Page 26

3
Review of Existing Knowledge for Blast-Effects Mitigation and Protective Design Technologies

Introduction

This chapter contains an assessment of the state of the art in blast-effects mitigation and protective design technologies and highlights blast-hardening capabilities developed by the military that are applicable to the civilian sector. The introductory section outlines the historical development of the existing knowledge base, the nature of explosions, and the induced physical behavior of materials. This is followed by a section on experimental and simple analytical approaches to blast-effects mitigation and structural design. The far-ranging needs of military organizations to protect key assets from enemy attack, together with the large cost and practical limitations of field testing, has led to a strong emphasis on developing analytical methods and advanced computer models for simulation of blast-effects on building structures. The third and fourth sections summarize applicable military design manuals and computational approaches, respectively, to predicting blast loads and the responses of structural systems. Although the majority of military design guidance is based on semi-empirical relationships, much of the existing knowledge of blast environments and the effects of high-intensity, short-duration loadings on the behavior of structures, systems, and components is embodied in sophisticated "first-principle" computer programs as well as in simplified, design oriented and semi-empirical software. Accordingly, the fifth section discusses specific computer programs, followed by a section on code validation which looks at comparisons of computational calculations of the response to an explosion to the actual experimental results. The



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 26
Page 26 3 Review of Existing Knowledge for Blast-Effects Mitigation and Protective Design Technologies Introduction This chapter contains an assessment of the state of the art in blast-effects mitigation and protective design technologies and highlights blast-hardening capabilities developed by the military that are applicable to the civilian sector. The introductory section outlines the historical development of the existing knowledge base, the nature of explosions, and the induced physical behavior of materials. This is followed by a section on experimental and simple analytical approaches to blast-effects mitigation and structural design. The far-ranging needs of military organizations to protect key assets from enemy attack, together with the large cost and practical limitations of field testing, has led to a strong emphasis on developing analytical methods and advanced computer models for simulation of blast-effects on building structures. The third and fourth sections summarize applicable military design manuals and computational approaches, respectively, to predicting blast loads and the responses of structural systems. Although the majority of military design guidance is based on semi-empirical relationships, much of the existing knowledge of blast environments and the effects of high-intensity, short-duration loadings on the behavior of structures, systems, and components is embodied in sophisticated "first-principle" computer programs as well as in simplified, design oriented and semi-empirical software. Accordingly, the fifth section discusses specific computer programs, followed by a section on code validation which looks at comparisons of computational calculations of the response to an explosion to the actual experimental results. The

OCR for page 26
Page 27 final section of the chapter reviews the applications of the computational methods to terrorist threats. Although not directly related to the subject of this study, two previous National Research Council (NRC) studies have dealt with terrorism and bombing: The first is The Embassy of the Future (NRC, 1986), which reviews the measures that could be applied to new foreign service buildings for all categories of threats, including bombs. The Department of State has since developed a large body of data and design guidelines related to countering terrorism through design. The second study is Protection of Federal Office Buildings Against Terrorism (NRC, 1988) which addresses the protection of existing domestic federal buildings from various threats, including bombs. Historical Background Explosive devices have been used for hundreds of years, yet comprehensive treatment of blast-effects and their mitigation appeared in the Western Hemisphere only during and after World War II. Following the war, the Office of Scientific Research and Development (National Defense Research Committee, 1946) produced the seminal, unclassified document on weapons and penetration capabilities. This report remains of value today. While the document is difficult to locate, some of the information it contains has been republished (Bangash, 1993). Of special interest in the latter publication is information that permits estimates to be made of the penetration resistance of structural elements of various materials to projectiles of many forms. World War II was the first international conflict that resulted in massive destruction of cities, mostly with high explosives, which also inflicted enormous casualties. In the latter stages of that war, the use of two nuclear weapons demonstrated the destructive capability of such weapons (Glasstone and Dolan, 1977). The accelerated arms race during the Cold War, from 1945 through 1990, led to research and development of modern nuclear weapons during this time period, and a major research program on protective structures and systems was established, largely sponsored by the Armed Forces Special Weapons Project, later renamed the Defense Atomic Support Agency, and thereafter the Defense Nuclear Agency (DNA). Throughout this period, DNA, in cooperation with the U.S. Air Force, the U.S. Navy, and the U.S. Army, conducted major test programs in laboratories and field test facilities. In addition, theoretical and experimental work sponsored by these agencies was carried out in various companies and universities. In the United States a comprehensive program of research over the past half century was undertaken to increase the blast resistance of military structures such as weapons storage facilities and command, control, and communication facilities. Much of this research was in response to deployment of ballistic and guided missile systems. In addressing both the nuclear threat and the threat of conven-

OCR for page 26
Page 28 tional weapons, a number of widely used manuals on protective structures were developed. These manuals reflect the improvements in engineering practice that occurred over the years (Newmark and Haltiwanger, 1962; Crawford et al., 1974; ASCE, 1985; Schuster et al., 1987). This progression of manuals concentrated on techniques for estimating the loadings from nuclear weapons explosions, attenuation of pressure effects in the air and ground, simple analytical techniques for design and proportioning of structural elements, guidelines for designing and analyzing equipment, and many other related topics. Although nuclear weapons effects were the primary focus of protective design through the 1970s, related work was ongoing to develop quantitative procedures for the design of structures subject to accidental explosion. Extensive research and development programs, including numerous full- and small-scale structural response and explosive effects tests, form the empirical basis for Structures to Resist the Effects of Accidental Explosions (U.S. Departments of the Army, Navy, and Air Force, 1969). Several other manuals dealing with blast effects and the design of protective structures have also been developed for use by the military services (U.S. Department of the Army, 1986, 1990; Drake et al., 1989; U.S. Department of Energy, 1992). Although none of these manuals is dedicated solely to the subject of this report, taken together, they provide a range of structural data and design procedures for protective structures useful to designers and which have applicability to the design of more explosion-resistant buildings. Increased terrorist activity throughout the world and in the United States has also led to a number of specialized studies in recent years. These studies and a variety of design manuals are discussed later in this chapter. Nature of Explosions Explosive materials are designed to release large amounts of energy in a short time. The explosion arises through the reaction of solid or liquid chemicals or vapor to form more stable products, primarily gases. A high explosive is one in which the speed of reaction (typically 5,000–8,000 m/s) is faster than the speed of sound in the explosive. High explosives produce a shock wave along with gas, and the characteristic duration of a high-explosive detonation is measured in microseconds (10-6 s). Explosives come in various forms, commonly called by names such as TNT, PETN, RDX, and other trade names (U.S. Department of the Army, 1986; McGraw-Hill Encyclopedia of Science & Technology, 1987). A common explosive employed in rack blasting, called ANFO, is composed of ammonium nitrate and diesel fuel oil—products that are readily available. Dynamite, of which there are many kinds, is also readily available, and theft and misdirection in shipping occur occasionally. The lethality of high explosives has been increasing since the nineteenth century (Johansson and Persson, 1970; Henrych, 1979; Baker et al., 1983; Dick et al., 1983; Fickett, 1985; McGraw-Hill Encyclopedia of Science &

OCR for page 26
Page 29 Technology, 1987). For someone bent on destruction, high explosives are relatively easy to make or acquire, and detonation, though technically more complicated, is not difficult for someone with even modest training in explosives. The effects of explosions on structures are directly related to stress-wave propagation as well as impact and missile penetration. In all close-in explosions, where shock waves must travel through the surrounding medium to cause damage to a facility, a realistic description of the wave-propagation phenomena is needed. The literature on these subjects can be divided into two groups: one group addresses the classical issue of wave propagation, with emphasis on linear-type problems, and the second group is more focused on nonlinear problems. (Various theoretical aspects of the explosive effects in materials are discussed in Achenbach, 1973; Whitham, 1974; Davis, 1988; Han and Yin, 1993; and Batsanov, 1994.) The effects of an explosion are varied. For explosions close to the targeted object, the pressure-driven effects occur quickly, on the order of microseconds to a few milliseconds. The air-blast loads are commonly subdivided into (1) loading due to the impinging shock front, its reflections, and the greatly increased hydrostatic pressure behind the front, all commonly denoted as overpressure; and (2) the dynamic pressures due to the particle velocity, or mass transfer, of the air. It is customary to characterize the pressure loadings in terms of scaled range, as given by Z = R/W1/3, in which Z is the scaled range, R is the radial distance between the explosion center and the target, and W is the explosive weight (normally expressed as an equivalent TNT weight). Units for charge weight and distance should be either pounds and feet or kilograms and meters. In the scaled-range concept, as long as the value of Z remains the same, the same parameters for the explosive effects (i.e., peak pressure, positive duration, etc.) should be obtained. If an explosion is confined by a chamber or room, the gas pressure increases rapidly to a sustained level and then decays by venting out. Under these conditions shock reflections occur and the overall effect can be greater than that of the incident shock. The effects of internal explosions can be devastating to buildings and their occupants, which supports evaluation of the possibility of mitigating blast-effects by controlled venting. There is a considerable body of knowledge available concerning blast-effects mitigating techniques for buildings subject to accidental explosion (U.S. Department of the Army, 1990), which may have applicability to the design of civilian office structures. There are three additional explosion-related phenomena relevant to this study, namely impact of objects propelled by the explosion environment (Jones, 1989), penetration of such objects (Zukas, 1990; Bangash, 1993), and ground-transmitted shock (U.S. Department of the Army, 1986, 1990; Drake et al., 1989; U.S. Department of Energy, 1992; DNA, 1995).

OCR for page 26
Page 30 Induced Physical Behavior If the explosion originates at a sufficiently great scaled range (i.e., a small charge or a large distance from a structure), then the structure will be loaded in a manner that leads to global deformation, meaning that all the elements provide some degree of resistance to the loading. The definition of the expected loading, and the provision of resisting elements to accommodate the loading, are the essence of dynamic design, analysis, and construction; these issues are addressed in the previously cited design manuals and by the computer codes discussed later in this chapter. If the explosion is sufficiently close to a wall or floor (that is, with a small scaled range), there can be gross disintegration, with either spalled fragments coming off the front and back sides or wall fragments themselves being propelled as missiles. These fragments can injure people, damage property, and, if structural support is sufficiently disrupted, cause the building to collapse. At intermediate scaled ranges, both global and localized response, including severe cracking, with near-face disintegration and spalling on the rear face, can be expected. When an explosion impinges on a structural element, a shock wave is transmitted internally at high speed; for example, dilatational waves (tension or compression) propagate at speeds of 2,700–3,400 m/s in typical concrete and 4,900–5,800 m/s in steel. At these speeds, reflections and refractions quickly occur within the material (within milliseconds), and, depending on the material properties, high-rate straining and major disintegration effects can occur. For example, under extremely high shock pressures, concrete, a relatively brittle material, tends to undergo multiple fractures which can lead to fragmentation. In steel, under similar conditions, depending on the material properties and geometry, yielding and fracture can be expected, especially if fabrication flaws are present, with fragmentation occurring in some cases. Primary, fragments are produced when a detonating explosive is in contact with a material such as concrete or steel. The initial velocity of the primary fragments depends in part on the detonation pressure. Secondary fragments are produced by the effect of the blast wave on materials not in contact with the explosive. Openings such as doors and windows require special design considerations if intrusion of the explosive shock wave is to be averted, or damage mitigated. Where high levels of blast-effects mitigation are sought, labyrinth ( and ) entrances, possibly with blast doors, as well as ventilation blast valves, can be used. As described earlier, explosions in a partially or fully confined space, as a room or garage, can be even more devastating, with higher pressures than would occur in free air and a longer duration of loading. In such situations significant damage can be expected. Other explosion-generated effects are also produced, such as fire (including smoldering fires), smoke, pressure damage to ears and other organs, and violent motion of the structure and its contents. Such shock-related motion can result in

OCR for page 26
Page 31 personal injury and equipment damage and cause the loss of lifelines such as utilities and communications cables. Experimental and Simple Analytical Approaches Theory and experiments are essential for predicting the blast effects of an explosion. Experimental data may be combined with certain aspects of explosion theory to properly characterize material behavior at high strain rates, which in turn can be employed in developing computational approaches for estimating structural and equipment reaction to an explosion. It is important to validate such computational approaches by experiment whenever possible. Some military testing programs have concentrated on the effects of nuclear weapons, frequently employing specially designed high-explosive devices to simulate either close-in or far-field effects of nuclear blast waves. The large-scale ANFO testing program using full-scale structures conducted by the military is one example of data collection specifically addressed to the issue of analytic verification. In these tests, approximately 450 metric tons of ANFO charges were used to produce overpressure loadings below approximately 0.69 megapascals (100 psi) corresponding to a 1 kt nuclear weapon; other high-explosive techniques were employed occasionally to obtain higher pressures. The most common method used in this type of testing has been to load reduced-scale structures with either weapons or special high-explosive devices. In many cases, testing conditions permit only one structural scale and charge per study rather than a range of scaled sizes and charge weights, and, for a variety of reasons, comprehensive post-test studies often have not been conducted. Although the voluminous data available from weapons-systems test procedures and related military experiments may require extensive analysis and interpretation before the results can be fully incorporated into the knowledge base, this data provides a rich source of information not readily available elsewhere.1 The Accident Data Base of the Department of Defense Explosives Safety Board (DDESB) is another valuable source of empirical data on explosive events. The DDESB accident reports analyze damage sustained by structures in accidental explosions and can provide considerable insight into the performance of structural elements following an explosion. 1The majority of tests were conducted to determine the survivability of existing or proposed blast-hardened U.S. facilities: other tests were conducted to determine the vulnerability of possible enemy facilities, hardened or otherwise. From a survivability perspective, test objectives dealt with determining margins of strength of fully described facilities, whereas from a vulnerability perspective, the objectives concerned failure modes of facilities whose characteristics were known only approximately. Thus, what was conservative in one type of test, tended to be unconservative in the other, and extreme care must be taken in evaluating specific results for application in civilian practice.

OCR for page 26
Page 32 The ability to design and construct safe buildings depends on an understanding of how such buildings behave under severe loads (i.e., cause-and-effect relationships). Unfortunately, past experiments, performed mainly by the military, have shown that supposedly identical high-explosive devices frequently produce significantly different loading environments. Consequently, progress in designing structures subject to explosive loading has been difficult to attain. A program of precision testing and code simulations in which loads are well defined and the outcome is well documented and assessed is clearly a prerequisite to further progress in designing blast-resistant buildings. The results of full-scale tests for strengthening existing civilian structures against terrorist attack were recently reported at an international symposium (Fouks, 1993). Results were obtained for windows, doors, and light ceilings and roofs, and appear to be limited to failure pressures; no computational analyses of these tests have been reported. The suitability of these results for computer program validation (as discussed later in this chapter) is questionable because of limited data. Two companion papers report on measures to increase the blast resistance of walls and ceilings (Eytan and Kolodkin, 1993) and windows and doors (Kolodkin and Eytan, 1993). There has also been extensive test programs on windows in the United States and the United Kingdom. Over the years, as a result of research coupled with test programs, computational approaches have been developed for estimating the responses and behavior of simple structures subjected to blast loading. In turn, based on experimental data, field-test observations and analytical procedures, a number of technical design manuals were developed, as described in the next section. Technical Design Manuals Structures to Resist the Effects of Accidental Explosions, TM 5-1300 (U.S. Departments of the Army, Navy, and Air Force, 1990). This manual appears to be the most widely used publication by both military and civilian organizations for designing structures to prevent the propagation of explosion and to provide protection for personnel and valuable equipment. It includes step-by-step analysis and design procedures, including information on such items as (1) blast, fragment, and shock-loading; (2) principles of dynamic analysis; (3) reinforced and structural steel design; and (4) a number of special design considerations, including information on tolerances and fragility, as well as shock isolation. Illustrative computations are also included in many cases. Guidance is provided for selection and design of security windows, doors, utility openings, and other components that must resist blast and forced-entry effects. The manual contains a valuable listing of relatively current references. Distribution is unlimited. A Manual for the Prediction of Blast and Fragment Loadings on Structures, DOE/TIC-11268 (U.S. Department of Energy, 1992). This manual provides guid-

OCR for page 26
Page 33 ance to the designers of facilities subject to accidental explosions and aids in the assessment of the explosion-resistant capabilities of existing buildings. It includes chapters on air blast; cratering and ground shock; fragment ballistics, including a thorough description of secondary debris hazards, secondary explosions, and the dynamic properties of materials. It is intended to be used in conjunction with other structural design manuals and provides a comprehensive listing of references. Distribution is unlimited. Protective Construction Design Manual, ESL-TR-87-57 (Air Force Engineering and Services Center, 1989). This manual provides procedures for the analysis and design of protective structures exposed to the effects of conventional (non-nuclear) weapons and is intended for use by engineers with basic knowledge of weapons effects, structural dynamics, and hardened protective structures. Chapters cover topics such as uncertainties in protective design, air-blast effects, fragment protection, loads on structures, resistance of structural elements, and dynamic responses of structures. Distribution is limited. Fundamentals of Protective Design for Conventional Weapons, TM 5-855-1 (U.S. Department of the Army, 1986). This manual provides procedures for the design and analysis of protective structures subjected to the effects of conventional weapons. It is intended for use by engineers involved in designing hardened facilities. It includes chapters on air-blast effects, fire, incendiary and chemical agents, loads on structures, and auxiliary systems (piping, air ducting, and electrical cable). Distribution is unlimited. Design of Structures to Resist Nuclear Weapons Effects, Manual 42 (ASCE, 1985). This manual was prepared for civilian use, and has been widely distributed throughout the world. It contains information on weapon detonation characteristics, radiation shielding, blast and shock-loadings, applicable limit-load theory, simplified dynamic analysis procedures, and design procedures for structures as well as equipment. Even though the procedures emphasized are perhaps oversimplified, the manual has a broad audience. Distribution is unlimited. The Design and Analysis of Hardened Structures to Conventional Weapons Effects (DAHS CWE) (DNA, 1995). This new Joint Services manual, written by a team of more than 200 experts in conventional weapons and protective structures engineering, supersedes U.S. Department of the Army TM 5-855-1, Fundamentals of Protective Design for Conventional Weapons (1986), and Air Force Engineering and Services Center ESL-TR-87-57, Protective Construction Design Manual (1989). The manual is based on state-of-the-art design information and methods for protective structures and includes new, recently analyzed and validated test data from the DNA test programs on conventional weapons effects,

OCR for page 26
Page 34 as well as design examples. Selected sections of the manual have unlimited distribution, but the manual as a whole has limited distribution. In parallel with this effort, a DAHS CWE hypertext system based on the DAHS CWE hardcopy manual is being developed. This electronic hypertext version is intended to transform the manual into an interactive computer product complete with text, figures, graphs, tables, equations, and a number of stand-alone computer codes such as BLASTX and FOIL (discussed later in this chapter). All DAHS CWE system software and codes will be produced and distributed on CD-ROM and will eventually operate on both DOS- and Unix-based platforms. The hypertext system is expected to be completed in the fall of 1995. Security Engineering, TM 5-853 (U.S. Department of the Army, 1993). The Department of Defense has recently shown an increased interest in applying systems engineering approaches to the design of military facilities for increased physical security against a range of threats, including terrorist attack. A three-volume security engineering manual has been developed that is intended for new construction and provides designers with guidance for protecting assets within facilities against a range of criminal, protester, terrorist, and subversive threats. Distribution is limited. Terrorist Vehicle Bomb Survivability Manual (Naval Civil Engineering Laboratory, 1988). This manual contains information on vehicle barriers and blast survivability for buildings. It provides information to aid owners in protecting their property, assets, and personnel against terrorist vehicle bombs. This manual includes information on access control, vehicle barrier systems and testing, and sample blast vulnerability analyses. Distribution is limited. Structural Design for Physical Security—State of the Practice Report (ASCE, 1995). This report is intended to be a comprehensive guide for civilian designers and planners who wish to incorporate physical security considerations into their designs or building retrofit efforts. Individual chapters are devoted to threat determination, load definition, structural systems behavior and design philosophy, structural components behavior and design, security window design, door design, utility opening design, and retrofitting existing structures. Publication is expected in 1995. Balanced Survivability Assessment (Cicolani, 1994) DNA has developed and uses a method of survivability assessment that also appears applicable to architectural design, particularly for retrofitting existing buildings. This method incorporates a comprehensive systems approach to survivability assessment. Its elements include consideration of a full-threat spectrum analysis, assessment of the capabilities of all the physical systems of the facility to meet the threat

OCR for page 26
Page 35 (systemwide and single-point assessment), with assessment identification of the likelihood of failure and consequences on mission readiness and capability. It appears to the committee that this approach could be adapted to civilian application with relative ease. Computational Techniques During the past 25 years, powerful computer programs have been developed for predicting blast loads and the resulting structural response. This section discusses the methods used in and the validation of these programs. To provide the nonexpert reader with some background on the need for validation, the classifications of semi-empirical and first-principle programs and linear and nonlinear problems are introduced. The purpose of this discussion is to support the theme that since blast evolution and response problems are highly nonlinear, validation of the computer programs by experiments is a necessity. The section also notes that a considerable degree of expertise is needed to use these programs effectively. DNA and other military organizations have conducted numerous tests and experiments for validating these computer programs over the years, and the need for validation is discussed later in this chapter. However, the structures in these validation tests and experiments were generally representative of military applications. It is not clear how relevant these previous tests are to civilian structures which are typically lighter in construction yet at the same time more structurally complex than the military structures tested. First-Principle and Semi-Empirical Methods Computer programs for the prediction of blast-effects can be subdivided in two groups: first-principle and semi-empirical. In first-principle programs, mathematical equations are solved that describe the basic laws of physics governing a particular problem. These principles are conservation of matter, momentum, and energy. In addition, mathematical relationships called constitutive equations, which describe the physical behavior of materials, are needed. If these equations are solved accurately with suitable mathematical models, they should predict the blast loads and structural response. However, there are several barriers to accurate prediction of the effects of an explosion through the use of first-principle programs. Among them are the following: • In the calculation of blast due to explosions in air, the response of the air often involves complicated phenomena, such as dust-air mixtures, boundary effects, and turbulence. Turbulent flow, for example, cannot be calculated without the addition of models governed by empirical parameters. • Calculation of the pressures imparted by a detonating explosive on the

OCR for page 26
Page 36 structure involves multiscale phenomena that are very difficult to deal with; such phenomena also occur in the structure during failure. • In the calculation of structural failure, the behavior of the materials is neither well understood nor readily characterized; in other words, accurate constitutive equations are not available for the materials, particularly in fracture or fragmentation. While these deficiencies in first-principle codes are often compensated for by the use of engineering judgment, the main objective of first-principle techniques is to provide predictions in new domains where the experience that makes engineering judgment possible is not available. Semi-empirical computational methods are based on simplified models of physical phenomena, which are developed through analysis of test results and application of engineering judgment. These methods rely on extensive data and case studies. They involve fewer equations and require far less computer time, which makes them more practical than first-principle codes for design purposes. The computer programs applied in the evaluation of explosive effects cover two physical disciplines: • computational fluid dynamics (CFD), which is used for the prediction of the air blast caused by the explosion and the pressures applied to surfaces exposed to the propagating air blast; and • computational solid mechanics (CSM), which deals with the prediction of the response of structures to loads. The pressures and the response of the structure are interrelated, and in many cases ''coupled'' analyses of the fluid and structure (where the fluid solution is obtained interactively with the structural solution) are needed. Coupled CSM-CFD solutions entail the use of much larger computer programs and are more costly, but they can provide more accurate predictions. Linear and Nonlinear Problems Computer models and programs have become indispensable in engineering design and development. The complexity and dependability of the models varies dramatically. To provide a perspective for blast programs, this section introduces several classifications. An important classification of computer simulation models and analyses is whether the governing equations and response are linear or nonlinear. Linear analysis is applicable when the displacements of a structure or medium are small and the stresses can be related to the deformation by linear relationships. Examples of linear analyses include acoustic-wave propagation and stress analyses of structures and machines under normal operating loads (referred to as

OCR for page 26
Page 38 • Many parameters need to be input, including artificial viscosities and choices among methodologies, and their selection requires considerable experience. • The development of inputs requires construction of detailed finite-element or finite-difference models, which entails selection of the features of the physical sites that are important and a knowledge of how to model them. • Considerable skill is required to evaluate the output, both as to its correctness and its appropriateness to the situation modeled; without such judgment, it is possible through a combination of modeling errors and poor interpretation to obtain erroneous or meaningless results. • Conversely, the generality of these programs and the relative ease with which details of a physical system can be incorporated in the simulation model, even by those without the requisite expertise, can lead to unwarranted confidence in the validity of specific results. Therefore, successful computational modeling of specific scenarios by engineers unfamiliar with these programs is difficult, if not impossible. Current research is seeking ways to make these and other sophisticated computer programs more intuitive and user-friendly. The committee learned of one such effort to develop a powerful methodology which facilitates simulation-based design by providing the user with an expert system to assist in the development of the model and specification of the parameters required for a computational simulation (SAIC, 1994). Such tools would make the programs developed for blast-effects simulation more accessible to engineers who are not experts in the programs themselves. Another impediment to the use of these programs is the magnitude of computational resources required. Some of the simulations take 20–100 hours on the most powerful supercomputers, such as the Cray C-90. Small-scale, two-dimensional calculations can often be made on workstations in a matter of minutes or hours, but the time required grows rapidly as the model increases in detail. Other Applications of Numerical Simulation Techniques Computational methods are proven techniques that are used extensively in commercial engineering design and evaluation. For example, in the automotive industry, crashworthiness design is based on computer simulations which can predict passenger acceleration levels, and hence whether the crash can be survived without serious injury. CFD programs, similar to those used for blast prediction, are used to design the air flow for cooling the engine compartment and for drag reduction. The use of computer models in the automotive industry has reduced the numbers of prototypes that must be built during the design cycle and thus shortened the time required from inception of design to production. The automotive industry is able to amortize the cost of computer modeling over the

OCR for page 26
Page 39 large number of units produced. Nonlinear computational mechanics is finding increasing application in design manufacturing processes, such as sheet-metal forming, extrusion, forging, and casting, and to simulate prototype tests, such as drop tests for electronic products and durability tests of safety-critical components. The methods used for prediction of blast effects and structural damage are identical to those used in manufacturing and some of the same computer programs are used in both areas. Computer Programs for Blast and Shock Effects Computational methods in the area of blast-effects mitigation are generally divided into those used for prediction of blast loads on the structure and those for calculation of structural response to the loads. Computational programs for blast prediction and structural response use both first-principle and semi-empirical methods. The governing equations of physics include conservation of momentum, conservation of energy, measures of stress and strain, and laws governing the relation between stress and strain, which depend on the physical properties of materials involved. To some extent, prediction in solids is simplified as compared to fluids, since there is no counterpart to turbulence; however, the dynamic behavior of solid materials generally is far more complicated than fluids, particularly when the structure fractures or fragments. Most explosion-induced structural response calculations are made in an uncoupled manner. This involves calculating blast loads as if the structure (and its components) were rigid and then applying these loads to a responding model of the structure. The shortcoming of this procedure is that when the blast field is obtained with a rigid model of the structure, the loads on the structure are often overpredicted, particularly if significant motion or failure of the structure occurs during the loading period. An example of this was the overprediction of pressures in a numerical simulation using the FEFLO program (see below) after the World Trade Center bombing. A current, active area of research is addressing the need for coupled calculations. In coupled calculations, the CFD model for blast-load prediction is solved simultaneously with the CSM model for structural response; that is, for a coupled calculation, the blast prediction program is linked with a structural response program. By accounting for the motion of the structure while the blast calculation proceeds, the pressures that arise due to motion and failure of the structure can be predicted more accurately. Several efforts in this direction are now under way. Under DNA sponsorship, FEFLO has been coupled to DYNA3D and its adaptivity features added to DYNA3D. This change allows FEFLO to be used for the blast calculation and DYNA3D) for the structural response calculation. Table 3-1 summarizes a partial listing of computer programs that are currently being used to model blast-effects on structures, with more detailed descrip-

OCR for page 26
Page 40 TABLE 3-1 Representative Computer Programs Used to Simulate Blast Effects and Structural Response Summary of Computational Codes Name Purpose Type Corporate Author Reference BLASTX Blast prediction Semi-empirical SAIC Britt and Lumsden, 1994 CTH Blast prediction First-principle Sandia National Laboratories McGlaun et al., 1990 FEFLO Blast prediction First-principle SAIC Baum et al., 1994 FOIL Blast prediction First-principle Applied Research Associates, Waterways Experiment Station Windham et al., 1993 HULL Blast prediction First-principle Orlando Technology, Inc. Gunger, 1992 SHARC Blast prediction First-principle Applied Research Associates, Inc. Hikida et al., 1988 DYNA3D Structural response First-principle Lawrence Livermore National Laboratory Whirley and Engelmann, 1993 EPSA-II Structural response First-principle Weidlinger Associates Atkatsh et al., 1994 FLEX Structural response Firm-principle Weidlinger Associates U.S. Department of Energy, 1992 ALEGRA Coupled analysis First-principle Sandia National Laboratories Budge and Peery, 1993 ALE3D Coupled analysis First-principle Lawrence Livermore National Laboratory American Society of Mechanical Engineers, 1993 DYNA3D/FEFLO Coupled analysis First-principle Lawrence Livermore National Laboratory/SAIC Löhner et al., 1995 FUSE Coupled analysis First-principle Weidlinger Associates Sandler and Rubin, 1990 MAZe Coupled analysis First-principle TRT Corporation Schlamp el al., 1995

OCR for page 26
Page 41 tions of the codes presented in Appendix B. Code validation has occurred in all cases, with the rigor of a particular validation depending on such factors as code maturity, urgency of application, availability of funding, etc. Software availaiblity to the private sector varies, with some codes being currently available for sale (e.g., FLEX), others classified as "export-controlled" (CTH), and some having limited availability due to national security classification. Corporate authors can provide availability information for particular codes. Except for a few cases where the code can be run on a personal computer (e.g., FLEX), these codes are generally of sufficient complexity to require workstation or mainframe hardware. Code Validation Prediction of the blast-induced pressure field on a structure and its response involves highly nonlinear behavior. Computational methods for blast-response prediction must therefore be validated by comparing calculations to experiments. It is important to note, however, that experimental validation applies only to the class of events (or domain of applicability) encompassed by the experiments. For example, an experiment involving explosive loading of an unreinforced concrete wall subject to a blast pulse of 20.67 megapascals (3,000 psi) peak pressure and with a duration of 10 ms could be used only to validate a computer program for predictions for structures with similar characteristics and stiffness loaded over a similar pressure pulse. During the past 30 years, many experiments and tests have been conducted by the Department of Defense for the purpose of code validation. At the beginning of this period, CFD and CSM developments were in their infancy, and few experiments were designed specifically to validate individual codes; tests mainly focused on determining response modes and failure mechanisms of representative military structures. The response of this class of structures may be significantly different from conventional civilian structures for the reasons mentioned earlier. DNA recently conducted a series of experiments to evaluate the effects of penetrating weapons on hardened, reinforced concrete structures and to validate first-principle blast and structural response programs. The test structure had reinforced concrete walls and horizontal slabs 45–60 cm thick and consisted of a series of rooms interconnected by corridors and doorlike openings. Tests were carried out at full scale and at one-third and one-sixth scale; in each experiment a high-explosive charge (representing the penetrated weapon) was detonated in one room, and blast pressures and response measurements were obtained in adjacent rooms. For this series of tests, the first-principle codes FEFLO, HULL, and SHARC (Hikida et al., 1988) were used for both blast propagation and structural response analyses. Initially, all calculations were two dimensional in that only the plane of the test structure was modeled. The results of the computations exhibited subtle differences, which were slight in the source chamber but showed substantial

OCR for page 26
Page 42 pressure differences in adjoining chambers. This result was attributed to the large effect of turbulence on the transmitted pressures triggered at the doorways. Subsequently, full three-dimensional calculations were made with both first-order (HULL) and second-order (FEFLO) programs. The former, without adaptivity2, underpredicted the pressures in the adjoining chambers, but did exhibit the delay in the pressure spikes observed in the experiments; the FEFLO calculation compared quite well with the measurements. The response of the structures was determined in an uncoupled manner with FLEX, DYNA3D, and TRT computer programs, using computed pressures as a loading on the structures. The predictions of failure did not match the experiment in any of the analyses. The discrepancy was ascribed to the difficulties in modeling the tensile failure of the concrete, an area of modeling that is still not well in hand. Recent tests performed by DNA on shallow-buried structures to validate an application of the DYNA3D and FLEX programs provide an example of the accuracy with which the deformation of a structure can be predicted when the structure remains homogeneous (i.e., does not encounter any tearing, fracture, or fragmentation). The tests were intended to study the interaction between the structure and the surrounding soil for different types of soil. A cylindrical structure was tested in both clay and sand. The cylinder and the surrounding soil were modeled in detail, and the pressures measured on the surface of the soil during the experiment were applied to the top surface of the model. The computations were made prior to the release of experimental response data, so there was no opportunity to "tune" the analysis. A comparison of the computed shapes of the cylinders in the sand and clay blackfills are shown in Figure 3-1. The deformation can be seen to depend markedly on the surrounding medium; the deformations in clay are much larger and exhibit a different pattern of deformation. The computation accurately predicted the difference in the response in soil and clay and the magnitude of the deformation, illustrating the predictive capabilities of first-principle programs in certain instances. Some programs, under certain conditions, have produced results in excellent agreement with observations. In other cases, the differences between observed and predicted results have been great. This points to the need not only for ongoing testing but also for care in the selection of an appropriate model and input parameters and equal care in the interpretation of the results. Once a model has been validated over an experimental range, its real value lies in the richness of the predictive domain thus established. For example, the effects of attributes such as various window shapes, geometrical configurations, percentage of reinforcement, 2Adaptivity refers to the ability of a simulation model to initiate modifications to the analytic procedure(s) during program execution based on the model's "sensing" the results of the ongoing analysis.

OCR for page 26
Page 43 TRANAL PREDICTION-CLAY TRANAL PREDICTION-SAND OBSERVED RESPONSE-CLAY OBSERVED RESPONSE-SAND Figure 3-1 Comparison of predicted and observed deformation of a buried structure inclay and sand backfills. and other variables can be studied by computer programs once they are validated. The designers of civilian structures can make use of these validated results in the quest for improved building performance. Applications of Computational Methods to Terrorist Threats Several of the programs described above have already been applied to the evaluation of terrorist threats and events. The FEFLO code was used to analyze the explosion in the World Trade Center (Baum et al., 1994). This computation required 10,000,000 equations and approximately 150 hours on the Cray C-90. The model of the geometry of the area affected by the explosion was extremely detailed; even the cars in the parking lot were modeled. Nonetheless, in the later stages of the calculation, after the shock passes and turbulence develops, even this detailed a mesh was found to be inadequate for resolving the turbulent flow field. This calculation overestimated the blast loading because it did not account

OCR for page 26
Page 44 for the failure of the floors, the modeling of which would require coupling the program with a structural response program. Several other analyses of the World Trade Center bombing were done with different suites of codes. One analysis using the CTH, BLASTX, and FLEX programs focused on interpreting explosive effects on the floor and roof slabs and failure patterns of the floor slabs. The analyses coupled the blast prediction with the floor response and showed that the principal mechanisms that limited the damage were due to the decrease in pressure and impulse that resulted from the progressive venting of the gas pressure and the large displacement of the slabs. Good correlations were obtained between computed and observed damage patterns in floor and ceiling slabs. Another analysis used BLASTX and some simple computer models of the floor slabs, but BLASTX proved inadequate for this class of problems because the long spanwise dimensions of the parking floors led to significant overpredictions of the impulse. The CTH, BLASTX, and FLEX suite of codes was also used to develop guidelines for the protection of U.S. embassies against explosions from terrorist attacks (Clemens and Watt, 1987; Nelson and Watt, 1989). Tests were conducted with a one-third scale model to validate the computational results. Both internal and external explosions were considered. The predicted peak displacements of the exterior wall were within 15 percent of the measured values, but the final displacements differed by almost a factor of two. Nonetheless, the analyses were used to develop guidelines for structural design, including layout and anchorage of beams and slabs, and recommendations for mitigating progressive collapse due to detonation of explosives hand-carried into the building. The CTH and FLEX codes were also used by Weidlinger Associates to perform a vulnerability study of a hotel due to a car bomb explosion in the JFK Airport east parking garage, and to develop protection schemes for the Rockefeller Center control room against the threat of briefcase bombs (Baron and Hinman, 1994). The FLEX and FUSE (Sandler and Rubin, 1990) codes were used to develop procedures for blast-hardening areas of a federal court building to explosions from package bombs; a similar study was made for the World Trade Center control center prior to the bombing. Summary Observations Over the past 50 years, design procedures have been developed for structures subjected to explosive blast loads. These procedures are based largely on synthesis of test data and simplified computational models and apply for the most part to the type of hardened structures found in military construction. These design procedures are codified in a variety of manuals and computer programs. While this body of knowledge can serve as a foundation for designing civilian buildings to be more blast-resistant, substantial effort will be required to make the methods more directly applicable and useful for civilian design professionals.

OCR for page 26
Page 45 The available computer programs for prediction of blast-effects and structural response have been described and classified in several ways. From the viewpoint of a user, the most important classification is whether the programs are semi-empirical or first-principle in nature. First-principle codes are more generally applicable, but they require the user to be well versed in structural dynamics and explosive behavior. For both classes of computer programs, validation by tests is imperative. To date, very little test data have been obtained for structures representative of civilian building design and construction practice. Application of these programs to blast-resistant design strategies in civilian buildings is of limited usefulness without appropriate experimental validation. Extending this technology to civilian design and construction practice affords a valuable opportunity to both solidify and advance the state of knowledge in this field. There are difficulties in understanding and mathematically modeling structural behavior in transition regions of response from predominately flexural behavior into domains dominated by boundary or punching shear, and ultimately to material disintegration. Also, material constitutive relationships are less well understood in these transition regions. Nevertheless, computations can give valuable information about the magnitude and type of damage. The pressures resulting from complex, nonspherical explosive charges (e.g., car bombs) are not well understood and carefully controlled experiments are vital for a better understanding and validation of computer programs. Wherever possible, tests of civilian buildings and component types should be conducted at sufficiently large scale to allow the use of actual design details, materials, and construction practice. Despite some success in re-creating the observed effects of actual bombings and the cited examples where numerical codes were applied to specific design problems, it is not clear that the routine application of these programs to civilian buildings will become widespread. The cost and complexity of the analysis, coupled with uncertainty regarding threat levels and the low probability that a specific building will actually become a target, all suggest that they will not. However, where these programs could prove very valuable is in testing a wide range of building types and structural details over a broad range of hypothetical explosion events. The knowledge gained from such testing, verified by experimentation, could transfer directly to civilian practice through manuals and other design aids, and ultimately into building codes, in much the same way as the application of seismic design principles has become routine. References Achenbach, J.D. 1973. Wave Propagation in Elastic Solids. Amsterdam. Netherlands: North-Holland Publishing Company. Air Force Engineering and Services Center. 1989. Protective Construction Design Manual. ESL-TR-87-57. Engineering & Services Laboratory, Tyndall Air Force Base, Florida.

OCR for page 26
Page 46 American Society of Mechanical Engineers. 1993. Computational Methods for Material Modelling. Benson and Asaro, eds.ASME Book No. H-00883. New York: American Society of Mechanical Engineers. ASCE (American Society of Civil Engineers). 1985. Design of Structures to Resist Nuclear Weapons Effects. Manual 42. Washington, D.C.: American Society of Civil Engineers. ASCE (American Society of Civil Engineers). 1995. Structural Design for Physical Security-State of the Practice Report. Washington, D.C.: American Society of Civil Engineers, in press. Atkatsh, R.S., K.K. Chan, K.G. Stultz, Jr., R. Dillworth. and M.L. Maranda. 1994. EPSA-II Theoretical Guidebook. Revision G. New York: Weidlinger Associates. Baker, W.E., P.A. Cox, P.S. Westine, J.J. Kulesz, and R.A. Strehlow. 1983. Explosion Hazards and Evaluation. New York: Elsevier. Bangash, M.Y.H. 1993. Impact and Explosion—Analysis and Design. Boca Raton, Florida: CRC Press. Baron, M.L., and E. Hinman. 1994. Technology Transfer for Anti-Terrorism Design, Some Current Examples. Presentation to the Committee on Feasibility of Applying Blast-Mitigating Technologies and Design Methodologies from Military Facilities to Civilian Buildings, National Research Council. New York: Weidlinger Associates. Batsanov, S.S. 1994. Effects of Explosions on Materials. New York: Springer-Verlag. Baum, J.D., H. Luo, and R. Löhner. 1994. Numerical Simulation of Blast in the World Trade Center. Paper presented at the American Institute of Aeronautics and Astronautics 33rd Aerospace Sciences Meeting, Reno, Nevada, January. Britt, J.R., and M.G. Lumsden. 1994. Internal Blast and Thermal Environment From Internal and External Explosions: A User's Guide for the BLASTX Code, Version 3.0. Vicksburg, Mississippi: U.S.A.C.E. Waterways Experiment Station. Budge, K.G., and J.S. Peery. 1993. R Hale: AMMALE shock physics code written in C++. International Journal of Impact Engineering 14:107–120. Cicolani, A. 1994. Balanced Survivability Assessment. Briefing material provided to the Committee on Feasibility of Applying Blast-Mitigating Technologies and Design Methodologies From Military Facilities to Civilian Buildings. National Research Council. Defense Nuclear Agency. Washington, D.C. Clemens, G.P., and J.M. Watt, Jr. 1987. Design of Structures to Resist Terrorist Attack, Report 3, 1/3-Scale Model Exterior Wall Test, and Exterior Bay Roof and Floor Slab Test. TR-SL-87-13. Washington, D.C.: U.S. Department of State. Crawford, E.R., C.J. Higgins, and E.H. Bultmann. 1974. The Air Force Manual for Design and Analysis of Hardened Structures. AFWL-TR-74-102. Kirtland Air Force Base. New Mexico: Air Force Weapons Laboratory. Davis, J.L. 1988. Wave Propagation in Solids and Fluids. New York: Springer-Verlag. Dick, R.A., L.R. Fletcher, and D.V. D'Andrea. 1983. Explosives and Blasting Procedures Manual. Information circular 8925. U.S. Department of the Interior. Washington, D.C.: NTIS. DNA (Defense Nuclear Agency). 1995. The Design and Analysis of Hardened Structures to Conventional Weapons Effects. DAHS CWE. Defense Nuclear Agency. in preparation. Drake, J.L., L.A. Twisdale, R.A. Frank. W.C. Dass, M.A. Rochefort. R.E. Walker, J.R. Britt. C.E. Murphy, T.R. Slawson, and R.H. Sues. 1989. Protective Construction Design Manual. ESL-TR-87-57. Air Force Engineering & Services Center, Tyndall Air Force Base. Florida. Eytan, R., and A. Kolodkin. 1993. Practical strengthening measures for existing structures to increase their blast resistance: walls and ceilings. Proceedings of the Sixth International Symposium on Interaction of Nonnuclear Munitions with Structures. Panama City, Florida. May. Fickett, W. 1985. Introduction to Detonation Theory. Berkeley: University of California Press. Fouks, Y. 1993. Full scale tests on strengthening measures for existing structures. Proceedings of the Sixth International Symposium on Interaction of Nonnuclear Munitions with Structures. Panama City. Florida, May.

OCR for page 26
Page 47 Glasstone, S., and P. Dolan. 1977. The Effects of Nuclear Weapons, 3rd ed. Washington, D.C.: U.S. Department of Defense and U.S. Department of Energy. Gunger, M. 1992. Progress on Tasks Under the Sympathetic Detonation Program. WL/MN-TR-91-85. Shalimar, Florida: Orlando Technology, Inc. Han, Z., and X. Yin. 1993. Shock Dynamics. Boston. Massachusetts: Kluwer Academic Publishers. Henrych, J. 1979. The Dynamics of Explosions and Its Use. New York: Elsevier. Hikida, S., R. Bell, and C. Needham. 1988. The SHARC Codes: Documentation and Sample Problems. SSS-R-89-9878. Albuquerque, New Mexico: S-Cubed division of Maxwell Laboratories. Distribution limited. Johansson, C.H., and P.A. Persson. 1970. Detonics of High Explosives. New York: Academic Press. Jones, N. 1989. Structural Impact. Cambridge, England: Cambridge University Press. Kolodkin, A., and R. Eytan. 1993. Practical strengthening measures for existing structures to increase their blast resistance: windows and doors. Proceedings of the Sixth International Symposium on Interaction of Nonnuclear Munitions with Structure. Panama City, Florida. May. Löliner, R., C. Yang, J. Cebral, J.D. Baum, H. Luo, D. Pelessone, and C. Charman. 1995. Fluid-Structure Interaction Using a Loose Coupling Algorithm and Adaptive Unstructured Grids. Paper presented at the 26th American Institute of Aeronautics and Astronautics Fluid Dynamics Conference, San Diego. California. June. McGlaun, J.M., S.L. Thompson, and M.G. Elrick, 1990. CTH: a three-dimensional shock physics code. International Journal of Impact Engineering 10:351–360. McGraw-Hill Encyclopedia of Science & Technology. 6th ed. 1987. Explosion and Explosives 6:521–528. National Defense Research Committee. 1946. Effects of Impact and Explosions. Office of Scientific Research and Development. Washington, D.C.: National Defense Research Committee. Naval Civil Engineering Laboratory. 1988. Terrorist Vehicle Bomb Survivability Manual. Port Hueneme. California: Naval Civil Engineering Laboratory. Nelson, D.H., and J.M. Watt, Jr. 1989. Design of Structures to Resist Terrorist Attack, Report 6. Response of an Embassy Lobby to an Internal Bomb. TR SL-87-13. Washington, D.C.: U.S. Department of State. Newmark, N.M.. and J.D. Haltiwanger. 1962. Air Force Design Manual. AFSWC-TDR-62-138. Kirtland Air Force Base. New Mexico: Air Force Special Weapons Center. NRC (National Research Council). 1986. The Embassy of the Future: Recommendations for the Design of Future U.S. Embassy Buildings. Building Research Board. Washington, D.C.: National Academy Press. NRC (National Research Council). 1988. The Protection of Federal Office Buildings Against Terrorism. Building Research Board. Washington, D.C.: National Academy Press. SAIC (Science Applications International Corporation). 1994. Simulation-Based Design. Presentation to the Committee on Feasibility of Applying Blast-Mitigating Technologies and Design Methodologies from Military Facilities to Civilian Buildings. National Research Council. Arlington, Virginia. Sandler, I.S., and D. Rubin. 1990. FUSE Calculations of Far-Field Water Shock Including Surface and Bottom Effects. Technical report for SAIC subcontract No. 29-90050-64. New York: Weidlinger Associates. Distribution limited to SAIC. Schlamp, R.J. P.J, Hassig, C.T. Nguyen, D.W. Hatfield, P.A. Hookham, and M. Rosenblatt. 1995. MAZe User's Manual. Los Angeles. California: TRT Corporation. Schuster, S.H., F. Sauer, and A.V. Cooper. 1987. The Air Force Manual for Design and Analysis of Hardened Structures. AFWL-TR-87-57. Kirtland Air Force Base. New Mexico: Air Force Weapons Laboratory. U.S. Department of the Army. 1986. Fundamentals of Protective Design for Conventional Weapons. TM 5-855-1. Washington, D.C.: Headquarters, U.S. Department of the Army.

OCR for page 26
Page 48 U.S. Department of the Army. 1969. Structures to Resist the Effects of Accidental Explosions. Army TM 5-1300, Navy NAVFAC P-397, AFR 88-22. Washington, D.C.: Departments of the Army. Navy, and Air Force. U.S. Department of the Army. 1990. Structures to Resist the Effects of Accidental Explosions. Army TM 5-1300, Navy NAVFAC P-397, AFR 88-22. Washington, D.C.: Departments of the Army, Navy, and Air Force. U.S. Department of the Army. 1993. Security Engineering. TM-5-853. Washington, D.C.: Headquarters, U.S. Department of the Army. U.S. Department of Energy. 1992. A Manual for the Prediction of Blast and Fragment Loadings on Structures. DOE/TIC-11268. Washington, D.C.: U.S. Department of Energy. Whirley, R.G., and B.E. Engelmann. 1993. DYNA3D: A nonlinear, explicit, three-dimensional finite element code for solid and structural mechanics user manual. Revision 1. UCRL-MA-107254-Rev. 1. Livermore. California: Lawrence Livermore National Laboratory. Whitham, G.B. 1974. Linear and Nonlinear Waves. New York: Wiley. Windham, J.E., H.D. Zimmerman, and R.E. Walker. 1993. Improved ground shock predictions for fully buried conventional weapons. Proceedings of Special Session of the Sixth International Symposium on the Interaction of Conventional Munitions with Protective Structures, Panama City. Florida, May. Zyukas, L. A., 1990. High Velocity Impact Dynamics. New York: Wiley Interscience.