Center Truck Performance on Low-Floor Light Rail Vehicles (2006)

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

52 The research was organized into 11 tasks as follows: Phase 1— • Task 1. Review of domestic and international literature. • Task 2. Collection of information from suppliers and tran- sit systems. • Task 3. Identification of the factors contributing to per- formance issues. • Task 4. A working paper covering Tasks 1 through 3. • Task 5. Identification and planning of the research needed in Phase 2. • Task 6. Production of the interim report covering Tasks 1 through 5. Phase 2— • Task 7. Execution of the approved research plan. • Task 8. Production of a working paper summarizing the results of Task 7. • Task 9. Development of the guidance required to mitigate performance issues. • Task 10. Development of recommendations for further research. • Task 11. Development of the final report. The notes that follow describe the work carried out on each task. Conclusions and results are summarized; text in the body of the report is not duplicated. Task 1. Literature Review The review focused on locating information as early as pos- sible that illuminated the following issues: • The principles applied in designing vehicles. • The extent to which the designs of center trucks and vehi- cle articulation have taken into account the environment within which they will be used, especially conditions that apply on U.S. and Canadian light rail systems. • The extent to which issues have arisen internationally, how they have been dealt with, and how successful any correc- tive measures have been. • Comparing conditions that may have caused issues inter- nationally with those that exist in the United States and Canada. • Finding data on wheel and rail wear where center trucks of this type have been used. • Finding out how vehicle manufacturers have responded to these issues and seeing if they have modified their designs. • Establishing if standards in the United States differ from other countries using these vehicles, and whether such dif- ferences might contribute to any known issues or might limit the potential for corrective action. • Obtaining details of the history of performance issues in order to identify any patterns. • Reviewing the documentation of vehicle and/or track modifications implemented by transit systems in order to address the issues. It was thought that the best material would be that pro- duced by the transit authorities and vehicle manufacturers themselves. Documents that report on these organizations’ activities will tend to be prone to error, and previous research and periodical articles based on second-hand reports will, in turn, be even more likely to be inaccurate. A “hierarchy” of documents in terms of their reliability was created in order to avoid this issue. The literature review was carried out by a team of engineers and technical specialists in this area working internationally and sharing information continually via email. Interfleet assessed literature using a form developed for this purpose. The form was used to record the essential information, providing a summary in English that could easily be referred to at any stage in the project. The form was A P P E N D I X A TCRP Research Project C-16 Scope of Work

53 also designed so that the data could be recorded in data base format. The overall view was that the amount of literature that is directly relevant and of high value to this research was fairly limited and that the main pieces of literature had probably been identified and reviewed. During Phase 2 further literature was identified—some of which was seen as being especially relevant, including the reports issued by TCRP Projects D-2 and D-7,which were now available. Only the MBTA and TriMet Portland appeared to have written up their experience in the United States and Canada and so expe- rience elsewhere could not be determined by a literature search when the research was undertaken.Some European systems had reported on their experience in operating vehicles of this type. Berlin and Duisburg had reported on experience with 100- percent low-floor vehicles and Essen had reported on experience of a 70-percent low-floor vehicle with four trucks. These cities had experienced issues different from those seen in the United States and the vehicles were not of the generic type under inves- tigation.Despite this, their findings about the principles involved helped to inform Task 3. Some of the lessons learned were expected to be applicable to the U.S. and Canadian vehicle types being studied. Some publications covering appropriate areas of theory were identified but, in an area where the key develop- ments are both novel and recent and where commercial secrecy remains, these were not significant except as a means of identi- fying the key factors. The main exception to this was the work carried on for TCRP Research Project D-7 and reported in TCRP Report 71, Volume 5, which considered performance issues with IRWs.All the literature reviewed is listed in Appendix F, together with a short summary of the content and its value. Although the literature review was useful and was performed as planned, the amount of published information directly rele- vant to this research project was very small. Although a more exhaustive search might identify further material, the costs of doing this would probably not justify the small benefit. The value, therefore, was in confirming that this research was the first to examine these issues in this particular way and, therefore, could provide useful guidance for the transit industry. Task 2. Transit Agency and Manufacturer’s Experience Priority was given to understanding the issues that have arisen in the United States and the operating environment within which LFLRVs operate in this country. To this end, a targeted questionnaire was produced for the transit agencies operating LFLRVs. This questionnaire was scheduled after significant progress had been made with Task 3, so that per- tinent questions could be asked. Some design and performance data were obtained from other sources, including internal libraries and files and telephone and email contacts. Although this and the preced- ing tasks targeted partial low-floor vehicles of a specific design, where appropriate, the experience of operators of 100-percent low-floor cars was also taken into account. This was done to estimate whether there were lessons to be learned from these vehicles that were applicable to center trucks on partial low-floor cars. As in the case of the literature review, the review of design and performance data was prioritized to first consider infor- mation with the highest potential utility. This was to maxi- mize the effectiveness of the task within resource constraints and in light of potential availability issues resulting from the sensitive nature of some of the relevant information. All seven transit agencies that operate this type of car were asked in the questionnaire for some specific information about these vehicles, track standards, experience associated with the use of the center trucks, and any mitigation the agencies might have introduced for overcoming issues. Six of the seven agen- cies provided detailed responses, along with supporting infor- mation such as drawings and maintenance procedures. This information proved to be of great value to this project. The responses covered all the systems that had the most experience of operating this vehicle type and that had experienced per- formance issues with this vehicle type before 2005. Further information was collected during Phase 2 by direct inquiry and as part of Tasks 7.2, 7.3, 7.7., and 7.12. Task 3. Contributing Factors The project scope identified the following potential per- formance issues: • Derailments, • Excessive wheel and rail wear, and • Reduced ride quality. The issues being observed were thought to be concurrent manifestations of multiple failure phenomena and, therefore, even more difficult to isolate and identify. Effort in Task 3 was concentrated on identifying those factors that appeared to adversely affect LFLRV dynamic performance. This identifi- cation was done by circulating ideas internally in a table; these ideas were then discussed and refined by experts based on their collective experience. This analysis was used both as the basis of the questionnaire designed to obtain key feedback from U.S. and Canadian transit agencies (Task 2) and, com- bined with these results, to formulate preliminary guidance. The initial findings proved sufficiently robust so as to at least indicate situations likely to cause issues that might easily be avoided, especially with new systems. The contributing factors were assessed iteratively in dis- cussions between the groups of technical experts. The results

54 provided the basis for investigations in Phase 2 that were then combined in order to produce Chapter 3 of this report. The initial assessment indicated that certain features of LFLRV design are likely to have a greater effect on the per- formance issue areas being investigated than would be the case with a traditional high-floor LRV. Based on the question- naire returns, this appeared to have led to significant diver- gence in the performance of different U.S. LFLRV designs. In Phase 2, the researchers proposed investigating these diver- gences in order to establish the optimum combination of parameters to minimize issues. At this stage, the following features appeared critical: • Wheel profile, in relation to rail profile; • Flange angle; • Flange height; • Truck and truck-to-body attachment detail design and tol- erances; and • LRV articulation design, including stabilizing links and dampers. Because lubrication was seen as beneficial, it was proposed to study this as well in Phase 2. Some of the more serious issues were clearly associated with track geometry. Four issues were apparent from Phase 1 work: • Wheel and rail profiles must be compatible. • Sharp curves (under 25 m/80 ft radius) should be avoided. Where they exist, special measures such as low speed restric- tions, track lubrication,and restraining rail may be necessary. • Switches and crossings need to provide as smooth as pos- sible transit for LFLRV wheels and wheels need to be ade- quately checked by restraining rail. • The maintenance standards need to keep the track param- eters within the tighter tolerances that may be necessary for LFLRVs. Task 4. Working Paper A working paper was prepared to bring together the prin- cipal data collected during the performance of Tasks 1 through 3 that was likely to influence subsequent work. It was used to develop the Phase 1 report, including discussions on the research needed in Phase 2 (Task 5 of Phase 1). Task 5. Identification and Planning of the Research Needed in Phase 2 The conclusions of Phase 1 were that Phase 2 should con- centrate on modeling the features identified as critical for LFLRV performance in combinations appropriate to U.S. applications. This was seen as giving most value from the resources available. As part of this modeling activity it was seen to be necessary to obtain further data and background information so that the results were truly representative. The modeling activity involved creating vehicle and track models that would cover the main combinations of design features. This assessment was checked against what could be seen happening in practice by visits to representative U.S. transit systems using LFLRVs. Task 6. The Interim Report Covering Tasks 1 through 5 The report was prepared and some comments were received from the Panel. Interfleet presented the report for discussion at the TCRP Project Panel C-16 meeting. The panel requested further information on the proposed Phase 2 Research Plan. A supplement was produced in order to pro- vide this information and replacing section 5 of the Interim Report and the associated appendices 8.7 and 8.8. The sup- plement included more detail about why ADAMS/Rail was selected as the modeling tool. Task 7. Execution of the Approved Research Plan Phase 2 concentrated on modeling the features identified as critical for LFLRV performance in combinations appro- priate to U.S. applications. Vehicle and track models were created to cover the main combinations of design features found on the generic LFLRV type under consideration. This activity formed the core of Task 7, which was divided into 14 subtasks. ADAMS/Rail was used to undertake detailed analysis of track and vehicle conditions. Rail vehicle models were created by entering the required assembly data into forms. ADAMS/Rail then used the data to automatically construct the subsystem models and full-system assemblies building a complete, parametric model of each light rail vehicle being studied. Tracks were also modeled in ADAMS/Rail, defining the track centerline by specifying the analytic layout parameters: curvature, cant, and gauge. Track-measured data was speci- fied as irregularity parameters: alignment, cross level, and gauge variation. The virtual vehicles were run through a series of kinematic, static, and dynamic tests to determine the vehi- cle’s stability and derailment safety. Information was requested and obtained from two other major U.S. and Canadian transit systems to determine why

55 they had not introduced this type of vehicle onto their systems. Task 7.1 Identification of the Parameters Required for the Vehicle Models Models were created of the following vehicle types: • A high-floor LRV closely based on the MBTA Type 7 car. • An LFLRV closely based on the MBTA Type 8 car, and • An LFLRV closely based on the Kinki Sharyo New Jersey, Newark Subway car. These represented a traditional high-floor articulated vehicle, a low-floor vehicle that has experienced some issues, and a low- floor vehicle that is reportedly successful. The design of all three vehicles conforms to the generic vehicle type outlined in the Research Project Statement but the detail designs of the three cars are known to differ substantially from one another. These basic models were subjected to a range of detail variations dur- ing the modeling work in order to increase the breadth of the results. These variations are discussed in Task 7.3. A list of all the information required to cover all the options to be modeled was prepared; this took into account the capa- bility and structure of the model being used. A technical memo was prepared with 3 parameter list documents and cir- culated among the team. These parameter lists were eventu- ally completed following the visits to transit systems (Task 7.12); in the meantime estimated values were used. Task 7.2 Collect Data on U.S. Vehicles The lists created in Task 7.1 identified missing parameters that were required by using a color code. Red indicated infor- mation definitely required; yellow indicated where it was pos- sible to make assumptions or use available unconfirmed data without serious effect on the results. Sensitivity analyses were performed to quantify the possible errors of the output. Before the visit to transit systems (see Task 7.12) a memo was produced that identified the most important questions to be answered. During the visit some new insights were gained and more accurate parameter values were obtained, mainly for the MBTA Type 7 cars. Task 7.3 Collect Data on Alternative Vehicle Designs The key parameters of the models were varied in order to gage the effects of introducing different design features. The purpose of this task was to gather information on alternative components or arrangements that exist in order to be able to create representative models of them. These design modifica- tions included • Articulation roll stiffness, • Pitch damping of articulation, • Different solutions for stabilizing the pitching mode of the body section above the center truck, • Variation in the lateral secondary suspension stiffness, and • Different wheel and rail profiles. These variations, which were easily implemented in the parameterized model, allowed study of the effect of design changes on the behavior of the vehicle. Although these are comparative rather than absolute analyses, they gave the basis to fulfill the project objectives. Task 7.4 Develop Vehicle Models Vehicle models were created for the vehicle types listed above. This work included coordination of data collection in Tasks 7.2 and 7.3 so as to achieve compatible treatment of each one. Figures A-1 through A-3 are representations of the three models. The MBTA Type 7 vehicle produced by Kinki Sharyo is a standard vehicle (not low-floor) with normal trucks and wheelsets. The main vehicle parameters are: Overall length 21.95 m Overall mass 35,500 kg empty, 48,670 kg full Axle load 9.3 t motor truck, 7.7 t trailer truck (max. load, from model) Pivot distance 7 m Wheel base 1.905 m Number of sections 2 Motor trucks (2) 2 standard wheelsets Trailer truck (1) 2 standard wheelsets, arranged between the car bodies, selfsteering, independent of car bodies Inter-car connection Common crown bearings, no dampers CENTER TRUCK Figure A-1. MBTA Type 7.

56 loop of analyses described under Sub-tasks 7.10 and 7.11 (ini- tial assessment). The others were performed following the visits to transit systems, which permitted the models to be refined. Task 7.6 Create the Equivalent Model Modifications The model and parameter options described in Sub-task 7.5 were implemented in the ADAMS/Rail template based vehicle models. Modifications concerning the structure of the vehicle (adding or removing elements) resulted in a new ver- sion of the model whereas parameter variations were covered by different sets of parameter files. For every option, an ADAMS solver input file was generated and the collection of input files was run as a batch queue.The MBTA Type 8 produced by Breda has a low-floor area in the central part including the center section (C-car) which is therefore equipped with independently rotating wheels. The pitching stability of the center section is achieved by anti-pitch spring elements. This vehicle type has experienced derailment problems and high wheel wear on the center section (C-car). The main vehicle parameters are: Overall length 21.95 m Overall mass 38,460 kg empty, 47,360 kg full Axle load 9.3 t motor truck, 6.8 t trailer truck (max. load, from model) Pivot distance 7.14 m Wheel base 1.905 m Number of sections 2 end sections, 1 short central section rigidly connected (in yaw) to the central truck Motor trucks (2) 2 standard wheelsets Trailer truck (1) 4 single wheels on 2 rigid cranked axles, pitching stabilized by a spring system Inter-car connection 2 spherical bearings at either end of the central truck, no dampers, roll inhibited by lateral bars at roof level The Kinki NJT LFLRV has a low-floor area in the central part of the vehicle including the center section (C-car) which is therefore equipped with independent rotating wheels. The pitching stability of the center section is achieved by a system of bars (Z-link) arranged on the roof of the vehicle. This vehicle is reported to perform satisfactorily. The main vehicle parameters are: Overall length 26.7 m. Overall mass 45,790 kg empty, 59,440 kg full. Axle load 10.2 t motor truck, 9.4 t trailer truck (max. load, from model). Pivot distance 10.15 m. Wheel base 1.9 m on motor truck, 1.8 m on trailer truck. Number of sections 2 end sections, 1 short central section rigidly connected (in yaw) to the central truck. Motor trucks (2) 2 standard wheelsets. Trailer truck (1) 4 single wheels on 2 rigid cranked axles. Inter-car connection 2 crown bearings with pitch hinge on either side of the central truck, a pair of dampers between central and end section, pitch stabilization of the central section by a system of bars at roof level (Z-link) connecting all three sections. Task 7.5 Creation of the Vehicle/Track Options Matrix The purpose of this sub-task was to define combinations of vehicle features and track layouts that gave a wide range of possibilities in order to identify those parameters that are crucial in terms of derailment prevention. Table A-1 shows which basic runs were undertaken using the existing vehicle designs: Both empty and fully laden vehicles were modeled; this is because although from the point of view of safety against derailment the empty vehicle is most critical, it is possible that other factors (like wheel/rail forces) might get critical at full load. Parameter variations were defined based on the assessment of possible alternative design options used in other vehicles (see Sub-task 7.3) and in order to cover uncertainties in the model inputs. Table A-2 provides an overview of all the analy- sis cases considered. Cases S1 to S3 were performed in the first Figure A-2. MBTA Type 8. Figure A-3. NJT.

57 Task 7.7 Collect specific U.S. Track Data Tables were produced to compare the track standards used on seven of the eight transit systems LFLRVs identified in Phase 1. Data for another modern light rail system was also included, because comprehensive information was in hand. The tables also showed equivalent U.S. national standards and the equivalent German standards. Notes were made of where practice varies from the U.S. standards, e.g. where tighter tol- erances are being applied. A table was produced for each of the following issues: • Wheel and rail profiles, • Sharp curves, • Switch and crossing transitions, • Track standard tolerances, • Gauge tolerance on tangent track, • Tangent track between curves, • Parallelism of rails, • Flangeway clearance on special trackwork, and • Track twist. The purpose was not to create comprehensive tables of data but more to identify where any notable discrepancies likely to have an effect on LFLRV operation might exist and allow rec- ommendations to be made where necessary. As part of this work a detailed comparison of typical U.S. practices with the equivalent German standard was also carried out. This showed that the German Standards and guidance (BOStrab/VDV) were not prescriptive in general terms but deal with a process that trained engineers need to apply to their systems, based on their own knowledge of them and of standards that have been adopted in the past. One rea- son for this is that most systems in Central Europe have existed for a long while and parameters vary as a conse- quence. It was, therefore, decided that it would be useful to give the MBTA track standards to engineers in the research team who had experience of accepting LFLRVs on a repre- sentative German system to determine if they would make changes based on BOStrab/VDV practice. This exercise high- lighted some specific differences that may be having an effect: • Smaller ballast not taking up forces as well, • Gauge variation, • Check rail clearances, and • Permitted number of defective ties. The detailed comparison is included in this report as Appendix E. Task 7.8 Track Model Features Decisions were made on what track conditions would be representative to model and what specific features should be included. Table A-3 shows the parameters of the first three track options. No. Vehicle/Condition Track 1 KS New Jersey LFLRV - empty TCRP D-7 reference New Jersey new MBTA "worst case" High speed track High speed track (worn) 2 KS New Jersey LFLRV - full (fully laden) TCRP D-7 reference New Jersey new MBTA "worst case" High speed track High speed track (worn) 3 MBTA No. 7 - empty TCRP D-7 reference New Jersey new MBTA "worst case" High speed track High speed track (worn) 4 MBTA No. 7 - full (fully laden) TCRP D-7 reference New Jersey new MBTA "worst case" High speed track High speed track (worn) 5 MBTA No. 8 - empty TCRP D-7 reference New Jersey new MBTA "worst case" High speed track High speed track (worn) 6 MBTA No. 8 - full (fully laden) TCRP D-7 reference New Jersey new MBTA "worst case" High speed track High speed track (worn) Notes: "TCRP D-7 reference" means the track parameters identified in TCRP Report 71, Volume 5 (Wu, Shu, Wilson), in the work on the investigation of wheel flange climb criteria for transit vehicles. "New Jersey new" refers to a track of the Hudson-Bergen line as defined in the track design manual. "MBTA "worst case"" refers to track parameters assuming extreme MBTA track geometry and track at the limits accepted in current MBTA maintenance standards. MBTA track was chosen because this is a long-established system in contrast with a typical new build. "High speed track" means a synthetic track definition defined specifically to study derailment at high speed on (nearly) straight track sections. "High speed track worn" is a variation with significant rail wear following the information obtained during the visit to transit systems. Table A-1. The basic modeling runs.

58 parameters on this behavior, a pair of track layouts was defined as follows: • Curve radius 2000 m allowing high speed but imposing a flange contact on one side • Lateral track irregularity of amplitude 3/4 inch over a length of 20 feet in order to see separately the effect for every truck • 115 RE rail profile (new rail) and a worn shape as described by the transit system engineers during the US visit. The wear shape and original rail profile are compared in Figure A-4. Data extracted from the TCRP Report 71,Volume 5, reference (Exhibits 41-6) as model input. The curvature was transformed so as to consist of circular curve sections joined by spiral sec- tions. The original curvature data was assumed to correspond to the formula: degree = 5729.578/R for R in feet. The superel- evation was transformed so as to consist of constant sections and linearly varying sections. Lateral and vertical track excita- tions for both rails were set to zero at X = 30 m (98 ft). Some of the reported derailment issues involving MBTA type 8 cars occurred on straight track sections at high speed at a time when wheel profiles with a flange angle of 63° were in use. In order to assess the influence of other system No. Vehicle / Condition Track(s) S1 KS New Jersey LFLRV Reduced roll stiffness (1/10) of the inter-car articulations MBTA “worst case” New Jersey new TCRP D-7 reference S2 KS New Jersey LFLRV Z-link (inter-car stabilizing mechanism) removed MBTA “worst case” S3 MBTA No. 8 Increased roll stiffness (x10) of the inter-car articulations MBTA “worst case” S11 KS New Jersey LFLRV - empty Wheel profile with 63° flange angle High speed track High speed track worn MBTA “worst case” S12 KS New Jersey LFLRV – empty Inter-car dampers removed TCRP D-7 reference MBTA “worst case” S13 KS New Jersey LFLRV - empty Inter-car dampers between all cars (2 sets of 2) TCRP D-7 reference MBTA “worst case” S14 KS New Jersey LFLRV – empty Z-link replaced by a pitch stabilizer spring TCRP D-7 reference MBTA “worst case” S15 KS New Jersey LFLRV - empty Wheel profile with 63° flange angle plus inter-car dampers between all cars (2 sets of 2) High speed track S16 KS New Jersey LFLRV – empty Wheel profile with 63° flange angle plus Z-link replaced by a pitch stabilizer spring High speed track S21 MBTA No. 8 - empty Wheel profile with 63° flange angle High speed track High speed track worn MBTA “worst case” S22 MBTA No. 8 - empty Inter-car damper analogue to KS New Jersey LFLRV TCRP D-7 reference MBTA “worst case” S23 MBTA No. 8 - empty Wheel profile with 63° flange angle plus inter-car damper analogue to KS New Jersey LFLRV High speed track S24 MBTA No. 8 - empty TCRP D-7 reference Two sets of inter-car dampers MBTA “worst case” S25 MBTA No. 8 - empty Wheel profile with 63° flange angle plus two sets of inter-car dampers High speed track S31 MBTA No. 7 - empty Wheel profile with 63° flange angle High speed track High speed track worn MBTA “worst case” Table A-2. Analysis cases.

59 track model. The continuously varying curvature was trans- formed into a sequence of sections with constant curvature and transition spirals (named “Standardized” in Figure A-5) and slightly modified to include a straight section around t=120 (“Standardized modified”). Task 7.10 Simulation Runs The basic simulation options given in Table A-2 and sen- sitivity cases S1 through S3 identified in Task 7.5 were run using ADAMS/Rail. A selection of typical results is shown in Tables A-4 through A-9. These results were obtained with preliminary model and track parameters. Wheels are num- bered in the form wij, where i is the axle index and j=1 is for right and j=2 for left wheels respectively (Figure A-6). Forces are shown in Newtons (N) (1 lbf = 4.45 N). Lateral forces are negative if the wheel is pushed inward (this is usually the case for an outer wheel in a curve). The curves modeled are right curves; therefore the leading outer wheel of the center truck is w32. Figure A-4. Comparison of wear shape and original rail profile. Figure A-5. Continuously varying curvature in sections. Vehicle w32 w41 NJT_Full -23 kN 2.5 kN Type 7_Full -27 kN -2.5kN Type 8_Full -17 kN 1.5kN Vehicle w32 w41 NJT_Full -23kN 3kN Type 7_Full -10kN 3kN Type 8_Full -18kN 2kN Table A-4. Wheel-rail lateral force (N), NJT new track. Table A-5. Wheel-rail lateral force (N), TCRP Report 71, Volume 5 track. Model NJT New Track TCRP Report 71, Volume 5 Reference MBTA worst case Track type Ballasted Ballasted Ballasted , Note A Rail profile 115RE 115RE 115RE, Note A Rail inclination 1:40 1:40 1:40 Track gauge 56.5in 56.5in 56.5in straight, 57.5in curve Where measured 5/8 in below rail head Curve radius 300ft See below 44ft Vertical curve (combined) See below 400ft Spiral length 130ft (2 of) See below 20ft (2 of) Circular length 180ft See below 46ft Superelevation 4in See below Note B Track irregularity See below Note C Design speed 23mph 38mph 4.5mph Note A. Although some of the worst conditions occur on embedded track the application of the same rail sections in all these cases allows the separation of rail profile issues from other track quality issues. Note B. Negative elevation 1.25 in (32 mm), starting at 36 m, fully established 40m to 50m, removed with end of curve spiral (based on MBTA limits for track maintenance). Note C. Track irregularity 1.25 in (32 mm) lateral on outer rail in inward direction, on a length of 31.5 feet (based on MBTA limits for track maintenance). Table A-3. Track option parameters. Task 7.9 Develop Track Models Three track models were created using the material from Sub-tasks 7.7 and 7.8 covering the basic options defined under Sub-task 7.5. Details of the track features are given in 7.8. After the visit to transit systems, a fourth track was added. This track contained a curve with large radius and a single lat- eral inward track irregularity on the outer rail. It allowed comparison of the vehicle models with respect to the derail- ment at high speed on straight tracks observed for the MBTA Type 8 central truck. The track models are defined in ADAMS/Rail as a sequence of straight and curved sections joined by spiral transitions. The software versions used in this project require a straight section between two curved sections of different radius. This is an acceptable restriction for most practical purposes but it required some adaptation for the TCRP Report D-7 reference

60 Task 7.11 Initial Assessment The analysis results obtained with the preliminary vehicle models showed a similar behavior of the vehicles. Although some observations may have depended on preliminary data, the following general tendencies were observed: • The MBTA worst-case track had the smallest curve radius and may therefore have produced the largest steady lateral forces caused by a large angle of attack. The maximum forces were obtained for the NJT LFLRV. • The highest peak lateral forces occur in the perturbed zone of the TCRP Report 71, Volume 5, track and again the max- imum was found for the NJT LFLRV. • The maximum L/V ratios are found in the perturbed zone of the TCRP Report 71, Volume 5, track and the vehicles with single wheels (MBTA Type 8 and NJT) were most affected. The MBTA Type 8 vehicle, which has reportedly encountered various difficulties in service (e.g., derailment and excessive wear), behaved quite satisfactorily in the simulation. The analysis results obtained for the sensitivity cases S1 to S3 showed only minor effects: • The modification of the roll stiffness of the articulations between the end sections and the central section with sin- gle wheels has no significant effect on the wheel unloading. This meant that the vehicle could be considered to be rigid in torsion for all the cases considered. • The removal of the Z-link did not affect wheel/rail forces. However it did lead to a progressive pitch deformation, which was however developing slowly caused by the high damping factor of the inter-car dampers. A means of pitch stabilization is required (Z-link or an alternative solution). Task 7.12 Visits to Transit Systems Visits were made to discuss four transit systems with lead- ing engineers involved in their operation and maintenance. Questions and agendas were prepared in advance. The team was able to have other discussions to discuss results, exchange and obtain information, and see and expe- rience the systems, vehicles, and maintenance facilities. Task 7.13 Simulation Runs The information gathered during the visit to the tran- sit systems allowed finalization of the vehicle models and completion of the model options as described in Sub- task 7.5. An overview of the simulation output is providedFigure A-6. Simulation run numbering. Vehicle w12 w21 w32 w41 NJT_Full -42kN -19.5kN -32kN -27kN * Type 7_Full -37kN -19kN -34kN -20kN Type 8_Full -37kN -20kN -15kN -14kN *Unsteady peak value Vehicle w32 w41 NJT_Empty 0.43 0.05 Type 7_Empty 0.50 0.07 Type 8_Empty 0.45 0.02 Vehicle w32 w41 NJT_Empty 0.93 0.22 Type7_Empty 0.51 0.10 Type 8_Empty 0.80 0.30 Table A-6. Wheel-rail typical lateral force (N), MBTA worst-case track. Table A-7. Maximum L/V, NJT new track. Table A-8. Maximum L/V, TCRP Report 71, Volume 5, track. Vehicle w32 w41 NJT_Empty 0.60 0.72 Type 7_Empty 0.62 0.50 Type 8_Empty 0.47 0.42 Table A-9. Maximum L/V, MBTA worst- case track.

61 in Tables A-10 to A-22 and Figures A-7 through A-9. The main outcome can be summarized as follows: • The overall behavior of the three vehicle types was found to be very similar. All the vehicles were able to negotiate small radius curves without excessive L/V ratios. • The vehicles with single wheel trucks (NJT LFLRV and MBTA Type 8) produce higher lateral forces in narrow curves as expected. This was expected to increase noise and wheel-rail wear. • The pitch stabilization of the cars with a center truck was solved in two different ways: The NJT vehicle had a link mechanism (Z-link) connecting the three car bodies at the roof level whereas the MBTA Type 8 had a torsion spring system located in the truck. The Z-link introduced some asymmetry, which resulted in the vehicle running eccentri- cally close to one rail even on straight tracks. Similar effects would be produced by manufacturing tolerances with any vehicle design. The MBTA Type 8 solution is prone to dynamic pitching modes caused by the damper design. • Inter-car dampers tended to reduce the pitching mode of the center section. The MBTA Type 8 cars did not have dampers but by adding them into the model they were found to be especially beneficial. The disadvantage was a limited increase in the lateral forces and therefore the L/V ratio. The best behavior was obtained with two pairs of dampers (both articulations damped). However the improvement was not significant. The modeling suggested that only one pair of dampers was required for the NJT LFLRV in order to limit the lateral wheel-rail forces. • The use of wheel profiles with low flange angles (63°) increased the tendency for derailment on all vehicles. There was a marked difference between the solutions with inde- pendent wheels and the standard wheelset truck of MBTA Type 7. The effect was accentuated when combined with worn rail profiles. The MBTA Type 8 vehicle, which was reported to have derail- ment issues, had a very similar behavior to the NJT LFLRV. In general, the lateral forces and L/V ratio were found to be even less critical. NJT/full -19.8kN -3.68KN Vehicle/Condition w32 w41 Type 7/empty -20kN -2.6kN Type 8/empty -12.8kN -6.1kN NJT/empty -17.6kN 2.7kN Type 7/full -30kN -6.45kN Type 8/full -14.3 kN -6.95kN Vehicle/Condition w32 w41 Type7/empty -4.5kN +8.2kN Type8/empty -30.8kN -12.2kN NJT/empty -20.8kN -13.3kN Type7/full -6.4kN -10.8kN Type 8/full -28kN -13.5kN NJT/full -32.8kN -20.1kN NJT/empty S12 -47.2kN -37kN NJT/empty S13 -28.6kN -10kN NJT/empty S14 -50.6kN -12.5kN Type 8/empty S22 41.2kN 11.3kN Type 8/empty S24 41.2kN 12.2kN Vehicle/Condition w32 w41 Type 7/empty -27.3kN -12.3kN Type 8/empty -10.8kN -10.5kN NJT/empty -21.8kN -31.2kN Type 7/full -44.7kN -18.35kN Type 8/full -17.6kN 12.9kN NJT/full -37.5kN -38.4kN NJT/empty S11 -28.5kN 30.1kN NJT/empty S12 -32.6kN 31.7kN NJT/empty S13 -34.8kN 32.3kN NJT/empty S14 -23.8kN 28.3kN Type 8/empty S22 22.5kN 24.5kN Type 8/empty S24 21.5kN 20.8kN Table A-10. Wheel-rail lateral force (N), NJT new track. Table A-11. Wheel-rail lateral force (N), TCRP Report 71, Volume 5, track. Table A-12. Wheel-rail lateral force (N), MBTA worst-case track. Vehicle/Condition w32 w41 Type 7/empty S31 -14.9kN -8.6kN Type 8/empty S21 -21.1kN -16.8kN NJT/empty S11 -27.1kN -10.6kN Type 8/empty S23 -20.6kN -6.8kN Type 8/empty S25 -19.6kN -6.8kN NJT/empty S15 -28.3kN -12.5kN NJT/empty S16 -25.3kN -12.5kN Table A-13. Wheel-rail lateral force (N), high-speed track. Vehicle/Condition w32 w41 Type 7/empty S31 -11.1kN -3.8kN Type 8/empty S21 -24.2kN -8.8kN NJT/empty S11 -27.1kN -10.3kN Table A-14. Wheel-rail lateral force (N), high-speed worn track.

62 Vehicle/Condition w32 w41 Type 7/empty -0.56 -0.1 Type 8/empty -0.44 -0.25 NJT/empty -0.49 -0.37 Type 7/full -0.56 -0.094 Type 8/full -0.44 -0.22 NJT/full -0.44 -0.13 Table A-15. L/V ratio, NJT new track. Vehicle/Condition w32 w41 Type 7/empty -0.2 -0.15 Type 8/empty -0.72 -0.42 NJT/empty -0.61 -0.33 Type 7/full -0.2 -0.24 Type 8/full -0.63 -0.39 NJT/full -0.56 -0.41 NJT/empty S12 -1.25 -0.95 NJT/empty S13 -0.7 -0.21 NJT/empty S14 -1.0 -0.3 Type 8/empty S22 -1.32 -0.44 Type 8/empty S24 -1.32 -0.38 Table A-16. L/V ratio, TCRP D-7 track. Vehicle/Condition w32 w41 Type 7/empty -0.84 -0.48 Type 8/empty -0.49 -0.44 NJT/empty -0.78 -0.87 Type 7/full -0.75 -0.50 Type 8/full -0.48 -0.44 NJT/full -0.68 -0.78 NJT/empty S11 -0.82 -0.83 NJT/empty S12 -1.0 -0.94 NJT/empty S13 -0.85 -0.94 NJT/empty S14 -0.72 -0.75 Type 8/empty S22 -0.78 -0.85 Type 8/empty S24 -0.75 -0.78 Vehicle/Condition w32 w41 Type 7/empty S31 -0.47 -0.13 Type 8/empty S21 -0.67 -0.28 NJT/empty S11 -0.74 -0.36 NJT/empty S15 -0.76 -0.42 NJT/empty S16 -0.74 -0.40 Type 8/empty S23 -0.67 -0.28 Type 8/empty S25 -0.63 -0.28 Table A-17. L/V ratio, MBTA worst-case track. Table A-18. L/V ratio, high-speed track. Vehicle/Condition w32 w41 Type 7/empty S31 none none Type 8/empty S21 none none NJT/empty S11 none none NJT/empty S15 2 (short duration) none NJT/empty S16 none none Type 8/empty S23 none none Type 8/empty S25 none none Vehicle/Condition w32 w41 Type 7/empty 1.4 (short duration) none Type 8/empty none 1.5 (short duration) NJT/empty none 1.5 (short duration) Table A-19. Wheel lift (mm), high-speed track. Table A-20. Wheel lift (mm), high-speed worn track. Vehicle/Condition w32 w41 Type 7/empty 1.1 0.07 Type 8/empty 1.0 0.15 NJT/empty 1.1 0.2 Type 7/full 1.0 0.07 Type 8/full 1.0 0.15 NJT/full 1.0 0.1 Table A-21. Angle of attack (degrees), NJT new track. Vehicle/Condition w32 w41 Type 7/empty 4.5 2.4 Type 8/empty 4.1 2.8 NJT/empty 4.3 2.8 Type 7/full 4.6 2.3 Type 8/full 4.1 2.8 NJT/full 4.3 2.6 NJT/empty S12 3.8 2.9 NJT/empty S13 4.3 2.5 NJT/empty S14 4.6 2.2 Type 8/empty S21 4.2 2.7 Type 8/empty S22 4.2 2.7 Type 8/empty S24 4.2 2.7 Table A-22. Angle of attack (degrees), MBTA worst-case track.

63 Figure A-7. Simulation Output 1. Figure A-8. Simulation Output 2. Figure A-9. Simulation Output 3.

64 Task 7.14 Final Assessment of Results The results of the modeling, other investigations carried out and conclusions of the visits to transit systems were assessed as a whole in order to produce the input required for Tasks 9 and 10. Chapter 3 of this report incorporates the results. The assessment process was based on the results of the modeling and other research and the experience and knowl- edge of individual team members. The discussion material in Chapter 3 is therefore an original collation of some of the lat- est thinking in this topic area by the experts involved. Drafts of the working paper (Task 8) were circulated and revised and each of the main contributory factors were discussed in depth at the workshop held in Cologne (Task 9). Task 8. Working Paper Summarizing the Results of Task 7 This was issued for distribution to the panel for comment. Task 9. Development of the Guidance Required to Mitigate Performance Issues A structured workshop took place in Cologne, Germany, at which the European experts in the team addressed both the issues associated with Task 9 and with Task 10. They concen- trated on two main issues. One was checking that the results from the modeling and theoretical studies were being cor- rectly interpreted, based on their own experience of theoret- ical work. The other was to feed in any European experience of solutions that were being recommended where there is limited experience of applying them in the United States and Canada. The guidance, which forms Chapter 3 of this report, was prepared in draft and discussed by the team as a whole. Having completed Tasks 7, 8, and 9, all the input required in order to make clear overall recommendations was now available. Task 10. Develop Recommendations for Further Research The research work was used as a basis for creating an ini- tial list that was then brainstormed at the Cologne workshop and further discussed by the team. Some issues were identi- fied. These tend to be of two types: • Issues that could not be fully explored within the limits of the current budget and • Areas of research in associated areas that are outside of the scope but might provide alternative solutions. Task 11. Develop the Final Report The preliminary draft final report was issued with three months allowed for the Panel to review it and for Interfleet to provide the final text.