Analysis Of Blast Loads On Bridge Substructures.pdf

Analysis of blast loads on bridge substructures K. Marchand1, E. B. Williamson2 & D. G. Winget3 1

Walter P. Moore and Associates Department of Civil Engineering, University of Texas 3 Department of Civil & Mechanical Engineering, U.S. Military Academy 2

Abstract The design of structures to resist blast loads has traditionally been considered only for essential government buildings, military structures, and petrochemical facilities. Until recently, however, little attention has been given to bridges. One strategically placed truck bomb on a critical bridge could result in significant loss of life, severe structural damage, and devastate an economy. Recent terrorist threats to bridges in California and New York have demonstrated the vulnerability of our transportation infrastructure and reinforced the need for bridge security. This paper summarizes the results of ongoing research to develop performance-based blast design standards tailored specifically for bridges. The goal of the research is to investigate economical, unobtrusive and effective methods to mitigate the risk of terrorist attacks against critical bridges. The potential effects of blast loads on bridge substructures are presented, and structural design and retrofit solutions to counter these effects are discussed. Case studies demonstrate the use of simple models to analyze concrete piers. The modeling concept, determination of peak overpressures, and inherent assumptions are described, and empirical deformation-based damage criteria that are used to estimate the level of damage are presented. Keywords: bridge, pier, terrorist, explosives, blast, breaching, retrofit.

1

Introduction

Most of the current state of knowledge for blast-resistant design of structures was developed with a focus on buildings. Very little information exists specifically for bridges, and bridge designers have not traditionally needed to consider security in their design process. Recent increases in the frequency and Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

152 Structures Under Shock and Impact VIII intensity of terrorist attacks, however, have highlighted the need for bridge security. In a 1997 report, Brian Jenkins describes over 550 terrorist attacks worldwide against transportation targets and states that “we have seen an increase in attacks on public transportation as terrorism has increased over the past quarter century and more recently as terrorists have demonstrated greater willingness to kill indiscriminately [6].” Because bridges tend to be a chokepoint in transportation networks, the destruction of some critical bridges could have devastating effects on both the transportation system and economy. Therefore, bridge designers in the United States are now investigating both physical security and structural hardening techniques that can be implemented to reduce the risk of terrorist attacks against critical bridges.

2

Performance-based blast design standards for bridges

Federal facilities in the United States are divided into protection categories based upon their importance and vulnerability, and each category has established minimum security standards [7]. Borrowing from this approach, bridges could also be divided into protection categories, based on their level of importance and a vulnerability assessment. Each bridge category could have a standard level of protection, with additional specific measures tailored to the different types of bridges in each category. These countermeasures could include physical security measures, such as closed-circuit television monitoring or guards, in addition to minimum design standards for blast loads. The authors have previously proposed a set of performance-based design criteria for terrorist threats against bridges [10]. These performance-based standards establish a baseline threat level for design loads and define the acceptable level of damage under these loads. The design loads and acceptable damage for each category are based on a balanced assessment of the threats, acceptable risks, and available resources. In order to implement these design standards, each critical bridge will need to be analyzed individually to determine its major vulnerabilities for all specified baseline threats. The individual bridge components will then either need to be redesigned or retrofitted to meet the specified performance standards. However, though individual analyses may be the best option for the most critical bridges, it can be very costly to implement for the lesser important bridges. Therefore, to meet the performance standards for these less important bridges, design charts tailored specifically for each bridge component could be developed. These design charts should be applicable to a wide variety of bridges, help to minimize costs and expedite the design or retrofit process, and still provide acceptable levels of protection against the baseline threats. Current research is being conducted by the authors to evaluate the relative effectiveness of potential structural retrofits, refine the performance-based standards, and develop general blast-resistant design guidelines specifically for bridges. This research is based on parameter studies using simplified single degree-of-freedom (SDOF) computer models. The simplified models allow for the analysis of a large number of load cases, bridge types, and structural configurations. Because they account for the dynamic properties of the structural

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

Structures Under Shock and Impact VIII

153

system and the blast loads, the models provide reasonable estimates of the relative effectiveness of different retrofit options. Furthermore, the approach taken is the one typically used by engineers that design structures for blast. The relative improvement in performance of alternative design and retrofit strategies is also being studied. Because the models require assuming a failure mode for each structural model, complicated modes of structural behavior or combinations of complex loading scenarios are not being analyzed. These scenarios will require a more detailed finite element analysis which couples the effects of the blast or impact loads with the structural response. Based on the current research, the most promising design approaches will be identified and can be verified with physical testing.

3

Blast load determination

When an explosion occurs below the deck of a bridge, bents and piers will be subjected to large lateral forces, possibly resulting in large deformations, shear or flexural failures. Additionally, concrete cratering and spalling from the blast wave impact may lead to significant losses of concrete, especially if the standoff distance is small. For above-deck explosions, the piers will experience increased loads in addition to gravity loads. However, these members are generally large enough to withstand these above-deck loads [8]. Therefore, the focus of the substructure analysis for the current research is on below-deck explosions. The baseline bridge chosen for analysis consists of two 26-meter (85-foot) main spans, with a 4.8-meter (16-foot) clearance and total deck width of 14 meters (46 feet). It contains three concrete piers per bent spaced at 5.2 meters (17 feet), each with a 91-centimeter (3-foot) diameter. Variations of this baseline were considered by varying the pier diameter, pier shape, and concrete strength. Two terrorist threat scenarios were chosen for analysis of the substructure, based on a preliminary vulnerability assessment [10]. These general courses of action consisted of a vehicle bomb below the deck and hand-placed charges in contact with the pier. For the vehicle bomb scenario, variations in charge weight and standoff were considered. It was expected that the vehicle bomb possesses the potential to produce large lateral blast pressures, resulting in localized breaching damage of the concrete and causing a flexural failure of the pier. However, if barriers are used to limit the vehicle standoff distance to the pier, small charges may still be placed directly on the pier. For the hand-placed charge scenario, two variations were considered: a charge placed on one side of the pier only and counterforce charges placed on each side of the pier. Though these contact charges will probably cause less of a flexural response than the vehicle bomb, they possess the potential to breach enough of the pier to render it incapable of supporting the dead loads. BlastX version 4.2.3.0 [1] was used to generate loads for all cases considered. For external explosions, BlastX computes the pressure on exterior walls, or on the ground surface, as well as the shock waves and mass flow propagated throughout a structure, using fast running analytical/empirical models. Shock waves dominate the blast environment of external explosions,

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

154 Structures Under Shock and Impact VIII while the flow of confined detonation product gases will likely also be important for explosions inside a confined area. To calculate the loads on the piers, a “room” was constructed in BlastX consisting only of the deck, ground, and a small “wall” for the pier. Blast waves were allowed to vent on the sides, and the multiple reflections between the ground and deck surfaces were included in the load effects. The effect of reflections near the abutment could also be considered. Traffic loads and girder reactions were neglected. This assumption is conservative for most piers due to increased flexural resistance under low levels of axial compressive forces. Only in cases where the axial load is high (not common for bridge piers) will this assumption not be conservative. In addition, it was assumed the pier cannot go into tension from uplift due to simple seating conditions of the girders. To calculate the reduced area of the piers due to breaching from local blast damage, ConWep V. 2.0.6.0 [3] was used (distribution limited to U.S. Government agencies and their contractors). ConWep’s breaching calculations are empirical and are developed from explosive tests conducted by the Army and military contractors. It performs airblast calculations, computes the breaching and fragment penetration, and determines ground cratering depth based on charge type and configuration, equivalent TNT weight, standoff distance, type of burst, target material type and strength, and target dimensions. Using BlastX, blast wave pressures were calculated at discrete heights along the pier for a 1.2-meter (4-foot) height of burst. As seen in Figure 1:, the pressure distribution can vary significantly along the height of the pier at any given time. This phenomenon is due to the reflected pressures off the ground and the reflected pressure buildup between the girders under the deck. In general, most of the impulse will have a parabolic distribution with the peak centered a few feet off the ground. However, reflections from the bottom of the deck acting against the pier will occur later in the load cycle, thereby causing higher pressures to occur towards the top of the pier (Figure 1:). This changing pressure distribution affects the load factor used in the structural response calculations, which is a multiplication factor used to transform the total load on the structural member to an equivalent uniform load for an SDOF, dynamic analysis. The load factor is calculated by equating the external work done by the applied load under an assumed displaced shape to the work done by a equivalent uniform load producing the same deflection, and can be determined from charts in numerous sources [4]. However, in this case, the load factor varies with time. Because the goal of the research was to perform numerous analyses using simplified models, an equivalent “average” pressure was calculated for the SDOF model instead of using load factors that change with time. Additionally, due to the uncertainty in charge weight and location, performing an analysis with a detailed level of load resolution is best suited for specific critical bridges using a finite element model. The “average” pressure was calculated by ensuring that the total average impulse over the entire height remained the same. As seen in Figure 2:, there are significant differences in a load prediction that accounts for reflections off the ground and pressure buildup in confined regions, and one that does not. Both pressure histories were computed for the same charge weight and

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

Structures Under Shock and Impact VIII

155

standoff. However, the hemispherical surface burst calculations by ConWep do not account for the wave reflections off the deck surface and corresponding pressure buildup, nor do they account for the burst height.

Figure 1: Pressure distribution along pier height at various times.

Figure 2: Comparison of BlastX and Conwep results. Blast loads calculated in BlastX also had to be adjusted to account for the pressure equalization of the blast wave around the back side of the pier, the reduced reflected pressures due to the curved surface of a round pier (Figure 3:), and the clearing time for the load acting on the impact face of the pier. The load curves were first calculated on the front and back side of the pier using BlastX (Figure 4:). To account for the pressure equalization, the back

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

156 Structures Under Shock and Impact VIII pressure load history was subtracted from the incident wave pressure-time history on the front side because they occur in opposite directions on the pier. Using Figure 5-3 from the DAHSCWE manual [4], the reduction of reflected pressure due to a curved surface could be calculated based on the changing angle of incidence (Figure 3:). It should be noted that the reduction factor is a function of the magnitude of the reflected pressures. To account for the curved column surface, the resulting pressures (Y-axis) from the net pressure-time history were graphically multiplied by a factor of 0.8 (Figure 5:).

Figure 3: Pressure equalization and reflected pressures on curved surface.

Figure 4: Accounting for pressure equalization. Finally, the clearing time was calculated by using the “3 transits to the nearest free edge” rule from the DAHSCWE manual [4]. According to the manual, the clearing time is equal to three times the distance from the center of the pier to the pier’s edge in the projected side view (3 times ¼ of the circumference) divided by the shock front velocity. The clearing time for the example shown in Figure 5: was calculated to be 0.72 msec. When added to the incident wave arrival time of 0.71 msec (Figure 5:), we obtain the final clearing time of 1.43 msec for the incident wave. Since the original pulse length is 2.44 – 0.71 = 1.73 msec (Figure 5:), the duration of the pressure history needs to be scaled to the incident pressure at this clearing time by multiplying the x-axis

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

Structures Under Shock and Impact VIII

157

by a factor of 0.72 / 1.73 = 0.42, which is the ratio of the adjusted to original pulse length. The curve was then shifted to the right to correspond to the original arrival time. The resulting pressure time history shown in Figure 6:, which clears the incident wave at 1.43 msec, can be used to determine the structural response.

Figure 5: Accounting for curved surface by reducing y axis.

Figure 6: Final pressure-time history. Another option for adjusting the pulse duration to account for the clearing time is to simply truncate the original pulse at the new clearing time. Truncating is more conservative, as it will preserve more impulse. In the specific case shown here, it will result in almost twice as much impulse.

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

158 Structures Under Shock and Impact VIII

4

Pier response

To calculate the flexural response of the piers to vehicle blast loads, SPAn32 version 1.2.6.9 [9] was used. Developed by the U.S. Army Corps of Engineers, this program performs an equivalent SDOF dynamic analysis, taking bi-linear material nonlinearity into account for the reinforcement. The blast load pressure history is specified as an equivalent, uniformly distributed load, and can be entered directly from a BlastX output file. SPAn32 calculates the equivalent SDOF stiffness and mass parameters based on the member geometry and material properties. Ultimate resistance is determined from the full plastic hinge capacity of the member. Though it is based primarily on the first principle solution to the ordinary differential equation of motion for the SDOF system, it makes some adjustments based on empirical data. These adjustments include incorporating dynamic increase factors to modify the material strengths based on the instantaneous calculated strain rate. Also, ultimate moment capacity of the concrete is based on the equivalent rectangular stress distribution from ACI 318, but is reduced once the member exceeds a rotation of 2 degrees (based on empirical data). Although the program will also take into account interaction with the ground shock, this feature was not used. The pier was modeled as an SDOF flexural member, fixed at the foundation and the pier cap. Although some rotation might occur at the foundation, it was assumed to be negligible due to large footings and the assumption of negligible ground cratering. Also, some sidesway may occur at the top of the pier. However, the support conditions were assumed to be such that sidesway of the bents would be counteracted by dead load and the axial stiffness of the girders. In addition, impact axial loads on the pier due to rigid body vertical translation and subsequent dropping of the deck were neglected. When determining the flexural response of the piers due to the blast pressure and reduced cross-sectional area from local damage, it was conservatively assumed that the cross-sectional area along the entire height of the pier was reduced to its minimum predicted diameter at the location of maximum breaching. Though this assumption will likely produce very conservative results, it allows us to avoid the time consuming task of calculating equivalent mass factors (transforming distributed masses to an equivalent lumped mass) and equivalent stiffnesses due to a reduced cross-section. As previously discussed, concrete breaching was estimated using ConWep. The cross-sectional area was reduced accordingly, and the resulting cross-section was analyzed in SPAn32 to determine the flexural response and calculate the maximum support rotations. Damage levels could then be predicted using deformation-based empirical data. The maximum support rotation for slight to moderate damage corresponds to 1.3 degrees. For moderate to heavy damage, 2 degrees was used. The pier was assumed to lose structural integrity at 3 degrees of rotation. These values were empirically determined from experiments [2] on concrete beam elements in flexure (including tension membrane effects), and were based primarily on member sizes that are commonly found in buildings.

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

Structures Under Shock and Impact VIII

159

Therefore, they may require adjustment in the future based on experimental data specifically for bridge piers. Using iterative analysis techniques, the required combination of vehicle standoff (achieved through barriers) and pier diameter can be calculated for a given scenario. Also, the protective effects of retrofit options can be predicted, such as FRP wraps and steel jacketing. For FRP, the ductility limits can be adjusted to correlate with experimental results for wrapped piers or columns. For steel jacketing, the flexural stiffness of the pier can be adjusted and the concrete breaching could probably be neglected assuming a sufficient thickness of steel is used. Lessons learned from the analysis include the fact that when neglecting breaching, concrete strength has little effect on the flexural response of the pier to below-deck blast loads. Reducing the strength from 5000 psi to 3000 psi generally resulted in less than a 10% increase in the maximum support rotation. However, this strength reduction does result in approximately a 30% increase in the amount of concrete breaching for the charge weights and standoff distances being considered in the current research. In scenarios where a complete breach is not achieved, the additional loss of concrete due to lower compressive strengths significantly reduced the flexural stiffness of the damaged pier. Additionally, the shape of the pier can have somewhat significant effects on the peak overpressures. A 91-cm (36-inch) square pier will have 25% higher peak overpressures than a round pier with the same diameter.

5

Conclusions

Initial results indicate that breaching failure of the concrete will usually govern the design, especially in cases of large truck bombs with limited standoff or for counterforce charges. Therefore, standoff protective measures such as the use of 360 degree barriers should be considered first whenever it does not interfere with the function of the bridge under consideration. Additionally, minimum pier diameters, or the use of steel jacketing (connected with high-strength bolts and epoxy or a layer of grout), should be specified to ensure that the design handplaced charge weight does not achieve a breach failure. Other strengthening options for piers include using fiber reinforced polymer wraps, incorporating tightly spaced spiral reinforcement, or embedding a steel column within a concrete pier. Generally, retrofit techniques should enhance concrete confinement, increase bending resistance and ductility, add protection against breaching, or a combination of these effects. As bridge engineers and planners learn to cope with the design and retrofit of bridges for security, it is necessary to introduce new design concepts. The use of performance-based design standards can help ensure each bridge has an appropriate level of protection, based on its importance and specific vulnerabilities. Additionally, design charts developed under the current research may assist bridge designers in meeting these standards without performing an individual detailed analysis for each bridge.

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X

160 Structures Under Shock and Impact VIII

References [1] [2] [3] [4] [5] [6]

[7] [8]

[9] [10]

BlastX Version 4.2.3.0, Science Applications International Corporation: San Diego, CA, 2001. Conrath, E.J., Krauthammer, T., Marchand, K.A., Mlakar, P.F., Structural Design for Physical Security: State of the Practice, ASCE: Reston, VA, 1999. ConWep Version 2.0.6.0, U.S. Army Engineer Research and Development Center: Vicksburg, MS., 2003 Defense Threat Reduction Agency (DTRA). (1997). Design & Analysis of Hardened Structures to Conventional Weapons Effects, Washington, D.C. (distribution restricted to US Government agencies and their contractors). Department of the Army, Structures to Resist the Effects of Accidental Explosions. Army TM 5-1300. U.S. Govt. Printing Office: Washington, D.C. (approved for public release), 1990. Jenkins, B. M., Protecting Public Surface Transportation and Patrons from Terrorist Activities: Case Studies of Best Security Practices and a Chronology of Attacks, MTI Report 97-4, Mineta Transportation Institute: San Jose, CA., 1997. Lewis, D., “Department of Defense Antiterrorism Construction Standards (Draft),” SAME/ASCE National Symposium on Comprehensive Force Protection, November 1 – 2, 2001, Charleston, SC., 1991. Ray, J., Personal correspondence with the author , Bridge Research Team Leader at the Army Corps of Engineers’ Engineering Research and Development Center, Waterways Experiment Station, Vicksburg, MS, 2002. SPAn32 Version 1.2.6.9, U.S. Army Corps of Engineers Omaha District: Omaha, NE., 2002. Williamson, E.B., and Winget, D.G., “Risk Management and Design of Critical Bridges for terrorist Attacks”, ASCE Journal of Bridge Engineering, ASCE: Reston, VA, To be published.

Structures Under Shock and Impact VIII, N. Jones & C. A. Brebbia (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-706-X