Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance (2022)

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.

19   Introduction The computer simulation analyses described here were conducted to help update the AASHTO LRFD (6) and the AASHTO Roadside Design Guide (15) for bridge rail geometric design. The FE simulations were conducted to identify potential snagging concerns for bridge rails with different geometric rail and post designs. Specifically, a matrix of simulations was performed to update the bridge rail geometric design criteria, which consists of post setback distance, vertical clear opening, and ratio of rail contact width to rail height. Intermediate Validation Pickup Truck Vehicle Model A computer model of a 2007 Chevrolet Silverado pickup truck was used to represent a MASH pickup truck, which was used for impact computer simulations against various bridge rail sys- tems. The original vehicle model was developed by the Center for Collision Safety and Analysis (CCSA). Over the years, Texas A&M Transportation Institute (TTI) has made small modifica- tions to the vehicle model to improve accuracy and robustness. A detailed tire model was also included in the vehicle model because there was expected to be significant interaction of the tire with various elements of the bridge rail models. An isolated model of the vehicle tire was previously validated against FMVSS139 plunger and bead unseating tests (12). It was necessary to perform an initial validation of the full vehicle model with the detailed tire model against full- scale MASH Test 3-11 crash tests with bridge rail systems. A full verification and validation of the vehicle model was performed after the crash tests were conducted. Two systems were selected for the initial validation of the vehicle model. Figure 2.1 shows the TxDOT T224 and T1W bridge rails that were previously tested according to MASH Test 3-11 (10, 11). The two systems repre- sented different bridge rail profiles, which would validate the performance of the vehicle model against two different bridge rail types. The TxDOT T224 bridge rail is similar to a closed-profile concrete bridge rail with small openings (i.e., concrete post-and-beam bridge rail), and the T1W bridge rail is similar to an open-profile metal bridge rail with a concrete curb and large openings (i.e., curb-mounted metal post-and-beam bridge rail). FE models of the two bridge rail systems were developed for the intermediate validation simu- lations. Figures 2.2 and 2.3 show the FE computer models of the TxDOT T224 and T1W bridge rails. The details of the computer model systems replicated the details of the tested bridge rail systems. For both systems, the bridge rail components were modeled as a representative rigid material. In both tests of the systems, the bridge rails had minimal dynamic deflection. C H A P T E R 2 Preliminary Evaluation of AASHTO Geometric Curves

20 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Figure 2.3. TxDOT T1W FE computer model. Figure 2.1. TxDOT T224 ( left) and TxDOT T1W (right) bridge rails. Figure 2.2. TxDOT T224 FE computer model. Simulations were first performed with the pickup truck impacting the TxDOT T224 bridge rail. The simulation impact speed and angle were 64.3 mph and 24.8 degrees, respectively. The impact conditions matched the full-scale crash test impact conditions. Figure 2.4 shows a com- parison of sequential frames from the simulation and full-scale crash test. The vehicle’s x, y, and z accelerations and roll, pitch, and yaw rates from the computer simulation were compared to the full-scale crash test. A quantitative comparison was performed using the roadside safety verification and validation program (RSVVP). Figure 2.5 shows the results of the multichannel evaluation. All metrics were considered a pass.

Preliminary Evaluation of AASHTO Geometric Curves 21   0.000 s 0.086 s 0.171 s 0.257 s Figure 2.4. Sequential photographs of T224 FE simulation and full-scale crash test. (continued on next page)

22 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0.340 s 0.426 s 0.512 s 0.597 s Figure 2.4. (Continued).

Figure 2.5. RSVVP multichannel evaluation for TxDOT T224 crash test and simulation.

24 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Table 2.1 shows a comparison of the occupant risk metrics, which are the occupant impact velocity (OIV), ridedown acceleration (RDA), roll angle, pitch angle, and yaw angle. The occu- pant risk values were similar between the computer simulation and full-scale crash test. The only occupant risk metric with a significant difference between the full-scale crash test and the simu- lation was the lateral RDA. The simulation predicted a value of 17.8 g’s, which was about 5 g’s higher than the crash test. This discrepancy has occurred in other simulations with the vehicle model impacting rigid barriers. It has been attributed to a stiffer rear suspension system in the vehicle model in comparison to the actual vehicles used for crash tests. Simulations were also performed with the pickup truck impacting the TxDOT T1W bridge rail. The simulation impact speed and angle were 62.0 mph and 25.0 degrees, respectively. The impact conditions matched the full-scale crash test impact conditions. Figure 2.6 shows a com- parison of sequential photographs from the simulation and full-scale crash test. Full-Scale Crash Test Computer Simulation Maximum Limit Longitudinal OIV (m/s) 12.19 Lateral RDA (g’s) −12.2 −17.8 20.49 Roll (deg.) 6.2 7.8 75 Pitch (deg.) 6.6 Lateral OIV (m/s) 12.197.88.8 6.3 4.7 Longitudinal RDA (g’s) −7.5 20.49−7.1 −4.2 75 Yaw (deg.) −36.3 −31.6 - Table 2.1. T224 full-scale crash test and simulation occupant risk comparisons. 0.000 s 0.090 s Figure 2.6. Sequential photographs of T1W FE simulation and full-scale crash test.

Preliminary Evaluation of AASHTO Geometric Curves 25   0.250 s 0.340 s 0.400 s 0.170 s Figure 2.6. (Continued).

26 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance The vehicle’s x, y, and z accelerations and roll, pitch, and yaw rates from the computer simu- lation were compared to the full-scale crash test. A quantitative comparison was performed using RSVVP. Figure 2.7 shows the results of the multichannel evaluation. All metrics were considered a pass. Table 2.2 shows a comparison of the occupant risk metrics. The occupant risk metrics are the OIV, RDA, roll angle, pitch angle, and yaw angle. The occupant risk values were largely similar between the computer simulation and full-scale crash test, except the longitudinal RDA and roll angle. For the longitudinal RDA, the computer simulation predicted a value approximately 5 g’s lower than the full-scale crash test. In the full-scale crash test, the pickup truck tire seemed to engage the curb and steel post slightly more than in the simulation, resulting in higher longitu- dinal RDA. The roll angle was higher in the computer simulations than in the full-scale crash test. This was also true in the T224 validation simulation, but it was closer to the full-scale crash test. The research team did not observe any distinct interactions in the simulation that may have caused this higher roll angle. It was likely due to a combination of differing vehicle suspension systems and no concrete damage or rail deformation. The RDA and roll angle values themselves were not near the MASH limits, so the research team did not consider this difference to be a significant issue with the pickup truck computer model. Overall, the pickup truck vehicle model with a detailed tire model performed adequately in comparison to the full-scale MASH Test 3-11 crash tests performed on the TxDOT T224 and T1W bridge rails. This vehicle model was used for the simulation cases described later in this chapter. Small Car Vehicle Model A computer model of a 2010 Toyota Yaris sedan was used to represent a MASH small car, which was used for impact computer simulations against various bridge rail systems. The original vehicle model was developed by the CCSA. Over the years, TTI has made small modifications to the vehicle model to improve the accuracy and robustness of the model. A detailed tire model was also included in the vehicle model because there was expected to be significant interaction of the tire with various elements of the bridge rail models. An isolated model of the vehicle tire was previously validated against FMVSS139 plunger and bead unseating tests (12). It was neces- sary to perform an initial validation of the full vehicle model with the detailed tire model against full-scale MASH Test 3-10 crash tests with bridge rail systems. A full verification and validation of the vehicle model was performed after the full-scale crash tests were conducted. Two systems were selected for the initial validation of the vehicle model. Figure 2.8 shows the TxDOT T224 and T1W bridge rails that were previously tested according to MASH Test 3-10 (10, 11). The two systems represented different bridge rail profiles, which would validate the performance of the vehicle model against two different bridge rail types. The TxDOT T224 bridge rail is similar to a closed-profile concrete bridge rail with small openings (i.e., concrete post-and-beam bridge rail), and the T1W bridge rail is similar to an open-profile metal bridge rail with a concrete curb and large openings (i.e., curb-mounted metal post-and-beam bridge rail). FE models of the two bridge rail systems were developed for the intermediate validation simu- lations. Figures 2.9 and 2.10 show the FE computer models of the TxDOT T224 and T1W bridge rails. The details of the computer model systems replicated the details of the tested bridge rail systems. For both systems, the bridge rail components were modeled as a representative rigid material. In both tests of the systems, the bridge rails had minimal dynamic deflection.

Figure 2.7. RSVVP multichannel evaluation for TxDOT T1W crash test and simulation.

28 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Full-Scale Crash Test Computer Simulation Maximum Limits Roll (deg.) −15.0 −22.5 75 Pitch (deg.) Longitudinal OIV (m/s) 6.6 12.195.9 Lateral RDA (g’s) 11.8 20.499.4 4.3 Lateral OIV (m/s) −7.8 12.19−7.3 Longitudinal RDA (g’s) −10.1 20.49−4.5 −5.6 75 Yaw (deg.) 37.1 32.6 - Table 2.2. T1W full-scale crash test and simulation occupant risk comparisons. Figure 2.8. TxDOT T224 ( left) and TxDOT T1W (right) bridge rails. Figure 2.9. TxDOT T224 FE computer model. Figure 2.10. TxDOT T1W FE computer model.

Preliminary Evaluation of AASHTO Geometric Curves 29   Simulations were first performed with the small car impacting the TxDOT T224 bridge rail. The simulation impact speed and angle were 62.6 mph and 25.1 degrees, respectively. The impact conditions matched the full-scale crash test impact conditions. Figure 2.11 shows a comparison of sequential frames from the simulation and full-scale crash test. The vehicle’s x, y, and z accelerations and roll, pitch, and yaw rates from the computer simula- tion were compared to the full-scale crash test. A quantitative comparison was performed using RSVVP. Figure 2.12 shows the results of the multichannel evaluation. All metrics were consid- ered a pass. Table 2.3 shows a comparison of the occupant risk metrics. The occupant risk values were similar between the computer simulation and full-scale crash test, except the longitudinal RDA. The computer simulation predicted a value approximately 10 g’s lower than the full-scale crash test. After extensive review of the crash test videos of the small car impacting the TxDOT T224, it was determined that the high longitudinal RDA was due to the vehicle tire and bumper inter- action with the opening in the bridge rail. Due to dust and debris, it was difficult to determine the exact nature of the interaction and what parts may be snagging. The high longitudinal RDA value from the crash test is not a typical occurrence for small car vehicles impacting rigid bridge rail systems. As such, the vehicle model was deemed appropriate for intermediate validation. Simulations were also performed with the small car impacting the TxDOT T1W bridge rail. The simulation impact speed and angle were 61.9 mph and 24.8 degrees, respectively. The impact conditions matched the full-scale crash test impact conditions. Figure 2.13 shows a comparison of sequential photographs from the simulation and full-scale crash test. 0.000 s 0.100 s Figure 2.11. Sequential photographs of T224 FE simulation and full-scale crash test. (continued on next page)

30 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0.400 s 0.450 s 0.200 s 0.300 s Figure 2.11. (Continued).

Figure 2.12. RSVVP multichannel evaluation for TxDOT T224 crash test and simulation.

32 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Full-Scale Crash Test Computer Simulation Longitudinal OIV (m/s) 9.6 7.6 Lateral OIV (m/s) 9.2 9.0 Longitudinal RDA (g’s) Lateral RDA (g’s) −14.1 −10.1 Roll (deg.) 7.3 Pitch (deg.) −7.3 −2.7 −3.5 −8.6 −5.9 Yaw (deg.) −52.4 −42.0 Table 2.3. T224 full-scale crash test and simulation occupant risk comparisons. 0.000 s 0.100 s Figure 2.13. Sequential photographs of T1W FE simulation and full-scale crash test. The vehicle’s x, y, and z accelerations and roll, pitch, and yaw rates from the computer simula- tion were compared to the full-scale crash test. A quantitative comparison was performed using RSVVP. Figure 2.14 shows the results of the multichannel evaluation. All metrics were consid- ered a pass. Table 2.4 shows a comparison of the occupant risk metrics. The occupant risk values were similar between the computer simulation and full-scale crash test, except the lateral RDA. The computer simulation predicted a value approximately 5 g’s higher than the full-scale crash test. This was likely due to a combination of the vehicle having a slightly smaller roll angle in the simulation and the vehicle computer model suspension being somewhat stiffer than the actual

0.300 s 0.400 s 0.500 s 0.200 s Figure 2.13. (Continued).

Figure 2.14. RSVVP multichannel evaluation for TxDOT T1W crash test and simulation.

Preliminary Evaluation of AASHTO Geometric Curves 35   vehicle. It was not deemed necessary to improve the vehicle model further based on the overall behavior of the vehicle and other occupant metrics being similar to the full-scale crash test. The vehicle model seemed to be slightly conservative in its prediction of lateral RDA. Overall, the small car vehicle model with a detailed tire model performed adequately in com- parison to the full-scale MASH Test 3-10 crash tests performed on the TxDOT T224 and T1W bridge rails. This vehicle model was used for the simulation cases described later in this chapter. Evaluation of AASHTO Post Setback and Snag Potential Relationships Evaluation Method This section details simulations conducted to analyze vehicle snagging interaction for the different bridge rail types. The bridge rail systems were configured to evaluate the current post setback and snag potential geometric relationships in AASHTO LRFD Section 13. Figure 2.15 shows the current geometric relationships and the different labeled regions. The geometric relationships may need to be modified due to change in MASH testing condi- tions. Computer simulations were used to generate information needed to evaluate the different regions for post setback and snag potential. The procedure to determine the proper location of the not recommended, marginal, and rec- ommended regions consisted of performing simulations at the current lines between the regions and analyzing the performance of the bridge rail systems at these locations. If the bridge rails Full-Scale Crash Test Computer Simulation Longitudinal OIV (m/s) 7.0 Lateral OIV (m/s) −9.5 Longitudinal RDA (g’s) −3.5 Lateral RDA (g’s) 11.0 Roll (deg.) −12.4 −7.8 Pitch (deg.) 6.0 Yaw (deg.) 7.3 −9.7 −5.4 5.7 −4.2 43.9 39.4 Table 2.4. T1W full-scale crash test and simulation occupant risk comparisons. 0 2 4 6 8 10 12 14 16 18 0 5 10 15 V er tic al C le ar O pe ni ng (i n) Post Setback Distance (in) High Snag Potential Low Snag Potential 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 5 10 15 R at io o f R ai l C on ta ct W id th to H ei gh t Post Setback Distance (in) Preferred Not Recommended Marginal Marginal Figure 2.15. AASHTO LRFD Section 13 figures A13.1.1-2 and A13.1.1-3 (6).

36 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance did not perform acceptably according to MASH evaluation criteria, then the line being evalu- ated was moved toward the opposing line. If the bridge rails performed acceptably according to MASH evaluation criteria, then the line being evaluated was adequate. In addition, if the bridge rail systems resulted in MASH OIV and RDA values below preferred MASH evaluation criteria, then the line was moved away from the opposing line. Figure 2.16 illustrates this procedure for the post setback figure not recommended/marginal line that would be evaluated. This procedure was repeated in a similar manner for the marginal/preferred line of the post setback figure and for both boundary lines on the snag potential figure. Different bridge rail types can present different snagging possibilities and produce different vehicle reactions upon impact. As such, the research team evaluated the post setback and snag potential figures with the previously described method for different bridge rail categories. These categories consisted of those previously developed by Silvestri-Dobrovolny et al. (7). The catego- ries were concrete post-and-beam, deck-mounted metal post-and-beam, curb-mounted metal post-and-beam, and parapet-mounted metal post-and-beam. The analysis per each category allowed a thorough investigation of snagging potential according to different bridge rail types. Bridge Rail Systems Evaluated The research team developed a three-dimensional concrete post-and-beam bridge rail model for each rail used in the FE simulations. The concrete post-and-beam bridge rail FE models were generated by varying the post setback distance, vertical clear opening, and rail contact width. Figure 2.17 shows a profile view of a concrete post-and-beam rail and the geometric dimensions that were varied for each bridge rail system. The concrete post-and-beam models were used for the FE simulations conducted to iden- tify potential snagging concerns for bridge rails with different geometric rail and post designs. Perform Simulations along Selected Not Recommended/Marginal Line Meets MASH Evaluation Criteria? Selected Not Recommended/Marginal Line Is Confirmed Move Not Recommended/Marginal Line toward Recommended/Marginal Line Yes No OIV and RDA below MASH Preferred Values? Move Not Recommended/Marginal Line Further into Not Recommended Region Yes No Figure 2.16. Method to evaluate not recommended and marginal regions for post setback criteria.

Preliminary Evaluation of AASHTO Geometric Curves 37   The concrete post-and-beam rails were modeled as rigid material. Figure 2.18 shows an FE model of one of the concrete post-and-beam bridge rail systems. The research team developed a three-dimensional metal post-and-beam bridge rail model for each rail used in the FE simulations. The metal post-and-beam bridge rail FE models were generated by varying the post setback distance, vertical clear opening, and rail contact width. Figure 2.19 shows a profile view of a deck-mounted metal post-and-beam rail and the geo- metric dimensions that were varied for each system. Figure 2.20 shows a profile view of a curb- or parapet-mounted metal post-and-beam rail and the geometric dimensions that were varied for each system. The metal post-and-beam models were used for the FE simulations conducted to identify potential snagging concerns for bridge rails with different geometric rail and post designs. The research team modeled the metal post-and-beam rails as rigid material. Figure 2.21 shows an FE model of one of the deck-mounted metal post-and-beam bridge rail systems. Figure 2.22 shows an FE model of one of the curb-mounted metal post-and-beam bridge rail systems. Fig- ure 2.23 shows an FE model of one of the parapet-mounted metal post-and-beam bridge rail systems. Each bridge rail model developed for the FE simulations corresponded to a single geometric data point on the AASHTO Post Setback vs. Ratio of Contact Width to Height figure and Post Setback vs. Vertical Clear Opening figure (see Figure 2.24). Figure 2.24 shows an example of the geometric data points of a deck-mounted metal post-and-beam system plotted on the AASHTO rail geometric figures. Figure 2.17. Profile view of a concrete post-and-beam rail showing the geometric variables. Figure 2.18. Concrete post-and-beam FE computer model.

38 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Figure 2.19. Profile view of a deck-mounted metal post-and- beam rail showing the geometric variables. Figure 2.20. Profile view of a curb- or parapet-mounted metal post-and-beam rail showing the geometric variables. Figure 2.21. Deck-mounted metal post-and-beam FE computer model. Figure 2.22. Curb-mounted metal post-and-beam FE computer model.

Preliminary Evaluation of AASHTO Geometric Curves 39   The research team selected several common dimensions for the bridge rail models. For all bridge rails, a post spacing of 10 ft. on center was used, and a total bridge rail model length of 75 ft. was selected for appropriate vehicle contact length during impact. A post size of W 6 × 25 was selected for all metal post-and-beam models since that size is commonly used for bridge rail systems and the rail sizes were determined based on the post setback distance, vertical clear opening, and rail contact width. In addition, the maximum clear opening (Cb) was selected as the bottom opening for all bridge rail systems since this opening had shown from previous test data Figure 2.23. Parapet-mounted metal post-and-beam FE computer model. Figure 2.24. Example of a deck-mounted post-and-beam system plotted on AASHTO rail geometric figures.

40 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance to be more critical in terms of vehicle snagging that can lead to high occupant risk factors. For the curb-mounted metal post-and-beam systems, a curb height of 6 in. was selected to allow for the highest probability of the vehicle tire snagging between the curb and bottom rail member. For the parapet-mounted metal post-and-beam systems, a parapet height of 18 in. was selected since this would be the minimum parapet height based on previously tested bridge rails. The research team selected the minimum parapet height to maximize the openings that could lead to vehicle snagging. These common dimensions were used for all bridge rail models. One dimension of the bridge rail systems that required investigation was the height. It was necessary to determine if shorter or taller bridge rail systems would result in a higher probability of vehicle snagging and higher occupant risk values. To investigate this, computer simulations were performed with two rigid deck-mounted metal post-and-beam bridge rail systems. One bridge rail system had a total height of 30 in., the other a total height of 42 in. Figure 2.25 shows both bridge rail systems that were used for the computer simulations. Both bridge rail models corresponded to the same geometric data point plotted in Figure 1.2. Other than the difference in height and extra rail, both bridge rail models were identical. Only MASH Test 3-10 was performed. MASH Test 3-11 is primarily performed to evaluate the struc- tural adequacy of the bridge rail systems; with the two systems being rigid, it was not beneficial to perform simulations with the pickup truck. The occupant risk evaluations of the two simulations are shown in Table 2.5. The simulations result in very similar occupant risk values. However, the simulation with the 30-in. bridge rail system resulted in a lateral RDA of 16.8 g’s, while the simula- tion with the 42-in. bridge rail system resulted in a lateral RDA of 11.1 g’s. Therefore, the shorter bridge rail system was deemed more critical for occupant risk. For most of the simulations, a bridge rail height between 29 and 33 in. was used. Some of the bridge rail systems required a taller bridge rail system in order to plot at specific locations on the two geometric figures. (a) 30-in. bridge rail system (b) 42-in. bridge rail system Figure 2.25. Comparison of 30- and 42-in. bridge rail systems. 30-in. Bridge Rail Height 42-in. Bridge Rail Height Longitudinal OIV (m/s) 1.0 2.4 Lateral OIV (m/s) 11.0 11.1 Longitudinal RDA (g’s) Lateral RDA (g’s) −16.8 −11.1 Roll (deg.) 11.6 13.6 Pitch (deg.) −5.0 −7.2 −6.2 −7.4 Yaw (deg.) 41.9 32.4 Table 2.5. Occupant risk values for 30- and 42-in. bridge rail computer simulations.

Preliminary Evaluation of AASHTO Geometric Curves 41   0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Simulation Cases Post Setback Distance (in.) Ra tio o f R ai l C on ta ct W id th to H ei gh t Figure 2.26. Post setback simulation cases for concrete post-and-beam bridge rails. 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Simulation Cases Post Setback Distance (in.) Ve rti ca l C le ar O pe ni ng (i n. ) Figure 2.27. Snag potential simulation cases for concrete post-and-beam bridge rails. With this information gathered for the various dimensions, the configurations for the bridge rail systems were determined. Figures 2.26 and 2.27 show the bridge rail configurations evaluated for concrete post-and- beam geometrics. Due to constraints with some of the dimensions, it was not possible to use bridge rail systems that would plot along the bottom curve of Figure 2.26. The points plotted in the figure show realistic bridge rail configurations for the different post setback values.

42 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Figures 2.28 and 2.29 show the bridge rail configurations evaluated for deck-mounted metal post-and-beam geometrics. Figures 2.30 and 2.31 show the bridge rail configurations evaluated for curb-mounted metal post-and-beam geometrics. Figures 2.32 and 2.33 show the bridge rail configurations evaluated for parapet-mounted metal post-and-beam geometrics. Due to constraints with some of the dimensions, it was 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Simulation Cases Post Setback Distance (in.) Ra tio o f R ai l C on ta ct W id th to H ei gh t Figure 2.28. Post setback simulation cases for deck-mounted metal post-and-beam bridge rails. 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Simulation Cases Post Setback Distance (in.) Ve rti ca l C le ar O pe ni ng (i n. ) Figure 2.29. Snag potential simulation cases for deck-mounted metal post-and-beam bridge rails.

Preliminary Evaluation of AASHTO Geometric Curves 43   not possible to use bridge rail systems that would plot along the bottom curve of Figure 2.32. The points plotted in the figure show realistic bridge rail configurations for the different post setback values. Figures 2.34 and 2.35 show a summary of all the bridge rail configurations evaluated for rail geometrics. Appendix B provides details on the dimensions for all the bridge rail systems. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Simulation Cases Post Setback Distance (in.) Ra tio o f R ai l C on ta ct W id th to H ei gh t Figure 2.30. Post setback simulation cases for curb-mounted metal post-and-beam bridge rails. 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Simulation Cases Post Setback Distance (in.) Ve rti ca l C le ar O pe ni ng (i n. ) Figure 2.31. Snag potential simulation cases for curb-mounted metal post-and-beam bridge rails.

44 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Simulation Cases Post Setback Distance (in.) Ra tio o f R ai l C on ta ct W id th to H ei gh t Figure 2.32. Post setback simulation cases for parapet-mounted metal post-and-beam bridge rails. 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Simulation Cases Post Setback Distance (in.) Ve rti ca l C le ar O pe ni ng (i n. ) Figure 2.33. Snag potential simulation cases for parapet-mounted metal post-and-beam bridge rails.

Preliminary Evaluation of AASHTO Geometric Curves 45   0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Concrete Post and Beam Metal Post and Beam Deck- Mounted Metal Post and Beam Curb- Mounted Metal Post and Beam Parapet- Mounted Post Setback Distance (in.) Ra tio o f R ai l C on ta ct W id th to H ei gh t Figure 2.34. Post setback simulation cases for all bridge rails. 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Concrete Post and Beam Metal Post and Beam Deck- Mounted Metal Post and Beam Curb- Mounted Metal Post and Beam Parapet- Mounted Post Setback Distance (in.) Ve rti ca l C le ar O pe ni ng (i n. ) Figure 2.35. Snag potential simulation cases for all bridge rails.

46 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Computer Simulation Results FE computer simulations were performed on the bridge rail models presented in the previous section. MASH Test 3-10 and Test 3-11 impact simulations were conducted on each bridge rail system. The impact speed and impact angle were 62 mph and 25 degrees, respectively. The impact locations were 3.6 ft. and 4.3 ft. upstream of the centerline of the post for MASH Tests 3-10 and 3-11, respectively. The FE simulations were conducted to identify potential snagging concerns for bridge rails with different geometric rail and post designs. Figures 2.36 and 2.37 show sequential gut and overhead views of an FE simulation performed on the CPB-SP-System01 bridge rail. A total of 52 simulations were performed for the concrete post-and-beam bridge rails. All simulations ran without error. No modifications were made to the vehicles during the simulations. Gut View 0.000 s 0.100 s 0.200 s 0.300 s Overhead ViewTime Figure 2.36. Sequential simulation frames gut and overhead views for concrete post-and-beam MASH Test 3-10.

Preliminary Evaluation of AASHTO Geometric Curves 47   0.500 s 0.600 s 0.400 s Gut View Overhead ViewTime Figure 2.36. (Continued). 0.000 s 0.100 s Gut View Overhead ViewTime Figure 2.37. Sequential simulation frames gut and overhead views for concrete post-and-beam MASH Test 3-11. (continued on next page)

48 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0.500 s 0.600 s 0.200 s 0.300 s 0.400 s Gut View Overhead ViewTime Figure 2.37. (Continued).

Preliminary Evaluation of AASHTO Geometric Curves 49   Figures 2.38 and 2.39 show sequential gut and overhead views of an FE simulation performed on the MPBD-SP-System01 bridge rail. A total of 48 simulations were performed for the deck- mounted metal post-and-beam rails. All simulations ran without error. No modifications were made to the vehicles during the simulations. Figures 2.40 and 2.41 show sequential gut and overhead views of an FE simulation performed on the MPBP-SP-System01 bridge rail. A total of 48 simulations were performed for the parapet- mounted metal post-and-beam bridge rails. All simulations ran without error. No modifications were made to the vehicles during the simulations. Figures 2.42 and 2.43 show sequential gut and overhead views of an FE simulation per- formed on the MPBC-SP-System01 bridge rail. A total of 48 simulations were performed for the curb-mounted metal post-and-beam bridge rails. All the MASH Test 3-11 simulations ran without error. No modifications were made to the pickup truck vehicle during the simulations. 0.000 s 0.100 s 0.200 s 0.300 s Gut View Overhead ViewTime Figure 2.38. Sequential simulation frames gut and overhead views for deck-mounted metal post-and-beam MASH Test 3-10. (continued on next page)

50 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0.500 s 0.600 s 0.400 s Gut View Overhead ViewTime Figure 2.38. (Continued). 0.000 s 0.100 s Gut View Overhead ViewTime Figure 2.39. Sequential simulation frames gut and overhead views for deck-mounted metal post-and-beam MASH Test 3-11.

Preliminary Evaluation of AASHTO Geometric Curves 51   0.500 s 0.600 s 0.200 s 0.300 s 0.400 s Gut View Overhead ViewTime Figure 2.39. (Continued).

52 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Gut View 0.000 s 0.100 s 0.200 s 0.300 s 0.400 s Overhead ViewTime Figure 2.40. Sequential simulation frames gut and overhead views for parapet-mounted metal post-and-beam MASH Test 3-10.

Preliminary Evaluation of AASHTO Geometric Curves 53   0.500 s 0.600 s Gut View Overhead ViewTime Figure 2.40. (Continued). 0.000 s 0.100 s 0.200 s Gut View Overhead ViewTime Figure 2.41. Sequential simulation frames gut and overhead views for parapet-mounted metal post-and-beam MASH Test 3-11. (continued on next page)

54 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0.500 s 0.600 s 0.300 s 0.400 s Gut View Overhead ViewTime Figure 2.41. (Continued).

Preliminary Evaluation of AASHTO Geometric Curves 55   0.000 s 0.100 s 0.200 s 0.300 s 0.400 s Gut View Overhead ViewTime Figure 2.42. Sequential simulation frames gut and overhead views for curb-mounted metal post-and-beam MASH Test 3-10. (continued on next page)

56 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0.500 s 0.600 s Gut View Overhead ViewTime Figure 2.42. (Continued). 0.000 s 0.100 s 0.200 s Gut View Overhead ViewTime Figure 2.43. Sequential simulation frames gut and overhead views for curb-mounted metal post-and-beam MASH Test 3-11.

Preliminary Evaluation of AASHTO Geometric Curves 57   0.500 s 0.600 s 0.300 s 0.400 s Gut View Overhead ViewTime Figure 2.43. (Continued). However, errors were encountered during the MASH Test 3-10 simulations. The small car vehicle tire was pushing through the opening between the bottom rail and the concrete curb and catch- ing on the steel post, causing the tire parts to snag and reach numerical instability. To combat this issue a few modifications were made to the vehicle model. First, the rubber material model for the vehicle tire was changed from MAT_ELASTIC to MAT_SIMPLIFIED_RUBBER/FOAM. This modification was made to use a slightly more robust material model. Second, material failure was added for tire and suspension parts experiencing high plastic strain. Most of these parts were located near the joints connecting the tire to the vehicle. A few additional changes were made to improve contact interaction between the vehicle and bridge rail system. After the modifications were made to the small car vehicle model, the MASH Test 3-10 simulations were completed without error.

58 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance Overall, a total of 196 simulations were conducted. No modifications were made to the pickup truck vehicle model during the simulation runs. Modifications were made to the small car vehicle model during the last set of simulations for the curb-mounted metal post-and-beam bridge rail systems. After completing the computer simulations for all the bridge rail systems, the results for each system were plotted on the snag potential and post setback figures. Figures 2.44 and 2.45 show 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Pass Max OIV/RDA Pass Preferred OIV/RDA Ve rti ca l C le ar O pe ni ng (i n. ) Post Setback Distance (in.) Figure 2.44. Snag potential simulation results for concrete post-and-beam bridge rails. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Pass Max OIV/RDA Pass Preferred OIV/RDA Fail OIV/RDA Ra tio o f R ai l C on ta ct W id th to H ei gh t Post Setback Distance (in.) Figure 2.45. Post setback simulation results for concrete post-and-beam bridge rails.

Preliminary Evaluation of AASHTO Geometric Curves 59   the results for the concrete post-and-beam bridge rail simulations. One simulation did result in a failed occupant risk metric: the lateral RDA for one of the bridge rail systems was 20.7 g’s in the MASH Test 3-10 simulation, which exceeded the 20.49 g MASH limit. The value did not exceed the limit by a significant margin, and it was noted in the intermediate validation that the small car seemed to be conservative and overpredicted the lateral RDA. Thus, the research team did not deem it necessary to adjust the geometric curves and reevaluate new ones based on the single simulation OIV failure. All other bridge rail configurations except two resulted in occu- pant risk values below the MASH maximum limit, but not below the preferred limit. Therefore, no additional simulations were conducted to evaluate new curves. The current geometric curves are adequate for concrete post-and-beam bridge rail systems. Figures 2.46 and 2.47 show the results for the deck-mounted metal post-and-beam bridge rail simulations. One simulation did result in a failed occupant risk: the lateral RDA for one of the bridge rail systems was 22.7 g’s in the MASH Test 3-10 simulation, which exceeded the 20.49 g MASH limit. The value did not exceed the limit by a significant margin, and it was noted in the intermediate validation that the small car seemed to be conservative and over-predict the lateral RDA. Thus, the research team did not deem it necessary to adjust the geometric curves and reevaluate new ones based on the single simulation RDA failure. All other bridge rail configu- rations except one resulted in occupant risk values below the MASH maximum limit, but not below the preferred limit. Therefore, no additional simulations were conducted to evaluate new curves. The current geometric curves are adequate for deck-mounted metal post-and-beam bridge rail systems. Figures 2.48 and 2.49 show the results for the curb-mounted metal post-and-beam bridge rail simulations. There were a significant number of simulations with occupant risk values exceed- ing the MASH limit. Some of these exceeded the OIV limit and some exceeded the RDA limit. According to the method outlined earlier in this chapter, additional simulations should be con- ducted with a new curve in the snag potential and post setback figures. However, due to needed validation of the small car vehicle model as discussed previously, the research team elected not to run additional simulations. Since the simulation results were validated later in Chapter 4, 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Pass Max OIV/RDA Pass Preferred OIV/RDA Fail OIV/RDA Ve rti ca l C le ar O pe ni ng (i n. ) Post Setback Distance (in.) Figure 2.46. Snag potential simulation results for deck-mounted metal post-and-beam bridge rails.

60 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Pass Max OIV/RDA Pass Preferred OIV/RDA Fail OIV/RDA Ra tio o f R ai l C on ta ct W id th to H ei gh t Post Setback Distance (in.) Figure 2.47. Post setback simulation results for deck-mounted metal post-and-beam bridge rails. 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Pass Max OIV/RDA Pass Preferred OIV/RDA Fail OIV/RDA Ve rti ca l C le ar O pe ni ng (i n. ) Post Setback Distance (in.) Figure 2.48. Snag potential simulation results for curb-mounted metal post-and-beam bridge rails.

Preliminary Evaluation of AASHTO Geometric Curves 61   0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Pass Max OIV/RDA Fail OIV/RDA Ra tio o f R ai l C on ta ct W id th to H ei gh t Post Setback Distance (in.) Figure 2.49. Post setback simulation results for curb-mounted metal post-and-beam bridge rails. the research team did not run the additional simulations to evaluate a new curve. This reduced the possibility of running more simulations that might not have been relevant and necessary after the validation effort. Figures 2.50 and 2.51 show the results for the parapet-mounted metal post-and-beam bridge rail simulations. For all bridge rail configurations except one, the occupant risk values were below the MASH maximum limit but not below the preferred limit. Therefore, no additional 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 Pass Max OIV/RDA Ve rti ca l C le ar O pe ni ng (i n. ) Post Setback Distance (in.) Figure 2.50. Snag potential simulation results for parapet-mounted metal post-and-beam bridge rails.

62 Evaluation of Bridge Rail Systems to Confirm AASHTO MASH Compliance 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Pass Max OIV/RDA Pass Preferred OIV/RDA Ra tio o f R ai l C on ta ct W id th to H ei gh t Post Setback Distance (in.) Figure 2.51. Post setback simulation results for parapet-mounted metal post-and-beam bridge rails. simulations were conducted to evaluate new curves. The current geometric curves are adequate for parapet-mounted metal post-and-beam bridge rail systems. Discussion of Results FE computer simulations were conducted to evaluate the current AASHTO geometric curves. Additional information was needed to verify the results from the 1100C impact simulations. Two full-scale crash tests were conducted to verify the behavior of the passenger car vehicle when impacting metal post-and-beam bridge rail systems with large vertical clear openings. The data from the crash tests were used to verify and validate the small car FE model. Based on the results of the computer simulations, MPBD-PS-System04 and MPBC-PS-System04 were selected for full-scale crash testing. Both systems were critical in terms of wheel snagging and occupant risk metrics (Figure 2.52). Figure 2.52. Small car tire interaction with bridge rail systems.

Preliminary Evaluation of AASHTO Geometric Curves 63   The geometry of both bridge rail systems resulted in points being plotted as shown in Fig- ure 2.53. While one system had a curb and the other was deck-mounted, both systems had the same post setback, vertical clear opening, and ratio of rail contact width to height. The bridge rails plot very near the high snag potential and not recommended regions for snag potential and post setback. Figure 2.53. Geometric plot for bridge rail systems. (a) Snag Potential (b) Post Setback 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 V er tic al C le ar O pe ni ng (i n. ) Low Snag Potential High Snag Potential 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 2 4 6 8 10 12 Ra tio o f R ai l C on ta ct W id th to H ei gh t Preferred Not Recommended Post Setback Distance (in.) Post Setback Distance (in.)