ARTICLE IN PRESS

Soil Dynamics and Earthquake Engineering 27 (2007) 200–211

www.elsevier.com/locate/soildyn

Stress analysis of buried steel pipelines at strike-slip fault crossings

Dimitrios K. Karamitros, George D. Bouckovalas, George P. Kouretzis
Department of Geotechnical Engineering, School of Civil Engineering, National Technical University of Athens, 9,
Iroon Polytehniou Street, 15780 Athens, Greece
Received 4 April 2006; received in revised form 27 July 2006; accepted 10 August 2006

Abstract

Existing analytical methods for the stress analysis of buried steel pipelines at crossings with active strike-slip faults depend on a number
of simplifications, which limit their applicability and may even lead to non-conservative results. The analytical methodology presented
herein maintains the well-established assumptions of existing methodologies, but also introduces a number of refinements in order to
achieve a more wide range of application without any major simplicity sacrifice. More specifically, it employs equations of equilibrium
and compatibility of displacements to derive the axial force applied on the pipeline and adopts a combination of beam-on-elastic-
foundation and elastic-beam theory to calculate the developing bending moment. Although indirectly, material and large-displacement
non-linearities are also taken into account, while the actual distribution of stresses on the pipeline cross-section is considered for the
calculation of the maximum design strain. The proposed methodology is evaluated against the results of a series of benchmark 3D non-
linear analyses with the finite element method. It is shown that fairly accurate predictions of pipeline strains may be obtained for a wide
range of crossing angles and fault movement magnitudes encountered in practice.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Buried steel pipelines; Strike-slip faults; Stress analysis; Design

1. Introduction provide ground for the use of simplified analytical

methodologies, at least for preliminary design and verifica-
Evaluation of the response of buried steel pipelines at tion purposes.
active fault crossings is among their top seismic design A simplified methodology which is widely used today
priorities. This is because the axial and bending strains for strike-slip and normal faults, is the one originally
induced to the pipeline by step-like permanent ground proposed by Kennedy et al. [4], and consequently adop-
deformation may become fairly large and lead to rupture, ted by the ASCE guidelines for the seismic design of
either due to tension or due to buckling. Apart from the pipelines [5]. Kennedy et al. extended the pioneering
detrimental effects that such a rupture can have to the work of Newmark and Hall [6], by taking into account
operation of critical lifeline systems [1,2], an irrecoverable soil-pipeline interaction in the transverse, as well as in
ecological disaster may also result from the leakage of the longitudinal directions. Focusing upon cases where
environmentally hazardous materials such as natural gas, the fault rupture provokes severe elongation of the
fuel or liquid waste. The currently available techniques of pipeline, so as tension is the prevailing mode of de-
numerical analysis (e.g. large scale Finite Element models) formation, Kennedy et al. analyzed the relationship
allow a rigorous solution of this problem, minimizing the between the axial tensile force, the bending moment and
number of necessary approximations [3]. Nevertheless, the the corresponding axial and bending strains and concluded
non-linear behavior of the pipeline steel, the soil-pipeline that:
interaction and the second order effects, induced by large
displacements, make such analyses rather demanding, and

the axial tensile force does not depend on the pipeline
Corresponding author. Tel.: +30 2107723870; fax: +30 2108068393. curvature, as long as bending strains do not exceed 80%
E-mail address: [email protected] (D.K. Karamitros). of the corresponding axial strains, and that

0267-7261/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.soildyn.2006.08.001
 ARTICLE IN PRESS
D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211 201

Nomenclature tu limit soil-pipeline friction force, per pipeline

length
a parameter of the Ramberg–Osgood stress– V shear force
strain curve w transverse horizontal displacement of the pipe-
As area of the pipeline cross-section line
C constant, used in Eq. (1b) x position along the pipeline longitudinal axis
Cr constant of rotational springs on point A and C b angle formed by the fault trace and the pipeline
D pipeline diameter axis
E Young’s modulus of the pipeline steel d transverse displacement of point B
E1 elastic Young’s modulus of the pipeline steel Df total fault displacement
E2 plastic Young’s modulus of the pipeline steel DLav available elongation
Ei initial Young’s modulus in the Ramberg–Os- DLreq required elongation
good stress–strain curve Dx fault displacement, parallel to the pipeline
Esec secant Young’s modulus of the pipeline steel longitudinal axis
F axial force Dy fault displacement, perpendicular to the pipe-
Fa axial force at the intersection of the pipeline line longitudinal axis
with the fault trace e1 yield strain of the pipeline steel
I moment of inertia of the pipeline cross section ea pipeline axial strain
k elastic constant of the transverse horizontal soil eb pipeline bending strain
springs emax maximum strain on the pipeline cross-section
L distance from the fault trace emin minimum strain on the pipeline cross-section
Lanch pipeline unanchored length y polar angle of the pipeline cross-section
Lc pipeline curved length l coefficient defined in Eq. (1c)
M bending moment s1 yield stress of the pipeline steel
Mmax maximum bending moment sy yield stress in the Ramberg–Osgood stress–
qu limit stress for transverse soil springs, per strain curve
pipeline length sa axial stress at the intersection of the pipeline
r parameter of the Ramberg–Osgood stress– with the fault trace
strain curve j angle of rotation
R radius of curvature j1 polar angle defining the part of the cross-
Rm mean pipeline radius section that is under yield due to tension
t pipeline thickness j2 polar angle defining the part of the cross-
section that is under yield due to compression

bending moments become negligible and the pipe predicted strains may become even one order of magnitude
behaves essentially as a cable, when the whole pipeline larger than the actual ones.
cross-section is under yield. Wang and Yeh [7] tried to overcome this shortcoming by
taking the pipeline bending stiffness into account. Their
On these grounds, they consequently ignored the pipeline methodology refers only to strike-slip faults and relies on
bending stiffness, claiming that even when the above partitioning of the pipeline into four (4) distinct seg-
criteria are not met, their analysis would overestimate the ments (Fig. 1): two (2) in the high curvature zone on both
pipeline curvature and the associated bending strains. sides of the fault trace, and another two (2) outside this
It becomes evident that the Kennedy et al. criteria are zone. The latter segments are treated as beams-on-elastic-
met only when the pipeline is subjected to large fault foundation, while the former ones are assumed to de-
movements and is able to undergo large tensile strains form as circular arcs, with a radius of curvature calculated
without rupture. However, in practice, special construction from the equations of equilibrium and the demand for
measures (e.g. stringent welding procedures, special inspec- continuity between adjacent segments. In this way, the
tion and confirmation through laboratory tests) are bending moment at the conjunction of each arc with its
required in order to satisfy this condition. In common neighboring elastic beam can be readily calculated and
cases, where such measures have not been approved, the consequently compared to the pipeline’s ultimate mo-
maximum allowable strain is seriously reduced compared ment capacity, in order to estimate a factor of safety
to that of the pipeline steel, in order to account for thermal against failure. In addition, a second factor of safety is
effects and metallurgical alterations induced by welding. calculated at the point of intersection with the fault trace,
For such strain levels (e.g. 0.5%), which are well out of the as the ratio of the pipeline ultimate strain over the
range specified by the Kennedy et al. criteria, analytically corresponding axial strain.
 ARTICLE IN PRESS
202 D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211

Beam on Elastic Transition Beam on Elastic relationship for the pipeline steel, combined with an
Foundation Zone Foundation iterative linear elastic solution scheme which uses the
secant Young’s modulus of the pipeline steel in order to
q Initial Pipe Position ensure compatibility between computed non-linear stresses
B A' ∆x
fp β and strains.
fs The strike-slip fault is taken as an inclined plane, i.e.
q pp ∆f ∆y
A with null thickness of rupture zone, so that the intersection
fs q of the pipeline axis with the fault trace on the ground
Pipe A Fault D E surface is reduced to a single point. The fault movement is
FA q
VA Displaced Pipe defined in a Cartesian coordinate system, where the x-axis
is collinear with the undeformed longitudinal axis of the
Anti-symmetrical Configuration
pipeline, while the y-axis is perpendicular to x in the
Fig. 1. Pipeline analysis model proposed by Wang and Yeh [7]. horizontal plane (Fig. 2). Subsequently, the fault move-
ment is analyzed into two Cartesian components, Dx and
Dy, interrelated through the angle formed by the x-axis and
Although clearly advanced, compared to the methodol- the fault trace (angle b in Fig. 2). In its present form, the
ogy of Kennedy et al. [4], the methodology of Wang and proposed method applies to crossing angles bp901,
Yeh also features some pitfalls. Namely: resulting in pipeline elongation.
Following the general concept originally introduced by

The axial force is merely taken into account for reducing Wang and Yeh [7], the pipeline is partitioned into four (4)
the pipeline’s ultimate moment capacity, while its segments, defined by the characteristic points A, B and C in
unfavorable contribution to bending stiffness is over- Fig. 3: Point B is the intersection of the pipeline axis with
looked. the fault trace, while points A and C are the closest points

The most unfavorable combination of axial and bending of the pipeline axis with zero y displacement. Then, the
strains does not necessarily develop at the end of the computation of combined axial and bending pipeline
high-curvature zone (points B and D in Fig. 1) as the strains proceeds in six (6) steps:
method assumes, but within this zone, closer to the
intersection with the fault trace (point A in Fig. 1).
1. Segments AA0 and CC0 are analyzed as beams-on-

The calculation of a safety factor in terms of bending
elastic-foundation in order to obtain the relation
moments may be misleading for displacement-controlled
between shear force, bending moment and rotation
problems, such as the one at hand, where strain or
angle at points A and C.
deformation acceptance criteria are more appropriate
2. Considering the boundary conditions determined in step
[3].
1, segments AB and BC are analyzed according to the
elastic-beam theory in order to derive the maximum
Based on the well established concepts inherited by
bending moment.
existing methodologies, such as the equations used by
3. The axial force on the pipeline, at the intersection with
Kennedy et al. [4] to quantify the effect of axial tension on
the fault trace (point B), is obtained by equalizing the
the pipeline curvature and the partitioning of the pipeline
required and the available pipeline elongation.
into four (4) segments first introduced by Wang and Yeh
4. Bending strains are calculated, accounting for geometric
[7], the proposed methodology attempts to eliminate the
second-order effects.
abovementioned setbacks by introducing a number of
5. The maximum pipeline strain is computed from the
critical refinements. Its validity and its range of application
demand for equilibrium between the externally applied
are evaluated through comparison with typical analytical
axial force and the internal stresses developing on the
predictions obtained with the methodologies of Kennedy et
pipeline cross-section.
al. [4] and Wang and Yeh [7], as well as with more accurate
numerical predictions, based on the 3D non-linear Finite
Element Method.
t
faul

2. Methodology outline
pipeline

In more detail, the proposed methodology computes y

axial and bending strains along the pipeline with the aid of ∆y
beam-on-elastic-foundation and elastic-beam theories, tak- x
ing into account the bending stiffness of the pipeline cross- β
section, as well as the soil-pipeline interaction effects in ∆x
both the axial and the transverse directions. Material non-
linearity is considered by assuming a bilinear stress–strain Fig. 2. Definition of axes x and y and fault displacements Dx and Dy.
 ARTICLE IN PRESS
D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211 203

Lc Lc
A'
A
B

∆y
C
C'
pipeline

fault
q(x)= - k w(x)
Lc qu
A'
A

δ=∆y/2
x Cr
w
B

δ=∆y/2
w
Cr x
C
qu C'
Lc
q(x) = - k w(x)

Fig. 3. Partitioning of the pipeline into four segments.

6. Finally, given the stress and strain distribution within where

the cross-section, an updated secant Young’s modulus is sffiffiffiffiffiffiffiffiffiffiffi
computed and steps 2–6 are repeated until convergence 4 k
l¼ , (1c)
is accomplished. 4E 1 I
where x is the distance from point A along the pipeline axis,
As in all similar design methodologies [4,6,7], the above w is the transverse horizontal displacement, E1 is the elastic
solution algorithm overlooks initial pipeline stresses due to Young’s modulus of the pipeline steel, I is the moment of
soil overburden, as they are fairly small compared to the inertia of the pipeline cross-section, and k is the constant of
ones developing due to fault movement. Furthermore, the transverse horizontal soil springs (Fig. 4a). According
pipeline and soil inertia effects are not taken into account to the ALA-ASCE [3] guidelines, computation of k can be
as the velocity of the sliding part of the fault is considered based on Hansen [11] and Trautmann and O’Rourke [12].
to be sufficiently small, while concurrent spatially variable Differentiation of Eq. (1a) yields the following relations
transient motion is neglected in the computations. Finally, between the shear force V A ¼ E 1 Iw000 A , the bending
it should be stressed that the proposed methodology moment M A ¼ E 1 Iw00A and the rotation fA ¼ w0A at
overlooks local buckling and section deformation effects point A:
[8], two phenomena which dominate pipeline behavior at M A ¼ ð2lE 1 IÞfA , (2)
large fault displacements. As a result, its application range
is limited within the allowable strains which are explicitely V A ¼ lM A . (3)
defined by design codes [3,9] in order to avoid such
Due to symmetry, similar relations apply for point C.
detrimental effects. For larger strain levels, rigorous
numerical methods, or large displacement approximate
3.2. Step 2
methods [10] should be used.
Due to symmetry, the analysis can be focused on
segment AB. This segment is modeled as an elastic beam,
3. Solution algorithm supported at point A by a rotational spring, whose
constant is calculated from Eq. (2) as C r ¼ 2lE 1 I, and at
3.1. Step 1 point B by a joint, which is displaced by half the transverse
component of the strike-slip fault movement, i.e. d ¼ Dy=2.
The differential equilibrium equation for the elastic line As a result, a uniformly distributed load qu is applied to the
of segment AA0 (Fig. 3) is: beam, equal to the limit value of soil reaction for transverse
E 1 Iw0000 þ kw ¼ 0. (1a) horizontal movement of the pipeline relatively to the
surrounding soil (Fig. 4b). According to the ALA–ASCE
Imposing w ¼ 0 for x ¼ 0 and w ! 0 for x ! 1, Eq. (1) [3] guidelines, the value of qu can be calculated from the
yields: properties of the backfill of the pipeline trench, based on
the relations proposed by Hansen [11] and Trautmann and
w ¼ Celx sin lx, (1b) O’Rourke [12].
 ARTICLE IN PRESS
204 D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211

q(x) = - k w(x)

VA

x ϕA
MA
A' A

(a)

Lc

Cr
ϕA
MA
A

δ=∆y/2
VA
B

qu

VB

VA

xmax

Mmax
MA

(b)

Fig. 4. Analysis model for pipeline segments 1 (a) and 2 (b).

Application of the elastic-beam theory yields the 24EIdC r þ 12EIqu L3c þ 3qu C r L4c
following bending moment and shear force reactions on VB ¼ . (6)
24EIL2c þ 8C r L3c
the supports A and B:
Eqs. (4)–(6) express the reaction forces of segments AB
24EIdC r  qu C r L4c and BC in terms of the curved length Lc of the beam, which
MA ¼ , (4)
24EILc þ 8C r L2c is not a priori known. However, substituting Eqs. (4) and
(5) into Eq. (3) yields:
24EIdC r  12EIqu L3c  5qu C r L4c
VA ¼ , (5)
24EIL2c þ 8C r L3c a5 L5c þ a4 L4c þ a3 L3c  a1 Lc  a0 ¼ 0, (7a)
 ARTICLE IN PRESS
D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211 205

where a0–a5 are known constants, equal to: where

a0 ¼ 24EIdC r , VB
xmax ¼ . (8b)
qu
a1 ¼ 24EIdC r l,
a3 ¼ 12EIqu ,
a4 ¼ 5qu C r ,
a5 ¼ qu C r l. ð7bÞ
3.3. Step 3
The above polynomial equation can be solved iteratively,
Similar to all existing methodologies [4,6,7], the axial
using the Newton–Raphson method, with a large initial
force at the intersection of the pipeline with the fault trace
value for Lc (e.g. 500 m). In this way, the values of MA, VA,
is calculated from the requirement for compatibility
and VB can be directly estimated, and the maximum
between the geometrically required and the stress-induced
bending moment developing on the pipeline can be
(available) pipeline elongation. The required elongation
consequently calculated from the elastic-beam theory, as:
DLreq is defined as the elongation imposed to the pipeline
qu x2max due to the fault movement. For the sake of simplicity, the
M max ¼ V B xmax  , (8a) elongation provoked by the Dy fault displacement compo-
2
nent may be neglected, as it is minimal compared to the
elongation due to the Dx component. Therefore:
DLreq  Dx. (9)
σ
On the other hand, the available elongation DLav is
E2 defined as the elongation resulting from the integration of
σ1 1 axial strains along the unanchored length, i.e. the length
over which slippage occurs between the pipeline and the
surrounding soil:
Z Lanch
DLav ¼ 2
ðLÞ dL, (10)
0
E1
where L is the distance from the fault trace, while the factor
2, by which the integral in Eq. (10) is multiplied, accounts
1 for the elongation on both sides of the fault trace.
The unanchored length Lanch may be calculated from the
ε1 equilibrium along the pipeline axis, assuming that axial
ε
pipeline stresses essentially become zero at the far end of
Fig. 5. Assumed bilinear stress-strain relationship for the pipeline steel. the unanchored length. In this way, Lanch is expressed as:
ε(L)

σa(point B) <σ1 σa(pointB) >σ1 ε(L)

ε1

L L
σ(L)

σ(L)

σ1

L L Lanch - L1
Lanch L1
Lanch

tu B Fa tu B
Fa
(a) (b)

Fig. 6. Linear (a) and non-linear (b) stress and strain variation along the pipeline’s unanchored length.
 ARTICLE IN PRESS
206 D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211

then plastic strains develop in the pipeline (Fig. 6b) and Eq.

dϕ
/2
Fa (10) becomes:
Z L1   Z Lanch 
sðLÞ  s1 sðLÞ
DLav ¼ 2
1 þ dL þ dL ,
0 E2 L1 E1
(16)
where
qu
ðsa  s1 ÞAs
L1 ¼ . (17)
tu
R

Combining Eq. (11), (12), (16) and (17), the maximum

developing tensile stress becomes:
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
tu

dϕ
s1 ðE 1  E 2 Þ þ s21 ðE 22  E 1 E 2 Þ þ E 21 E 2 DLreq

/2
R
As
sa ¼ .
E1
Fa (18)
dϕ

Regardless of the axial stress level, the corresponding

axial force is equal to:
F a ¼ sa A s . (19)
Fig. 7. Effect of the applied axial force on the pipeline curvature.

F a sa As 3.4. Step 4
Lanch ¼ ¼ , (11)
tu tu
According to the elastic beam theory, bending strains on
where Fa and sa are the axial force and stress developing in the pipeline can be calculated as:
the intersection of the pipeline axis with the fault trace, As
M max D
is the area of the pipeline cross-section and tu is the limit
Ib ¼ , (20)
friction due to the slippage of the pipeline relatively to the 2EI
surrounding soil [3]. where D is the external pipeline diameter.
The distribution of strains along the unanchored length The above equation is accurate for small fault displace-
can be consequently derived from the corresponding ments, while for larger fault displacements, geometrical
stresses, assuming a bilinear stress–strain relationship for second-order effects must be also taken into account. To
the pipeline steel (Fig. 5). Assuming further that axial simplify this relatively complex problem, the bending
stresses attenuate linearly with the distance from the fault stiffness of the pipeline may be approximately neglected
trace, due to the constant value of the limit friction force, [4], so that bending strains can be computed geometrically
tensile stresses along the pipeline axis can be expressed as: as:
tu D=2
sðLÞ ¼ sa  L. (12)
II
b ¼ . (21)
As R
For the case where the axial tensile stress sa is below the The radius of curvature R results from the equilibrium of
yield limit s1 (Fig. 6a), Eq. (10) can be re-written as: the forces acting on an infinitesimal part of the pipeline’s
Z Lanch curved length (Fig. 7):
sðLÞ s2 As
DLav ¼ 2 dL ¼ a . (13) Fa
0 E1 E 1 tu R¼ (22)
qu
Therefore, for DLav ¼ DLreq , the maximum tensile stress
becomes: and finally:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qu D
E 1 tu DLreq
II
b ¼ . (23)
sa ¼ . (14) 2F a
As
Eq. (23) indicates that bending strains induced by
If the required elongation is larger than the one second-order effects are inversely proportional to the axial
corresponding to sa ¼ s1 , i.e. when: force applied on the pipeline. In other words, for small
fault displacements, bending strains computed by Eq. (23)
s21 As tend to become infinite. This is due to the fact that Eq. (22)
DLreq 4 (15)
E 1 tu has been derived assuming that the pipeline bending
 ARTICLE IN PRESS
D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211 207

εmax = εa+εb σmax

θ ε1 σ1

ϕ1 εa σa

π-ϕ2

-ε1 -σ1

εmin = εa-εb σmin

Fig. 8. Non-linear stress and strain distribution on the pipeline cross-section.

stiffness is equal to zero, which is approximately true only Bearing in mind the above, the strain distribution on the
when the whole pipeline cross-section is under yield. cross-section is given by Eq. (25):
According to the above, the actual bending strain eb lays
¼
a þ
b cos y, (25)
between
Ib and
II I
b , asymptotically approaching
b as fault
II
displacements tend to zero and
b in the opposite case. where the angle y is the polar angle of the cross-section,
Here, it is approximately assumed that: defined in Fig. 8. The corresponding distribution of stresses
on the pipeline cross-section is given by Eq. (26):
1 1 1 8
¼ þ . (24) s þ E 2 ð

1 Þ; 0pyof1 ;

b
Ib
II
b
>
< 1
s ¼ E 1
; f1 pypp  f2 ; (26)
>
: s þ E ð
þ
Þ; p  f oypp;
1 2 1 2
3.5. Step 5 where the angles j1,2 define the portion of the cross-section
that is under yield (Fig. 8), and are calculated as:
Existing methodologies [4,6,7] calculate the axial strain 8
1 
a
directly from the axial stress, using the adopted stress–- >
> p; o  1;
>
>
b
strain relation for the pipeline steel. When strains in the >
>  
<
1 
a
1 
a
pipeline cross-section remain in the elastic range, this ; 1p p1;
f1;2 ¼ arccos (27)
assumption is valid for every point along the pipeline axis. >
>
b
b
>
>
1 
a
However, when yielding occurs, it is accurate only at the >
>
: 0; 1o :
intersection of the pipeline with the fault trace, where the
b
bending strain is zero. In the vicinity of the cross-section
where the maximum bending strain occurs, axial strain The total axial force is calculated by integrating the
increases locally, so that the integral of stresses on the stresses over the cross-section, as:
Z p
pipeline cross-section remains equal to the applied axial
F ¼2 sRm t dy
force. 0
This effect of curvature on the relation between the 
) F ¼ 2Rm t E 1 p
a  ðE 1  E 2 Þðf1 þ f2 Þ
a
applied axial force and the corresponding axial strain was
þ ðE 1  E 2 Þðf1  f2 Þ
1
in fact acknowledged by Kennedy et al. [4], but the 
Ramberg–Osgood stress–strain curve, which they adopted ðE 1  E 2 Þðsin f1  sin f2 Þ
b , ð28Þ
to approximate steel behavior, inhibited the derivation of a where
simple relation to quantify this effect. Therefore, Kennedy
et al. were limited to performing a numerical investigation Dt
Rm ¼ . (29)
of the strain range for which this interaction could be 2
ignored. The axial strain of the pipeline can be derived from the
In the present work, the interaction between axial and demand for equilibrium, by equating the axial force
bending strains is quantified by determining the exact computed using Eq. (28), to the one calculated using Eq.
distribution of strains and stresses on the pipeline cross- (19). Note that Eq. (19) applies strictly at the pipeline’s
section. For this purpose, the beam-theory assumption of intersection with the fault trace, but it is approximately
plane cross-sections is embraced, while the stress–strain extended to the neighboring position of the maximum
curve of the pipeline steel is considered to be bilinear bending moment. The solution of the system of Eqs. (27)
(Fig. 5). and (28) results in a complex formula for ea, which can be
 ARTICLE IN PRESS
208 D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211

Table 1
API5L-X65 steel properties considered in the numerical analyses

Yield stress (s1) 490 MPa

Failure stress (s2) 531 MPa
Shell elements Failure strain (e2) 4.0%
Elastic Young’s modulus (E1) 210 GPa
Rigid element Yield strain (
1 ¼ s1 =E 1 ) 0.233%
Plastic Young’s modulus (E 2 ¼ ðs2  s1 Þ=ð
2 
1 Þ) 1.088 GPa
Beam elements

Table 2
Parameters of the Ramberg–Osgood stress–strain curve for steel type
API5L-X65

Initial Young’s modulus (E1) 210 GPa

Yield stress (sy) 490 MPa
a 38.32
r 31.50
Axial soil springs
Horizontal soil springs
Table 3
Vertical soil springs Soil spring properties considered in the numerical analyses

Fig. 9. 3D model used for analyses with the Finite Element Method: the Yield Yield
junction between shell and beam part. force displacement
(kN/m) (mm)

solved iteratively, using the Newton–Raphson method. Axial (friction) springs 40.5 3.0
Namely, using an initial value of
0a ¼ 0, the axial strain on Transverse horizontal springs 318.6 11.4
Vertical springs (upward movement) 52.0 2.2
each iteration can be calculated as:
Vertical springs (downward movement) 1360.0 100.0

F ð
ka Þ  F a

kþ1 ¼
ka   , (30)
a
dF =d
a 
a ¼
k 3.6. Step 6
a

where The analysis of segments AB and BC, from which the

 maximum bending moment emerges, is based on the elastic
dF beam theory and is not taking into account the non-linear
¼ 2Rm t E 1 p  ðE 1  E 2 Þðf1 þ f2 Þ  ðE 1  E 2 Þ
d
a behavior of the pipeline steel. Since the steel stress–strain
   
df1 df2 df1 df2 relationship is considered to be bilinear (Fig. 5), a series of
 þ
a þ ðE 1  E 2 Þ 
1 equivalent linear calculation loops is performed, employing
d
a d
a d
a d
a
   a procedure for readjusting the secant Young’s modulus of
df1 df
ðE 1  E 2 Þ cos f1  2 cos f2
b ð31Þ the pipeline steel on each loop.
d
a d
a
More specifically, using the already defined stress
and distribution on the pipeline cross-section, the correspond-
ing bending moment can be calculated using Eq. (33):
8
> 1 Z p
>
> 
sin f ;
b sin f1;2 p  0:01;
>
> M¼2 sRm tRm cos y dy
>
>
b 1;2
>
< 100; 0:01o
b sin f1;2 p0;
0

df1;2 E 1 p
b
¼ (32) 2
d
a >
> 100; 0o
b sin f1;2 p0:01; ) M ¼ 2Rm t
2
 ðE 1  E 2 Þðsin f1  sin f2 Þ
a
>
>
>
> 1
>
> ðE 1  E 2 Þðf1 þ f2 Þ
b
: 
sin f ; 0:01o
b sin f1;2 : þ ðE 1  E 2 Þðsin f1 þ sin f2 Þ
1 
b 1;2 2

ðE 1  E 2 Þðsin 2f1 þ sin 2f2 Þ
b
The limits of df1;2 =d
a proposed in Eq. (32) are more or  . ð33Þ
4
less arbitrary, but significantly accelerate the iterative
process. Therefore, the secant modulus for the next iteration can
Having calculated the axial strain at the position of be calculated as:
maximum bending strain, the minimum and maximum  
Mð
a ;
b ÞD Mð
a ;
b ÞD 1 1
longitudinal strains can be subsequently computed as their 0
E sec ¼ ¼  (34)
algebraic sum (i.e.
max;min ¼
a 
b ). 2I
Ib 2I
b
IIb
 ARTICLE IN PRESS
D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211 209

β=30° β=45° β=60°

3
Kennedy et al. [4]
Wang & Yeh [7]
2
εa,fault (%)

Proposed Method
Numerical
1

2
εa,max (%)

2
εb (%)

2
εmax (%)

0
0 0.5 1 1.5 2 0 0.5 1 1.5 2 0 0.5 1 1.5 2
∆f / D ∆f / D ∆f / D

Fig. 10. Comparison of the results of the proposed analytical methodology with the results of the numerical analyses and the predictions of the Kennedy
et al. [4] and Wang and Yeh [7] methods.

and steps 2–6 are repeated, until convergence is accom- trace (i.e. a total length of 100 m) modeled as a cylindrical
plished. shell, and the remaining 450 m part (i.e. a total length of
900 m) modeled as a beam (Fig. 9). The shell perimeter was
4. Validation of the proposed methodology discretized into 16 equal sized quadrilateral shell elements,
each of 0.20 m length. CQUAD4 type elements were used,
To validate the results of the proposed methodology, namely isoparametric quadrilateral shell elements, with
analytical predictions are compared to the results from a both bending and membrane stiffness [13]. The beam part
series of 3D non-linear numerical analyses with the Finite was discretized with 0.50 m long CBEAM type beam
Element Method, performed with the commercial code elements, with extension, bending and shearing stiffness
MSC/NASTRAN [13]. For this purpose, a typical high- [13].
pressure natural gas pipeline was considered, featuring an To simulate soil-pipeline interaction effects, each node of
external diameter of 0.9144 m (36 in), a wall thickness of the model was connected to axial, transverse horizontal
0.0119 m (0.469 in), and a total length of 1000 m. and vertical soil springs, modeled as elastic-perfectly plastic
A hybrid model was used for the simulation of the CROD type rod elements [13]. Thus, the Finite Element
pipeline, with a part of 50 m along both sides of the fault Model used herein consisted of a total number of 32,252
 ARTICLE IN PRESS
210 D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211

nodes and 32,236 elements and had 58,710 degrees of displacements, after the yield strain of the pipeline steel has
freedom. been exceeded, while they grossly under-predict
a;max at
The pipeline steel was of the API5L-X65 type, with a smaller displacements. Note that observed differences are
bilinear elasto-plastic stress–strain curve (Fig. 5) and the larger at intermediate displacement levels, of the order of
properties listed in Table 1. Note that the numerical Df  1D, which are commonly encountered in practice.
analyses were also performed using an equivalent Ram- Focusing next on bending strains eb, a good overall
berg–Osgood stress–strain curve, i.e.: agreement is observed again between the proposed analytical
   r  method and the numerical analyses. The Kennedy et al. [4]
s a jsj

¼ 1þ (35) method proves accurate in the region of large displacements
Ei rþ1 sy
(Df =D41:5), i.e. when the criteria for the applicability of the
and the properties listed in Table 2. However, they showed method are met, but seriously over-predicts eb for smaller
negligible divergence from the bi-linear elasto-plastic fault displacements. The Wang and Yeh [7] method under-
analyses and are not presented herein. predicts eb for the entire range of fault displacements analyzed
The properties of the soil-springs (Table 3) were herein, mainly because it neglects the effect of axial tension on
calculated according to the ALA–ASCE [3] guidelines, the pipeline bending stiffness.
assuming that the pipeline top is buried under 1.30 m of The maximum longitudinal strains
max ¼
a;max þ
b are
medium-density sand with friction angle j ¼ 361 and unit probably the best criterion for the evaluation of the
weight g ¼ 18 kN=m2 . The fault movement was applied proposed methodology, as they form the basis of pipeline
statically at the sliding part of the fault, as a permanent design. From the last row of Fig. 10, it may be observed
displacement of the free end of the corresponding soil- that the good overall performance of the proposed method,
springs. acknowledged in the previous comparisons, applies here as
The results of the numerical analyses are presented in well. As expected, the Kennedy et al. [4] method over-
Fig. 10 in comparison with analytical predictions of the predicts maximum strains for small fault displacements.
proposed methodology, as well as the analytical methodol- This trend is reversed at intermediate levels of fault
ogies of Kennedy et al. [4] and Wang and Yeh [7]. The displacement, and the divergence is gradually reduced as
comparison is shown in terms of: displacements increase. Finally, the methodology of Wang
and Yeh [7] provides accurate results only in the region of

the axial strain at the intersection of the pipeline with large displacements, where axial tension is the prevailing
the fault trace, mode of deformation. For small and intermediate fault

the maximum axial strain, displacements emax is consistently under-estimated.

the bending strain and

the maximum total strain. 5. Conclusion

Three (3) different fault cases are examined, with An improved analytical methodology has been devel-
intersection angles of b ¼ 301, 451 and 601. In each case, oped for the stress analysis of buried steel pipelines
the analysis proceeded incrementally to a final fault crossing active strike-slip faults. It is based on firm
displacement Df ¼ 2D, with D being the pipeline’s external assumptions adopted in the existing analytical methodol-
diameter. Note that bending strains corresponding to the ogies of Kennedy et al. [4] and Wang and Yeh [7], but
Wang and Yeh [7] method were derived from Eq. (21), proceeds further:
consistently with the assumption that pipeline segments AB
and BC deform as circular arcs (Fig. 1).
to analyze the curved part of the pipeline with the aid of
The top row of Fig. 10 shows the comparison between elastic-beam theory, in order to locate the most
analytical and numerical predictions of axial strain
a;fault at unfavorable combination of axial and bending strains,
the pipeline-fault trace intersection. The agreement appears and
fairly good, for all analytical methods. As the main
to consider the actual stress distribution on the pipeline
assumption for computing
a;fault is the compatibility cross-section, in order to account for the effect of
between the geometrically required and the stress induced curvature on axial strains and calculate the design
(available) elongation of the pipeline, the observed agree- maximum strain.
ment is essentially considered as a solid verification of this
assumption. Comparison with the results of benchmark numerical
The second row of Fig. 10 refers to the overall maximum analyses, performed over a wide range of fault displace-
axial strain
a;max , which does not necessarily develop at the ments (Df =D ¼ 0C2) and three different intersection
pipeline-fault trace intersection, as the existing analytical angles (b ¼ 301, 451 and 601), showed a remarkable overall
methods imply. In this case, a good overall agreement is agreement, with minor deviations which did not exceed
observed only between numerical results and analytical about 10%.
predictions with the proposed methodology. The existing Acknowledging that there is no end to the refinements that
analytical methods approach the numerical solution at large can be applied to simplified analytical methodologies, it needs
 ARTICLE IN PRESS
D.K. Karamitros et al. / Soil Dynamics and Earthquake Engineering 27 (2007) 200–211 211

to be stressed out that the above modifications considerably [2] Uzarski J, Arnold C. Chi-Chi, Taiwan, Earthquake of September 21,
improve the accuracy of analytical predictions, especially for 1999, Reconnaissance Report. Earthquake Spectra, Professional
J EERI 2001;17(Suppl. A).
small and medium fault displacements, while they still permit a
[3] American Lifelines Alliance—ASCE. Guidelines for the Design
simple analytical solution algorithm to be developed. In fact, of Buried Steel Pipe, July 2001 (with addenda through February
although more complicated than the most commonly used 2005).
today method of Kennedy et al. [4], the computational [4] Kennedy RP, Chow AW, Williamson RA. Fault movement effects on
algorithm of the proposed methodology remains relatively buried oil pipeline. Transport Eng J ASCE 1977;103:617–33.
simple and stable, and can be easily programmed for quick [5] ASCE Technical Council on Lifeline Earthquake Engineering.
Differential Ground Movement Effects on Buried Pipelines. Guide-
application. For instance, such a computer code may be down- lines Seismic Des Oil Gas Pipeline Syst 1984:150–228.
loaded from http://users.civil.ntua.gr/gbouck/en/publications. [6] Newmark NM, Hall WJ. Pipeline design to resist large fault
htm. displacement. In: Proceedings of the US National Conference on
Note that, in its present form, the proposed method Earthquake Engineering. Ann Arbor: University of Michigan; 1975.
applies to intersection angles bp901 resulting in elongation p. 416–25.
[7] Wang LRL, Yeh Y. A refined seismic analysis and design of buried
of the pipeline. Furthermore, it does not account for the pipeline for fault movement. Earthquake Eng Struct Dyn
effects of local buckling and section deformation. There- 1985;13:75–96.
fore, its application should not be extended beyond the [8] Calladine CR. Theory of shell structures. Cambridge: Cambridge
strain limits explicitely defined by design codes in order to University Press; 1983.
[9] CEN European Committee for Standardisation, Eurocode 8:
mitigate such phenomena.
Design of structures for earthquake resistance, Part 4: Silos, tanks
and pipelines, Draft No 2, Ref. No. EN1998-4: 2003 (E), December
Acknowledgements 2003.
[10] Takada S, Hassani N, Fukuda K. A new proposal for simplified
design of buried steel pipes crossing active faults. Earthquake Eng
This research is supported by the ‘‘EPEAEK II— Struct Dyn 2001;30:1243–57.
Pythagoras’’ Grant, co-funded by the European Social [11] Hansen JB. The ultimate resistance of rigid piles against transversal
Fund and the Hellenic Ministry of Education. forces. Bulletin 12. Copenhagen, Denmark: Danish Geotechnical
Institute; 1961.
[12] Trautmann CH, O’Rourke TD. Behavior of pipe in dry sand under
References lateral and uplift loading. Geotechnical Engineering Report 83-6.
Cornell University, Ithaca, New York, 1983.
[1] EERI. The Izmit (Kocaeli), Turkey Earthquake of August 17, 1999. [13] The MacNeal—Schwendler Corporation. MSC/NASTRAN for
EERI Special Earthquake Report, 1999. Windows: Reference Manual, 1994.