Design of Roadside Barrier Systems Placed on MSE Retaining Walls (2010)

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.

50 The objectives of the bogie tests include quantification of the movement of the barrier, coping, and moment slab system and measurement of the force distributions in the reinforcement strips due to a design impact load. To help plan the bogie test, wall and finite element models were developed and impact simulations using the bogie impactor were performed using LS-DYNA. The results of bogie tests were used to develop the design guidelines for scenarios that include a traffic barrier mounted on the edge of an MSE wall. 5.1 5 ft High MSE Wall and Barrier Test Plan Half of the test wall was constructed using 2.43 m (8 ft) long reinforcement strips, while the other half was constructed with 4.88 m (16 ft) long reinforcement strips. The 2.43 m (8 ft) long reinforcement strip represents the minimum length allowed in current practice and, therefore, constitutes the critical case for assessing wall displacement during a barrier impact. Such lengths are commonly used in short-height wall segments such as at the beginning or ending of an elevated overpass structure. At the minimum 2.43 m (8 ft) length, current design proce- dures typically require a density of six reinforcement strips per wall panel (three in each of two different horizontal layers of reinforcement). The other half of the wall was constructed using 4.88 m (16 ft) long reinforcement strips. This length of reinforcement is a practical maximum length used in many MSE wall installations as wall height increases. The increased length increases the pullout resistance of the reinforcement. Therefore, a wall section with 4.88 m (16 ft) strips will con- stitute the critical case for assessing the magnitude and distri- bution of impact loads in the reinforcement. A summary of the bogie impact test plan is shown in Table 5.1. Two different barrier types were used in the test program: a N.J. safety shape (Test 1) and a vertical barrier (Tests 2, 3, and 4). Bogie Tests 1 (N.J. shape) and 4 (vertical barrier) were conducted over the portion of the wall with 4.88 m (16 ft) steel strip reinforcement. Bogie Tests 2 and 3 involved impacts into vertical concrete barriers placed over wall seg- ments with 2.43 m (8 ft) steel strip and bar mat reinforcement, respectively. The bogie test installation was planned on the premise that multiple impacts could be conducted on barrier segments con- nected to the same moment slab. This plan was accepted with the understanding that if excessive motion of the barrier– moment slab system occurred during the first test associated with a given moment slab, the ability to conduct subsequent impact tests of other barrier sections connected to the same moment slab could be compromised. Further, if the motion resulted in contact with a wall panel, the integrity of the wall might be compromised. Thus, the impact speed of the bogie had to be carefully selected to achieve the desired result of iden- tifying the failure mechanism of the barrier–moment slab sys- tem without imparting an unnecessarily high degree of damage to the underlying wall. 5.1.1 Calculation of MSE Wall Capacity AASHTO LRFD (2) was used to estimate the forces expected on the reinforcement strips due to both gravity and impact loads for the 1.52 m (5 ft) high MSE wall. This information ultimately was compared to forces estimated through simula- tion and measured in the bogie tests as new design procedures were developed. In the 2.43 m (8 ft) long strip case, unfactored resistances were calculated to be 6.595 kN (1.483 kips) (F* = 1.837) at the uppermost layer and 12.019 kN (2.702 kips) (F* = 1.674) at the second layer. A density of three strips per layer per panel was used. The unfactored load per strip due to gravity was calcu- lated to be 2.53 kN (0.569 kips) at the uppermost layer and 4.87 kN (1.095 kips) at the second layer. In this analysis, the traffic surcharge was not considered. The unfactored load per strip due to the impact was calculated to be 1.903 kN (0.428 kips) at the uppermost layer and 1.293 kN (0.291 kips) C H A P T E R 5 5 ft High MSE Wall and Barrier Study

at the second layer. Therefore, the unfactored total load per strip was 4.433 kN (1.0 kips) at the uppermost layer and 6.163 kN (1.43 kips) at the second layer. A summary of resistance and load per strip is presented in Table 5.2. In the 4.88 m (16 ft) long strip case, unfactored resistances were calculated to be 13.19 kN (2.965 kips) (F* = 1.837) at the uppermost layer and 24.038 kN (5.404 kips) (F* = 1.674) at the second layer. A density of 2 strips per layer per panel was used. The unfactored load per strip due to gravity was calculated to be 3.795 kN (0.853 kips) at the uppermost layer and 7.306 kN (1.642 kips) at the second layer. In this analysis, the traffic sur- charge was not considered. The unfactored load per strip due to the impact was calculated to be 2.854 kN (0.642 kips) at the uppermost layer and 1.939 kN (0.436 kips) at the second layer. Therefore, the unfactored total load per strip was 6.65 kN (1.49 kips) at the uppermost layer and 9.245 kN (2.09 kips) at the second layer. A summary of resistance and load per strip is presented in Table 5.3. The detailed calculations for designing the MSE test wall are provided in Appendix A. 5.1.2 Calculation of Barrier Capacity It is important to be able to quantify the ultimate strength of the barrier sections used in the bogie test matrix to aid in analy- ses of the barrier–moment slab systems. For example, knowing the ultimate strength of the vertical wall section permitted the overturning test to be planned at an impact speed that will cause substantial movement (i.e., rotation) of the barrier– moment slab system without causing failure of the barrier. Also, the bogie impact speed for the planned bogie tests of the different barrier sections atop a 1.52 m (5 ft) tall MSE wall was selected such that the generated impact force would exceed the capacity of the barrier. In this way, the failure mode of the pre- cast barrier unit can be identified and the maximum impact load will be applied to the supporting MSE wall and its rein- forcement. For these reasons, the strength of the selected N.J. safety shape and vertical wall barrier sections was computed. In regard to the analysis of the strength of a barrier, most references containing concrete barrier design information have used the yield line analysis approach (AASHTO LRFD A13.3.1). Yield line theory considers the plastic strength of all the railing system components along with barrier geometry, material strengths, applied loading, and strength of the sup- porting bridge structure. Steel rail systems, concrete rail systems, or a combination rail composed of a steel rail on a concrete barrier can be evaluated using these design procedures. Based on yield line theory, the limiting ultimate capacity of the railing systems used in the test program was calculated. This procedure assumes that the underlying support structure for the barrier section (e.g., deck, foundation, etc.) has sufficient strength to develop the base moment capacity of the barrier such that the failure occurs in the barrier rather than the sup- port structure. Although this may not necessarily be the case for the precast barrier–coping sections, the yield line analysis approach was used to establish what would be considered a maximum barrier capacity. This ultimate capacity was then compared to design forces derived from simulations to deter- mine appropriate impact speeds for the bogie vehicle tests. The ultimate load capacity calculated following the assumed yield line failure mechanism was 515.06 kN (115.79 kips) for the 813 mm (32 in.) tall N.J. barrier. This load capacity is much 51 Test Sequence Barrier Type Moment Slab Width Barrier Length Reinforcement Type Reinforcement Length Test 1 New Jersey 4.5 ft 10 ft Steel strips 16 ft Test 2 Vertical wall 4.5 ft 10 ft Bar mats 8 ft Test 3 Vertical wall 4.5 ft 10 ft Steel strips 8 ft Test 4 Vertical wall 4.5 ft 10 ft Steel strips 16 ft Layer (1) TStatic Static Load (kips) (2) TDynamic Dynamic Load (kips) (3)=(1)+(2) TTotal Total Load (kips) R Resistance (kips) Top 0.569 0.428 0.997 1.483(F* = 1.837) Second 1.095 0.291 1.386 2.702(F* = 1.674) Layer (1) TStatic Static Load (kips) (2) TDynamic Dynamic Load (kips) (3)=(1)+(2) TTotal Total Load (kips) R Resistance (kips) Top 0.853 0.642 1.495 2.965(F* = 1.837) Second 1.642 0.436 2.078 5.404(F* = 1.674) Table 5.1. Bogie test plan. Table 5.2. Unfactored resistance and force in case of MSE wall with 8 ft long strip. Table 5.3. Unfactored resistance and force in case of MSE wall with 16 ft long strip.

higher that observed in previous dynamic bogie testing. The moment capacity of the “toe” of the safety shape is large and, thus, typically restricts failure to the upper wall portion of the barrier. The ultimate load capacity of the upper wall portion of the N.J. barrier was calculated to be 329.26 kN (74.02 kips). The length of the failure mechanism calculated for the N.J. bar- rier section analyzed was 2.23 m (7.3 ft). The ultimate load capacity calculated following the assumed yield line failure mechanism was 325.34 kN (73.14 kips) for the 0.69 m (27 in.) tall vertical wall barrier. Note that the height of load application assumed for calculation of the ultimate bar- rier load capacities was the top of the barrier. The length of the failure mechanism calculated for the vertical wall barrier sec- tion analyzed was 1.65 m (5.4 ft). This indicates that, provided the coping has sufficient capacity to develop the ultimate strength of the barrier, the 3.05 m (10 ft) section length selected for evaluation in the bogie tests should be sufficient for devel- oping the primary failure mechanism for each barrier type. 5.2 Finite Element Analysis The complex nonlinear interactions that occur during an impact event are difficult to capture through conventional ana- lytical means. Therefore, an explicit nonlinear finite element methodology was used to evaluate the dynamic impact per- formance of the representative barrier–moment slab–MSE wall configurations considered in the test plan. 5.2.1 Modeling Methodology The methodology followed to model the barrier on top of the MSE wall, and then simulate bogie impacts, consisted of the following steps: 1. Construct a finite element model of the barrier and MSE wall. 2. Initialize the model of MSE wall and barrier to account for gravitational loading. 3. Simulate the bogie impact against the barrier. 4. Compare results with test data and calibrate the MSE wall and barrier finite element model if needed. 5. Identify any further investigation needed. The details of the finite element analysis are presented in the following sections. Geometry and Meshing The finite element representation of the MSE wall planned for use in the bogie test program consists of the following major components: • Precast concrete barrier–coping sections • Cast-in-place moment slab • Steel reinforcement in the barrier, moment slab, and wall panels • MSE wall including the backfill soil, concrete wall panels, level-up concrete, pedestal, and wall reinforcement • Accelerometers on the barrier and moment slab. The total length of the MSE wall model was 9.14 m (30 ft), which represented the length of one moment slab section. Three 3.05 m (10 ft) long barrier–coping sections were attached to the 9.14 m (30 ft) long moment slab. Both full and half- panels were used to construct the wall (see Figure 5.1). The concrete barrier and moment slab were modeled using solid elements, as were various components of the MSE wall including the soil, wall panels, leveling pad, and pedestal. Three-dimensional beam elements with six degrees at each end were used to model the steel rebar inside the barriers, moment slab, and wall panels. The steel strip reinforcement for the MSE wall was modeled using shell elements with 4 mm (0.16 in.) thickness and 50.8 mm (2 in.) width. The elements of impact barrier located in the middle of the system were meshed using an element characteristic size of about 50 mm (2 in.) to capture the barrier deformation and damage due to the impact with more accuracy. The two other barriers were meshed more coarsely to reduce compu- tational cost of the simulations since these barriers do not interact with the bogie. The soil was modeled as three compo- nents: the reinforced backfill, the overburden soil, and the side soil for modeling continuity at the edges of the moment slab as shown in Figures 5.1 and 5.2. The soil elements located beneath the barrier and moment slab were meshed rela- tively fine using an element characteristic size ranging from 50 mm (2 in.) to 101.6 mm (4 in.) to improve the robust- ness of the contact between the coping and top edge of the soil and better capture the load transfer from the barrier to the soil during the impact. The overburden soil and side bound- ary soil are rather coarsely meshed, while a finer mesh is used for the reinforced backfill to capture gravity and impact loads distributed into the soil through the MSE wall and the barrier with more accuracy. 52 Figure 5.1. Three-dimensional view of a MSE wall and barriers model with a bogie.

Contact Although LS-DYNA features some of the most advanced contact algorithms available, capturing interaction between solid and beam or shell elements is rather complex. The requirement of matching nodes to merge the reinforcing steel inside the concrete continuum would dictate the creation of elements with poor aspect ratios and the creation of unneces- sarily small element sizes, which has a significant effect on time step control (24). To mitigate this problem, a different connec- tion scheme was utilized between the barrier and the steel rein- forcement that permits a more regular, uniform mesh of the concrete to be used throughout the barrier. The steel reinforcements are coupled (rather than merged) to the surrounding concrete continuum to prevent the creation of poor quality elements. This coupling was achieved using the *CONSTRAINED_LAGRANGE_ IN_SOLID feature in LS-DYNA. The use of this coupling permits the concrete mesh to be constructed without consideration of the location of steel reinforcement. The steel reinforcement is treated as a slave material that is coupled with a master material composed of the moment slab and barrier concrete. The slave parts (i.e., steel reinforcing bars) can be placed anywhere inside the mas- ter continuum part without any special mesh accommodation. The soil and its reinforcement were methodologically modeled in a manner similar to the steel reinforcement in concrete to capture salient responses of the MSE wall. Reinforcing steel in the barriers sections and moment slab and steel reinforcement strips in the soil are shown in Figures 5.3 and 5.4. Another coupling mechanism, *CONTACT_TIED_ SHELL_EDGE_TO_SURFACE, was defined to account for the connection between the panel and steel reinforcement. The interface between the soil and concrete was modeled using contacts and/or constraints to capture interface (i.e., con- tact) forces generated between the concrete structure and the MSE wall. The contact friction was based on the estimated soil internal friction angle. Using a soil friction angle φ of 35 degrees, the contact friction was calculated to be 0.7 (tan φ). Material Models and Model Parameters Concrete and Steel Material. There are several material options to be considered for modeling the concrete structures in LS-DYNA. These material options range from the very sim- ple elastic material to a nonlinear damage material model. The elastic material option can be useful in modeling areas that will not be subjected to significant stress in order to reduce compu- tational costs of the simulation. If this approach is used, appro- priate checks must be made to ensure the tensile stress in the concrete does not exceed its failure threshold (24). The outside barrier sections were modeled using elastic material (designated as MAT type 1 in LS-DYNA) as shown in Table 5.4. However, the middle barrier that was subjected to direct impact was modeled using a nonlinear response con- crete material model definition. In LS-DYNA, it is designated 53 Figure 5.2. Side view of a MSE wall and barriers with a bogie. Figure 5.3. Rebars detail in N.J. barrier and moment slab. Figure 5.4. Interface between soil and strip shell element.

as material MAT type 159 developed by APTEK (24). This method to explicitly model concrete is more sophisticated but computationally expensive. In this model, a brittle material like concrete will lose (at a given rate) its ability to carry load when a specified damage/failure threshold is reached. This feature is very useful because it provides a more accurate representation of the failure mechanism of the concrete components, and better prediction of the impact load trans- fer. The parameters of MAT type 159 can be assigned using two additional concrete properties: the unconfined compres- sive strength of concrete (f ′c) and the maximum aggregate size, which was taken as 25.4 mm (1 in.). The moment slab, the wall panels, the pedestal, and the level-up concrete were modeled using an elastic material model definition (MAT type 1). All steel rebars and steel strips were modeled using a piecewise linear plasticity material model (MAT type 24) that is representative of the actual stress-strain relationship of the material using the properties shown in Table 5.5. Steel rebar exhibits rate effects, and yields in a duc- tile manner until it breaks at an ultimate strain greater than approximately 20%. Before yielding, the material is assumed to be linearly elastic. After yielding, the steel can undergo plastic deformation and strain hardening. Soil Material. The soil was modeled using the two- invariant geological cap material model (MAT type 25) (20). The advantage of the cap model over other models such as the Drucker–Prager formulation is the ability to model plastic com- paction. In these models, all purely volumetric response is elas- tic until the stress point hits the cap surface. Therefore, plastic volumetric strain (compaction) is generated at a rate controlled by the hardening law. Thus, in addition to controlling the amount of dilatancy, the introduction of the cap surface adds another experimentally observed response characteristic of geological materials into the model (25, 26). The cap model is defined in terms of the first stress invariant I1 = trace(σ) = σ11 + σ22 + σ33 and the second deviatoric stress invariant J2 = 1⁄2SijSij = 1⁄2(s211 + s 2 22 + s 2 33), where σ is the stress ten- sor and Sij = σ1 − σ3 is the deviatoric stress tensor. The yield sur- face of the cap model consists of three regions (Figure 5.5): a failure envelope f1(σ), an elliptical cap f2(σ, κ), and a tension cutoff region f3(σ), where κ is the hardening parameter. The functional forms of the three surfaces are the following: In the elastic region, Fe(I1) can be expressed as: where the yield surface was determined by the parameters α, θ, γ, and β, which are usually evaluated by fitting a curve through failure data taken from a set of triaxial compression tests. F I e Ie I1 1 1 5 4( ) = − +−α γ θβ ( )- 3 03 1 1. : ,Tension cutoff region forf T I Iσ( ) = − = = T ( )5 3- 2 02 2 1. : , ,Cap surface region f J F ID cσ κ κ( ) = − ( ) = , ( )for -L I Xκ κ( ) ≤ < ( )1 5 2 1 01 2 1. :Failure surface region f J F ID eσ( ) = − ( ) = , ( )for -T I L≤ < ( )1 5 1κ 54 Concrete E (psi) (lb/in3) f’c (psi) Elastic (MAT type 1) 3.62E+6 0.17 0.084 – Damage (MAT type 159) – – 0.084 4,000 E is Young’s modulus, is Poisson’s ratio, is the mass density, and f’c is the compressive strength. Table 5.4. Material properties of concrete model. E (psi) (lb/in3) Yield Stress (psi) Steel (MAT type 24) 30.46E+6 0.29 0.28 60,000 E is Young’s modulus, is Poisson’s ratio, and is the mass density. Table 5.5. Material properties of steel model. Source: Hallquist (20) Figure 5.5. Yield surface of the cap model.

In Equation 5-2, Fc(I1, κ) can be expressed as: where X(κ) is the intersection of the cap surface with the I1 axis and the hardening parameter κ is related to the plastic volume change v p through the hardening law v p W D X X= − − ( )−( )[ ]{ }1 5 80exp ( )κ - L κ κ κ κ ( ) = > ≤ ⎧⎨⎪⎩⎪ if if - 0 0 0 5 7( ) X RFeκ κ κ( ) = + ( ) ( )5 6- F I R X L I Lc 1 2 1 21 5 5, ( )κ κ κ κ( ) = ( )− ( )[ ] − − ( )[ ] - where the values of parameters W and D are found from hydro-static compression test data. The value of R is the ratio of major to minor axes of the quarter ellipse defining the cap surface. The parameters used in the numerical simulation are shown in Table 5.6. To understand the failure behavior of the cap soil material, the various soil properties were collected as given in Table 5.6. Two different soil properties, McCormick Ranch Sand (27) and elasto-plastic soil parameters given in NCHRP Report 556 (28), were compared to verify the cap property used in this study. The cap models for each case were plotted as shown in Figure 5.6. In the failure surface [f1(σ)] and tension cutoff region [f3(σ)], the three soils show good agreement, but in the elliptical cap [f2(σ, κ)], the soil material used in simulation shows a larger cap surface area than the other soils due to the large R. 55 Simulation McCormickRanch Sand NCHRP 556 K (MPa) 22.219 459.676 52.19 Elasticity G (MPa) 7.407 275.792 24.087 (MPa) 4.154 0.00186 0.01 (MPa-1) 0.0647 0.09718 0 (MPa) 4.055 0.00117 0 Plasticity (radian) 0 0.02 0.2925 W 0.08266 0.064 0.023 D (MPa-1) 0.239 0.00725 0.87 R 28.0 2.5 4.0 Hardening Law X0 (MPa) –2.819 1.20658 0.01593 Tension Cut T (MPa) 0 –2.06843 0 Table 5.6. Comparison of cap soil properties. 0 0.2 0.4 0.6 0.8 1 -5 0 5 10 15 20 I1 (MPa) SQ RT (J 2 D ) ( M Pa ) Cap Model used in Simulation McCormick Ranch Sand NCHRP 556 Figure 5.6. Comparison of cap models.

Bogie Vehicle The Texas Transportation Institute (TTI) test bogie is a 2,268 kg (5,000 lb) vehicle configured with three 304.8 mm (12 in.) diameter crushable steel cylinders on its nose assem- bly as shown in Figure 5.7. A spreader beam is attached across the three cylinders as shown in Figure 5.8. A wood block is attached to the face of the spreader beam to help dampen high-frequency noise during an impact. The bogie has an accelerometer installed at its CG. The finite element model of the bogie consisted of a simple representation of the vehicle chassis with a more detailed rep- resentation of the crushable nose assembly. Similar to the test bogie, an accelerometer was placed at the CG of bogie model. The finite element model of the bogie consists of 3,935 ele- ments and 4,645 nodes. Initialization of the Model for Gravitational Loading The MSE wall and barrier model was initialized to account for gravitational loading on the soil mass. Gravity loading effects soil pressure on the wall panels and the buildup of ini- tial stresses in the steel strips. This step had to be done prior to vehicular impact on the barrier. It was achieved by gradually ramping up gravity on the system while imposing a diminish- ing mass damping on the soil mass to prevent oscillatory forces from developing. The gravity loading and damping on the soil are shown in Figure 5.9. The weight of the system was measured and used as a con- vergence criterion for the steady state solution for the MSE wall model with 4.88 m (16 ft) strips as shown in Figure 5.10. For example, the total mass of the finite element model with the N.J. barrier on top of the MSE wall with 4.88 m (16 ft) long strips is 277,549 kg (19,018 slug or 611,890 lb mass). The weight of the system is calculated to be 2,721.6 kN (611.85 kips) using the mass of the finite element model and the acceleration of gravity. Therefore, after accounting for gravitational load, the weight of the model system should converge to the calcu- lated system weight. The weight of the finite element model was 2,717.7 kN (610.96 kips) at the end of initialization step. A reasonable agreement shows that the weight of the finite ele- 56 (a) 5,000 lb TTI test bogie. (b) 5,000 lb bogie model. Accelerometer Accelerometer Figure 5.7. 5,000 lb bogie model. Figure 5.8. Detailed crushable cylinders of test bogie and numerical simulation.

57 0 0.2 0.4 0.6 0.8 1 1.2 0 50 100 150 200 250 300 350 400 450 0 0.1 0.150.05 0.2 0.25 0.3 0.35 G ra vi ty (g ) D am pi ng (s ec - 1 ) Time (sec) Damping Gravity Figure 5.9. Gravity and damping of the MSE wall in steady state condition. 0 0.1 0.150.05 0.2 0.25 0.3 0.35 Time (sec) 0 500 1000 1500 2000 2500 3000 W ei gh t ( kN ) Simulation weight Calculated weight Figure 5.10. System weight of the MSE wall model. ment model approached the calculated weight of the model system as shown in Figure 5.10. 5.2.2 Finite Element Model: Boundary Conditions Initially, the overburden soil on the moment slab was defined to be continuous across the front traffic edge of the moment slab. Along the sides of the moment slab, the soil was discontinuous and only constrained in the longitudinal direc- tion (y-direction in Figure 5.11) to retain it in place and prop- erly account for mass and inertial effects. Thus, the first attempted model did not account for any fric- tion or shear strength that might exist along the sides of the moment slab [Figure 5.12(a)]. The displacement of the barrier–moment slab system using this condition was greater than expected based on field experience with these systems. Additionally, the forces estimated in the strips were well above those computed based on current AASHTO LRFD design practice. The lateral displacement at the bottom edge of the coping was predicted to be of sufficient magnitude to contact and apply substantial force to the recessed wall panel (see the circle in Figure 5.11). This contact will undoubtedly increase the forces in the wall reinforcement and has the potential to

fracture the wall panel and/or result in sufficient movement of the panel to cause pullout of the reinforcing strips. It was the- orized that some of this “excessive movement” in the barrier– moment slab system might be attributed to the model’s neglect of friction along the sides of the overburden soil and moment slab. The sensitivity of the dynamic behavior of the system to the boundary conditions along its sides was investi- gated through additional simulations. In practice, the boundary conditions for a barrier–moment slab system can vary from installation to installation. In some systems, adjacent moment slabs are doweled together—a practice that greatly enhances the shear resistance at the edge of the moment slab. If dowel bars are not used, the moment slab sections act independently of one another but there will still be some shear capacity at the interface due to frictional contact. In both cases, the shear strength of the overburden 58 Moment Slab Figure 5.11. System reaction force of the MSE wall model. (b) Friction wall condition (c) Side soil condition (a) No friction boundary condition Figure 5.12. Comparison of simulation with different boundary condition.

soil and overlying pavement surface that is continuous across the moment slab interface will increase the overall resistance of the system to movement. A second model was constructed with frictional contact on each side of the barrier–moment slab system and soil [Figure 5.12(b)]. The static and dynamic coefficients of friction were defined based on angles of internal friction of 35 degrees and 31 degrees, respectively. As indicated by the results of these simulations shown in rows 1 and 2 of Table 5.7, friction at the ends of the moment slab and overburden soil can have a pronounced effect on the displacement and rotation of the barrier–moment slab system under impact load. The two simulations are thought to bracket the barrier movement likely to be observed in the bogie tests. Given the range of movement observed at the bottom of the coping in the previous simulations, an additional simulation effort was undertaken in an attempt to more precisely define the soil and moment slab boundary conditions as shown in Figure 5.12(c). The overburden soil was extended across the edges of the moment slab, and a soil continuum was modeled adjacent to the ends of the moment slab (replacing the previ- ously modeled side wall). The shear strength of the overburden soil was captured through the soil material model rather than through a defined frictional contact. The interface between the moment slab and adjacent soil were defined through definition of material contacts with friction coefficients assigned based on the angle of internal friction of the soil. Additionally, the row of elements directly beneath the moment slab and adjacent to the bottom edge of the coping were removed from the soil mesh [see the circle in Figure 5.12(c)]. Based on analyses of the previous simulations, it was theorized that these elements may be providing artificially high resistance to rotation of the barrier–moment slab system. A comparison of displacements for the barrier–coping section obtained from the different simulations is shown in Table 5.7. The greatest movement was obtained for the model without side friction. The lowest displacements at the top and bottom of the barrier–coping section were obtained from the model in which the ends of the overburden soil and moment slab were in frictional contact with side walls. The model with continuous overburden soil across the ends of the moment slab fell between the other two. The fact that the simulation with continuous overburden soil is closer to the results of the model without side friction than the model with frictional side walls is at least partially due to the removal of a row of elements below the moment slab adjacent to the bottom edge of the coping. Table 5.7 also shows a comparison of the displacement of the reinforcement strip directly under the point of impact, which is the one that experiences the highest impact load. The displacements at the end of the strip range from 1.7 mm (0.065 in.) to 3.7 mm (0.13 in.). These displacements are due to transfer of the barrier impact load into the MSE wall (through the backfill soil) and do not include any movement that may arise because of direct contact between the coping and wall panels. While it is arguable which of the simulations most closely resembles reality, they collectively raised concern regarding the possibility of strip pullout. Movement of the strip along its entire length would limit the magnitude of the impact force. In other words, the maximum force due to impact that can be measured in the strips is limited by their pullout resis- tance. If pullout occurs, the maximum forces imparted to the reinforcement strips due to a barrier impact will be reduced. Without further test data to validate the soil model and boundary conditions, it was unknown which of the modeling methodologies most closely resembles the actual system. Data derived from the bogie tests were later used to calibrate and validate the finite element model so that additional simula- tions can be conducted with more confidence to support the development of new design guidelines and predict the per- formance of the barrier–moment slab system under a full- scale vehicular impact. 5.2.3 Simulated Impact into Barrier Placed on MSE Wall with 8 ft Long Strip The simulated bogie vehicle hit the middle vertical bar- rier section at a speed of 32.67 km/h (20.3 mph). The point of impact was slightly offset from the centerline of the middle barrier section to align with one of the reinforcement strips (strip D1). To enable comparison of forces and displacements, selected strip locations were assigned an alphanumeric desig- nator that describes its horizontal position relative to the bogie impact point and its vertical reinforcement layer. The location designator used is based on a density of three strips per layer per panel. For example, strip D1 is positioned beneath the impact point in the first (i.e., upper) layer of reinforcement as shown in Figure 5.13. Sequential images from the simulation are shown in Fig- ure 5.14. The maximum displacement at the top of the mid- dle barrier occurred at 0.07 sec. The maximum 50 msec average impact load was 346.38 kN (77.87 kips) at 0.04 sec as 59 Middle barrier at top Middle barrier at bottom Strip at front at impact location Strip at end at impact location No Side Friction 7.35 in. 4.25 in. 0.16 in. (4.0 mm) 0.13 in. (3.7 mm) Side Wall 4.99 in. 0.03 in. 0.07 in.(2.0 mm) 0.065 in. (1.7 mm) Continuous Soil 6.69 in. 3.30 in. 0.15 in. (4.0 mm) 0.128 in. (3.3 mm) Table 5.7. Comparison of displacements for bogie test models.

60 A1C1 A2C2 B1 B2 D1F1 D2F2 E1 E2 Figure 5.13. Strip location indicator. Figure 5.14. Sequence image of model during impact. (a) 0 sec (b) 0.07 sec (c) 0.125 sec (d) 0.32 sec (final)

shown in Figure 5.15. The maximum dynamic displacement was 99 mm (3.9 in.) at the top of the barrier and 44.5 mm (1.75 in.) at the bottom of the coping. The permanent dis- placement was 54 mm (2.1 in.) at the top of the barrier and 33 mm (1.29 in.) at the bottom of the coping. The maximum dynamic displacement of the panel was 4.3 mm (0.17 in.) at the upper reinforcement layer and 1 mm (0.04 in.) at the sec- ond layer. The permanent displacement of the panel was 2.3 mm (0.09 in.) at the upper layer and 0.4 mm (0.02 in.) at the second layer (see Figure 5.16). As shown in Figure 5.17, the damage profile that develops in the simulated barrier is similar to that observed in previous tests of N.J. profile barriers. It occurs above the toe of the bar- rier and has a parabolic shape. However, because of the short [3.05 m (10 ft)] length of the precast barrier section that was modeled, much of the damage eventually radiates out to the free ends of the section. The strip load in the numerical simulation consists of the initialized static load and the dynamic impact load. Figure 5.18 shows the raw data of the total load in selected reinforcement strips. The maximum instantaneous load for strip D1 was 21.31 kN (4.79 kips). The total maximum 50 msec tensile load for strip D1 was 15.44 kN (3.47 kips) at a distance of 178 mm (7 in.) from the face of the panel, which corresponds to the planned location of strain gages in the test installation (Figure 5.19). The distributions of maximum load along the reinforce- ment strips at the uppermost reinforcement layer (D1) are pre- sented in Figure 5.20. As expected, the maximum load in a given strip occurs near its attachment to the wall panel and the 61 0 10 20 30 40 50 60 70 80 90 0 0.05 0.1 0.15 0.2 0.25 Im pa ct F o rc e (ki ps ) Time (sec) 0 1 2 3 4 0 0.05 0.1 0.15 0.2 0.25 0.3 H or iz . D ip l. of B ar ri er a nd P an el (in .) Time (sec) Top of Barrier Bottom of Barrier Top Layer of Strip Bottom Layer of Strip Figure 5.15. Impact load. Figure 5.16. Displacement of barrier and panel.

62 (a) (b) Figure 5.17. Concrete damage profile on (a) front side and (b) back side of the barrier.

63 -3 -2 -1 0 1 2 3 4 5 6 7 0 0.05 0.1 0.15 0.2 0.25 0.3 Fo rc e (ki ps ) Time (sec) D-1 D-2 F-1 F-2 -2 -1 0 1 2 3 4 5 Fo rc e (ki ps ) Time (sec) D-1 D-2 F-1 F-2 0 0.05 0.1 0.15 0.2 0.25 0.3 0 2 4 6 8 10 12 0 1 2 3 4 5 6 7 8 Fo rc e (ki ps ) Strip Length from the Surface of Panel (ft) Figure 5.18. Raw data of load on the strip. Figure 5.19. 50 msec average load on the strip. Figure 5.20. Distribution of load on the strip (D1) at 0.087 sec.

load decreases along its length. The loads in the strips attached to the wall panel below the point of impact in the middle of the barrier section have similar distributions. Table 5.8 shows the absolute maximum load in the strips at the location 178 mm (7 in.) from the wall end of the reinforcement. The strain on the wall panel surface was evaluated to check the resistance of the panel-to-barrier impact loads. Figure 5.21 shows the strains predicted in the numerical simulation at strip D1. The maximum compressive strain was 0.0047 at 0.095 sec. The maximum bending moment (Mi) per unit length of wall at the top level of reinforcement can be calculated by a simple model as shown in Figure 5.22. The moment (Mi) is generated from the horizontal shear load (H) on top of the panel times the moment arm (l) and the vertical load (V) that is transferred from the barrier to the panel during the impact times the moment arm (t/2) where t is the thickness of the panel. Because the simulation corresponded to a peak dynamic force of 346.38 kN (77.87 kips) when the design calls for 240 kN (54 kips), the results of the numerical simulation were decreased by the ratio of 240 kN / 346.38 kN (54 kips / 77.87 kip) The distribution of shear load, vertical load, and bending moment along the panel at the time of peak force during the impact is shown in Figures 5.23, 5.24, and 5.25, respectively. In this case, the forces due to the rebars in the panel were neglected. The predicted shear load was 2.62 kN (0.59 kips) at the top of the panel during the impact due to the friction between the leveling pad and the panel. The vertical load transferred by the barrier was 9.61 kN (2.16 kips). 5.2.4 Simulated Impact into Barrier Placed on MSE Wall with 16 ft Long Strip The simulated bogie vehicle impacted the middle N.J. bar- rier section at a speed of 32.19 km/h (20 mph). The point of 64 Location on Strip Load D1 D2 F1 F2 Total max. load (kips) 4.79 1.36 6.17 1.45 Total 50 msec average load (kips) 3.47 0.6 4.2 0.68 Table 5.8. Load in strips at 7 in. location from the face of the wall panel. -0.005 -0.004 -0.003 -0.002 -0.001 0 0.001 0 0.05 0.1 0.15 0.2 0.25 0.3 St ra in (in ./i n .) Time (sec) Figure 5.21. Panel strain at strip D1 in MSE wall with 8 ft long strips. Rotation Point, B Mi Barrier Moment Slab l FS1 H V t C Figure 5.22. Free body diagram on the panel (rotation point B).

impact was slightly offset from the centerline of the middle bar- rier section to align with one of the reinforcement strips (D1). To enable comparison of forces and displacements, the same alphanumeric designators shown in Figure 5.13 were used Sequential images from the simulation are shown in Fig- ure 5.26. The maximum displacement at the top of the mid- dle barrier occurred at 0.07 sec. The maximum 50 msec average impact load was 359.06 kN (80.72 kips) at 0.038 sec (Figure 5.27). The maximum dynamic displacement was 108.7 mm (4.28 in.) at the top of the barrier and 33.3 mm (1.31 in.) at the bottom of the coping. The permanent dis- placement was 41.1 mm (1.62 in.) at the top of the barrier and 19.6 mm (0.77 in.) at the bottom of the coping. The maximum dynamic displacement of the panel was 5.6 mm (0.22 in.) at the upper reinforcement layer and 2.5 mm (0.1 in.) at the second layer. The permanent displacement of the panel was 1.5 mm (0.06 in.) at the upper layer and 0.8 mm (0.03 in.) at the second layer (see Figure 5.28). The damage profile that developed in the barrier is shown in Figure 5.29. Figure 5.30 shows the raw data of the total load in selected reinforcement strips. The maximum instantaneous load for strip D1 was 39.86 kN (8.96 kips) at a distance of 178 mm 65 0.59 0.49 0.29 -0.96 -0.60 -0.04 0.07 0.10 0.12 0 1 2 3 4 5 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 Pa ne l H ei gh t ( ft) Shear Force(kips/ft) Figure 5.23. Shear load on the panel. -2.16 -2.61 -3.69 -3.20 -1.54 -2.12 -2.09 -2.16 -2.17 0 1 2 3 4 5 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 Pa ne l H ei gh t ( ft) Vertical Force(kips/ft) Figure 5.24. Vertical load on the panel.

66 0.00 -0 .0 7 -0 .2 3 -0 .2 8 -0 .4 6 -0 .4 8 -0 .4 2 -0 .1 9 -0 .1 2 0 1 2 3 4 5 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 Pa n el H eig ht (ft ) Bending Moment (kips-ft/ft) Figure 5.25. Bending moment on the panel. Figure 5.26. Sequence image of model during impact.

(7 in.) from the face of the panel which corresponds to the planned location of strain gages in the test installation. The total maximum 50 msec tensile load in the wall reinforcement was 28.96 kN (6.51 kips) (Figure 5.31). The distribution of maximum load along strip D1 is pre- sented in Figure 5.32. As expected, the maximum load occurs near its attachment to the wall panel and the load decreases along its length. The loads in the strips attached to the wall panel below the point of impact in the middle of the barrier section have similar distributions. Table 5.9 shows the sum- mary of the load for the strips at a distance of 178 mm (7 in.) from the wall end of the reinforcement. As in the previous simulation, the strain in the wall panel was evaluated to determine its ability to resist barrier impact loads. The results are shown in Figure 5.33. The maximum compressive strain was 0.0021 at 0.045 sec. The maximum bending moment (Mi) per unit length of wall at the location of the top level of reinforcement can be calcu- lated by the simple model shown previously in Figure 5.22. The moment (Mi) is generated from the horizontal shear load (H) on top of the panel times the moment arm (l) and the vertical load (V) which is transferred from the barrier to the panel during the impact times the moment arm (t/2) where t is the thickness of the panel. Because the simulation corresponded to a peak dynamic force of 359.06 kN (80.72 kips) when the design calls for 240 kN (54 kips), the results of the numerical simulation were decreased by the ratio of 240 kN / 359.06 kN (54 kips / 80.72 kip). 67 0 10 20 30 40 50 60 70 80 90 0 0.05 0.1 0.15 0.2 0.25 Im pa ct F or ce (k ips ) Time (sec) Figure 5.27. Impact load. -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Time (sec) D ip la ce m en t ( in. ) Barrier Top Barrier Bottom Top Layer of Strip Bottom Layer of Strip Figure 5.28. Displacement of barrier and panel.

68 (a) (b) Figure 5.29. Concrete damage profile on (a) front side and (b) back side of the barrier.

-4 -2 0 2 4 6 8 0 0.05 0.1 0.15 0.2 0.25 0.3 Time (sec) Fo rc e (k ips ) D1 D2 F1 F2 Figure 5.30. Raw data of load on the strip. -4 -2 0 2 4 6 8 0 0.05 0.1 0.15 0.2 0.25 0.3 Time (sec) Fo rc e (ki ps ) D1 D2 F1 F2 Figure 5.31. 50 msec average load on the strip. 0 1 2 3 4 5 6 7 8 9 0 2 4 6 8 10 12 14 16 Fo rc e (ki ps ) Strip Length from the Surface of Panel (ft) Figure 5.32. Distribution of load on the strip (D1) along the strip length.

The distribution of the shear load, vertical load, and bending moment along the panel at the time of peak force during the impact is shown in Figures 5.34, 5.35, and 5.36, respectively. The predicted shear load was 6.98 kN (1.57 kips) at the top of the panel during the impact due to the friction between the lev- eling pad and the panel. The vertical load transferred by the bar- rier was 76.2 kN (17.13 kips). Using the shear and vertical loads, the bending moment was calculated to be 18.15 kN-m/m (4.15 kips-ft/ft), which is higher than the calculated strength of 70 Location on Strip Load D1 D2 F1 F2 Total max. load (kips) 8.96 4.59 8.35 5.22 Total 50 msec average load (kips) 6.51 3.31 5.95 3.83 Table 5.9. Load in strips at 7 in. location from the face of the wall panel. -0.0025 -0.002 -0.0015 -0.001 -0.0005 0 0.0005 0 0.05 0.1 0.15 0.2 0.25 St ra in (in ./in .) Time (sec) Figure 5.33. Panel strain at strip D1 in MSE wall with 16 ft long strip. 1.57 1.48 1.33 -0.35 -0.68 -0.50 -1.15 -0.75 0.95 0 1 2 3 4 5 -1.5 -1 -0.5 0 0.5 1.51 2 Shear Force(kips/ft) Pa ne l H ei gh t ( ft) Figure 5.34. Shear load on the panel.

the panel by ACI specifications (29) (12.9 kN or 2.9 kips) (details of the calculations are presented in Appendix J). 5.3 Bogie Test 5.3.1 5 ft High MSE Wall Construction and Test Installation An elevation of the bogie test wall is shown in Figure 5.37. The total length of bogie test was approximately 18.29 m (60 ft). The MSE wall on which the six precast barrier and coping sections were placed was approximately 1.52 m (5 ft) tall and comprised full and half-panel sections that were approximately 1.52 m (5 ft) wide. The wall panels were placed on a 304.8 mm (1 ft) wide × 15.24 mm (6 in.) thick concrete leveling pad. The MSE wall had two layers of rein- forcement at depths of 262 mm (0.86 ft) and 1.01 m (3.32 ft) below the bottom of the moment slab. The vertical distance between the two reinforcement layers was 0.75 m (2.46 ft). Half of the wall was constructed with 2.43 m (8 ft) rein- forcement with a density of three strips per layer per panel and the other half with 4.88 m (16 ft) reinforcement with a density of two strips per layer per panel. The wall panels were recessed inside the coping sections a distance of 203.2 mm (8 in.). The precast barrier–coping sec- tions rested on a 101.6 mm (4 in.) thick leveling course of 71 -17.13 -17.21 -22.19 -23.22 -15.00 -17.17 -17.97 -18.08 -18.28 0 1 2 3 4 5 -25 -20 -15 -10 -5 0 Vertical Force kips/ft Pa ne l H ei gh t ( ft) Figure 5.35. Vertical load on the panel. 4.15 4.15 4.02 3.07 2.64 2.63 2.22 2.13 2.70 0 1 2 3 4 5 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 Moment (kips-ft)/ft Pa ne l H ei gh t ( ft) Figure 5.36. Bending moment on the panel.

concrete placed on top of the wall panels. The moment slab connecting the precast barrier–coping sections was cast in place in two 9.14 m (30 ft) lengths. The two 1.37 m (4.5 ft) wide, 9.14 m (30 ft) long moment slabs were connected to one another using two No. 9 shear dowels. The backfill for the wall was crushed rock that met the specifications for TxDOT Type A backfill (30). The estimated friction angle for the crushed rock was 35 degrees and the unit weight was 20 kN/m3 (0.125 kips/ft3). The backfill was com- pacted in 0.15 m (6 in.) layers with 10 passes of a 1,320 kg (2,905 lb), 0.89 m (35 in.) wide drum roller. Also, the surface layer of soil was recompacted after each test. A grain size analysis was performed for the backfill material to determine the relative proportions of different grain sizes as shown in Figure 5.38. The particle diameters corresponding to 10% fines (D10) and 60% fines (D60) were 0.075 mm and 6.8 mm, respectively. The coefficient of uniformity [Cu (= D60/D10)] was determined to be 90.67, therefore, the friction factor (F*) at ground level was determined to be 2.0 in accordance with AASHTO LRFD (see Figure 2.5). Selected reinforcement in the MSE wall was instrumented with strain gages to capture the tensile forces transmitted into the reinforcement during the dynamic bogie vehicle impacts. A total of eight strain gages were used for each reference test. 72 4.85' 6 SPACES @ 10' 60' 59'-10 9/16" LENGTH OF BARRIERS LENGTH OF WALL PANELS (1) (4) (3) 8.0' or 16.0'STEEL STRIPS OR BAR MATS LENGTH REFERENCE NUMBER 2' 5'-7 1/2" 2.53' 2.46' 1.20' 2.67' 3/4" 4'-1 3/8" 4'-3 1/2" 3/4" NJ Vertical Vertical NJVertical (2) 16-ft Strip 16-ft Strip 8-ft Strip 8-ft Strip8-ft Bar mats No test plan Half connector Test Order Figure 5.37. Overall elevation of installation for bogie tests. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0.01 0.10 1.00 10.00 Pe rc en t p as si ng (% ) Grain size (mm) Figure 5.38. Particle size distribution curve of the backfill for bogie tests.

The placement of these strain gages was selected to measure the maximum tensile load in each layer of reinforcement as well as give an indication of the distribution of forces in the lateral, longitudinal, and vertical directions. Five strain gages were used on the upper reinforcement layer, and three strain gages were placed on the lower reinforcement layer. Two strain gages were used on both layers of reinforcement adja- cent to the wall panel at the point of impact to provide some redundancy at the location expected to experience maximum tensile loading. A contact switch was placed on the top edge of the traffic face (inside face) of the concrete leveling pad cast on top of the wall panels inside the recess of the coping. The switch indicates the time (referenced from impact) at which the barrier slides and/or rotates sufficiently for the coping to contact the wall panel/level up concrete. The full-height wall panel below the point of impact on the barrier was instrumented with five concrete strain gages to capture normal strains in the panel induced from impact loads transmitted into the MSE wall and generated from direct contact of the barrier–coping section with the top of the wall panel. Two strain gages were placed in a horizontal position along the length of the panel just below the anchorages for the upper layer of reinforcement (region of maximum negative moment) that are below and immediately adjacent to the point of impact. These are the anchorage locations associated with the instrumented reinforcement. Three strain gages were placed in a vertical position along the height of the panel. A strain gage was placed adjacent to the anchorage locations for the upper and lower layer of reinforcement at the point of impact, and one strain gage was placed in the center of the panel between the two layers of reinforcement (region of maximum positive moment). An accelerometer was mounted behind each barrier section at the height of impact to help analyze its dynamic response. An accelerometer also was placed on the end of each of the two 9.14 m (30 ft) long moment slabs at their midpoints to measure any acceleration or motion imparted to the moment slab during impact. Additionally, the bogie vehicle was instru- mented with an accelerometer at its CG. Angular and translational displacements and/or rotation of the barrier and wall panels were determined from high-speed video operating at 1,000 frames/s. Displacement gages were placed at the top and bottom of the precast barrier–coping section and at the upper and lower strip locations on the wall panel to assist with the displacement analysis. String lines were placed behind the barrier and wall to mea- sure their permanent deflection after impact. The four corner points on the barrier–coping sections and five points on each wall panel were measured. The distance from the string lines to the barrier and panel reference points were measured after each test. A 2,268 kg (5,000 lb) bogie vehicle hit each bogie test sec- tion at a speed of approximately 35.41 km/h (22 mph) for the N.J. barrier and 32.19 km/h (20 mph) for the vertical con- crete barriers. Prior numerical simulation results indicated that these velocities would provide sufficient energy to fail the barrier–coping section. By loading each barrier–coping section beyond its ultimate strength, the maximum impact loading transferred into the MSE wall would be measured. The test sequence was selected such that the first two tests involved hitting the barrier segments in the middle of each moment slab. The other two tests were conducted on similar vertical concrete barriers located on the end of each moment slab with different strip length and density. The precast barrier–coping sections, concrete wall panels, and steel strip wall reinforcement were provided by RECO at no cost to the project. RECO also provided supervision of the construction of the wall. The bar mat reinforcement that was used in one of the reference tests was provided by Foster Geo- technical at no cost to the project. Detailed drawings of the ref- erence bogie test installation and construction procedure are presented in Appendix C and Appendix D, respectively, which are available from the NCHRP Report 663 summary web page on the TRB website (www.trb.org) by searching for “NCHRP Report 663”. 5.3.2 Bogie Test 1: New Jersey Barrier with 16 ft Strips The 2,268 kg (5,000 lb) bogie vehicle, shown in Figure 5.39, hit the reference point of the N.J. profile barrier head-on at a speed of 35.08 km/h (21.8 mph). The reference impact point was 101.6 mm (4 in.) from the top of the barrier and 12.7 mm (0.5 in.) from the middle of the barrier to correspond to the location of the reinforcement in the underlying wall. 73 Figure 5.39. Bogie Test 1: N.J. barrier with 16 ft long strip.

Data from Accelerometers The accelerations at the top of the barrier and end of the moment slab exceeded the range set for the accelerometers at these locations. Therefore, a portion of the signal represent- ing the peak acceleration of the barrier and moment slab was truncated as shown in Figure 5.40(b) and (c). Consequently, the maximum 50 msec average acceleration provides lower values than expected. To prevent the reoccurrence of this problem, the range of accelerometer was increased for sub- sequent tests. Data obtained from the bogie-mounted accelerometer were analyzed and the results are presented in Figure 5.41. As shown in Figure 5.41(b), the maximum 50 msec average deceleration was 14.45 g. Based on this acceleration and the mass of the bogie, the maximum 50 msec average impact force was cal- culated to be 326.5 kN (73.4 kips) [see Figure 5.41(a)]. The velocity–time and horizontal displacement–time histories of the bogie are shown in Figure 41(c) and (d), respectively. These time histories were calculated through integration of the accel- eration data. The maximum 50 msec average acceleration of the barrier, as measured by the accelerometer at the top of the barrier, was 7.35 g in the direction of impact [see Figure 5.42(a)]. The velocity–time history of the barrier, as calculated by integra- tion of the raw acceleration data, is shown in Figure 5.42(b). The displacement–time history obtained from integration of the velocity history was known to have some inherent error due to the truncation of data from the barrier accelerometer. Figure 5.42(c) presents displacement–time history from both double integration of acceleration data and from analysis of the high-speed film which is considered to be more accurate. The maximum 50 msec average acceleration of the moment slab is shown in Figure 5.43(a). The velocity–time and horizon- tal displacement–time histories of the moment slab are shown in Figure 5.43(b) and (c), respectively. The velocity–time his- 74 (b) Barrier (c) Moment slab (a) Bogie Figure 5.40. Raw acceleration data of bogie, barrier, and moment slab (Test 1).

tory and displacement–time histories were calculated by inte- gration of the acceleration data. Targets affixed to the displacement bars attached to the top and bottom of the barrier–coping section (see Figures 5.44 and 5.45) were used as reference points to determine angular and translational displacement of the barrier from analysis of high-speed film. From the film analysis, the maximum dynamic displacement of the barrier was 156 mm (6.14 in.) at the top of the barrier and 28.5 mm (1.12 in.) at the bottom of the coping. The permanent displacement of the barrier was 76.2 mm (3 in.) at the top of the barrier and 14 mm (0.55 in.) at the bottom of the coping. Two additional targets affixed to the displacement bars attached to the wall panel at locations corresponding to the upper and lower layers of wall reinforcement were used to determine angular and translational displacement of the panel from analysis of high-speed film. From the film analysis, the maximum dynamic displacement of the panel was 16 mm (0.63 in.) at the upper reinforcement layer of the panel. The permanent displacement of the panel was 6.1 mm (0.24 in.) at the upper reinforcement layer. No movement was mea- sured at the bottom reinforcement layer of the panel. The corresponding displacement–time histories for the barrier– coping section and wall panel are shown in Figure 5.46. Fig- ure 5.47 shows the force–displacement curve for the top of the barrier. Load in Reinforcement Strips As mentioned previously, the wall reinforcement was instru- mented with a total of eight strain gages as shown in Figure 5.48 to capture the tensile forces transmitted into the reinforcement during the dynamic bogie vehicle impacts. To enable com- parison of loads on the reinforcement strips, the strain gage locations were assigned a unique numeric designator. The first number indicates the location of the strain gage, and the 75 (a) Impact Force (b) Deceleration (c) Velocity (d) Displacement Figure 5.41. Force, acceleration, velocity, and displacement of bogie (Test 1).

second number indicates the reinforcement layer. For exam- ple, with reference to Figure 5.48, gage location 1-1 is posi- tioned away from the wall on a strip beneath the impact point in the first (upper) layer of reinforcement. Note that two strain gages were used at locations 2-1 and 2-2 adjacent to the wall panel at the point of impact to provide some measurement redundancy at the location expected to experience maximum tensile loading. One gage was placed on top of the reinforcement and one gage was placed on the bot- tom of the reinforcement. Measurements obtained from the strain gages during testing indicated that the reinforcement experienced some bending in addition to tensile loading. The strain gages on the top and bottom of the reinforcement enabled the bending to be canceled out and the average tensile force in the reinforcement to be calculated. Because of the bending, the average tensile loads obtained at gage locations 2-1 and 2-2 were given more credibility in the analysis of the test data and guideline development than the other locations. Note that these locations correspond to the point of impact and thus are expected to be the location of maximum loading in the reinforcement. These expectations were generally con- firmed by the numerical simulations. Raw data obtained from the strain gages on the strips were analyzed and the results are presented in Figure 5.49. The 50 msec average of the raw data was analyzed to obtain design loads for the strips, and the results are presented in Figure 5.50. The ultimate load obtained for the N.J. barrier was 326.5 kN (73.4 kips), which exceeds the barrier design load of 240 kN (54 kips) prescribed by AASHTO LRFD. To obtain the load on the strips when the barrier impact force equaled the 240 kN (54 kips) design force, the data from the bogie accelerometer [Figure 5.41(a)] were used to find the time at which the design force was reached (0.0198 sec). This time of design force as well as the time of maximum load (0.0331 sec) are shown on Fig- ures 5.49 and 5.50. A complicating factor in the analysis is that the loads in the strips continued to increase after the maximum impact force in the barrier was reached. In other words, the time at which the maximum impact load occurred does not correspond to the time at which the maximum load in the strip occurred. 76 (a) Acceleration (b) Velocity (c) Displacement Figure 5.42. Acceleration, velocity, and displacement of barrier (Test 1).

A contact switch placed on the top edge of the level-up con- crete on top of the wall panels inside the recess of the coping indicated that the coping contacted the wall panel from 0.0784 to 0.1186 sec, which, as shown in Figure 5.50, corresponds to a period of time after maximum impact load. Thus, the barrier–coping section continued in motion under its own momentum as the impact loads were decreasing. This motion likely contributes to the increase in loads in the strips beyond the time of maximum impact load. It is assumed that an impact of lesser severity will follow a similar pattern of behavior. For example, if an impact pro- duces a maximum force of 240 kN (54 kips), one would expect the loads in the strips to increase beyond the values correspond- ing to the time of maximum impact load. Thus, it is not nec- essarily appropriate to use the strip load corresponding to the time at which the 240 kN (54 kips) design load was reached in the bogie impact tests as the design strip load. Assuming the increase in strip load is proportional to the barrier impact load, the design strip loads corresponding to a design impact load of 240 kN (54 kips) can be estimated as follows: Table 5.10 presents a summary of the strip loads from the first bogie impact test including the maximum force, maxi- mum 50 msec average force, and an estimate of the maximum 50 msec average force for a 240 kN (54 kips) design impact. Note that only gage locations 2-1 and 2-2 had two strain gages that could be used to account for bending. To enable compar- ison with other cases, the estimated design load for the strip was expressed in kips per foot of wall. Permanent Deflection of Barrier and Panels The string lines located 1.22 m (4 ft) from the face of the wall panels were used to measure the permanent deflection of barriers and panels at different elevations after bogie vehi- cle impact. After bogie vehicle impact of the N.J. barrier sec- tion, the permanent deflection was measured to be 83 mm (3.27 in.) and 70 mm (2.87 in.) at the top corners of the barrier Estimated Strip Load Maximum Strip L= × 54 73 4. oad -( )5 9 77 (a) Acceleration (b) Velocity (c) Displacement Figure 5.43. Acceleration, velocity, and displacement of moment slab (Test 1).

78 2' -6 1/ 2" 2' -5 1/ 2" 1' -2 1/ 2" 4' -6 " 8' or 16' B ogi e 0. 5' x1' (T YP .) 2' 5" #6 @ 10 " 5- #4 BARS @ 4 EQ . SP AC E 3" 3/ 4" 7" TY P. 43" TY P. 8" 2" 10" 4" 3" 20 " 8" C oncr et e Pa d fo r to e-syst em 6" Di sp l ace me nt Ba rs are lo ca te d on th e li ne of im pac t po in t 18 " 45 1/ 2" 1: 1 TY P. St ri ng Li ne fo r me asurem ent of pe rmanent di sp la cem ent of bar ri er a nd panel s 4' 5' -6 " : A cce le ro me te r (2 ) : Di sp la ce me nt ba r (4 ) : St rain gauge (8 on the strips, 5 on the panel) : Ta pe sw it ch (1 ) Figure 5.44. Side view of installation (Test 1). Figure 5.45. Location of displacement bars affixed on the barrier and panels (Test 1). Figure 5.46. Horizontal displacement of barrier and panel measured from film (Test 1). Figure 5.47. Force–displacement of the top of the barrier (Test 1).

79 Figure 5.48. Location of strain gages and labeling (Test 1). 2-1 (2 gages) 1-1 3-1 4-1 WallPanel WallPanelBarrier Barrier Impact Point 4-2 Impact Point 2-2 (2 gages) (a) Upper layer (b) Bottom layer Figure 5.49. Load on the strips (Raw data, Test 1). Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ip s) -2 0 2 4 6 8 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 73 kips Tape Switch 54 kips 73 kips Tape Switch Figure 5.50. Load on the strips (50 msec avg., Test 1). Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ips ) -2 0 2 4 6 8 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 73 kips Tape Switch 54 kips 73 kips Tape Switch

and 20 mm (0.79 ins) and 13 mm (0.51 in.) at the bottom corners of the coping as shown in Figure 5.51. The perma- nent deflection obtained from the film analysis of the motion of the targets affixed to the barrier–coping section was 76.2 mm (3 in.) at top of the barrier and 14 mm (0.55 in.) at the bottom of the coping. Note that the adjacent barrier–coping sections moved approximately 20 mm (0.79 in.) due to their connection to the same 9.14 m (30 ft) moment slab as the N.J. barrier section that was impacted. The permanent defection of the wall panels was measured as shown in Figure 5.51. The maximum permanent movement measured in the wall panels beneath the impact barrier section was approximately 6 mm (0.24 in.). Note that negative values indicate movement toward the traffic side of the barrier. Such movement may be the result of the panel being loaded eccentrically and rotating. Panel Analysis The wall reinforcement was instrumented with a total of five strain gages to capture the strains in the panel during the bogie impacts as shown in Figure 5.52. The maximum com- pressive strain was 0.0018 and occurred at 0.123 sec (see Fig- ure 5.53) at the location of the uppermost layer of strip. Note that positive values for the vertical direction strain gages indicate movement toward the traffic side of the barrier. The strains at the horizontal centerline of the panel and at the second layer of strips were less than 0.0002. Component Damage Damage to the barrier–coping section resulting from the bogie impact is shown in Figure 5.54. Cracks were observed across the entire length of the N.J. barrier section at a height of approximately 381 mm (15 in.) above ground or above the “toe” of the barrier. The vertical reinforcing bars were exposed along some of these cracks due to fracture and spalling of the concrete. Although difficult to see in the pictures, the cracks radiated upward on either side of the barrier centerline in a U-shaped pattern observed in previous testing of safety shape barriers. This damage mechanism, which had a maximum length of 2.6 m (8.43 ft), was not as pronounced as in past test- 80 Upper Layer (kips) Bottom Layer (kips) Impact Point, Behind (1-1) Impact Point, Front (2-1) Next to Impact Point, Behind (3-1) Next to Impact Point, Front (4-1) Impact Point, Front (2-2) Next to Impact Point, Front (4-2) Maximum load from raw data (t = 0.0685 sec) 4.33 7.78 6.69 5.51 2.9 3.68 Maximum 50 msec avg. load 3.67 7.23* (7.95 top, 6.43 bot.) 5.86 5.83 –1.2* (–0.48 top, –1.92 bot.) 2.71 Estimated design load per strip 2.70 5.29 4.31 4.29 –0.88 2.00 Estimated design load per foot of wall 1.11 2.17 1.77 1.76 –0.36 0.82 * Average of top and bottom loads Table 5.10. Load on the wall reinforcement (Test 1). NJ & 16-ft Strips 5 0 2 0 -2 0 4 -6 0 -2 20 21 83 70 20 13 21 0 6 0 2 3 3 0 -8 -6 -2 -1 1 0 1 1 -2 5 15 1/2" Impact Point Figure 5.51. Permanent deflection of barrier and panels (Test 1, units: mm).

ing because the short length of this precast barrier section caused other failure modes to occur at similar failure loads. Cracking in the soil was observed approximately 1.22 m (48 in.) from the front face of the barrier, which corresponds with the location of the end of the moment slab. The crack, shown in Figure 5.54(c), was about 19 mm (0.75 in.) wide and extended along the entire length of the 9.14 m (30 ft) long moment slab. Although numerous cracks were observed on the back side of the barrier [see Figure 5.54(d)], they were not as wide or pronounced as those on the front of the barrier. Damage to the panel beneath the point of impact on the bar- rier is shown in Figure 5.55. The leveling concrete on top of the wall panel was broken and shifted over the front edge of the panel due to contact with the inside face of the coping. The bonding of the leveling concrete to the top of the wall panel caused the top corner of the wall panel to spall as shown in Fig- ure 5.55(b) and (c). 5.3.3 Bogie Test 2: Vertical Concrete Barrier with 8 ft Bar Mats The second bogie test was conducted on a vertical concrete barrier connected to the mid-span of the undisturbed 9.14 m (30 ft) moment slab adjacent to the moment slab used in Test 1. This vertical barrier section was located above a wall segment that was reinforced with 2.43 m (8 ft) long bar mats. The 2,268 kg (5,000 lb) bogie vehicle, shown in Figure 5.56, impacted the reference point of the vertical barrier head-on at a speed of approximately 32.7 km/h (20.3 mph). The reference point was along the top edge of the barrier and approximately 0.37 m (14 5⁄8 in.) from its centerline to coincide with the location of one of the two bar mats comprising each of the two layers of wall reinforcement below the barrier. Data from Accelerometers As previously discussed, the range of the accelerometers attached to the top of the barrier and end of the moment slab was increased after Test 1. However, as shown in Figure 5.57(b), some of the accelerations still exceeded the revised 81 3 3/4" TPY. AA A-A 3 3/4" TPY. 17" 17" 29-1/8" 14-3/8" 14-1/8" 14-3/8" 14-1/8" Figure 5.52. Location of concrete strain gages (Test 1). -0.002 -0.0015 -0.001 -0.0005 0 0.0005 0 0.05 0.1 0.15 0.250.2 0.3 St ra in (in ./in .) Time (sec) Vertical_top Vertical_middle Vertical_bottom Horiz._right Horiz._left Figure 5.53. Strain on the panel (Test 1).

82 (a) Front side of the barrier Impact Point 8.43 ft 15-in. (b) Closeup of front side of the barrier Figure 5.54. Damage to barrier after Test 1.

range set for the barrier accelerometer. Consequently, these data must be analyzed with appropriate caution. To prevent further occurrence of this problem, the range of the barrier accelerometer was increased substantially for subsequent tests. Data obtained from the bogie-mounted accelerometer were analyzed and the results are presented in Figure 5.58. As shown in Figure 5.58(b), the maximum 50 msec average deceleration was 13.01 g. Based on this acceleration and the mass of the bogie, the maximum 50 msec average impact force was calculated to be 294.03 kN (66.1 kips) [see Figure 5.58(a)]. The velocity–time and horizontal displacement–time histories of the bogie are shown in Figure 5.58(c) and (d), respectively. These time histories were calculated through integration of the acceleration data. The maximum 50 msec average acceleration of the barrier, as measured by the accelerometer at the top of the barrier, was 10.71 g in the direction of impact [see Figure 5.59(a)]. The velocity–time history of the barrier, as calculated by integra- tion of the raw acceleration data, is shown in Figure 5.59(b). The displacement–time history obtained from integration of the velocity history is shown in Figure 5.59(c). The raw acceleration–time history for the moment slab is shown in Figure 5.57(c). The increased range of the accelero- meter was sufficient to obtain the peak acceleration of the moment slab. However, after impact, the accelerometer had a non-zero offset. The problem was traced to a connection issue that was resolved prior to the next test on this moment slab. The 50 msec average acceleration for the moment slab and the 83 (c) Side view of the barrier (d) Back view of the barrier Figure 5.54. (Continued).

associated velocity–time and displacement–time histories are shown in Figure 5.60. Targets affixed to the displacement bars attached to the top and bottom of the barrier–coping section (see Figures 5.61 and 5.62) were used as reference points to determine angular and translational displacement of the barrier from analysis of high- speed film. From the film analysis, the maximum dynamic dis- placement of the barrier was 153.4 mm (6.04 in.) at the top of the barrier and 24 mm (0.93 in.) at the bottom of the coping. The permanent displacement of the barrier was 101.6 mm (4 in.) at the top of the barrier and 12.7 mm (0.5 in.) at the bot- tom of the coping. Two additional targets affixed to the displacement bars attached to the wall panel at locations corresponding to the upper and lower layers of wall reinforcement were used to determine angular and translational displacement of the panel from analysis of high-speed film. From the film analysis, the maximum dynamic displacement of the panel was 94 mm (0.37 in.) at the upper reinforcement layer of the panel and 2.54 mm (0.1 in.) at the bottom reinforcement layer. The per- manent displacement of the panel was 5.08 mm (0.2 in.) at the upper reinforcement layer and 0.51 mm (0.02 in.) at the bot- tom reinforcement layer. The corresponding displacement–time histories for the barrier–coping section and wall panel are shown in Figure 5.63. Figure 5.64 shows the force–displacement curve for the top of the barrier. Load in Reinforcement Bar Mats A total of eight strain gages were used to instrument the bar mats to capture the tensile forces transmitted into the rein- forcement during the dynamic bogie vehicle impact. The locations of strain gages were assigned a numeric designator 84 (a) Surface of panel (b) Top of panel inside of recessed coping Bottom of Coping Surface of Panel Leveling Pad (c) Closeup of panel inside of recessed coping Leveling Pad Surface of Panel Figure 5.55. Damage to panel and leveling pad (Test 1). Figure 5.56. Test 2: vertical concrete barrier with 8 ft bar mats.

as shown in Figure 5.65. Note that two strain gages were used at locations 2-1 and 2-2 adjacent to the wall panel at the point of impact to provide some measurement redundancy at the location expected to experience maximum tensile loading. One gage was placed on top of the reinforcement and one gage was placed on the bottom of the reinforcement. Mea- surements obtained from the strain gages during testing indi- cated that the reinforcement experienced some bending in addition to tensile loading. The strain gages on the top and bottom of the reinforcement enabled the bending to be can- celed out and the average tensile force in the reinforcement to be calculated. Raw data obtained from the strain gages on the bar mats were analyzed and the results are presented in Figure 5.66. It can be seen in this figure that the contact switch placed on the top edge of the level-up concrete on top of the wall panels inside the recess of the coping indicated that the coping con- tacted the wall panel from 0.0806 to 0.1798 sec. The 50 msec average of the raw data was analyzed to obtain design loads for the strips, and the results are presented in Figure 5.67. As with the strips in Test 1, there was some increase in force in the bar mats after the time of maximum barrier impact load. The maximum 50 msec average design strip loads corresponding to a design impact load of 240 kN (54 kips) were estimated as follows: where 294.03 kN (66.1 kips) is the maximum 50 msec average impact load measured for the vertical wall barrier over 2.43 m (8 ft) bar mats. The interior wires of a bar mat are more highly stressed than exterior wires due to the soil interaction on the cross bars between the longitudinal wires. Because the strain gages were installed in the exterior wire, the interior wires may be subject to tensions that are 1.34 times the exterior wire tensions. Estimated Strip Load Maximum Strip L= × 54 66 1. oad -( )5 10 85 (a) Bogie (b) Barrier (c) Moment slab Figure 5.57. Raw acceleration data of bogie, barrier, and moment slab (Test 2).

86 (c) Velocity (d) Displacement (a) Impact force (b) Deceleration Figure 5.58. Force, acceleration, velocity, and displacement of bogie (Test 2).

87 (b) Velocity (c) Displacement (a) Acceleration Figure 5.59. Acceleration, velocity, and displacement of barrier (Test 2).

88 (b) Velocity (c) Displacement (a) Acceleration Figure 5.60. Acceleration, velocity, and displacement of moment slab (Test 2). Bo gi e #6 @ 10" 5- #4 BARS @ 4 EQ. SPAC E 2' -6 1/2" 2' -5 1/2" 1'-2 1/2 " 4' -6 " 3" 3/4 " 7" TYP. 43" TYP . 8" 2" 10 " 4" 8' or 16' 3" 20" 8" 2' Concrete Pa d 6" 5" 18" 45 1/2 " Displacement Bars are located on the line of impact point 1:1 TYP. String Line for measurement of permanent di splacement of barrier and panel s : Accelerometer (2) : Displacement bar (4) : Strain gauge (8 on the strips, 5 on the panel) : Tape switch (1) 4' 5'-6" Figure 5.61. Side view of installation (Test 2, 3, and 4).

89 Figure 5.62. Location of displacement bars affixed on the barrier and panels (Test 2). Figure 5.63. Horizontal displacement of barrier and panel measured from film (Test 2). Figure 5.64. Force–displacement of the top of the barrier (Test 2). 1-1 3-1 4-1 Barrier Impact Point Wall Panel 2-1 (2 gages) 4-2 Barrier Impact Point Wall Panel 2-2 (2 gages) (a) Upper layer (b) Bottom layer Figure 5.65. Location of strain gages and labeling (Test 2). Table 5.11 presents a summary of the forces in the bar mat obtained from the second bogie impact test including the maximum force, maximum 50 msec average force, and an estimate of the maximum 50 msec average force for a 240 kN (54 kips) design impact. To enable comparison with other cases, the estimated design load for the strips was expressed in kips per foot of wall. Permanent Deflection of Barrier and Panels The string lines located 1.22 m (4 ft) from the face of wall panels were used to measure the permanent deflection of bar- riers and panels after the bogie vehicle impact at different ele- vations. The permanent deflection was measured to be 99 mm (3.9 in.) and 45 mm (1.77 in.) at the top and 13 mm (0.51 in.) and 18 mm (0.71 in.) at the bottom corners of the coping as shown in Figure 5.68. Note that the reference impact point was 0.37 m (14.625 in.) left of the centerline of the barrier– coping section as shown in Figure 5.68 to align with the instru- mented bar mats. This location corresponded to the side of the barrier with greater movement. Note that the barrier–coping sections to the left and right of the section that was hit had permanent movement at the top

Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ips ) 0 1 2 3 4 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 66.1 kips Tape Switch 54 kips 66.1 kips Tape Switch Figure 5.66. Load on the strips (Raw data, Test 2). Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ip s) 0 1 2 3 4 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 66.1 kips Tape Switch 54 kips 66.1 kips Tape Switch Figure 5.67. Load on the strips (50 msec avg., Test 2). Upper layer (kips) Bottom layer (kips) Impact Point, Behind (1-1) Impact Point, Front (2-1) Next to Impact Point, Behind (3-1) Next to Impact Point, Front (4-1) Impact Point, Front (2-2) Next to Impact Point, Front (4-2) Maximum load from raw data (t = 0.072 sec) 1.86 1.57 3.52 2.21 0.11 0.72 Maximum 50 msec avg. load (t = 0.0607 sec) 1.76 1.54* (1.44 top 1.63 bot.) 3.06 2.07 0.076* (–0.16 top 0.31 bot.) 0.31 Estimated design load of interior wire 1.931 1.68 3.349 2.271 0.083 0.344 Estimated design load per foot of wall 2.77 2.41 4.8 3.258 0.119 0.494 * Average of top and bottom loads Table 5.11. Load on the wall reinforcement (Test 2).

of the barrier of 25 mm (1 in.) and 32 mm (1.26 in.), respec- tively, due to their connection to the same 9.14 m (30 ft) moment slab as the vertical barrier section that was hit. The permanent defection of the wall panels was measured as shown in Figure 5.68. The maximum permanent movement mea- sured in the wall panels beneath the impact barrier section was approximately 5 mm (0.2 in.). Panel Analysis The wall reinforcement was instrumented with a total of five strain gages to capture the resistance of the panel during the bogie impacts as shown in Figure 5.69. The maximum compressive strain of 0.00016 occurred at 0.056 sec (see Fig- ure 5.70) at the location of the uppermost layer of strips. Note that positive values for the vertical direction strain gages indi- cate movement toward the traffic side of the barrier. The strains at the horizontal centerline of the panel and at the sec- ond layer of strips were 0.00012 and 0.00007, respectively. Component Damage Damage to the vertical wall barrier–coping section resulting from the bogie impact is shown in Figure 5.71. Left of the point of impact, the vertical barrier failed in a classical yield line pat- tern by developing a vertical “hinge” line at the point of impact and a diagonal hinge line radiating from the bottom to the top of the barrier. However, because the bogie impact point was off- set 0.37 m (14.625 in.) from the centerline of the barrier to align with the bar mat, there was insufficient strength to develop a similar yield line failure on the right side of the barrier. Rather, a lower strength cantilever bending moment controlled failure mode on the right side of the barrier. Longitudinal cracks devel- oping from this failure mode were observed at the groundline connection between the barrier and coping and approximately 152 mm (6 in.) above the groundline from the point of impact to the end of the vertical barrier [see Figure 5.71(b)]. Cracking in the soil was observed approximately 1.22 m (48 in.) from the front face of the barrier, which corresponds with the location of the end of the moment slab. The crack, shown in Figure 5.71(c), was about 13 mm (0.5 in.) wide and extended along the entire length of the 9.15 m (30 ft) long moment slab. Damage to the back of the vertical barrier, shown in Figure 5.71(e) and (f), mirrors that on the front face of the barrier. Most pronounced are the diagonal hinge line and the longitudinal crack at the interface between the vertical barrier and coping. Damage to the panel beneath the point of impact on the barrier is shown in Figure 5.72. As shown in Figure 5.72(a), the surface of the panel showed little distress. The leveling con- crete on top of the wall panel was broken and shifted over the 91 Vertical & 8-ft Bar mats 11 17 99 18 32 20 2 1 2 0 0 1 62 0 0 2 -1 0 5 4 4 3 -1 0 0 9 -3 5 13 4 7 61 0 0 25 6 45 13 27 16 14 5/8" Impact Point Figure 5.68. Permanent deflection of barrier and panels (Test 2, units: mm). 2 1/2'' TYP. 14-3/8'' 14-3/8''17'' 29-1/8'' 22-7/8''22-7/8'' 25-1/2'' 2 1/2'' TYP. Figure 5.69. Location of concrete strain gages (Test 2).

front edge of the panel due to contact with the inside face of the coping. The bonding of the leveling concrete to the top of the wall panel caused the top edge of the wall panel to spall as shown in Figure 5.72(b). 5.3.4 Bogie Test 3: Vertical Concrete Barrier with 8 ft Strips Upon completion of the first two reference tests, the soil on and around the two moment slabs was recompacted for the following tests. The third bogie test was conducted on a ver- tical concrete barrier connected to the end of the 9.14 m (30 ft) moment slab above the wall section with 2.43 m (8 ft) long steel reinforcement strips. The 2,268 kg (5,000 lb) bogie vehicle, shown in Figure 5.73, hit the reference point of the vertical barrier head-on at a speed of approximately 32.5 km/h (20.2 mph). The reference point was along the top edge of the bar- rier and approximately 121 mm (4.75 in.) from its centerline to coincide with the location of a steel strip in the wall rein- forcement below the barrier. Data from Accelerometers The increased range used for the barrier and moment slab accelerometers was sufficient to capture the accelerations arising from the bogie impact. The raw acceleration data for the bogie, barrier, and moment slab are shown in Figure 5.74. Data obtained from the bogie-mounted accelerometer were analyzed and the results are presented in Figure 5.75. As shown in Figure 5.75(b), the maximum 50 msec average deceleration was 13.82 g. Based on this acceleration and the mass of the bogie, the maximum 50 msec average impact force was cal- culated to be 312.13 kN (70.17 kips) [see Figure 5.75(a)]. The velocity–time and horizontal displacement–time histories of the bogie are shown in Figure 5.75(c) and (d), respectively. These time histories were calculated through integration of the acceleration data. The maximum 50 msec average acceleration of the barrier, as measured by the accelerometer at the top of the barrier, was 10.16 g in the direction of impact [see Figure 5.76(a)]. The velocity–time history of the barrier, as calculated by integra- tion of the raw acceleration data, is shown in Figure 5.76(b). The displacement–time history obtained from integration of the velocity history is shown in Figure 5.76(c). The 50 msec average acceleration for the moment slab is shown in Figure 5.77(a). The velocity–time and vertical displacement–time histories of the moment slab are shown in Figure 5.77(b) and (c), respectively. The velocity–time history and displacement–time histories were calculated by integra- tion of the raw acceleration data. Targets affixed to the displacement bars attached to the top and bottom of the barrier–coping section (see Figures 5.61 and 5.78) were used as reference points to determine angular and translational displacement of the barrier from analysis of high- speed film. From the film analysis, the maximum dynamic dis- placement of the barrier was 131 mm (5.17 in.) at the top of the barrier and 30 mm (1.16 in.) at the bottom of the coping. The permanent displacement of the barrier was 63.5 mm (2.5 in.) at the top of the barrier and 15 mm (0.6 in.) at the bottom of the coping. 92 -0.0002 -0.00015 -0.0001 -0.00005 0 0.00005 0 0.05 0.1 0.15 0.250.2 0.3 St ra in (in ./in .) Time (sec) Vertical_top Vertical_middle Vertical_bottom Horiz._right Horiz._left Figure 5.70. Strain on the panel (Test 2).

93 (a) Front view of the barrier Impact Point 6.88 ft (b) Close-up front side of the barrier Figure 5.71. Cracks on the barrier after test (Test 2). (continued on next page)

94 (c) Side view of the barrier (d) Closeup side view of the barrier (e) Back view of the barrier Figure 5.71. (Continued).

95 (f) Closeup back view of the barrier Figure 5.71. (Continued). (a) Surface of panel (b) Inside of panel Bottom of Coping Surface of Panel Leveling Pad Figure 5.72. Cracks on the panel and leveling pad after test (Test 2). Figure 5.73. Test 3: Vertical wall barrier with 8 ft long strip. Two additional targets affixed to the displacement bars attached to the wall panel at locations corresponding to the upper and lower layers of wall reinforcement were used to determine angular and translational displacement of the wall panel from analysis of high-speed film. From the film analysis, the maximum dynamic displacement of the panel was 23 mm (0.92 in.) at the upper reinforcement layer and 5 mm (0.19 in.) at the bottom reinforcement layer. The per- manent displacement of the panel was 14 mm (0.55 in.) at the upper reinforcement layer and 5 mm (0.18 in.) at the bottom reinforcement layer. The corresponding displacement–time histories for the barrier–coping section and wall panel obtained from film analysis are shown in Figure 5.79. Figure 5.80 shows the impact force–displacement curve for the top of the barrier.

96 (b) Barrier (c) Moment slab (a) Bogie Figure 5.74. Raw acceleration data of bogie, barrier, and moment slab (Test 3).

97 (c) Velocity (d) Displacement (a) Impact force (b) Deceleration Figure 5.75. Force, acceleration, velocity, and displacement of bogie (Test 3).

98 (b) Velocity (c) Displacement (a) Acceleration Figure 5.76. Acceleration, velocity, and displacement of barrier (Test 3). (b) Velocity (c) Displacement (a) Acceleration Figure 5.77. Acceleration, velocity, and displacement of moment slab (Test 3).

99 Load in the Reinforcement Strips A total of eight strain gages were used to instrument the strips to capture the tensile forces transmitted into the reinforcement during the dynamic bogie vehicle impact. To enable compari- son of loads on the strips, the locations of strain gages were assigned a numeric designator as shown in Figure 5.81. Note that two strain gages were used at locations 2-1 and 2-2 adja- cent to the wall panel at the point of impact to provide some measurement redundancy at the location expected to experi- ence maximum tensile loading. One gage was placed on top of the reinforcement and one gage was placed on the bottom of the reinforcement. Measurements obtained from the strain gages during testing indicated that the reinforcement experi- enced some bending in addition to tensile loading. The strain gages on the top and bottom of the reinforcement enabled the bending to be canceled out and the average tensile force in the reinforcement to be calculated. Raw data obtained from the strain gages on the bar mats were analyzed and the results are presented in Figure 5.82. A con- tact switch placed on the top edge of the level-up concrete on Figure 5.78. Location of displacement bars affixed on the barrier and panels (Test 3). Figure 5.79. Horizontal displacement of barrier and panel measured from film (Test 3). Figure 5.80. Force–displacement of the top of the barrier (Test 3). 1-1 3-1 4-1 Barrier Wall Panel Impact Point 2-1 (2 gages) 4-2 Barrier Wall Panel Impact Point 2-2 (2 gages) (a) Upper layer (b) Bottom layer Figure 5.81. Location of strain gages and labeling (Test 3).

in this figure, there was some increase in force in the strips observed after the time of maximum barrier impact load, but the increase was not as significant as that seen in Test 1. The maximum 50 msec average design strip loads corresponding to a design impact load of 240 kN (54 kips) were estimated as follows: top of the wall panels inside the recess of the coping activated twice from 0.0522 to 0.115 sec and from 0.1605 to 0.2385 sec, which indicates that the coping was in contact with the wall panel and/or leveling concrete during these times. The 50 msec average of the raw data was analyzed to obtain design loads for the strips, and the results are presented in Figure 5.83. As shown 100 Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ips ) -2 0 2 4 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 70.17 kips Tape Switch 54 kips 70.17 kips Tape Switch Tape Switch Figure 5.82. Load on the strips (Raw data, Test 3). Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ips ) -2 0 2 4 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 70.17 kips Tape Switch 54 kips 70.17 kips Tape Switch Tape Switch Figure 5.83. Load on the strips (50 msec avg., Test 3).

101 where 312.13 kN (70.17 kips) is the maximum 50 msec average impact load measured for the vertical wall barrier over 2.43 m (8 ft) strips. Table 5.12 presents a summary of the forces in the steel strips obtained from the third bogie impact test including the maximum force, maximum 50 msec average force, and an estimate of the maximum 50 msec average force for a 240 kN (54 kips) design impact. To enable comparison with other cases, the estimated design load for the strips was expressed in kips per foot of wall. Permanent Deflection of Barrier and Panels The string lines located 1.22 m (4 ft) from the face of the wall panels were used to measure the permanent deflection of barriers and panels after bogie vehicle impact. After the bogie vehicle impact of the vertical barrier, the permanent deflection was measured to be 63 mm (2.48 in.) and 53 mm (2.09 in.) at Estimated Strip Load Maximum Strip= × 54 70 17. Load -( )5 11 the top corners of the barrier and 17 mm (0.7 in.) and 15 mm (0.59 in.) at the bottom corners of the coping as shown in Fig- ure 5.84. Note that the reference impact point was offset 95 mm (3.75 in.) left from the centerline of the barrier–coping section as shown in Figure 5.84 to align with the steel strip reinforce- ment. The left side of the barrier on Figure 5.84 was therefore slightly closer to the reference impact point. The barrier–coping sections to the left of the section that was impacted had permanent movement at the top of the barrier of 24 mm (0.95 in.). This indicates that the shear dow- els are effective in transferring load to the adjacent moment slab. The permanent defection of the wall panels was measured as shown in Figure 5.84. The maximum permanent move- ment measured in the wall panels beneath the impact barrier section was approximately 16 mm (0.63 in.), which was about three times the movement observed in the previous tests with the 4.88 m (16 ft) strips and 2.43 m (8 ft) bar mats. The mag- nitude of movement appears to indicate that the strips reached their pullout capacity. This conclusion is supported by the lower loads measured in the strips for this test compared to the previous tests. Upper layer (kips) Bottom layer (kips) Impact Point, Behind (1-1) Impact Point, Front (2-1) Next to Impact Point, Behind (3-1) Next to Impact Point, Front (4-1) Impact Point, Front (2-2) Next to Impact Point, Front (4-2) Maximum load from raw data (t = 0.0643 sec) –0.11 1.76 0.37 0.53 1.38 4.46 Maximum 50 msec avg. load (t = 0.0635 sec) 0.24 2.13* (–0.3 top 4.56 bot.) 0.06 0.79 1.19* (0.98 top 1.4 bot.) 3.71 Estimated design load 0.18 1.64 0.05 0.60 0.92 2.85 Estimated design load per foot of wall 0.11 1.01 0.03 0.37 0.57 1.76 * Average of top and bottom loads Table 5.12. Load on the wall reinforcement (Test 3). Vertical & 8-ft Strips 2 2 -3 5 3 6 8 21 63 15 15 6 5 1 -1 0 2 0 3 12 0 1 1 3 0 2 5 1 -1 9 2 2 -1 14 1 8 16 3 14 18 2 3 3 7 1 4 4 -3 1 1 -2 -1 0 24 8 53 17 9 14 12 3 3/4" Impact Point Figure 5.84. Permanent deflection of barrier and panels (Test 3, units: mm).

Panel Analysis The wall reinforcement was instrumented with a total of five strain gages to capture the resistance of the panel during the bogie impacts as shown in Figure 5.85. The maximum compressive strain of 0.00022 occurred at 0.056 sec (see Fig- ure 5.86) at the location of the uppermost layer of strips. Note that positive values for the vertical direction strain gages indi- cate movement toward the traffic side of the barrier. The strains at the horizontal centerline of the panel and at the second layer of strips were 0.00018 and 0.00014, respectively. Component Damage Damage to the vertical wall barrier–coping section result- ing from the bogie impact is shown in Figure 5.87. Although difficult to discern from the photos because the cracks are not large, the vertical barrier failed in a classical yield line pat- tern by developing a vertical “hinge” line at the point of impact and two diagonal hinge lines radiating from the bottom to the top of the barrier on either side of impact. As shown in Fig- ure 5.87(b), three cracks were observed along the diagonal hinge lines. The length between the two inside most cracks on either side of the impact point was 1.98 m (6.51 ft). The lengths between the middle and outer sets of cracks were 2.19 m (7.17 ft) and 2.52 m (8.28 ft), respectively. Cracking in the soil was observed approximately 1.22 m (48 in.) from the front face of the barrier, which corresponds with the location of the end of the moment slab. The cracking, 102 2 1/2'' TYP. 14-3/8'' 14-3/8''17'' 29-1/8'' 22-7/8''22-7/8'' 25-1/2'' 2 1/2'' TYP. Figure 5.85. Location of concrete strain gages (Test 3). -0.0002 -0.00015 -0.00025 -0.0001 0.0001 -0.00005 0 0.00005 0 0.05 0.1 0.15 0.250.2 0.3 St ra in (in ./in .) Time (sec) Vertical_top Vertical_middle Vertical_bottom Horiz._right Horiz._left Figure 5.86. Strain on the panel (Test 3).

103 (a) Front view of the barrier Impact Point 8.28 ft 7.17 ft 6.51 ft (b) Closeup front view of the barrier (1) (2) (3) Figure 5.87. Cracks on the barrier after test (Test 3). (continued on next page)

104 (c) Side view of the barrier (d) Top view of the barrier (e) Back view of the barrier Figure 5.87. (Continued).

105 shown in Figure 5.87(c), extended 18.29 m (60 ft) along the entire length of both moment slabs. This indicates that the two No. 9 shear dowels placed between the moment slabs were able to transfer substantial load to the adjacent moment slab. Damage to the back of the vertical barrier is shown in Figure 5.87(e). Damage to the panel beneath the point of impact on the barrier is shown in Figure 5.88. As shown in this figure, the surface of the panel showed little distress. 5.3.5 Bogie Test 4: Vertical Concrete Barrier with 16 ft Strips Test 4 was similar to Test 3 with the exception that the ver- tical barrier was installed over a segment of wall having 4.88 m (16 ft) long reinforcement strips. As in Test 3, the vertical bar- rier was connected to the end of a 9.14 m (30 ft) moment slab doweled to the adjacent moment slab using two No. 9 bars. The 2,268 kg (5,000 lb) bogie vehicle as shown in Figure 5.89 hit the center of the vertical barrier head-on at a speed of approximately 32.5 km/h (20.2 mph), which was the same as reference Test 3. The reference point was along the top edge of the barrier and approximately 41.3 mm (1.625 in.) from its centerline to coincide with the location of a steel strip in the wall reinforcement below the barrier. Data from Accelerometers The raw acceleration data for the bogie, barrier, and moment slab are shown in Figure 5.90. Data obtained from the bogie-mounted accelerometer were analyzed and the results are presented in Figure 5.91. As shown in Figure 5.91(b), the maximum 50 msec average deceleration was 12.69 g. Based on this acceleration and the mass of the bogie, the maximum 50 msec average impact force was calculated to be 286.55 kN (64.42 kips) [see Figure 5.91(a)]. The velocity–time and horizontal displacement–time histories of the bogie are shown in Figure 5.91(c) and (d), respectively. These time histories were calculated through integration of the accel- eration data. The maximum 50 msec average acceleration of the barrier, as measured by the accelerometer at the top of the barrier, was 13.04 g in the direction of impact [see Figure 5.92(a)]. The velocity–time history of the barrier, as calculated by inte- gration of the raw acceleration data, is shown in Figure 5.92(b). The displacement–time history obtained from integration of the velocity history is shown in Figure 5.92(c). The 50 msec average acceleration for the moment slab is shown in Figure 5.93(a). The velocity–time history of the moment slab is shown in Figure 5.93(b). The velocity–time his- tory was calculated by integration of the raw acceleration data. Targets affixed to the displacement bars attached to the top and bottom of the barrier–coping section (see Figures 5.61 and 5.94) were used as reference points to determine angular and translational displacement of the barrier from analysis of high- speed film. From the film analysis, the maximum dynamic displacement of the barrier was 153 mm (6.02 in.) at the top of the barrier and 17.5 mm (0.69 in.) at the bottom of the cop- ing. The permanent displacement of the barrier was 76.2 mm (3 in.) at the top of the barrier and 5.6 mm (0.22 in.) at the bot- tom of the coping. Two additional targets affixed to the displacement bars attached to the wall panel at locations corresponding to the upper and lower layers of wall reinforcement were used to Figure 5.88. Panel surface after test (Test 3). Figure 5.89. Test 4: Vertical wall barrier with 16 ft long strips.

106 (b) Barrier (c) Moment slab (a) Bogie Figure 5.90. Raw acceleration data of bogie, barrier, and moment slab (Test 4).

107 (c) Velocity (d) Displacement (a) Impact force (b) Deceleration Figure 5.91. Force, acceleration, velocity, and displacement of bogie (Test 4).

108 (b) Velocity (c) Displacement (a) Acceleration Figure 5.92. Acceleration, velocity, and displacement of barrier (Test 4). (b) Velocity(a) Acceleration Figure 5.93. Acceleration and velocity of moment slab (Test 4).

109 determine angular and translational displacement of the wall panel from analysis of high-speed film. From the film analysis, the maximum dynamic displacement of the panel was 7.6 mm (0.3 in.) at the upper reinforcement layer of the panel and 1.8 mm (0.07 in.) at the bottom reinforcement layer. The per- manent displacement of the panel was 1.8 mm (0.07 in.). There was little discernable movement of the panel at the bottom reinforcement layer. The corresponding displacement–time histories for the barrier–coping section and wall panel obtained from film analysis are shown in Figure 5.95. Figure 5.96 shows the impact force–displacement curve for the top of the barrier. Load in the Reinforcement Strips A total of eight strain gages were used to instrument the strips to capture the tensile forces transmitted into the rein- forcement during the dynamic bogie vehicle impact. To enable comparison of loads on the strips, the locations of strain gages were assigned a numeric designator as shown in Figure 5.97. Note that two strain gages were used at locations 2-1 and 2-2 Figure 5.94. Location of displacement bars affixed on the barrier and panels (Test 4). Figure 5.95. Horizontal displacement of barrier and panel measured from film (Test 4). Figure 5.96. Force–displacement of the top of the barrier (Test 4). 2-1 (2 gages) 1-1 3-1 4-1 Barrier Wall Panel Impact Point 2-2 (2 gages) 4-2 Barrier Wall Panel Impact Point (a) Upper layer (b) Bottom layer Figure 5.97. Location of strain gages and labeling (Test 4).

110 Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ips ) 0 2 4 6 8 10 12 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 64.42 kips Tape Switch 54 kips 64.42 kips Tape Switch Tape Switch Figure 5.98. Load on the strips (Raw data, Test 4). adjacent to the wall panel at the point of impact to provide some measurement redundancy at the location expected to experience maximum tensile loading. One gage was placed on top of the reinforcement and one gage was placed on the bot- tom of the reinforcement. Measurements obtained from the strain gages during testing indicated that the reinforcement experienced some bending in addition to tensile loading. The strain gages on the top and bottom of the reinforcement enabled the bending to be canceled out and the average tensile force in the reinforcement to be calculated. Raw data obtained from the strain gages on the bar mats were analyzed and the results are presented in Figure 5.98. A contact switch placed on the top edge of the level-up concrete on top of the wall panels inside the recess of the coping activated twice from 0.0655 to 0.1183 sec and from 0.173 to 0.1802 sec, which indicates that the coping was in contact with the wall panel and/or leveling concrete during these times. The 50 msec average of the raw data was analyzed to obtain design loads for the strips, and the results are presented in Figure 5.99. As shown in this figure, there was some increase in force in the strips observed after the time of maximum barrier impact load, but the increase was not as significant as that seen in Test 1. The maximum 50 msec average design strip loads correspond- ing to a design impact load of 240 kN (54 kips) were estimated as follows: Estimated Strip Load Maximum Strip= × 54 64 42. Load -( )5 12 where 286.55 kN (64.42 kips) is the maximum 50 msec aver- age impact load measured for the vertical wall barrier over 4.88 m (16 ft) strips. Table 5.13 presents a summary of the forces in the steel strips obtained from the fourth bogie impact test including the maximum force, maximum 50 msec average force, and an estimate of the maximum 50 msec average force for a 240 kN (54 kips) design impact. To enable comparison with other cases, the estimated design load for the strips was expressed in kips per foot of wall. Permanent Deflection of Barrier and Panels The string lines located 4 ft from the face of wall panels were used to measure the permanent deflection of barriers and panels after bogie vehicle impact at different elevations. After bogie vehicle impact of the vertical barrier, the permanent deflection was measured to be 67 mm (2.64 in.) and 68 mm (2.68 in.) at the top corners of the barrier and 2 mm (0.08 in.) and 8 mm (0.31 in.) at the bottom corners of the coping as shown in Figure 5.100. Note that the reference impact point was offset 41.3 mm (1.625 in.) left of the centerline of the barrier–coping section as shown in Figure 5.100 to align with the steel strip reinforcement. The permanent defection of the wall panels was measured as shown in Figure 5.100. The maximum permanent move- ment measured in the wall panels beneath the impact barrier section was only about 4 mm (0.16 in.), which is considerably

111 Time (sec) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Lo ad (k ips ) 0 2 4 6 8 10 12 1-1 2-1(2 gages) 3-1 4-1 2-2(2 gages) 4-2 54 kips 64.42 kips Tape Switch 54 kips 64.42 kips Tape Switch Tape Switch Figure 5.99. Load on the strips (50 msec avg., Test 4). Upper layer (kips) Bottom layer (kips) Impact Point, Behind (1-1) Impact Point, Front (2-1) Next to Impact Point, Behind (3-1) Next to Impact Point, Front (4-1) Impact Point, Front (2-2) Next to Impact Point, Front (4-2) Maximum load from raw data (t = 0.054 sec) 11.59 8.08 10.25 12.51 –0.16 1.25 Maximum 50 msec avg. load (t = 0.0645 sec) 10.75 7.46* (7.53 top 7.40 bot.) 7.20 9.58 0.15* (–2.44 top 2.74 bot.) 0.73 Estimated design load 9.01 6.25 6.03 8.03 0.13 0.62 Estimated design load per foot of wall 3.70 2.57 2.48 3.30 0.05 0.25 * Average of top and bottom loads Table 5.13. Load on the wall reinforcement (Test 4). Vertical & 16-ft Strips 1 2 1 4 0 1 -2 0 5 0 0 3 11 7 3 10 67 8 2 4 -1 0 0 0 1 -1 0 -8 -5 0 0 3 0 2 1 0 4 1 -6 -1 5 6 1 3 5 -1 3 0 0 -1 0 2 0 68 2 -1 11 4 1 5/8" Impact Point Figure 5.100. Permanent deflection of barrier and panels (Test 4, units: mm).

less than the movement observed in the previous test of the vertical barrier with the 8 ft strips. Panel Analysis The wall reinforcement was instrumented with a total of five strain gages to capture the resistance of the panel during the bogie impacts as shown in Figure 5.101. The maximum com- pressive strain of 0.0024 occurred at 0.11 sec (see Figure 5.102) at the location of uppermost layer of strips. Note that positive values for the vertical direction strain gages indicate movement toward the traffic side of the barrier. As shown in Figure 5.102, the strain dropped suddenly after 0.05 sec. The strains at the horizontal centerline of the panel and at the second layer of strips were 0.00014 and 0.00008, respectively. Component Damage Damage to the vertical wall barrier–coping section result- ing from the bogie impact is shown in Figure 5.103. Although difficult to discern from the photos because some of the crack widths are not large, the vertical barrier has characteristics of a classical yield line pattern by developing a vertical “hinge” line at the point of impact and two diagonal hinge lines radi- ating from the bottom to the top of the barrier on either side of impact. As shown in Figure 5.103(b), several cracks were observed along the diagonal hinge lines. The length between the two inside most cracks on either side of the impact point was 1.76 m (5.77 ft). There were also signs of a flexural-type failure mode with the entire length of the barrier cracked near the groundline at the connection between the barrier and coping and also 0.51 m (20 in.) above ground. Cracking in the soil was observed approximately 1.22 m (48 in.) from the front face of the barrier, which corresponds with the location of the end of the moment slab. The cracking, 112 3 3/4" TPY. AA A-A 3 3/4" TPY. 17" 17" 29-1/8" 14-3/8" 14-1/8" 14-3/8" 14-1/8" Figure 5.101. Location of concrete strain gages (Test 4). -0.002 -0.0015 -0.0025 -0.001 -0.0005 0 0.0005 0 0.05 0.1 0.15 0.250.2 0.3 St ra in (in ./in .) Time (sec) Vertical_top Vertical_middle Vertical_bottom Horiz._right Horiz._left Figure 5.102. Strain on the Panel (Test 4).

113 (a) Front view of the barrier Impact Point 5.77 ft 20-in. Figure 5.103. Cracks on the barrier after test (Test 4). (continued on next page) (b) Closeup front view of the barrier

114 (c) Side view of the barrier (d) Top view of the barrier (e) Back view of the barrier Figure 5.103. (Continued).

115 shown in Figure 5.103(c), extended 18.29 m (60 ft) along the entire length of both moment slabs. This indicates that the two No. 9 shear dowels placed between the moment slabs were able to transfer substantial load to the adjacent moment slab. Damage to the back of the vertical barrier is shown in Figure 5.103(e). Damage to the panel beneath the point of impact on the barrier is shown in Figure 5.104. Note that the panel is cracked along its length at an elevation corresponding to the upper layer of reinforcement. It appears the additional resistance provided by the 4.88 m (16 ft) strips enabled more load to be transferred to the wall panel. 5.3.6 Damage of Moment Slab after Test After the bogie impact test, the overburden soil was removed to permit inspection of the moment slab and the connection between the coping and moment slab. After impact, cracks were observed in the top of the moment slab close to its connec- tion to the coping as shown in Figure 5.105. Transverse cracks were found in the moment slab at locations corresponding to the joints as shown in Figure 5.106. 5.4 Summary of Bogie Tests Four reference tests were conducted as summarized in Table 5.14. The impact speeds of bogie vehicle varied from 32.5 km/h (20.2 mph) to 35.08 km/h (21.8 mph). The barrier types used were a 0.81 m (32 in.) tall N.J. shape barrier (Test 1) and a 0.69 m (27 in.) tall vertical wall barrier (Test 2 through Test 4). Wall reinforcement types included 4.88 m (16 ft) steel strips at a density of four per panel (Test 1 and 4), a 2.43 m (8 ft) bar mat (Test 2), and 2.43 m (8 ft) steel strips at a density of six per panel (Test 3). The maximum 50 msec average impact load on the barriers varied from 286.55 kN (64.42 kips) to 326.5 kN (73.4 kips) and Figure 5.104. Cracks on the panel after test (Test 4). (a) Bogie Test 1 (b) Bogie Test 2 (c) Bogie Test 3 (d) Bogie Test 4 Figure 5.105. Cracks in the moment slab below barrier.

maximum dynamic horizontal displacement of the panel at the bottom layer of reinforcement varied from 0 mm to 5 mm (0.19 in.). The permanent movements of the target locations were obtained in two ways: high-speed film analysis and distances from the reference string line stretched in front of the wall. The string line permanent measurements consisted of measuring the distance with a tape measure from the target to the string before and after each test. The permanent horizontal displace- ment at the top of the barrier varied from 6.3 mm (0.25 in.) to 99 mm (3.9 in.). The permanent horizontal displacement at the bottom of the barrier varied from 8 mm (0.31 in.) to 20 mm (0.79 in.). The permanent horizontal displacement of the panel at the level of the top row of reinforcement varied from 1 mm (0.04 in.) to 16 mm (0.63 in.). The permanent horizontal displacement of the panel at the level of the bottom row of reinforcement varied from 0 mm to 4.1 mm (0.16 in.). Even though the wall systems were subjected to loads higher than design conditions, all movements were consid- ered acceptable from a performance point of view. The wall system composed of the 2.44 m (8 ft) strip reinforcement (Test 3) had the highest panel movements, while the lowest movements were recorded for the configuration that incor- porated 4.88 m (16 ft) strips and the vertical wall barrier (Test 4). In Test 4, the top panel exhibited a horizontal hair- line fracture crack along a line corresponding to the location of the top layer of reinforcement. It is possible that this dam- age occurred as a result of accumulated and repeated move- ment of the coping, thereby decreasing the clearance between the coping and the panel. As a result, the coping engaged the panel earlier in the dynamic event and may have lead to higher load. 5.5 Comparison of Test and Numerical Simulation A comparison between the results of Test 1 [N.J. barrier with 4.88 m (16 ft) long strip] and the numerical simulations was performed to determine if further calibration of the numerical model was needed. The calibrated model was used in the sub- sequent study of the 3.05 m (10 ft) high MSE wall and barrier described in Chapter 6. To enable comparison of forces and dis- placements, selected strip locations have been assigned an alphanumeric designator that describes their horizontal posi- tion relative to the bogie impact point and the corresponding vertical reinforcement layer (see Figure 5.107). As shown in Figures 5.108 and 5.109, the damage profile that develops in the simulated barrier is similar to that observed in the test in that it occurs above the toe of the barrier and has a parabolic shape. However, due to the short [3.05 m (10 ft)] length of the precast barrier section that was modeled, much 116 (a) Test 1 and Test 4 (b) Test 2 and Test 3 Test 4 Test 1 Test 3 Test 2 (c) Test 3 and Test 4 Test 3 Test 4 Figure 5.106. Cracks in the moment slab near joint between barriers. are all higher than the 240 kN (54 kips) dynamic force associ- ated with the design of barriers for AASHTO TL-3 and TL-4. Table 5.14 also presents the dynamic and permanent deflec- tion at the top and bottom of the barrier and at the upper and lower layer of reinforcement. The maximum dynamic horizon- tal displacement at the top of the barrier varied from 131 mm (5.17 in.) to 156 mm (6.14 in.). The maximum dynamic hor- izontal displacement at the bottom of the barrier varied from 18 mm (0.69 in.) to 30 mm (1.16 in.). The maximum dynamic horizontal displacement of the panel at the top layer of rein- forcement varied from 8 mm (0.3 in.) to 23 mm (0.92 in.). The

117 Test 1 Test 2 Test 3 Test 4* Test Installation Barrier Type New Jersey Vertical Wall Vertical Wall Vertical Wall Reinforcement 16 ft long strip (4 per panel) 8 ft long bar mat 8 ft long strip (6 per panel) 16 ft long strip (4 per panel) Speed of Bogie 21.8 mph 20.3 mph 20.19 mph 20.19 mph Test Results Peak Acceleration Bogie 14.45 g 13.00 g 13.82 g 12.69 g Barrier 7.36 g 10.71 g 10.16 g 13.04 g Moment Slab 1.84 g N/A 1.00 g N/A Impact Force 73.4 kips 66.1 kips 70.17 kips 64.42 kips Displacement Top of Barrier Dynamic 6.14 in. 6.04 in. 5.17 in. 6.02 in. Permanent 3.00 in. 4.00 in. 2.50 in. 3.00 in. Bottom of Coping Dynamic 1.12 in. 0.93 in. 1.16 in. 0.69 in. Permanent 0.55 in. 0.50 in. 0.60 in. 0.22 in. Panel (Upper Layer) Dynamic 0.63 in. 0.37 in. 0.92 in. 0.30 in. Permanent 0.24 in. 0.20 in. 0.55 in. 0.07 in. Panel (Second Layer) Dynamic 0.00 in. 0.10 in. 0.19 in. 0.07 in. Permanent 0.00 in. 0.02 in. 0.18 in. 0.00 in. Loads in Strips Upper Layer Max. 50 msec** 7.23 kips 1.54 kips 2.13 kips 7.46 kips Design Load 5.29 kips (2.17 kips/ft) 1.68 kips (2.14 kips/ft) 1.64 kips (1.01 kips/ft) 6.25 kips (2.57 kips/ft) Second Layer Max. 50 msec** –1.20 kips 0.08 kips 1.19 kips 0.15 kips Design Load –0.88 kips (–0.36 kips/ft) 0.08 kips (0.12 kips/ft) 0.92 kips (0.57 kips/ft) 0.13 kips (0.05 kips/ft) Total Design Load*** 1.81 kips/ft 2.26 kips/ft 1.58 kips/ft 2.62 kips/ft * Test Section 4 was between Test Sections 1 and 3 and Test 4 was carried out after Tests 1 and 3. Residual deformations from Tests 1 and 3 may have influenced the results of Test 4. ** Average of top and bottom loads. *** Average of loads in the strips at upper and second layers. Table 5.14. Bogie test results. A1C1 A2C2 B1 B2 D1F1 D2F2 E1 E2 Figure 5.107. Strip location indicator.

118 (a) (b) Figure 5.108. Concrete damage profile on front side of (a) Test 1 and (b) simulation. of the damage eventually radiates out to the free ends of the section. The maximum 50 msec average impact loads on the bar- rier were 326.5 kN (73.4 kips) from Test 1 and 365.1 kN (81.85 kips) from the simulation as shown in Figure 5.110. The comparison of the horizontal displacement of the barrier and the wall panel is shown in Figure 5.111. The strip loads in the simulation include the static load due to the earth pressure and the dynamic load due to the barrier impact. To compare the simulation results to the test, the static load in the strips was calculated based on AASHTO LRFD and subtracted from the simulation result. The static loads in the upper and lower layers of reinforcement were computed to be 3.69 kN (0.83 kips) and 7.07 kN (1.59 kips), respectively (Table 5.15). Figure 5.112 shows the comparison of the raw data of load on the strip. In the simulation, the maximum dynamic load in the strip was calculated to be 33.23 kN (7.47 kips) [total load 36.92 kN (8.3 kips)—strip load 3.69 kN (0.83 kips)]. The maximum dynamic load measured in the strip in Test 1 was 34.7 kN (7.8 kips) at 0.0675 sec. The load was shown to drop down at this time in both cases and then rebound. The 50 msec average of the forces in the strip with maximum load

119 (a) (b) Figure 5.109. Concrete damage profile on side view of (a) Test 1 and (b) simulation.

120 0 10 20 30 40 50 60 70 80 90 0 0.1 0.150.05 0.2 0.25 Time (sec) Im pa ct F or ce (k ips ) Test1 Simulation (a) 0 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 H or iz . D ip la ce m en t o f B a rr ier (in . ) Time (sec) (b) Barrier Top Barrier Bottom Top Layer of Strip Bottom Layer of Strip Figure 5.110. Impact load. Figure 5.111. Displacement of barrier and panel of (a) Test 1 and (b) simulation.

121 Layer Static Load by AASHTO LRFD (kips) Dynamic Load Measured (kips) Total Load Measured (kips) Total Load by Simulation (kips) Top 0.83 7.19 (raw) 5.29 (50 msec avg.) 8.02 (raw) 6.12 (50 msec avg.) 8.30 (raw) 5.22 (50 msec avg.) Second 1.59 –1.20 (raw) –0.88 (50 msec avg.) – 8.30 (raw) 3.83 (50 msec avg.) Table 5.15. Total loads on the wall reinforcement. is shown in Figure 5.113. The maximum loads were shown to be 27.58 kN (6.2 kips) at 0.05 sec in Bogie Test 1 and 25.22 kN (5.67 kips) [total load 28.91 kN (6.5 kips)—strip load 3.69 kN (0.83 kips)] at 0.045 sec in the simulation. The strain on the wall panel was evaluated as shown in Fig- ure 5.114. The maximum compressive strain in the simulation wall panel was 0.0021 at 0.045 sec. The simulation reasonably captured the rate of strain increase and maximum strain in the panel, but did not capture a delay in the response of the panel that occurred during the first 0.05 sec of the tests with 4.88 m (16 ft) long reinforcement strips (Tests 1 and 4). These bogie impact simulations and tests were used to sup- port the development of design guidelines and predict the performance of the barrier–moment slab system and MSE wall in the full-scale crash test. These efforts are described in the following chapters of the report. Time (sec) Test1 Simulation -4 -2 0 2 4 6 8 0 0.05 0.1 0.15 0.250.2 0.3 Fo rc e (k ips ) -4 -2 0 2 4 6 8 0 0.05 0.1 0.15 0.250.2 0.3 Time (sec) Fo rc e (k ips ) Test 1 Simulation Figure 5.112. Comparison of raw data of load on the strip. Figure 5.113. Comparison of 50 msec average data of load on the strip.

122 -0.0025 -0.002 -0.0015 -0.001 -0.0005 0 0.0005 0 0.05 0.1 0.15 0.2 0.25 Time (sec) St ra in (in ./in .) Test1 Test4 Simulation Figure 5.114. Panel strain at D1.