Development and research of a SS-level universal double-beam assembled bridge barrier

A wide variety of bridge barriers are used on highways. The bearing capacities of different types of deck slabs are measured in this study by applying the yield line theory to determine the worst-case scenario. An improved configuration for the worst-case barrier base and deck slab is developed, namely the universal double-beam assembled bridge barrier, which can enhance the safety performance of exiting concrete-base bridge barriers. According to the simulated impact test results, the new barrier meets SS-level requirements in terms of containment, redirective, and buffering performance as specified in the Standard for Safety Performance Evaluation of Highway Barriers (JTG B05-01-2013). The barrier structure’s compatibility with various bases is also analyzed. SS-level impact tests are conducted on real vehicles, including cars, buss, and trucks. The results show that the safety performance of the new barrier configuration reaches SS-level, and the barrier is universally compatible with concrete bases with a height of 63 cm and above.

The combination of finite element calculation and actual vehicle collision tests is an effective method for studying the protective capacity of guardrails 6 .Ray et al., summarized the design and analysis of the Annisquam River Bridge railing to satisfy the requirements of National Cooperative Highway Research Program (NCHRP) Report 350 7 Test Level 3. Design of a previously accepted Minnesota Test Level 3 bridge railing was used as the starting point.A baseline finite-element model of a crash-tested Minnesota bridge railing is developed and validated against the full-scale crash test results, using the non-linear dynamic finite-element program LS-DYNA 8 .Yao et al., presented a novel style of assembled rolling guardrail, and a vehicle-guardrail collision numerical model is built in LS-DYNA.The performance of Beam-column guardrail (B-C.G), Assembled Guardrail (A.G), and Assembled rolling guardrail (A-R.G) is evaluated 9 .
In this paper, the worst-case barrier base and bridge deck of a highway are studied, and a universal improvement scheme is developed.The novelty of this article lies in the investigation and analysis of existing bridge guardrail bases and bridge decks.The yield line theory is used to calculate the bearing capacity of different bridge guardrail bases and bridge decks, and the most unfavorable concrete bases and bridge decks are selected.Based on this, the bridge guardrail structure design method is used to carry out the improvement plan design, and the protection ability of the improvement plan is verified through finite element simulation and full-scale vehicle collision tests.

Current bridge barrier bases
There are three types of bridge barriers on the discussed highway, which are assembled barrier, concrete barrier, and corrugated beam barrier.Through a comparative study of the design drawings of the barriers and a field survey, the worst-case concrete base is determined.
According to the Appendix D of the Design Specifications for Highway Safety Facilities (JTG/T D81-2017) 10 , the bearing capacity of concrete barriers mainly comes from reinforcement and concrete, based on theoretical calculations.The theoretical calculation method was introduced in the Technical Guide for Enhancing Highway and Bridge Safety Performance 4 as a suitable approach for measuring the bearing capacity of reinforced concrete flexural members with a single-steel rectangular section, as shown in Fig. 1.
Flexural capacity of the concrete base of the bridge barrier: M-the standard value of bending moment generated by the load on this section; A s -the cross-sectional area of longitudinally stressed steel bars in the tensile zone; f yk -the standard value of yield strength of steel bars; h 0 -the effective height of the cross-section; x-the height of the compression zone calculated according to the equivalent rectangular stress graph.
The analysis shows that the bearing capacity of assembled barrier and concrete barrier is directly related to the height of concrete when there is little difference in reinforcement.The height of the concrete bases of the three types of bridge barriers on the studied highway is 81 cm, 63 cm, and 68.5 cm, respectively, as illustrated in Figs. 2 and 3.After using the rebound method to test the concrete strength and carbonization depth, the estimated minimum strength value of the guardrail base concrete is 18 MPa after conversion, as shown in Figs. 4  and 5.The reinforcement effect of a barrier is directly proportional to its height.Generally, the reinforcement area is larger for higher barriers.Thus, according to the equation, the shortest base can be regarded as the worst case, i.e. the concrete base measuring 63 cm in height.

Current deck slabs
A site inspection and review of the design drawings reveal that the deck slabs using simply supported T-beams are at least 10 cm thick.According to the calculation method for the bearing capacity of bridge decks in Appendix D of the Design Specifications for Highway Safety Facilities (JTG/T D81-2017) 10 , when a large vehicle collides Figure 1.Rectangular section with single-steel reinforcement.
with the existing bridge barriers, the barriers will bear the impact force and transmit the collision moment to the deck slab, which will greatly impact the overall safety performance of the barriers while causing irreparable damage to the bridge.By investigating various types of bridge decks of the highway, the one with the thinnest edge and the longest cantilever length is selected, as shown in Fig. 6.
According to the theoretical calculation method of yield line recommended by JTG/T D 81-2017 10 , the bearing capacity of the bridge deck was calculated, and the weakest carrying capacity case is the T-beam bridge deck with an edge thickness of 10 cm and a cantilever length of 94 cm (Table 1).

Design of the universally applicable improved configuration
Following the review of current bridge barrier bases and deck slabs, an improved configuration was developed based on the current barrier with a 63 cm-high concrete base and a T-beam bridge deck with a thickness of 10 cm and a cantilever length of 94 cm.The improved configuration has removed the upper beam and added a new beam column, forming an assembled barrier with a column spacing of 2 m, as shown in Fig. 7.The ①②③ steel bars are the hoops of the barrier, with a spacing of 150 mm.The ④ steel bar connects the ① and ② steel bars    The Design Specifications for Highway Safety Facilities (JTG D81-2017) put forward construction requirements for bridge barriers 3 .The distances between the center of the upper beam and the bridge deck, and between the center of the lower beam and the bridge deck are 1300 and 1000 mm, respectively.The beams are all made of the same material.The resistance ratio between the beams is the same as their modulus ratio.The weighted average height Y of the load-bearing beams from the bridge deck can be calculated as the weighted average height of each beam's center.The total height meets the requirement that the height of a SS-level assembled bridge barrier should be at least 120 cm.www.nature.com/scientificreports/By adding spacer blocks between the upright posts and cross beams, the upright posts' back distance is increased to 300 mm and the clearance between the cross beams increased to 180 mm, thereby reducing the possibility of the upright posts being directly impacted by wheels, bumpers, or engine covers.The total height of the beam in contact with the vehicle is 240 mm, and its ratio to the height of the column The splicing sleeve of the beam is made of the same material as the beam, with a length of 500 mm, more than twice as wide as the beam.The connecting bolts are 10.9S bolts.
The design of the universal double-beam assembled bridge barrier complies with the Design Specifications for Highway Safety Facilities (JTG/T D81-2017) 10 .

Safety performance evaluation indicators
The Standard for Safety Performance Evaluation of Highway Barriers (JTG B05-01-2013) specifies the evaluation indicators of barriers as well as experimental conditions for real-vehicle impact tests 11 .
During the verification tests of the improved bridge barrier configuration, the impact point for the small and buss and the truck shall be at 1/3 of the distance from the starting point of the standard barrier section along the driving direction.The setup of the tested barrier and the impact point is illustrated in Fig. 8.
Containment, redirective and buffering performance of the barrier are measured and analyzed based on the results of the impact tests.
Containment refers to that the barrier shall be capable of containing the vehicle from crossing through, climbing over, or riding over it after an impact.Components of the tested barrier and their breakaway parts shall not penetrate the vehicle's passenger compartment.
Buffering requires that the occupant impact velocity shall not exceed 12 m/s both longitudinally and transversely.In no case shall the longitudinal or transverse component of the post-impact acceleration exceed 200 m/ s 2 .
(2)  www.nature.com/scientificreports/ The collision speed of passengers is calculated according to Eq. ( 3): v x,y -passenger collision velocities in longitudinal (x-direction) and transverse (y-direction) directions; a x,y -acceleration at the center of gravity of the vehicle in the longitudinal (x-direction) and transverse (y-direction) directions; t * -the time when the hypothetical passenger's head collides with the interior of the passenger com- partment is taken as the time when the hypothetical passenger's head moves 0.6 m longitudinally (x direction) or 0.3 m laterally (y direction) inside the passenger compartment, calculated according to Eq. ( 4): X = 0.6 m, Y = 0.3 m, t * is the smaller value of t * x and t * y obtained by satisfying the integration equations in the x and y directions.
According to the acceleration data of the collision test vehicle's center of gravity, t* can be calculated from Eq. ( 4), and then the longitudinal and transverse components of the passenger collision velocity can be calculated from Eq. ( 3).The absolute value should be less than or equal to 12 m/s.During the process of vehicle collision with the guardrail, after the hypothetical collision between the passenger's head and the interior of the passenger compartment, the longitudinal and transverse components of the acceleration at the center of gravity of the vehicle can be calculated at intervals of 10 ms.The maximum absolute value should be less than or equal to 200 m/s 2 .
Redirecting refers to that the vehicle shall not roll over after an impact, and the wheel marks of the vehicle after leaving the exit point shall meet the requirements for redirective exit box.
Table 2 displays the test conditions for SS-level barriers of the numerical model, the real-vehicle and JTG B05-01-2013 11 .Real-vehicle and numerical-vehicle are shown in Fig. 9.
Table 3 displays the vehicle technical parameters for SS-level barriers of the numerical model, the real-vehicle and JTG B05-01-2013 11 .The vehicle technical parameters is shown in Fig. 10.

Establishment of the numerical and the full-scale barrier
The universal double-row beams and upright posts are made of isotropic Q355.Having been applied external force of 355 MPa which is yield strength, the material enters a plastic stage until it is stretched and broken.Parameters of the material are defined by using the lsdyna keyword *MAT_PIECEWISE_LINEAR_PLASTICITY, with a density of 7.85 kg/m 3 , Poisson's ratio of 0.3, elastic modulus of 210 GPa and yield strength of 355 MPa.Section characteristics are defined by using the lsdyna keyword *SECTION_SHELL.Beams, upright posts, and spacer blocks are set as Belytschko-Tsay shell units with a thickness of 6 mm.The concrete model is selected for both the barrier and deck slab, which are defined by using the keyword *MAT_CSCM_CONCRETE.The barrier concrete is 18 MPa in terms of strength grade, and the deck concrete is C40, which standard value of axial compressive strength is 26.8 MPa.The characteristics of the cross-section are set as entity units by using *SECTION_SOLID.The reinforcement is simulated as a line element by using *SECTION_BEAM, and the material is Q235 steel with a yield strength of 235 MPa.The contact between the embedded bar bolt, steel bar, and concrete is calculated by applying the Lagrange algorithm defined by the lsdyna keyword of *CONSTRAINED_LAGRANGE_IN_SOLID.The single-surface contact is defined by *CONTACT_AUTOMATIC_SINGLE_SURFACE for the upper structure and lower barrier.The surface-to-surface contact between the barrier and vehicle is defined by *CONTACT_ AUTOMATIC_SURFACE_TO_SURFACE.Constrain the 6 degrees of freedom of the end nodes of barriers and bridge decks in the length direction to simulate infinite length boundary conditions, which is defined by * BOUNDARY_SPC, that insert 1for translational constraint in x-y-z-direction and rotational constraint about x-y-z-axis.The established finite element model is shown in Fig. 11.
As shown in Fig. 12, the universal double-beam assembled bridge barriers are constructed by binding steel bars around the formwork, pouring the bridge deck, pouring the concrete barrier base, and planting the steel bars.The strength test value of the "base concrete" in the guardrail test section for collision testing is 16.5 MPa; the strength test value of the guardrail test section "bridge deck" used for collision testing is 39.7 MPa.The performance parameters of guardrail materials is shown in Table 4.For every type of vehicle, there are two rows of screenshots which are taken shot in 0.2 s respectively from left to right.Obviously, the car, bus, and truck all smoothly leave the impact zone after the impact, without crossing, climbing up, or riding over the barrier, as shown in Fig. 13.Thus, the containment performance complies with the requirements.And it is true that the travelling track of FE analysis result and full-scale vehicle impact record are nearly the same, which indicates the consistence of simulation and real test in this paper.
The centroid acceleration of the car is extracted.Define the X direction as the initial speed direction of the vehicle, and the Y direction as its vertical direction.The maximum acceleration is on the Y direction with a peak value of 70.9 m/s 2 , indicating that the barrier's buffering performance meets relevant requirements (as shown in Fig. 14).
The actual vehicle collision test tested the acceleration at the center of mass of the car.Using data processing software EXCEL, the acceleration curve is drawn by filtering at a frequency of 60 Hz and taking one test value every 9 values, so as to achieve an average value of 10 ms.The time-history curves of the longitudinal and transverse components of centroid acceleration of a car after the impact are shown in Fig. 15.When value of Y which represent the transverse displacement of driver is equal to 0.6 m, it means that the driver is likely to collide with car window.The larger of the acceleration of this moment is, the more danger the drive is.Displacement is calculated by integrating acceleration.When transverse displacement is 0.6 m, the maximum absolute values of   According to the experimental data, the absolute values of the longitudinal and transverse components of the vehicle after impacting the barrier are calculated as v x = 6.0 m/s and v y = 7.3 m/s, respectively, less than 12 m/s, which shows that the occupant impact velocities is meeting the relevant requirements.The results indicate that the barrier's buffering effect complies with the standards.The barrier's deformation after impacting a truck is analyzed, the maximum dynamic deformation is 581 mm, and is 690 mm in the full-scale test.The difference in guardrail deformation values between computer simulation analysis and actual vehicle collision tests is only 10 cm, and the calculation results are close, as shown in Fig. 16.
Analyze the state of guardrails after different vehicles collide with them.After the impact of a small passenger car and the simulation collision test, there is a small area of concrete damage to the guardrail base.And, in the    After the medium-sized bus collided with the guardrail, in the simulation test, the large deformation length of the crossbeam is 6 m, and the guardrail base is damaged within a range of 6 m.In the actual vehicle collision test, four flange columns are damaged, and the concrete base is damaged with a length of 6 m; After a large truck collided with a guardrail, in the simulation test, the large deformation length of the crossbeam is 6 m, and the damaged length of the concrete base is 6 m, and this is also true in actual vehicle collisions, as shown in Fig. 17.
After the impact of a medium-sized bus, the barrier concrete is most severely damaged.Therefore, the damage status of the bridge deck after a medium-sized bus collides with the barrier is analyzed.As shown in Fig. 18, after the medium-sized bus collided with the guardrail, the damage marks on the guardrail concrete did not extend to the bridge deck, the bridge deck was intact and undamaged, and the connection between the guardrail and the bridge deck was firmly connected in the simulation test and the actual vehicle collisions.
According to the simulation results and the real-vehicle impact test results, the universal double-beam assembled barrier meets all SS-level requirements in terms of containment, redirective and buffering, the universal double-beam assembled bridge barrier can effectively handle the impact of car, bus and truck, and the overall safety performance reaches Level VI (SS).

Figure 2 .
Figure 2. Concrete base with a height of 81 cm.

Figure 3 .
Figure 3. Concrete bases with heights of 63 cm and 68.5 cm.

Figure 4 .
Figure 4. Detection of concrete strength using rebound method.

Figure 5 .
Figure 5. Detecting the depth of concrete carbonation.
A H = 240 670 = 0.35 .The back distance of the upright posts complies with relevant standards.

Figure 7 .
Figure 7. Cross-sectional view of the universal double-beam assembled bridge barrier.

Figure 8 .
Figure 8. Impact point on the standard barrier section.

Figure 13 .
Figure 13.Comparisons of FE analysis and full-scale vehicle impact (from left to right).

Figure 14 .
Figure 14.Acceleration of a car.

Figure 15 .
Figure 15.Time history curves of centroid acceleration of a car after collision.

Table 1 .
Bearing capacity of deck slabs on an expressway bridge. S/

analysis and Full-scale vehicle impact test verification SS
-level vehicle impact tests are conducted in accordance with JTG B05-01-2013 to verify the performance of the improved universal double-beam assembled bridge barrier.

Table 4 .
Performance parameters of guardrail materials.
ParametersField test values Numerical-vehicle Real-vehicle full-scale collision test of the actual vehicle, there are little obvious signs of damage to the guardrail base and crossbeam.This indicates that computer simulation tests can effectively improve the likelihood of passing real vehicle collision tests.