BACKGROUND
    In 1989, Texas Transportation Institute developed the single-slope barrier for the State Department of Highways and Public Transportation [1]. The development of the single-slope barrier was done under the NCHRP Report 230 criteria. Several full-scale crash tests were performed to evaluate the performance of the barrier. Two of the tests were conducted for use of the barrier in a temporary installation. Thirty-foot, free-standing single-slope barrier segments were connected using the rebar-grid connection in one test and the angle-splice connection in another test. The barriers were impact with a 2043-kg vehicle at 62 mph and 15-degree impact angle in both the tests. The barrier deflection in the test with the rebar-grid connection and the angle-splice connection was six and seven inches, respectively. Two more tests were performed with the single-slope barrier segments in a permanent configuration. In these tests, the 30-ft. barrier segments were keyed into a 1-inch thick asphalt layer. The asphalt layer was 5-ft. wide on the field side of the barrier and 1-ft. wide on the traffic side. Rebar-grid connection was used to connect adjacent barriers. The connection was grouted with cement to provide more resistance to barrier movement. The barrier was impacted with a 2043-kg vehicle at 63.1 mph and 26.5 degrees in one test, and with a 817-kg vehicle at 60.7 mph and 19.9 degrees in another test. No barrier deflection was observed in both tests.
    It should be noted that the impact severity, which is the measure of the kinetic energy being imparted on the barrier laterally, was significantly less in the tests with free-standing barriers when compared to the impact severity prescribed in the update to the NHCRP Report 350 criteria. The impact angle in the updated NCHRP Report 350 is 25-degrees, which increases the impact severity by 2.67 times. Thus higher deflections should be expected for the same barrier system if crash tested under the updated NCHRP Report 350 criteria. It should also be noted that most pooled fund states use barrier segments that are less than 30-feet in length. Reducing segment length is also expected to increase barrier deflection as shorter segments have lower mass and allow more relative rotation due to the presence of more connections.
    In 1987, Graham et al. conducted a survey of different states on the use of portable concrete barriers [2]. A preliminary analysis of some of the commonly used barrier connections was also performed.

OBJECTIVE
    Develop and test a design for a rigid concrete barrier that can be placed in front of slopes as steep as 1.5H:1V or on top of an MSE wall. The design would not require a moment slab and minimize the amount of widening behind the barrier. The design feature would be a concrete barrier with a single slope face. It should be assumed that the barrier would be precast with a 20 ft maximum length. Adjacent barrier segments will be connected using the grouted rebar-grid connection. The barrier would be tested using the proposed update to the NCHRP Report 350 criteria.

BENEFITS
    This project would develop a practical design that meets the new crash test criteria. It will save considerable cost compared to a moment slab or added widening.

IMPLEMENTATION
    TTI will provide all the supporting information and written discussion for submitting a request to FHWA for acceptance of the concrete barrier system for use on the National Highway System.
    The research will provide information and documentation on testing of the concrete barrier system so that design and operational standards can be further reviewed and evaluated. Detailed engineering drawings that will facilitate development of standards sheets and specifications will be provided to each participating state.
    Drawings provided for Task Force 13 documents will further support implementation of the research.

WORK PLAN
    TTI researchers will evaluate the performance of the grouted rebar-grid connection to design a single-slope barrier that can be used in front of a slope as steep as 1.5H:1V or on top of an MSE wall. The grouted rebar-grid connection will be evaluated by conducting a 90-degree surrogate bogie vehicle impact test of a single grouted rebar-grid joint with two single-slope barrier segments. The response of the grouted rebar-grid connection will then be incorporated into a full scale finite element model of a single-slope barrier system with 20-ft barrier segments. Simulation analysis will be performed to evaluate the performance of the barrier system under the updated NCHRP Report 350 TL-3 impact conditions. If the simulation results indicate an acceptable overall deflection of the barrier system, the researchers will perform a full scale crash test to verify simulation results. A successful test would imply that the 20-ft single-slope concrete barrier system with grouted rebar grid connection can be used in front of slopes and on MSE walls without any embedment. If however the simulation results indicate that the free-standing single-slope barrier will result in unacceptable deflections, further component level testing and simulation analysis will be performed prior to the full-scale crash test. In this scenario, quasi static push tests will be performed to determine the resistance force offered by different depths of soil layers behind the barrier. The force-deflection response measured from these tests will be incorporated into the full-scale finite element model to determine the performance of the single-slope barriers system when embedded in various depths of soil. Following are the details of the proposed work plan broken into various tasks.

Task 1: Evaluation of Grouted Rebar-Grid Connection
    Single-slope barrier with grouted rebar-grid connection is generally used in permanent applications where the barrier is keyed into a layer of asphalt. Prior testing has shown that such permanent configuration behaves rigidly with no barrier deflection [1].The performance of the barrier in a situation when it is free standing (i.e. without keying into asphalt layer) is not known. The deflection of the barrier will depend on the strength of the grouted rebar-grid connection. Since it is difficult to analytically determine the response of the grouted rebar-grid connection under impact load, the researchers will conduct a component level bogie impact test to evaluate the strength of a grouted rebar-grid connection in a 20-ft single slope barrier. The setup of the test is shown in figure 1.

Figure 1: Setup for bogie impact testing.

    Two barrier segments will be used in the testing with a 5000-lb bogie impacting the grouted rebar-grid connection as shown in figure 1. An obstruction will be used at the far ends of the barrier such that the barrier-ends away from the rebar-grid connection can rotate without deflecting laterally. The bogie vehicle will be instrumented with uniaxial accelerometers to obtain acceleration versus time data for the impact event. From this data, force-deflection relationships, ultimate dynamic load capacity, and total energy dissipated during the impact will be computed. The response of the rebar grid connection determined in the testing will be used to calibrate the response of the grouted rebar-grid connection in the full scale finite element model.

Task 2: Simulation Analysis of Free-Standing Barrier
    This task will be divided into two sub-tasks as follows.
Task 2-A: Simulation of the Bogie Test
    A finite element model of the system used for conducting the bogie test will be developed using the commercially available LS-DYNA code. LS-DYNA is a general-purpose, explicit finite element code used to analyze the nonlinear dynamic response of three-dimensional structures. Previous crash-testing of the single-slope barrier has indicated that concrete damage in the vicinity of the joint was not significant [1]. Therefore the barrier segments in the finite element model will be modeled using most rigid material representation. Concrete failure will thus not be included in the model. The model of the barrier segments will incorporate the slots that host the grouted rebar-grid. However, instead of incorporating an explicit model of the grouted rebar-grid, the slots will be filled with an elastic material having some initial properties. A bogie vehicle impact will be simulated using a previously developed model of the TTI’s surrogate bogie vehicle. The mass and the velocity of the bogie will be matched to that of the crash test and the acceleration versus time response will be measured using simulation results. The response will be compared to that obtained in the actual bogie test. Based on this comparison, the material properties of the elastic material used to fill the slots will be calibrated. This process will repeated until reasonable correlation has been obtained between the test and simulation results.
Task 2-B: Full-scale Vehicle Impact Simulation
    Once the properties of the barrier connection in bogie impact model has been calibrated to match the response of the grouted rebar-grid connection, the model will be expanded to develop the full barrier system model comprising of five 20-ft single-slope barrier segments with a total length of 100 feet. A vehicle impact simulation will then be performed using the updated NCHRP Report 350 test-level 3 impact conditions (i.e. 25-degree impact with 2270-kg pickup truck at 100 km/h). The deflection of the barrier will be evaluated from simulation results. If the simulation results indicate very insignificant or no deflection, the researchers will skip Tasks 3 and 4 and move on to Task 5, which is to conduct a full-scale crash test to verify simulation results. If however the simulation results indicate high barrier deflection, the researchers will continue with tasks 3 and 4 as described below. The researchers will convey simulation results to the technical representative and seek approval before conducting the crash test.
    It should be noted that in the update to the NCHRP Report 350, the pickup truck design vehicle has been changed from the 2000-kg, ¾-ton, standard cab pickup truck to a 2270-kg, ½-ton, 4-door pickup truck. While a public domain finite element model of the 2000-kg pick up truck is available to the researchers in the roadside safety community, no such model is available for the 2270-kg, ½-ton, 4-door pickup truck. The FHWA has recently funded development of a finite element model for a ½-ton, Chevy Silverado, 4-door, pickup truck, which meets the design test vehicle requirements of the update to NCHRP Report 350. However, the work on developing this model is underway and may not complete within the timeframe of this research project.
    In the past, TTI researchers have made simple modifications to the existing 2000-kg pickup truck model for use in studies where the dynamics of the vehicle were not very critical compared to the higher mass of the updated NCHRP Report 350 vehicle. These modifications incorporate increasing the mass of the existing 2000-kg pickup truck model by distributing additional mass over different parts of the vehicle and bringing the total vehicle mass to 2270-kg.
    In the absence of a pickup truck model that meets the updated NCHRP Report 350 criteria, the researchers will employ the above mentioned methodology of using a surrogate 2270-kg pickup truck model for evaluating the single-slope barrier design. Doing so will enable the researchers to impart the same level of impact energy into the barrier system as is required by the update NCHRP Report 350. It is expected that vehicle dynamics response of the surrogate 2270-kg vehicle will not match the response measured in the crash test. However, previous testing of the single-slope barrier has shown that the vehicle remains fairly stable during the impact [1]. Thus the vehicle dynamic characteristics are not as critical and accounting for the increased vehicle mass is expected to enable a successful evaluation of the barrier system for the updated NCHRP Report 350 criteria.

Task 3: Static Testing for Evaluation of Soil Resistance
    If the simulation results in Task 2 indicate that the free standing single slope barrier with grouted rebar-grid connection results in deflections that are not acceptable, the researchers will evaluate embedding the barrier in soil to reduce barrier deflection.
    To incorporate soil resistance to barrier movement in the simulation model, the researchers will determine the resistance force offered by different depths of compacted soil layers behind the barrier using quasi-static testing. The setup for this quasi-static testing is shown in figure 2. A hydraulic cylinder will be used to push on a 2-ft wide and a 15-inch tall rigid frame that will match the profile of the single-slope barrier. The soil in front of the rigid frame will be compacted in 1-ft wide and three, six, and ten inch deep layers. Thus a total of three quasi-static tests will be conducted to evaluate soil resistance for a range of soil depths. A 1.5H:1V slope will be constructed behind the 1-ft wide soil layer as shown in figure 2. A load cell will be used in conjunction with the hydraulic cylinder to measure the force-deflection response of different soil depths.

Figure 2: Setup for quasi-static testing to determine soil resistance.

Task 4: Simulation Analysis with Embedded Barrier
    The force-deflection response of the soil layer measured in Task 3 will be used to incorporate soil resistance in the finite element model of the single-slope barrier system developed in Task 2. The soil force will be applied by modeling non-linear springs along the base of the barrier on the field side. These springs will be assigned force-deflection properties measured in Task 3. Initially a simulation will be performed using the force-deflection properties of the 3-inch deep soil layer. If the deflections are still large, the force-deflection properties of the 6-inch and 10-inch soil layers will be subsequently used to reduce deflection by increasing soil resistance.

Task 5: Full-Scale Crash Testing and Final Report
    The researchers will perform test 3-11 of updated NCHRP Report 350 (2270-kg vehicle, 100 km/hr, 25 deg) on the design finalized from either Task 2 or Task 4. The test will be performed to verify simulation results. It is argued that this is the critical test for this design and the test with smaller 1100-kg is not needed. Due to higher impact energy, the test with the 2270 kg pickup truck will result in greater lateral deflection and help evaluate connection strength and the tendency of the barriers to rotate. An impact resulting from the lighter 1100-kg passenger car under same impact speed and angle will not result in any increase in lateral deflection of the barrier nor will it impart a higher force on the barrier to evaluate connection strength and barrier rotation. Thus the test will be conducted with the 2270-kg pickup only.
    TTI will provide the test facility, test vehicle, instrumentation of the vehicle, high-speed film, video, still photographs, and a final report suitable for submittal to Federal Highway Administration (FHWA).

 
Final Report on the Embedded Barrier in Front of Slopes or on MSE Walls
 
Quarterly Progress Reports:
December 2008 Progress Report
October 2008 Progress Report
June 2008 Progress Report
March 2008 Progress Report

____________________
[1] Beason, W.L., Ross, H.E. Jr., Perera, H.S., Campise, W., and Bullard, D.L., Jr. (1989). “Development of a Single-slope Concrete Median Barrier.” Texas Transportation Institute, Texas.
[2] Graham, J.L., Loumiet, J.R., and Migletz, J. (1987). “Portable Concrete Barrier Connectors.” Graham-Migletz Enterprises, Inc., Missouri.