Development and Implementation of the Simplified MGS Stiffness Transition

Transcription

1 Duplication for publication or sale is strictly prohibited without prior written permission of the Transportation Research Board Paper No Development and Implementation of the Simplified MGS Stiffness Transition by Karla A. Lechtenberg, M.S.M.E., E.I.T. Midwest Roadside Safety Facility University of Nebraska-Lincoln 130 Whittier Building 2200 Vine Street Lincoln, Nebraska Phone: (402) Fax: (402) (Corresponding Author) Scott K. Rosenbaugh, M.S.C.E., E.I.T. Midwest Roadside Safety Facility University of Nebraska-Lincoln 130 Whittier Building 2200 Vine Street Lincoln, Nebraska Phone: (402) Fax: (402) Robert W. Bielenberg, M.S.M.E., E.I.T. Midwest Roadside Safety Facility University of Nebraska-Lincoln 130 Whittier Building 2200 Vine Street Lincoln, Nebraska Phone: (402) Fax: (402) Mario Mongiardini, Ph.D. Midwest Roadside Safety Facility University of Nebraska-Lincoln 130 Whittier Building 2200 Vine Street Lincoln, Nebraska Phone: (402) Fax: (402) Ronald K. Faller, Ph.D., P.E. Midwest Roadside Safety Facility University of Nebraska-Lincoln 130 Whittier Building 2200 Vine Street Lincoln, Nebraska Phone: (402) Fax: (402) Francisco Daniel Benicio de Albuquerque, M.S.C.E. Midwest Roadside Safety Facility University of Nebraska-Lincoln 130 Whittier Building 2200 Vine Street Lincoln, Nebraska Phone: (402) Fax: (402) Submitted to Transportation Research Board 91 st Annual Meeting January 22-26, 2012 Washington, D.C. November 15, 2011 Length of Paper: 5,640 (text) + 2,250 (7 figures and 2 tables) = 7,890 words

2 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 2 ABSTRACT Due to the varied lateral stiffness between bridge rails and approach guardrail systems, vehicle pocketing or wheel snag may occur near rigid bridge rail ends. An approach guardrail transition (AGT) is used to mitigate this potential hazard by providing a gradual increase in the lateral barrier stiffness between the W-beam guardrail system and the bridge railing. However, these transitions can also cause a propensity for vehicle pocketing or wheel snag if the change in lateral stiffness occurs too rapidly. Recently, an NCHRP 350 stiffness transition was developed for use with the Midwest Guardrail System (MGS) and a stiff thrie-beam AGT with successful testing performed close to the upstream end of the AGT. It was designed with three different steel post sizes, one of which was non-standard for State Departments of Transportation (DOTs). Thus, a simplified version of the original MGS stiffness transition was developed utilizing two common sizes of steel posts and was full-scale crash tested according to Test Level 3 (TL-3) set forth in the Manual for Assessing Safety Hardware (MASH). Subsequently, an equivalent woodpost version of the simplified MGS stiffness transition was also developed using dynamic post properties obtained from bogie testing and numerical simulations. Recommendations were also made regarding the attachment of the stiffness transitions to prior Federal Highway Administration (FHWA) accepted thrie beam, bridge rail approach guardrail transitions. Keywords: Bridge Rail, Approach Guardrail Transition, MGS, Roadside Safety, Crash Testing, MASH, TL-3.

3 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 3 INTRODUCTION Throughout the United States, State Highway Departments commonly use standard strong-post, W-beam guardrail systems to prevent errant vehicles from leaving the roadway and encountering safety hazards beyond the roadway edge. One of the more common applications for these barriers is to prevent errant vehicles from impacting blunt ends of bridge railings and their associated drop offs. Although strong-post, W-beam systems are generally considered to be semi-rigid, these barriers are much more flexible than most bridge railing systems. Consequently, vehicle pocketing or wheel snag may occur at the point of attachment to a rigid bridge rail end. Thus, an approach guardrail transition (AGT) region is added between the semirigid W-beam guardrail system and the stiff bridge railing system to provide a more gradual change in lateral barrier stiffness. Traditionally, AGTs have been comprised of some combination of reduced post spacing, longer posts, additional rail elements, a curb incorporated under the barrier, and thrie beam guardrail. These additional elements provide increased stiffness in the AGT system and prevent vehicles from impacting the upstream end of the bridge rail. However, the upstream end of the typical AGT can provide a critical location where vehicle pocketing or wheel snag may occur due to rapid changes in lateral barrier stiffness. Prior full-scale crash tests involving impacts near the upstream ends of typical AGTs have exhibited vehicle pocketing, wheel snag, rail ruptures, and vehicle instabilities [1-4]. In 2007, the Midwest Roadside Safety Facility (MwRSF) developed a new gradual stiffness transition between the Midwest Guardrail System (MGS) and a stiff thrie beam, bridge rail approach transition system to provide the gradual change in lateral barrier stiffness, as shown in Figure 1a [5-7]. The MGS stiffness transition utilized an asymmetrical W-beam to thrie beam transition element to connect the MGS to one of the stiffest FHWA-accepted thrie beam transition systems used by State Departments of Transportation (DOTs). This stiffness transition was successfully crash tested near the upstream end of the AGT according to Test Level 3 (TL-3) of National Cooperative Highway Research Program (NCHRP) Report No. 350 [8]. The MGS stiffness transition provided an adaption from the 31-in. (787-mm) high MGS to the upstream end of a crashworthy thrie beam approach transition. However, the MGS stiffness transition utilized three types of steel posts, including non-standard W6x12 (W152x17.9) steel guardrail posts that were not currently used by most State DOTs. Thus, the system was viewed as being too complicated. Further, stiffness transitions had not yet been crash tested and evaluated under the newly-adopted guidelines found in the Manual for Assessing Safety Hardware (MASH) [9]. RESEARCH OBJECTIVE The objectives of this research project were to: (1) develop a simplified stiffness transition between the MGS and a thrie beam AGT using only standard W6x15 (W152x22.3) and W6x9 (W152x13.4) steel posts; (2) develop a wood-post alternative to the simplified steel-post MGS stiffness transition utilizing the same rail elements and only replacing the steel posts; and (3) provide recommendations regarding the attachment of the stiffness transitions to Federal Highway Administration (FHWA) accepted thrie beam, bridge rail approach guardrail

4 (a) Original MGS Stiffness Transition (Non-Standardd Post) (b) Simplified Steel-Post MGS Stiffness Transition (c) Simplified Wood-Post MGS Stiffness Transition FIGURE 1 MGS Stiffness Transition Designs. Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 4

5 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 5 transitions. The simplified stiffness transition systems were to be evaluated according to the TL- 3 safety performance criteria set forth in MASH. MGS STIFFNESS TRANSITION STEEL POST VERSION Background and Scope The basic design of the stiffness transition included standard MGS, a very stiff thrie beam bridge rail, a previously-accepted thrie beam AGT system, and the asymmetrical W-beam to thrie beam transition element [5-7]. The asymmetrical transition element utilized the same basic philosophy as the symmetrical transition element (i.e., the gradual introduction of a new peak between the two existing peaks in the W-beam). The thrie beam AGT consisted of nested 12-gauge (2.66-mm thick) thrie beam attached to W6x15 (W152x22.3) steel posts at half-post or 37½-in. spacings, which represented a critical configuration (one of the stiffest AGT) after reviewing the previously-accepted FHWA AGT systems [10-11]. Dynamic post testing and BARRIER VII [12] were utilized to determine the optimum, simplified transition design. Dynamic Post Testing Prior research conducted by MwRSF has determined the dynamic properties of W6x9 (W152x13.4) steel posts. However, the dynamic properties of the larger W6x15 (W152x22.3) steel posts used in the selected approach transition system had not been determined [10-11]. Thus, dynamic component testing was utilized to determine the post-soil behavior of W6x15 (W152x22.3) steel posts placed in compacted, coarse, crushed limestone material meeting America Association of State Highway and Transportation Officials (AASHTO) standard soil designation M 147, Grade B, as recommended in MASH. Two dynamic bogie tests were performed on W6x15 (W152x22.3) steel posts embedded 54 in. (1,372 mm) in soil and impacted at a height of 24⅞ in. (632 mm). For each bogie test, raw acceleration data was acquired and filtered, and then force-displacement and energydisplacement graphs were plotted. From the energy-displacement graphs, the average post-soil forces were calculated for various displacements at the center rail height. Each test showed very similar results with maximum deflections of approximately 19.5 in. (495 mm) and average forces of between 10.9 and 14.7 kips (48.5 and 65.6 kn) from 10 in. (254 mm ) of deflection until reaching their maximum deflections. Thus, the typical dynamic response of a W6x15 (W152x22.3) steel post embedded in soil were provided from the test results. System Design Using BARRIER VII Analysis As a vehicle approaches the stiffened, semi-rigid AGT from the less stiff W-beam guardrail, a potential for vehicle pocketing exists. Pocketing occurs when the lateral deflection of the downstream guardrail system is much less than the upstream barrier system, thus causing a sharp bend in the guardrail system as the impacting vehicle approaches the stiffer region. The sharp bend produces a high longitudinal force on the vehicle that can create excessive decelerations or cause vehicle override of the barrier near the pocket and/or rollover. BARRIER VII simulations [13] were conducted to calibrate the barrier model with the full-scale test results from the original FHWA-accepted bridge rail AGT, the original MGS

6 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 6 stiffness transition (non-standard post), and the MGS. The force-deflection characteristics of W6x15 (W152x22.3) steel post embedded in soil obtained during the dynamic bogie tests were utilized in the proposed stiffness transition model. The initial model components (post deflection at failure and effective coefficient of friction) were adjusted such that the system defections, vehicle time to parallel, and vehicle exit time matched the results from the full-scale tests within 10 percent. Each transition configuration was then subjected to simulated MASH TL-3 pickup truck impacts with a speed of mph (100 km/h) and an angle of 25 degrees. In order to analyze the entire approach transition, the impact point was moved along the length of the stiffness transition system at 9⅜-in. (238-mm) intervals from 12.5 ft (3.8 m) upstream of the first reduced post spacing downstream through the W6x15 (W152x22.3) steel posts, which are part of the original FHWA-accepted thrie beam bridge rail transition. The BARRIER VII simulations were utilized to predict the severity of pocketing on various transition configurations, as shown in Table 1. The maximum pocketing angle was quantified for seven different stiffness transition configurations. Each configuration utilized a different combination of post quantities, post spacing, and system lengths. Prior research noted a desire to limit the maximum guardrail pocketing angle to approximately 23 degrees [5-6]. However, due to the larger and more stable MASH pickup truck as compared to the NCHRP 350 pickup truck, it was believed that the critical pocket angle was likely closer to 30 degrees. Simulated pocketing angles were measured using linear regression fit to five consecutive points along the barrier (i.e., over a distance of 37½ in. or 953 mm). Full input parameters for the BARRIER VII models and analysis results are available in the referenced report [13]. Designs K and L were identified as the preferred stiffness transition alternatives. These options were two of the shorter configurations, neither required nesting of W-beam rail between the standard MGS and the stiffness transition, and both restricted the pocketing angles to less than 30 degrees. Designs K and L were similar, except Design L had one additional post at halfpost spacing on the upstream end. However, the maximum pocketing angles were 29.6 and 23.9 degrees for Designs K and L, respectively. Thus, Design K was considered a more aggressive design, while Design L was considered a more conservative design. Design K, as shown in Figures 1b and 2, was selected for testing by the sponsoring states since it was one of the shortest transitions and provided the lowest cost alternative. Full-Scale Crash Testing Test No. MWTSP-2 Pickup Truck Test For test no. MWTSP-2, a 5,158-lb (2,340-kg) pickup truck with a dummy or surrogate occupant seated in the right-front seat, impacted the simplified MGS stiffness transition system at a speed of 61.2 mph (98.5 km/h) and at an angle of 26.3 degrees. A summary of the test results and timesequential photographs are shown in Figure 3. The critical impact point (CIP) was determined to be 75 in. (1,905 mm) upstream from the centerline of post no. 9 from a detailed BARRIER VII analysis discussed in the referenced report [13]. The actual point of impact was 74½ in. (1,892 mm) upstream from the centerline of post no. 9. At sec after impact, the vehicle became parallel to the guardrail with a speed of 41.4 mph (66.6 km/h). At sec, the vehicle exited the guardrail at an angle of 22.0 degrees and at a speed of 37.3 mph (60.0 km/h), and its trajectory did not violate the bounds of the exit box. The vehicle was smoothly redirected even though the vehicle encountered moderate pitching and rolling. The exterior vehicle damage,

7 TABLE 1 Summary of Stiffness Transitions Combined with Thrie Beam Approach Guardrail Transitions. System Parameters Design D Design F Design G Design H Design J Design K Design L Length of Transition 1 Length of Nested W-beam Rail Length of 10-Gauge W-to-Thrie Rail Length of Thrie Beam Rail Length of Nested Thrie Beam Rail 40 ft 7½ in. (12.38 m) ft 6 in. (3.81 m) 12 ft 6 in. (3.81 m) 40 ft 7½ in. (12.38 m) 25 ft (7.62 m) 12 ft 6 in. (3.81 m) 28 ft 1½ in. (8.57 m) 12 ft 6 in. (3.81 m) 12 ft 6 in. (3.81 m) 34 ft 4½ in. (10.48 m) 12 ft 6 in. (3.81 m) 12 ft 6 in. (3.81 m) 34 ft 4½ in. (10.48 m) 25 ft (7.62 m) 12 ft 6 in. (3.81 m) 28 ft 1½ in. (8.57 m) 34 ft 4½ in. (10.48 m) ft 6 in. (3.81 m) 12 ft 6 in. (3.81 m) No. of W6x9 Posts No. of W6x15 Posts First reduced post spacing to first bridge post. 2 - Posts located within the reduced post spacing region (i.e., the post at the beginning of standard post spacing is included). Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 7

8 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Alberquedue 8 FIGURE 2 Simplified Steel-Post MGS Stiffness Transition System.

9 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Alberquedue 9 FIGURE 3 Summary of Test Results and Sequential Photographs, Test MWTSP-2.

10 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 10 as shown in Figure 4, was moderate, and the interior occupant compartment deformations were minimal, with a maximum of 1¼ in. (32 mm), consequently not violating the limits provided in MASH. As shown in Figure 4, damage to the barrier was moderate, consisting mostly of deformed W-beam and guardrail posts as well as contact marks on guardrail and posts. The maximum lateral dynamic rail and post deflections were 30.7 in. (779 mm) at the centerline of post no. 10 and 32.8 in. (833 mm) at the centerline of post no. 11, respectively, as determined from high-speed digital video analysis. The working width of the system was 51.6 in. (1,310mm), as determined from high-speed digital video analysis. All occupant risk measures were well below recommended values, and the test vehicle showed no tendency to roll over. Therefore, test no. MWTSP-2 was determined to be acceptable according to the TL-3 safety performance criteria found in MASH. Test No. MWTSP-3 Small Car Test For test no. MWTSP-3, a 2,591-lb (1,175-kg) small car with a dummy or surrogate occupant seated in the right-front seat, impacted the simplified MGS stiffness transition system at a speedof 61.0 mph (98.2 km/h) and at an angle of 25.7 degrees. A summary of the test results and time-sequential photographs are shown in Figure 5. The CIP was selected to maximize pocketing and the probability of wedging the vehicle under the asymmetrical W-beam to thrie beam transition element. From a detailed BARRIER VII analysis discussed in the referenced report, the CIP was determined to be 93¾ in. (2,381 mm) upstream from the centerline of post no. 9 [13]. The actual point of impact occurred at the targeted impact point. By sec after impact, the vehicle appeared to be redirecting. At this point, the vehicle stopped redirecting and yawed about the front of the vehicle as it remained in contact with the system. At sec and after yawing significantly, the vehicle exited the guardrail at an angle of 68.7 degrees and at a speed of 5.8 mph (9.4 km/h). The exterior vehicle damage, as shown in Figure 6, was moderate, and the interior occupant compartment deformations were minimal, with a maximum deformation of 1¾ in. (44 mm), consequently not violating the limits provided in MASH. As shown in Figure 6, damage to the barrier was moderate, consisting mostly of deformed W-beam rail and steel guardrail posts as well as contact marks on guardrail and posts. The maximum lateral dynamic rail and post deflections were 25.9 in. (658 mm) at the mid-span between post nos. 9 and 10 and 18.5 in. (470 mm) at the centerline of post no. 11, respectively, as determined from high-speed digital video analysis. The working width of the system was 39.8 in. (1,011 mm), as determined from high-speed digital video analysis. All occupant risk measures were well below recommended values, and the test vehicle showed no tendency to roll over. Therefore, test no. MWTSP-3 was determined to be acceptable according to the TL-3 safety performance criteria found in MASH. MGS STIFFNESS TRANSITION WOOD POST VERSION Dynamic Post Testing For many years, the roadside safety community has considered 6-in. x 8-in. (152-mm x 203-mm) wood posts and W6x9 (W152x13.4) steel posts interchangeable for 6-ft (1.8-m) long guardrail posts. Similar performance between W6x9 (W152x13.4) steel and 6-in. x 8-in. (152-mm x 203- mm) wood guardrail posts has been documented in both dynamic component testing and fullscale testing [14-16]. Thus, 6-in. x 8-in. (152-mm x 203-mm) wood posts were selected as an

11 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 11 FIGURE 4 Vehicle Damage and Barrier Damage, Test MWTSP-2. Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg and Alberquedue 16

12 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Alberquedue 12 FIGURE 5 Summary of Test Results and Sequential Photographs, Test MWTSP-3.

13 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Alberquedue 13 FIGURE 6 Vehicle Damage and Barrier Damage, Test MWTSP-3.

14 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 14 alternative for the W6x9 (W152x13.4) steel posts found in the MGS to thrie beam stiffness transition. However, it was necessary to find an equivalent wood post alternative for the W6x15 (W152x22.3) steel posts utilized in the steel-post MGS stiffness transition. Thus, a series of dynamic component tests were conducted on W6x15 (W152x22.3) steel posts and various large cross-section, wood posts [17]. Several southern yellow pine (SYP) sections were investigated, and the test results are summarized in Table 2. Variability in wood strength proved to be a difficult hurdle to overcome as multiple posts of the same cross section, embedment depth, and wood grade occasionally resulted in different failure mechanisms. After inspection, many of the fractured posts contained wood defects which are inevitable in large cross-sectional wood posts. Consequently, posts used in actual installations would be expected to have natural defects that may cause premature post fracture. Fractured posts absorb less energy and do not provide any resistance after fracture. This lack of resistance can have a negative effect on the safety performance of the sensitive region of the stiffness transition. Therefore, the 8-in. x 8-in. (203-mm x 203-mm) and 6-in. x 10-in. (152-mmx 254-mm) wood posts were removed from this equivalent consideration as they showed a greater propensity to fracture prematurely during the component testing program. Considering the average force for each of the remaining post sizes, the 6.5-ft (2.0-m) long 8-in. x 10-in. (203-mm x 254-mm) wood posts provided average force characteristics that best match those of W6x15 (W152x22.3) steel posts, as shown in Table 2. At 10 in. and 15 in. (254 mm and 381 mm) of deflection, the 8-in. x 10-in. (203-mm x 254-mm) wood posts averaged 15.5 percent and 1.1 percent higher than the steel posts, respectively. Further, the researchers would rather error conservatively toward strong as opposed to being too weak and opening up the probability of snagging on the bridge rail. Therefore, based on the component testing program, the 8-in. x 10-in. (203-mm x 254-mm) wood post with an embedment depth of 48 in. (1,219 mm) best resembled the performance of the W6x15 (W152x22.3) steel transition post and was recommended for further analysis in the wood-post MGS stiffness transition. System Performance Evaluated Using BARRIER VII The simplified, wood-post MGS stiffness transition was not evaluated with full-scale vehicle crash testing but instead through the use of dynamic component testing of posts embedded in soil combined with BARRIER VII computer simulations. This approach was similar to that used in a prior study that determined a wood-post equivalent for the 9-ft (2.7-m) long W6x9 (W152x13.4) steel posts in the MGS adjacent to a steep slope which received FHWA acceptance [18]. At a minimum, the simplified wood-post transition system was required to perform as well as the simplified steel-post transition system in terms of the following three criteria: (1) maximum deflection; (2) maximum pocketing angle; and (3) propensity for vehicle snag on posts. These performance characteristics were the same as those used to compare the various configurations during the development of the original steel-post stiffness transition. Thus, the wood-post transition system was evaluated in the same manner. Full input parameters for the BARRIER VII models as well as the simulation results are available in the referenced report [17]. Following the full-scale crash tests, BARRIER VII simulations were conducted to calibrate the barrier model with the full-scale test results (test nos. MWTSP-2 and MWTSP-3). After calibration of the steel-post stiffness transition barrier model, the characteristics of the 6-in. x 8-in. (152-mm x 203-mm) and 8-in. x 10-in. (203-mm 254-mm) wood posts replaced the steel

15 TABLE 2 Bogie Test Results and Averages by Post Size. Test No. MGSATB-5 MGSATB-6 MGSATB-7 MGSATB-8 MGSATB-9 MGSATB-11 MGSATB-12 Post Type in. x in. (mm x mm) Steel W6x15 (W152x22.3) Steel W6x15 (W152x22.3) SYP 8x8 (203 x 203) SYP 8x8 (203 x 203) SYP 8x10 (203 x 254) SYP 8x10 (203 x 254) SYP 8x10 (203 x 254) Post Length ft (m) 7 (2.1) 7 (2.1) 7 (2.1) 7 (2.1) Embed. Depth in. (mm) 54 (1,372) 54 (1,372) Failure Type Post Yielding & Rotation in Soil Post Yielding & Rotation in Soil Average W6x15, 7 Long 54 (1,372) 54 (1,372) Post Fracture Post Fracture Average SYP 8 x8, 7 Long 7 (2.1) 7 (2.1) 7 (2.1) 54 (1,372) 54 (1,372) 54 (1,372) Post Fracture Rotation in Soil Rotation in Soil Average SYP 8 x10, 7 Long Total Energy kip-in. (kj) (39.7) (38.9) (39.3) 73.0 (8.3) 66.8 (7.5) 69.9 (7.9) 37.3 (4.2) (35.2) (31.1) (23.5) Average 15" kip kip kip (kn) (kn) (kn) (68.6) (75.3) (75.3) (71.9) (79.7) (80.2) (70.3) (77.5) (77.7) 10.9 (48.4) 12.7 (56.4) 11.8 (52.4) 7.5* (33.2)* 21.6 (95.9) 18.1 (80.4) 15.7 (69.8) 7.5* (33.2)* 6.9* (30.5)* 7.2 (31.8) NA NA NA NA 25.1 (111.5) 20.8 (92.7) 23.0 (102.1) Percent Change from W6x15 15" NA -58.6% NA NA NA NA 32.2% NA Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Alberquedue 15

16 TABLE 2 Bogie Test Results and Averages by Post Size (continued). Test No. MGSATB-13 MGSATB-14 MGSATB-15 MGSATB-16 MGSATB-17 MGSATB-10 MGSATB-18 MGSATB-19 MGSATB-20 Post Type in. x in. (mm x mm) SYP 8x10 (203 x 254) SYP 8x10 (203 x 254) SYP 8x10 (203 x 254) SYP 8x10 (203 x 254) SYP 8x10 (203 x 254) SYP 10x10 (254 x 254) SYP 6x10 (152 x 254) SYP 6x10 (152 x 254) SYP 6x10 (152 x 254) Post Length ft (m) 6.5 (2.0) 6.5 (2.0) 6.5 (2.0) 6.5 (2.0) 6.5 (2.0) Embed. Depth in. (mm) 48 (1,219) 48 (1,219) 48 (1,219) 48 (1,219) 48 (1,219) Failure Type Rotation in Soil Rotation in Soil Rotation in Soil Post Fracture Rotation in Soil Average SYP 8 x10, 6.5 Long 7 (2.1) 54 (1,372) Rotation in Soil Average SYP 10 x10, 7 Long 7 (2.1) 7 (2.1) 7 (2.1) * Fracture had already been initiated. 52 (1,321) 52 (1,321) 52 (1,321) Rotation in Soil Post Fracture Post Fracture Average SYP 6 x10, 7 Long Total Energy kip-in. (kj) (33.8) (32.0) (36.7) (22.0) (32.3) (31.4) (34.7) (34.7) (36.7) (14.0) 28.5 (3.2) (18.0) Average 15" kip kip kip (kn) (kn) (kn) (60.9) (65.1) (67.4) (69.3) (76.7) (75.9) (88.9) (109.2) (92.6) * (89.3) (85.9)* NA (104.2) (109.7) NA (82.5) (89.3) (78.6) 25.6 (114.0) 25.6 (114.0) 14.7 (65.6) 11.8 (52.6) 5.5* (24.6)* 10.7 (47.6) 28.2 (125.5) 28.2 (125.5) 17.7 (78.8) 11.5* (51.3)* NA Percent Change from W6x15 15" 15.5% 1.1% NA 62.1% NA 18.4 (81.9) NA NA NA 14.6 (65.0) 18.4 (81.69) -16.1% 5.1% Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Alberquedue 16

17 posts to create a wood-post transition model. Both the steel- and wood-post versions of the stiffness transition model were subjected to simulated MASH TL-3 impacts at 45 different impact points throughout the length of the MGS stiffness transition. The results of the BARRIER VII analysis showed that the simplified wood-post transition system consistently provided reductions in maximum deflection, vehicle pocketing angle, and propensity for wheel snag. The predicted maximum deflections for the wood-post system were consistently 15 to 30 percent lower than the simplified steel-post system. The wood-post system also consistently showed a 5 to 25 percent reduction in the maximum pocketing angle, reducing the risk of vehicle instability. Finally, the propensity for wheel snag was found to be lower for the wood-post system. The reduction in system deflection significantly reduced the estimated wheel snag for the 6-in. x 8-in. (152-mm x 203-mm) wood post. However, wheel snag on the larger 8-in. x 10-in. (203-mm x 254-mm) wood transition posts was found to be closer to that estimated for the steel W6x15 (W152x22.3) steel posts. The slight increase (less than 0.31 in. or 7.9 mm) in potential wheel snag at a limited number of locations was not considered a major concern, and the wood-post MGS stiffness transition system was expected to perform satisfactorily for all safety criteria. Thus, substituting the wood posts for the steel posts in the simplified stiffness transition did not adversely affect the safety performance of the successfullytested, simplified stiffness transition. Details of the equivalent wood-post stiffness transition are shown in Figure 1c and further documented in the referenced report [17]. SUMMARY AND CONCLUSIONS A new, simplified steel-post stiffness transition between the MGS and a previously FHWAaccepted thrie beam approach guardrail transition was developed and satisfied all TL-3 safety performance criteria presented in MASH. This simplified stiffness transition consists of standard steel posts and an asymmetric W-beam-to-thrie beam transition element. A very stiff thrie beam guardrail transition was used during the full-scale crash testing program. Following the successful development of the simplified steel-post stiffness transition, a wood-post equivalent stiffness transition was developed and evaluated. BARRIER VII computer simulations were utilized to evaluate the behavior of the wood post stiffness transition system. It is believed that the wood-post transition system would also satisfy the MASH TL-3 performance criteria since the BARRIER VII analysis showed that the wood-post stiffness transition system behaved similarly (without increases in deflections, pocketing, or snag) to the crash tested steel-post stiffness transition. IMPLEMENTATION OF THE SIMPLIFIED MGS STIFFNESS TRANSITIONS Recall, the simplified MGS stiffness transition was tested in combination with a very stiff thrie beam AGT, specifically the Missouri thrie beam and channel bridge railing [10]. Similarly, the wood-post version of the simplified stiffness transition was assessed using numerical simulations based on the same stiff thrie-beam AGT. This stiff AGT represented a worst-case scenario in terms of the probability for vehicle snag and pocketing due to its high lateral barrier stiffness. Since the AGT was very stiff, both the simplified steel- and wood-post MGS stiffness transitions should be adaptable to most crashworthy thrie beam AGT designs. Implementation of the crash tested, simplified stiffness transition to the AGT of the Missouri thrie beam and channel bridge railing as well as the wood-post stiffness transition are

18 shown in Figures 1b and 1c, respectively. Details for attaching the simplified MGS stiffness transitions developed in this study to common types of existing thrie-beam AGTs for bridge railings is presented herein with additional adaptations available in the referenced reports [13,17]. Steel-Post MGS Stiffness Transition - Thrie Beam/Tube Bridge Rail The original AGT to thrie beam and tube bridge rail for use on transverse glulam timber deck bridges and the adaptation of the simplified, steel-post stiffness transition are shown in Figure 7a [19-22]. The original AGT was 18 ft 9 in. (5.7 m) long, while the adapted system was configured to be 28 ft 1½ in. (8.6 m) long, as measured from the centerline of the first post at half-post spacing to the centerline of the splice between the thrie beam and the bridge rail. A 75- in. (1,905-mm) long, 12-gauge (2.66-mm thick) thrie beam section is used between the asymmetrical stiffness transition element and the 10-gauge (3.42-mm thick) thrie beam section of the AGT. Four out of five of the 7-ft (2.1-m) long W6x15 (W152x22.3) steel posts from the original AGT system were retained. Upstream from these four posts, eight 6-ft (1.8-m) long W6x9 (W152x13.4) steel posts are installed. The first four of these eight posts, starting from the downstream end, are placed at quarter-post spacing and the last four are placed at half-post spacing. Next, another set of four 6-ft (1.8-m) long, W6x9 (W152x13.4) steel posts are installed at half-post spacing, such that the first upstream post at half-post spacing is located 37½ in. (953 mm) upstream of the centerline of the splice between the MGS and the asymmetrical stiffness transition element. Wood-Post MGS Stiffness Transition - Iowa Transition to NJ Shape Concrete Bridge Railing The original Iowa AGT to New Jersey safety-shape concrete parapet and the adaptation of the simplified, wood-post stiffness transition are shown in Figure 7b [23-26]. The original AGT was 18 ft 9 in. (5.7 m) long, while the adapted system was configured to be 28 ft 1½ in. (8.6 m) long, as measured from the centerline of the first post at half-post spacing to the centerline of the splice between the thrie beam and the concrete parapet. In the adapted system, an additional 6-ft 3-in. (1.9-m) long segment of 12-gauge (2.66-mm thick) thrie beam was placed downstream of the asymmetrical transition element. The six 7-ft (2.1-m) long, 6-in. x 8-in. (152-mm x 203-mm) wood posts from the original AGT were retained. Upstream from these six posts, a total of ten 6-ft (1.8-m) long, 6-in. x 8-in. (152-mm x 203-mm) wood posts were placed. The first four of these ten posts are placed at quarter-post spacing and the last six are placed at half-post spacing. The first of the six posts at half-post spacing was located 37½ in. (953 mm) upstream from the centerline of the splice between the MGS and the asymmetrical stiffness transition element.

19 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerqu 19 Original Approach Guardrail Transition (STTR Test Series) Adapted Simplified Steel-Post Stiffness Transition (a) Transition to Thrie Beam and Tube Bridge Railing Steel-Post Version Original Approach Transition to NJ-Shape Concrete Bridge Railing (Iowa) Adapted Simplified Wood-Post Stiffness Transition (b) Transition to NJ-Shape Concrete Bridge Railing Wood-Post Version FIGURE 7 Implementation Examples for the Simplified MGS Stiffness Transitions: (a) Steel-Post Version and (b) Wood-Post Version.

20 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 20 General Implementation Guidelines The simplified stiffness transition was tested with a length of 43 ft 9 in. (13.3 m) of standard MGS between the stiffness transition and the upstream end anchorage. Guardrail terminals have been designed, crash tested, and evaluated for use when directly attached to semi-rigid W-beam guardrail systems instead of the stiff AGTs. However, the placement of the upstream end anchorage too close to the MGS stiffness transition may negatively affect system performance, and result in excessive barrier deflections, vehicle pocketing, wheel snag on posts, vehicle-tobarrier override, or other vehicle instabilities. Therefore, the following implementation guidelines should be followed (and are the same) for both the steel-post and wood-post MGS stiffness transitions. 1. A recommended minimum length of 12 ft 6 in. (3.8 m) for standard MGS is to be installed between the upstream end of the asymmetrical transition element and the interior end of an acceptable TL-3 guardrail end terminal. 2. A recommended minimum barrier length of 46 ft 10½ in. (14.3 m) is to be installed beyond the upstream end of the asymmetrical transition element, which includes standard MGS, a crashworthy guardrail end terminal, and an acceptable anchorage system. 3. For flared guardrail applications, a minimum length of 25 ft (7.6 m) is recommended between the upstream end of the asymmetrical transition element and the start of the flared section (i.e. bend between flare and tangent sections). It is unknown as to whether a non-blocked version of the MGS will perform in an acceptable manner when installed adjacent to the new stiffness transition. Since the safety performance of the non-blocked MGS in conjunction with the new stiffness transition can only be verified through the use of full-scale crash testing, it is recommended that a minimum of 25 ft (7.6 m) of standard MGS with spacer blocks be placed adjacent to the new stiffness transition prior to transitioning to other non-blocked, 31-in. (787-mm) tall, W-beam guardrail systems. The new stiffness transition was developed and evaluated for use with a thrie beam approach guardrail transition where all posts were installed on level terrain. Therefore, this stiffness transition should be implemented with a minimum of 2 ft (0.61 m) of level or gentlysloped fill placed behind the posts, unless special design provisions are made to account for decreased post-soil resistance. Recently, FHWA acceptance requests were submitted for both the steel- and wood-post simplified stiffness transitioned developed and evaluated in this study as well as the adaptations to other thrie beam AGTs. DISCLAIMER The contents of this report reflect the views of the authors who are responsible for the facts and the accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of the FHWA nor the state highway departments participating in the Midwest

21 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 21 States Regional Pooled Fund Program. This report does not constitute a standard, specification, or regulation. ACKNOWLEDGMENTS The authors wish to acknowledge several sources that made a contribution to this project: (1) the Midwest States Regional Pooled Fund Program for sponsoring this research project; (2) IMH Products, Inc. for donating the asymmetrical W-beam to thrie beam transition elements; and (3) MwRSF personnel for constructing the barriers and conducting the crash tests. REFERENCES 1. Bryden, J.E. and Phillips, R.G., Performance of a Thrie Beam Steel-Post Bridge-Rail System. Report No. FHWA/NY/RR-85/118, Submitted to the Office of Research, Development, and Technology, Federal Highway Administration, Performed by New York State Department of Transportation, February Bryden, J.E. and Phillips, R.G., Performance of a Thrie Beam Steel-Post Bridge-Rail System, Transportation Research Record No. 1024, Transportation Research Board, Washington, D.C., Mak, K.K. and Menges, W.L., Testing and Evaluation of the W-beam Transition (on Steel Posts with Timber Blockouts) to the Vertical Flared-Back Concrete Bridge Parapet, Report No. FHWA-RD , Submitted to the Office of Research, Development, and Technology, Federal Highway Administration, Performed by Texas Transportation Institute, Texas A&M University, College Station, Texas, November Polivka, K.A., Faller, R.K., Sicking, D.L., Reid, J.D., Rohde, J.R., and Holloway, J.C., Crash Testing of Missouri s W-Beam to Thrie beam Transition Element, Final report to the Midwest States Regional Pooled Fund Program, Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, September 12, Polivka, K.A., Eller, C.M., Faller, R.K., Sicking, D.L., Rohde, J.R., Reid, J.D., Bielenberg, B.W., and Allison, E.M., Development of the Midwest Guardrail System (MGS) W-Beam to Thrie Beam Transition Element, Final Report to the Midwest States Regional Pooled Fund Program, Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, November 26, Polivka, K.A., Coon, B.A., Sicking, D.L., Faller, R.K., Bielenberg, R.W., Rohde, J.R., and Reid, J.D., Midwest Guardrail System (MGS) W-Beam to Thrie Beam Transition, Transportation Research Record No. 2025, Transportation Research Board, Washington, D.C., 2007.

22 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque Nicol, D.A., Federal Highway Administration (FHWA) NCHRP Report No. 350 Acceptance Letter B-187 of the Transition from MGS to Asymmetrical Thrie Beam, To R. K. Faller, Midwest Roadside Safety Facility, February 13, Ross, H.E., Sicking, D.L., Zimmer, R.A., and Michie, J.D., Recommended Procedures for the Safety Performance Evaluation of Highway Features, National Cooperative Highway Research Program (NCHRP) Report No. 350, Transportation Research Board, Washington, D.C., Manual for Assessing Safety Hardware (MASH), American Association of State Highway and Transportation Officials (AASHTO), Washington, D.C., Polivka, K.A., Faller, R.K., Reid, J.D., Sicking, D.L., Rohde, J.R., Keller, E.A., and Holloway, J.C., Development of an Approach Guardrail Transition Attached to a Thrie Beam and Channel Bridge Railing. Research Report No. TRP , Final Report to the Midwest States Regional Pooled Fund Program, Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, March 21, Horne, D.A., Federal Highway Administration (FHWA) NCHRP Report No. 350 Acceptance Letter of the W-Beam Transition to Thrie Beam and Channel, To R. K. Faller, Midwest Roadside Safety Facility, June 4, Powell, G.H., Barrier VII: A Computer Program for Evaluation of Automobile Barrier Systems, Report No. FHWA-RD-73-51, Department of Civil Engineering, University of California, Berkeley, California, April Rosenbaugh, S.K., Lechtenberg, K.A., Faller, R.K., Sicking, D.L., Bielenberg, R.B., and Reid, J.D., Development of the MGS Approach Guardrail Transition Using Standardized Steel Posts, Final Report to the Midwest States Regional Pooled Fund Program, Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska- Lincoln, Lincoln, Nebraska, December 21, Homan, D.M., Thiele, J.C., Faller, R.K., Rosenbaugh, S.K., Rohde, J.R., Arens, S.W., Lechtenberg, K.A., Sicking, D.L., and Reid, J.D., Investigation and Dynamic Testing of Wood and Steel Posts MGS on a Wire-Faced MSE Wall, Draft Report to the Federal Highway Administration, Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, June Polivka, K.A., Faller, R.K., Sicking, D.L., Rohde, J.R., Bielenberg, B.W., and Reid, J.D., Performance Evaluation of the Midwest Guardrail System Update to NCHRP 350 Test No with 28 C.G. Height (2214MG-2), Final Report to the National Cooperative Highway Research Program (NCHRP), Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, October 11, Stolle, C.J., Lechtenberg, K.A., Faller, R.K., Rosenbaugh, S.K., Sicking, D.L., and Reid, J.D., Evaluation of the Midwest Guardrail System (MGS) with White Pine Wood Posts, Final Report to the Wisconsin Department of Transportation, Research Report No. TRP-

23 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque , Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, March 28, Rosenbaugh, S.K., Schrum K.D., Faller, R.K., Lechtenberg, K.A., Sicking, D.L., and Reid, J.D., Development of the Wood-Post MGS Approach Transition. Final Report to the Midwest States Regional Pooled Fund Program, Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, October 18, Griffith, M.S., Federal Highway Administration (FHWA) NCHRP Report No. 350 Acceptance Letter B-211A of the Wood Post Alternative to the Steel Post Midwest Guardrail System installed adjacent to a 2H:1V Fill Slope, To D. L. Sicking, Midwest Roadside Safety Facility, June 10, Polivka, K.A., Faller, R.K., Rosson, B.T., Ritter, M.A., Fowler, M.D., and Keller, E.A., Two Test Level 4 Bridge Railing and Transition Systems for Transverse Glue-Laminated Timber Decks, Draft Report to the U.S. Department of Agriculture, Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska - Lincoln, Lincoln, Nebraska, January 30, Faller, R.K., Ritter, M.A., Rosson, B.T., Fowler, M.D., and Duwadi, S.R., Two Test Level 4 Bridge Railing and Transition Systems for Transverse Timber Deck Bridges, Transportation Research Record No. 1696, Transportation Research Board, Washington, D.C., Fowler, M.D., Design and Testing of a Test Level 4 Bridge Railing for Transverse Glulam Timber Deck Bridges, M.S. Thesis, University of Nebraska-Lincoln, May Baxter, J.R., Federal Highway Administration (FHWA) NCHRP Report No. 350 Acceptance Letter B-138 of TL-2 and TL-4 Railings and Transitions for Timber Bridge Decks, To R. K. Faller, Midwest Roadside Safety Facility, August 4, Faller, R.K., Reid, J.D., Rohde, J.R., Sicking, D.L., and Keller, E.A., Two Approach Guardrail Transitions for Concrete Safety Shape Barriers, Final Report to the Midwest States Regional Pooled Fund Program, Research Report No. TRP , Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, May 15, Faller, R.K., Reid, J.D., and Rohde, J.R., Approach Guardrail Transition for Concrete Safety Shape Barriers, Transportation Research Record No. 1647, Transportation Research Board, Washington, D.C., November Horne, D.A., Federal Highway Administration (FHWA) NCHRP Report No. 350 Acceptance Letter B-47 of Steel Post and Wood Post Thrie Beam Transitions to Concrete Parapets, To R. K. Faller, Midwest Roadside Safety Facility, March 6, Horne, D.A., Federal Highway Administration (FHWA) NCHRP Report No. 350 Conditional Acceptance Letter B-47A of Steel Post and Wood Post Thrie Beam

24 Lechtenberg, Mongiardini, Rosenbaugh, Faller, Bielenberg, and Albuquerque 24 Transitions to Vertical Wall, To D. L. Sicking, Midwest Roadside Safety Facility, May 28, 1999.