National Academies Press: OpenBook

Criteria for Restoration of Longitudinal Barriers, Phase II (2021)

Chapter: Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity

« Previous: Chapter 13. Anchor Strength Quantified in Terms of Anchor Damage
Page 375
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 375
Page 376
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 376
Page 377
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 377
Page 378
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 378
Page 379
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 379
Page 380
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 380
Page 381
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 381
Page 382
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 382
Page 383
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 383
Page 384
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 384
Page 385
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 385
Page 386
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 386
Page 387
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 387
Page 388
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 388
Page 389
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 389
Page 390
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 390
Page 391
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 391
Page 392
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 392
Page 393
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 393
Page 394
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 394
Page 395
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 395
Page 396
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 396
Page 397
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 397
Page 398
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 398
Page 399
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 399
Page 400
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 400
Page 401
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 401
Page 402
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 402
Page 403
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 403
Page 404
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 404
Page 405
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 405
Page 406
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 406
Page 407
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 407
Page 408
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 408
Page 409
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 409
Page 410
Suggested Citation:"Chapter 14. Effects of W-Beam Splice Damage on Rail Capacity." National Academies of Sciences, Engineering, and Medicine. 2021. Criteria for Restoration of Longitudinal Barriers, Phase II. Washington, DC: The National Academies Press. doi: 10.17226/26321.
×
Page 410

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.

375 CHAPTER 14 – EFFECTS OF W-BEAM SPLICE DAMAGE ON RAIL CAPACITY The effects of crash-induced damage on the residual capacity of w-beam splices were evaluated using physical impact testing. In Report 656 a single damage mode was evaluated for the w-beam splice, which involved a “cut out” of w-beam material around the lower upstream splice bolt. From review of full-scale tests (see Chapter 3) it was determined that a more critical location for splice rupture may be at the downstream splice-bolts. The goal of this study was to supplement the splice damage assessment criteria in Report 656 by quantifying the effects of additional damage modes on the capacity of the splice connection. The most difficult challenge of this task was in determining how to measure the degree of damage for many of the damage modes, particularly in regards to field assessments. There are several obvious damage modes, such as a missing bolts, that are both easy to assess in the field and also easy to quantify; however, damage modes for splice connections are usually more complicated and are more commonly caused by high stress concentrations in the splice connection and small tears/cuts in the w-beam caused by the w-beam contacting other components with relatively sharp edges during crash events. Since small tears in the splice bolt holes are likely to be hidden by the bolt head and/or nut, the degree of damage must be quantified by parameters such as degree of flattening, angle of bend in the splice, measure of separation between the two w-beam elements at the upstream and downstream ends of the splice, noticeable indentation/gouging of bolt head into w-beam (e.g., rotation of splice bolt), and slip of bolt in splice-bolt-hole. Most of these damage modes would be difficult to fabricate in the laboratory; therefore, damaged sections of w-beam from crash sites in the state of Maine were collected and assessed for use in this study. Research Approach The basic research approach involved: 1. Identifying and measuring the visible damage modes from a number of crash- damaged splice connections from field installations, 2. Correlating the measurable visible damage to the potential for crack initiation in the splice-bolt holes via disassembly and non-destructive inspection techniques (e.g., ultrasonic or magnetic particle) (those specimens selected for pendulum testing were not disassembled), 3. Use pendulum testing to measure the reduction in splice capacity due to various damage modes and levels of damage for a select number of damaged splices, and 4. Associate various splice damage modes with “priority for repair” for inclusion in a Field Guide to assist DOT maintenance personnel in making decisions about repairing damaged guardrail installations.

376 Procurement and Assessment of Damaged W-Beam Splices A total of fifteen damaged splice samples were provided to the study by the Maine Department of Transportation (MEDOT); however, a number of those did not have sufficient damage for consideration in the test program. The damaged splice samples were extracted from various crash-damaged field installations in Oxford County, Maine. Most of the damaged articles were delivered fully-assembled to nearby MEDOT maintenance facilities. The research team traveled to these collection-sites to identify and measure the various damages for each specimen. Inspection of Fully Assembled Splice (General Field-Type Measurements) The first phase of this assessment involved measurement-procedures that could readily be performed by maintenance personnel at a crash site and included: 1. Percent Rail Flattening/Crush – The condition of rail flattening was measured at each of the four quadrants of the splice connection, since it was not known how deformation at each of these locations would affect capacity. In particular, measurements were taken at the: a. Downstream, top section – The distance from the center of the w-beam to the top edge was measured from the back-side of the w-beam splice at the downstream splice bolts, as shown in Figure 325. b. Downstream, bottom section – The distance from the center of the w- beam to the bottom edge was measured from the back-side of the w-beam splice at the downstream splice bolts. c. Upstream, top section – Measurement was taken in the same manner as described above at the upstream splice bolts on the upper half of the w- beam splice. d. Upstream, bottom section – Measurement was taken in the same manner as described above at the upstream splice bolts on the lower half of the w- beam splice.  The degree of flattening was then computed using the following equation: 𝑃𝑒𝑟𝑐𝑒𝑛𝑡 𝑅𝑎𝑖𝑙 𝐹𝑙𝑎𝑡𝑡𝑒𝑛𝑖𝑛𝑔 = 𝑥 − 𝑥0 𝑥𝑓𝑙𝑎𝑡 − 𝑥0 ∗ 100  Where x is the deformed ½-cross-sectional height from the center of the w-beam to the top-edge of the rail measured from the back-side of the w- beam splice; x0 is the undeformed ½-cross-section height taken as 6.125 inches; xflat is the ½-cross-section height of the w-beam when completely flattened taken as 9.6 inches.  For cases where the value of x was less than x0 (i.e., vertical crushing of the splice), the degree of crush was computed as the percent difference of the deformed height (x) to that of the undeformed height (x0) using the following relationship: 𝑃𝑒𝑟𝑐𝑒𝑛𝑡 𝑅𝑎𝑖𝑙 𝐶𝑟𝑢𝑠ℎ = 𝑥 − 𝑥0 𝑥0 ∗ 100

377 2. Rail Separation – Two types of rail separation were measured: a. The lateral gap opening at the upstream edge of the splice (traffic face of rail) and at the downstream edge of the splice (back face of rail). The measurements were taken at eight points, adjacent to the splice-bolt locations, as illustrated in Figure 326. b. The longitudinal slip in the splice – taken as the distance measured from the original position of the rail (discernible in most cases) to the current position, as illustrated in Figure 327. 3. Condition of Splice Bolts – The condition for each splice bolt was recorded as missing, loose, gouging, or torn-through. The term gouging refers to the condition of a splice-bolt rotated about the vertical axis of the rail, where the edge of the bolt-head is pressed into the w-beam panel. This damage mode generally results from high tensile load in the rail during a crash event. An example of “light gouging” is shown in Figure 328 evidenced by visible rotation of the splice-bolt. 4. Condition of Splice Bolt Holes – Each splice bolt hole was inspected for visible stretching and tearing. Figure 329 shows an example of a splice bolt hole with evident stretching (resulting from high tension loading in the rail during crash) which was measured by taking the longitudinal distance from the edge of the splice bolt to its apparent original position. In some cases the only visual cue to indicate that a bolt hole is “stretched” is when the edge of the bolt hole is visible beside the bolt head. 5. Condition of Post Bolt Holes – Each post bolt hole was inspected for visible damages which included (listed in order of increasing severity) stretching, horizontal tears, and vertical tears. Figure 330 shows an example of a splice bolt hole with evident stretching and a horizontal tear. In this case, the worse condition was recorded, which was the horizontal tear. 6. Vertical and Horizontal Tears – When a tear was evident within the splice region (i.e., region encompassing the out-to-out distance from the ends of the rail sections in the splice) the length of the tear was measured and reported in the appropriate category. The possible categories include tears in the upstream or downstream panels at the top section, bottom section, middle section, top edge or bottom edge, as illustrated in Figure 331. For tears that intersect the top or bottom edge of the rail, in which case the tear locations would also fall into the category of top section or bottom section, only the “edge” category was reported - since edge tears in general have been shown to be the more critical condition.[Gabler10]

378 Figure 325. Rail flattening – Rail flattening and rail crush at each quadrant of the splice is measured from the center of the w-beam to the top/bottom edge at the splice bolts. Figure 326. Rail separation - Lateral gap between rail elements at splice connection (measurements taken at locations adjacent to the eight splice bolts).

379 Figure 327. Rail separation – Longitudinal slip in splice connection. Figure 328. Gouging – Gouging of splice-bolts into w-beam panel is evidenced by rotation of the bolt about the vertical axis. Final Longitudinal Position Original Longitudinal Position

380 Figure 329. Splice bolt hole stretching – Stretching of the splice bolt hole is evidenced in this photo by the apparent longitudinal movement of the splice bolt relative to its original position. Figure 330. Horizontal tear in post bolt hole – This photo shows evidence of both stretching and a horizontal tear in the post bolt hole.

381 Figure 331. Horizontal and vertical tears – tears were measured and categorized according to location on upstream and downstream panel as denoted here. A damage summary sheet was created for each damaged splice specimen (see Appendix R). The summary sheets include damage measurements as described above and photos of the damaged splices. When there were no measurable damages for a particular damage mode, the corresponding damage-field was left blank in the assessment table. Tables 81 through 85 show the results of the damage measurements. Each table focuses on a single damage mode and lists the specimens in descending order with respect to the maximum level of the corresponding damage. Since splice damage generally involves multiple damage modes, each damage-mode table also includes a summary of all other damages identified for reference. Regarding the physical test program, there were not a sufficient number of specimens with a single mode of damage for each damage mode case, which made it difficult to accurately quantify the effect of each isolated damage mode on capacity of the splice. For example, there was some amount of rail flattening, rail crush or both in essentially all damage splice cases, as shown in Table 81. So, when evaluating the effects of any specific damage mode on the capacity of the splice, the response was also influenced by the effects of any other existing damages, such as flattening/crushing. It also appears that some of the damage modes are strongly linked with others. For example, from Table 83, high levels of rail separation generally occur in combination with high levels of rail flattening and also gouging of the splice bolts into the w-beam (e.g., compare columns for “Max Separ.”, “Rail Flattening” and “Gouging” in Table 83). There were no apparent correlations between rail tears (i.e., non-hole tears) and other splice damage modes (at least from the damaged specimens collected in this study). From visual inspection, these horizontal and vertical tears in the splice were likely caused by a part on the vehicle cutting into the rail during the crash (see Figure 331 for example). Splice Region Upper Section Bottom Section Downstream Panel Upstream PanelMiddle Section Top Edge Bottom Edge 4.5 in tear

382 Table 81. Summary of “rail flattening” damage mode for MEDOT crash-damaged splice specimens – listed in descending order, w.r.t maximum degree of flattening. Long. Lateral Loose Missing Gouging Torn Stretched Torn Vertical Horizontal 4A5-ME010 97.1% -2.0% 86.3% -4.1% 97.1% 0.20 0.55 * 4A5-ME006 89.9% 68.3% 97.1% 82.7% 97.1% N.A. N.A. N.A. N.A. 4A5-ME013 93.5% 3.6% 89.9% 3.6% 93.5% 0.20 0.35 * 11.8 4A5-ME008 86.3% 10.8% 75.5% 10.8% 86.3% 0.20 0.28 4A5-ME011 82.7% 10.8% 68.3% 18.0% 82.7% N.A. N.A. N.A. N.A. 4A5-ME007 75.5% 61.2% 82.7% 61.2% 82.7% 0.50 * 4A5-ME001 68.3% 3.6% 68.3% -14.3% 68.3% 0.20 0.13 * * 4A5-ME005 18.0% 68.3% 10.8% 32.4% 68.3% 0.12 0.25 4.5 4A5-ME017 32.4% 68.3% 46.8% 61.2% 68.3% 0.25 0.25 * 22.5 4A5-ME003 61.2% 10.8% 25.2% -2.0% 61.2% 0.25 1.0 4A5-ME018 46.8% 39.6% 39.6% 32.4% 46.8% 0.25 0.50 * * 4A5-ME002 25.2% 3.6% 3.6% 39.6% 39.6% N.A. N.A. N.A. N.A. * 5.0 4A5-ME014 25.2% 7.3% 25.2% -8.7% 25.2% 0.28 * 11.8 4A5-ME004 -26.5% -30.6% -30.6% 3.6% 3.6% 0.25 4A5-ME015 0.0% 0.0% 0.0% -4.1% * Other Damages Rail Separation (in) Post Bolt Hole Tears (in)Splice Bolts and Hole Max Deg. of Flattening % Rail Flattening / Crush Specimen No. @ upstream bolts @ downstream bolts Top Section Bottom Section Top Section Bottom Section

383 Table 82. Summary of “rail crush” damage mode for MEDOT crash-damaged splice specimens – listed in descending order, w.r.t maximum degree of crush. Long. Lateral Loose Missing Gouging Torn Stretched Torn Vertical Horizontal 4A5-ME004 -26.5% -30.6% -30.6% 3.6% -30.6% 0.250 4A5-ME001 68.3% 3.6% 68.3% -14.3% -14.3% 0.197 0.125 * * 4A5-ME014 25.2% 7.3% 25.2% -8.7% -8.7% 0.276 * 11.8 4A5-ME010 97.1% -2.0% 86.3% -4.1% -4.1% 0.197 0.551 * 4A5-ME015 0.0% 0.0% 0.0% -4.1% -4.1% * 4A5-ME003 61.2% 10.8% 25.2% -2.0% -2.0% 0.250 1.0 4A5-ME002 25.2% 3.6% 3.6% 39.6% N.A. N.A. N.A. N.A. * 5.0 4A5-ME005 18.0% 68.3% 10.8% 32.4% 0.118 0.250 4.5 4A5-ME006 89.9% 68.3% 97.1% 82.7% N.A. N.A. N.A. N.A. 4A5-ME007 75.5% 61.2% 82.7% 61.2% 0.500 * 4A5-ME008 86.3% 10.8% 75.5% 10.8% 0.197 0.276 4A5-ME011 82.7% 10.8% 68.3% 18.0% N.A. N.A. N.A. N.A. 4A5-ME013 93.5% 3.6% 89.9% 3.6% 0.197 0.354 * 11.8 4A5-ME017 32.4% 68.3% 46.8% 61.2% 0.250 0.250 * 22.5 4A5-ME018 46.8% 39.6% 39.6% 32.4% 0.250 0.500 * * @ upstream bolts @ downstream bolts Other Damages Specimen No. Max Deg. of Crush Top Section Bottom Section Top Section Bottom Section Tears (in)Splice Bolts and Hole % Rail Flattening / Crush Rail Separation (in) Post Bolt Hole

384 Table 83. Summary of “rail separation” damage mode for MEDOT crash-damaged splice specimens – listed in descending order, w.r.t maximum separation. @ bolt1 @ bolt2 @ bolt3 @ bolt4 @ bolt5 @ bolt6 @ bolt7 @ bolt8 Loose Missing Gouging Torn Stretched Torn Vertical Horizontal 4A5-ME010 0.236 0.551 0.157 0.118 0.551 0.197 97% -4% * 4A5-ME007 0.5 0.3125 0.25 0.500 83% * 4A5-ME018 0.5 0.5 0.500 0.250 47% * * 4A5-ME013 0.276 0.354 0.157 0.354 0.197 94% * 11.8 4A5-ME008 0.276 0.118 0.276 0.197 86% 4A5-ME003 0.25 0.25 0.250 61% -2% 1.0 4A5-ME004 0.25 0.250 4% -31% 4A5-ME005 0.25 0.250 0.118 68% 4.5 4A5-ME017 0.25 0.25 0.25 0.250 0.250 68% * 22.5 4A5-ME001 0.125 0.125 0.197 68% -14% * * 4A5-ME002 N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. 0.000 N.A. 40% N.A. N.A. * 5.0 4A5-ME006 N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. 0.000 N.A. 97% N.A. N.A. 4A5-ME011 N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. 0.000 N.A. 83% N.A. N.A. 4A5-ME014 0.000 0.276 25% -9% * 11.8 4A5-ME015 0.000 0% -4% * Post Bolt Hole Specimen No. Max. Separ. (in) Rail Separation Tears (in) Gap at Downstream End (in) Gap at Upstream End (in) Rail Flattening (%) Rail Crush (%) Long. Slip (in) Other Damages Splice Bolts and Hole

385 Table 84. Summary of “visible splice bolt and splice bolt hole” damage mode for MEDOT crash-damaged splice specimens. #1 #2 #3 #4 #5 #6 #7 #8 #1 #2 #3 #4 #5 #6 #7 #8 #1 #2 #3 #4 #5 #6 #7 #8 #1 #2 #3 #4 #5 #6 #7 #8 Long. (in) Lateral (in) H-Tear V-Tear Vertical (in) Horizontal (in) 4A5-ME001 * * * 68% -14% 0.20 0.13 * 4A5-ME007 * * * 83% 0.50 4A5-ME010 * * 97% -4% 0.20 0.55 4A5-ME013 * * * * 94% 0.20 0.35 11.8 4A5-ME014 * 25% -9% 0.28 11.8 4A5-ME015 * 0% -4% 4A5-ME003 61% -2% 0.25 1.0 4A5-ME004 4% -31% 0.25 4A5-ME005 68% 0.12 0.25 4.5 4A5-ME008 86% 0.20 0.28 4A5-ME002 - - - - - - - - - - - - - - - - * 40% N.A. N.A. 5.0 4A5-ME006 - - - - - - - - - - - - - - - - 97% N.A. N.A. 4A5-ME011 - - - - - - - - - - - - - - - - 83% N.A. N.A. 4A5-ME018 * * * * * 47% 0.25 0.50 4A5-ME017 * * * * * 68% 0.25 0.25 22.5 Post Bolt Hole Non Bolt-Hole Tears Specimen No. Visible Splice Bolt and Splice-Bolt Hole Damage Other Damages Loose Bolts Missing Bolts Visible Gouging Visible Tear Rail Flattening (%) Rail Crush (%) Rail Separation

386 Table 85. Summary of “non-bolt-hole tear” damage mode for MEDOT crash-damaged splice specimens – listed in descending order w.r.t maximum vertical tear, then maximum horizontal tear. Top edge (in) Bot. edge (in) Mid. Sect. (in) Top Sect. (in) Bot. Sect. (in) Top edge (in) Bot. edge (in) Top Sect (in) Bot. Sect. (in) Mid. Sect (in) Long. (in) Lateral (in) Loose Missing Gouging Hole Torn H-Tear (in) V-Tear (in) 4A5-ME003 1.0 1.0 0.0 61% -2% 0.25 4A5-ME017 0.0 22.5 22.5 68% 0.25 0.25 * 4A5-ME013 0.0 11.8 11.8 94% 0.20 0.35 * 4A5-ME014 0.0 11.8 11.8 25% -9% 0.28 * 4A5-ME002 0.0 5.0 5.0 40% N.A. N.A. N.A. N.A. * 4A5-ME005 0.0 4.5 4.5 68% 0.12 0.25 4A5-ME001 0.0 0.0 68% -14% 0.20 0.13 * * 4A5-ME004 0.0 0.0 4% -31% 0.25 4A5-ME006 0.0 0.0 97% N.A. N.A. N.A. N.A. 4A5-ME007 0.0 0.0 83% 0.50 * 4A5-ME008 0.0 0.0 86% 0.20 0.28 4A5-ME010 0.0 0.0 97% -4% 0.20 0.55 * 4A5-ME011 0.0 0.0 83% N.A. N.A. N.A. N.A. 4A5-ME015 0.0 0.0 0% -4% * 4A5-ME012 4A5-ME016 4A5-ME018 * * Splice Bolts and Hole Other Damages Rail Separation Post Bolt Hole Specimen No. Vertical Tears Horzontal Tears Rail Flattening (%) Rail Crush (%) Max. Vertical (in) Max. Horz. (in) Non-Bolt Hole Tears

387 Inspection of Unassembled Splice (for Hidden Damages in the Splice-Holes) The next step of the research approach involved disassembly of the damaged splice samples in order to assess damages to the splice-bolt holes. These areas cannot be directly inspected in the field since they are hidden by the splice-bolt head and nut. The purpose of this task was to correlate visible, measurable damages to the potential for crack initiation in the splice-bolt holes. The intent was to first select specimens for use in the pendulum test program. Those specimens would not be disassembled for inspection because it was highly unlikely that the damaged splice could be reassembled exactly to its original state, which included residual stress in the splice restrained by the bolted connection. Once the specimens for the pendulum test program were selected, the remaining specimens would then be disassembled and inspected for damages around the edges of the splice-bolt holes via non-destructive inspection techniques. Any damage mode, as well as the level of damage for that mode, that resulted in notable damage to the splice-bolt holes would be determined. The potential for crack initiation in the splice-bolt holes could then be correlated to the various types of measurable damage modes. It was decided, however, that this task would not be performed due to the low number of samples collected for the study. Instead, all available damaged splice specimens were used for the pendulum test program. Physical Testing Equipment and Instrumentation Pendulum Device The striker used in the tests was a 4,360-lb concrete pendulum with a semi-rigid nose, which is shown in Figure 332. The semi-rigid nose was developed by researchers from Virginia Tech during the first phase of this study and was fabricated from a wooden block and covered with sheet metal.[Gabler10] The radius of chamfer at the center of the impactor face was 6 inches, which was based on measurements of a 2006 Chevrolet 1500 pickup truck. [Gabler10] Figure 332. 4,360-lb pendulum device with semi-rigid nose.

388 Accelerometers The pendulum was instrumented with three accelerometers mounted onto the backside of the pendulum mass. Accelerometers 1 and 3 recorded data in the x-direction (forward direction) and Accelerometer 2 recorded data in the z-direction (vertical direction). Figure 333 provides a schematic showing the locations of the accelerometers. Figure 333. Schematic of the accelerometer instrumentation for the pendulum tests. Photography Cameras The tests were also recorded using five high-speed cameras with an operating speed of 500 frames per second and two digital video cameras ( Ì´ 60 fps). Figure 334 provides the specifications and the general placement of the high-speed cameras for the tests. The accelerometers and the high-speed video were triggered upon impact using pressure tape switches when the pendulum contacted the post. The test setup and results were also documented with pre- and post-test photographs.

389 Figure 334. High-speed camera specifications and placement. Test Setup The original plan was to use the same test setup that Gabler et al used for testing the damaged splice in Phase I.[Gabler10] After critical review of the test videos from Gabler’s study, it was determined that the test setup tended to result in a tensile rupture of the splice, which is not generally the mode of failure mode witnessed in full-scale tests. Because of the short length of rail and the “rigid” constraint on the boundaries in Gabler’s study, the rail developed very high tensile forces at relatively low rail deflections. Figure 335 shows sequential views extracted from the high-speed video in one of Gabler’s tests. The upper image shows the test article immediately before splice rupture and the lower image is the test article shortly after rupture.

390 Figure 335. Sequential views of a pendulum test performed in Gabler’s study. Unfortunately, the length of rail could not be extended beyond these limits due to the limited space at the test site. One alternative would be to design the boundary conditions to mimic the longitudinal and lateral stiffness behavior of an extended guardrail, but such an effort was beyond the scope of this study. It was therefore decided to modify the test setup such that the rail would bend at the post (typical of a normal pocket that develops in a crash event), as illustrated in Figure 336. Figure 336. Pocketing during full-scale crash test C08C3-27.2.[Fleck08b] The test set up is shown in Figure 337. The test article was a 13-ft long section of w- beam rail with a w-beam splice located at 65 inches from the downstream end. Each end of the rail was constrained from longitudinal displacement using two 0.75-inch diameter cables (i.e., AASHTO-AGC-ARTBA FCA01) fastened onto the ends of the rail. On the downstream end, the cables were fastened to two standard cable anchor brackets (i.e., AASHTO-AGC-ARTBA

391 FPA01). Each bracket was bolted onto the rail using eight ½-inch bolts and nuts with 1-1/16 inch diameter washers under the bolt heads and two 2.5 x 15 x ¼-inch steel bearing plates under the nuts, as shown in Figure 338. On the upstream end, the cables were fastened to the rail by welding three modified anchor cable brackets directly to the end of the w-beam, as shown in Figure 339. The two legs of each of the cable anchor brackets were removed and a continuous weld along the top and bottom side of each bracket was used to fasten the bracket to the rail. The test setup involved a single W6x16 structural steel post with wood blockout installed downstream of the impact location at the splice connection. There was no post installed upstream of the impact point in order to permit the maximum amount of deflection prior to developing full tension in the rail. The rigid W6x16 post was used to limit the amount of deflection of the post and force the bending mode of the splice around the post. The post was 72 inches long and was embedded 44 inches in the soil. The 6x8x12 inch routed wood blockout was used to separate the w-beam rail from the post, and a standard 10-inch long 5/8-inch diameter bolt and nut (i.e., AASHTO-AGC-ARTBA FBB03) was used to fasten the rail to the blockout and post. The post- bolt was positioned at the downstream end of the slotted hole in the w-beam to emulate the typical position of the bolt resulting from an impact upstream of the splice, as shown in Figure 340. The rail height was 28 inches measured from the ground to the top of the rail. The soil for all tests conformed to Grading B of AASHTO M147-95 and was compacted in 6-inch lifts using a pneumatic tamper. The density, moisture content and degree of compaction of the soil was measured in front of and behind the post after each compaction process using a Troxler-Model 3440 Surface Moisture-Density Gauge. There were a total of twelve readings which were averaged to determine the effective soil conditions. Figure 337. Test set-up for Test Series 14004.

392 Figure 338. Anchoring of the downstream end of the rail for Test Series 14004. Figure 339. Anchoring of the upstream end of the rail for Test Series 14004. Figure 340. Typical mounting position for post-bolt was on the downstream side of the slotted hole on the rail.

393 Impact Conditions The 4,360-lb pendulum struck the w-beam rail at 37.5 inches upstream of the splice connection. The nominal impact speed was 20.5 mph, which resulted in 734 kip-in of kinetic energy for the striker. Several preliminary tests were performed to determine the appropriate impact and boundary conditions for the study. The results of those preliminary tests are included in the test summary sheets in Appendix S as Tests 14004A-E. Note that the test conditions for Test 14004E were consistent with those used in the primary test program, except that the post- bolt was positioned at the upstream end of the slotted hole in the rail. This test involved a new, undamaged rail splice connection, but resulted in relatively low capacity. Since it could not be determined if the position of the post bolt affected the results, the test was excluded from the study. Scope Three damage modes were investigated: (1) rail flattening, (2) longitudinal slip in the splice, and (3) splice separation (gap between panels at the downstream splice bolts). There were a total of fifteen tests, including the five preliminary tests performed using “spare” test specimens to help finalize the test setup and impact conditions. The complete test matrix is shown in Table 86. Because of the general nature of crash damage, most of the test specimens contained multiple damage modes. The primary damage mode evaluated for each test case is shown in Table 86. In most cases the primary damage mode corresponded to the predominate damage mode observed for that case. Of the ten primary tests (Tests 14004F-O), two were performed on splices with significant rail flattening, two were performed on splices with notable longitudinal slip in the splice connection, one was performed on a splice with significant rail crush, two were performed on splices with significant lateral separation of the w-beam rails at the downstream splice bolts, and two were performed on undamaged (new) splices. One supplemental test (i.e., Test 14004M) was also performed which involved a damaged splice with a combination of three damage modes with notable damage levels (i.e., 0.35 inches of splice separation, 0.2 inch of longitudinal slip, and 93.5% flattening).

394 Table 86. Pendulum test matrix for the splice damage study. Mass (kips) Velocity (mph) Energy (kip-in) 14004M 4A5-ME013 Rail Flattening Long. Slip Splice Separation Horizontal Tear Splice-Bolts Gouging 94% 0.2 inches 0.35 inches 11.81 in. 4 locations 4.36 20.5 734 14004G 4A5-ME011 Rail Flattening 83% 4.36 20.5 734 14004F 4A5-ME001 Rail Flattening Rail Crush Long. Slip Splice Separation Loose Splice-Bolts 68% 14% 0.20 inches 0.13 inches 3 locations 4.36 20.5 734 14004H 4A5-ME014 Long. Slip Rail Flattening Rail Crush Loose Splice-Bolts 0.28 inches 25% 9% 1 location 4.36 20.5 734 14004I 4A5-ME017 Long. Slip Rail Flattening Horizontal Tear Splice-Bolts Gouging 0.25 inches 68% 22.8 inches 5 locations 4.36 20.5 734 Rail Crush 14004J 4A5-ME004 Rail Crush Splice Separation 31% 0.25 inches 4.36 20.5 734 14004K 4A5-ME010 Splice Separation Rail Flattening Rail Crush Long. Slip Splice-Bolts Gouging 0.55 inch 97% 4% 0.2 inches 2 locations 4.36 20.5 734 14004L 4A5-ME007 Splice Separation Rail Flattening Splice-Bolts Gouging 0.5 inch 83% 3 locations 4.36 20.5 734 14004N New02 - - 4.36 20.5 734 14004O New03 - - 4.36 20.5 734 Undamaged Target Impact Conditions Long. Slip in Splice Lateral Gap in Splice Primary Damage Mode Complete Damage Summary Rail Flattening Specimen #Test # Damage Level

395 Results A summary of the test results for Test Series 14004 are shown in Table 87, including damage mode case, damage level, impact velocity, failure mode, peak force during the impact, and the maximum energy absorbed by the guardrail system up to the point of rail rupture. The resulting peak impact force and peak energy for each case is also shown graphically in Figures 341 and 342. The impact force and energy vs. rail deflection for the various damage mode cases is compared to the results for the undamaged rail in Figures 343 through 346. Sequential views of the tests are shown in Figures 347 through 351. The individual test-summary sheets are included in Appendix S. Table 87. Results of pendulum test series 14004. The force, energy and displacement values were back-calculated using the accelerometer data from the impacting mass, and thus relate indirectly to the lateral motion of the rail at the point of impact. The force values reported herein should not be interpreted as tensile forces in the splice. Also, the reported energy values are a function of deformations within the overall guardrail test section, which include not only the deformations of the rail and splice, but also the energy dissipated through deformations of the post and soil. Impact Velocity (mph) Energy (kip-in) Failure Mode Max Force (kips) Max Energy (kip/in) 14004M 4A5-ME013 Rail Flattening Long. Slip Splice Separation Horizontal Tear Splice-Bolts Gouging 94% 0.2 inches 0.35 inches 11.81 in. 4 locations 20.6 741.6 Splice-Bolt Tear -Out 48.2 503 14004G 4A5-ME011 Rail Flattening 83% 21.2 734.0 Splice-Bolt Tear -Out 45.4 550 14004F 4A5-ME001 Rail Flattening Rail Crush Long. Slip Splice Separation Loose Splice-Bolts 68% 14% 0.20 inches 0.13 inches 3 locations 20.2 734.0 Rail Tear 43.2 437 14004H 4A5-ME014 Long. Slip Rail Flattening Rail Crush Loose Splice-Bolts 0.28 inches 25% 9% 1 location 19.7 734.0 Splice-Bolt Tear -Out 41.7 443 14004I 4A5-ME017 Long. Slip Rail Flattening Horizontal Tear Splice-Bolts Gouging 0.25 inches 68% 22.8 inches 5 locations 20.3 734.0 Boundary Failure 50.1 542 Rail Crush 14004J 4A5-ME004 Rail Crush Splice Separation 31% 0.25 inches 20.2 734.0 Boundary Failure 52.1 561 14004K 4A5-ME010 Splice Separation Rail Flattening Rail Crush Long. Slip Splice-Bolts Gouging 0.55 inch 97% 4% 0.2 inches 2 locations 20.5 734.0 Splice-Bolt Tear -Out 39.5 501 14004L 4A5-ME007 Splice Separation Rail Flattening Splice-Bolts Gouging 0.5 inch 83% 3 locations 20.6 734.0 Splice-Bolt Tear -Out 54.6 512 14004N New02 - - 20.4 734.0 Splice-Bolt Tear -Out 49.9 573 14004O New03 - - 20.6 734.0 Splice-Bolt Tear -Out 46.1 520 ResultsPrimary Damage Mode Test # Specimen # Complete Damage Summary Damage Level Rail Flattening Long. Slip in Splice Lateral Gap in Splice Undamaged I p t Conditions

396 Figure 341. Peak impact force for each damage mode case investigated in test series 14004F-O. Figure 342. Peak impact energy for each damage mode case investigated in test series 14004F-O.  N OFG H I J K L M H I J K L N O F G M

397 Figure 343. (a) Force vs. Deflection and (b) Energy vs. Deflection curves for damaged splices with flattened cross-section compared to undamaged cases. 14004F 14004G 14004M (b) (a) Undamaged - 94% flattened - 68% flattened - 83% flattened

398 Figure 344. (a) Force vs. Deflection and (b) Energy vs. Deflection curves for damaged splices with longitudinal slip compared to undamaged cases. (b) (a) Undamaged - 0.28 inches slip - 0.25 inches slip 14004H 14004I

399 Figure 345. (a) Force vs. Deflection and (b) Energy vs. Deflection curves for damaged splice with vertical crush compared to undamaged cases. (b) (a) Undamaged - 31% crush 14004J

400 Figure 346. (a) Force vs. Deflection and (b) Energy vs. Deflection curves for damaged splice with combination of lateral rail separation and flattened rail compared to undamaged cases. (b) (a) Undamaged 14004K 14004L - 0.55” lat. separation - 97% flattened - 0.50” lat. separation - 83% flattened

401 Figure 347. Sequential views of Tests 14004F and 14004G. 0.072 sec 0.018 sec 0.036 sec 0.054 sec 0.090 sec Test 14004F Test 14004G

402 Figure 348. Sequential views of Tests 14004H and 14004I. 0.072 sec 0.018 sec 0.036 sec 0.054 sec 0.090 sec Test 14004H Test 14004I (rail tear at left boundary)

403 Figure 349. Sequential views of Tests 14004J and 14004K. 0.072 sec 0.018 sec 0.036 sec 0.054 sec Test 14004J (cable rupture at left boundary) 0.090 sec Test 14004K

404 Figure 350. Sequential views of Tests 14004L and 14004M. 0.072 sec 0.018 sec 0.036 sec 0.054 sec 0.090 sec Test 14004L Test 14004M

405 Figure 351. Sequential views of Tests 14004N and 14004O. 0.072 sec 0.018 sec 0.036 sec 0.054 sec 0.090 sec Test 14004N Test 14004O

406 Discussion of Test Results Pendulum impact tests were performed to evaluate the effects of crash-damage on the capacity of w-beam rail splices. For standard strong-post guardrail systems, such as the modified G4(1S) and the G4(2W) guardrails, the w-beam splices are located at the guardrail posts. During impact, one or more of the splice connections are often subjected to multidirectional loading as the splice bends around a guardrail post while the rail is under relatively high tensile force. As a result, relatively high stress concentrations occur at the splice-bolt holes which can sometimes initiate a small fracture or tear at those locations. Once a tear is initiated, the tension in the rail may cause the tear to propagate vertically through the w-beam section causing the rail to rupture. The pendulum tests performed in this study were designed to create “pocketing” of the rail immediately upstream of a splice connection at the guardrail post in an attempt to emulate typical real-world conditions that lead to splice ruptures. Unfortunately, the tests were not able to accurately duplicate these conditions. Figure 352 shows an overhead view from Test 14004K demonstrating the typical pocketing behavior that occurred in the tests. The image was taken from one of the high-speed cameras at 0.002 seconds before rupture. The degree of pocketing was moderate and may have been sufficient for this study had the tension in the rail not developed so abruptly and so severely. Only one test (i.e., Test 14004F) resulted in rail rupture due to tearing the rail through its cross-section. In all other tests the rupture was due to the splice-bolts tearing through the splice-bolts holes or due to boundary failure. Figure 352. Overhead view of Test 14004K from high-speed video camera. This was basically the same issue that Gabler faced in Phase I, i.e., having a relatively short length of rail with a fixed longitudinal constraint at each end.[Gabler10] As a result very high tensile forces developed in the rail at relatively low rail deflections during the impact tests. As discussed earlier, the test setup did not permit a longer length of rail because of the limited space within the pendulum test area. So, future pendulum impact studies on rail splices should consider designing boundary conditions to more accurately emulate the longitudinal and lateral stiffness behavior of an extended guardrail. Even though the tests did not produce the type of loading that was intended for the splice study, the results still provided some insight into the effects of the various damages on the tensile capacity of the splice connection. These are discussed in the following sections. Rail Flattening The results from Tests 14004F, 14004G and 14004M were used to evaluate the effects of flattening of the rail at the splice connection. Referring to Table 87, the only test article that had rail flattening as the sole damage mode was that of Test 14004G with 83% flattening at the top

407 section of the rail splice. The force and energy curves shown in Figure 343 indicate that this test article resulted in essentially the same capacity as the undamaged test article in Test 14004O. Test 14004M had the highest degree of rail flattening (i.e., 94%) but resulted in no reduction in load capacity and negligible reduction in energy capacity compared to the undamaged cases. Test 14004F, on the other hand, had only 68% flattening but resulted in a 6 percent decrease in load capacity and a 16 percent decrease in energy capacity. Test articles 14004M and 14004F both had the same amount of longitudinal slip in the splice (i.e., 0.2 inches); but 14004F, which resulted in the highest capacity, had greater lateral separation between the rail elements (i.e., 0.35 inches vs. 0.13 inches). The most notable outlier in Test 14004M was the fact that three of the splice-bolts were loose. It is possible that the loose bolts caused higher forces at one or more of the neighboring splice bolt connections which lead to the premature rupture of the splice. Recall that 14004F was the only test article that failed by tearing the rail through the cross-section, with the tear passing through all the downstream splice bolt holes, as shown in Figure 353. Figure 353. Results of Test 14004F with rail tear passing through all four downstream splice-bolt holes. Longitudinal Displacement (Slip) in the Splice Connection The test articles for Tests 14004H and 14004I included the highest levels of lateral separation between rail elements at the downstream edge of the splice connection. The results from these tests yielded mixed results regarding the effects of longitudinal slip on the capacity of the splice, as shown in Table 87 and Figure 344. Test 14004H which had 0.28 inches of longitudinal slip in the splice resulted in 9.5 percent decrease in load capacity and 15 percent decrease in energy capacity. This test specimen also included a 25 percent flattened rail and one loose splice-bolt. Test 14004I, which included 0.25 inches of longitudinal slip, showed no decrease in capacity. This test article also included a 68 percent flattened rail, 0.25 inches of lateral separation between the two rails at the downstream edge of the splice and a large

408 horizontal tear downstream of the splice connection. Since the horizontal tear was located downstream of the splice connection, its effect was considered negligible in the test. It was also concluded that rail flattening was not likely the cause for reduced rail capacity based on the results from the previous group of tests (i.e., Tests 14004M, G and F). As in Test 14004F, it appears that the most notable outlier in Test 14004H was a loose splice-bolt. Vertical Crush There was only one test specimen available that had significant vertical crush of the splice. The damages for Test Article 14004J included 31 percent rail crush and 0.25 inches of lateral separation of the two rail sections at the downstream edge of the splice. The results from the test showed slightly higher load capacity for this damage mode than either of the undamaged cases, as shown in Table 87 and Figure 345. Although there was only a single test conducted for this damage mode, it was concluded that rail crush does not adversely affect rail capacity. Lateral Separation of Rails at Splice Tests 14004K and 14004L both included significant lateral separation between the two rail elements at the downstream edge of the splice (i.e., 0.55 inches and 0.5 inches, respectively). Both test articles also included significant rail flattening (i.e., 97% and 83%, respectively) as well as slight gouging of the splice bolts into the w-beam rail. In addition, Test 14004K also included 0.2 inches of longitudinal slip. The results from these two tests were inconclusive. Test 14004L resulted in little or no decrease in capacity. Test 1400K, on the other hand, resulted in a 14 percent decrease in load capacity and a 4 percent decrease in energy capacity (refer to Table 87 and Figure 346). It was apparent, however, from the sudden drop in force shown in Figure 346 that the splice failed at approximately 30 inches of lateral rail displacement, which corresponded to approximately 470 kip-in of energy absorbed by the rail system (i.e., 10% less energy than the undamaged case). The additional energy absorbed in Test 14004K after splice rupture was due to the post-bolt tearing through the w-beam, based on review of the high-speed test video. It should be noted that the Test Article 14004K did include additional damage modes, but the level of damage was considered low to moderate for those cases. In fact the damages of Test Article 14004K are very similar to those of Test Article 14004M, with the exception of the lateral separation of the rails at the splice; but the results of Test 14004M showed no reduction of splice capacity for that case. Conclusions The effects of crash-induced damages on the residual capacity of w-beam splices were evaluated using dynamic impact testing. The tests were carried out at the Federal Outdoor Impact Laboratory at the Turner-Fairbank Highway Research Center in McLean, Virginia. The tests involved a 4,360-lb pendulum impacting a representative section of w-beam rail with various types and levels of splice damage at an impact speed of approximately 20.5 mph. The test samples were extracted from various crash-damaged field installations by the Maine DOT. The damaged rail specimens were delivered with the splice connections intact (i.e., still assembled) in order to preserve the original damage state of the splice, which include residual stresses in the splice restrained by the bolted connection. Four primary splice damage modes were evaluated: (1) rail flattening, (2) rail crush, (3) longitudinal displacement (slip) between the rail elements at the splice and (4) lateral separation

409 (gap) between the two rail elements at the downstream edge of the splice. A total of 10 test specimens were used in the evaluations. The results from the test program were not very conclusive, due in large part to the low number of test samples available. Further complicating the effort was the fact that, due to the nature of crash damaged rails, each test sample included multiple damage modes - making it difficult to accurately quantify the effects of each isolated mode. The results from the limited test data, however, indicated that rail fattening and rail crush do not significantly affect capacity of the splice connection. The results from the tests involving longitudinal displacement (slip) between the two rail elements at the splice connection indicated that, for the levels of damage investigated in this study (e.g., ¼-inch displacement), this damage mode also did not significantly affect rail capacity. Since longitudinal displacement in the splice is directly associated with high tensile loads in the rail, then relatively high magnitudes of longitudinal displacement would signify an increased potential for damage in the w-beam material at the bolt holes. With that said, such high levels of tension in the rail would likely be associated with other high level damage modes, such as excessive lateral deflections of the w- beam rail or post, that would have already denoted high priority for repair. In other words, once guardrail damage has been deemed severe enough to warrant repair, further assessment of the splice would be moot. Although the damage mode of “loose splice-bolts” was not directly evaluated in the study, two of the test specimens included one or more loose bolts. In both of these test cases there was notable reduction in splice capacity. Based on those results, it is recommended that repair is warranted when one or more splice-bolts are loose or missing. This particular damage mode is not easily evidenced by such obvious signs as crash damage, thus loose bolts will not be easy to identify in the field without conducting close, detailed inspections of the splice. Since it is not feasible to perform such inspections on a routine basis, care should be taken to ensure that the splice-bolts are properly tightened when installing or repairing w-beam rails. Recommendations The following recommendations are based on the results presented herein and on the recommendations by Gabler et al. in NCHRP Report 656.[Gabler10] As a result of these studies, the research team recommends that the repair threshold for splice damage include splice-bolts missing, loose, damaged, severely gouging or torn through the rail, or visibly missing any rail material under the bolt. When any of these damages occur at a single splice-bolt location the recommended repair priority is medium. When any of these conditions occur at two or more splice-bolt locations the recommended repair priority is high. A summary of these recommendations are located in Table 88. Further, all other damage assessment criteria presented throughout this study related to w-beam railing also apply to the w-beam splices; including rail height, rail flattening, rail crush, lateral rail deflection, holes in the rail, horizontal tears and vertical tears.

410 Table 88. Recommendations for assessment criteria for w-beam splice damage. Repair Threshold Relative Priority - Missing - Visibly missing any underlying rail - Severely gouging rail - Torn through rail - Damaged - Loose - Height - Flattening - Crush - Deflection - Non-manufactured holes - Horizontal Tears - Verical Tears Damage Mode Two or more splice-bolt locations At a single splice-bolt location High Medium Splice Bolts W-Beam Refer to assessment criteria for each coresponding damage mode for the w-beam rail element.

Next: Chapter 15. Development of Field Guide Materials and Field Testing »
Criteria for Restoration of Longitudinal Barriers, Phase II Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Guardrails are an important feature of the roadside that are used to shield errant motorists from becoming involved in even more catastrophic crashes by redirecting vehicles away from fixed hazards such as trees and poles and terrain hazards such as steep roadside slopes and fill embankments.

The TRB National Cooperative Highway Research Program's NCHRP Web-Only Document 304: Criteria for Restoration of Longitudinal Barriers, Phase II develops a Field Guide to assist maintenance personnel in making decisions about repairing damaged guardrail installations.

Supplementary material to the document is Appendices A-S.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!