3

Review of Program Activities

Before commenting on specific technical areas, the committee wishes to point out one important programmatic issue DTRA should address as the program matures. At present, industrial contractors and government laboratories, drawn almost exclusively from DTRA's protective-construction and blast-hardening design community, are implementing the BMSP. Therefore, although this group has first-rate technical capabilities, the committee recommends that the contractor base be expanded to include the commercial architectural and engineering communities, as well as specialists in injury prevention, disaster medicine, and technology transfer, especially for the planning phases of the program. The committee's concern is that, as a group, the existing corps of contractors may be concerned mostly with traditional military and defense objectives, which may not reflect the emphasis of the BMSP on nonstructural solutions, injury reduction methods, and improved rescue and recovery techniques. In addition, although some of the BMSP's contractors are active in key organizations involved in the development of building codes and commercial engineering design, many of them are not. As a result, the BMSP may not be benefiting from complementary developments in the commercial sector. The committee considers this a significant, but easily remedied, problem and strongly recommends that DTRA broaden its contractor base.

STRUCTURAL ISSUES

Progressive Collapse

Progressive structural collapse is a principal, if not the leading, cause of injury and death in building failures, regardless of the source of the loading (e.g., bomb, earthquake, internal explosion). For this reason, predicting and designing to prevent the progressive collapse of a building under a specified attack scenario is (and should be) a primary objective of the BMSP. After considering whether the study of progressive collapse should be addressed through physical testing or computation and analysis, the committee concluded that both are necessary.

This dual experimental approach could investigate the behavior of complete structures or structural components. However, because buildings are complex systems that can have large variances between design specifications and as-built conditions, a test structure may not accurately mimic the progressive failure of a real building. Although an experienced engineer can often estimate the likelihood that a specific building will collapse by superimposing a damage scenario on the design, because of the variances described above, the actual progress of a collapse is essentially a stochastic process. For example, following major earthquakes, nonstructural building components, such as mechanical piping, partition walls, equipment, heavy-duty storage facilities (shelving and file cabinets), and curtain walls that can transfer some of the dead load to lower levels, have been observed to keep buildings standing that would otherwise have been expected to fail (Loizeaux, 1999). This phenomenom, (i.e., a complete



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program 3 Review of Program Activities Before commenting on specific technical areas, the committee wishes to point out one important programmatic issue DTRA should address as the program matures. At present, industrial contractors and government laboratories, drawn almost exclusively from DTRA's protective-construction and blast-hardening design community, are implementing the BMSP. Therefore, although this group has first-rate technical capabilities, the committee recommends that the contractor base be expanded to include the commercial architectural and engineering communities, as well as specialists in injury prevention, disaster medicine, and technology transfer, especially for the planning phases of the program. The committee's concern is that, as a group, the existing corps of contractors may be concerned mostly with traditional military and defense objectives, which may not reflect the emphasis of the BMSP on nonstructural solutions, injury reduction methods, and improved rescue and recovery techniques. In addition, although some of the BMSP's contractors are active in key organizations involved in the development of building codes and commercial engineering design, many of them are not. As a result, the BMSP may not be benefiting from complementary developments in the commercial sector. The committee considers this a significant, but easily remedied, problem and strongly recommends that DTRA broaden its contractor base. STRUCTURAL ISSUES Progressive Collapse Progressive structural collapse is a principal, if not the leading, cause of injury and death in building failures, regardless of the source of the loading (e.g., bomb, earthquake, internal explosion). For this reason, predicting and designing to prevent the progressive collapse of a building under a specified attack scenario is (and should be) a primary objective of the BMSP. After considering whether the study of progressive collapse should be addressed through physical testing or computation and analysis, the committee concluded that both are necessary. This dual experimental approach could investigate the behavior of complete structures or structural components. However, because buildings are complex systems that can have large variances between design specifications and as-built conditions, a test structure may not accurately mimic the progressive failure of a real building. Although an experienced engineer can often estimate the likelihood that a specific building will collapse by superimposing a damage scenario on the design, because of the variances described above, the actual progress of a collapse is essentially a stochastic process. For example, following major earthquakes, nonstructural building components, such as mechanical piping, partition walls, equipment, heavy-duty storage facilities (shelving and file cabinets), and curtain walls that can transfer some of the dead load to lower levels, have been observed to keep buildings standing that would otherwise have been expected to fail (Loizeaux, 1999). This phenomenom, (i.e., a complete

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program progressive failure that is just barely contained) has been observed in Mexico, California, Japan, Guam, and more recently in Turkey and Taiwan. Unfortunately, although nonstructural elements obviously have an important role to play in determining individual outcomes, their random contribution to preventing a collapse cannot be easily included in structural models. (Random factors are discussed in more detail later in this chapter.) Because of fiscal realities, only a limited number of full-scale tests can be conducted. For this reason, the committee believes that testing complete, full-scale structures is not a practical way to gather knowledge about progressive collapse. Nevertheless, experimental data to improve the collapse-resistance of buildings is important. Therefore, reinforcement details in columns, beams, and floors and their interconnections can be tested at the component level for various blast loadings, and the information gained can then be used to determine additional measures for minimizing the extent of collapse. The committee believes that physical testing of components at both full scale and reduced scale is a cost-effective means of adding valuable information to the knowledge base for reducing the likelihood of progressive collapse and is a valid area of investigation for the BMSP. The committee recommends that basic research into progressive collapse also be supported at academic institutions—either through the BMSP directly or through BMSP's support of other cooperative arrangements. Computational Modeling Advances in parallel computing have increased both the size and speed of computational tools (on the order of 102 to 103 ), and consequently, computational modeling is changing rapidly. As the speed of central processing units continues to increase, hardware is outpacing software, which in turn is outpacing the availability of experimental data for validating models. Despite this computational capability, attempts to develop an accurate model of progressive collapse have been unsuccessful, however. In the explosives demolition business (which provides many opportunities to compare predictions with observed results under relatively controlled conditions), predemolition predictions for a structural frame using computational models do not generally compare well with the results obtained under actual field conditions (Loizeaux, 1999). In addition to random factors, one of the difficulties in modeling progressive collapse caused by a bomb is order-of-magnitude differences in time scales between an explosive event and the response of the building. For example, the time frame in which damage occurs depends on the size of the explosion and the mechanism of resistance of the structural elements (floor slabs, walls, columns, etc.) and ranges from milliseconds to seconds. If the significant damage mechanism is the long-term dynamic response of the building, rather than just the explosion, millions of time steps will be required to simulate an explosion-induced collapse. By comparison, the time scale for earthquake-induced building response is on the same order as the seismic event that caused it. Consequently, modeling earthquake-induced damage will require far fewer time steps (Attaway, 1999). Analytical models can still be useful for evaluating different designs, however, and, with continued refinements in hardware, software, and test data, they could be used to improve predictions of collapse mechanisms. The committee recommends that DTRA take advantage of the advances in parallel-processor computing made by DoD and the U.S. Department of Energy to improve the capability and ease of use of computational tools for predicting structural

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program responses to bomb blasts. As the complexity of analytical models increases, verification and validation (V&V) of the resulting codes and solution techniques will be necessary. The experimental database being developed by the BMSP should be designed from the outset with the requirements for V&V in mind. Small-Scale versus Full-Scale Testing Reduced-scale testing offers an excellent, lower cost means of learning about failure modes and loading parameters and obtaining other data that can be used to test results of computational models. For these purposes, testing reduced-scale models and components is far more cost-effective than full-scale testing. However, reduced-scale testing also has limitations, such as scale effects (i.e., no model, no matter its size, can reproduce the prototype building exactly). If the critical local response is well understood, reduced-scale tests can be very effective for determining behavioral phenomena of the system. For example, the flexural behavior of a complex slab or wall system subjected to out-of-plane loading can be reliably investigated with small-scale models (but not so small that the behavior of the component material is altered significantly). This cannot be done confidently, however, if the failure is, or would be, in shear. A very small specimen might have a very high shear strength leading to a flexural failure, even though a larger scale test would have been terminated by shear. Scaling down connection details can be difficult, or even impossible. Reduced-scale tests of the response of floor slabs can yield much information, but gravity effects cannot be scaled conveniently. Therefore, it may not be possible to determine in a small-scale test if a floor slab would have collapsed onto the floor below and triggered a progressive collapse. Scale effects for buildings in the five-story range are probably not significant, but for buildings of ten or more stories, column sizes may be large enough to have a significant effect. An excellent discussion of the effects of scale on experimental results can be found in Design of Model Test Program for a Buried Field Shelter (Newmark & Associates, 1965). The committee believes that the BMSP should include a mix of full-scale and reduced-scale component testing. From an analytical and modeling standpoint, reduced-scale tests are valuable because they are simpler in size, cost, and complexity of the test setup. However, one of the challenges in validating a numerical method is identifying which components can be modeled and validated independently; models of phenomena that involve complex effects cannot be validated by a single type of measurement, or even by a single test. For example, when modeling the deformation of a floor slab or column under heavy static load, the analyst can assess the accuracy of the model for both the concrete and the reinforcing steel. However, if something goes wrong with a fully coupled model of dynamic blast loads on a complicated structural geometry, it is difficult to determine which part of the model is incorrect. The interpretation of test results requires critical judgement by the analyst. Questions about system behavior can be addressed with small-scale structures, provided the experimenter understands the behavior of details and is careful when comparing the results of calculations to tests. For example, in some instances, agreement may be more apparent than real because the experimental results in a particular application, or at that stage of loading, may be insensitive to the assumptions made in developing the theoretical model. Nevertheless, the data will be

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program generally useful, and, because the cost of the small-scale experiments is relatively low, tests can be repeated to increase confidence in the results. Problems that could be investigated at reduced scale include the effects of a floor slab failing and impacting lower floors, the effects of interior walls redirecting a blast wave to the floor or ceiling, and the effects of rebar splices on the strength of floor slabs. In a postattack analysis of the Alfred P. Murrah Federal Building in Oklahoma City, the authors speculated that confinement of the concrete would have increased the toughness of the columns (Corley et al., 1998). For this reason, the committee believes a series of tests of quarter-scale models would be useful for determining the effects of transverse reinforcement on column toughness. Confinement levels should be those currently used in spirally reinforced columns and in special moment frames. Damaged Test Articles In one test conducted as part of the BMSP, a column in CTS-1 was severely damaged and, by most definitions, failed. However, for various reasons, the CTS-1 structure did not collapse. Based on this event, and other experiences of committee members, the committee concluded that the residual strength of damaged columns and other components is not well understood but could be of great value for predicting the performance of buildings, their potential for progressive collapse, and for determining the stability of damaged buildings during search and rescue operations. For example, following earthquake-induced shear failures in columns, the columns sometimes remain standing and sometimes collapse, depending on the location of inclined cracks, the bending stiffness of longitudinal bars, and the location/size of the ties (Sozen, 1999). If the bars have sufficient lateral support to prevent buckling and to resist the dead load on the column along with whatever concrete remains, the building will not collapse. Because of limitations on what is known and the vagaries of construction, the results are very difficult to predict. The committee recommends that the BMSP undertake a series of column tests to cause various levels of damage and measure their residual strength. Full-scale columns representative of those in buildings of ten stories or more should also be tested. Because there are so many possible combinations of rebar and column dimensions, the tests should be designed to bound the behavior of a representative design. Although building a database of responses of many column sizes and shapes would be valuable, a database that could be used to validate numerical models would be even more valuable in the long term. Once numerical methods have been shown to be capable of predicting the observed response, then numerical methods could be used to explore the residual strengths of damaged components. Component Testing The continuity of structural members in both steel and reinforced concrete buildings is highly dependent on reliable steel connections. Although a considerable amount of data on the performance of mechanical connectors under earthquake loads are available, little, if any, public information is available on their performance under blast loads. Consequently, a series of tests to evaluate how well common rebar splices and connections can function after being damaged by blasts would be useful. Data could be gathered, for example, on the behavior of large bars in

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program mechanical splices, welded connections, lap splices, or Type 2 mechanical splices. Because steel connections, like welding details, cannot be accurately scaled down for reduced-scale testing, full-scale testing would be necessary for steel construction components. The BMSP program has a unique opportunity to clarify relationships between earthquake-resistant building designs and blast-resistant building designs, especially on a component/subsystem basis. Although there have been discussions in the design community of the benefits of seismic designs to blast resistance, a recent paper comparing the design requirements showed that even a design that meets Seismic Zone 4 requirements may not result in a building resistant to blast effects (Ettouney et al., 1998). The key to improved performance appears to be in the type and location of structural detailing (Woodson and Krauthammer, 1998). Therefore, the committee believes that BMSP should plan to conduct a series of component tests to identify detailing methods for blast-resistant construction. The results could be used to inform the building code process and could ultimately be incorporated into design guides, such as Building Code Requirements for Structural Concrete (ACI, 1999). Comparative costs of alternative designs could also be developed to encourage the adoption of these techniques in commercial buildings. Component testing would also be useful for validating high-fidelity numerical models. Although not all combinations of splices and joint types can be accounted for in a series of tests, examples of different joint types under different loading conditions could be used to develop benchmarks for testing models. Once confidence in sophisticated numerical models has been established, they could be used to calibrate the simple structural models used in the design process. Next Class of Structures The BMSP has focused on the testing and analysis of generic flat-slab structures because of their perceived vulnerability to blast damage and progressive collapse. As part of the committee's evaluation, the value of conducting full-scale testing on another generic structure type—possibly a steel frame building—was considered. For example, if an economical, blast-resistant steel frame design could be identified, it could have widespread commercial appeal because it could be erected and enclosed rapidly. However, although the generic building approach seems to be a reasonable experimental paradigm, the committee concluded that DTRA should not proceed with another controlled test structure for several reasons. First, many of the typical structure types used in military construction domestically and overseas (e.g., multistory, load-bearing masonry; high-bay, long-span structures with load-bearing walls, such as gymnasiums and dining halls; preengineered metal buildings; and multistory wood structures), although simple structurally, are vulnerable to even fairly small bombs and usually have high occupancies. The observed performance of unreinforced masonry and other lightly constructed buildings, such as wood and preengineered metal, in earthquakes and high wind conditions has shown that they are very vulnerable to extreme loading. Thus, the potential for injury and death to a large number of occupants is probably higher in these buildings than in larger, more complex structures. On this basis alone, the BMSP would be justified in investigating these structures, and the committee recommends that DTRA consider including a series of component and reduced-scale tests on masonry structures in the BMSP. The test program should include unreinforced masonry for benchmarking purposes and a range of

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program reinforcement techniques, including fully reinforced masonry, to improve protection. Following analysis and interpretation of the test results for both unreinforced and strengthened masonry, DTRA could conduct another series of tests on construction typical of long-span buildings. The committee also questions the validity of the results obtained from repetitive blast loading of CTS-1. After the initial blast test, the structure became a de facto damaged structure. At that point, although the structure still has considerable value as a reaction frame for testing full-scale structural components and nonstructural elements, performance data for the overall building system obtained in subsequent tests will necessarily be compromised to some degree by preexisting damage. (For this reason, the committee strongly recommends that CTS-1 not be tested to failure.) The relative value of testing full-scale structures versus analysis, computational modeling, and component testing should also be reevaluated. From the standpoint of cost-effectiveness, the construction of another full-scale test structure is not justified at this time. Although a full-scale test structure could be valuable in a “proof-of-principle ” demonstration, the committee believes that the next full-scale test structure should not be designed or constructed until a thorough study has been done to identify the knowledge gap(s) that should be addressed. Internal versus External Blasts The security community has expressed concerns about the vulnerability of interior columns and adjacent floor slabs to small satchel and suitcase bombs (~50 pounds) with little or no standoff (Dadazzio, 1999). Information on the damage potential of these small devices can be obtained from component and reduced-scale tests. Full-scale testing of a complete structure provides only a few data points, but with reduced-scale testing, repeated tests can be and should be run to reduce the considerable statistical variance in these events. In conjunction with good analytical models, reduced-scale testing (for reinforced-concrete columns) or full-scale component testing (for steel columns) can provide a basis for estimating the potential for structural collapse. By studying how joint construction affects the strength of components under explosive loads, it may be possible to create design guidelines that would lead to better methods for joining floor slabs and columns This information could then be used to estimate the potential for collapse. In recent years, the American Concrete Institute committee (ACI-318) that deals with code requirements for structural concrete has added several requirements for detailing reinforcement to maintain building integrity. The committee believes the BMSP would benefit from obtaining data evaluating the reinforcement patterns of existing buildings and the reinforcement patterns now required by ACI-318, on the assumption that current ACI requirements have increased the resistance to progressive collapse following removal of a single column. If the BMSP tests prove otherwise, then other details to increase resistance could be evaluated. For example, some data are available on the resistance of slender elements to unusual loads over long spans, but investigations have not been made of catenary action with representative details. As demonstrated by the World Trade Center bombing, a van filled with explosives in an enclosed parking area can do a great deal of damage. In these cases, the likely mechanism of collapse is very different than in external building explosions. Instead of loading an exterior column(s), an interior blast directly loads floor slabs, both above and below. Loss of the floor slabs destroys the lateral support of the remaining columns, further weakening the structure and

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program rendering it prone to progressive collapse. Many parking garages are constructed of post-tensioned concrete elements (e.g., long-span, quad tees) that carry most of the load in continuous steel cables that may be very sensitive to progressive collapse if they are damaged by an explosion. Because of widespread concerns about vehicle bombs, the committee believes that DTRA should consider investigating parking garages as part of the BMSP. Existing Buildings as Test Articles The committee debated the value of using an existing building scheduled to be demolished as a test specimen. On the one hand, existing buildings may offer a technical advantage over purpose-built test structures because they are constructed by typical contractors following the plans and specifications of typical engineers using standards generally imposed at the time of the design. On the other hand, because of the inherent variability in construction details, they will be difficult to characterize from an analytical standpoint. Several problems would have to be addressed before buildings scheduled for demolition could be considered as viable experimental options. The most significant issue is that the building must be demolished within a few seconds after the test because of the very high probability that the structure will become unstable and pose a danger to the demolition crew. In addition, collecting data would require high-speed video, contact gauges, and displacement measurements—all before collapse occurred (a duration of seconds). Nevertheless, a purpose-built test structure at ERDC or White Sands, although easier to characterize, may be perceived as a specimen created especially for the test and, therefore, not representative, regardless of whether it was constructed in accordance with current procedures and standards. Retrofits Information about the design of new buildings can only be put into place as they are constructed. Because the inventory of existing buildings (many of which have vulnerabilities to blast loadings) far exceeds the number of new buildings, the committee believes that the BMSP should put a high priority on developing and evaluating retrofitting techniques—particularly creative conceptual retrofitting measures that would prevent a life-threatening progressive collapse following a blast that has destroyed the load-carrying capacity of a limited number of structural components. This recommendation is based partly on the relatively long service lives of buildings. For example, when the California Hospital Seismic Safety Act became law in 1972, hospitals already constructed were expected to be replaced by new facilities through attrition. However, as of 1990, more than 66 percent of California's hospitals were still not in compliance with the 1972 standard and were not be expected to be functional after a major earthquake (Office of Statewide Health Planning and Development, 1990). Many retrofitting methods, including wrapping columns with carbon and glass fibers or jacketing them with steel plate and grout, are currently under consideration for inclusion in the BMSP. Methods for using carbon fiber as floor reinforcement against uplift have also been proposed. However, the absence of test data to support design calculations has seriously hampered the introduction of new, potentially life-saving technologies. Testing retrofitting

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program methods will require that the behavior of the building without the retrofitting be determined first. The baseline behavior can then be used to assess the cost benefit of a given retrofitting design. Testing and analysis of the reduced-scale structures at ERDC and the full-scale CTS-1 have demonstrated that some component behavior can be studied independently of overall structural behavior. A cost-effective way of adding a large amount of information to the design database would be to conduct a few carefully designed experiments to test a large number of retrofitting methods simultaneously. The BMSP' s first priority should be evaluating available techniques, such as fiber wrapping and steel jacketing of columns, followed by techniques for improving the performance of floor slabs, particularly by strengthening them against uplift pressures. Technical analyses should be accompanied by economic data to aid engineers and other participants in the decision-making process. DTRA might consider partnering and cost-sharing arrangements with the private sector to accelerate the testing and deployment of emerging retrofitting technologies. Multihazard Mitigation The identification and assessment of design features and materials that can improve performance over a range of hazards (i.e., earthquake, fire, extreme wind events, chemical and biological agents) could have an ancillary benefit of improving building performance. Because design features that provide multihazard resistance are likely to generate more interest among designers and manufacturers than design features that promise only blast resistance, multihazard features could ultimately reduce the cost and increase the application of improved building practices and products. The BMSP should investigate construction techniques that not only mitigate blast effects but also permit the rapid repair, recovery, and continued use of damaged buildings. Obviously, preventing progressive collapse is the primary goal, but if other building systems, such as electrical service and distribution, air conditioning, and fire and life-safety systems, can be restored promptly, the diminished loss of revenue might make mitigation cost effective for owners. Therefore, the committee believes that life-safety system elements should be included in the testing program of the BMSP. Components that could be tested include fire alarm control panels and sensors, fire-suppression components, and emergency lighting fixtures. At the same time, protective enclosures around critical building systems to reduce shock effects on electronic equipment and utilities could also be tested. If these are effective, they could both enhance the ruggedness of communications and prevent secondary fires and explosions from, for example, damaged gas valves. A coordinated program of testing and analysis could determine minimum distances between redundant building systems (e.g., backup and main power boards) and provide a technical and economic basis for providing redundant building systems. The results of the program could contribute not only to safer designs for high-hazard locations but could also support postattack rescue and recovery operations.

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program REDUCING INJURIES THROUGH NONSTRUCTURAL APPROACHES Nonstructural Systems The BMSP Master Plan states: The protection of occupants of buildings from terrorist bomb attacks can be enhanced by an appropriate balance between better security procedures including the enforcement of increased standoff distances and the use of blast mitigation techniques. This program addresses only the blast hardening and blast mitigation aspects of the problem including design of blast walls, blast loading modification, and structural hardening (DTRA, 1999). In general, the BMSP has focused its efforts on building structures and related exterior wall components (i.e., structural approaches and physical design methods) for mitigating the blast effects, and consequently reducing the injuries and deaths, caused by terrorist bombs. Although the committee recognizes that the first task in designing a building to sustain a bomb blast is to prevent progressive collapse, once that has been achieved, minimizing loss of life and property from other blast effects becomes paramount. Substantial harm can be done even if structural collapse does not occur. For example, the board that reviewed the embassy bombings in Nairobi and Dar es Salaam in 1998 found: The damage to the embassy was massive, especially internally. Although there was little structural damage to the five story reinforced concrete building, the explosion reduced much of the interior to rubble—destroying windows, window frames, internal office partitions and other fixtures on the rear side of the building. The secondary fragmentation from flying glass, internal concrete block walls, furniture, and fixtures caused most of the embassy casualties. (U.S. Department of State, 1999). Although evidence that window glazing is a major contributor to blast-related injuries and also causes many deaths is overwhelming (Gans and Kennedy, 1996; Mallonee et al., 1996), the potential of nonstructural internal building configurations and components to cause and mitigate injury is not well understood. Some attempt has been made to address this problem for earthquakes. For example, a report by the Federal Emergency Management Agency, Identification and Reduction of Nonstructural Earthquake Hazards in Schools, states: Nonstructural hazards can occur in every part of a building and include everything but the columns, beams, floors, load-bearing walls, and foundations. Common nonstructural items include ceilings, lights, windows, office equipment, computers, files, air conditioners, electrical equipment, furnishings, and anything stored on shelves or hung on walls. In an earthquake, nonstructural elements may become unhooked, dislodged, thrown about, and tipped over; this can cause injury and loss of life, extensive damage, and interruption of operations. (FEMA, 1993). The Building and Fire Research Laboratory of the National Institute of Standards and Technology has reported that existing building codes do not adequately address the performance of nonstructural building components in earthquakes (Phan and Taylor, 1996). This report recommends research to assess the adequacy of current building codes and identify necessary improvements; to develop techniques to mitigate damage to ceiling components designed to

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program older codes; and to develop uniform guidelines for the design, installation, and restraint methods for ceiling components. Types of architectural partition and anchoring, interior building systems, and other nonstructural components can protect occupants by deflecting blast energy and shielding them from flying debris. At the same time, they also have the potential to become harmful debris themselves. The committee believes that the BMSP should focus more attention on the behavior of nonstructural systems in the blast environment, including tests of the effectiveness of various types of interior partitions or perimeter zones of “soft” space in protecting occupants and contents, as well as comparisons of floor-based systems of mechanical and electrical distribution and typical overhead systems. The committee believes that CTS-1 can be used effectively for further testing of both structural and nonstructural elements. Outside protective features, such as perimeter walls of various heights and construction located at various standoff distances, berms, and other landscape features could also be tested with CTS-1 as the structure to be protected. These tests could identify combinations of building surface and internal features that would reduce blast effects to acceptable levels. The testing program should take into account that buildings are not just structures but interconnected series of systems that can work together to improve performance and increase safety. Modeling for Injury Prediction The modeling for injury prediction in the BMSP Program Plan appears to be based on a good understanding of the technical issues and tasks involved. However, the committee questions the quality of existing empirical data on human injuries caused by building failures and, thus, their suitability for use in epidemiological analyses. Both the quantity and quality of data used in analyses and validations will be critical to the development of reliable models. The epidemiologic and engineering literature on risk factors for physical injuries from natural and man-made disasters that involve building failures (e.g., earthquakes, tornadoes, hurricanes, volcanic explosions) is a potentially rich source of data and should be reviewed (e.g., Abrams et al., 1998; Jones et al., 1990; Tanaka et al., 1999). This literature includes data on injuries caused by sources other than structural components of buildings, including occupant behavior and damage to building contents and nonstructural components. The engineering literature, which is more extensive than the epidemiologic literature on injuries caused by building damage, should also be reviewed. However, because most engineers are not trained in epidemiologic methods designed to reduce bias in the interpretation of data, the BMSP should involve both epidemiologists and engineers in developing injury prediction models. The committee notes that the BMSP's injury modeling does not include the so-called “crush syndrome,” a class of serious and fatal injuries related to building damage and collapse that occurs when heavy objects, such as collapsed brick walls, pin individuals down. Crush-syndrome victims may suffocate rapidly or suffer life-threatening internal injuries even in the absence of blast-related trauma (Better, 1999). Most injury assessment tools do not include crush-syndrome casualties, but the committee believes crush syndrome should be included in the BMSP's efforts to model blast injuries. Modeling and testing can contribute a great deal to an understanding of how to prevent injuries to the occupants of buildings. Postevent field studies are valuable tools for collecting

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program information under less-than-ideal, real-world conditions. These types of studies have been done after both seismic events and bomb blasts (Durkin and Theil, 1992; Mallonee et al., 1996). A key aspect of collecting data is interviewing rescuers, victims, and bystanders (both injured and uninjured) to assess the factors that contributed to their situations and outcomes, including the rapidity, safety, and ease of rescue. However, access to these building occupants decreases rapidly over time, as do memories of the events. Therefore, the BMSP could provide a valuable tool by formalizing an institutionalized process that could be quickly engaged for data collection, analysis, and dissemination. The development of this process before an event occurs would accelerate the mobilization and deployment of trained investigative team(s), with prearranged funding and logistics, that could collect data on building damage, occupant injuries, and rescue difficulties in a standardized format. Injury data should not be limited to occupant injuries but should include injuries to rescuers as well (e.g., physical trauma and hazardous exposures [asbestos, chemicals, sewage, electrical wiring, etc.]). If these data were evaluated for many events by disaster epidemiologists, they could be valuable for continued improvements in design and construction, as well as rescue and recovery (Wagner et al., 1994). The Centers for Disease Control and Prevention, which fielded ad-hoc teams in the aftermath of the Oklahoma City and African embassy bombings, could be a valuable partner in an institutionalized program of injury reduction through improved design and rescue and recovery methods. The bombing of the World Trade Center in 1993 demonstrated that fire and smoke propagation following a bomb explosion can be a significant cause of injuries to the building occupants (NRC, 1995). The flammability and smoke generation potential of building materials, furniture, and common office products have been studied extensively by many fire research organizations, including the Building and Fire Research Laboratory of the National Institute of Standards and Technology (BFRL). This work is included in FIREDOC, a web searchable bibliographic database of 55,000 holdings maintained by BFRL (BFRL, 2000). The committee believes that FIREDOC is a potentially rich source of data that can be useful for extending the range of injury prediction models to include injuries related to fire and smoke propagation. Rescue and Recovery Experience has shown that the most effective disaster planning is based on good evidence and data from past events (Auf der Heide, 1989). Critical data on blast injuries in buildings include: geographical patterns of injuries and their associations with design features, building materials, and building contents possibilities of escape from, or survival in, an attacked building accessibility to rescuers knowing where to look for survivors Although damage-resistant building designs are critical in preventing injuries from terrorist attacks and other disasters, the ease and rapidity with which trapped or injured occupants can be extricated is also important. Key factors are how easily occupants can be located, whether they are able to evacuate the building, and whether rescuers can safely enter areas of the collapsed structure to render aid. The BMSP should evaluate these factors in

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program cooperation with urban search and rescue teams and other rescue units to plan and carry out simulated rescue and recovery operations to refine or improve techniques. INTERNATIONAL PROGRAMS To date, the committee has heard only passing references to joint programs between DTRA and similar agencies in Israel and the United Kingdom. Although these programs are not funded by TSWG, the committee believes that similar work being carried out in other countries may offer DTRA opportunities to enhance program elements that are funded by TSWG, such as the BMSP. Some measure of international information sharing is currently being accomplished through the International Symposium on Interaction of the Effects of Munitions on Structures, a biennial event jointly sponsored by DTRA and the German Federal Ministry of Defence. The committee will review this activity, as well as other means for DTRA to incorporate the results of studies in other countries into its technology transfer efforts. CONCLUSIONS AND RECOMMENDATIONS Conclusion 2. Although the Program Master Plan includes many activities that could yield worthwhile benefits, the committee identified several modifications to the BMSP to be considered in the next program cycle. Recommendation 2a. All analytical and experimental activities should be designed to test a specific hypothesis about the outcome. With respect to full-scale tests, parametric studies should be conducted to determine what could be learned from the test on the basis of the proposed instrumentation. Recommendation 2b. The program should take full advantage of the advances in parallel-processor computing made by the U.S. Department of Defense and the U.S. Department of Energy to improve the capability and ease of use of computational tools for predicting structural responses to bomb blasts. Recommendation 2c. The residual strength of blast-damaged structural components should be investigated more fully. For example, tests of full scale columns representative of buildings ten stories and more should be included, as well as a series of tests to evaluate how well common rebar splices and connections can function after being damaged by blasts. Recommendation 2d. The Blast Mitigation for Structures Program should consider conducting a series of tests on masonry structures, including tests of unreinforced masonry for benchmarking purposes and tests of a range of reinforcement techniques to improve protection. A series of tests on construction typical of long-span buildings should also be considered. Recommendation 2e. The Blast Mitigation for Structures Program should place a higher priority on the development and evaluation of retrofitting techniques —particularly on creative

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program conceptual retrofitting measures that would prevent a life-threatening progressive collapse following a blast. Recommendation 2f. The Blast Mitigation for Structures Program should focus more attention on the behavior of nonstructural systems in the blast environment, including (1) tests of the effectiveness of various types of interior partitions or perimeter zones of “soft” space in protecting occupants and contents, and (2) comparisons of floor-based systems of mechanical and electrical distribution and typical overhead systems. Recommendation 2g. The Blast Mitigation for Structures Program should evaluate the key factors affecting the ease and rapidity with which trapped or injured occupants can be extricated from damaged buildings and whether rescuers can safely enter areas of the collapsed structure to render aid. In cooperation with urban search and rescue teams, the program should support simulated rescue and recovery operations to refine or improve rescue techniques. Conclusion 3. The design and engineering approaches favored by the industrial contractors and government laboratories that are implementing the BMSP may be more appropriate to traditional military and defense objectives despite the emphasis of the BMSP on nonstructural solutions, injury reduction, and improved rescue and recovery techniques. Recommendation 3. The contractor base should be broadened to increase the representation of the commercial architectural and engineering communities, as well as specialists in injury prevention, disaster medicine, and technology transfer, particularly in the planning phases of the program. Conclusion 4. The committee is in complete agreement with the BMSP's emphasis on determining progressive-collapse vulnerability of buildings in selected attack scenarios but believes this ability would be improved by fuller coordination of research activities. Recommendation 4. The Defense Threat Reduction Agency should adopt a balanced approach toward understanding and preventing the progressive collapse of buildings. This approach should include coordinated physical testing, experimentation, and analyses and should guide the planning of research activities and the interpretation and synthesis of the results. Conclusion 5. Full-scale testing of structural systems has been overemphasized at this relatively early stage of the program at the expense of reduced-scale testing, the development of retrofitting techniques for existing buildings, the testing of nonstructural building systems, and the investigation of technologies related to injury prevention. Recommendation 5. The Defense Threat Reduction Agency should not construct another full-scale test structure until the results of previous experiments on Controlled Test Structure-1 (CTS-1) have been fully analyzed and understood. At this stage of the program, DTRA should rely more on experiments with scaled elements and scaled assemblies of elements wherever scale effects are well understood.

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program Conclusion 6. Controlled Test Structure-1 has been underutilized so far; although it has been damaged in previous tests, it still has considerable value for testing full-scale structural components and nonstructural elements. Recommendation 6. Controlled Test Structure-1 should not be tested to failure because it can still be used as a reaction frame for component tests. Conclusion 7. Although, the inventory of existing buildings vulnerable to blast damage far exceeds the number of new buildings that will be constructed in the foreseeable future, the BMSP appears to have placed more emphasis on methods applicable to new construction than on retrofitting techniques for existing structures. Recommendation 7. The development of tools for conducting vulnerability assessments and strengthening existing buildings should be given a higher priority. Resources should also be allocated to investigating construction techniques that permit the rapid rehabilitation of blast-damaged buildings. Conclusion 8. The Blast Mitigation for Structures Program has a unique opportunity to determine how requirements and techniques for earthquake-resistant designs could apply to blast-resistant designs, as well as to identify and assess design features and materials that could improve building performance over a range of hazards (e.g.., earthquake, fire, flood, and extreme wind) that could impact the safety of the occupants. Recommendation 8. The Blast Mitigation for Structures Program should incorporate activities with the maximum potential for multihazard mitigation. Because design features that provide multihazard resistance are likely to generate more interest among designers and manufacturers than design features that promise only blast resistance, multihazard features could ultimately reduce the cost and increase the application of improved building practices and products. Conclusion 9. Data on blast-related injuries and building damage are limited and, therefore, have hindered the development of statistically valid damage-prediction and epidemiological models. Recommendation 9. The Blast Mitigation for Structures Program should initiate an institutionalized process that can be quickly mobilized for collecting critical data related to blast damage and injuries in buildings that are subject to bomb damage. REFERENCES Abrams, J., L. Bourque, J. Kraus, C. Peek-Asa, D. Vimalachandra, and J. Yu. 1998. Fatal and hospitalized injuries resulting from the 1994 Northridge earthquake. International Journal of Epidemiology27: 459–465. ACI (American Concrete Institute). 1999. Building Code Requirements for Structural Concrete and Commentary. Farmington Hills, Mich.: American Concrete Institute.

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program Attaway, S. 1999. Personal communication from Stephen Attaway, Distinguished Member of the Technical Staff, Sandia National Laboratories, Albuquerque, New Mexico, to Richard Little, director, Board on Infrastructure and the Constructed Environment, National Research Council, Washington, D.C., November 11, 1999. Auf der Heide, E. 1989. Disaster Response: Principles of Preparation and Coordination. St. Louis, Mo.: C.V. Mosby Company. Better, O.S. 1999. Rescue and salvage of casualties suffering from the crush syndrome after mass disasters. Military Medicine164(5): 366–369. BFRL (Building and Fire Research Laboratory), 2000. FIREDOC. Available on line at: http://fris.nist.gov/cgi-bin/starfinder/0?path=firedoc.txt&id=anon&pas=anon&OK =OK Corley, W.G., P. Mlakar Sr., M. Sozen, and C. Thornton. 1998. The Oklahoma City bombing: summary and recommendations for multihazard mitigation. Journal of Performance of Constructed Facilities12(3): 100–112. Dadazzio, R. 1999. Comment by Raymond Dadazzio, Weidlinger Associates, at the DTRA/TSWG Program Review on Structural Collapse, Alexandria, Virginia, November 4, 1999. DTRA (Defense Threat Reduction Agency). 1999. Blast Mitigation for Structures Program Master Plan, June 1999. Alexandria, Va.: Defense Threat Reduction Agency. Durkin, M.E., and C.C. Theil Jr. 1992. Improving measures to reduce earthquake casualties. Earthquake Spectra 7(1): 95–113. Ettouney, M., R. Smilowitz, and R. Dadazzio. 1998. Comparison between Design Requirements of Earthquake and Blast Events. Paper T210-1 in Structural Engineering World Wide 1998, edited by N.K. Srivastava. Available on CD-ROM from Elsevier Science, New York. FEMA (Federal Emergency Management Agency). 1993. Identification and Reduction of Nonstructural Earthquake Hazards in Schools. FEMA-241. Washington, D.C.: Federal Emergency Management Agency. Gans, L., and T. Kennedy. 1996. Management of unique clinical entities in disaster medicine. Emergency Medicine Clinics of North America14(2): 301–326. Jones, N.P., F. Krimgold, E. Noji, and G. Smith. 1990. Considerations in the epidemiology of earthquake injuries. Earthquake Spectra6(3): 507–528. Loizeaux, M. 1999. Personal communication from Mark Loizeaux, CEO, Controlled Demolition Incorporated, Phoenix, Maryland, to Richard Little, director, Board on Infrastructure and

OCR for page 15
Blast Mitigation for Structures: 1999 Status Report on the DTRA/TSWG Program the Constructed Environment, National Research Council, Washington, D.C., November 12, 1999. Mallonee, S., S. Shariat, G. Stennies, R. Waxweiler, D. Hogan, F. Jordan. 1996.Physical injuries and fatalities resulting from the Oklahoma City bombing.Journal of the American Medical Association 5: 382–387. Newmark and Associates. 1965. Design of Model Test Program for a Buried Field Shelter.Contract Report No. I-110. Report prepared for U.S. Army Engineer Waterways Experiment Station. Vicksburg, Miss.: U.S. Army Corps of Engineers. NRC (National Research Council) 1995. Protecting Buildings from Bomb Damage: Transfer of Blast-Effects Mitigation Technologies from Military to Civilian Applications. Washington, D.C.: National Academy Press. Office of Statewide Health Planning and Development. 1990. A Recommended Program to Seismically Strengthen Pre-Hospital Act Hospital Facilities: A Response to Milestone 4, Initiative 1.2, “California at Risk,” December 1990. Sacramento, Calif. Building Safety Board, Office of Statewide Health Planning and Development, Division of Facilities Development and Financing. Phan, L.T., and A.W. Taylor, 1996. State of the Art Report on Seismic Design Requirements for Nonstructural Building Components. NISTIR 5857. Gaithersburg, Md.. U.S. Department of Commerce. Sozen, M. 1999. Personal communication from Mete Sozen, Kettelhut Distinguished Professor of Structural Engineering, Purdue University, to Richard Little, director, Board on Infrastructure and the Constructed Environment, National Research Council, Washington, D.C., December 11, 1999. Tanaka H., J. Oda, A. Iwai, Y. Kuwagata, T. Matsuoka, M. Takaoka, M. Kishi, F. Morimoto, K. Ishikawa, Y. Mizushima, Y. Nakata, H. Yamamura, A. Hiraide, T. Shimazu, and T. Yoshioka. 1999. Morbidity and mortality of hospitalized patients after the 1995 Hanshin-Awaji earthquake. American Journal of Emergency Medicine 17(2): 186–191. U.S. Department of State. 1999. Report of the Accountability Review Boards on the Embassy Bombings in Nairobi and Dar es Salaam on August 7, 1998, January 1999. Available on line at: http://www.state.gov/www/regions/africa/accountability report.html Wagner, R.M., N.P. Jones, and G.S. Smith. 1994. Risk factors for casualty in earthquakes: the application of epidemiologic principles to structural engineering. Structural Safety13: 177–200. Woodson, S.C., and T. Krauthammer. 1998. Recent Developments in Blast-Resistant Structural Detailing. Paper T210-3 in Structural Engineering World Wide 1998, edited by N.K. Srivastava. Available on CD-ROM from Elsevier Science, New York.