FLEXURAL PERFORMANCE REQUIREMENTS FOR PRESTRESSED CONCRETE TIES BY FACTORING By P. Jeffrey McQueen, P.E., Consulting Engineer Abstract The AREMA Manual for Railway Engineering, Chapter 30, Part 4, gives recommendations for the design, production and use of prestressed concrete ties in North America railroad tracks. This paper summarizes the results, conclusions and recommendations reached by AREA Committee 10, Concrete ties, subsequently by AREMA Committee 30, Part 4, Concrete Ties of the AREMA Manual. The Committees, through investigation over an extended period of time considered the track service loading as well as laboratory testing and material performance. A simplified method for the calculation of concrete tie flexural requirements is presented in order to aid the engineer in the determination of such values. Introduction * * * Since the first railway technical committees were formed in the United States in the late 1960's for the purpose of making recommendations covering the design and use of prestressed concrete ties under American railway operating conditions, a dilemma existed in respect to the appropriate loading values to be used in the determination of concrete tie structural design. From the onset, the basis for the design flexural capacity of concrete ties was approached rationally using the parameters of maximum axle load, impact factor, tie length, tie spacing and distribution factor in the calculations. By assuming the eventual ballast reaction as uniform, bending moments were calculated and tables prepared for four different tie lengths and four different tie spacings. As had been the practice in the United States, concrete ties were produced to meet the criteria and installed in numerous locations across the country. In the early days of concrete tie development, data derived from instrumented ties was used to verify the assumptions used for the determination of flexural capacity. Although the figures used for flexural capacity calculations should have produced crack-free ties, inspection of ties installed in test installations revealed that this was not the case and most ties exhibited positive moment railseat cracks within a very short span of time after installation.
Strain gauges used for data collection apparently did not give an accurate representation of the loading spectrum and it was postulated that the initial designation of an impact factor of 50% was too low and therefore it was increased, firstly to 100%, then to 150%, then to the current figure of 200%. There was usually a five-year period which elapsed between discovery of railseat positive cracking, reporting to the committees, discussion of possible causes, arrival at a consensus for increased design values and eventual approval and publishing in the American Railway Engineering s Manual for Railway Engineering, Chapter 10, now AREMA, Chapter 30, Part 4. The ten years between 1970 and 1980 saw many test installations and with the advent of Amtrak s decision to use prestressed concrete ties in the upgrading of their track in the Northeast Corridor between Washington, New York and Boston, an important stage was reached in United States concrete tie development as a definitive investigation into the cracking in the NEC was initiated by the Federal Railway Administration in conjunction with Amtrak. The study, conducted by Battelle s Columbus Laboratories, positively identified for the first time in the United States the nature and magnitude of high-amplitude impact loads resulting from a small percentage of wheel defects. In addition to the NEC study, FRA-Battelle conducted an investigation of concrete tie flexural strains and bending moments in four revenue railroads; namely, the Atchison Topeka and Santa Fe at Streator, Illinois, Amtrak and Conrail at Aberdeen, Maryland, the Chessie System at Richmond, Virginia, the Norfolk and Western at Roanoke, Virginia, and the FAST test track at the Transportation Test Center at Pueblo, Colorado. Figure 21 shows plots of railseat bending moments in concrete ties at all sites. The investigation generally involved data collection for 40,000 axle loads of traffic and showed a wide variation in railseat bending moments, related directly to vehicle wheel tread condition and, in some cases, bad bearings. In many cases, the cracking threshold of the ties, designed to meet the then current AREA recommendations, was exceeded but in the less than 0.1 percent frequency of occurrence level. Even at this low frequency of occurrence, each tie in the southern section of the NEC was subjected to loads in excess of the cracking threshold once every 1 to 2 days while in the coal-haul Norfolk and Western railroad, with its predominance of 100 ton cars, exceedance would occur as often as five times each day. It is generally well known that the FRA-Amtrak-Battelle investigations and recommendations produced criteria for strain-attenuating railseat pads having the ability to reduce railseat impacts by at least 25 percent thus reducing railseat bending moments to below the cracking threshold. In addition, the use of track side impact detectors enabled the defective wheels to be captured and corrected in the NEC. Although there has been a general move by the freight railroads toward better wheel and bearing maintenance necessitated by the destructive forces derived from 100-120 ton cars having wheel tread irregularities, high impact loads can still be expected. In the period 1980-82, AREA Committee 10 revised the recommendations of Chapter 10 in respect to flexural performance for concrete ties. Values were then shown only for one tie spacing, 24 inches and three tie lengths, 8'-3", 8'-6" and 9'-0". The reasoning behind this change was that ties of these lengths and spaced at 24 inches has been shown to perform satisfactorily if their flexural
strength met the requirements given previously by Chapter 10 of the AREA Manual for ties spaced at 30 inches. In 1982, AREA Committee 10 assigned the investigation of the influence of tie spacing and axle loads on tie design to a subcommittee. The recommendations of this subcommittee were subsequently approved by Committee 10 and published in Chapter 10 of the AREA Manual. The Factoring Method The United States operates approximately 240,000 miles of railroad track to which can be added 45,000 miles in Canada and 13,000 miles in Mexico. Thus, the main North American railway system, not including transit properties, comprises approximately 300,000 miles of standard gauge railroad track in a wide variety of terrain and environmental conditions ranging from desert to the sub-arctic. The traffic is predominantly freight with speeds up to 79 mph, but Amtrak now runs passenger trains at speeds of 150 mph in the NEC and both Canada and Mexico run passenger trains at lesser speeds. In 1982, there were upwards of 8 million concrete ties in service in North America and as their use was predicated on economic justification, they were generally chosen to be installed in heavy traffic lines most often in hill-curve territory. Thus a performance base could be established by comparing the ties subjected to the loading resulting from trains operating under a wide range of service conditions. Furthermore, as the criteria for tie performance had successively been upgraded over a 15-year period, it was possible to relate tie performance to design parameters and service conditions. The first approach of the sub-committee was to use the standard published criteria given by AREA Chapter 10 for ties of the three lengths spaced at 24 inch centers and apply factors as follows: 1. Distribution factor related to tie spacing. 2. Impact factor 200%. 3. Speed factor. 4. Wheel tread defect factor. 5. 100 ton car factor. 6. Rail condition factor. 7. Support condition track modulus factor. Each of factors 3 through 7 were assigned values of 0.9, 1.0, or 1.1. with the three values reflecting conditions of lesser influence (0.9) general influence (1.0), i.e., the conditions under which the majority of railroads operate and greater influence (1.1), due to more severe conditions. While the choice of the values appears to be somewhat simplistic, it should be noted that the general influence factor (1.0) is derived from the satisfactory performance of ties designed to meet AREA criteria under the severe North American railroad operating conditions. As there were five factors which required a choice by judgment on the part of the engineer, a selection of one factor low (or high) would only have a marginal effect on the design criteria. Moreover, estimates generally would be expected to be on the high side due to the trend toward increasing car loads. After much discussion, Committee 10 rejected the five factors as being too subjective and
requested further study by the sub-committee. On reflection it was decided to limit the factors to the two conditions which the railroad engineer knows well. They are simply, annual gross tonnage and speed. As previously noted, the FRA-Battelle correlation study recorded strains from instrumented ties in four revenue service railroads and the FAST track at Pueblo and, therefore, gave a good representation of the loading spectrum. The strain readings enabled the calculation of tie bending moments. The sub-committee then proposed that the collected data be used as a guide for the preparation of a factoring chart, see Figures 30-4-3 and 30-4-4 of the current AREMA Manual, Chapter 30, Part 4. While it may be argued that speed and tonnage factors probably are not linear, sample values taken directly from concrete ties in service in the United States, Canada and Mexico were used to validate the charts which are used as follows. From Figure 30-4-4, the positive bending moment at the railseat is read directly for a particular tie length and spacing, for example, an 8'-6" tie spaced at 24 inches would have a 300 inch kip positive moment flexural performance requirement. Then, by the use of Figure 1.4.1.2, speed and tonnage factors are obtained. As an example, for a maximum speed of 40 mph, the speed factor is 0.8 and a maximum tonnage of 60 MGT, the tonnage factor would be 1.0. The resulting design positive moment flexural capacity would then be: 300 x 0.8 x 1.0 = 240 inch kips. For the other tie flexural requirements, railseat negative and tie center, positive and negative, ratios related to the positive railseat moment are used, based on previously recorded data to calculate the required values. Why use factors? As an approximation, for every 50 inch kips of positive moment flexural capacity under or over 300 inch kips, the cost of a concrete tie in the United States will vary by about 10% of the base cost not including hardware components, i.e., about $4.00 per tie. The economic justification is therefore apparent. Under U.S. operating conditions which will include transient loading peaks, some cracking at the railseat must be expected. On the other hand, many thousands of ties installed in U.S. railroads designed to lower loading criteria, are performing in a satisfactory manner with some having over 1,000 million gross tons of traffic Even though cracked, the reason that they perform satisfactorily was the requirement in Chapter 10 of the AREA manual and now Chapter 30 of the AREMA Manual, that all designs to be qualified must be subjected to three million cycles of test loading in a laboratory at a 10% overload in the cracked condition. Ties meeting this requirement, therefore, demonstrate specific fatigue strength and tendon bond capacity. It should be noted that the Figures quoted in the AREMA Manual already have had some factors exceeded by the heavy-haul railroads and Amtrak. If deemed appropriate, i.e., by service investigations which employ strain gauges/wheel impact detectors, the speed and tonnages factors may easily be adjusted. In a like manner, many transit properties are using the figures to arrive at their own flexural design criteria by proportioning their axle loads, tonnages and speeds. It is not to be assumed that the AREMA Committee is giving approval for the use of cracked
ties, but rather to accept the fact that some cracking may occur under peak loads due to wheel and rail anomalies. These cracks need not be detrimental to tie performance and life so long as they are recognized and accommodated within the concrete-tendon-bond system and peak moments reduced by a strain attenuating, flexible rail fastening system. The factoring method allows savings to be applied to tie design due to variations in the site-specific loading spectrum, rather than having only one tie design based on the maximum loading conditions for the whole country. MCQ-710-10