.11 GENERAL WAY STRUCTURE For on grade or elevated steel structure, a reduction in wayside noise could be achieved with wayside sound barrier walls. The wall should be somewhat higher than the top of the wheels and of sufficient mass to provide sound transmission loss at frequencies of undercar noise. As noted elsewhere, walls would have to satisfy many non-acoustical constraints. Subway tunnels are highly reverberant. Absorptive, sprayed-on materials on the side walls and ceiling of the tunnel would be an alternate to installing additional insulation in the car body. A section of tunnel between two adjacent stations should be evalua¬ ted prior to installation in the approximately 3.7 mile length of subway. 4.12 MARKET-FRANKFORD LINE NOISE SUMMARY A graphic summary of community, station and in-car noise on the Market-Frankford line is presented in Figure 3—5. Levels have been grouped into six 5 dBA ranges. 70-75, 75-80, 80-85, 85-90, and 95-100 dBA. Wayside measurements were made at a distance of 15M from the near track, station noise recorded at the center of a stopped train and in-car data was taken in the second car of a multi-car train for two round trips. A summary of roadbed type for the Market-Frankford line is shown in Figure 3—6. 3-48 •—. C +J 3 ST R E h li) FAI RMOUNT G I RARD BERKS VORK-DAUPHIN HUNT!NGTON SOM ERSET ALLEGHENY TIOGA ERIE-TORR. CHURCH ORTHODOX B R! DGE-P RATT 3-50 FIGURE 3-6. SEPTA MARKET-FRANKFORD SUBWAY-ELEVATED ROADBED SCHEMATIC In-car noise presented represents steady state plateau levels reached between stations. In the 1960 Budd cars, these levels are established primarily by wheel/rail noise, with propulsion system noise generally audible only during acceleration and deceleration. When trains are stopped, the noise of the air comfort blowers is audible. Substantial wheel squeal is generated leaving 69th Street east of Millbourne, east of 46th Street entering the subway portal, north of 2nd Street, north of York-Dauphin and near Church station. In addition, there are several other locations where squeal is generated of a shorter duration or lower amplitude, generally between York-Dauphin and the Bridge Street terminal. A commuter riding the line from terminal to terminal would experi¬ ence noise plateau levels in the 70-75 dBA group 4 percent of the time, from 75-80, 54 percent of the time, from 80-85, 28 percent of the time, from 85-90, 11 percent of the time and from 90-95, 3 percent of the time. Levels in the subway are from 6-8 dBA greater than elsewhere on the system. The majority of stations on the system (19) are located on elevated steel structure. Of these, 17 have a noise environment during train arrival or departure in the noise group from 85-90 dBA. The underground stations all exceed this level with eight of the nine subway stations ranking in the 95-100 dBA group. Two at-grade stations, Millbourne and Bridge Street, are within the 80-85 dBA group. Wheel/rail impact, roar (and, at some locations, squeal) and propulsion system noise all contribute to the acoustical environment of a commuter on the station platform. Community noise levels on the system can be described almost entirely from measure¬ ments taken adjacent to elevated structure. Roadbed on the Market-Frankford line is located on elevated structure over 62 percent of the route, with the subway portion representing 34 percent. The remaining 4 percent is on-grade. Data taken both in West Philadelphia (54th and Market) and North Philadelphia (Hart Lane) is in the 85-90 dBA interval, with the on-grade wayside levels occurring in the grouping from 75-80 dBA. The remaining groups illustrate squeal levels at the 69th Street terminal return loop (80-85) and York-Dauphin curve on the elevated structure (95-100 dBA). Where the roadbed is located in the subway, noise levels due to train operation are not measurable in the community. 3-51 4. MINIMUM COST NOISE REDUCTION METHODOLOGY A. METHODOLOGY APPLIED TO INVESTIGATION The methodology used in this investigation to determine the minimum cost of reducing the various transit lines noise environments is that described in detail in Reference 2 conducted by the Transportation Systems Center of the Department of Transportation. A short review of those elements most important in the process of determining minimum costs follows. The grouping together of various track segments which have similar noise characteristics is essential to the quantification of the method. This data is based on, but expanded in detail from, the transit line summaries developed for the various transit systems and reported in Interim Report No. 1, Volumes 1 through 4 (Reference 1). Along with the grouping of the various track segments, an identification is made of major transit line noise sources and paths to the wayside, station platform and in-car receivers of the noise. Similar combinations are grouped together as noise scenarios. For each scenario thus determined, an estimate of the noise at each receiver contributed by each source via each path is made. The combination of data on noise abatement techniques and costs with noise control scenarios yields the noise reduction achievable to meet a selected noise goal for minimum cost. This latter step is performed using an iterative process by means of a computer algorithm described in Reference 3. 1. URBAN RAIL NOISE SOURCES, PATHS, AND RECEIVERS Even on a relatively compact noise source like the cars of a transit system there are several individually discernable sound sources and many paths by which their noise reaches a listener. Generally, only a few of the sources and paths are significant in establishing the total noise emitted. Table 4-1 is a summary of the sources and paths considered in the present program along with their coded reference numbers as used in subsequent calculations. Figure 4—1 shows the complexity of the total noise arriving at a receiver. It illustrates the various paths along different track segments for the case of an in-car noise receiver. For example, path P3 would be considered of prime importance for the determination of in- car noise levels when the car is in a tunnel. Its contribution would overshadow that made by path P9 since exterior reverberation from path P9 is limited to ballast and roadbed surfaces. At grade, or on elevated embank¬ ment however, path P9 would be a higher contributor than path P3 because of the absence of reflecting tunnel walls. Determination of the relative importance of the various sources and paths would be considerably enhanced by diagnostic tests and data analysis of the individual car equipment noise source, if such data were available. Tests of this nature were beyond the scope of this program. 4-1 TABLE 4-1. REFERENCE CODE LIST FOR NOISE CONTROL SCENARIO SOURCES AND PATHS NOISE CONTROL COSTS MINIMIZATION NOISE SOURCES DEFINED IN PROGRAM SOURCE DESCRIPTION 1 Rail Joints 2 Rail Roughness 3 Wheel Roughness 4 Power Pickup 5 Propulsion 6 Auxiliary Mechanical Noise 7 Auxiliary Airborne Noise 8 Air Brake Vents 9 Door Operation 10 Wheel Screech 11 Brake Screech 12 Flange Singing 13 Wheel Flats NOISE PATHS DEFINED IN PROGRAM PATH DESCRIPTION 1 Direct Radiation Under Car to Community 2 Structureborne to Car + Interior Reverberation 3 Exterior Reverberation + Car XMSN Loss + Interior Reverberation (Tunnel) 4 Direct to Car Exterior + Car XMSN Loss + Interior Reverberation 5 Direct Radiation Under Car to Station Patron 6 Station Reverberation to Patron 7 Structureborne + Secondary Radiation to Station Patron 8 Direct Radiation to Car Interior 9 Exterior Reverberation + Car XMSN Loss + Interior Reverberation (At Grade) 10 Direct Radiation to Car Exterior (At Grade) 11 Structureborne Secondary Radiation to Community 12 Station Sidewall Reflection to Patron, Subway or In Cut 4-2 PATH ELEMENT IDENTIFICATION P a = Structure Borne P b = Interior Reverberation P c = Exterior Reverberation Pd = Car Transmission Loss P e = Direct From Exterior Pf = Direct From Interior PATH PROPAGATION .. P 2 = P a + Pb -P3 = Pc + Pd + Pb (Tunnel) ----P4 = p e + Pd + Pb - P8 = Pf --— P9 = Pc + Pd + Pb (At Grade) FIGURE 4-1. ILLUSTRATION OF IN-CAR NOISE PATH COMBINATION USED IN CONTROL SCENARIOS 4-3 2. FORMULATION OF NOISE CONTROL SCENARIOS Development of a noise scenario involves separating the total transit line noise environ¬ ment into source-path contributions. All similar track segments are grouped together, as shown in F igure 4-2, for example. The table below this is divided to show which in-car noise plateau levels can be grouped together on the basis of track construction, vehicle speed, and geometric similarities of the various noise propagation paths. The various noise scenarios are here identified by the symbols R1, R2, etc. The same technique of grouping similar noise source and path elements together, is followed in defining scenarios for station platform and wayside community environments. 3. DETERMINATION OF MINIMUM-COST NOISE CONTROL OPTIONS The assessment of transit line noise control options which result in the minimum cost of reducing noise at a particular receiver is initiated by determining in detail the contribu¬ tion of each significant source and its appropriate paths. Table 4-2 is a sample scenario breakdown for the five sources and five paths considered applicable to subway, bridge and elevated embankment track construction at PATCO. Close inspection indicates that a predominant noise source on this car is the propulsion system. The noise of this system reaches a patron along path P3 primarily in subways, but by path P4 when the car is on elevated embankment. This information, together with the data developed on noise abatement techniques and costs (see Section 3), is then used to achieve a listing of noise reduction versus minimum treatment costs. Table 4-3 is a listing of the various transit system noise control fixes considered applicable to some or all of the transit systems under study. The cost of attenuating the sources reaching the receiver by one or more paths is calculated for a combination of treatments to arrive at a minimum cost. Applying this method to all track segments of a transit line then yields the minimum cost to achieve a uniform noise goal along the entire line. B. SYSTEM NOISE REDUCTION The methodology described in Section 4-A has been applied to each of the systems in the study. The minimum cost program has been performed separately for each category (commuter, station and community.) The resulting costs were tabulated with the associated line and car fixes, and the cost of noise reduction plotted in 1 dBA increments to illustrate where significant increases in cost occur. The actual values placed on each of the line and car fixes have undoubtedly changed from the time the study was prepared, and therefore may not represent actual costs at the present. How¬ ever, if it is assumed that the costs do remain correct relative to each other, the trends illus¬ trated in the noise-value curves will not change to any great extent. 4-4 aiOMNBQNn QNV1HSV aiiUNoaavH XNO IAI _LS 3M QOOMS9N ITIOO 3HN3AV AH 2d 33 AVMavoaa nVH AXIO 3 3a Ida NHXNVad N I IAIVfN3a X3» av Id—8 xsnooi oi-6 jjsnoon zi—zi i xsnooi 91-si Noise Group R3 R8 c ^ o LO CJ -r- 03 CM 03 4-> jO • • • O r^. i S- CJ 1— o 1— QJ r-'. oo x: ' o co +-> 4-> CO cn qj CO CO C QJ #3 #\ Q) u_ LO CO _l —' 0) Q. _a cn C3 O V CM • O O s- QC QC / 2 : cd / j c - X*C o f o CJ -r- CJ 03 0 • O CJ « y r- CO 03 +J i— CO *3" CO 03 r>. 1 / s- LO (— QJ u r"» oo — J) .c ' • • • • 0 — +-> CO *3" CO o 'nr cn qj CQ CM CD CO CM Cj C QJ i— CM r— CM CM Q) U— rs r\ r\ -O 00 _J — CO CO LO « _JQ qj so. cn 3 • 1 - o o s~ V ^ P- sr crs d c o O # r- 03 lo 03 +J CM 'O’ LO CM CO J- +J CO O CM cn aj co co i— d C QJ r— CM l— a> u_ A A ft io _j —- CO LO CM u s- 03 C-J C NCMLOOOOLO CM r-> co c . r~- -0 CO CQ C TD d- ~o -o h-H — 1 4-> 1 4-> c C 03 c o •r- > aj -r- 03 CJ 03 JO O 03 o s- E CJ +-> i — CM CO LO CO • • • cn o m Q- aj cn 03 CJ r— "O -*-> QJ S- Q) CSJ •f— so C/3 1— OO ie »—1 — -JX o o CJ 03 m 03 > QJ -m +-> 4-J 1— CJ QJ 03 cn o no 03 03 CL 2 -a CJ r— c— !- >> JO o_ Q_ 1— 1— =3 C- c CO CQ 4-5 FIGURE 4-2. SAMPLE NOISE CONTROL SCENARIO DEFINITION - PATCO IN-CAR NOISE DOUBLE CARS. TABLE 4-2. SAMPLE SCENARIO BREAKDOWN - PATCO IN-CAR NOISE LEVELS - DOUBLE CARS. LU UJ _J LU > CD -Q 3 CO > CD -Q 3 CO Cl) OT jD "u. CO JO E LU LU < CO 3 I CO _J LU > LU _J LU CO LU O cc D O CO V ° X CC 3 LU < < Px ID O O LO co "d- co co co CNI CO O LO LO O 'T in LO CD CO O LO LO <3- co lo co co px co o lo co CO LO CO CO CO § y <1 px io co o CO CO 0 CN CO O'— LO -cr co lo CO O LO lo lo o 'St CN CO O CO LO CO CN •3- z o CO _l Z> CL o cc Q_ r- 05 O 'tf co id oo rx CO 'f ID O is 0 S 0 O CO o rx co co LO 0 O t rx px rx CO LO LU O LU Z> 5 o ^ cc CN O T- rx io rx LO CO px LO CO O co co co t— o CO LO CO LO rx r- to CO CO CO rx X < o cc cc CN O r- rx lo rx LO CO lo co o CD 'Sf CO CO r- Kt C) CO LO CO LO rx r- io CO CO CO r-x CO cc X H CO CN O LO LO rx. co rx. co lo rx r- io io px px Px fx. CO LO X I- < QL — CN CO ■'3" CO < Q_ CL Q- Q_ -z CN CO 'St CO < Q- CL Q_ Q_ -z CN 'sp CO 05 < Q. Q. 1 O. 3 CN ^ ffl O) < Q- 0- Ol 0- < CQ "O 3' UJ —1 I- UJ < > —I LU CL —I CN CO rx px CN rx rx px cc < z LU CJ CO X I— o z LU o CO LO o CO- lo" o CO lo' CO 'd- X ^r CO cc CN cc CO X cc 4-6 TABLE 4-3. REFERENCE CODE LIST FOR NOISE CONTROL SCENARIO SYSTEM FIXES SYSTEM FIXES DEFINED IN PROGRAM REFERENCE NO. DESCRIPTION 1 Damped Wheels (Viscous) 2 Resilient Wheels 3 Interior Car Absorption 4 Acoustic Sealing of Car 5 Wheel Trueing 6 New Car 7 Brake Vent Mufflers 8 Brake Maintenance 9 Weld Rails 10 Improved Joints 11 Rail Grinding 12 Resilient Fasteners 13 Non-Absorptive Barriers 14 Absorptive Barriers 15 Absorptive Sidewalls in Tunnel 16 Ceiling Absorption 17 Wall Absorption 18 Station, Under Platform Absorption 19 Station, Absorption on Concrete Invert 20 Absorptive Barrier Between Trains 21 Damping Steel Elevated Structure 22 Damped Wheels (Friction) 23 Propulsion System 24 Blower 25 Modified Propulsion System 26 Ballast in Concrete Tunnel Roadbed 27 Open Track Gage 28 Under-Car Absrption 29 Modified Brake Air Compressor 4-7 1. BAY AREA RAPID TRANSIT DISTRICT (a) Wayside Noise Reduction The potential noise abatement techniques for rail rapid transit noise reduction pre¬ sented in Table 4-3 was reviewed for applicability to the BART system. The tech¬ niques selected are summarized in Table 4-4. These techniques along with the noise sources and paths which they reduce, their costs and noise reduction are pre¬ sented in Table 4-5. The analysis result which selected minimum costs for wayside noise reduction to goals of 90, 85 and 80 dBA are presented in Figure A3 and tabulated in Table 4-6. A goal of 90 dBA could be achieved on the entire system with rail grinding on Scenario 6, a high speed section of elevated structure. The lower goals require sound barrier walls, at least on some track sections as well as car fixes, such as undercar absorption. Compared with other systems in this study, noise abatement costs are generally equally effective over the entire range of goals achievable in the study (see Figure A3). Although there are step increases in costs in reducing noise levels from 88 to 87 dBA and 79 to 78 dBA, the curve indicates a relatively even trend from 91 to 78 dBA. (b) Station Noise Reduction The noise reduction techniques presented in Table 4-3 were reviewed for applicability to BART stations. All techniques that were considered to be potentially applicable to station noise reduction are shown in Table 4-7. The techniques along with noise sources and paths which they reduce, the associated noise reduction and costs are presented in Table 4-8. Computer selected techniques to reduce station noise levels to 80 dBA are shown in Figure A4 and tabulated in Table 4-9. Levels of 75 dBA for all stations was not achievable by the methods under study. Levels below the 85 dBA already exist on the BART system. The 80 dBA goal could be achieved by adding absorptive material to the undercar areas of the BART car and absorptive (ballast) on the concrete invert in the subway station areas. (c) In-Car Noise Reduction System noise control fixes selected for application to BART in-car noise reduction presented in the listing shown in Table A9a. The noise abatement techniques, related sound and paths and the noise reduction valves are unit costs assumed for the analysis are presented in Table A10. 4-8 TABLE 4-4. POTENTIAL WAYSIDE NOISE REDUCTION TECHNIQUES - BART TABLE BART WAYSIDE POTENTIAL NOISE REDUCTION TECHNIQUES Scenario Noise Reduction Technique Car Resilient wheels Wheel trueing New car with resilient wheels Modified propulsion system Undercar absorption Scenario Reference Max NO. Speed dBA @ 50 Feet 1 At grade 80 86 Rail grinding 2 70 85 Absorptive sound barrier wall 3 60 83 Non-absorptive sound barrier 4 50 80 5 < 40 73 6 Elevated So 91 Rail grinding 7 70 90 Non-absorptive sound barrier wall 8 6o 88 Absorptive sound barrier wall 9 50 36 10 < 4o 84 4-9 TABLE 4-5. WAYSIDE NOISE ABATEMENT TECHNIQUES - BART TECHNIQUE COSTS Scenario Source (S) Reduction Fixed Cost Unit Cost Reference Path (P) (dBA) ($) ($) No. 2 Resilient wheels 2 0 9,600 Source =3 5 Wheel trueing 0 300 Source = 5 2 6 New car w/resilnt whls 0 450,000 Source = 3 2 Source = 5 1 Source = 6 5 11 Rail grinding 0 1 Source = 2 6 13 Non-absorptive sound barrier wall 0 20 Path = 1 6 14 Absorptive sound barrier wall 0 30 Path = 1 9 25 Mod propulsion system 0 10,000 Source = 5 1 28 Undercar absorption 0 280 Path = 1 1 13 Non-absorptive sound barrier 0 20 Path = 1 9 4-10 MILLIONS OF DOLLARS FIGURE 4-3. MINIMUM COSTS TO REDUCE WAYSIDE NOISE - BAY AREA RAPID TRANSIT 4-11 TABLE 4-6. COMPUTER SELECTED TECHNIUQUES FOR WAYSIDE NOISE REDUCTION - BART Scenario Wayside Noise Goal Car Fix Reference Line No. Fix Cost $ 90 None 6 Rail grinding 69,200 85 Undercar absorption 6 Non-absorption barrier wall sound 1,730,400 7 Non-absorption barrier wall sound 8 Rail grinding 80 Wheel truing 1 Non-absorption barrier sound 4,605,500 Undercar absorption 2 Non-absorption barrier sound 3 Rail grinding \ 6 Absorption sound barrier wall 7 Rail grinding non-absorption barrier wall sound 8 Non-absorption barrier wall sound 9 Non-absorption barrier wall sound 10 Rail grinding 4-12 TABLE 4-7. POTENTIAL STATION NOISE REDUCTION TECHNIQUES - BART Scenario Reference No. Wheel trueing New car with resilient wheels Modified propulsion system Undercar absorption dee nano Noise Reduction Technique Car Resilient wheels 11 Sta 1 Center platform Aerial concrete structure 12 Sta 2 Side Platform 13 Sta 3 Center platform Ground BT 14 Sta 3 Center platform Ground BT 15 Sta k Side Platform Ground BT 16 Sta 6 Center platform Subway, concrete Rail grinding Absorption on concrete invert Rail grinding Absorption on concrete invert Rail grinding Absorption on concrete invert Rail grinding Absorption on concrete invert Rail grinding Absorptive sidewalls - tunnel Absorption on concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls - tunnel Absorption on concrete invert Ballast in tunnel TABLE 4-8. STATION NOISE ABATEMENT TECHNIQUES - BART TECHNIQUE Scenario Reference Source (S) No. Path (P) Reduction (dBA) COSTS Fixed Cost (?) Unit Cost ($) 2 Resilient wheels Source = 3 2 0 5 o 11 Wheel trueing Source =3 2 Source =13 5 New car with resilient whls Source =3 2 Source =5 1 Source = 6 5 Rail grinding Source =2 6 0 300 0 450,000 0 0 15 19 25 26 Absorptive sidewalls tunnel Path =12 5 Station absorption concerete invert Path =5 1 Path =6 5 Mod propulsion system Source =5 2 0 0 0 Ballast in tunnel Path =12 5 0 60 2 10,000 2 28 Undercar absorption Path = 5 0 200 4-14 FIGURE 4-4. MINIMUM COSTS TO REDUCE STATION NOISE - BAY AREA RAPID TRANSIT 4-15 TABLE 4-9. COMPUTER SELECTED TECHNIQUES FOR STATION NOISE REDUCTION - BART Wayside Noise Goal (dBA) Scenario Car Reference Line Fix No. Fix Cost ($) 80 Undercar absorption 16 Absorption on concrete invert 73,600 4-16 TABLE 4—9a. POTENTIAL IN-CAR NOISE REDUCTION TECHNIQUES - BART Car Car Subway Subway Subway Subway Subway Subway Subway Scenario Noise Reduction Technique Resilient wheels Isolation of floor Isolation and balancing aux equip. Wheel trueing New car w/resilient wheels Mod. propulsion system Undercar absorption 69,000 ft Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel 1,600 ft Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel 8,000 ft Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel 6,700 ft Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel 8,900 ft Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel 9,300 ft Rail grinding Absorptive sidewalss in tunnel Station absorption, concrete invert Ballast in tunnel 2,000 ft Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel 4-17 TABLE 4—9a. (Continued) Scenario Car At Grade At Grade At Grade 102,000 ft At Grade 16,500 ft 5,500 ft At Grade 11,650 7,600 ft Elevated 69,200 ft Elevated 17,000 ft Elevated 8,900 ft Elevated 14,950 ft Noise Reduction Technique Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel Rail grinding Absorptive sidewalls in tunnel Station absorption, concrete invert Ballast in tunnel 4-18 TABLE 4—9a. (Continued) Scenario Car Noise Abatement Technique Elevated 9,400 ft Rail grinding Absorptive sidewalls in tunnel Station absorption concrete invert Ballast in tunnel 4-19 TABLE 4-10. IN-CAR NOISE ABATEMENT TECHNIQUES - BART Scenario Reference No. 2 3 4 5 6 11 15 19 25 26 27 Source (s) Path (P) TECHNIQUE Reduction (dBA) COSTS Fixed ($) 0 0 0 0 0 0 0 0 0 0 0 Cost/Unit ($) 9,600 50,000 10,000 300 450,000 1 240 8 10,000 8 280 Resilient wheels Source =3 2 Isolation of Floor Source =6 5 Isolation & Bal Aux Eqpmt Source =6 4 Wheel Trueing Source =3 1 New Car with resilient whls Source =3 2 Source =5 5 Source =6 5 Source =7 2 Rail Grinding Source =2 3 Absorptive Sidewalls tunnel Path =3 5 Station absorption concrete invert Path =3 5 Mod Propulsion System Source =5 5 Ballast in Tunnel Path =3 6 Undercar Absorption Path =3 2 Path =9 4 4-20 The results of the computer analysis to select the minimum cost to reduce in-car noise levels to 80, 75and 71 dBA are shown in Figure 4-5 and Table 4-11. The goal of 71 dBA was the lowest level achievable for the total BART system. The results of the analysis suggest basic approaches for in-car noise reduction and should not be regarded as a specific acoustical treatment. No attempt has been made, for example, to define an absorptive treatment for stations which is satis¬ factory with regard to durability vandal resistance or the like since the present study has the broad purpose of identifying areas where accoustic treatment is required. 2. CLEVELAND TRANSIT SYSTEM (a) Wayside Noise Reduction The list of potential noise abatement techniques presented in Table 4-3 has been reviewed for applicability to the CTS Airport Rapid Line. Those techniques which were considered applicable are summarized in Table 4-12 and also presented in Table 4-13 along with the noise sources and paths which they reduce, the assumed noise reduction and costs. The sources (SI, S2 . . . . ) and paths (PI, P2 . . . . ) refer to sources and paths listed in Table 4-1. The results of the computer analysis to determine the minimum cost to reduce way- side noise levels to selected goals of 95, 90, 85 and 80 dBA are shown in Figure 4-6 and tabulated in Table 4-14. An initial requirement for wayside noise reduction is a more frequent truing of wheels. Progressive reductions in noise would require a noise reduction program for the propulsion system and rail grinding. In addition, wheel flange tones would have to be eliminated as a noise source — the method adopted was to increase the rail gage — and to provide wayside barriers along a large segment of the route. Whether wayside noise goals below 85 dBA are required along the CTS right-of-way, where residents are already pre-conditioned to railroad traffic is questionable. In fact, a reduction in CTS traffic noise without an accompanying reduction in Penn Central and Norfolk and Western system noise would be of little value to wayside residents. As a minimum, this question would require further study prior to any committment of funds. One recommendation would be to calculate Day-Night Levels for several wayside locations without Airport Rapid passby noise events to determine the impact on nearby residents. (b) Station Noise Reduction The scenario system fixes listed in Table 4-3 were reviewed for applicability to the Airport Rapid station noise reduction and the resulting fixes listed by scenario in Table 4-15. These noise control techniques along with the noise sources and paths which they reduce, the assumed noise reduction and costs are presented in Table 4-16. The numbers identified by each source (SI, S2 . . . . ) and paths (PI, P2 . . . . ) refer to the source and paths listed in Table 4-1. 4-21 FIGURE 4-5. MINIMUM COSTS TO REDUCE IN-CAR NOISE BAY AREA RAPID TRANSIT 4-22 TABLE 4-11. COMPUTER SELECTED TECHNIQUES FOR IN-CAR NOISE REDUCTION - BART In-Car Noise Goal (dBA) Car Fix Scenario Reference No. Line Fix Cost ($) 80 Undercar Absorpt. None 70,000 75 Undercar Absorp. 1 2 3 Station Absorpt, Cone. Invert 1,250,800 Ballast in Tunnel Ballast in Tunnel Station Absorpt, Cone. Invert 71 Resilient Wheels 1 Isolation & Bal Aux Equip. Wheel Trueing Mod. Propulsion System 2 Undercar Absorption 3 4 5 13 Rail grinding 8,881,200 Station Absorpt, Cone. Invert Ballast in Tunnel Rail grinding Ballast in Tunnel Ballast in Tunnel Rail grinding Rail grinding Rail grinding 4-23 TABLE 4-12. POTENTIAL WAYSIDE NOISE REDUCTION TECHNIQUES - CTS Scenario Reference No. 1 2 V Scenario Noise .Abatement Technique Car Visco elastic damped wheels Resilient wheels Wheel trueiig New car Friction damped wheels New propulsion system Modified propulsion system At grade and elevated embankment In cut Rail grinding Non-absorptive barriers Absorptive barriers Open track gauge Rail grinding Non-absorptive barriers Absorptive barriers Open track gauge 4-24 TABLE 4-13. WAYSIDE NOISE ABATEMENT TECHNIQUES - CTS TECHNIQUE COSTS Scenario Reference Source (S) Reduction Fixed Cost Unit Cost No. Path (P) (dBA) ($) ($) 1 Viscous damped, wheels 0 1,200 Source = 2 2 Source = 3 2 Source = 13 2 2 Resilient wheels 0 9,600 Source = 2 3 , Source = 3 3 Source = 13 3 5 Wheel trueing 0 1,000 Source = 13 20 Source = 3 6 6 New car 0 600,000 Source = 13 20 Source = 3 6 Source = 5 10 Path =1 l 11 Rail grinding 0 1 Source = 2 5 13 Non-absorptive barrier 0 100 Path = 1 7 l4 Absorptive barrier 0 150 Path = 1 10 22 Friction damped wheels 0 4oo Source = 2 2 Source = 3 2 Source = 13 1 23 New prop system 0 80,000 Source = 5 7 25 Mod prop system 0 4,000 Source = 5 y 27 Open track gauge 0 5 Source = 12 20 4-25 SOUND LEVEL - dBA SOUND LEVEL - dBA 100 CTS WAYSIDE MILLIONS OF DOLLARS FIGURE 4-6. MINIMUM COSTS TO REDUCE WAYSIDE NOISE - CTS 4-26 FIGURE 4-14. COMPUTER SELECTED TECHNIQUES FOR WAYSIDE NOISE REDUCTION - CTS PULLMAN CARS Wayside Scenario Noise Goal Car Reference Line Cost (dBA) Fix No. Fix ($) 95 Wheel trueing None 20,000 90 Wheel trueing 1 Rail grinding 382,360 Modified propulsion system Open track gauge 85 Wheel trueing 1 Non absorp. barr. 5,108,900 Friction damped wheels Rail grinding Modified propulsion system 2 Open track gauge Rail grinding 8 o Wheel trueing 1 Absorp. barriers 12,36U,36 o Friction damped wheels Open track gauge Modified propulsion system 2 Non-absorption barrier 4-27 TABLE 4-15. STATION NOISE REDUCTION TECHNIQUES - CTS Scenario Noise Abatement Technique Car Visco-elastic damped wheels Resilient wheels Wheel trueing New car Friction damped wheels New propulsion system Modified propulsion system Elevated embankment Ceuiter platform Rail grinding Absorptive sidewalls Station ceiling absorption Wall absorption Under-platform absorption Open track gauge In cut Side platform Rail grinding Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Open track gauge Subway Two center platforms Rail grinding Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Open track gauge In cut Center platform Rail grinding Absorptive sidewalls Ceiling absorption Wall absorption Under-platform Absorption Open track gauge At grade Center platform Rail grinding Absorptive sidewalls Ceiling absorption Wall absorption Uncer-platform absorption Open track gauge Subway Center oletform Rail grinding Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Open track gauge 4-28 TABLE 4-16. STATION NOISE ABATEMENT TECHNIQUES - CTS TECHNIQUE Scenario Reference Source (S) No. Path (?) Reduction Fixed (dBA) ($) COSTS Per Unit ($) 1 2 5 £ c 11 15 16 17 18 22 23 25 27 Viscous damped wheels Source =2 2 Source =3 2 Source =13 2 Resilient wheels Source =2 3 Source =3 3 Source =13 3 Vdieel trueing Source =3 6 Source =13 20 New car Source =3 2 Source =5 10 Source =6 10 Source = 13 20 Path =5 1 Rail grinding Source =2 5 Absorptive sidewalls tunnel, cut Path ‘= 12 8 Sta ceiling absorption Path =6 2 Sta wall absorption Path =6 2 Sta under platform absorption Path =6 3 Friction damped wheels Source =2 2 Source =3 2 Source =13 1 New propulsion system Source =5 7 Mod propulsion system Source =5 5 Open track gauge Source = 12 20 0 0 0 0 0 0 0 0 0 0 0 0 0 1,200 1,000 600,000 1 35 20 20 35 LOO 80,000 L,000 £ 4-29 The results of the computer analysis to determine the minimum cost to reduce all station noise levels on the system to goals of 85, 80 and 75 dBA are shown in Figure 4-7 and tabulated in Table 4-17. The car and line fixed itemized in Table 4-18 in general are self explanatory. The comment regarding noise of the Penn Central, Norfolk and Western railroad lines which was made in the wayside noise reduction section, applies equally to station noise. The noise levels of long line, high speed trains which closely parallel the Rapid total, mash the noise of the Rapid cars when the passby of transit train coincides with fast moving frieght. Thus, the costs shown in the accompanying tables and figures do not recognize these independent noise sounds for those stations where the railroad and transit systems operate parallel to one another. These stations are in the majority at CTS. (c) In-Car Noise Reduction The system noise control fixes selected for application to in-car noise reduction are shown in the listing of Table 4-18. The noise abatement techniques, the sources and paths which they relate to and the noise reduction values assumed for this analysis and unit costs are presented in Table 4-19. The results of the computer analysis to select the minimum cost to reduce in car noise levels to 85, 80, 75 and 70 dBA are shown in Figure 4-8 and tabulated in Table 4-20. Goals of 85 and 80 dBA require fixes to the car only, while the 75 and 70 dBA levels would require rail grinding as well as opening of the track gage in several scenarios. Note that a reduction to 70 dBA would require an estimated expenditure 100 times the cost to achieve a 75 dBA level in the car. PORT AUTHORITY TRANSIT CORPORATION (a) Wayside Noise Reduction The list of potential system fixes itemized in Table 4-3 was reviewed for applicability to the PATCO system, and the resulting figures listed by scenarios in Table 4-21. The selected techniques, along with the noise sources and paths which they reduce, the assumed noise reduction (in dBA) and costs are presented in Table 4-22. The numbers identified by each source and paths refer to the sources and paths listed in Table 4-1. The results of the computer analysis to determine the minimum cost to reduce way- side noise levels to selected goals of 90 and 85 dBA are shown in Figures 4—9 and tabulated in Table 4—23. To reduce noise levels in the community to 90 dBA, would require (1) modification of the existing propulsion system to lower noise levels and (2) grinding the rail on the system. A goal of 85 dBA would require, in addition to rail grinding, the installation of absorptive barriers along the system, but this would eliminate the requirement for a modified propulsion system. Goals of 75 and 80 dBA were not achievable with the noise abatement techniques included in the program and would require unrealistically low levels for wheel/rail, car specifications. 4-30 SOUND LEVEL - dBA SOUND LEVEL - dBA 90 80 FIGURE 4-7. MINIMUM COSTS TO REDUCE STATION NOISE - CTS 4-31 TABLE 4-17. COMPUTER SELECTED TECHNIQUES FOR STATION NOISE REDUCTION - CTS Station Noise Goal (dBA) Car Fix 85 Friction damped wheels 80 Wheel trueing Scenario Reference No. Line Fix 16 Rail grinding Open track gauge 16 Rail grinding Absorp. sidewalls Under -platform absorp Open track gauge Wheel trueing Rail grinding Friction damped wheels 12 Open track gauge New propulsion system 13 Rail grinding 14 Rail grinding 15 Rail grinding 16 Rail grinding Absorp. sidewalls Ceiling absorpt. Open track gauge Cost ($) 9,800 L2,800 1 , 658,960 4-32 TABLE 4-18. POTENTIAL IN-CAR NOISE REDUCTION TECHNIQUES - CTS Scenario Reference No. 21 22 23 2k 25 26 Scenario Car At grade At grade In cut In cut Subway Elevated embankment Noise Abatement Technique Visco elastic damped wheels Resilient wheels Interior car absorption Acoustic sealing of car Wheel trueing New car Friction damped wheels New propulsion system Modified propulsion system Rail grinding Absorptive sidewalls Open track gauge Rail grinding Absorptive sidewalls Open track gague Rail grinding Absorptive sidewalls Open track gauge Rail grinding Absorptive sidewalls Open track gauge Rail grinding Absorptive sidewalls Open track gauge Rail grinding Absorptive sidewalls Open track gauge 4-33 TABLE 4-19. IN-CAR NOISE ABATEMENT TECHNIQUES - CTS TECHNIQUE COSTS Scenario Reference Source (S) Reduction Fixed Per Unit No. Path (P ) (dBA) (?) ($) 1 Viscous damped wheels 0 1,200 Source = 2 2 Source = 3 2 Source = 13 2 o d Resilient wheels 0 9,600 Source = 2 3 Source = 3 3 Source = 13 3 R mJ Interior car absorption 0 1,000 Path = 2 3 Path = 3 3 Path = 4 3 Path = 9 3 4 Acoustic sealing car 0 500 Path = 3 5 Path = 4 5 Path = 9 5 5 Wheel trueing 0 1,000 Source = 13 20 Source = 3 6 6 New car 0 600,000 Source = 13 20 Source = 3 6 Source = 5 10 Source = 6 10 Path = 2 3 Path = 3 Path = 4 o o Path = 9 3 , 11 Rail grinding 0 1 Source = 2 5 15 Absorptive sidewalls tunnel 0 35 Path = 3 5 22 Friction damped wheels 0 400 Source = 2 2 Source = 3 2 Source = 13 1 23 New propulsion system 0 30,000 Source = 5 7 25 Mod propulsion system 0 4,000 Source = 5 4 27 Open track gauge 0 5 Source = 12 20 4-34 SOUND LEVEL - dBA SOUND LEVEL - dBA 100 90 80 FIGURE 4-8. MINIMUM COST TO REDUCE IN CAR NOISE - CTS 4-35 TABLE 4-20. COMPUTER SELECTED TECHNIQUES FOR IN-CAR NOISE REDUCTION - CTS In Car Scenario 'Toise Coal Reference Cost (dBA) Car* Fix No. Line Fix ($) 85 Acoust. seal, of car None 44,000 80 Internal car absorpt. None 167,200 Acoust. seal of car Friction damped wheels 75 Internal car absorp. 21 Rail grinding 524,260 Acoust. sealing of car Wheel trueing Open track gauge 23 Rail grinding Open track gauge 70 New car 21 Rail grinding 53,162,000 Open track gauge 22 Rail grinding *— Rail grinding Open track gauge 26 Rail grinding 4-36 TABLE 4-21. POTENTIAL WAYSIDE NOISE REDUCTION TECHNIQUES - PATCO Scenario Noise Abatement Technique Car Visco elastic damped wheels New car Friction damped wheels New propulsion system Equipment blower acoustic treatment Modified propulsion system Elevated embankment Rail grinding Non-absorptive barrier Absorptive barrier Concrete viaduct Rail grinding Non-absorptive barrier Absorptive barrier In cut Rail grinding Non-absorptive barrier Absorptive barrier 4-37 TABLE 4-22. POTENTIAL WAYSIDE NOISE ABATEMENT TECHNIQUES - PATCO TECHNIQUE COSTS Scenario Reference Source (S) Reduction Fixed Cost Unit Cost No. Path (P) (dBA) ($) ($) 1 Viscous damped wheels 0 1,200 Source = 2 1 Source = 3 1 6 New car 0 600,000 Source = 3 1 Source = 5 10 Sourch - 7 5 Path = 1 1 11 Rail grinding 5,000 1 Source = 2 3 13 Non-absorp barriers 0 100 Path = 1 7 l4 Absorp barriers 0 150 Path = 1 10 22 Friction damped wheels 0 400 Sourch = 2 1 Source = 3 1 23 New propulsion system 0 80,000 Source = 5 8 24 Equipment blower 0 6,000 Source = 7 5 2p Modified propulsion system 0 4,000 Source = 5 5 4-38 PATCO WAYSIDE FIGURE 4-9. MINIMUM COSTS TO REDUCE WAYSIDE NOISE - PATCO 4-39 TABLE 4-23. COMPUTER SELECTED TECHNIQUES FOR WAYSIDE NOISE REDUCTION - PATCO Wayside Noise Goal (dBA) Car Fix Line Fix 90 Modified propulsion system 85 ' None Rail grinding Rail grinding Absorptive barriers Cost ($) 309,752 722.552 (b) Station Noise Reduction The scenario system fixes listed in Table 4-3 were reviewed for application to station noise recution and the resulting fixes listed by scenario in Table 4-24. These noise control techniques along with the noise sources and paths which they reduce, the assumed noise reduction and costs are presented in Table 4-8. The numbers identi¬ fied by each source and paths refer to the sources and paths listed in Table 4-25. The results of the computer analysis to determine the minimum cost to reduce all station noise levels on the system to 90 and 85 dBA are shown in Figure 4-10 and tabulated in Table 4-26. To provide for all station noise levels on the system to be 90 dBA or lower requires absorptive sidewalls in the tunnels for Scenarios 12 and 13 — the line stations on the system where trains operate at normal subway speeds. A goal of 85 dBA or less for all stations involves the same scenarios, but in addition to absorptive tunnel sidewalls, requires rail grinding and absorption between rails on the concrete invert, and for Scenario 13 the same treatment without rail grinding. It was not possible to achieve station noise levels in the subway less than 82 dBA by the methods (fixes) listed. To do so would require car noise levels which are substantially lower than the existing Budd cars as well as any of the other cars currently under construction for other transit systems even with the stringent noise control requirements which have been imposed on them. Noise levels of 80 dBA are achievable at stations located at grade or on elevated embankment, but a uniform goal of less than 80 dBA throughout the system might not be achievable utilizing existing technology. (c) In-Car Noise Reduction The system noise control fixes selected for application to in-car noise reduction are presented in the listing shown in Table 4-27. The resulting techniques, the sources and paths related to them, the noise reduction assigned to them for this analysis and the associated fixed and unit costs are presented in Table 4-28. The results of the computer analysis to determine the minimum cost to reduce in car noise levels to 80, 75 and 70 dBA are shown in Figure 4-11 and tabulated in Table 4-29. For uniform noise goals to be achieved, both car and line fixes must be installed. Not unexpectedly, the lines fixes involve subway scenarios only. Plateau levels of 66 dBA might be achievable, as shown in Table 4-29, but may not be desirable from a speech privacy standpoint. 4. BROAD STREET SUBWAY (a) Wayside Noise Reduction Potential noise abatement techniques presented in Table 4-3 were reviewed for applicability to the short length of the Broad Street line which is located on grade 4-41 TABLE 4-24. POTENTIAL STATION NOISE REDUCTION TECHNIQUES - PATCO Scenario Scenario Reference No. Car 11 Subway-Terminal (Jointed track) 12 Subway-Line (Jointed rail) 13 Subway-r Line (Welded) 17 Subway-Line Slow speed (Welded) lb. Elevated embankment (Welded) Noise Abatement Technique Visco elastic damped wheels New car Friction damped wheels New propulsion system Equipment blower acoustic treatment Modified propulsion system Rail grinding Absorptive tunnel sidewalls Station ceiling absorption Station wall absorption Station under-platform absorption Station absorption - concrete invert Rail grinding Absorptive tunnel sidewalls Station ceiling absorption Station wall absorption Station under-platform absorption Station absorption - concrete invert Rail grinding Absorptive tunnel sidewalls Station ceiling absorption Station wall absorption Station under-platform absorption Station absorption - concrete invert Rail grinding Absorptive tunnel sidewalls Station ceiling absorption Station wall absorption Station under-platform absorption Station absorption - concrete invert Rail grinding Absorptive tunnel sidewalls Station ceiling absorption Station wail absorption Station under-platform absorption Station absorption - concrete invert 4-42 TABLE 4—24. (Continued) Scenario Reference No. Scenario Noise Abatement Techniques 15 Concrete Viaduct (Welded) Rail grinding Absorptive tunnel sidewalls Station ceiling absorption Station wall absorption Station under-platform absorption Station absorption - concrete invert 16 In-Cut (Welded) Rail grinding Absorptive tunnel sidewalls Station ceiling absorption Station wall absorption Station under-platform absorption Station absorption - concrete invert 4-43 1 6 11 15 16 17 18 19 22 23 24 25 TABLE 4-25. STATION NOISE ABATEMENT TECHNIQUES - PATCO TECHNIQUE COSTS Source (S) Path (P) Reduction (dBA) Fixed ($) Viscous damped wheels Source =2 1 Source =3 1 New car Source =2 1 Source =3 1 Source =5 10 Source =7 5 Path =5 1 Rail grinding Source =2 3 Absorptive sidewalls tunnel Path =12 8 Station ceiling absorption Path =6 * 2 Station wall absorption Path =6 2 Station under-platform absorption Path =6 3 Station absorption concrete invert Path =6 6 Friction damped wheels Source =2 1 Source =3 1 New propulsion system Source =5 8 Equipment blower Source =7 5 Modified propulsion system Source =5 5 0 0 5,000 0 0 0 0 0 0 0 0 0 Per Unit ($) 1,200 600,000 1 35 20 20 25 5 400 80,000 6,000 4,000 4-44 PAT CO STATION FIGURE 4-10. MINIMUM COSTS TO REDUCE STATION NOISE - PATCO 4-45 TABLE 4-26. COMPUTER SELECTED TECHNIQUES FOR STATION NOISE REDUCTION - PATCO Station Scenario oise Goal Reference Cost (dBA) Car Fix No. Line Fix 9CNone . 12 Absorptive sidewalls tunnel 81,200 13 Absorptive sidewalls tunnel 35Frict. damped wheels 12 Rail grinding 429,540 Absorptive sidewalls tunnel Sta abs. cone. inv. Modified propulsion system 13 Absorptive sidewalls tunnel Sta. abs. cone. inv. 82l T ew car 12 Rail grinding 4 5,134,340 13 Rail grinding Abs. sidewalls tunnel Sta. ceil. abs. Sta. abs. cone. inv. 4-46 TABLE 4-27. POTENTIAL IN-CAR NOISE REDUCTION TECHNIQUES - PATCO Scenario Reference Scenario Noise Abatement Technique No. Car Visco Elactic damped wheels Interior car absorption Acoustic sealing of car New car Friction damped wheels New propulsion system Equipment blower Modified propulsion system 21 Subway-Line (Jointed rail) 31681 Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 22 Subway-Line (Jointed rail) 7392 1 Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 23 Subway-Line (Welded rail) 7392 Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 24 Subway- (Welded rail) 8448 Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 25 Bridge Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 26 Elevated embankment 64416 1 Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 27 Elevated embankment 23232 1 Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 4-47 TABLE 4—27. (Continued) Scenario Reference No. Scenario Noise Abatement Technique 28 Concrete viaduct 29 In Cut 4224 30 In Cut 6336 Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel Weld rails Rail grinding Absorptive sidewalls in tunnel Ballast in tunnel 4-48 TABLE 4-28. IN-CAR NOISE ABATEMENT TECHNIQUES - PATCO TECHNIQUE COSTS Scenario Reference Source (S) Reduction Fixed Cost No. Path (P) (dBA) ($) Unit Cost ($) 1 Viscous damped wheels Source =2 1 Source =3 1 3 Interior car absorption Path =2 3 Path =3 3 Path = 4 3 Path =3 3 Path =9 3 k Acoustic sealing of car (Doo Path =3 5 6 New car Source =3 1 Source = 5 10 Source =6 5 Source =7 5 Path = 2. 3 Path =3 8 Path =4 8 Path =9 8 9 Weld rails Source =1 7 11 Rail grinding Source =2 3 15 Absorptive sidewalls in tunnel Path =3 5 22 Friction damped wheels Source =2 1 Source =3 1 23 New propulsion system Source =5 7 24 Equipment blower Source =7 5 25 Modified propulsion system Source =5 4 26 Ballast in tunnel Path =3 6 0 0 0 0 0 5,000 0 0 0 0 0 0 1,200 1,000 500 600,000 25 1 35 L00 80,000 6,000 4,000 5 4-49 COST FOR NOISE ABATEMENT PATCO-IN CAR 90 FIGURE 4-11. MINIMUM COST TO REDUCE IN CAR NOISE - PATCO TABLE 4-29. COMPUTER SELECTED TECHNIQUES FOR IN-CAR NOISE REDUCTION - PATCO In Car Scenario Noise Goal Reference Cost (dBA) Car Fix No. Line Fix ($) 80 Acoustic sealing of car None 37,500 75 Interior car absorption 21 Ballast in tunnel 165,300 Acoustic sealing of car 23 Ballast in tunnel 70 Interior car absorption 21 Ballast in tunnel 6,195,300 Acoustic sealing of car 23 Ballast in tunnel Friction damped wheels New propulsion system 66 New car 21 Rail grinding 45,060,968 Ballast in tunnel 23 Ballast in tunnel 4-51 near Fern Rock station.. The potential techniques are shown in Table 4-30 and also in Table 4-31 along with the noise sources and paths which they reduce and their assumed noise reduction and costs. The computer selected techniques to reduce community wayside noise to goals of 75, 70, 65 and 60 dBA are shown in Figure 4-12 and tabulated in Table 4-32. Although, goals below 75 dBA were generally not considered for this program, the analysis showed that levels of 60 dBA could be achieved with the techniques available to the analysis. The reduction in wayside noise predicted by the analysis would benefit residents along a short section of right-of-way only — 1,200 ft. Note that an expenditure of $120,000 for non-absorptive barriers would immediately reduce wayside levels at 50 ft to below 70 dBA. Reductions in levels below this goal increases rapidly. (b) Station Noise Reduction The noise reduction techniques presented in Table 4-3 were reviewed for applicability to station noise reduction on the Broad Street subway. All the potential techniques considered applicable to the system are shown in Table 4-33. These techniques, along with the noise sources and paths which they produce, the assumed noise reduc¬ tion and costs are presented in Table 4-34. The computer selected techniques to reduce station noise levels to goals of 95, 90 and 85 dBA are shown in Figure 4-13 and tabulated in Table 4-35. Levels below 85 dBA were not achievable by the methods under study. Although wheel truing is already being performed on the Market-Frankford line on a regular basis — each car is rotated through this aspect of the maintenance program on an annual basis — the achievement of a goal of 90 dBA has selected this technique on a more frequent basis as a method for achieving this noise level. A level of 85 dBA for a station noise goal would require a procurement of new cars specifically designed with stringent noise goals in the procurement specifications. In addition, improved rail joints, rail grinding and for some stations, absorptive treatment of the station itself, would be required.. (c) In-Car Noise Reduction The potential system noise control fixes selected for in-car noise reduction are shown in the listing of Table 4-36. Noise abatement techniques, the sources and paths which they relate to and the noise reduction values and unit costs assumed for the analysis are presented in Table 4-29. Results of the computer analysis to select the minimum cost to reduce in-car noise levels to goals of 95, 90, 85, and 80 are shown in Figure 4-14 and Table 4-38. Although data used in the computer program was based on measurements of the 1928 cars, these levels were assumed to apply to all cars in the system in order to simplify the analysis. 4-52 TABLE 4-30. POTENTIAL WAYSIDE NOISE REDUCTION TECHNIQUES - SEPTA BROAD STREET SUBWAY Scenario Reference No. Scenario Noise Abatement Technique Car 1 At grade Visco elastic damped wheels Resilient wheels Wheel trueing New car Friction damped wheels New propulsion sysaem Modified propulsion system Modified brake air compressor Weld rail Improve joints Rail grinding Non absorptive barriers Assorptive barriers 4-53 TABLE 4-31. WAYSIDE NOISE ABATEMENT TECHNIQUES - SEPTA BROAD STREET SUBWAY Scenario Reference Source (S) No. p a th (P) TECHNIQUE Reduction (dBA) COSTS Fixed Cost ($) Unit Cost ($) 1 Viscous damped wheels 0 1,200 Source = 2 1 Source = 3 1 Source = 13 1 2 Resilient wheels 0 9,600 Source = 2 2 Source = 3 2 Source = 13 2 5 Wheel trueing 0 1,000 Source = 3 6 Source = 13 15 6 New car 0 600,000 Source = 3 2 Source = 5 15 Source = 6 10 Source = 13 15 9 Weld rails 0 25 Source = 1 7 10 Improve joints 0 2 Source = 1 2 11 Rail grinding 0 1 Source = 2 4 13 Non absorptive barriers 0 100 Path = 1 7 14 Absorptive barriers 0 150 Path * = 1 10 22 Friction damped wheels 0 400 Source = 2 1 Source = 3 1 Source = 13 1 23 New propulsion system 0 40,000 Source = 5 7 C\J Modified propulsion system 0 2,000 Source = 5 4 29 Modified brake air compressor 0 300 Source = 6 10 4-54 SOUND LEVEL - dBA SOUND LEVEL - dBA FIGURE 4-12. MINIMUM COSTS TO REDUCE WAYSIDE NOISE - SEPTA BROAD STREET SUBWAY 4-55 TABLE 4-38. COMPUTER SELECTED TECHNIQUES FOR WAYSIDE NOISE REDUCTION - SEPTA BROAD STREET SUBWAY Wayside Noise Goal (cLBA) Car Fix Scenario Reference No. Line Fix Cost ($) 75 None 1 Non absorptive barriers 120,000 70 None 1 Non absorptive barriers 120,000 65 Nile el trueing 1 Rail grinding Absorptive barriers 381,200 65 Nile el trueing New propulsion Modified brake 1 system air comp. Weld rails Railing grinding Absorption barriers 8,471,200 * 4-56 TABLE 4-33. POTENTIAL STATION NOISE REDUCTION TECHNIQUES - SEPTA BROAD STREET SUBWAY Scenario Scenario Noise Abatement Technique Reference No. Car Visco-elastic damped wheels Resilient wheels Wheel trueing Friction damped wheels New propulsion system Modified propulsion system Modified brake air compressor 11 At grade terminal Weld rails Improve joints Grind rails Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Concrete invert absorption 12 Subway terminal Weld rails Improve joints Grind rails Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Concrete invert absorption 13 New subway station Weld rails Improve joints Grind rails Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Concrete invert absorption 14 Type B station Weld rails Improve joints Grind rails Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Concrete invert absorption 4-57 TABLE 4—33. (Continued) Scenario Reference No. Scenario 15 Special Type A 16 Type B station 17 Type B station Noise Abatement Technique Weld rails Improve joints Grind rails Ab sorptive sidevails Ceiling absorption Wall absorption Under-platform absorption Concrete invert absorption Weld rails Improve joints Grind rails Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Concrete invert absorption Weld rails Improve joints Grind rails Absorptive sidewalls Ceiling absorption Wall absorption Under-platform absorption Concrete invert absorption 4-58 TABLE 4-34. STATION NOISE ABATEMENT TECHNIQUES - SEPTA BROAD STREET SUBWAY TECHNIQUE COSTS Scenario Reference Source (S) Reduction Fixed Per Unit No. Path (P) (dBA) ($) ($) 1 Viscous damped wheels 0 1,200 Source = 2 1 Source = 3 1 Source = 13 1 2 Resilient wheels 0 9,600 Source = 2 2 Source = 3 2 Source = 13 2 5 Wheel trueing 0 1,000 Source = 3 6 Source = 13 15 6 New car 0 600,000 Source = 3 2 Source = 5 15 Source = 6 10 Source = 13 15 Path = 5 1 9 Weld rails 0 25 Source = 1 7 10 Improve joints 0 2 Source = 1 2 n Rail grinding 0 1 Source = 2 4 15 Absorptive sidewalls tunnel 0 35 Path = 12 8 16 Station ceiling absorption 0 34 Path = 6 2 17 Station wall absorption 0 20 Path = 6 2 18 Station under-platform absorption 0 27 Path = 6 3 ' 19 Station absorption concrete invert 0 8 Path = 6 6 , 22 Friction damped wheels 0 400 Source = 2 1 Source = 3 1 Source = 13 1 23 New propulsion system 0 40,000 Source = 5 7 25 Modified propulsion system 0 2,000 Source = 5 4 29 Modified brake air compressor . 0 300 Source = 6 10 4-59 95 90 FIGURE 4-13. MINIMUM COSTS TO REDUCE STATION NOISE SEPTA BROAD STREET SUBWAY 4-60 TABLE 4-35. COMPUTER SELECTED TECHNIQUES FOR STATION NOISE REDUCTION - SEPTA BROAD STREET SUBWAY Station Scenario Noise Goal Reference (dBA) Car Fix No. Line Fix 95 None 17 Rail grinding Absorp. cone, invert 90 Wheel trueing 17 Rail grinding Mod. propulsion system Absorp. sidewalls 0 Absorp. cone, invert 85 New car 13 Improve joints Grind rail 14 Absorp. cone, invert 15 Absorp. cone, invert 16 Improve joints Grind rail 17 Weld rails Grind rails Absorp. sidewalls Under platform absorp. Absorp. cone, invert Cost ($) 34,650 769,400 4o4,8oo 4-61 TABLE 4-36. POTENTIAL IN-CAR NOISE REDUCTION TECHNIQUES - SEPTA BROAD STREET SUBWAY Scenario Reference No. Scenario Car 21 2 Track Ballast and tie at grade 22 k Track Concrete invert 23 h Track Concrete invert 2b 2 Track Concrete invert 25 2 Track Concrete invert Boise Abatement Technique Visco-elastic damped wheels Resilient wheels Interior car absorption Acoustic seeling of car wheel trueing New car Friction damped wheels New propulsion system Modified propulsion system Modified brake air compressor Weld rails Improve joints Rail grinding Resilient fasteners Absorptive sidewalls Ballast in tunnel Weld rails Improve joints Rail grinding Resilient fasteners Absorptive sidewalls Ballast in tunnel Weld rails Improve joints Rail grinding Resilient fasteners Absorptive sidewalls Ballast in tunnel Weld rails Improve joints Rail grinding Resilient fasteners Absorptive sidewalls Ballast in tunnel Weld rails Improve joints Rail grinding Resilient fasteners Absorptive sidewalls Ballast in tunnel 4-62 Scenario Reference No. 1 2 3 4 5 6 9 10 11 12 15 22 23 25 26 29 TABLE 4-37. IN-CAR NOISE ABATEMENT TECHNIQUES - SEPTA BROAD STREET SUBWAY Technique Costs Source (s) Reduction Fixed Cost Unit Cost Path (P) (dBA) ($) ($) Viscous damped wheels 0 1,200 Source = 2 1 Source = 3 1 Source = 13 1 Resilient wheels 0 9,600 Source = 2 2 Source = 3 2 Source = 13 2 Interior car absorption 0 1,200 Path = 2 4 Path = 3 4 Path = 4 4 Path = 9 4 Acoustic sealing of car 0 500 Path = 3 5 Wheel trueing 0 1,000 Source = 3 3 Source = 13 15 New car 0 600,000 Source = 3 2 Source = 5 15 Source = 6 10 4 Source = 13 15 Path = 2 3 Path = 3 8 Path = 4 8 Path = 9 8 Weld rails 0 25 Source = 1 7 Improve joints 0 d. Source = 1 d Rail grinding 0 1 Source = 2 4 Resilient fasteners 0 10 Path = 3 3 Absorptive sidewalls tunnel 0 40 Path = 3 5 Friction damped wheels 0 400 Source = 2 1 Source = 3 1 Source = 13 1 New propulsion system 0 40,000 Source = 5 7 Modified propulsion system 0 2,000 Source = 5 4 Ballast in tunnel 0 5 Path = 3 6 Modified brake air compressor 0 300 Source = 6 10 4-63 SOUND LEVEL - dBA SOUND LEVEL - dBA FIGURE 4-14. MINIMUM COST TO REDUCE IN-CAR NOISE - SEPTA BROAD STREET SUBWAY 4-64 TABLE 4-32. COMPUTER SELECTED TECHNIQUES FOR IN-CAR NOISE REDUCTION - SEPTA BROAD STREET SUBWAY In Car Scenario Noise Goal Reference (dBA) Car Fix No. Line Fix 95 None 25 Ballast in tunnel 90 Interior car absorption 85 Interior car absorption Acoustic sealing of car Modified propulsion 80 Interior car absorption Acoustic sealing of car Wheel trueing Friction damped wheels New propulsion system Mod. brake air compressor 25 Ballast in tunnel 25 Rail grinding Ballast in tunnel 23 Rail grinding 2h Rail grinding 25 Weld rails Rail grinding Absorptive sidewalls Ballast in tunnel Cost ($) 68,640 308,6U0 822,368 9,777,712 4-65 The methods chosen do not always account for other system restraints. For example, the car fix of Interior Car Absorption' does not recognize maintenance, durability, vandal resistance or other obvious characteristics which would have to be satisfied for such an approach to be adopted. The analysis offers basic approached to noise reduction only and does not attempt to define a specific acoustic treatment. Note that a goal reduction from 80 to 77 dBA requires an order of magnitude change in costs to achieve. 5. MARKET-FRANKFORD LINE (a) Wayside Noise Reduction The list of potential noise abatement techniques presented in Table 4-3 has been reviewed for applicability to the Market Frankford wayside community. The applicable techniques are summarized in Table 4-39 and also presented in Table 4-40 with the appropriate noise sources and paths involved, the assumed noise reductions and costs. The sources and paths refer to those listed in Table 4-1. The results of the computer analysis to determine minimum costs to reduce wayside noise levels to selected goals of 90, 85, and 80 dBA are shown in Figure 4-15 and tabulated in Table 4-41. For purposes of the analysis, it was assumed that all elevated structure will eventually permit vehicle speeds that were in use during the survey along the Market Street section of the line. Thus, although noise levels measured along elevated structure north of 2nd Street were 2-3 dBA lower than between 46th and 63rd Streets, the data input to the computer analysis was increased by a similar amount to account for speed restrictions during construction along the northern segment of the line. An analysis which should be performed prior to initiating any wayside noise reduc¬ tion program is the determination of the Equivalent Noise Levels (Leq) or Day Night Level (Ldn) for the community without transit system operation. This would identify the noise level below which the transit system is not contributing to com¬ munity noise. It would also set a lower limit for transit system noise goals until such time as other community noise sources are identified by regulation for maximum permissible noise. This type of investigation is not within the scope of the current program. Reduction of wayside community noise levels to 90 and 85 dBA are achieved by the line and car fixes identified in Table 4-41. A reduction to 80 dBA would require installation of resilient wheels on all cars of the system and a program for even more frequent truing of wheels than currently in effect. Reference to the modified propulsion system indicates the requirement for a reduction in noise level of that unit. Note on the curve of Figure 4-15 that a goal of 82 dBA could be achieved for significantly less expenditures (approximately 1/10th) than an 80 dBA goal. 4-66 TABLE 4-39. POTENTIAL WAYSIDE NOISE REDUCTION TECHNIQUES - SEPTA MARKET-FRANKFORD LINE Scenario Reference No. Scenario Noise Abatement Technique Car 1 Way 1 elevated Visco-elastic damped wheels Resilient wheels Wheel trueing New car Friction damped wheels New propulsion system Modified propulsion system Modified brake air compressor Weld rails Improve joints Rail grinding Non-absorptive barrier Absorptive barrier 2 V/ay 2 elevated Way 3 On-grade 4 Way 4 On-grade Weld rails Improve joints Rail grinding Non-absorptive barriers Absorptive barriers Weld rails Improve joints Rail grinding Non-absorptive barriers Absorptive barriers Weld rails Improve joints Rail grinding Non-absorptive barriers Absorptive barriers 4-67 TABLE 4-40. WAYSIDE NOISE ABATEMENT TECHNIQUES AND COSTS - SEPTA MARKET-FRANKFORD LINE Scenario Reference No. Technique Costs Source (S) Path (P) Reduction (dBA) Fixed cost ($) Unit Cost 1 r— P 6 o y 10 11 13 14 22 23 25 29 Viscous damped wheels Source = 2 Source = 3 Source = 13 Resilient wheels Source = 2 Source = 3 Source = 13 Wheel trueing Source = 3 Source = 13 New car Source = 3 Source = 5 Source = 7 Source = 13 Path - 1 Weld rails Source = 1 Improve joints Source = 1 Rail grinding Source = 2 2 O c. 5 3 3 3 8 3 10 8 8 1 10 Ton-absorptive barriers Path =1 7 Absorptive barriers Path ‘ = 1 10 Friction damped wheels Source =2 2 Source =3 2 Source =13 1 New propulsion system Source =5 7 Modified propulsion system Source =5 5 Modified brake air compressor Source =6 10 0 0 0 0 0 0 0 0 0 0 0 0 0 1,200 1,000 600,000 25 o £ 1 100 U00 80,000 4,000 300 4-68 SOUND LEVEL - dBA SOUND LEVEL - dBA 95 -i 90- FIGURE 4-15. MINIMUM COSTS TO REDUCE WAYSIDE NOISE - SEPTA MARKET-FRANKFORD LINE 4-69 TABLE 4-41. COMPUTER SELECTED TECHNIQUES FOR WAYSIDE NOISE REDUCTION - SEPTA MARKET-FRANKFORD LINE Wayside Noise Goal (dBA) Cor Fix Scenario Reference No. Line Fix Cost ($) 90 None 2 Weld rails Rail grinding 13,000 8 5 Friction damped wheels 1 Improve rail joints 246,292 2 Improve rail joints Rail grinding Non-absorptive barriers 30 Resilient wheels Wheel trueing Modified propulsion system 1 Non-absorptive barriers 3,353,400 2 Weld rails Rail grinding absorptive barriers V 4-70 (b) Station Noise Reduction The system fixes listed in Table 4-3 were reviewed for applicability to Market- Frankford station noise reduction. The resulting techniques are listed by scenario in Table 4-42 and these techniques along with the noise sources and paths which they reduce. The assumed noise reductions and associated costs are presented in Table 4-43. The numbers identified by each source and path refer to those listed in Table 4-1. The results of the computer analysis to determine the minimum cost to reduce all stations noise levels on the system to goals of 95 to 90 dBA are shown in Figure 4-16 and tabulated in Table 4-44. Goals below 90 dBA at stations were not achievable by the techniques presented in Table 4-42. A goal of 95 dBA could be achieved with no change to the existing cars by either of two sets of line fixes, as shown in Table 4-33. A 90 dBA level would require substational absorption in the subway stations. Reduction to goals below this level could be realized at above ground stations, but achieving a uniform level 85 dBA at all stations would be limited by the inability to reduce noise in subway stations to 85 dBA. Note that a reduction in noise below 87 dBA would require the expendi¬ ture of over six times the funds required for a goal of 88 dBA. (c) In-Car Noise Reduction The system noise control fixes selected for application to in-car noise reduction are shown in the listing of Table 4-45. The noise abatement techniques, the sources and paths which they relate to and the noise reduction values and unit costs assumed for this analysis are presented in Table 4-46. The results of the computer analysis to select the minimum cost to reduce in car noise levels to 85, 80 and 75 dBA are shown in Figure 4-17 and tabulated in Table 4-47. A reduction in interior noise to 85 dBA would require acoustic sealing of car doors with no line fixes identified. A goal of 80 dBA would require added absorption within the car, in addition to door sealing and improved rail joints in the subway scenario where welded rail is not already employed. Reduction to 75 dBA would require nearly four times the expenditures required to achieve 80 dBA and the car and line fixes would be as identified in Table 4-47. » 4-71 TABLE 4-42. POTENTIAL STATION NOISE REDUCTION TECHNIQUES - SEPTA MARKET-FRANKFORD LINE Scenario Reference No. Scenario Car 11 Station 1 - elevated 12 Station 2 - at grade 13 Station 3 - elevated l4 Station 4 - at grade 15 Station 5 - subway 16 Station 6 - subway Noise Abatement Technique Visco-elastic damped wheels Resilient wheels Wheel trueing New car Friction damped wheels Non-propulsion system Modified propulsion system Rail grinding Ceiling absorption Wall absorption Under platform absorption Absorption on concrete invert Rail grinding Ceiling absorption Wall absorption Under platform absorption Absorption on concrete invert Rail grinding Ceiling absorption Wall absorption Under platform absorption Absorption on concrete invert Rail grinding Ceiling absorption Wall absorption Under platform absorption Absorption on concrete invert Rail grinding Absorptive sidewalls of tunnel Ceiling absorption Wall absorption Under platform absorption Absorption on concrete invert Rail grinding Absorptive sidewalls of tunnel Ceiling absorption Wall absorption Under platform Absorption on concrete invert TABLE 4-43. STATION NOISE ABATEMENT TECHNIQUES - SEPTA MARKET-FRANKFORD LINE Scenario Reference No. Technique Costs Source (S) Reduction Fixed Per Unit Path (P) (dBA) ($) ($) 1 Viscous damped wheels 0 1,200 Source = 2 2 Source = 3 2 Source = 13 2 2 Resilient wheels Source = 2 3 0 9,600 Source = 3 3 Source = 13 3 5 Wheel trueing Source = 3 3 0 1,000 Source = 13 8 6 New car Source = 3 6 0 600,000 Source = 5 10 Source = 6 10 Source = 13 20 11 Rail grinding Source = . 2 3 0 1 15 Absorptive sidewalls tunnel, cut 0 35 Path = 12 8 16 Station ceiling absorption 0 20 Path = 6 2 17 Station wall absorption Path = 6 2 0 20 18 Station under platform absorption 0 35 Path = 6 3 19 Station absorption on concrete invert 0 5 Path = 5 3 22 Friction damped wheels Source = 2 2 0 bOO Source = 3 2 Source = 13 1 23 New Propulsion system Source = 5 7 0 80,000 25 Modified propulsion system 0 L,000 Source = 5 5 4-73 35 50 100 _±_ 150 0 MILLIONS OF DOLLARS FIGURE 4-16. MINIMUM COSTS TO REDUCE STATION NOISE - SEPTA MARKET-FRANKFORD LINE 4-74 TABLE 4-44. COMPUTER SELECTED TECHNIQUES FOR STATION NOISE REDUCTION - SEPTA MARKET-FRANKFORD LINE Station Noise Goal (cLBA) Car Fix Scenario Reference No. Line Fix 95 None 6 Ceiling absorption Absorption on concrete invert or Wall absorption Absorption on concrete invert 90 Friction damped wheels Modified propulsion system 6 Rail grinding Absorptive sidewalls in tunnel Ceiling absorption Wall absorption Under platform absorption Absorption on concrete invert Cost ($) 70,000 ,526,000 4-75 TABLE 4-45. POTENTIAL IN-CAR NOISE REDUCTION TECHNIQUES - SEPTA MARKET-FRANKFORD LINE Scenario Reference Scenario No. Car 21 At-grade Ballast and tie track 2 track Noise Abatement Technique Visco-elastic damped wheels Resilient wheels Interior car absorption Acoustic sealing of doors Wheel trueing New car Friction damped wheels New propulsion system Modified equipment blower Modified propulsion system Modified brake air compressor Weld rails Improve Joints Rail grinding Resilient fasteners 22 At-grade Ballast and tie track 2 track 23 Elevated Ballast and tie track 2 tracks 2h Elevated Ballast and tie track 2 tracks 25 Subway Concrete invert 2 tracks 26 Subway Concrete invert 2 tracks Weld rails Improve Joints Rail grinding Resilient fasteners Weld rails Improve Joints Rail grinding Resilient fasteners Damping - elevated steel structure Weld rails Improve Joints Rail grinding Resilient fasteners Damping elevated steel structure Weld rails Improve Joints Rail grinding Resilient fasteners Absorptive sidewalls - tunnel Ballast in tunnel Weld rails Improve Joints Rail grinding Resilient fasteners Absorptive sidewalls - tunnel Ballast in tunnel 4-76 TABLE 4-46. IN-CAR NOISE ABATEMENT TECHNIQUES - SEPTA MARKET-FRANKFORD LINE Scenario Reference Source (S) Technique Reduction No. Path (P) (dBA) Costs Fixed Cost ($) Unit Cost ($) 1 Viscous damped wheels Source = 2 1 Source =3 1 Source =13 1 2 Resilient wheels Source =2 2 Source =3 2 Source =13 2 3 Interior car absorption Path =2 4 Path =3 4 Path =4 4 Path =8 4 4 Acoustic sealing doors Path =3 5 Path =9 5 5 Wheel trueing Source =3 3 Source =13 8 6 New car Source =3 2 Source =5 10 Source =6 8 Source =13 8 9 Weld rails Source =1 10 10 Improve joints Source =1 6 11 Rail grinding Source^ =2 4 12 Resilient fasteners Path =3 3 Path =9 3 15 Absorptive sidewalls tunnel Path =3 5 21 Damping steel elevated structure Path' =4 3 Path =9 3 22 Friction damped wheels Source =2 1 Source =3 1 Source =13 1 23 New propulsion system Source =5 7 24 Modified equipment blower Source =6 4 25 Modified propulsion system Source =5 4 26 Ballast in tunnel Path =3 6 29 Modified brake air compressor Source =6 10 0 1,200 9,600 0 1,200 0 0 0 500 1,000 600,000 0 0 0 0 25 2 1 10 0 40 0 50 0 400 0 0 0 0 0 4o,ooo 250 2,000 5 300 4-77 FIGURE 4-17. MINIMUM COSTS TO REDUCE IN-CAR NOISE - SEPTA MARKET-FRANKFORD LINE 4-78 TABLE 4-47. COMPUTER SELECTED TECHNIQUES FOR IN-CAR NOISE REDUCTION - SEPTA MARKET-FRANKFORD LINE In Car Scenario Noise Goal Reference Cost (d£A) Car Fix No. Line Fix ($) 35 Acoustic sealing of doors None 136,500 80 Interior car absorption ■25 Improve rail joints 533,796 Acoustic sealing of doors 75 Interior car absorption 2b Rail grinding 2 , 102,988 Acoustic sealing of doors 25 Improve joints r lieel trueing Rail grinding Friction damped wheels Resilient fasteners Modified propulsion system Ballast in tunnel * 4-79 5. PROGRAM SUMMARY 5.1 BACKGROUND AND PURPOSE The Urban Mass Transportation Administration (UMTA), through the Transportation Systems Center (TSC) has funded programs to assess the noise produced by urban rail transit operations in the United States and has been appraising noise reduction methods and costs. The characterization of the noise climate on each rail transit system will provide UMTA with data upon which priorities and funding decisions can be based. The noise assessment program allows comparisons among systems, among different types of equipment or track structure on the same system, between measured noise levels and potential regulations, and before-and- after noise control measures. The program has had three distinct elements: a. Noise climate assessment b. Review of abatement technique options c. Estimates of costs of abatement to specified noise level goals The assessment task, which has been reported in four volumes for the subject program (Ref 1) provides data which has been used to identify noise sources and paths and their relative contri¬ bution to the total acoustical signature for wayside, station and in-car locations. Noise abate¬ ment techniques which are applicable in general to rail transit systems have been summarized and then applied in more specific terms to each of the systems under study. To assess each of the systems, each rapid transit system included in the study was surveyed and classified by vehicle, station type, roadbed construction and community location. Representa¬ tive measurement locations were then defined for each of these categories as well as for other locations with specific singularities. Similar approaches have been adopted by investigators for other noise assessment studies conducted recently (eg. Reference 2). Coordination of all these programs has been accomplished through the Transportation Systems Center. The type of measurements made were not intended to be used for detailed diagnostic studies whereby predominant noise sources and paths are investigated, the contribution of specific vehicle component sources identified, or the effect of vehicle speed on the various parameters investigated. However, they did serve to describe the existing system noise climate and provide sufficient information to determine the first order noise abatement techniques described pre¬ viously in this report which have been used to satisfy reduced noise level criteria. The abate¬ ment techniques have been divided into two general categories: (1) rail vehicle and (2) track, roadbed and associated structures. As previously noted, the techniques were tested on opera¬ tional transit lines and can be implemented on other existing systems, where practical con¬ straints don't rule otherwise. The noise reduction effectiveness noted for each technique is based on field data taken on one or more transit systems. 5-1 "Use or disclosure of proposal data is subject to the restriction on the Title page of this Proposal. (1966 DEC)". These techniques were then reviewed for specific application to each transit property under study and a discussion prepared on the effectiveness of each method as applied to vehicle noise reduction and track and way structure noise reduction. Having identified noise levels (the assessment program) and abatement techniques applicable to each system, costs were estimated for implementing each technique and the minimum costs determined to reduce noise levels on each system to specified goals. Actual costs are difficult to estimate. For example, the cost of a new car was set at $600,000 for this study. Obviously this figure will vary according to the requirements of a specific car specification, but it is in the range of current car costs. Costs for applying a specific technique to reduce noise for commuters, employees and the community, along with a value (in dB) for the amount of noise reduction each technique would achieve, were input to the cost/abatement program. The program was run separately for community, station and in-car scenarios, and not for the entire system as a single case. Thus some costs might be duplicated if the costs for community, station and in-car noise reduction were summed. For example, if in-car noise reduction required rail grinding to achieve a specific goal, the cost for rail grinding for station and community noise reduction would not be required ) a second time for station noise reduction or a third time for the community. This type of cost editing has not been performed. For each noise source, an estimate was made of its contribution to the total noise at a given receiver location, whether it was for a person in the community, on the station platform or in the transit vehicle. The paths by which the noise arrived at the listener (receiver) were also determined and the acoustical significance of each assessed. This information was then input to the TSC cost-abatement algrorithm and the results summarized. 5-2 6. CONCLUSIONS AND RECOMMENDATIONS 4 Several conclusions have been drawn from the noise environment surveys which formed the first series of reports for this program, as well as the present study which encompassed identification of proven noise abatement techniques and the application of these techniques at minimum costs to the transit systems surveyed. These conclusions are summarized below. The order in which they are listed is not considered significant. 1. Cost increments to reduce in-car, station and community noise levels are generally non¬ linear with noise reduction achieved. Frequently, the first incremental cost is the most effective in reducing system levels. 2. Noise goals below 80 dBA for the entire system, whether in-car, station or community, are generally not obtainable through practical modifications to existing equipment. Q>> Although costs and techniques to achieve uniform noise goals were identified by the pro¬ gram, an approach which allows some variations in noise throughout the sytem might be more cost effective. For example, not all communities may require noise criteria as stringent as others, particularly in cases where the residential community is located at a greater distance from the system than where transit noise can be heard. In some instances, however, it may be desirable to plan for future development along the system and provide uniformity low noise levels at the wayside even though segments of the transit property doesn't currently border residential communities. Adoption of a specific policy probably is a task to be resolved only through the agencies from which funding originates. 4. The subject program, which surveyed existing noise sources, was not sufficiently detailed and complete to quantify all car sources and paths for a detailed noise reduction program. The program does form the background which points the way to more detailed surveys of specific noise sources. Any additional work would require narrowband analysis (a resolution of 1/3 octane band or letter) with only the final solutions expressed in terms of dBA. 5. The costs used in the study were estimated by the authors based on discussions with transit authority engineers, car builder marketing experts and acoustical consultants. They should be recognized for what they are — estimates of costs at the time the study was performed. These costs can be utilized for order-of-magnitude planning, for relative ranking of prob¬ lems and for determining priorities, but they must not be interpreted as absolute costs since the level of effort afforded that portion of the study was insufficient for other than the purposes noted. The study suggests several areas for continued investigation. 1 . The results of this and similar surveys of rail rapid transit noise should be evaluted and coordinated with a study on the effects of community noise of an integrated transporta¬ tion network which includes automobiles, buses, trains, etc. for each of the cities operat¬ ing rail transit systems. (The integration of urban rail transit noisevQn a national basis is currently underway at the Transportation Systems Center.) - S 6-1 2. A study similar to the subject program should be implemented for selected light rail systems throughout the United States. Noise surveys at cities such as Newark, Cleveland, Pittsburgh and Philadelphia, that already operate light rail systems, could be used for projecting the acoustical impact in cities where light rail is contemplated. Noise surveys in Boston and San Francisco before and after the introduction of new light rail vehicles would also be useful in evaluating the impact of noise in the city ultimately selected by DOT for a light rail demonstration program. 3. Since transit system noise is frequently only one of many noise sources effecting a community, a study of Day-Night Noise Levels should be performed for various com¬ munity locations on each transit system, simulating (1) no train operation, and (2) operation with cars which display reduced exterior noise levels. Some initial work of this nature has already been performed for the BART system. 4. Output from the long and short versions of the TSC cost/abatement algorithm are not in agreement. The program should be reworked to permit the short version to be utilized for further studies since the long output is often cumbersome. The output format is very cumbersom to handle physically, and conceptual simplifications should be made in this area. 6-2 REFERENCES 1. Spencer, R.H.; and Hinterkeuser, E.G.: "Noise Assessment Report — Assessment of Urban Rail Noise Climates and Abatement Options", Interim Report No. 1 (Draft), Boeing Report No.'s D339-10015 — 1, 2, 3 and 4, July 1975. r *. 2. Kurzweil, L.G.; Lotz, R.; and Apgar, E.G.: "Noise Assessment and Abatement in Rapid Transit f- Systems", D.O.T. Report No. UMTA-MA-06-0025-74-8, September 1974. 3. Baritz, S.: "Updated Version of Computer Program for Minimizing Noise Control Cost on Rail Systems", Report No. KHL-TSC-75-1244-6791, March 1975. 4. "Urban Rapid Rail Vehicle and Systems Program", Annual Report, July 1974, Boeing Vertol Company. 5. G. P. Wilson and S. L. Wolfe, "BART Car Floor Vibration Measurements", Final Report, March 4, 1974, Wilson, Ihrig & Associates, Inc. 6. G. P. Wilson, "BARTD Prototype Car 107 Noise Tests with Standard, Damped and Resilient Wheels", June 1972, Wilson, Ihrig & Associates, Inc. 7. G. P. Wilson, "Noise Levels from Operations of CTA Rail Transit Trains", August 1975, Wilson, Ihrig & Associates, Inc. 8. R. J. Remington, M. J. Rudd, and I. L. Ver, "Wheel/Rail Noise and Vibration Control", November ♦ ^ 1974, Bolt, Beranek and Newman, Inc. b. * 9. F. Kirschner, "New Developments in the Control of Railroad Wheel Screech Noise", paper given at Inter-Noise 72, October 1972. 10. A. Senko, "PONYA-Terminal Noise", Report submitted to the Port of New York Authority, March 8, 1971, Goodfriend-Ostergaard Associates. 7-1 11. G. P. Wilson and T. Kihlman, “Tests of Sound Absorption Treatment in WMATA Metro Subway v Structures", May 1975, Wilson, Ihrig & Associates, Inc. 12. “San Francisco Bay Area Rapid Transit District Demonstration Project Technical Report No. 8, Acoustic Studies", Parsons Brinckerhoff-Tudor-Bechtel, June 1968. 13. G. P. Wilson, “Aerial Structure Sound Barrier Walls", January 1973, Wilson, Ihrig & Associates, Inc. UNIVERSITY OF ILLINOIS-URBAN A 3 0112 067171485 » { I i i ■m, t \ v I ♦ ! I » 1 i 7-2