Design Guidance for Blast-Resistant Glazing

Abstract

This paper reviews current design standards and test methods for blast-resistant glazing design and compares a typical design outcome with that from comprehensive finite-element (FE) analysis. Design standards are conservative and are limited to the design of relatively small glazed panels. Standard test methods are expensive, create environmental pollution, and can classify the hazard ratings of only smaller glazed panels. Here the design of a laminated glass (LG) panel is carried out according to an existing design standard, and then its performance is examined using comprehensive FE modeling and analysis. Finite-element results indicate that both glass panes crack, the interlayer yields with little damage, and the sealant joints do not fail for the designed blast load. This failure pattern satisfies some of the requirements for minimal hazard rating in the design standard. It is evident that interlayer thickness and material properties are important during the post-crack stage of an LG panel, but they are not accounted for in the design standards. The new information generated in this paper will contribute toward an enhanced blast design of LG panels.

Full text

This is the author’s version of a work that was submitted/accepted for pub-
lication in the following source:
Hidallana-Gamage, H., Thambiratnam, D., & Perera, N.
(2015)
Design Guidance for Blast-Resistant Glazing.
Journal of Architectural Engineering, Journal of Architectural Engineering,
21(3).
This file was downloaded from:
Notice:Changes introduced as a result of publishing processes such as
copy-editing and formatting may not be reflected in this document. For a
definitive version of this work, please refer to the published source:
http://doi.org/10.1061/(ASCE)AE.1943-5568.0000161

Design Guidance for Blast-Resistant Glazing
Hasitha D. Hidallana-Gamage1; David P. Thambiratnam, F.ASCE2; and Nimal J. Perera3
Abstract: This paper reviews current design standards and test methods for blast-resistant glazing design and compares a typical design
outcome with that from comprehensive finite-element (FE) analysis. Design standards are conservative and are limited to the design of
relatively small glazed panels. Standard test methods are expensive, create environmental pollution, and can classify the hazard ratings of
only smaller glazed panels. Here the design of a laminated glass (LG) panel is carried out according to an existing design standard, and
then its performance is examined using comprehensive FE modeling and analysis. Finite-element results indicate that both glass panes
crack, the interlayer yields with little damage, and the sealant joints do not fail for the designed blast load. This failure pattern satisfies
some of the requirements for minimal hazard rating in the design standard. It is evident that interlayer thickness and material properties are
important during the post-crack stage of an LG panel, but they are not accounted for in the design standards. The new information
generated in this paper will contribute toward an enhanced blast design of LG panels. DOI: 10.1061/(ASCE)AE.1943-5568.0000161.
© 2015 American Society of Civil Engineers.
Author keywords: Design standards; Test methods; Laminated glass; Blast loads; Finite-element modeling; Interlayer properties.
Introduction
Designing building facades to withstand blast loads has become a
major concern because of ever-increasing terrorist attacks. Glazed
facades are preferred in buildings by engineers and architects
because of their architectural features and aesthetical aspects. Most
buildings have glazed facades 4–10 m high without any
structural framework in ground floor lobby areas. These lower
levels are the most vulnerable to blast events; more than 80–90%
of blast-related injuries are due to flying glazed fragments and
facade pieces. If building facades disintegrate, the direct blast
pressure entering the building can cause injuries to occupants and
even structural collapse. Blast-resistant glazing should therefore
be used in buildings to minimize, if not eliminate, the hazard from
potential terrorist attacks.
Laminated glass (LG) consists of two or more glass plies
permanently bonded with one or more polymer interlayers. It
has superior blast resistance compared with monolithic glass
and is hence mostly used in blast-resistant glazing. The major
advantage of LG is that even if the glass cracks, the interlayer
holds the glass fragments instead of forming free-flying shards.
Upon fracture, annealed and heat-strengthened glass produce
large shards that adhere well to the interlayer and thus reduce the
amount of flying and falling glass shards. The use of annealed or
heat-strengthened glass types in LG, instead of fully tempered
glass, has thus been recommended (Norville and Conrath 2001).
Polyvinyl butyral (PVB) is commonly used as the interlayer
material in LG glass, but some stiffer interlayer materials such as
ionoplast are also used in practice (Ledbetter et al. 2006).
Laminated glass panels are fixed to window frames by means of
structural sealant joints, where silicone and rubber are common
sealant materials.
This paper reviews the latest design standards and docu-
ments used in blast-resistant glazing design, such as ASTM
F2248-09 (ASTM 2010a), Unified Facilities Criteria (UFC)
4-010-01 (DoD 2013), UFC 3-340-02 (DoD 2008), U.K.
Glazing Hazard Guide (1997), and Protective Design Center—
Technical Report (PDC-TR 10-02) (2012). Standard test methods
such as ASTM F1642-04 (ASTM 2010b), ISO 16933 (ISO
2007a), and ISO 16934 (ISO 2007b) are also reviewed in the
paper. Shortcomings and limitations of those design standards and
test methods are briefly discussed. The authors have developed
and validated a rigorous numerical procedure using LS-DYNA
(Hallquist 2006)finite-element (FE) code to study the blast
response of LG panels. This paper extends their previous research
work to the application of their modeling techniques to the
analysis of LG panels under blast loads. The comprehensive
information provided through such analysis will not only enhance
the understanding of the blast response of LG panels but also
facilitate their design.
The design of an LG panel is first carried out according to
ASTM F2248-09 (ASTM 2010a), and its performance is examined
using an FE model. The results of FE analysis are used to examine
the failure of glass, interlayer, and sealant joints. Finite-element
predictions are used to examine whether the LG panel has
achieved the desired level of protection according to ASTM
F2248-09 (ASTM 2010a). The energy absorption of different
components is studied and the importance of the interlayer
properties is highlighted, because they are not accounted for in the
current design standards. In practice, engineers do blast testing to
check their designs that they have carried out according to design
standards. The modeling techniques presented in this paper may be
used to complement and supplement existing standards for the
design of LG panels, where applicable, and also as a solution when
they are not applicable, thereby reducing the costs, risks, and
environmental pollution involved with blast testing.
1Ph.D. Student, Faculty of Science & Engineering, Queensland Univ.
Technology, GPO Box 2434, 2 George St., Brisbane, QLD 4001,
Australia (corresponding author). E-mail: hasithagamage@yahoo.com
2Professor, Faculty of Science & Engineering, Queensland Univ. of
Technology, GPO Box 2434, 2 George St., Brisbane, QLD 4001,
Australia. E-mail: d.thambiratnam@qut.edu.au
3Adjunct Professor, Faculty of Science & Engineering, Queensland
Univ. of Technology, GPO Box 2434, 2 George St., Brisbane, QLD 4001,
Australia. E-mail: nimal.perera@robertbird.com.au
Note. This manuscript was submitted on February 13, 2014; approved
on August 8, 2014; published online on April 2, 2015. Discussion period
open until September 2, 2015; separate discussions must be submitted for
individual papers. This paper is part of the Journal of Architectural
Engineering, © ASCE, ISSN 1076-0431/04015003(13)/$25.00.
© ASCE 04015003-1 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

Design Standards for Blast-Resistant Glazing
ASTM F2248-09 (ASTM 2010a), UFC 4-010-01 (DoD 2013),
UFC 3-340-02 (DoD 2008), U.K. Glazing Hazard Guide (1997),
and PDC-TR 10-02 (2012) are the latest standards and documents
used in blast-resistant glazing design. Existing design standards
and their limitations are reviewed in the following paragraphs.
ASTM F2248-09
ASTM F2248-09 (ASTM 2010a) provides a framework for
designing blast-resistant glazing using either single LG or insula-
ted glass fabricated with LG. This standard recommends using
either annealed or heat-strengthened glass types for the glass panes
rather than fully tempered glass, which has shown poor post-blast
performance during blast testing. Even though different interlayer
materials are used in practice, the information provided in ASTM
F2248-09 (ASTM 2010a) applies only to LG fabricated with a
PVB interlayer. For a given charge weight and standoff distance,
the 3-s-duration-equivalent design load should be selected from
the chart shown in Fig. 1, which is given in this standard. This
chart was developed using the results from a number of blast tests
conducted on LG panels for hemispherical charge weights at
ground level.
After determining the 3-s-duration-equivalent design load, a
relevant chart as shown in Fig. 2should be selected from ASTM
E1300-09a (ASTM 2009) to obtain the thickness of the LG.
ASTM F2248-09 (ASTM 2010a) recommends using either
structural silicone sealant or an adhesive glazing tape to fix the
glazing to the supporting frame. The width (bite) of the structural
silicone sealant bed should be at least equal to or greater than
10 mm or the nominal thickness of the glass panes but less than
twice the nominal thickness of the glass panes to which it adheres.
The minimum thickness of the structural silicone sealant bed
should be 5 mm. The glazing tape should be within two to four
times the thickness of the glass pane.
Framing members are designed to withstand a load twice the
load resistance of the attached glazing, and the edge deflection of
the glazing should be less than L/60 (L=length of the supported
edge). The framing system supporting the glazing should be
attached mechanically to the structural framing system by means
of fasteners that should be designed to resist a uniform load acting
on the glazing. The design load of the fasteners should be two
times the magnitude of the load resistance of the glazing if the
maximum air blast pressure is greater than one-half the magni-
tude of the load resistance of the glazing. On the other hand, the
fasteners should be designed for a load equal to the load resistance
of the glazing if the maximum air blast pressure is less than
Fig. 1. Graphical relationship between standoff distance, TNT charge mass, and 3-s-equivalent design load (reprinted, with permission, from
ASTM 2010a, F 2248-09, “Standard practice for specifying an equivalent 3-second duration design loading for blast resistant glazing fabricated
with laminated glass,”copyright ASTM International, 100 Barr Harbor Drive, West Conshohocken, PA 19428. A copy of the complete standard
may be obtained from ASTM International, www.astm.org.)
Fig. 2. Non-factored load chart for 6.0 mm (0.25″) glass with four
sides simply supported (reprinted, with permission, from ASTM 2009,
E 1300-09a, “Standard practice for determining load resistance of
glass in buildings,”copyright ASTM International, 100 Barr Harbor
Drive, West Conshohocken, PA 19428. A copy of the complete
standard may be obtained from ASTM International, www.astm.org.)
© ASCE 04015003-2 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

one-half the magnitude of the load resistance of the glazing. The
guidelines given in ASTM F2248-09 (ASTM 2010a) ensure that
blast-resistant glazing fails by tearing of the interlayer rather than
failure at the supports. Blast-resistant glazing designed with this
standard performs to minimal hazard as defined in ASTM F1642-04
(ASTM 2010b). When an LG panel fails under a minimal hazard,
it is expected to fracture but should remain in the frame with-
out any failure at the sealant joints and the supportive frame. The
design guidelines in ASTM F2248-09 (ASTM 2010a) have been
explained in detail by Norville and Conrath (2006).
However, ASTM F2248-09 (ASTM 2010a) has some
limitations with respect to designing blast-resistant glazing, and
these are explained in this paper. The chart in Fig. 1can be used to
calculate the 3-s-duration design loads only for charge weights in
the range 4.5–910 kg TNT and for standoff distances in the range
6–120 m. The design charts available in ASTM E1300-09a
(ASTM 2009) were developed only for LG panels having PVB as
the interlayer material without accounting for the thickness of the
interlayer. Therefore, with respect to the blast response of LG
panels, this standard accounts for neither the effects due to vari-
ations in the thickness of the interlayer nor the effects of different
interlayer materials with varied material properties. The charts
available in ASTM E1300-09a (ASTM 2009) can be used to
design only LG panels having a maximum length of approxi-
mately 5 m and width of approximately 4 m. A conservative
design approach based on static analysis is used to design window
frames, fasteners, and other supporting elements.
UFC Standards
The U.S. Department of Defense (DoD), has developed UFC stan-
dards that are applicable to the design of blast-resistant windows.
The latest versions of two of the UFC standards, UFC 4-010-01
(DoD 2013) and UFC 3-340-02 (DoD 2008), are reviewed in the
following paragraphs. The former describes an approach to design-
ing blast-resistant windows basically with LG, whereas the latter
provides a design approach using monolithic fully tempered glass.
UFC 4-010-01
The latest version of UFC 4-010-01 (DoD 2013), published in
October 2013, supersedes its previous versions published in 2003,
2007, and 2012. UFC 4-010-01 (DoD 2013)defines different
levels of protections known as below antiterrorism standards, very
low, low, medium, and high, which correspond to high hazard,
low hazard, very low hazard, minimal hazard, and no hazard, res-
pectively, according to the glazing hazard ratings defined in
ASTM F1642-04 (ASTM 2010b). Two baseline explosives
(Explosive Weights I and II) are defined in the standard, and
their magnitudes are not mentioned publicly for security reasons.
Department of Defense buildings are divided into different
categories wherein minimum and conventional construction stand-
off distances are given for each building category (refer to Table
B-1 in UFC 4-010-01). Conventional construction standoff dis-
tance implies the minimum standoff distance required by a DoD
building to achieve either a very low or a low level of protection
without any measures for blast resistance. Buildings must at least
satisfy the minimum standoff distance requirement, and those that
do not meet conventional construction standoff distances or that
require a higher level of protection should be designed for the
potential blast threat at the available standoff distance.
According to UFC 4-010-01 (DoD 2013), windows and
skylights in buildings that require blast resistance must be fabri-
cated with LG, and they may be designed for a credible blast load
by dynamic analysis, testing, or the approach given in ASTM
F2248-09 (ASTM 2010a). Dynamic analysis may be conducted by
using computer programs as described in PDC-TR 10-02 (2012).
Blast testing should be conducted according to ASTM F1642-04
(ASTM 2010a), which described later in this paper. The design
approach given in ASTM F2248-09 (ASTM 2010a) ensures
a medium level of protection according to UFC 4-010-01 (DoD
2013) (minimal hazard according to ASTM F1642-04), which
described earlier in this paper. The minimum interlayer thickness
should be 0.76 mm, and the design of the glass pane thickness,
structural sealant joints, frames, and fasteners is carried out
according to ASTM F2248-09 (ASTM 2010a).
In addition to LG, polycarbonate windows may also be used in
blast-resistant windows, in which case the frame bite (width of
the structural sealant joints) should be no less than 1.5 times the
polycarbonate thickness. The design should be carried out using 1
for both load and strength reduction factors for all methods of
analysis referenced in UFC 4-010-01 (DoD 2013).
UFC 3-340-02
UFC 3-340-02 (DoD 2008) provides a framework for designing
glazed facades with monolithic fully tempered glass to withstand
blast loads. The blast load is treated as a triangular load, and a
simplified single degree of freedom (SDOF) model is used to
simulate the dynamic response of the glass panels. The glass
panels are analyzed using large-deflection plate theory because the
panel deflections are large compared with the thickness of the
panel. The maximum allowable principal tensile stress of glazing
is used as 16,000 psi (110 MPa) in the standard. Design deflection
is the center deflection that corresponds to the maximum principal
tensile stress at any point in the glass panel.
UFC 3-340-02 (DoD 2008) provides several charts as shown in
Fig. 3to determine the required glass pane thickness for a given
blast overpressure and positive load duration. The charts were
developed for fully tempered glass panels having different aspect
ratios between 1 and 4 and glass thicknesses of 1 /4″(6.35 mm),
5/16″(7.94 mm), 3 /8″(9.53 mm), 1/2″(12.7 mm), 5/8″(15.88
mm), and 3/4″(19.05 mm). In addition to charts, a set of formulas
is given in the standard for the design of blast-resistant glazing.
Framing members should be designed for the load transferred from
the glass panel, and the static design blast load should be applied
to all exposed members. The relative displacement of the framing
members is limited to 1 /264 of its span or 1 /8″(3.18 mm), which-
Fig. 3. Peak blast pressure capacity for tempered glass panes with an
aspect ratio (a/b) of 1 and a thickness of 6 mm (0.25″) (image courtesy
of DoD 2008)
© ASCE 04015003-3 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

ever is smaller. The maximum stresses in the framing members
and fasteners are limited to fm/1:65 and fm/2, respectively, where
fm=yield strength of the frame material.
The limitations of the UFC 3-340-02 (DoD 2008)standardare
briefly described here. This standard uses a simplified SDOF
analysis method to study the blast response of glazed panels by
accounting for the positive phase of the blast load only. The design
charts developed in this standard are applicable to monolithic fully
tempered glass only. The maximum length and width of the glass
panels that could be designed with UFC 3-340-02 (DoD 2008)are
limited to approximately 3 and 1.5 m, respectively. However, it
should be noted that the generalized analysis and design method
given in this standard can be applied to the design of LG or any
glazing type with different sizes if the corresponding load–resistance
curve is determined from an analytical or experimental study.
U.K. Glazing Hazard Guide
The U.K. Glazing Hazard Guide (1997) provides a more realistic
approach for designing glazed facades with LG panels by
accounting for both their pre-crack and post-crack behavior
under blast loads. This guide idealizes an LG panel as an SDOF
system and undertakes a time–history analysis for a given blast
threat. The pre-crack resistance function is derived using large
deflection plate theory by considering the dynamic breaking
strength of glass. The dynamic breaking strengths used in the
design are 80 MPa for annealed glass and 200 MPa for fully
tempered glass. The post-crack resistance function is derived
considering the membrane action of the PVB interlayer but
neglecting the stiffness of the cracked glass panes. On the basis of
the extensive blast tests conducted for common window sizes used
in the United Kingdom (about 1:25 × 1:55 m), it has been shown
that an approximately 200-mm central deflection will cause the
limit of tearing in the PVB interlayer.
The U.K. Glazing Hazard Guide (1997) provides a set of
diagrams called pressure–impulse diagrams (P–I diagrams) for
common window sizes used in the United Kingdom that can be
used to evaluate their performance under a given blast loading. Fig.
4shows the P–I diagram for a typical LG panel (Smith 2001). Each
contour line in the diagram connects P–I pairs giving the same
deflection and stress, and those are called iso-damage lines. The
lower contour line represents the P–I pairs causing initial cracking
of the glass pane, whereas the upper contour line represents P–I
pairs causing tearing of the PVB interlayer. Known blast threats
can be marked on the P–I diagram of a selected LG panel to
estimate its performance under the blast threat. The panel edges
should be securely held in robust frames by using structural
silicone sealant with a width (bite) of approximately 35 mm.
Support reactions can be obtained using equivalent SDOF factors
for two-way spanning simply supported panes with a uniform load.
This guide has some limitations as briefly described here. It is
restricted to military use, giving limited access to external users.
The authors could not find a copy of this document, so limited
information is given in this paper. The U.K. Glazing Hazard Guide
(1997) is limited to a few window sizes used in the United
Kingdom and is therefore of limited application in designing blast-
resistant glazing for real buildings.
PDC-TR 10-02
PDC-TR 10-02 (2012) presents engineering guidelines and cost-
effective solutions for the design of window systems to reduce
their fragment hazards from blast loads. This report describes two
design approaches, known as the static and the dynamic approach,
that may be used to design single glazing units or insulated glazing
units fabricated with LG to withstand blast loads. The static design
approach is the same approach presented in ASTM F2248-09
(ASTM 2010a), which is used in conjunction with ASTM E1300-
09a (ASTM 2009) to select an appropriate glass type and
thickness. This approach is described earlier in this paper and is
also described in detail in PDC-TR 10-02 (2012) with some
working examples.
The dynamic analysis and design of blast-resistant glazing may
be carried out by using the available FE codes or computer pro-
grams recognized by the blast community. The authors prefer
the use of FE codes when analyzing and designing glazing under
blast loads, and their approach is described later. However,
PDC-TR 10-02 (2012) provides some useful information about
computer programs and their applications for designing window
systems under blast loads. SBEDS-W and WinGARD are two such
computer programs: SBEDS-W is available from the PDC, and
WinGARD is available from the Whole Building Design Guide.
These programs are based on SDOF analysis, which means that
their approach is an iterative process of selecting the initial glazing
or member size and then repeating the analysis until the window
system is found to have an acceptable response. The dynamic design
procedure based on the SBEDS-W computer program is described in
detail in PDC-TR 10-02 (2012).
Computer programs used in blast-resistant glazing design have
some limitations as described here. One of the major limitations is
that the design outcome will be extremely conservative because it
is based on simplified SDOF analysis. At the same time, com-
prehensive knowledge and understanding of the computer program
is required to achieve a feasible and safe design. In addition these
programs generate output results in numbers, in contrast with the
FE codes, which makes it possible to observe the predicted re-
sponse and the failure pattern.
Standard Test Methods for Blast-Resistant Glazing
Standard test methods provide guidelines for classifying the hazard
rating of glazed panels depending on their performance under blast
loads. These test methods can be classified into two types: arena air
blast test and shock tube test. An arena air blast test is carried out in
an open environment and is expensive compared with the shock
tube test, but it tests several test panels simultaneously. A shock
tube test is carried out in a closed tube and is not a realistic test but
is capable of reproducing the same shock repeatedly. Standards
formulated by ASTM and ISO are available for both test types, and
they are explained in the following paragraphs.
Reflected Pressure
Reflected Impulse
Initial crack of glass
Tearing of the interlayer
Blast threat
Fig. 4. P–I diagram with iso-damage curves for a typical LG panel
(adapted from Smith 2001)
© ASCE 04015003-4 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

ASTM F1642-04
The ASTM F1642-04 (ASTM 2010b) test method provides a
structured procedure for testing and rating all glazing, glazing
systems, and glazing retrofit systems, including but not limited
to those fabricated from glass, plastic, glass-clad plastics, LG,
glass/plastic glazing materials, and film-backed glass. The hazard
rating of a glazing system is determined on the basis of the severity
of fragments generated during blast testing. The severity of the
fragments is determined by considering the number, size, and
location of the fragments observed after the test. A fragment is
defined as any particle having a united dimension of 2.5 cm or
greater that is calculated by adding its width, length, and thickness.
ASTM F1642-04 (ASTM 2010b) provides six rating criteria
known as no break, no hazard, minimal hazard, very low hazard,
low hazard, and high hazard, and those are explained in the
standard. Testing can be conducted using either arena air blast or
shock tube test types from which the blast load is obtained. This
standard requires at least three test specimens representative of a
glazing or glazing system to be tested at a given blast load, and an
additional specimen should be used for pretest measurements.
Pressure transducers are used to record the blast pressure on the
test panel during testing.
ISO 16933 and ISO 16934
The ISO provides two standard test methods that may be used to test
and classify the performance of glazing systems under blast loads.
They are ISO 16933 (ISO 2007a) and ISO 16934 (ISO 2007b). The
former is based on the arena air blast test, whereas the latter is based
on the shock tube test. ISO 16933 (ISO 2007a) covers a broad range
of blast parameters, incorporating seven standard blasts simulating
vehicle bombs and seven standard blasts simulating smaller hand-
carried satchel bombs. On the other hand, ISO 16934 (ISO 2007b)is
applicable to blast waves generated in a shock tube facility that
simulate the reflected pressures and impulses generated from high-
explosive detonations of approximately 30–2,500 kg of TNT at
standoff distances of approximately 35–50 m. Both standards pro-
vide a structured procedure to test security glazing, including those
fabricated from glass, plastic glazing sheet materials, glass-clad
plastics, LG, insulated glass, glass/plastic glazing materials, and
film-backed glass.
A minimum of three test specimens, each (1,100 ± 5 mm) ×
(900 ± 5 mm), should be tested at a given level of air blast for
the purpose of classification according to these standards. Test
specimens should be clamped to the test frame using rubber strips of
thickness 4 ± 0:5 mm, width 50 ± 5 mm, and hardness 50 ± 10
IRHD in accordance with ISO 48 (ISO 1994). However, non-
standard test specimens can only be tested, not classified, according
to these standards. They provide six hazard ratings, A–F (no break,
no hazard, minimum hazard, very low hazard, low hazard, and high
hazard), based on the severity of the fragments and hazard effects
as evidenced by the distribution of the fragments and the damage to
the witness panel occurring during the blast test. These hazard
ratings are described in detail in both standards and are similar to
those given in ASTM F1642-04 (ASTM 2010b).
Limitations of the Test Methods
The major limitation of these test methods is the high cost invol-
ved with blast testing. Most universities and government organi-
zations do not have sufficient funds and space to conduct blast
testing. As previously explained, all these standards require at least
three specimens to be tested under a given blast load because
repetitive testing is required to accurately predict the behavior and
the failure of a glazed panel under a blast load. On the other hand,
these test methods are valid for small test specimens with standard
dimensions, and large glazed panels used in most buildings cannot
be classified according to these standards. Health and safety issues
and environmental pollution are some other negative effects of
blast testing.
Most design standards provide useful information for designing
blast-resistant glazing using LG windows. However, current de-
sign standards and test methods have some limitations, which have
already been discussed briefly. This emphasizes the need for a new
analytical procedure for the design of glazing to withstand blast
loads. Numerical analysis with FE codes is a feasible method
that has been used to investigate the behavior of LG panels under
blast loads (Chung et al. 2010;Weggel and Zapata 2008;Weggel
et al. 2007;Seica et al. 2011). This approach is presented in the
following paragraphs.
Finite-Element Modeling of LG
Laminated glass panels are thin structures wherein the thickness is
small compared with the in-plane dimensions, and they can be
modeled with either two-dimensional (2D) shell elements or three-
dimensional (3D) solid elements. Nonlinear dynamic analyses
have been conducted using FE codes with explicit capabilities
such as LS-DYNA, ABAQUS, ANSYS, and EUROPLEXUX to study
the blast response of LG panels. However, most of the research has
been unable to account for the post-crack load-carrying capacity of
LG as well as the effects of structural sealant joints. The authors
have developed a rigorous and reliable procedure for studying the
blast response of LG by overcoming these limitations. These
modeling techniques are described in detail in their previous re-
search work (Hidallana-Gamage et al. 2013a,b) and are briefly
described in this paper.
Modeling Techniques
In the present study, LG panels are modeled with 3D constant-
stress solid elements using LS-DYNA FE code (Hallquist
2006) incorporating material model 110 (MAT_HOLMQUIST_
CERAMICS) for the glass and material model 24 (MAT_
PIECEWISE_LINEAR_PLASTICITY) for the PVB interlayer
and the structural sealant joints. Material model 110 was deve-
loped using the Johnson–Holmquist (JH-2) material model, which
has been widely used to model brittle materials such as concrete,
ceramic, glass, and rock subjected to high pressures, large strains,
and high strain rates. The JH-2 material model was developed with
a set of mathematical equations, and they are explained in detail in
the literature (Cronin et al. 2003;Johnson and Holmquist 1993;
Holmquist et al. 1995).
Polymeric interlayers such as PVB show viscoelastic behavior
under loads with long durations, where their shear modulus changes
over time. However, change in the shear modulus of PVB is neg-
ligible under short-duration loads (approximately 100 ms), and thus
PVB can be analyzed as an elastic–plastic material under blast loads
(Larcher et al. 2012;Hidallana-Gamage et al. 2013a,b;Wei and
Dharani 2006a;Wei et al. 2006b). The behavior of structural sealant
joints can also be treated as elastic–plastic under blast loads. Both
PVB and structural sealant joints are modeled with material model
24, which is widely used to model polymeric materials with elastic–
plastic properties. These material models can account for high-
strain-rate effects, and the authors have confirmed the validity of
these material models for analyzing the behavior of LG under blast
loads (Hidallana-Gamage et al. 2013a,b).
© ASCE 04015003-5 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

Failure Analysis of Materials
Different failure theories are used in practice to predict the failure
of engineering materials. For brittle materials such as glass, the
first principal stress (σ11 is usually used to examine failure. Glass
is considered to have failed if σ11 exceeds the dynamic breaking
strength of glass (Tb), which should be approximately 80 MPa for
annealed glass under blast loads (Hooper et al. 2012;Seica et al.
2011). However, glass is not a homogeneous material and may
break at a lower strength than the expected theoretical values
owing to the presence of surface flaws and microcracks (Netherton
and Stewart 2009). Both PVB and structural sealant materials
show ductile behavior where the von Mises stress (σv) is used to
examine the failure. In the present study, they are considered to
have failed if σv> the failure stress of the material. The authors
have described these failure theories and their application to FE
modeling in detail in their previous research work (Hidallana-
Gamage et al. 2013b).
Comparison of Results
ASTM F2248-09 (ASTM 2010a) provides adequate provisions to
design a complete facade system, including window glazing, sealant
joints, window frame, fasteners, and other supportive elements. It is
also referred to in some of the other design standards and reports
such as UFC 4-010-01 (DoD 2013) and PDC-TR 10-02 (2012).
This paper therefore considers the design guidelines given in ASTM
F2248-09 (ASTM 2010a). The design of an LG panel with a
length of 1.1 m, a width of 0.9 m, and a thickness of 7.5 mm
(3 mm glass + 1:5 mm PVB + 3 mm glass) is carried out accor-
ding to the guidelines given in ASTM F2248-09 (ASTM 2010a)
and ASTM E1300-09a (ASTM 2009) standards. Then the perfor-
mance and the failure pattern of the LG panel are examined under
the design blast load using the results from FE analysis, and the
results are discussed.
Design Based on ASTM F2248-09
The 3-s-duration-equivalent design load for an LG panel with a
length of 1.1 m, a width of 0.9 m, and a thickness of 7.5 mm (having
a 6-mm glass thickness) was found to be approximately 4 kPa ac-
cording to the specified chart given in ASTM E1300-09a (ASTM
2009). This was obtained from the chart given in Fig. 2,whichwas
developed for four-sided simply supported PVB laminate with a
6-mm total glass thickness. According to Fig. 1(ASTM 2010a), the
3-s-duration-equivalent design load of 4 kPa is produced by a blast
threat of an 18-kg TNT equivalent charge weight at a standoff
distance of approximately 13 m. The relevant reflected blast wave
parameters for the blast threat were found from UFC 3-340-02 (DoD
2008) by using the chart developed for hemispherical surface
explosions. The maximum blast overpressure, positive phase
duration, and blast impulse were found to be approximately 88.3
kPa, 10.2 ms, and 301 kPa ms, respectively.
The blast overpressure time–history curve was obtained using
the Friedlander equation, which is given in Eq. (1), where p(t)=
instantaneous overpressure at time t,pa=atmosphere pressure,
pm=peak pressure when t=0, p0=(pm−pa)=peak overpressure
at t=0, td=positive pressure duration, and α=decay factor.
Here the atmospheric pressure was assumed to be 0 kPa, and
hence p0=pm=88:3 kPa and td=10:2 ms. The integration of p(t)
during the time tdgives the blast impulse, which is approximately
301 kPa, and αwas found to be approximately 1.35. Fig. 5shows
the design blast overpressure time–history curve obtained from the
Friedlander equation. Only the positive phase of the blast load
is considered in this study, whereas the negative phase will have
amoreinfluence on flexible structures such as cable net facades
(Teich et al. 2011).
p(t)=p0(1 −t/td)e−αt/td (1)
According to ASTM F2248-09 (ASTM 2010a), the LG panel
should be fixed to the frame using structural sealant joints having a
minimum thickness of 5 mm and a width (bite) of 10–12 mm. In
the present study, it is assumed that the LG panel is fixed to the
frame using structural sealant joints having a thickness of 5 mm
and a width of 10 mm.
Finite-Element Modeling
Finite-element modeling was conducted using LS-DYNA FE code
incorporating 3D constant-stress solid elements as explained ear-
lier. One-quarter of the panel was analyzed using symmetry, as-
suming that the blast load was uniformly distributed over the entire
front glass pane. The glass, interlayer, and sealant joints were
accounted for in the FE model, and the sealant joints were as-
sumed to be fixed to a rigid base, neglecting deformations in the
frame for simplicity. This may be a conservative approach because
flexible window frames will reduce the stresses in glazed panels by
absorbing some energy (Weggel and Zapata 2008). A 3D view and
a sectional view at the supports of the FE model are shown in
Figs. 6(a and b), respectively.
The material properties of glass and the JH-2 material constants
required for the material model 110 were obtained from the litera-
ture (Cronin et al. 2003;Johnson and Holmquist 1993;Holmquist
et al. 1995;Hooper et al. 2012) and those used in the ana-
lysis are presented in Table 1. The density, Young’s modulus, and
Poisson’s ratio for the glass were taken as 2,530 kg /m3, 7 GPa,
and 0.22, respectively (Hooper et al. 2012). In their previous
research work, the authors showed that the tensile strength (T)of
glass used with the material model should be approximately 60–65
MPa for annealed glass (Hidallana-Gamage et al. 2013a,b).
The material properties of the interlayer and structural sealant
used in the analysis are summarized in Table 2. The density,
Young’s modulus, and Poisson’s ratio for the PVB interlayer were
taken as 1,100 kg /m3, 530 MPa, and 0.485, respectively (Hooper
et al. 2012). The PVB was treated as an elastic–plastic material
wherein its yield stress, failure stress, and failure strain were taken
as 11 MPa, 28 MPa, and 2.0, respectively (Larcher et al. 2012).
The density and Poisson’s ratio for the silicone sealant were taken
as 1,100 kg/m3and 0.495, respectively, and its Young’s modulus
0
10
20
30
40
50
60
70
80
90
0 1 2 3 4 5 6 7 8 9 10 11
Reflected Overpressure (kPa)
Time (ms)
Fig. 5. Blast overpressure time–history curve
© ASCE 04015003-6 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

was taken as 2.3 MPa by assuming a hardness of approximately
50 IRHD in accordance with ISO 48 (ISO 1994). The yield stress,
failure stress, and failure strain of the silicone sealant were taken as
2.3 MPa, 3.5 MPa, and 2.5, respectively.
Results of FE Analysis
The results of the FE analysis for the mid-span deflection, fracture,
and crack propagation of the glass panes; the stress variations and
failure analyses; and finally the energy absorption of the glass, the
interlayer, and the sealant joints are presented and described in this
paper. Only one-quarter of the LG panel was modeled, but gra-
phical views of the entire model are shown using the reflection
option in the LS-DYNA FE code.
Mid-Span Deflection
The authors have shown that the tensile strength (T) of glass has a
considerable influence on the blast response of LG panels and
have confirmed that it should be approximately 60–65 MPa for
annealed glass. Fig. 7compares the deflection–time history curves
at the center of the panel for FE models having a Tof 60 and 65
MPa. The FE model with a Tof 60 MPa gives a maximum de-
flection of approximately 145 mm at approximately 16.5 ms,
whereas that with a Tof 65 MPa gives a maximum deflection of
approximately 120 mm at approximately 14.5 ms. The deflection–
time history curves are identical up to approximately 6 ms, and the
FE model with a Tof 60 MPa gives a higher deflection thereafter.
This is because the FE model with a Tof 60 MPa shows more
damage to the glass panes than that with a Tof 65 MPa. The FE
model with a Tof 60 MPa is therefore used as a conservative
approach for the detailed analysis in this paper.
Fracture and Stress Analysis of Glass Panes
The fracture and crack propagation of the glass panes are studied
and presented in this paper. Glass elements failed and were deleted
along the fracture lines, exposing the PVB elements at those lo-
cations. Figs. 8and 9show the fracture and crack propagation of
10 mm
(a)
(b)
Rubber sealant (5 mm)
Rubber sealant (5 mm)
Glass (3 mm)
PVB (1.52 mm)
Glass (3 mm)
Fig. 6. Finite-element model of the LG panel: (a) 3D view; (b) sec-
tional view at the support
Table 1. Material Properties and JH-2 Material Constants of Glass Used
in the FE Analyses
Material property/JH-2 constant Value
Density (ρ) 2,530 kg/m3
Young’s modulus (E) 72 GPa
Poisson’s ratio (υ) 0.22
Strength constants
A0.93
B0.2
C0.003
M1.0
N0.77
Ref. strain rate (EPSI) 1.0
Tensile strength (T) 60 MPa
Failure strain 0.0024
Normalized fractured strength 0.5
HEL 5.95 GPa
HEL pressure 2.92 GPa
HEL strength 4.5 GPa
Damage constants
D1 0.043
D2 0.85
Equation of state
K1 (bulk modulus) 45.4 GPa
K2 −138 GPa
K3 290 GPa
β1.0
HEL =Hugoniot elastic limit.
Table 2. Material Properties of PVB and Rubber Sealant Used in the FE
Analyses
Material property PVB Rubber
Density (ρ) 1,100 kg/m31,100 kg /m3
Young’s modulus (E) (MPa) 530 2.3
Poisson’s ratio (υ) 0.485 0.495
Yield stress (MPa) 11 2.3
Failure stress (MPa) 28 3.5
Failure strain 2 2.5
0
25
50
75
100
125
150
0 5 10 15 20 25
Deflection (mm)
Time (ms)
Glass, T=60 MPa
Glass, T=65 MPa
Fig. 7. Deflection time–history curve at the center of the LG panel for
different Tof glass
© ASCE 04015003-7 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)
(c) (d)
Fig. 8. Crack propagation of the top glass pane: (a) at 0 ms; (b) at 5 ms; (c) at 10 ms; (d) at 15 ms
(a) (b)
(c) (d)
Fig. 9. Crack propagation of the bottom glass pane: (a) at 0 ms; (b) at 5 ms; (c) at 10 ms; (d) at 15 ms
© ASCE 04015003-8 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

the top and bottom glass panes, respectively, at different times.
Both the top and bottom glass panes show similar crack patterns,
but the bottom glass pane has slightly more cracks compared with
the top glass pane. There is a considerable increase in the crack
propagation over time until approximately 15 ms, and there is no
noticeable increase in the cracks thereafter in both glass panes.
Fewer cracks are formed along the edges, showing no signs of
damage or failure at the supports.
Fig. 10 shows the variation in first principal stress (σ11 ) in the
top glass pane at different times. Initially, σ11 increases along the
diagonals and at the middle portion of the top glass pane, and then
the region with high stress expands throughout the top glass pane.
Fig. 11 shows the variation of σ11 in the bottom glass pane at
different times. There is an increase in σ11 along the edges and at
the middle portion of the bottom glass pane initially, and then the re-
ion with high stress expands throughout the bottom glass pane. It is
evident that σ11 increases and goes beyond even 90 MPa along the
fracture lines, confirming the failure of the glass at those locations.
Stress–Strain Analysis of the Interlayer
Fig. 12 shows the variation in von Mises stress (σv) at the bottom
of the interlayer at different times. Initially, σvincreases along the
edges and at the middle of the bottom surface of the interlayer,
similar to the bottom glass pane. When the glass cracks, the inter-
layer stretches at those locations, and therefore there is an increase
in σvalong the fracture lines of the glass panes. Fig. 13(a) shows
the variation in the plastic strain at the bottom of the interlayer at
20 ms. The interlayer has not exceeded its yield stress at most
locations except those along the two vertical fracture lines where
the glass elements failed and were deleted from the FE model.
Three PVB elements as shown in Fig. 13(a) are used for detailed
analysis, and Fig. 13(b) illustrates the variation in σvfor those
elements. The elements have exceeded their yield stress, which is
approximately 11 MPa, but only elements 1 and 2 have reached
their failure stress, which is approximately 28 MPa. Those ele-
ments that have exceeded the failure stress have negligible stress
thereafter, confirming their failure. However, no major damage has
occurred to the interlayer for the treated blast load.
Stress Analysis of Sealant Joints
The results of the FE analysis indicate that the sealant joints along
the long edge of the LG panel have high stresses compared with
those along the short edge. The authors have shown that sealant
joints at the middle of the long edge have high stresses compared
with other parts (Hidallana-Gamage et al. 2013b), and this is
supported by Figs. 8and 9, which show large deformations in the
sealant joints at those locations. Fig. 14(a) shows the critical
sealant elements at the middle of the long edge of the LG panel,
and Fig. 14(b) illustrates the variation in σvof those elements. The
elements have reached their yield stress, which is approximately
2.3 MPa, but none of the sealant elements in the FE model exceed
the failure stress of approximately 3.5 MPa. This confirms that
there cannot be any failure at the sealant joints of the LG panel for
the treated blast load.
(a) (b)
(c)
Fig. 10. First principal stress (σ11) variation in the top glass pane: (a) at 3 ms; (b) at 7 ms; (c) at 20 ms
© ASCE 04015003-9 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)
(c)
Fig. 11. First principal stress (σ11) variation in the bottom glass pane: (a) at 3 ms; (b) at 7 ms; (c) at 20 ms
(a) (b)
(c)
Fig. 12. Von Mises stress (σv) variation at the bottom surface of the interlayer: (a) at 3 ms; (b) at 7 ms; (c) at 20 ms
© ASCE 04015003-10 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

Energy Absorption
Fig. 15 compares the total energy absorption of the glass, the inter-
layer, and the sealant joints for 25 ms, where the total energy is the
summation of the internal and kinetic energies. Initially, the glass
panes absorb most of the energy, reaching a maximum of approxi-
mately 170 J at 5 ms, and their energy absorption decreases gra-
dually and remains around 10 J after 15 ms. The energy absorp-
tion of the interlayer increases gradually until approximately
15 ms, where it reaches its maximum of approximately 200 J and
then remains around that thereafter. The energy absorption of the
sealant joints increases to approximately 50 J at approximately
11 ms and then decreases slightly and remains around 30–40 J after
20 ms. After the glass breaks, the interlayer absorbs most of the
blast energy and contributes approximately 80% to the total energy
absorption after approximately 20 ms. This confirms the importance
of the interlayer of an LG panel during the post-crack stage.
Summary and Conclusion
This paper critically analyzed the latest standards and documents
used for blast-resistant glazing design, such as ASTM F2248-09
(ASTM 2010a), UFC 4-010-01 (DoD 2013), UFC 3-340-02 (DoD
2008), U.K. Glazing Hazard Guide (1997), and PDC-TR 10-02
(2012). Most of them facilitate designing glazed panels with LG
except UFC 3-340-02 (DoD 2008), which is limited to design with
fully tempered glass. Those standards are conservative because
they are based on simplified SDOF analysis. They provide
provisions for designing glazed panels with limited dimensions
and hence cannot be used in designing large facades in real
buildings. Dynamic analysis of glazed panels can be carried out
with computer programs as described in PDC-TR 10-02 (2012),
but comprehensive knowledge and understanding of the computer
programs is required for a feasible and safe design.
Standard test methods are used to classify the hazard rating of
glazed panels depending on their performance under blast loads.
This paper reviewed commonly used standard test methods such as
ASTM F1642-04 (ASTM 2010b), ISO 16933 (ISO 2007a), and
ISO 16934 (ISO 2007b). These test methods are expensive and
can be used only to classify the hazard rating of glazed panels with
limited dimensions. Blast testing causes health and safety issues
and environmental pollution. Numerical analysis of LG panels
with respect to blast loads will produce a comprehensive set of
information and hence provide a viable option to analyze and
design LG to withstand blast loads.
A comprehensive numerical procedure with LS-DYNA FE code
was used to study the blast response of LG panels. The glass,
interlayer, and sealant joints were modeled with 3D constant-stress
solid elements, assuming the window frame to be a rigid base for
simplicity. An LG panel with a length of 1.1 m, a width of
0.9 m, and a thickness of 7.5 mm (3 mm glass + 1:5 mm PVB +
3 mm glass) was designed according to the guidelines given in
Element 1
Element 2
Element 3
0
5
10
15
20
25
30
0 5 10 15 20 25
Von Mises Stress (MPa)
Time (ms)
Element 1
Element 2
Element 3
(a)
(b)
Fig. 13. Stress and strain analysis of critical PVB elements: (a) variation of the plastic strain at the bottom surface of the interlayer at 20 ms; (b) von
Mises stress (σv) variation of the critical PVB elements
© ASCE 04015003-11 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

ASTM F2248-09 (ASTM 2010a) and ASTM E1300-09a (ASTM
2009) standards. The design blast load for the LG panel was found
from UFC 3-340-02 (DoD 2008), and its performance was
examined with the FE model.
The results of FE analysis indicate that both glass panes
fractured under the blast load. The interlayer had high stresses
along the fracture lines, but no major failure could be seen in the
interlayer except along the two vertical fracture lines, where the
interlayer tore by reaching the failure stress. The sealant joints
at the middle of the long edge had high stresses, but no failure
was seen anywhere in the sealant joints. An LG panel, when
designed according to the provisions given in ASTM F2248-09
(ASTM 2010a), should perform to minimal hazard as defined in
ASTM F1642-04 (ASTM 2010b). An LG panel that fails minimal
hazard should fracture but should remain in the frame without
any failure at the sealant joints and the supportive frame. Finite-
element predictions for the failure pattern of the LG panel agreed
reasonably well with those expected from ASTM F2248-09
(ASTM 2010a).
The interlayer absorbed approximately 80% of the energy
from the treated blast load after the glass had broken. The
thickness and material properties of the interlayer have a major
influence on the post-crack behavior of LG, but they are not
accounted for in ASTM F2248-09 (ASTM 2010a)andother
design standards. Similarly, the width and thickness of the sealant
joints as well as the properties of the glass would have an impact
on the blast response of LG panels. Current modeling techniques
may be used to study the influence of the material and geometric
properties of glass, interlayers, and sealant joints to improve the
performance of LG panels under blast loads. This will enable
engineers to better design blast-resistant glazing with LG within
economic constraints.
As previously shown, numerical analysis with FE codes offers
a viable method for blast-resistant glazing design of LG. Com-
prehensive numerical models such as the one developed in this
paper can simulate deflections, glass fracture, stress–strain varia-
tion, and the energy absorption of constituent components in
LG panels. The results of FE analysis can be used to examine the
failure of the glass, interlayer, and sealant joints and hence to
evaluate the performance of the entire LG panel under blast
loading. The comprehensive information provided through such
analysis will not only enhance understanding of the blast response
of LG panels but also facilitate their design. The modeling tech-
niques presented in this paper can therefore be used in blast-
resistant glazing design as a supportive tool for design standards,
and also as a solution when they are not applicable, thus reducing
cost and avoiding the safety issues and environmental pollution
involved with blast testing.
References
ASTM. (2009). “Standard practice for determining load resistance of glass
in buildings.”E 1300-09a, West Conshohocken, PA.
ASTM. (2010a). “Standard practice for specifying an equivalent 3-second
duration design loading for blast resistant glazing fabricated with la-
minated glass.”2248-09 AF, West Conshohocken, PA.
ASTM. (2010b). “Standard test method for glazing and window systems
subject to air blast loadings.”1642-04, West Conshohocken, PA.
Chung, J. H., Consolazio, G. R., Dinan, R. J., and Rinehart, S. A. (2010).
“Finite-element analysis of fluid-structure interaction in a blast-re-
sistant window system.”J. Struct. Eng., 10.1061/(ASCE)ST.1943-
541X.0000100, 297–306.
Cronin, D. S., Bui, K., Kaufmann, C., McIntosh, G., and Berstad T.
(2003). “Implementation and validation of the Johnson-Holmquist
0
0.5
1
1.5
2
2.5
3
0 5 10 15 20 25
Von MisesStress (MPa)
Time (ms)
Element 4
Element 5
Element 5
Element 4
(b)
(a)
Fig. 14. Stress analysis of critical sealant elements: (a) view at the
middle of the long edge of the LG panel showing critical sealant
elements; (b) von Mises stress (σv) variation of the critical sealant
elements
0
50
100
150
200
250
0 5 10 15 20 25
Energy Absorption (J)
Time (ms)
Glass PVB Sealant
Fig. 15. Energy absorption of different components of the LG panel
© ASCE 04015003-12 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.

ceramic material model in LS-DYNA.”4th European LS-DYNA Users
Conf., Dynamore, Ulm, Germany.
Dept. of Defense (DoD). (2008). “Structures to resist the effect of acci-
dental explosions.”UFC 3-340-02, Washington, DC.
Dept. of Defense (DoD). (2013). “Unified Facilities Criteria DOD minimum
antiterrorism standards for buildings.”UFC 4-010-01, Washington, DC.
Hallquist, J. O. (2006). LS-DYNA version 970 theory manual, Livermore
Software Technology Corporation, Livermore, CA.
Hidallana-Gamage, H. D., Thambiratnam, D. P., and Perera, N. J. (2013a).
“Computational analysis of laminated glass panels under blast loads:
A comparison of two dimensional and three dimensional modeling
approaches.”Int. J. Eng. Sci., 2(8), 69–79.
Hidallana-Gamage, H. D., Thambiratnam, D. P., and Perera, N. J. (2013b).
“Failure analysis of laminated glass panels subjected to blast loads.”
Eng. Failure Anal., 36, 14–29.
Holmquist, T. J., Johnson, G. R., Lopatin, C., Grady, D. E., and Hertel, E. S. Jr.
(1995). High strain rate properties and constitutive modeling of glass,
Sandia National Labs, Albuquerque, NM.
Hooper, P. A., Sukhram, R. A. M., Blackman, B. R. K., and Dear, J. P.
(2012). “On the blast resistance of laminated glass.”Int. J. Solids
Struct., 49(6), 899–918.
ISO 48. (1994). Rubber, vulcanized or thermoplastic: Determination of
hardness (hardness between 10 IRHD and 100 IRHD), Geneva.
ISO 16933. (2007a). Glass in building—Explosion-resistant security
glazing—Test and classification for arena air-blast loading, Geneva.
ISO 16934. (2007b). Glass in building—Explosion resistant security
glazing—Test and classification by shock tube loading, Geneva.
Johnson, G. R., and Holmquist, T. J. (1993). “An improved computational
constitutive model for brittle materials.”AIP Conf. Proc., High-
Pressure Science and Technology, American Institute of Physics,
Colorado Springs, CO, 981–984.
Larcher, M., Solomos, G., Casadei, F., and Gebbeken, N. (2012). “Experi-
mental and numerical investigations of laminated glass subjected to
blast loading.”Int. J. Impact Eng., 39(1), 42–50.
Ledbetter, S. R., Walker, A. R., and Keiller, A. P. (2006). “Structural use of
glass.”J. Archit. Eng., 10.1061/(ASCE)1076-0431(2006)12:3(137),
137–149.
Netherton, M. D., and Stewart M. G. (2009). “The effects of explosive
blast load variability on safety hazard and damage risks for monolithic
window glazing.”Int. J. Impact Eng., 36(12), 1346–1354.
Norville, H., and Conrath, E. (2001). “Considerations for blast-resistant
glazing design.”J. Archit. Eng., 10.1061/(ASCE)1076-0431(2001)7:3
(80), 80–86.
Norville, H. S., and Conrath, E. J. (2006). “Blast-resistant glazing design.”
J. Archit. Eng., 10.1061/(ASCE)1076-0431(2006)12:3(129), 129–136.
Protective Design Center. (2012). “Blast resistant design methodology for
window systems designed statically and dynamically.”PDC-TR 10-02,
U.S. Army Corps of Engineers, Omaha, NE.
Seica, M. V., Krynski, M., Walker, M., and Packer, J. A. (2011).
“Analysis of dynamic response of architectural glazing subject to blast
loading.”J. Archit. Eng., 10.1061/(ASCE)AE.1943-5568.0000035,
59–74.
Smith, D. (2001). “Glazing for injury alleviation under blast loading–
United Kingdom practice.”Proc., 7th Int. Glass Conf. Glass Proces-
sing Days, Tampere, Finland, 335–340.
Teich, M., Warnstedt, P., and Gebbeken, N. (2011). “The influence of
negative phase loading on cable net facade response.”J. Archit. Eng.,
10.1061/(ASCE)AE.1943-5568.0000083, 276–284.
U.K. Glazing Hazard Guide. (1997). “Glazing hazard guide, cubicle stand-
offs, tables and charts.”SSG/EP/4/97, SAFE/SSG, Explosive Protec-
tion, London.
Weggel, D. C., and Zapata, B. J. (2008). “Laminated glass curtain
walls and laminated glass lites subjected to low-level blast loading.”J.
Struct. Eng., 10.1061/(ASCE)0733-9445(2008)134:3(466), 466–477.
Weggel, D. C., Zapata, B. J., and Kiefer M. J. (2007). “Properties and
dynamic behavior of glass curtain walls with split screw spline mul-
lions.”J. Struct. Eng., 10.1061/(ASCE)0733-9445(2007)133:10
(1415), 1415–1425.
Wei, J., and Dharani, L. R. (2006a). “Response of laminated architectural
glazing subjected to blast loading.”Int. J. Impact Eng., 32(12),
2032–2047.
Wei, J., Shetty, M. S., and Dharani, L. R. (2006b). “Stress characteristics
of a laminated architectural glazing subjected to blast loading.”Com-
put. Struct., 84(10–11), 699–707.
© ASCE 04015003-13 J. Archit. Eng.
J. Archit. Eng. 2015.21.
Downloaded from ascelibrary.org by Queensland University of Technology on 09/01/15. Copyright ASCE. For personal use only; all rights reserved.