Case Studies in Structural Engineering 6 (2016) 53–62

Contents lists available at ScienceDirect

Case Studies in Structural Engineering

journal homepage: www.elsevier.com/locate/csse

An approximate method of dynamic amplification factor for

alternate load path in redundancy and progressive collapse
linear static analysis for steel truss bridges
Hoang Trong Khuyen a , Eiji Iwasaki b,∗
a
Graduate School of Engineering, Nagaoka University of Technology, 1603-1 Kamitomioka Nagaoka, Niigata, Japan
b
Department of Civil and Environmental Engineering, Nagaoka University of Technology, 1603-1 Kamitomioka Nagaoka, Niigata, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Linear static analysis with an alternate load path using dynamic amplification factor (DAF)
Received 22 March 2016 is often used for redundancy and progressive collapse analysis of steel truss bridges to
Received in revised form 30 May 2016 avoid using the more time-consuming dynamic analysis. This study presents an empirical
Accepted 6 June 2016
equation to calculate the DAF for this type of analysis against the initial sudden member
Available online 7 June 2016
fracture. Currently, this analysis employs an approximate model with a single degree of
freedom to calculate the DAF. With a 5% damping ratio, the constant DAF of 1.854 is used
Keywords:
for all types of steel truss bridges. However, this approach is inaccurate because the DAF
Dynamic amplification factor
varies between bridges and with the location of the fractured members as well. Considering
Redundancy
Progressive collapse some of the approaches developed for building structures but adapting them to steel truss
Steel truss bridges bridges, this paper proposes an empirical equation that allows for the computation of the
Alternate load path DAF from the maximum norm stress is /iy in static linear elastic analysis of the damaged
model with a member removal. A total of 30 illustrative cases for two typical steel truss
bridges are investigated to obtain the data points for the empirical equation. The proposed
empirical equation is the enveloped line offset from the best fit line for the data points in
illustrative cases.
© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Progressive collapse is the spread of an initial local failure from element to element, member to member, eventually
resulting in the collapse of a part, entire structures or a disproportionately large part [1]. A sudden member failure is a
dynamic event in which the structural motion is initiated by energy released by the sudden loss of a load-carrying member.
Four methods, including linear static analysis, nonlinear static, linear dynamic and nonlinear dynamic methodologies, are
available for redundancy and progressive collapse analysis of structures for the sudden fracture of a member or component
[2,3]. The event of a sudden member fracture relates to both the primary loading, which causes the initial fracture, and
impact loading, which causes structural motions after the initial fracture. The dynamic method is a direct solution to address
impact loading and the dissipation procedure of the energy induced by the initial member fracture. This approach is accurate,
but it requires much intensive computation with time-history transient analysis. Static analysis with an alternate load path,

∗ Corresponding author.
E-mail addresses: s115081@stn.nagaokaut.ac.jp (H.T. Khuyen), iwas@nagaokaut.ac.jp (E. Iwasaki).

http://dx.doi.org/10.1016/j.csse.2016.06.001
2214-3998/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
54 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62

Fig. 1. Illustration of the procedure to obtain data points of DAF versus max normalized stress and how to find the empirical DAF formula.

which amplifies the primary loading with a dynamic amplification factor (DAF) to form the impact loading, is an alternative
approach for analysis without using dynamic analysis.
Currently, linear redundancy and progressive collapse linear static analysis of steel truss bridges have employed a single
degree of freedom (SDOF) model to conventionally calculate the DAF [4,5]. With a 5% damping ratio, the conventional DAF is
1.854, constant for all bridges. This approach is conservative because the bridge system acts as multiple degrees of freedom
instead of a single degree of freedom. The DAF varies between bridges and with the location of the fractured members, as
well.
To consider a model with multiple degrees of freedom, Goto et al. propose the root mean square mode combination
method to approximate the DAF [6]. This approach is moderately accurate and requires some correction factors. Although
no other studies have yet been published about the approximation of the DAF for bridge systems, such approaches by Liu
[7], McKay et al. [8], DoD, U.S. [9], and Stevens et al. [10] that approximate DAF in a building system are valuable. McKay
et al., DoD, U.S., and Stevens et al. propose different linear functions of norm rotation, which is the ratio of the total member
rotation to the member-yield rotation, to compute the DAF of steel buildings. On the other hand, Liu computes the DAF by
using the function of max(Mu /Mp ), where the max operator is applied to all affected beams that are directly adjacent to
and above the removed column. Mu and Mp are the factored moment demand under the original unamplified static gravity
loads and the factored plastic moment capacity, respectively, of an affected beam. These approaches may be limited to only
one building system because the norm rotation and Mu /Mp are critical parameters for the behavior of a building system. In
a steel truss bridge system, when a member fractures, in addition to axial force, the members are also subject to bending
moments. Considering this behavior, this study proposed the DAF as a function of the maximum norm stress ␴is /␴iy , where
␴is and ␴iy are stress in a static analysis and the yield stress of bridge members. In this paper, a total of 30 illustrative cases
are investigated in 3D models. The empirical equation to calculate the DAF was defined as the enveloped line offset from
the best fit line for the data points from illustrative cases.

2. New DAF calculation method and analysis procedure

The empirical equation to calculate the DAF is defined as a function of the maximum norm stress ␴is /␴iy , where ␴is and
␴iy are the stress in static analysis and yield stress of the ith bridge member. For a given member fracture scenario, the DAF is
obtained by the stress DAF and is then confirmed by an alternate static analysis with amplified loading using the calculated
amplification factor to ensure that the structural responses best match those from the linear dynamic analysis. The process
of computing the DAF in a given damaged scenario undertakes the following procedure, as in Fig. 1.
Step 1: Statically apply the primary load GL, as defined in Section 3.3, to the damaged bridge, remove the member that
is being fractured and perform the static linear elastic analysis. Then, measure the norm stress␴is /␴iy of the members and
the maximum ␴is /␴iy , where ␴is and ␴iy are stresses in the static analysis and yield stress of the ith bridge member.
 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62 55

Step 2: Carry out the linear dynamic analysis to obtain the position of maximum vertical deflection during time history.
The time loading is shown in Fig. 7. To do so, the analysis involves a linear elastic analysis of the intact bridge with the
primary loading GL to measure the end forces F of the to-be-removed member. The analysis then statically applies the
primary loading GL and end forces in the damaged bridge model with a member removed to make sure that the deflection
of the model is the same as the intact bridge just before the damage. In this step, the physical function of the to-be-removed
member is replaced by its end forces. Then, the end force F is suddenly forced down to zero to make a sudden fracture of the
member. The next step is to record the chord deflection and demand-capacity ratio R. In the standard definition, the DAF is
the ratio of deflections or stresses in the dynamic analysis of deflections or stresses in the static analysis [11]. Hence, in this
study, the DAF is calculated through Eq. (1). Normalizing the stress by the yielded stress, Eq. (1) can be rewritten as Eq. (2).
The system DAF is the slope of the best fit line of the data points of ␴idm /␴iy and ␴is /␴iy . This is the trial DAF, which needs
to be confirmed in Step 3.
␴idm
DAFi = (1)
␴is
␴idm /␴iy Nidm /Ni0 + Mxidm /Mxi0 + Myidm /Myi0
DAFi = = (2)
␴is /␴iy Nis /Ni0 + Mxis /Mxi0 + Myis /Myi0
where idm , Nidm , Mxidm , and Myidm are the maximum dynamic stress of a member, dynamic axial force, and bending
moments of the ith member, respectively, after member failure in the dynamic analysis at the time step when vertical
displacement reaches the peak. At this time step, the stress reaches the maximum value as well.  is , Nis , Mxis , and Myis
are the corresponding static stress, static corresponding axial force, and moments of the ith member, respectively, after the
member failure in the linear static analysis. ␴iy is the yield stress of the ith member. Ni0 , Mxi0 , and Myi0 are the initial yield
capacity of the section of the ith member.
Step 3: Alternate static analysis is executed with an amplified load, which is multiplied by the trial DAF in Step 2 to confirm
whether the deflection and demand-capacity ratio, R, are the same as in the linear dynamic analysis. The R is calculated by
Eqs. (3) and (4) [3,4].
N Mx My
R= + + (3)
NP Mpx Mpy
P 1 Meqx 1 Meqy
R= + + (4)
Pu 1 − P/Pex Mpx 1 − P/Pey Mpy
In the above equations, N and P are the tension and compression forces, respectively, and Mx and My are bending moments
around the strong and weak axis, respectively. Np , Mpx and Mpy are the plastic axial strength and full plastic moment
strengths around the strong and weak axis, respectively. Pe is the Euler buckling load. Pu is the ultimate compression
strength associated with the global buckling. The equivalent uniform moments, Meqx , and Meqy are used to convert the
linear distributed moment state into a uniform moment condition.

3. Illustrative application to steel truss bridges

3.1. Geometry of illustrative bridges

In this study, the DAF calculation method is illustrated using two steel truss bridges. To apply it to real bridge conditions,
two original bridge designs in Japan were investigated. The first bridge, Bridge A, is a steel deck-through truss type. The
bridge has a single span that is 90.0 m in length and 9.0 m in width. The concrete deck thickness is 200 mm. The member
details are presented in Fig. 2, Fig. 4 and Table 1. The second bridge, Bridge B, is a steel upper deck truss type. The bridge
span is 73.8 m, and the deck width is 8.30 m. The details of Bridge B are shown in Fig. 3, Fig. 4 and Table 2. The concrete deck
thickness is 200 mm. In both bridges, the structural steel is Japanese steel SM490A with yield stress ␴y = 315 MPa.

3.2. Structural modeling

The illustrative bridges were modeled on a 3D model by finite element analysis software Diana 9.4.3 to obtain the actual
behavior of the structures when a member is fractured. Fig. 5 presents the original models of the illustrative bridges before
removing the member. The structural steel was modeled using 3D beam elements with 3 nodes and six degrees of freedom
at each node. The truss members are designed to have pin connection at the truss joints. However, with the existence of large
gusset plates and numerous bolts and rivets, those truss joints can retain bending moments. Hence, a beam element with
a rigid truss joint better reflects the structural response to loading [12]. The concrete deck was modeled using curved shell
elements with 8 nodes per element. The eccentricities among the structural members were addressed by rigid beams to
provide an accurate structural layout as well as the deformation relationship of the members. E and ␴y are the elastic modulus
and yield stress, respectively, of the structural steel. This study assumed structures with elastic behaviors to calculate the
DAF, which is used in the linear static analysis with the alternate load path. Rayleigh damping was used, with the mass
damping coefficient ␣ ¯ forced to zero. The stiffness damping coefficient ␤ ¯ was defined by assuming a 5% damping ratio for
56 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62

4250 4250

UL

UL

UL

UL

8500
EP
UL UL UL UL

(a) Upper laterals

U1 U2 U3 U4

11000

D5
D1

D3

D4
D4
D2

D4

D4

D4
L1 L1 L2 L3 L4
90000=10000x9
2800

(b) Main truss

S1 S2 S2 S2 S2

F2
2800

F1
F2
F1
F2
F1
F2
F1
F2
F1
(c) Stringers and diaphragms
8500

CB

CB

CB

CB

CB
LL LL LL LL LL
3 3 3 2 1

(d) Floor beams and lower laterals

Fig. 2. Geometry of Bridge A.

UL1 UL1 UL1 UL1 UL1 UL1

6000

(a) Upper laterals

6000

ST2 ST2 ST2 ST2 ST2 ST1

F1

F2

F2

F2

F2

F2

F2

(b) Floor beams and stringers

U1 U1 U1 U2 U2 U3
D3
D1

D5
7000

D2

D4

D6
V1

V1

V1

L1 L2 L3
73800=12300x6
(c) Main truss
LL1 LL1 LL2 LL2 LL2 LL2 SW1 SW1 SW1 SW2
6000

(d) Lower laterals

Fig. 3. Geometry of Bridge B.

Upper Flg. Upper Flg. Upper Flg. Upper Flg. Upper Flg.
Web
3

Web
Web

Web

Web

Web

Web
Web

Web

35
35

Lower Flg. Lower Flg. Lower Flg. Lower Flg. Lower Flg. Lower Flg.
Box-1 shape Box-2 shape Box-3 shape I, H shape C shape T shape

Fig. 4. Member shape configuration.

the first vibration mode. A 5% damping ratio is the value used in the SDOF model to derive the conventional DAF of 1.854. A
5% damping is also the most acceptable damping ratio for the bridge structural system.

3.3. Loadings

The event of a member fracture relates two types of loadings, including primary loading GL and impact loading. The
primary loading is used to cause the initial member fracture. It may be overweight trucks, traffic collision, corrosion or
fatigue cracks. The impact loading, on the other hand, is the generated loading due to structural motion, which is in turn
 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62 57

Table 1
Details of Bridge A.

Member Shape Web (mm) Upper Flg. (mm) Lower Flg. (mm)

U1 Box-1 575 × 14 570 × 10 500 × 10

U2 Box-1 575 × 14 570 × 18 500 × 18
U3 Box-1 575 × 22 570 × 18 500 × 18
U4 Box-1 575 × 22 570 × 22 500 × 22

L1 Box-2 540 × 14 500 × 10 570 × 10

L2 Box-2 540 × 14 500 × 14 570 × 14
L3 Box-2 540 × 15 500 × 18 570 × 18
L4 Box-2 540 × 15 500 × 19 570 × 19

D1 Box-1 575 × 14 570 × 10 500 × 10

D2 H 472 × 14 445 × 14 445 × 14
D3 Box-3 478 × 10 445 × 11 445 × 11
D4 H 472 × 9 445 × 14 445 × 14
D5 Box-3 480 × 9 445 × 10 445 × 10

EP I 1200 × 9 300 × 14 300 × 14

CB I 1300 × 9 350 × 15 350 × 12
UL I 176 × 9 200 × 12 200 × 12

LL1 T 226 × 11 – 300 × 15

LL2 T 179 × 10 – 300 × 16
LL2 I 163 × 10 300 × 16 300 × 16

S1 I 800 × 9 240 × 15 240 × 15

S2 I 800 × 9 240 × 13 240 × 13

F1 C 174 × 8 90 × 13 90 × 13
F2 C 274 × 8 90 × 13 90 × 13

Table 2
Details of Bridge B.

Member Shape Web (mm) Upper Flg. (mm) Lower Flg. (mm)

U1 Box-1 650 × 13 470 × 16 390 × 13

U2 Box-1 650 × 15 470 × 22 390 × 22
U3 Box-1 650 × 16 470 × 25 390 × 25

L1 Box-2 400 × 9 390 × 9 470 × 11

L2 Box-2 400 × 13 390 × 16 470 × 19
L3 Box-2 400 × 16 390 × 22 470 × 25

D1 Box-3 390 × 16 330 × 16 330 × 16

D2 H 390 × 13 300 × 13 300 × 13
D3 Box-3 390 × 16 300 × 16 300 × 16
D4 H 358 × 9 280 × 13 280 × 13
D5 Box-3 390 × 10 280 × 9 280 × 9
D6 H 370 × 9 280 × 9 280 × 9

V1 Box-3 390 × 10 280 × 9 280 × 9

F1 I 800 × 9 200 × 12 200 × 12

F2 I 800 × 9 230 × 12 200 × 12

UL1 T 176 × 8 – 110 × 8

LL1 T 178 × 8 – 110 × 10
LL2 T 176 × 8 – 110 × 8
SW1 T 204 × 10 – 134 × 12
SW2 I 176 × 9 200 × 12 200 × 12

S1 I 650 × 9 240 × 11 230 × 10

S2 I 650 × 9 240 × 11 200 × 10

SW3 T 204 × 10 – 134 × 12

SW4 T 176 × 8 – 110 × 8

due to instantaneous geometrical change after the initial fracture. In static analysis, the impact loading is addressed by
the alternate load path using a dynamic amplification factor. This study does not consider the required primary loadings
for redundancy and progressive analysis of truss bridges. Instead, the load primary loading is assumed to be at some level
of standard design load GL = 1.0D + 0.5(L + I); 1.0D + 0.75(L + I); 1.0D + 1.0(L + I); 1.0D + 1.25(L + I); or 1.0D + 1.5(L + I), where D,
L, and I represent the dead load, live load and live load dynamic allowance, respectively. The loading levels are selected
to ensure that the Eigen-mode patterns of the bridge are the same when a member fractures. The loading in this study,
58 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62

Fig. 5. Original models of illustrative bridges: (a) Bridge A; (b) Bridge B.

5500 2000 750 5500 500 750

p1+p2 or p2
p1+p2 or p2 (p1+p2)/2 or p2/2
(p1+p2)/2 or p2/2 p2

silewalk silewalk

Damaged truss Notes: p1=10kN/m2; p2=3.5kN/m2

Damaged truss Notes: p1=10kN/m2; p2=3.4kN/m2

(a) (b)

8500 10000 Target member p1+p2

Target member
p2 p2
p1 +p 2
p2

10000x9=90000 26500 10000 37300

(c) (d)
Fig. 6. Live load distribution: (a) Bridge A from the transverse direction; (a) Bridge B from the transverse direction; (c) Bridge A from the longitudinal
direction; (d) Bridge B from the longitudinal direction.

Fig. 7. Member fractures and time loadings.

as shown in Fig. 6, is prescribed by the Japanese Standard for Highway Bridges because illustrative bridges were designed
by this specification [13]. In dynamic analysis, the model mass includes the self-weight of the steel structures, which are
represented by the mass density, and the equivalent mass of the live load level, including its dynamic allowance and the
weight of the concrete slab converted to stringer joints, as shown in Fig. 5 [6]. The time loading is included as in Fig. 7. The
sectional force F is suddenly forced down to zero at fracture time t, while the vertical load GL is kept constant to express
the sudden fracture. An additional series of dynamic analyses with various values of t showed that when t is less than
0.01 s, the results of dynamic analysis are the same as those shown in Fig. 8. In this study, the time t was set to 0.0001 s
for all study cases.
 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62 59

0 0.5 1 1.5 2 2.5 3

0
Fractured member 0.5s
-0.05

Vertial Displacement Y (m)

-0.1 0.2s

-0.15 0.1s

-0.2 0.01s
-0.25 0.001s
-0.3
Used in analyses 0.0001s
-0.35 Same results 0.00001s
-0.4
Time (sec)

Fig. 8. Effect of fractured time t.

Fracture member

Case1 Case2 Case3 Case4 Case5 Case6

Bridge A Bridge B

Fig. 9. Member fracture scenarios.

0 3 6 9 12 15
0
Linear dynamic 5%
-0.05
Vertial displacement Y (m)

-0.1 Static
-0.15

-0.2

-0.25

-0.3

-0.35 max. displacement (time step 19)

-0.4
Time (sec)

Fig. 10. Deflection in linear dynamic analysis for Case 1 at GL = 1.0 D + 1.0(L + I).

3.4. Member fracture scenarios

In steel truss bridges, members are fractured due to overloading, traffic collision, fatigue cracks, corrosion, etc. Because
of these causes of damage, the compression members buckle and lose their capacity. The buckling is not a sudden failure;
hence, the case of a sudden loss of the capacity of a compression member does not occur. However, when a tension member
initiates the cracks, then the cracks may be prolonged instantly, leading the entire section to fracture and suddenly lose its
capacity. Hence, in this study, six tension members that represent the typically damaged scenarios were investigated as in
Fig. 9. The fracture was assumed to occur at five primary loading states as in Section 3.3. The DAF of a total of 30 illustrative
cases was calculated.

3.5. Results and discussion

The results from Case 1 at loading GL = 1.0D + 1.0(L + I) are proposed to illustrate the above calculation process. Fig. 10
plots the time history of the vertical displacement of the mid-span damaged truss side in the linear dynamic analysis in Step
2. The time history of the vertical displacement attained its peak at time step 19. Hence, the sectional force at time-step 19
was used to define the norm stress in dynamic analysis␴idm /␴iy . Combining with ␴is /␴iy in the static analysis in Step 1, the
60 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62

2
y = 1.3232x
1.5 R² = 0.9983
1

0.5

σidm/σiy
0

-0.5

-1

-1.5 DAFtrial=1.3232

-2
-1.5 -1 -0.5 0 0.5 1 1.5
σis/σiy

Fig. 11. DAF for Case 1 at GL = 1.0D + 1.0(L + I).

0.15
0.1
0.05
0
Displacement (m)

-0.05 0 10 20 30 40 50 60 70 80

-0.1
-0.15
-0.2
-0.25 Dynamic Y Dynamic Z
-0.3 Alternate Y Alternate Z
-0.35
Distance (m)

Fig. 12. Deflection of upper chords in intact side for Case 1 at GL = 1.0 D + 1.0(L + I).

0.15
0.1
0.05
0
Displacement (m)

0 10 20 30 40 50 60 70 80
-0.05
-0.1 Dynamic Y
Dynamic Z
-0.15
Alternate Y
-0.2
Alternate Z
-0.25
-0.3
-0.35
Distance (m)

Fig. 13. Deflection of upper chords in damaged site for Case 1 at GL = 1.0 D + 1.0(L + I).

data points of the end sections of all truss members are plotted in Fig. 11. It can be seen that the data points are located
in a linear line. This result indicates that even though a model of multiple degrees of freedom was used, the DAF of the
bridge system can be expressed with a single value. This value is defined as the DAFtrial , which is shown in Fig. 1. In this
illustrative case, the value of the DAFtrial is 1.3232, which is the slope of the best fit line of the data points. Figs. 12 and 13
depict the comparison of the deflections of the upper chord between the linear static analysis with the alternate load path
in Step 3 compared and the dynamic analysis in Step 2. Fig. 14 shows the comparison of R between the alternate load path
analysis and dynamic analysis. We find that the deflection and R in the alternate load path analysis are similar to dynamic
analysis. Hence, the final DAF is decided to be equal to the DAFtrial . A similar process was applied to a total of 30 cases in two
illustrative bridges. Fig. 15 displays the data points of the final DAF and the max ␴is /␴iy of all illustrative cases. The proposed
 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62 61

Intact truss side Damaged truss side

Fractured member

1.1
1.0

5
8
1.5

1 .0
1.1

7
1
9
1.42 1.05
dynamic

6
1.01

4
1.1
1.1
1.6

1.0

6
2
1.41 1.05
alternate

Fig. 14. Demand-capacity ratio R for Case 1 at GL = 1.0 D + 1.0(L + I).

1.8
1.7 Proposed: y = 0.255x + 1.00
Bridge A
1.6
Bridge B
1.5
Data fit
DIF

1.4
1.3
1.2
1.1 y = 0.2548x + 0.9609
R² = 0.9361
1
0.3 0.6 0.9 1.2 1.5 1.8
max σis/σiy

Fig. 15. Approximation of dynamic amplification factor.

Table 3
Comparison between the accurate DAF and approximate DAF for study cases at GL = 1.0D+ 1.0(L + I).

Cases Exact DAF (dynamic) Empirical equation

max is /iy DAF (empirical equation) Error

Case 1 1.325 1.3079 1.327 0.8%

Case 2 1.216 0.9539 1.238 1.4%
Case 3 1.083 0.5724 1.143 4.2%
Case 4 1.150 0.8122 1.203 5.0%
Case 5 1.272 1.2738 1.318 4.1%
Case 6 1.325 1.3040 1.326 0.5%

empirical equation is the enveloped line offset from the best fit line of the data points. Table 3 compares the DAF calculated
by the proposed empirical equation to the exact DAF defined by dynamic analysis at GL = 1.0D + 1.0(L + I). It is shown that
the empirical equation derived a good approximation to the exact DAF from the dynamic analysis.

4. Conclusions

Data points of dynamic amplification factor DAF versus max ␴is /␴iy of a total of 30 illustrative cases in two typical truss
bridges showed that the DAF can be approximated by a function of max ␴is /␴iy. The empirical equation to calculate the DAF
for redundancy and progressive collapse analysis of steel truss bridges against the initial fracture of a member was proposed.
The empirical equation means an advance to the field of providing an approximation, which could help in the analysis of
a dynamic event. With this method, the analyses can all be static, becoming an easy problem that is quick to solve with
accuracy.
It is also noted that, although the proposed empirical equation for DAF calculation is expected to be applicable to any
steel truss bridge, the conclusions and the findings of the empirical equation derived in this paper are solely based on the
illustrative cases of only two types of steel truss bridges. Hence, the validation of the empirical equation to general steel
truss bridges may not be immediately filled; therefore, further research is needed. Additionally, the value of max ␴is /␴iy
in the data points for the empirical equation ranged from 0.4 to 1.8. Although this value rarely extends beyond this range
in modern bridge conditions, this situation needs further investigation. However, when the norm stress is large, the linear
62 H.T. Khuyen, E. Iwasaki / Case Studies in Structural Engineering 6 (2016) 53–62

analysis may be inaccurate. In this case, it is recommended to use a nonlinear analysis with a dynamic amplified factor
derived from nonlinear dynamics to analyze the structure.

References

[1] ASCE, Minimum Design Loads for Buildings and Other Structures, ASCE/SEI 7-10, Reston, VA, 2010.
[2] S. Marjanishvili, E. Agnew, Comparison of various procedures for progressive collapse analysis, J. Perform. Constr. Facil. 20 (4) (2006) 365–374.
[3] S.M. Marjanishvili, Progressive analysis procedure for progressive collapse, J. Perform. Constr. Facil. 18 (2) (2004) 79–85.
[4] Fatigue Evaluation and Redundancy Analysis, Bridge No.9340, I-35W over Mississippi Rive, URS Corporation, 2006.
[5] H. Nagatani, N. Akashi, T. Matsuda, M. Yasuda, H. Ishii, M. Miyamoto, Y. Obata, H. Hirayama, Y. Okui, Structural redundancy analysis for steel truss
bridges in Japan, J. Jpn. Soc. Civil Eng. Div. A (Doboku Gakkai Ronbunshuu A) 65 (2) (2008) 410–425.
[6] Y. Goto, N. Kawanishi, I. Honda, Dynamic stress amplification caused by sudden failure of tension members in steel truss bridges, J. Struct. Eng. ASCE
137 (8) (2011) 850–861.
[7] M. Liu, A new dynamic increase factor for nonlinear static alternate path analysis of building frames against progressive collapse, Eng. Struct. 48
(2013) 666–673.
[8] McKay, K. Marchand, M. Diaz, Alternate path method in progressive collapse analysis: variation of dynamic and nonlinear load increase factors, ASCE
Pract. Period. Struct. Des. Constr. 17 (4) (2012) 152–160.
[9] DoD U.S, UFC 4-023-03: Design of Buildings to Resist Progressive Collapse, US Department of Defense, Washington, DC USA, 2009.
[10] D. Stevens, B. Crowder, B. Hall, K. Marchand, Unified progressive collapse design requirements for DOD and GSA, ASCE Struct. Congr. 2008 (2008)
1–10.
[11] K. Chopra, Dynamics of Structures: Theory and Applications, Prentice-Hall, New Jersey USA, 1995.
[12] R.S. Nagavi, A.E. Aktan, Nonlinear behavior of heavy class steel truss bridge, J. Struct. Eng. 129 (8) (2003) 1113–1121.
[13] JRA, Specification for Highway Bridges—Part I &II, Maruzen, Tokyo, 2002.