Design of Track Transitions (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.

Research Results Digest 79 October 2006 INTRODUCTION This digest reviews and analyzes vari- ous track transition designs among ballasted and nonballasted track forms and structures and offers guidance to improve track and op- erating performance. The research is based on similar work conducted for freight rail- roads, modified, as necessary, for the transit operating environment. The results should be of interest to engineers involved in the design, construction, maintenance, and op- eration of rail transit systems. SUMMARY In rail transit systems, at-grade ballasted track frequently changes to a nonballasted track configuration or to ballasted track on a structure. The abrupt change in track sup- port that can occur at these locations is often associated with accelerated rates of track geometry and component degradation, high maintenance demand, and poor ride quality. Accordingly, a number of techniques have been proposed to improve track perfor- mance by providing a transition to smooth the stiffness interface between the dissimi- lar track types. A review of typical transi- tion designs, as found in the existing litera- ture, and analyses of representative designs are the subjects of this digest. A review of published material dealing with track transition problems and solu- tions was undertaken as the initial phase of the study. The literature indicated that tran- sitions were designed to (1) equalize the stiffness and rail deflection of the ballasted and nonballasted tracks, usually by con- trolling the resilience of the rail on the non- ballasted track, or (2) provide a gradual in- crease in the stiffness of the ballasted track to match that of the nonballasted track. Several designs seek to increase the stiffness of the ballasted track by placing a structural element, such as concrete slabs or an asphalt pavement layer, between the track granular layer (ballast/subballast layers) and the subgrade. These structural layers are generally tapered or stepped to allow a gradual increase, or ramping up, of the stiffness within about 20 ft of the non- ballasted track interface. Other designs seek to match the stiffness/deflection characteristics of the nonballasted track to the ballasted ap- proach track using elastomeric pads at the rail seat or beneath the tie plates. This tech- nique requires measurement of the bal- lasted approach track to determine its nom- inal stiffness and track modulus values and testing of the rail/tie pad stiffness charac- teristics to ensure that the pad stiffness matches the approach track modulus at the DESIGN OF TRACK TRANSITIONS This digest summarizes the results of TCRP Project D-7/Task 15. The digest was prepared by the Transportation Technology Center, Inc. (TTCI) in Pueblo, Colorado. David Read and Dingqing Li served as principal authors. Subject Areas: VI Public Transit, VII Rail Responsible Senior Program Officer: Christopher W. Jenks TRANSIT COOPERATIVE RESEARCH PROGRAM Sponsored by the Federal Transit Administration C O N T E N T S Introduction, 1 Summary, 1 Background, 3 Definitions, 3 Track Transition Literature Review, 4 Analysis of Representative Track Transition Designs, 16 Discussion of Transition Designs, 29 References, 36

appropriate wheel loading. Elastomeric materials have also been placed on the bottoms of ties installed on ballast deck bridges to equalize the stiffness/ deflection of the bridge and approach tracks. The following performance improvements were noted in case studies from the literature review: • Use of longer ties and a concrete approach slab by the Metropolitan Atlanta Rapid Transit Au- thority (MARTA) to transition from ballasted at-grade track to direct-fixation structures. • Transition from at-grade ballasted track to a direct-fixation structure on a commuter/ intercity passenger service railway in the United Kingdom using an approach slab along with vertically adjustable direct-fixation fasteners to allow design tamping of the bal- lasted approach track. • Installation of stone columns to strengthen and improve the drainage of a weak bridge ap- proach subgrade on a Union Pacific main line. • Use of a transition grade crossing system de- signed to smooth the track modulus across the approach to a highway crossing and reduce im- pact rail loads at the crossing on New Jersey Transit’s Atlantic City line. • Installation of tie pads on open wood-tie bridge decks having stiffness/resiliency characteris- tics designed to match the track modulus of the approach track on Amtrak’s northeast corridor (NEC) and on a Norfolk Southern mainline with freight/intercity passenger service. • Reducing the track modulus on a Union Pa- cific ballast deck bridge by replacing the exist- ing concrete deck ties with composite (plastic) ties or with concrete ties with a rubber pad cast into the tie bottom. The importance of following geotechnical best practices regarding soil selection, compaction, and drainage of the approach subgrade was also dis- cussed in a number of papers, especially highway re- search papers. A properly designed and constructed subgrade will have a nominal stiffness adequate for the applied load environment, will tend to perform consistently through wet and dry cycles, and will not be prone to differential settlement. These attributes make it easier to match the vertical response of the at-grade track and the track on a structure. It should be made clear that much of the litera- ture reviewed was based on research performed on freight and intercity passenger tracks. There was not much literature generated from transit research. Although the higher wheel loads and speeds of freight/ intercity rail traffic create more intense track transi- tion problems than rail transit, the basic track perfor- mance issues are similar. Therefore, the experiences and results of research projects involving freight and intercity passenger tracks are considered applicable to the transit environment. Following the literature review, a number of representative track transition designs were ana- lyzed using the GEOTRACK computer model. GEOTRACK is a well-established and validated model that predicts a quasi-static response of the track to an applied vertical wheel load. The analysis produced track modulus and verti- cal rail deflection values for a variety of track con- figurations: wood and concrete ties on low-, average-, and high-stiffness subgrades; at-grade track with concrete approach slabs and hot mix asphalt (HMA) underlayment; direct-fixation track with typical fastener pad vertical stiffness values; open deck bridges with wood ties; and ballast deck bridges with concrete ties. Three wheel loads considered to be representa- tive of the rail transit environment were analyzed: 12,000, 15,000, and 22,500 lb. The 12,000-lb load was intended to represent light rail operations, the 15,000-lb load is the static weight of a Metro North cab car with full seated passenger load (Kentner et al. 1994, p. 270), and the 22,500-lb wheel load repre- sents the Metro North static wheel load plus a 50% dynamic factor. Results of the GEOTRACK analysis were as follows: • Matching the rail deflection on direct-fixation track to the deflection of the at-grade ballasted track, through careful design and specifica- tion of the direct-fixation fastener vertical stiffness, provides the best possibility for an effective and seamless transition between the two track configurations. However, bal- lasted track on low-stiffness subgrades also requires strengthening with either a concrete approach slab or HMA underlayment to match the direct-fixation track. Otherwise, the pad stiffness of the direct-fixation track would need to be unreasonably low. • A concrete approach slab placed between the ballast and subballast layers was the most ef- 2

fective technique for increasing ballasted track stiffness. HMA underlayment installed between the ballast and subgrade also produced benefits to low-strength track, but it was not as effective as concrete in increasing the stiffness of track on very low-stiffness subgrades. • Increasing the subgrade stiffness reduced the differences between concrete slab and HMA layer thicknesses. • Placing additional rails on the ties of the bal- lasted track to increase the stiffness of the track panel had modest benefits for low-stiffness subgrades. This condition often exists when bridge guard rails extend past the abutment onto the approach track. • Other changes to the track superstructure, such as reduced tie spacing, installation of longer ties, or installation of ties with larger cross sec- tions had an insignificant effect on track mod- ulus or rail deflections and, therefore, would not be especially effective transition designs. BACKGROUND The metropolitan environments in which rail transit systems operate require the placement track not only in at-grade ballasted configurations, but also on bridges and elevated structures and in tunnels and street pavements. Locations where the at-grade bal- lasted track changes to a structure are often associ- ated with accelerated rates of geometry and compo- nent degradation, high maintenance demand, and poor ride quality. In addition to deterioration of the track surface, alignment, and cross level, compo- nent problems can include exposed tie ends and re- duced crib ballast from ballast migration, tie skew- ing and bunching, cracked concrete ties, accelerated plate cutting of wood ties, gage widening and loss of rail cant, deterioration of ballast from pumping and frequent tamping, and accelerated rail surface fatigue. The track interface at bridge abutments, grade crossings, slab/embedded track, and turnouts/rail crossings are potential problem areas, and it is gen- erally recognized that effective transition designs may be required to optimize track performance at these locations. This digest presents the results and conclusions of an investigation into track transition designs. The investigation included a review of available litera- ture from the railway industry and an analysis of designs thought to be representative of, and applica- ble to, the rail transit environment. DEFINITIONS Definitions for terms used throughout this digest are listed below. Approach Slab—A reinforced concrete slab in- stalled as a structural element in the track substructure to increase the stiffness/modulus of the track. Most slabs are reinforced concrete and are designed either with a taper to gradually increase the stiffness over an approach distance of about 20 ft, or are uniform in thickness but placed at an angle with tapering of the ballast depth to achieve the same ramping effect. At Grade—Track that is constructed on a pre- pared soil subgrade foundation. Ballasted/Nonballasted Track—Ballasted track has a layer of aggregate between the ties and the sub- grade to distribute the applied wheel loads to the un- derlying layers; provide vertical, lateral, and longi- tudinal resistance to track panel movement; to drain moisture away from the ties; and to facilitate sur- facing and lining of the track. Ballasted track is usu- ally at grade, but it may be located on a structure (as in the case of ballast deck bridges). Nonballasted track designs vary, but, in the context of this digest, nonballasted track will be considered to be a direct- fixation track form. Damping—The capacity to attenuate, diminish, and/or control oscillations or deflections of an element of a system expressed as a unit of force that is dissi- pated per unit of distance and unit of time (lb/in./sec). Track damping is provided primarily by the resilience of rail seat and tie pads, by the resilience of the bal- last layer, and by the friction between ties and ballast. Track that is highly resilient has more damping than track that is less resilient. Deep Pile Foundation—Foundations of aerial structure that are driven to bedrock. Design Tamping—A track surfacing technique developed in the United Kingdom in which the track is over-lifted to compensate for the rapid rate of ini- tial settlement. Direct-Fixation Track—Nonballasted track in which the rail is mounted directly to a concrete base—such as the deck of an aerial structure, a tunnel invert, or an at-grade slab—with a direct-fixation fas- tening system. Elastomer—Polymer materials having the elas- tic properties of natural rubber. 3

Fastener Stiffness—The combined stiffness, expressed as the unit of applied force per unit of de- flection (lb/in.), of the fastening system and the tie at a specific applied load. For wood-tie track with steel-tie plates and no tie pads, the fastener stiffness is basically the compressibility of the wood. Fastener stiffness of concrete-tie track is primarily the stiff- ness of the rail seat pad and pads on the tie bottom, if used. The stiffness of concrete-tie pads can vary between 300 and 2,000 kip/in. Fastener stiffness on direct-fixation track consists of the resilience of the elastomeric elements of the fastening system. Typi- cal direct-fixation track fastener stiffness values are between 100 and 300 kip/in. GEOTRACK Model—A computer model that represents the track as a multilayered elastic struc- ture and predicts the quasi-static response of the track to an applied wheel load. Input parameters include rail, tie, and substructure layer definitions as well as wheel load. Output parameters include rail deflections, track modulus, tie/ballast/subgrade pres- sures, and tie bending moments. Reference is made to GEOTRACK several times in this digest’s literature review (see “Track Transition Literature Review”) and is the basis of the analysis described in the section titled “Analysis of Representative Track Transition Designs.” A more detailed description of GEOTRACK is also included in this section. Hot Mix Asphalt (HMA) Underlayment— A layer of asphalt pavement that is installed in bal- lasted track as a structural element in the substruc- ture to increase the bearing capacity of the subgrade. Typical HMA layer thickness varies between 8 and 12 in. and can be installed between the ballast and subballast layers or directly on the subgrade in lieu of a granular subballast layer. HMA is a mixture of aggregate and bitumen, and its stiffness proper- ties can be designed by varying the ratio of the con- stituents and the aggregate particle size distribu- tion. Recommended use of HMA in the rail transit environment is available online from the Asphalt Institute. Resilient Modulus (Er)—A geotechnical param- eter that is expressed as a unit of force per unit of area (ksi) and is used to define the elastic response of a soil to load. In the context of this digest, Er can be thought of as being equivalent to the modulus of elasticity. Typical values range from 2 ksi for a low- strength soil, such as a high-plasticity clay, to 20 ksi for granular soil that has been placed at optimum den- sity. Er is also used to describe the resilient behavior of aggregate materials such as ballast and pavements. Track Stiffness/Track Modulus—Track stiff- ness is the ratio of an applied vertical force to the vertical deflection of the rail and is expressed as a unit of force per unit of deflection (lb/in.). Track mod- ulus is the supporting unit of force per unit length of rail per unit rail deflection (lb/in./in.). Track stiffness includes the bending stiffness of the rail, whereas track modulus is concerned only with the support condition below the rail. A further discussion of these parameters is included in the section titled “Track Stiffness and Modulus.” TRACK TRANSITION LITERATURE REVIEW Track transition issues affect all types of rail op- eration, and a number of papers have been written defining the causes and/or proposing solutions. The results of a limited number of case studies have also been documented. The purpose of this section is to summarize the existing literature in terms of problem definitions, case studies, and recommended designs and proposed mitigation techniques. Please note that although a few papers are specific to rail transit, much of the existing literature is related to the freight and intercity passenger rail environments. Problem Definition According to Li and Davis (2005) and Li et al. (2003), track transition problems, specifically prob- lems at bridge approaches, can be attributed to the following factors: • An abrupt change in the vertical stiffness of the track causes the wheel to experience an equally abrupt change in elevation because of the uneven track deflection. The change in ele- vation causes vertical acceleration of the ve- hicle mass that generates an increase in the applied loading. This mechanism can be self- perpetuating as the dynamic loads increase the differential deflections and settlement lead- ing to even higher forces (Kerr and Moroney 1993; Frohling et al. 1995; Hunt and Winkler 1997). The effect of the load increase depends on the direction of the train. When the train is moving from a higher to a lower stiffness con- dition—such as exiting a bridge deck, grade crossing, or tunnel invert—the dynamic load is applied to the lower-stiffness track, increasing the rate of settlement. This condition is char- 4

acterized by deterioration of the track geome- try, ballast migration, and tie movement on the lower-stiffness track, as shown in Figure 1. When the train is moving from a lower- to higher-stiffness track, the load increase oc- curs on the high-stiffness side of the transi- tion over a short distance and is more of an impact loading. In this situation, typical prob- lems are rail surface fatigue, tie deteriora- tion, and rail seat pad deterioration as Figure 2 shows. In addition to the track stiffness change, the damage potential at track transitions is related to vehicle axle loads, speeds, and sus- pension characteristics. • Even if the dynamic effects are minimal, at- grade ballasted track may inherently settle more than ballasted track on a structure or direct- fixation track, creating a dip in the surface at the transition. This is especially true when the structure abutment is built on a deep pile foun- dation where settlement is negligible. • Settlement of at-grade track can be highly variable because of geotechnical issues affect- ing the subgrade performance such as low- strength soils, deficient soil placement and com- paction, poor drainage, and erosion (Briaud et al. 1997; Smekal 1997; Hoppe 2001). Envi- ronmental factors such as wet/dry and freeze/ thaw cycles also affect subgrade settlement behavior. Sasaoka and Davis (2005) categorize track tran- sition problems and solution approaches in terms of differential settlement, track stiffness, and damping changes that are intrinsic to the different structures. Using analytical techniques, an optimum damping value of 300 lb/in./sec/tie/rail was suggested for rail- way track that is adequately resilient and capable of efficiently distributing dynamic loads, particularly the higher-frequency impact loads. Field tests, how- ever, showed a value of 50 lb/in./sec/tie/rail to be typ- ical of stiff structures such as ballast and open deck bridges. Increased track damping on these structures will attenuate the dynamic loading at transitions. It is clear that the above issues are related and whether considered from the viewpoint of uneven track stiffness and deflections or differential settle- ment driven primarily by geotechnical conditions, the goal of any technique intended to improve the per- formance of transition track is to minimize dynamic loads by equalizing or smoothing the vertical sup- port condition and the dissipation of dynamic energy across the transition. Track Stiffness and Modulus This section briefly discusses the terms “track stiffness” and “track modulus.” Track stiffness (k) is the ratio of the applied wheel load (P) to rail de- flection (y): Hay (1982) and others define track modulus as the supporting force per unit length of rail per unit k P y= 5 Figure 1 Typical differential settlement of a freight railroad ballasted track bridge approach. Figure 2 Cracked concrete ties at the abutment of a freight railroad ballast deck bridge caused by impact loads.

deflection. The relationship between track stiffness and track modulus is defined with continuous beam on elastic foundation analysis as where u = the track modulus (lb/in./in.), E = the rail modulus of elasticity, and I = the rail moment of inertia. It is important to note the fundamental difference between track stiffness and track modulus: track stiff- ness includes all track components, including the rail, whereas the track modulus calculation excludes the flexural stiffness of the rail and only represents the rail support condition. Track modulus is consid- ered to be an important indicator of track quality and strength and is a required term in many track design calculations. Although ballasted track modulus is not often measured directly, as is the case with track geom- etry, measured track modulus values that have been published for specific track configurations in the freight operating environment (Kerr and Moroney 1993; Hay 1982; Read et al. 1994) indi- cate that moduli of 2,500 lb/in./in. or higher are typical of stable track structures, and values less than 1,500 lb/in./in. would be indicative of track prone to significant rail deflection and rapid track geometry degradation. To equate these numbers to rail transit, reference is made to Chapter 4 of TCRP Report 57: Track Design Handbook for Light Rail Transit, in which similar values are listed (Parsons Brinckerhoff Quade & Douglas, Inc. 2000). TCRP Report 57 gives typical modulus values for good- quality, timber-tie ballasted track as 2,000 to 2,500 lb/ in./in. and 5,000 to 8,000 lb/in./in. for concrete- tie track. Extremely high track modulus can also adversely affect track performance. According to Redden et al. (2002), track modulus values higher than 10,000 lb/ in./in. are undesirable because of the propensity for increased dynamic loads. Because the track is a resilient load distribution system, a decrease in resilience caused by a stiff support condition also decreases the transfer of wheel loads to adjacent ties, thereby increasing rail seat forces and ballast pressures. Lack of resilience also tends to am- plify impact rail forces that are generated by wheel and rail surface anomalies and the high-frequency u k EI= ( ) ( )4 3 1 364 rail vibrations associated with them. These high- frequency vibrations are often associated with cor- rugation development (Ahlbeck 1990; Hay 1982) and can generate undesirable noise and vibration conditions. As stated, track modulus represents the overall stiffness of the rail support system including rail fas- teners and pads, ties, ballast, and subgrade. A pa- rametric study performed by Selig and Li (1994), using the GEOTRACK model, indicated that stiff- ness of the subgrade was the most influential param- eter of ballasted track modulus. Secondary influence parameters included the granular layer (ballast and subballast) thickness, rail fastener pad stiffness, and tie type (wood or concrete). Tie spacing and tie di- mensions had minimal influence on the modulus. These findings implied that (1) maintenance activi- ties not directly related to improvement of the sub- grade, such as surfacing and tie renewals, will not significantly affect the track modulus and (2) envi- ronmental conditions that may affect subgrade prop- erties and strength, such as wet/dry and freeze/thaw cycles, can substantially change track modulus on a seasonal basis. The modulus of direct-fixation track is almost entirely a function of the stiffness and resilience of the elastomeric elements in the rail fastening system. The modulus of direct-fixation track is, therefore, much more consistent and easier to estimate than that of at-grade ballasted track. Transition Problems Test Results The following section presents the results of tests sponsored by the Association of American Railroads and the Federal Railroad Administration on freight railroad transition problems. Track Geometry Degradation (Differential Settlement) Figure 3 shows a comparison of the results of tests on average track settlement on four ballast deck railroad bridges and their approaches (Li and Davis 2005). As illustrated, the approaches experi- enced more track geometry degradation than the tracks on the bridges and the open tracks. The set- tlement of the track on the bridges was approxi- mately one-third of the settlement from the bridge approaches. Figure 4 further illustrates the differ- ential nature of track settlement in the approach 6

7TRACK on BRIDGE Se ttl em en t (i n) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 OPEN TRACKAPPROACH Figure 3 Comparison of track settlement accumulated over a maintenance interval (elevation change of unloaded rails). Distance from Bridge Abutment (ft) Se ttl em en t (i n) –20 0.0 0.5 1.0 1.5 2.0 2.5 Site 1 Site 2 Site 3 Site 4 20151050–5–10–15 Figure 4 Settlement in approach areas (track settlement on bridges not shown, negative and positive distance indicate two approaches for each bridge). Distance (ft) D ef le ct io n (in ) 1,200 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Bridge 1,6501,6001,5501,5001,4501,4001,3501,3001,250 Figure 5 Loaded track deflection profile. areas (settlement results versus distance from the bridge abutment). Figures 3 and 4 show accumulated track geom- etry degradation (differential track settlement). These results were measured from the unloaded rail sur- faces using survey equipment. Figure 5 shows the deflection profile results obtained under the TLV (Track Loading Vehicle) moving test load (40-kip wheel load) for one of the four sites tested. The re- sults were obtained after a surfacing maintenance operation, when the unloaded track profiles were “smooth.” Nevertheless, as illustrated in Figure 5, the approaches still showed large and variable track deflections under load, indicating an apparent factor contributing to poor vehicle/track interactions. Note Distance (ft) Tr ac k M od ul us (lb /in /in ) 900 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 Concrete bridge 1,4001,3001,2001,1001,000 Figure 6 Track modulus test results (Site 1). that deflection results shown in Figure 5 included not only the contribution from the ballast, subbal- last, and subgrade layers, but also the contribution of possible gaps and slacks between ties and ballast, which would close under the loaded condition. Track Modulus Figures 6 and 7 show track modulus test results obtained for two railroad ballast deck concrete bridges (with concrete ties) and their approaches (Li and Davis 2005). As shown, the track structure on concrete bridges had high stiffness characteristics. On average, the measured track modulus on these bridges was ap- proximately 10,000 lbs/in./in., which, as mentioned previously, is too high to accommodate desirable vehicle/track dynamic interaction. In addition, the change of track stiffness between bridge and approach was also too high (by a factor of 2, on average).

• Lowering the stiffness on the high side of the transition. Increasing Track Stiffness with Long Ties One of the oldest, simplest, and most widely used transition designs is installation of a series of in- creasingly longer ties on the ballasted track side of the transition. A typical layout is found in Plan No. 913-52 of the American Railway Engineering and Maintenance of Way Association (AREMA) Portfolio of Trackwork Plans (AREMA 2005a) and is shown in Figure 9. This method assumes the track stiffness is in- creased by the larger bearing area of the ties. How- ever, as Kerr and Moroney (1993) point out, its ef- fectiveness depends on uniform density of the ballast beneath the tie from the gage-side rail seat to the end of the tie (i.e., uniform tamping in this area). Longer ties may also exceed the embankment width on nar- row bridge approaches, allowing ballast to migrate from the tie ends. Using GEOTRACK analysis, Sussman and Selig (1998) indicate that although a longer tie may engage a larger ballast bearing area, it does little to increase the track stiffness. To increase stiffness, they recom- mend longer ties at reduced spacing and/or increas- ing the tie cross section, which in effect creates a stiffer track panel. MARTA Variable Length Timber-Tie Transition A case study was published by Patel and Jordan (1996), involving the Metropolitan Atlanta Rapid Transit Authority (MARTA), in which four 10-ft timber ties followed by four 11-ft and four 12-ft tim- 8 Distance (ft) Tr ac k M od ul us (lb /in /in ) 1,200 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 Concrete bridge 1,7001,6001,5001,4001,300 Figure 7 Track modulus test results (Site 3). Note: x = distance, k = stiffness Figure 8 Transition remedy in which the stiffness step change is modified with a gradual increase in stiffness. Discussion of Transition Remedies In the literature, a number of remedies have been proposed or used to provide gradual stiffness transi- tion. The following is a summary and discussion of those remedies. Kerr and Moroney (1993) Transition Categories Kerr and Moroney (1993) propose the following three categories of track transition remedies: • Smoothing the stiffness/modulus step change at the interface by gradually increasing stiff- ness on the lower-stiffness side of the transi- tion, as shown in Figure 8. • Increasing the bending of the rail-tie struc- ture (track panel) on the low-stiffness side of the transition.

ber ties were installed at 24-in. centers as a transition between ballasted at-grade, concrete-tie track and direct-fixation structures. The transition also in- cluded a 20-ft-long concrete transitional slab on the ballasted track approach. After modeling a number of options with GEO- TRACK, the design shown in Figure 10 was chosen for the test. Patel and Jordan (1996) indicate that the variable length design reduced maintenance costs by a factor of 3 when compared to designs that in- cluded the approach slab but not long ties. The vari- able length design has been adopted for future new construction. HMA Underlayment The positive performance of an HMA pave- ment layer placed between the subgrade and ballast to reinforce weak subgrades is well documented in Rose 1998, Rose et al. 2002, and Li et al. 2001. These studies indicate that when properly designed and installed an HMA layer will reduce subgrade stresses and differential settlement and extend track maintenance cycles. Because it is a structural layer, HMA can reduce subgrade stresses to levels that will not exceed the compressive strength of low-strength soils. However, in tests on the Union Pacific Railroad, Li and Davis (2005) found that HMA, placed on the approach to a ballast deck concrete bridge with a well-compacted subgrade, did not reduce the geometry deterioration of the approach compared with a similar approach without HMA. In the Li and Davis 2005 study, the track modulus of the approach with HMA was about 6,000 lb/in./in., which was very similar to the mod- ulus of the non-HMA approach. The modulus on the ballast deck bridge in both cases was between 9,000 and 12,000 lb/in./in. The test data indicated that the HMA layer provided little improvement to a subgrade 9 Source: AREMA 2005a. Figure 9 AREMA Plan No. 913-52 approach ties for open deck bridges and trestles.

with high load-bearing capacity, and the differential settlement seen on the approaches was caused pri- marily by settlement in the ballast layer rather than the subgrade. These results suggest that HMA and other meth- ods used to improve performance of weak subgrades, such as geocell and soil cement, will not improve ballast performance on stiff subgrades. For cases in which the approach track stiffness is already high, it would appear that trying to further increase the ap- proach stiffness is not as effective as reducing the stiffness of the bridge track. Increasing Approach Stiffness at Grade Crossing A transition to improve ride quality and mainte- nance demand at the approach to a grade crossing is described by Zarembski and Palese (2003). In this case, a transition grade crossing design was devel- oped, installed, and tested on New Jersey Transit’s 10 Source: Parsons Brinckerhoff Quade & Douglas, Inc. 2000. Figure 10 MARTA variable length timber-tie transition design.

Atlantic City line. The design was developed with the aid of an analytical model and provides a transi- tion from low-modulus “parent” track to a high- modulus, concrete-panel grade crossing in the fol- lowing steps: 1. Standard track with spikes and wood ties, 2. Wood ties with Pandrol clips, 3. 10-ft ties with Pandrol clips and single 8-ft field-side crossing panel installed between the rails, 4. 10-ft ties with Pandrol clips and 8-ft gage-side crossing panel installed between the rails, and 5. Full 24-ft crossing. Measurements of track modulus and vehicle ac- celeration taken before and after installation of the transition grade crossing indicated that the transition was effective at smoothing the track stiffness differ- ence and that a 77% reduction in the dynamic over- loading in the crossing had been achieved. Additional Rails The German Federal Railways have developed a design for the InterCity Express (ICE) high-speed lines on which lengths of rails are installed between the running rails and on the field side of the running rails to stiffen the ballasted track panel (Kerr and Moroney 1993). This condition often exists by de- fault, when guard rails installed on open deck bridges extend beyond the abutment to the ballasted track. Concrete Bridge Approach Slabs A reinforced concrete slab that rests on the abut- ment or slab structure and is tapered toward the at- grade end is often used at transitions to direct-fixation aerial structures and tunnel/subway inverts. AREMA recommends using a slab that is a minimum of 20 ft long and that is tapered from 18 in. at the structure end to 12 in. at the at-grade end. TCRP Report 57 (Parsons Brinckerhoff Quade & Douglas 2000) shows a slab that is 12 in. thick and 20 ft long over which the bal- last depth tapers from 12 in. at the structure end to 14 in. at the at-grade end (see Figure 11). General specifications for an approach slab design, based on a successful trial in the United Kingdom, are provided by Sharpe et al. (2002). In addition to the slab, this design calls for vertical adjustment of the rail on the direct-fixation bridge deck. The adjustable fasteners permit the rail on the ballasted side to be raised higher than the desired final elevation and to settle to the desired final ele- vation (design tamping). The paper indicates that incorporating the design-tamping capability has im- proved the transition performance over that of an approach slab by itself. The use of approach slabs is also a common high- way transition practice (Briaud et. al. 1997). The most successful highway slabs have slope changes of 1/200 or less, which is more gradual than railway designs, which are typically 2-in. changes over 20 ft or 1/120. Slab Track Approach Concrete approach slabs 25 ft in length were in- stalled at the Transportation Technology Center (TTC) in Pueblo, Colorado, to provide the transition from at-grade, concrete-tie track to a 500-ft-long concrete slab track test section (Bilow and Li 2005). The cast-in-place, 12-in.-thick reinforced concrete approach slab, prior to construction of the slab track, is shown in Figure 12. This transition design uses concrete ties with about 16 in. of ballast between the ties and the approach slab. The slab also has vertical walls to confine the ballast shoulder below the sub- grade level. Track modulus data taken on the completed track (see Figure 13) showed the modulus at the approach slabs to be more than two times the mod- ulus of the slab itself. In this case, the stiffness of the slab track direct fastening system had been suc- cessfully designed to approximate the nominal mod- ulus of the surrounding wood-tie track (approxi- mately 2,500 lb/in./in.). But the approach slab transition was over designed, creating an unneces- sarily high (6,000 to 7,000 lb/in./in.) track modu- lus at the interface. Stone Columns Stone columns (geo-piers) were installed at the Union Pacific Cedar River bridge approach for long- term performance monitoring (Davis et al. 2003). A stone column is simply a hole, 30 in. in diameter and 7 ft deep, that is bored into the subgrade beneath the rail seat and backfilled with aggregate material that is compacted in 6-in. layers. In this case, 10 pairs of columns spaced longitudinally at 5-ft centers were installed (see Figures 14 and 15). Stone columns are designed to strengthen and enhance drainage of weak subgrades. The test results have been positive, 11

12 Source: Parsons Brinckerhoff Quade & Douglas 2000. Figure 11 Transition design from TCRP Report 57: Track Design Handbook for Light Rail Transit.

13 Figure 12 Slab track transition at TTC. Figure 13 TTC slab track modulus data showing increase in modulus of the approach slabs. with no record of maintenance at the site during the first year of service. Piles In addition to stone columns, Li et al. (2003) in- dicate that other types of piles, including concrete, timber, and sand columns are accepted methods of stabilizing weak subgrades. Unless the end of the pile is on a firm foundation, skin friction provides most of the load transfer capacity. Therefore, the pile’s effectiveness will depend on its length, and different lengths can be used to smooth the stiffness of the approach. Other Geotechnical Considerations The use of stone columns, HMA, soil cement, geosynthetic materials, and piles are all techniques Figure 14 Hole boring in approach subgrade for stone column. Figure 15 Stone columns installed in approach subgrade. that can be used to reduce differential settlement of an approach track by reinforcing or stabilizing a weak subgrade. However, consideration should also be given to maximizing the subgrade performance, es- pecially during construction, with established geo- technical best practices such as the following: • Determining the soil characteristics prior to construction by performing in situ testing. • Using select noncohesive soils or applying ad- mixtures to existing soils if needed to improve subgrade strength. • Maintaining optimum moisture content and using correct compaction techniques for the soil type being placed, as well as ensuring adequate compaction when placing soil next to structures such as abutment backwalls.

the pads should match the track modulus of the at- grade approach using a methodology described by Kerr and Bathurst (2000), or the stiffness of the pads should meet damping requirements that at- tenuate high-frequency impact loads (Sasaoka and Davis 2005). The target vertical pad stiffness in Kerr and Bathurst 2000 is equal to the fastener spacing mul- tiplied by the desired track modulus. For example, the pad stiffness for a direct-fixation structure with fasteners at 30-in. centers needed to match a modu- lus of 3,000 lb/in./in. on the approach track would be 90,000 lb/in. The pad stiffness for the same ap- proach modulus to an open deck bridge with ties at 16-in. centers would be 48,000 lb/in. TCRP Report 57 (Parsons Brinckerhoff Quade & Douglas 2000) gives typical rail vertical stiffness values for direct- fixation track as 75,000 to 150,000 lb/in. CRP-CD-3: Performance of Direct-Fixation Track Software: Design Guidelines and Software (Battelle 1999) in- dicates that pad spring rates below 100,000 lb/in. ease the transition to ballasted track. It should be kept in mind that these vertical stiffness values are based on deflection of the rail at maximum wheel load and include deflection of the tie and structure in addition to the pad. In the case of direct-fixation concrete structures, deflection of the concrete is negligible; however, in the case of wood or composite tie decks, compression of the tie material may represent a substantial part of the total rail deflection. Reducing Track Stiffness on Ballast Deck Bridges Several test sites were established on a high- density freight route to determine the effectiveness of various tie materials at reducing the track stiff- ness on ballast deck bridges (Sasaoka et al. 2005). In all cases, concrete ties were installed on the ap- proach and on the ballast decks. Track measurements showed that modulus values on the bridges exceeded 8,000 lb/in./in. and were 2,000 to 3,500 lb/in./in. higher than modulus values on the approaches. Two methods were tested to reduce the bridge track mod- ulus: (1) replacing concrete ties with composite (plastic) ties on the bridge deck and (2) installing concrete ties on the bridge deck with 1-in.-thick rub- ber pads cast into the bottom of the ties. Figure 16 shows track modulus measured on the approaches and decks of bridges with three different tie types (concrete, composite, and concrete with rub- 14 • Ensuring maximum and uniform soil density by performing adequate soil density testing dur- ing construction. • Removing ruts, crowning or sloping the sub- grade surface, and/or using edge drains at the toe of the ballast section to prevent pocketing of free water in the track granular layer. • Lowering ground water levels or installing cut- off layers if needed to prevent capillary move- ment of ground water upward into cohesive soil embankments. • Allowing for adequate embankment width to accommodate the ballast/subballast depth. • Allowing for adequate embankment slope angles or the use of benches, retaining walls, or sheet piles for slope stability and control of erosion. A case where minimal maintenance has been per- formed on the approaches to an open deck steel bridge subjected to 40-ton axle load traffic is referenced by Joy et al. (2001). The approach embankments were constructed with a silty-sand material that was well compacted, and the paper stated that the performance of the approach was relatively good because of the embankment strength, width, and drainage. The track granular layer should also be adequate in terms of ballast and subballast material quality, layer depth, and cross section (Li et al. 2003). Gran- ular layer recommendations include the following: • 12-in. ballast layer depth, • Well-compacted subballast layer conforming to AREMA specifications in Chapter 1, Sec- tion 2.11, of the AREMA Manual for Railway Engineering (AREMA 2005b). • Total granular layer depth (ballast plus sub- ballast) using the formula in Chapter 1, Sec- tion 2.11.2.3, of the AREMA Manual for Rail- way Engineering (AREMA 2005b), and • The use of wing walls attached to the back wall of the abutment or other methods to con- tain the ballast and prevent migration. Rail Seat Pads on Open Deck Bridges and Direct-Fixation Structures One category of track transition remedies in- volves reduction of the track stiffness on the stiff or structure side of the transition. This can be accom- plished with elastomeric pads placed between the rail and rail seat. To be effective, the stiffness of

ber ties). As can be seen in Figure 16, both compos- ite ties and concrete ties with rubber pads were suc- cessful at reducing the modulus on the bridge. The composite ties equalized the modulus of the bridge and the modulus of the approach, and the rubber pads reduced the modulus of the bridge by a factor of 2.8. Rubber Tie Mats Another technique to reduce the stiffness on a ballast bridge deck was developed in Japan in the 1970s for the Shinkansen high-speed network. Ac- cording to Li et al. (2003), rubber mats were placed between the ties and ballast to reduce dynamic loads and ballast deterioration. The shape of the mats was designed to achieve a specific spring rate, and re- sults of extensive testing indicated that the mats were effective in reducing ballast wear. There was no mention, however, of how well the mats attenuated the dynamic loads or the quality of their long-term performance. Summary of Findings Descriptions of a variety of track transition de- signs and remedies were found in the literature re- viewed. In most cases, the techniques were aimed at either increasing the stiffness of the approach track or decreasing the stiffness and adding damping to the stiff track. Case studies in which at least initial performance improvements were noted included the following: • Use of longer ties and a concrete approach slab by MARTA to transition from ballasted at-grade track to direct-fixation structures. • Transition from at-grade ballasted track to a direct-fixation structure on a commuter/ intercity passenger service railway in the United Kingdom using an approach slab along with vertically adjustable direct-fixation fas- teners to allow design tamping of the ballasted approach track. • Installation of stone columns to strengthen and improve the drainage of a weak bridge ap- proach subgrade on a Union Pacific main line. • Design of a transition grade crossing system to smooth the track modulus across the approach to a highway crossing and reduce impact rail loads at the crossing on the New Jersey Transit Atlantic City line. • Installation of tie pads on open wood-tie bridge decks having stiffness/resilience characteris- tics designed to match the track modulus of the approach track on Amtrak’s northeast cor- ridor and on a Norfolk Southern main line with freight/intercity passenger service. • Reduction of the track modulus on a Union Pa- cific ballast deck bridge by replacing the exist- ing concrete deck ties with composite (plastic) ties or with concrete ties with a rubber pad cast into the tie bottom. The importance of following geotechnical best practices regarding soil selection, compaction, and 15 Tr a ck M od ul us (lb /in /in ) 12,000 10,000 8,000 6,000 4,000 2,000 0 Concrete ties Composite ties Concrete ties w/rubber pads Bridge Modulus Approach Modulus Figure 16 Comparison of track modulus values for different ballast deck bridge tie types.

drainage were also discussed in a number of studies. Properly designed and constructed subgrades can greatly minimize track transition problems. ANALYSIS OF REPRESENTATIVE TRACK TRANSITION DESIGNS Introduction In this section of the report, typical transition methods and conditions are analyzed using the GEOTRACK model. The model predicts a number of track response parameters, including vertical rail deflections (y), track modulus (u), and ballast and subgrade pressures for various track configurations, component properties, and wheel loads. The objective of the analysis is to determine the response of specific transition designs, based on the following track input variables, to representative rail transit wheel loadings. Representative transition configurations are the following: • At-grade ballasted track to direct-fixation aerial structure, • At-grade ballasted track to open deck bridge, • At-grade ballasted track to ballast deck bridge, • At-grade ballasted track with concrete approach slab to direct-fixation aerial structure, • At-grade ballasted track with HMA layer to direct-fixation aerial structure, • At-grade ballasted track with additional rails to direct-fixation aerial structure, and • At-grade ballasted track with AREMA long- tie approach to direct-fixation aerial structure. Track input variables are the following: • At-grade ballasted track: – 7-in. × 9-in. × 8.5-ft wood ties at 20-in. spacing, – 7.5-in. × 10-in. × 8.25-ft concrete ties at 28- in. spacing, – 12-in. ballast layer, – 8-in. subballast layer, and – Low, average, and high subgrade stiffness values (resilient modulus values of 2, 10, and 20 ksi, respectively). • Direct-fixation track: – Fasteners at 30-in. spacing and – Fastener stiffness values of 100, 150, 200, and 300 kip/in. • Open deck bridge: – Wood ties at 16-in. spacing. • Ballast deck bridge: – Concrete ties with 10-mm resilient tie pad at 28-in. spacing, – Concrete ties with 1-in.-thick resilient tie bottom pads at 28-in. spacing, and – 8- and 12-in. ballast layer. • HMA underlayment: – 8- and 12-in. layer. GEOTRACK Description (Selig and Waters 1994) The GEOTRACK computer model predicts the quasi-static response of the track to applied wheel loads. GEOTRACK represents the rail and ties as linear elastic beams that are connected with linear springs. The ties are supported on a multi- layer elastic system that represents various ele- ments of the track substructure. The rail can span up to 17 ties, and up to 4 wheel loads can be applied on the rail. The rail is defined by weight (lb/yd), cross area (A), modulus of elasticity (E), and moment of iner- tia (I). The fastener stiffness is defined as the verti- cal spring rate of the rail seat pad and, in the case of wood ties, includes compression of the wood. Ties are defined by length, cross section, weight, spacing, moment of inertia, and modulus of elasticity. The substructure is represented by as many as five elastic layers of defined depth with the depth of the bottom layer always being infinite. In addition to depth, each layer is defined by its resilient modulus (Er), which can be thought of as the soil’s modulus of elasticity, Poisson’s ratio (v), and material density. The GEOTRACK model treats the applied wheel loads as a vertical component only. Although GEOTRACK allows multiple wheel loads, only single wheel loads were used in this analysis. Three wheel loads were analyzed: 12,000, 15,000, and 22,500 lb. The 12,000-lb load was intended to represent light rail operations, the 15,000-lb load was based on the static weight of a Metro North cab car with full seated passenger load (Kentner et al. 1997, p. 270), and the 22,500-lb wheel load represents the Metro North static wheel load plus a 50% dynamic factor. The component property values used in the analy- sis are listed in Table 1. 16

17 Table 1 Basic GEOTRACK input properties used in the analysis Light rail Commuter car with full load Commuter car with 50% dynamic factor 12 15 22.5 Modulus of elasticity (E) (ksi) Moment of inertia (I) (in4) Cross-sectional area (A) (in2) Gage rail center-to-center (in) Weight (lb/yd) 30,00 65.9 11.25 59.25 115 Cross section (in x in) Length (in) Weight (lb) Spacing (in) E (ksi) I (in4) Fastener stiffness (kip/in) Wood 7 x 9 102 220 20 1,500 257 400 Concrete 7.5 x 10 99 600 28 4,500 469 1,000; 300 Direct- Fixation* 7.5 x 10 99 600 30 4,500 469 100; 150; 200; 300 Density (lb/cubic ft) Poissonís ratio Resilient modulus (Er) (ksi) Depth (in) Ballast 110 0.3 40 12 Subballast 120 0.35 25 8 Density (lb/cubic ft) Poisson’s ratio Er (ksi) Depth (in) Low Stiffness 90 0.35 2 infinite Average Stiffness 110 0.35 10 infinite High Stiffness 120 0.35 20 infinite Density (pcf) Poisson’s ratio Er (ksi) Depth (in) 150 0.35 100 infinite Density (lb/cubic ft) Poisson’s ratio Er (ksi) Depth (in) 145 0.3 800 8; 12 Density (lb/cubic ft) Poisson’s ratio E (ksi) Depth (in) 150 0.4 4,500 8; 12; 18 * Direct-fixation parameters represent the plinth. Single Wheel Load (kips) Ties and Fasteners Rail Granular Layers Subgrade Layers Bedrock Layer HMA Layer Concrete Slab

Analysis of Representative Track Configurations In this section, GEOTRACK model outputs of the vertical rail deflection, track modulus, and bal- last and subgrade pressures calculated for the repre- sentative track configurations are presented. In each case, the track configuration is described, and the rail deflections and modulus values from the 15,000-lb load along with the significant component properties are shown graphically. All the output values are listed in a table format. At-Grade Ballasted Track Conventional ballasted track on a subgrade foundation was modeled for three different sub- grade conditions: (1) low-stiffness subgrade (Er = 2 ksi), (2) average-stiffness subgrade (Er = 10 ksi), and (3) high-stiffness subgrade (Er = 20 ksi). The sub- grade Er values were based on test data from TTC in Pueblo, Colorado, and other sources (Read et al. 1994). Table 2 lists the rail deflections, track modulus values, ballast pressures at the top of the ballast layer, and subgrade pressures at the top of the subgrade layer for wheel loads of 12, 15, and 22.5 kips. The layer properties, rail deflection, and track modulus values are shown for wood and concrete ties in Fig- ures 17 and 18, respectively. Direct-Fixation Track Analysis Direct-fixation track on an aerial structure was modeled with fasteners spaced at 30-in. centers. The plinths were represented as the ties, the concrete deck slab and girders were represented as a 72-in. concrete layer, and the foundation was represented as a bedrock subgrade. Fastener stiffness values of 100, 150, 200, and 300 kip/in. were included in the analysis. Table 3 lists the deflection and track modulus values for wheel loads of 12, 15, and 22.5 kips. The layer properties, rail deflection, and track mod- ulus values for each fastener stiffness are shown in Figure 19. Open and Ballasted Deck Bridge Analysis Open deck bridges with wood ties attached to a steel superstructure and ballast deck bridges with wood and concrete ties were modeled similarly to the direct-fixation structure with the bridge super- structure sitting on a deep foundation at bedrock. The open deck bridge was modeled with wood ties at 16-in. centers and with and without a tie pad of 100 kip/in. stiffness. The ballast deck bridge was modeled with concrete ties at ballast depths of 12 and 8 in. and wood ties at ballast depth of 12 in. The concrete ties were equipped with a 10-mm studded rubber tie pad with stiffness of 300 kip/in. Concrete 18 Table 2 At-grade ballasted track rail deflection, track modulus, ballast stress, and subgrade stress data Concrete Ties Wood Ties 12 kips 15 kips 22.5 kips 12 kips 15 kips 22.5 kips High-Stiffness Rail Deflection (in) 0.021 0.027 0.040 0.026 0.032 0.048 Subgrade Er  20 ksi Modulus (lb/in/in) 9,236 9,236 9,236 7,269 7,269 7,269 Ballast Stress (psi) 16.9 21.1 31.6 14.4 18.0 26.9 Subgrade Stress (psi) 3.7 4.6 7.0 3.6 4.5 6.8 Average-Stiffness Rail Deflection (in) 0.030 0.038 0.057 0.033 0.042 0.063 Subgrade Er  10 ksi Modulus (lb/in/in) 5,757 5,757 5,757 5,100 5,100 5,100 Ballast Stress (psi) 15.8 19.8 29.6 13.9 17.4 26.1 Subgrade Stress (psi) 3.3 4.2 6.2 3.4 4.2 6.3 Low-Stiffness Rail Deflection (in) 0.091 0.114 0.170 0.096 0.120 0.180 Subgrade Er  2 ksi Modulus (lb/in/in) 1,336 1,336 1,336 1,249 1,249 1,249 Ballast Stress (psi) 20.0 25.0 37.6 11.4 14.3 21.4 Subgrade Stress (psi) 1.9 2.4 3.6 2.0 2.5 3.8

19 Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) 15,000 lb Low-stiffness subgrade Er = 2 ksi Average-stiffness subgrade Er = 10 ksi High-stiffness subgrade Er = 20 ksi 8” subballast layer 12” ballast layer 115 RE 7” x 9” x 8’6” Wood tie 20” spacing –0.14 –0.12 –0.1 –0.08 –0.06 –0.04 –0.02 0 subgrade Er = 20 ksi subgrade Er = 10 ksi subgrade Er = 2 ksi u = 1.249 ksi u = 5.100 ksi u = 7.269 ksi –200 20015010050–50 0–100–150 Figure 17 Rail deflection and track modulus values for at-grade wood-tie ballasted track under 15-kip wheel loading and low-, average-, and high-stiffness subgrades.

20 15,000 lb Low-stiffness subgrade Er = 2 ksi Average-stiffness subgrade Er = 10 ksi High-stiffness subgrade Er = 20 ksi 12” ballast layer 8” subballast layer 115 RE 8’3” Concrete tie 28” spacing Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) subgrade E = 20,000 psi subgrade E = 10,000 psi subgrade E = 2,000 psi –0.12 –0.1 –0.08 –0.06 –0.04 -0.02 u = 1.336 ksi u = 5.757 ksi u = 9.236 ksi –200 20015010050–50 0–100–150 0 Figure 18 Rail deflection and track modulus values for at-grade concrete-tie ballasted track under 15-kip wheel loading and low-, average-, and high-stiffness subgrades. Table 3 Direct-fixation aerial structure rail deflection and track modulus data 12 kips 15 kips 22.5 kips 100 kip/in Fastener Stiffness Rail Deflection (in) 0.046 0.057 0.086 Modulus (lb/in/in) 3,330 3,330 3,330 150 kip/in Fastener Stiffness Rail Deflection (in) 0.034 0.042 0.063 Modulus (lb/in/in) 4,997 4,997 4,997 200 kip/in Fastener Stiffness Rail Deflection (in) 0.027 0.034 0.051 Modulus (lb/in/in) 6,668 6,668 6,668 300 kip/in Fastener Stiffness Rail Deflection (in) 0.020 0.025 0.038 Modulus (lb/in/in) 10,018 10,018 10,018

21 15,000 lb 115 RE Concrete deck and girder layer = 72” Bedrock Er = 100 ksi Fastener spacing 30” spacing Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) –0.07 –0.06 –0.05 –0.04 –0.03 –0.02 –0.01 0 0.01 Pad stiffness = 100 kip/in Pad stiffness = 150 kip/in Pad stiffness = 200 kip/in Pad stiffness = 300 kip/in u = 3.330 ksi u = 10.018 ksi u = 4.997 ksi u = 6.668 ksi –200 20015010050–50 0–100–150 Figure 19 Rail deflection and track modulus values for direct-fixation track and various pad stiffnesses on an aerial structure under 15-kip wheel loading.

ties were also modeled with a rubber pad of 100 kip/ in. stiffness bonded to the tie bottom. Table 4 lists the deflection and track modulus values for wheel loads of 12, 15, and 22.5 kips. The layer properties, rail deflection basin, and track modulus values for each fastener stiffness are shown for the ballast deck bridge in Figure 20. Concrete Approach Slab A 20-ft-long reinforced concrete approach slab placed between the ballast and subballast layers was modeled for wood- and concrete-tie track. Variables in the analysis included average and low-stiffness subgrade (Er = 10 and 2 ksi), tie type, and slab thick- ness. An approach slab on a high-stiffness subgrade was not analyzed as the track modulus values would greatly exceed 10,000 lb/in./in., which is considered to be excessive. Tables 5 and 6 list the rail deflections, track modulus values, ballast pressures at the top of the ballast layer, and subgrade pressures at the top of the subgrade layer for wheel loads of 12, 15, and 22.5 kips. The layer properties, rail deflection basin, and track modulus values are shown in Figures 21 and 22. HMA Underlayment An HMA layer placed between the ballast and subballast layers as an approach transition was modeled for wood- and concrete-tie track. Vari- ables in the analysis included low-, average-, and high-stiffness subgrade and tie type. Substructure layers included 12-in. ballast and 8-in. subballast. Tables 7 and 8 list the rail deflection, track mod- ulus, ballast pressure at the top of the ballast layer and subgrade pressure at the top of the subgrade layer for wheel loads of 12, 15, and 22.5 kips. The layer prop- erties, rail deflection basin, and track modulus values are shown in Figures 23 and 24. Additional Rails The transition design in which two additional rails are added to the track panel to increase the stiffness of the panel and reduce rail deflection was analyzed. In this case, the rail was doubled in weight, area, and moment of inertia to simulate the additional rail. Vari- ables in the analysis included subgrade stiffness (low, average, and high as before) and HMA layer depth. Table 9 lists the rail deflection, track modulus, ballast pressure at the top of the ballast layer, and 22 Table 4 Open and ballast deck bridge rail deflection and track modulus data 12 kips 15 kips 22.5 kips Ballast Deck Bridge with 8-in Ballast Rail Deflection (in) 0.021 0.026 0.039 Depth and Concrete Ties Modulus (lb/in/in) 9,595 9,595 9,595 Ballast Deck Bridge with 8-in Ballast Depth Rail Deflection (in) 0.045 0.056 0.084 and Rubber Pad on Bottom of Concrete Ties Modulus (lb/in/in) 3,432 3,432 3,432 Ballast Deck Bridge with 12-in Rail Deflection (in) 0.021 0.026 0.040 Ballast Depth and Concrete Ties Modulus (lb/in/in) 9,336 9,336 9,336 Ballast Deck Bridge with 12-in Ballast Rail Deflection (in) 0.045 0.057 0.085 Depth and Rubber Pad on Bottom Modulus (lb/in/in) 3,398 3,398 3,398 of Concrete Ties Ballast Deck Bridge with 12-in Ballast Rail Deflection (in) 0.020 0.025 0.037 Depth and Wood Ties Modulus (lb/in/in) 10,315 10,315 10,315 Open Deck Bridge with Wood Ties Rail Deflection (in) 0.013 0.016 0.025 Modulus (lb/in/in) 17,287 17,287 17,287 Open Deck Bridge with Wood Ties Rail Deflection (in) 0.030 0.037 0.056 and 100-kip/in Stiffness Tie Pad Modulus (lb/in/in) 5,903 5,903 5,903

subgrade pressure at the top of the subgrade layer for wheel loads of 12, 15, and 22.5 kips. Transition Analysis A number of graphs are presented in this sec- tion to compare the track modulus and rail deflec- tion values for the various modeled track configu- rations. In each case, the at-grade ballasted track is shown on the left of the graph, the direct-fixation or bridge structure is shown on the right side of the graph, and the transition designs are shown be- tween the two. Track Modulus Transition In Figure 25, the track modulus of concrete- tie track is compared to the modulus of a direct- fixation structure and ballast deck bridge with con- crete ties. The at-grade ballasted track data shows the track modulus range for subgrade resilient moduli of 2, 10, and 20 ksi. Concrete approach slabs of 8-, 12-, and 18-in. thicknesses and HMA layers of 8- and 12-in. thicknesses are also included in Figure 25 to show their effectiveness at increasing the track mod- ulus of low-stiffness and average subgrades. Fig- ure 26 shows a similar graph for wood-tie track. 23 15,000 lb 115 RE Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) 0.02 Concrete Ties Concrete Ties w/Bottom Pad Wood Ties 12” ballast layer Bedrock Er = 100 ksi Concrete ties 28” spacing Wood ties 20” spacing 72” concrete deck and cap -0.14 -0.12 -0.1 -0.08 -0.06 -0.04 -0.02 0 u = 3.398 ksi u = 9.336 ksi u = 10.315 ksi –200 20015010050–50 0–100–150 Figure 20 Rail deflection and track modulus values for wood and concrete ties on ballast deck bridge under 15-kip wheel load.

Rail Deflection Transition In addition to track modulus, it is useful to com- pare the rail deflections of the various track configu- rations. Because track modulus is a power function of the rail deflection, as shown in Figure 27, small am- plitude deflections tend to correspond to increasingly higher modulus values. Considering track modulus alone may, therefore, exaggerate the transition re- quirements as compared with the rail deflection. In Figures 28 through 30, rail deflections for the various track transitions that were calculated for concrete track under wheel loads of 12, 15, and 22.5 kips are shown. The layouts of the various tran- sition configurations are similar to the track modulus graphs, with the at-grade track on the left, the struc- tures on the right, and the transition designs in the middle. The wood-tie data are shown in Figures 31 through 33. Analysis of Tie Length, Tie Cross Section, and Tie Spacing Increased tie length and cross section and de- creased tie spacing are often considered to be effective track transition configurations. GEOTRACK model- ing, however, indicated that these transition methods had little, if any, benefit in terms of reduced rail de- flection or increased track modulus. However, there may be benefits in terms of other performance criteria not discussed in this digest. 24 Table 5 Rail deflection, modulus, ballast stress, and subgrade stress data for a concrete approach slab with concrete ties 8-in Slab 12-in Slab 18-in Slab 12 15 22.5 12 15 22.5 12 15 22.5 kips kips kips kips kips kips kips kips kips Average- Rail Deflection (in) 0.025 0.031 0.046 0.022 0.027 0.041 0.019 0.024 0.036 Stiffness Modulus (lb/in/in) 7,622 7,622 7,622 8,896 8,896 8,896 10,759 10,759 10,759 Subgrade Ballast Stress (psi) 17.3 21.7 32.5 18.4 23.0 34.6 14.0 17.5 26.3 Subgrade Stress (psi) 1.3 1.6 2.7 0.9 1.1 1.6 0.6 0.7 1.0 Low- Rail Deflection (in) 0.061 0.076 0.114 0.049 0.061 0.091 0.037 0.047 0.070 Stiffness Modulus (lb/in/in) 2,288 2,288 2,288 3,085 3,085 3,085 4,404 4,404 4,404 Subgrade Ballast Stress (psi) 16.8 2.1 27.4 18.6 23.2 34.8 13.8 17.3 24.9 Subgrade Stress (psi) 0.7 0.8 1.3 0.4 0.5 0.8 0.2 0.3 0.5 Table 6 Rail deflection, modulus, ballast stress, and subgrade stress data for a concrete approach slab with wood ties 8-in Slab 12-in Slab 18-in slab 12 15 22.5 12 15 22.5 12 15 22.5 kips kips kips kips kips kips kips kips kips Average- Rail Deflection (in) 0.029 0.036 0.054 0.026 0.033 0.049 0.023 0.029 0.044 Stiffness Modulus (lb/in/in) 6,129 6,129 6,129 6,967 6,967 6,967 8,138 8,138 8,138 Subgrade Ballast Stress (psi) 16.5 19.1 31.0 18.8 23.5 35.3 15.1 18.9 28.3 Subgrade Stress (psi) 1.3 1.6 2.4 0.9 1.1 1.6 0.5 0.7 1.0 Low- Rail Deflection (in) 0.065 0.082 0.122 0.053 0.066 0.100 0.042 0.053 0.079 Stiffness Modulus (lb/in/in) 2,080 2,080 2,080 2,730 2,730 2,730 3,718 3,718 3,718 Subgrade Ballast Stress (psi) 14.9 18.6 27.9 18.6 23.3 35.0 16.3 20.4 30.6 Subgrade Stress (psi) 0.7 0.8 1.3 0.4 0.5 0.8 0.2 0.3 0.4

Matching Fastener Stiffness with Subgrade Resilient Modulus Subgrade Er values that are compatible with fas- tener stiffness values in terms of track modulus and rail deflection are shown in Table 10. The compatibil- ity criterion used in Table 10 was ± 0.01 in. for rail de- flection. This analysis shows the subgrade conditions with wood or concrete ties that most closely match the resilience of the rail on a structure at typical fastener stiffnesses and that, therefore, would minimize the need for an additional transition design. Matching Slab and HMA Transition Designs to Fastener Stiffness In Table 11, the track modulus and rail deflections of the concrete approach slab or HMA underlayment transition designs are matched to the fastener stiffness values on the structure using the same compatibility 25 15,000 lb 115 RE Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) 0 12” ballast layer 8” subballast layer Wood tie 20” spacing -0.09 -0.08 -0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 8" slab Er 2 ksi subgrade 12" slab Er 2 ksi subgrade 18" slab Er 2 ksi subgrade 8" slab Er 10 ksi subgrade 12" slab Er 10 ksi subgrade 18" slab Er 10 ksi subgrade u = 2.080 ksi u = 2.730 ksi u = 3.718 ksi u = 6.129 ksi u = 6.967 ksi u = 8.138 ksi 8”, 12”, and 18” reinforced concrete slab Low-stiffness Er = 2 ksi subgrade Average-stiffness subgrade Er = 10 ksi –200 20015010050–50 0–100–150 Figure 21 Rail deflection and track modulus values for 8-, 12-, and 18-in. depth concrete approach slab transitions with wood ties on low- and average-stiffness subgrades under 15-kip wheel load.

criterion (± 0.01 in. for rail deflection) that was used in Table 10. This analysis shows the transition de- signs that most closely match the resilience of the rail on a structure at typical fastener stiffnesses for given subgrade Er and tie types. Analysis Conclusions The following conclusions were drawn from the GEOTRACK analysis of representative track tran- sition designs: • The track modulus and rail deflection of at- grade ballasted track were dominated by the subgrade stiffness. Therefore, to significantly increase the stiffness of at-grade ballasted track, modification or reinforcement of the subgrade is required. • The track modulus and rail deflection of direct- fixation track and track on ballast deck/open deck bridges are dominated by the vertical fas- tener stiffness as provided by elastomeric ele- ments at the rail seat or tie bottom. Direct- 26 15,000 lb 115 RE Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) 0 12” ballast layer 8” subballast layer Concrete tie 28” spacing 8" slab Er 2 ksi subgrade 12" slab Er 2 ksi subgrade 18" slab Er 2 ksi subgrade 8" slab Er 10 ksi subgrade 12" slab Er 10 ksi subgrade 18" slab Er 10 ksi subgrade 8”, 12”, and 18” reinforced concrete slab Low-stiffness Er = 2 ksi subgrade Average-stiffness subgrade Er = 10 ksi –0.1 –0.09 –0.08 –0.07 –0.06 –0.05 –0.04 –0.03 –0.02 –0.01 0 u = 2.288 ksi u = 3.085 ksi u = 4.404 ksi u = 7.622 ksi u = 8.896 ksi u = 10.759 ksi –200 20015010050–50 0–100–150 Figure 22 Rail deflection and track modulus values for 8-, 12-, and 18-in. depth concrete approach slab transitions with concrete ties on low- and average-stiffness subgrades under 15-kip wheel load.

fixation fastener stiffness between 100 and 200 kip/in. matches the rail deflections of track on subgrades with resilient moduli between 5 and 15 ksi. • Concrete approach slabs placed between the ballast and subballast layers produced the most substantial track modulus/rail deflection benefits. HMA underlayment also provided benefits, but was not as effective as concrete in terms of layer thickness. For example, a 12-in.-thick HMA layer produced about the same decrease in rail deflections as an 8-in. concrete slab on subgrades with resilient mod- ulus less than 5 ksi. • Increasing subgrade resilient modulus reduced the potential benefit from increasing the thick- ness of concrete slabs or HMA layers. For ex- ample, rail deflections were reduced 0.06 in. by an 8-in. concrete slab on a 2-ksi resilient mod- ulus and reduced 0.10 in. by an 18-in. slab on the same subgrade, giving a difference of 0.04 in. between the two thicknesses. The dif- 27 Table 7 Rail deflection, modulus, ballast stress, and subgrade stress data for HMA underlayment with concrete-tie track 8-in Layer 12-in Layer Concrete Ties 12 kips 15 kips 22.5 kips 12 kips 15 kips 22.5 kips High-Stiffness Rail Deflection (in) 0.020 0.024 0.037 0.019 0.023 0.035 Subgrade Modulus (lb/in/in) 10,368 10,368 10,368 11,059 11,059 11,059 Ballast Stress (psi) 17.2 21.5 32.2 16.2 20.3 30.4 Subgrade Stress (psi) 2.4 3.0 4.5 2.1 2.6 3.8 Average-Stiffness Rail Deflection (in) 0.026 0.034 0.050 0.024 0.031 0.047 Subgrade Modulus (lb/in/in) 6,819 6,819 6,819 7,492 7,492 7,492 Ballast Stress (psi) 16.2 22.2 33.3 15.6 21.6 32.5 Subgrade Stress (psi) 2.0 2.5 3.8 1.7 2.1 3.2 Low-Stiffness Rail Deflection (in) 0.071 0.090 0.135 0.062 0.079 0.118 Subgrade Modulus (lb/in/in) 1,834 1,834 1,834 2,190 2,190 2,190 Ballast Stress (psi) 17.1 23.5 35.2 17.1 24.0 35.9 Subgrade Stress (psi) 1.1 1.4 2.0 0.9 1.0 1.7 Table 8 Rail deflection, modulus, ballast stress, and subgrade stress data for HMA underlayment with wood-tie track 8-in Layer 12-in Layer Wood Ties 12 kips 15 kips 22.5 kips 12 kips 15 kips 22.5 kips High-Stiffness Rail Deflection (in) 0.024 0.030 0.045 0.023 0.029 0.043 Subgrade Modulus (lb/in/in) 7,960 7,960 7,960 8,311 8,311 8,311 Ballast Stress (psi) 17.2 21.5 32.3 17.1 21.3 32.0 Subgrade Stress (psi) 2.4 2.9 4.4 2.0 2.5 3.8 Average-Stiffness Rail Deflection (in) 0.031 0.039 0.058 0.029 0.037 0.055 Subgrade Modulus (lb/in/in) 5,580 5,580 5,580 6,038 6,038 6,038 Ballast Stress (psi) 16.5 20.7 31.0 17.8 22.2 33.3 Subgrade Stress (psi) 2.0 2.5 3.7 1.6 2.1 3.1 Low-Stiffness Rail Deflection (in) 0.077 0.096 0.144 0.067 0.084 0.126 Subgrade Modulus (lb/in/in) 1,682 1,682 1,682 1,994 1,994 1,994 Ballast Stress (psi) 14.6 27.3 18.2 17.2 21.5 32.2 Subgrade Stress (psi) 1.1 2.1 1.4 0.9 1.1 1.7

ference between the deflections on a 10-ksi subgrade, however, was 0.01 in., or about one- fourth that of the 2-ksi subgrade. There was almost no deflection difference between the 8- and 12-in. HMA layers on a 10-ksi subgrade. • Placing two additional rails on the ties to in- crease the stiffness of the track panel produced modest benefits with subgrades having resilient modulus less than 5 ksi. • Other changes to the track superstructure, such as reduced tie spacing, installation of longer ties, or using ties with larger cross sections, had an insignificant effect on the modulus or rail deflections. • The analyzed transit wheel loads generated dif- ferentials in the rail deflection amplitudes at the transition interface that were less than 0.1 in. for almost all the transition conditions ana- lyzed. The largest rail deflection differential (0.15 in.) was seen at the transition between a ballasted track with very weak subgrade and an open deck bridge with a commuter car dynamic 28 15,000 lb 115 RE Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) 0 12” ballast layer 8” subballast layer Concrete tie 28” spacing 8" slab Er 2 ksi subgrade 12" slab Er 2 ksi subgrade 18" slab Er 2 ksi subgrade 8" slab Er 10 ksi subgrade 12" slab Er 10 ksi subgrade 18" slab Er 10 ksi subgrade 8” and 12” HMA layer Low-stiffness subgrade Er = 2 ksi Average-stiffness subgrade Er = 10 ksi High-stiffness subgrade Er = 20 ksi -0.1 -0.09 -0.08 -0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 u = 2.288 ksi u = 3.085 ksi u = 4.404 ksi u = 7.622 ksi u = 8.896 ksi u = 10.759 ksi –200 20015010050–50 0–100–150 Figure 23 Rail deflection and track modulus values for 8-, 12-, and 18-in. depth HMA underlayment transitions on low-, average-, and high-stiffness subgrades with concrete ties under 15-kip wheel load.

Figure 24 Rail deflection and track modulus values for 8-, 12-, and 18-in. depth HMA underlayment transitions on low-, average-, and high-stiffness subgrades with wood ties under 15-kip wheel load. 29 15,000 lb 115 RE Distance (in) Ve rt ic al R ai l D ef le ct io n (in ) 12” ballast layer 8” subballast layer Wood tie 20” spacing 8" HMA layer Er = 2 ksi 8" HMA layer Er = 10 ksi 8" HMA layer Er = 20 ksi 12" HMA Layer Er = 2 ksi 12" HMA layer Er = 10 ksi 12" HMA layer Er = 20 ksi 8” and 12” HMA layer Low-stiffness subgrade Er = 2 ksi Average-stiffness subgrade Er = 10 ksi High-stiffness subgrade Er = 20 ksi -0.1 -0.09 -0.08 -0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 u = 1.682 ksi u = 1.994 ksi u = 5.580 ksi u = 6.038 ksi u = 7.960 ksi u = 8.311 ksi 15010050–50 0–100–150 load of 22.5 kips. This same transition under the light rail 12-kip wheel load produced a de- flection differential of about 0.09 in. Typical deflection differentials between ballasted track on average-stiffness subgrades and direct- fixation structures were less than 0.04 in. • Embedded track configurations were not ana- lyzed. As currently structured, GEOTRACK is incapable of modeling a rail with continuous support, as is the case with embedded track. DISCUSSION OF TRANSITION DESIGNS The following transition designs are considered to be the most efficient for rail transit applications based on the preceding literature review and GEO- TRACK analysis: • Matching the vertical fastener stiffness of direct-fixation track, ballast deck, or open deck bridges to the track modulus and rail deflection behavior of the at-grade ballasted track,

30 Table 9 Rail deflection, modulus, ballast stress, and subgrade stress data for additional rail design Concrete Ties Wood Ties Concrete Ties 12 kips 15 kips 22.5 kips 12 kips 15 kips 22.5 kips High-Stiffness Rail Deflection (in) 0.019 0.024 0.036 0.023 0.028 0.043 Subgrade Modulus (lb/in/in) 8,538 8,538 8,538 6,770 6,770 6,770 Ballast Stress (psi) 14.5 17.8 26.7 12.0 15.0 22.5 Subgrade Stress (psi) 3.0 4.0 6.0 3.2 4.0 5.9 Average-Stiffness Rail Deflection (in) 0.027 0.034 0.052 0.031 0.039 0.059 Subgrade Modulus (lb/in/in) 5,234 5,234 5,234 4,420 4,420 4,420 Ballast Stress (psi) 14.9 18.7 28.0 11.2 14.0 21.1 Subgrade Stress (psi) 2.7 3.4 5.1 2.7 3.4 5.1 Low-Stiffness Rail Deflection (in) 0.083 0.104 0.156 0.088 0.109 0.164 Subgrade Modulus (lb/in/in) 1,195 1,195 1,195 1,118 1,118 1,118 Ballast Stress (psi) 18.1 22.7 34.0 9.1 11.4 17.1 Subgrade Stress (psi) 1.6 2.0 3.0 1.6 2.1 3.1 0 2,000 4,000 6,000 8,000 10,000 12,000 At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Direct- fixation aerial structure Ballast deck bridge Tr ac k M od ul us (lb /in /in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness Nominal concrete tie ballasted track modulus Nominal direct-fixation/ballast deck bridge modulus Figure 25 Comparison of track modulus values for concrete-tie track transition configurations.

31 Tr ac k M od ul us (lb /in /in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness Nominal wood tie ballasted track modulus Nominal direct-fixation/ballast deck bridge modulus 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000 At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Direct- fixation aerial structure Ballast deck bridge Open deck bridge Figure 26 Comparison of track modulus values for wood-tie track transition configurations. y = 0.0737x-1.3333 0 5 10 15 20 25 30 35 40 0 115 RE Rail Deflection (in) Tr ac k M od ul us U (k si) 0.250.20.150.10.05 Figure 27 Track modulus as a function of rail deflection under 15-kip load. quate resilience to transition to ballasted track on an average-stiffness subgrade. • The use of resilient tie pads with a nominal stiffness of 100 kip/in. on open deck timber bridges provides adequate resilience to transi- tion to ballasted track on an average subgrade. • Low-stiffness subgrades with Er values less than 5 ksi require some modification in addi- tion to the controlled resilience of the structure track. These subgrades are typically made up of cohesive soils (clays and silts) with mois- ture contents higher than optimum. Increasing the modulus of track on a low-stiffness sub- grade requires modification of the physical state of the soil and/or installation of a struc- tural reinforcing layer between the ballast and subgrade such as HMA underlayment or a concrete approach slab. The introduction of a structural layer, however, creates an additional interface point at the end of the slab/layer that is away from the structure, and this interface may require a second transition design in the form of increased ballast depth or stepping the layer thickness to be implemented. • Avoiding the creation of weak subgrade con- ditions during new construction by careful soil selection and the application of geotech- nical best practices is recommended. without modification of the at-grade track, pro- vides the most efficient and cost-effective de- sign. Direct-fixation fasteners with stiffness values between 100 and 200 kip/in. are com- patible with ballasted tracks with average- stiffness subgrades (Er values between 5 and 15 ksi). The analysis showed the rail deflection differentials for these designs to be less than 0.04 in. for all three wheel loads. • The use of 10-mm concrete-tie pads with a nominal stiffness of 200 to 300 kip/in. on bal- last deck bridge concrete ties provides ade-

32 Ve rti ca l R ai l D ef le ct io n Un de r 1 2- kip W he el L oa d (in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness –0.1 –0.09 –0.08 –0.07 –0.06 –0.05 –0.04 –0.03 –0.02 –0.01 0 At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Additional rails Direct- fixation structure Ballast deck bridge Figure 28 Comparison of rail deflection data for concrete-tie track transition configurations under 12-kip wheel loading. Ve rti ca l R ai l D ef le ct io n Un de r 1 5- kip W he el L oa d (in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Additional rails Direct- fixation structure Ballast deck bridge –0.12 –0.1 –0.08 –0.06 –0.04 –0.02 0 Figure 29 Comparison of rail deflection data for concrete-tie track transition configurations under 15-kip wheel loading.

33 Figure 30 Comparison of rail deflection data for concrete-tie track transition configurations under 22.5-kip wheel loading. Ve rti ca l R ai l D ef le ct io n Un de r 2 2. 5- kip W he el L oa d (in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Additional rails Direct- fixation structure Ballast deck bridge –0.12 –0.1 –0.08 –0.06 –0.04 –0.02 0 –0.12 –0.1 –0.08 –0.06 –0.04 –0.02 0 At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Additional rails Direct- fixation structure Ballast deck bridge Open deck bridge Ve rti ca l R ai l D ef le ct io n Un de r 1 2- kip W he el L oa d (in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness Figure 31 Comparison of rail deflection data for wood-tie track transition configurations under 12-kip wheel loading.

34 At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Additional rails Direct- fixation structure Ballast deck bridge Open deck bridge Ve rti ca l R ai l D ef le ct io n Un de r 1 5- kip W he el L oa d (in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness –0.14 –0.12 –0.1 –0.08 –0.06 –0.04 –0.02 0 At-grade ballasted track 8" concrete slab 12" concrete slab 18" concrete slab 8" HMA layer 12" HMA layer Additional rails Direct- fixation structure Ballast deck bridge Open deck bridge Ve rti ca l R ai l D ef le ct io n Un de r 2 2. 5- kip W he el L oa d (in ) Low-stiffness subgrade Er=2ksi Average-stiffness subgrade Er=10ksi High-stiffness subgrade Er=20ksi 100 kip/in fastener stiffness 150 kip/in fastener stiffness 200 kip/in fastener stiffness 300 kip/in fastener stiffness –0.2 –0.18 –0.16 –0.14 –0.12 –0.1 –0.08 –0.06 –0.04 –0.02 0 Figure 32 Comparison of rail deflection data for wood-tie track transition configurations under 15-kip wheel loading. Figure 33 Comparison of rail deflection data for wood-tie track transition configurations under 22.5-kip wheel loading.

35 Table 10 Matching subgrade resilient modulus to fastener stiffness on structures Tie Type/ Direct-Fixation and Ballast Deck Open Deck Bridge Subgrade Er (ksi) Bridge Fastener Stiffness (kip/in) Fastener Stiffness (kip/in)* Wood/6 100 50 Wood/8 150 75 Wood/10 150 100 Wood/12 150–200 100 Wood/15 200 100 Wood/20 300 200 Concrete/5 100 NA** Concrete/8 150 NA Concrete/10 150–200 NA Concrete/12 200 NA Concrete/15 200–300 NA Concrete/20 300 NA *Wood ties with elastomeric pad between the tie plate and tie. **NA = not applicable. Table 11 Matching subgrade resilient modulus with transition design to fastener stiffness on structures Tie Type/ Direct-Fixation and Open Deck Bridge Subgrade Transition Type/ Ballast DeckBridge Fastener Stiffness Er (ksi) Thickness (in) Fastener Stiffness (kip/in) (kip/in) Wood/2 Concrete slab/18 100 75 Wood/6 Concrete slab/18 200 100 Wood/6 Concrete slab/12 150 100 Wood/6 Concrete slab/8 150 75 Wood/10 Concrete slab/18 200–300 150 Wood/10 Concrete slab/12 200 150 Wood/10 Concrete slab/8 200 100 Wood/15 Concrete slab/12 300 150 Wood/15 Concrete slab/8 200 150 Wood/6 HMA/12 150 75 Wood/6 HMA/8 125 75 Wood/10 HMA/12 200 100 Wood/10 HMA/8 150–200 100 Wood/15 HMA/12 200 150 Wood/15 HMA/8 200 150 Wood/20 HMA/12 200–300 150 Wood/20 HMA/8 200–300 150 Concrete/2 Concrete slab/18 150 NA* Concrete/2 Concrete slab/12 100 NA Concrete/6 Concrete slab/18 200 NA Concrete/6 Concrete slab/12 200 NA Concrete/6 Concrete slab/8 200 NA Concrete/10 Concrete slab/18 300 NA Concrete/10 Concrete slab/12 300 NA Concrete/10 Concrete slab/8 200 NA Concrete/15 Concrete slab/8 300 NA Concrete/6 HMA/12 150 NA Concrete/6 HMA/8 125 NA Concrete/10 HMA/12 200–300 NA Concrete/10 HMA/8 200 NA Concrete/15 HMA/8 300 NA *NA = not applicable.

REFERENCES Ahlbeck, D. R. and L. E. Daniels. 1990. “A Review of Rail Corrugation Processes under Different Operating Modes.” In Technical Papers Presented at the 1990 ASME/IEEE Joint Railroad Conference (April 17–19, 1990, Chicago, IL). AEG Westinghouse Transporta- tion Systems, pp. 13–17. AREMA. 2005a. Plan 913-52. In “Portfolio of Trackwork Plans.” AREMA. AREMA. 2005b. AREMA Manual for Railway Engineer- ing. AREMA. Battelle. 1999. CRP-CD-3: Performance of Direct- Fixation Track Software: Design Guidelines and Software. Transportation Research Board, National Research Council, Washington, DC. Bilow, D. N. and D. Li. 2005. “Concrete Slab Track Test on the High Tonnage Loop at the Transportation Technology Center.” In Proceedings of the 2005 AREMA Annual Conference. AREMA. Briaud, J., R. W. James, and S. B. Hoffman. 1997. NCHRP Synthesis 234: Settlement of Bridge Ap- proaches (the Bump at the End of the Bridge). Trans- portation Research Board, National Research Coun- cil, Washington, DC. Davis, D., D. Otter, D. Li., and S. Singh. 2003. “Bridge Approach Performance in Revenue Service.” Railway Track and Structures, Vol. 99, No. 12, pp. 18–20. Frohling, R. D., H. Sheffel, and W. Ebersohn. 1995. “The Vertical Dynamic Response of a Rail Vehicle Caused by Track Stiffness Variations along the Track.” The Dynamics of Vehicles on Roads and on Tracks: Pro- ceedings of 14th IAVSD Symposium (August 21–25, 1995, Ann Arbor, MI). International Association for Vehicle Systems Dynamics. Hay, W. W. 1982. Railroad Engineering. 2d. ed. John Wiley and Sons, New York. Hoppe, E. J. 2001. “The Use, Design, and Control of Bridge Approach Slabs.” Routes and Roads, No. 312, October, pp. 24–33. Hunt, H. and M. Winkler. 1997. “Settlement of Railway Track near Bridge Abutments.” Proceedings of the Institution of Civil Engineers, Transport, Vol. 123, No. 1, February 1997, pp. 68–73. Joy, R. B., D. C. Oliva, D. E. Otter, B. E. Doe, and S. Uppal. 2001. “FAST Bridge Tests 1997–1999.” Research Report R-948. Association of American Railroads, pp. 35–44. Kentner, P., B. Brundige, J. C. Thorpe, J. Winfield, W. W. Kratville, and L. J. O’Connor (Eds.). 1997. The Car and Locomotive Cyclopedia. 6th ed. Simmon- Boardman Books, Inc. Kerr, A. D. and L. A. Bathurst. 2000. “Pads Ease Track Transitions.” Railway Track and Structures, Vol. 96, No. 8, pp. 57–64. Kerr, A. D. and B. E. Moroney. 1993. “Track Transition Problems and Remedies.” In Bulletin 742, American Railway Engineering Association, pp. 267–298. Li, D. and D. Davis. 2005. “Transition of Railroad Bridge Approaches.” Journal of Geotechnical and Geoenvironmental Engineering, Vol. 131, No. 11, pp. 1392–1398. Li, D., J. Rose, H. Lees, and D. Davis. 2001. “Hot-Mix As- phalt Trackbed Performance Evaluation at Alps, New Mexico.” In Technology Digest TD 01-015. Associa- tion of American Railroads, Transportation Technol- ogy Center, Inc. Li, D., L. Smith, B. Doe, D. Otter, and S. Uppal. 2003. “Study of Bridge Approach and Track Transition Degradation—Factors and Mitigation.” Federal Rail- road Administration Contract DTFR53-01-H-00305. Parsons Brinckerhoff Quade & Douglas, Inc. 2000. TCRP Report 57: Track Design Handbook for Light Rail Transit. Transportation Research Board, Na- tional Research Council, Washington, DC. Patel, N. and H. Jordan. 1996. “Ballasted Track Transi- tions.” In Proceedings of the 1996 Rapid Transit Con- ference of the American Public Transit Association (June 2–6, 1996, Atlanta, GA). APTA, pp. 116–126. Read, D., S. Chrismer, W. Ebersöhn, and E. Selig. 1994. “Track Modulus Measurements at the Pueblo Soft Subgrade Site.” In Transportation Research Record 1470. Transportation Research Board, National Re- search Council, Washington, DC, pp. 55–64. Redden, J. W., E. T. Selig, and A. M. Zarembski. 2002. “Stiff Track Modulus Considerations.” Railway Track and Structures, Vol. 98, No. 2, pp. 25–30. Rose, J. 1998. “Long-Term Performances, Tests, and Evaluations of Asphalt Trackbeds.” In Proceedings (AREMA) 1998 Track and Structures (September 1998, Chicago, IL). AREMA. Rose, J., L. A. Walker, and D. Li. 2002. “Heavy Haul As- phalt Underlayment Trackbeds: Pressure, Deflection, Materials, Properties, Measurements.” In Proceed- ings, Railway Engineering 2002. Engineering Tech- nics Press. Edinburgh. Sasaoka, C. D. and D. Davis. 2005. “Implementing Track Transition Solutions for Heavy Axle Load Service.” In Proceedings of the AREMA 2005 Annual Confer- ence. AREMA. Sasaoka, C. D., D. D. Davis, K. Koch, R. P. Reiff, and W. GeMeiner. 2005. “Implementing Track Transition So- lutions.” In Technology Digest TD-05-01. Association of American Railroads, Transportation Technology Center, Inc. Selig, E. T. and D. Li. 1994. “Track Modulus: Its Meaning and Factors Influencing It.” In Transportation Research Record 1470. Transportation Research Board, Na- tional Research Council, Washington DC, pp. 47–54. 36

Selig, E. T. and J. M. Waters. 1994. Track Geotechnology and Substructure Management. Thomas Telford, Ltd. Sharpe, P., R. J. Armitage, W. G. Heggie, and A. Rogers. 2002. “Innovative Design of Transition Zones.” In Proceedings, Railway Engineering 2002. Engineer- ing Technics Press. Edinburgh. Smekal, A. “Transition Structures of Railway Bridges,” Proceedings of the 4th World Congress of Railway Research (WCRR’97), Vol. B, pp. 149–158, Novem- ber 1997. Sussman, T. R. and E. T. Selig. 1998. “Track Component Contributions to Track Stiffness.” E. T. Selig, Inc. Amherst, MA. Zarembski, A. M. and J. Palese. 2003. “Transitions Elim- inate Impact at Crossing.” Railway Track and Struc- tures, Vol. 99, No. 8, pp. 28–30. 37

Transportation Research Board 500 Fifth Street, NW Washington, DC 20001 These digests are issued in order to increase awareness of research results emanating from projects in the Cooperative Research Programs (CRP). Persons wanting to pursue the project subject matter in greater depth should contact the CRP Staff, Transportation Research Board of the National Academies, 500 Fifth Street, NW, Washington, DC 20001. COPYRIGHT PERMISSION Authors herein are responsible for the authenticity of their materials and for obtaining written permissions from publishers or persons who own the copyright to any previously published or copyrighted material used herein. Cooperative Research Programs (CRP) grants permission to reproduce material in this publication for classroom and not-for-profit purposes. Permission is given with the understanding that none of the material will be used to imply TRB, AASHTO, FAA, FHWA, FMCSA, FTA, or Transit Development Corporation endorsement of a particular product, method, or practice. It is expected that those reproducing the material in this document for educational and not-for-profit uses will give appropriate acknowledgment of the source of any reprinted or reproduced material. For other uses of the material, request permission from CRP.