Mechanical Behavior of Steel
Pipe Bends: An Overview
Spyros A. Karamanos
Department of Mechanical Engineering,
University of Thessaly,
Volos 38334, Greece
e-mail: skara@mie.uth.gr
1
An overview of the mechanical behavior of steel pipe (elbows) is offered, based on previously reported analytical solutions, numerical results, and experimental data. The behavior of pipe bends is characterized by significant deformations and stresses, quite higher
than the ones developed in straight pipes with the same cross section. Under bending
loading (in-plane and out-of-plane), the main feature of the response is cross-sectional
ovalization, which influences bending capacity and is affected by the level of internal
pressure. Bends subjected to cyclic in-plane bending exhibit fatigue damage, leading to
base metal cracking at the elbow flank. Using advanced finite-element tools, the response
of pipe elbows in buried pipelines subjected to ground-induced actions is also addressed,
with emphasis on soil–pipeline interaction. Finally, the efficiency of special-purpose
finite elements for modeling pipes and elbows is briefly discussed.
[DOI: 10.1115/1.4031940]
Introduction
Pipe bends, often referred to as elbows, are curved pipe parts
widely used in piping systems of industrial plants or power stations (Fig. 1(a)). Their mechanical behavior, compared with
straight pipe segments, is significantly more flexible and associated with significantly higher stresses and strains, and very pronounced cross-sectional deformation, referred to as “ovalization.”
Because of their flexibility, they can accommodate thermal expansions and absorb other externally induced loading, but they are
considered as critical components for the structural integrity of
piping systems. For the case of extreme loading conditions, their
mechanical response is characterized by a biaxial state of stress
and strain, which may lead to pipe elbow failure, in a mode quite
different than the one expected in straight pipes.
Pipe elbows in industrial piping systems have relatively small radius of bend curvature R with respect to pipe diameter D (Fig. 2).
In ASME 16.9 standard [1], pipe elbows of diameter ranging from
1=2 to 48 in. are specified, and the ratio of the longitudinal radius R
over the pipe diameter D is equal to either 1 (“short radius” bends)
or 1.5 (“long radius” bends). Similar pipe elbow dimensions are
specified in the relevant European standard EN 10253-1.
For the case of large-diameter steel hydrocarbon transmission
pipelines, pipe elbows or bends are used to accommodate the pipe
in the direction of pipeline alignment (Fig. 1(b)). Large-diameter
pipeline bends can be either “hot” or “cold” bends. Cold pipe
bends, sometimes referred to as “field bends,” are manufactured at
the construction site from straight pipes, with the use of special
bending devices and are commonly employed in pipeline construction practice. For typical large-diameter pipes (i.e., from 36
to 56 in.), the R/D ratio of the bend radius R over the pipe diameter D may range between 30 and 45. In the case of abrupt changes
of pipeline direction, small radius “hot bends” are also employed,
usually manufactured through induction heating and bending; a
typical value of the R/D ratio for those hot bends is equal to 5,
which allows for pigging inspection operations in the course of
pipeline maintenance.
The present paper offers an overview of the response of steel
pipe bends (elbows) under structural loading, in the presence of
pressure, in an attempt to identify and summarize their main features of mechanical behavior and the corresponding principal failure modes under various loading conditions. In Sec. 2, some
important features stemming from the small-strain elastic analysis
of pipe elbows subjected to in-plane and out-of-plane bending are
Contributed by the Pressure Vessel and Piping Division of ASME for publication
in the JOURNAL OF PRESSURE VESSEL TECHNOLOGY. Manuscript received July 21, 2015;
final manuscript received October 11, 2015; published online April 28, 2016. Assoc.
Editor: Chong-Shien Tsai.
Journal of Pressure Vessel Technology
presented. The behavior of pipe bends under severe monotonic
loading is presented in Sec. 3, whereas in Sec. 4 the response of
pipe bends under strong cyclic loading is discussed; in both sections, reference to experimental results is made. The particular
case of bends in buried pipelines under ground-induced actions is
briefly discussed in Sec. 5, addressing the issue of soil–pipeline
interaction and presenting some novel results. In Sec. 6, an
overview of special-purpose elements is described, suitable for
the efficient modeling of elbows in piping systems. Finally, some
important conclusions from the previous sections are summarized
in Sec. 7.
2
Small-Strain Elastic Behavior
The stress analysis of piping systems and pipelines constitutes
an essential part of pipe design procedure under operating loads.
In this design process, elbows (bends) are considered as critical
components, exhibiting significantly higher stresses, deformations, and bending flexibility than straight pipes of the same crosssectional properties. The configuration of a typical elbow, widely
employed in piping systems, is shown in Fig. 2. The geometry is
characterized by the pipe diameter D, the pipe thickness t, the
bend radius R, and is usually expressed through by the dimensionless parameters k or h, also defined in Fig. 2. Parameter k arises in
analytical calculations, as the ones described briefly below, and
parameter h is often employed in design standards, such as the
ASME B31 standards [2,3] or EN 13480-3 [4].
The early analytical work of Von Karman [5] can be used for
understanding the particular mechanical behavior of pipe bends.
Von Karman, employed a simple two-dimensional formulation
that couples longitudinal strain energy due to bending, with hoop
strain energy due to ovalization, assumes inextensional ring
theory, and considers a simple doubly symmetric trigonometric
function for the radial displacement of arbitrary point of the elbow
cross section w, as shown in Fig. 2
wðhÞ ¼ a cos 2h
(1)
Minimization of the strain energy results in the following expression for the moment–curvature relationship in the absence of
internal pressure, written below in a nondimensional form
m ¼ pj 1
9
10 þ 12k2
(2)
where m and j are the dimensionless values pof
moment
ffiffiffiffiffiffiffiffiffiffiffiffiffi
M and curvature k, normalized by Me ¼ Ert2 = 1 2 and
C 2016 by ASME
Copyright V
AUGUST 2016, Vol. 138 / 041203-1
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 1 Pipe elbows (bends) in (a) piping systems of industrial plants and (b) buried
gas pipelines
pffiffiffiffiffiffiffiffiffiffiffiffiffi
kN ¼ t= ðr 2 1 2 Þ, respectively. The elbow geometry is
expressed in terms of the dimensionless elbow parameter k (see
Fig. 2). From Eq. (2), one may readily obtain the flexibility factor
g, defined as the ratio of elbow bending flexibility over the flexibility of the corresponding straight pipe, as follows:
g¼ 1
9
10 þ 12k2
1
(3)
A more enhanced energy formulation can be employed to
describe in more detail the mechanical behavior of elastic elbows,
proposed by Rodabaugh and George [6], including the effects of
internal pressure. This formulation is a generalization of the Von
Karman [5] solution. It assumes uniform cross-sectional deformation along the elbow axis, and the total potential energy of the
bent and pressurized elbow is written in terms of the radial and
the tangential displacements of the arbitrary point A, denoted
w; v, respectively, as shown in Fig. 2, which are discretized
through series of doubly symmetric trigonometric functions and
the corresponding generalized coordinates an
vðhÞ ¼
1
X
wðhÞ ¼
an sin 2nh
(4)
n¼1
1
X
2an cos 2nh
(5)
n¼1
Solution is obtained by minimization of the potential energy,
and the results are expressed in terms of the flexibility factor g.
Furthermore, the longitudinal and hoop stresses and the ovalization of the cross section are also computed. Despite the fact that
this solution has been presented in the late 1950s, it still constitutes the basis of current pipe design specifications, such as the
ASME B31 standards [2,3] or EN 13480-3 [4]. The main outcome
of the above analyses is that the response is governed by crosssectional ovalization, shown schematically in Fig. 3; under inplane bending, the elbow flattens in an oval shape symmetric with
respect to the plane of bending for closing or opening bending.
However, the flattening pattern under opening moments is
opposite to the one that occurs under closing moments; the pattern
observed in opening moments is referred to as “reverse
ovalization” and plays a significant role in the mechanical
response of elbows. In the case of out-of-plane bending, crosssectional flattening occurs at 45-deg with respect to the plane of
bending.
Figure 4 presents the flexibility factor in terms of dimensionless
parameter k and the pressure level in the pipe elbow. Pressure is
expressed in terms of the dimensionless pressure parameter
041203-2 / Vol. 138, AUGUST 2016
w ¼ pR=Ert. The flexibility values of Fig. 4 indicate that the
curved pipe (elbow) is substantially more flexible than the corresponding straight pipe. The flexibility factor is higher in the
absence of internal pressure and is reduced when internal pressure
is raised. Figure 5 depicts the longitudinal and the circumferential
stresses around the pipe cross section, assuming elastic behavior
of the elbow material, as a function of the distance y from the pipe
axis. The elbow under consideration has outside diameter
D ¼ 165 mm, thickness t ¼ 3 mm, and R/D ratio equal to 3, subjected to a bending moment equal to 10 kN m. It is interesting to
note that the maximum circumferential stress is higher than the
maximum longitudinal stress. Furthermore, the maximum longitudinal stress is considerably higher than the maximum stress of a
straight pipe with the same cross section and does not occur at the
top or the bottom of the cross section. Cross-sectional ovalization
is the main reason for this behavior.
The above observations clearly show the special features of
pipe bend behavior; it is substantially different than the behavior
of a straight pipe, in terms of both stiffness and stress. One should
notice that the above analysis has several limitations: it assumes
elastic material response, small displacements, and constant curvature along the pipe. Those assumptions are discussed below.
In practical applications, elbows are connected to straight pipe
segments, resulting in a variation of deformation and ovalization
along the elbow, with the maximum deformation in the middle
cross section of the elbow, and the assumption of constant curvature along the bend is not applicable. In such a case, a numerical
simulation is required. Figure 6 shows a setup used for in-plane
and out-of-plane bending loading of a 90-deg pipe elbow, with
diameter D ¼ 160 mm and thickness t ¼ 3 mm. The response of
Fig. 2 Pipe elbow geometry, cross-sectional displacements,
and biaxial state of stress
Transactions of the ASME
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 3 Schematic representation of ovalization in (a) in-plane closing moments, (b) out-of-plane opening moments reverse
ovalization, and (c) out-of-plane bending
that considers geometric and material nonlinearities is necessary.
It is noted that an analysis of pipe elbows with only material nonlinearities may not be appropriate for obtaining reliable results;
cross-sectional distortion is quite significant and the ensuing geometric nonlinearities due to large displacements should be taken
into account in the analysis.
3
Fig. 4 Flexibility factor of pipe elbows with respect to the
elbow parameter k
this pipe, subjected to in-plane closing bending moments for different levels of pressure, is shown in Fig. 7. The results show
the increase of stiffness with increasing level of internal pressure, and the experimental results are compared quite successfully with numerical results from a finite-element model that
employs shell elements and simulates the experimental procedure of Fig. 6 [7].
Furthermore, the assumptions of elastic material behavior and
small displacements are valid only for low levels of loading, associated with operational loading conditions. In the case of severe
loading conditions, e.g., in the course of a strong seismic event,
the elbow behavior at the ultimate limit state is characterized by
large displacements and significant inelastic material deformations. To predict numerically the mechanical behavior of steel
elbows under severe loading conditions, a numerical simulation
Behavior Under Severe Monotonic Loading
Previous works (e.g., Refs. [7–9]) have pointed out that under
severe loading conditions, the development of significant crosssectional ovalization is associated with the development of significant
stresses and strains both in the circumferential and the longitudinal directions. In addition, it has been recognized that the response
under closing bending moments is substantially different than the
response under opening bending moments. This is mainly due to
the different signs of ovalization due to opening bending
moments, referred to as reverse ovalization, described in Section 2
and depicted in Fig. 3(b); this reverse ovalization increases crosssectional height and results in higher bending moment capacity.
Sobel and Newman [10,11] and Dhalla [12] reported experiments on the bending response of elbows through a series of tests
on 16-in. diameter 90 deg elbows (D/t ¼ 39 and R/D ¼ 1.5) under
in-plane closing moments. The test data were compared with
numerical results from shell elements and simplified elbow elements. Gresnigt et al. [13,14] reported test data on five 30 deg,
five 60 deg, and one 90 deg steel elbows (R/D ¼ 3) under bending
and pressure. The 30 deg and 60 deg specimens were tested under
inelastic in-plane bending, whereas the 90 deg specimen was subjected to out-of-plane bending. An analytical model for the
elastic–plastic cross-sectional deformation of elbows was also
developed by Gresnigt et al. [14] and Gresnigt and van Foeken
[15], introducing a correction factor to account for the influence
Fig. 5 Elastic analysis of a 90-deg standalone elbow (k 5 0.23); variation of (a) longitudinal stresses and (b) hoop stresses at
external pipe wall with respect to the cross-sectional height; comparison of the analytical solution of Rodabaugh and George
[6] with finite-element results
Journal of Pressure Vessel Technology
AUGUST 2016, Vol. 138 / 041203-3
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 6 (a) Experimental setup for testing a 160-mm-diameter
pipe elbow under in-plane and out-of-plane bending and (b)
schematic representation of the 90 deg specimens [7]
of the adjacent straight pipe segments. Greenstreet [16] investigated experimentally the response of carbon steel and stainless
steel pipe elbows, under in-plane and out-of-plane bending loading conditions, in the presence of internal pressure. Hilsenkopf
et al. [17] reported test data on thin-walled (D/t ¼ 89.5) stainless
steel elbows and thick-walled (D/t ¼ 13.4) ferritic elbows under
both in-plane and out-of-plane bending, in connection with their
functional capability. Suzuki and Nasu [18] conducted two inplane closing moment tests on a 12-in. 90 deg elbow (D/t ¼ 46.3)
and on a 24-in. 90 deg elbow (D/t ¼ 64.9) and compared the test
data with numerical predictions from four-node shell elements.
More recently, Tan et al. [19] reported one closing in-plane
moment test and one opening in-plane moment test on 90 deg
thick stainless steel elbows (D/t ¼ 10.5) and compared their measurements with finite-element analysis results.
The development of computational methods (e.g., finite elements) enabled the numerical investigation of elbow response and
the prediction of ultimate capacity. To model elbow deformation
at the ultimate limit state, a nonlinear analysis accounting for both
material and geometric nonlinearities is necessary. Using the
special-purpose “elbow” element ELBOW31B of ABAQUS, Shalaby
and Younan [8,20] analyzed standalone 90 deg steel elbows
(R/D ¼ 1.5) for a wide range of diameter-to-thickness ratios
(15.5 D/t 97), under in-plane bending (opening and closing
moments) and internal pressure. In subsequent papers, Mourad
and Younan [21,22] analyzed pressurized standalone 90 deg steel
elbow segments (R/D ¼ 1.5) under out-of-plane bending for a
wide range of diameter-to-thickness ratios (15.5 D/t 97),
using special-purpose elbow element ELBOW32 of ABAQUS software. In those investigations, only the curved part of the pipe was
analyzed, neglecting the effects of the adjacent straight parts.
Chattopadhyay et al. [23] employed general-purpose program
NISA to analyze thick 90 deg elbows (D/t 25) under in-plane
bending, through 20-node fully integrated solid elements,
accounting for the effects of the adjacent straight parts. Using a
curve-fitting procedure, simplified formulae were proposed for the
collapse (limit) moment capacity in terms of pressure and the
bend factor h. Karamanos et al. [9] presented a numerical study of
steel elbow response under in-plane bending. Emphasis was given
on the buckling failure of nonpressurized thin-walled elbows, and
a good comparison was found between numerical results and test
measurements reported by Gresnigt et al. [13,14]. In a subsequent
work, Karamanos et al. [7] extended the work of Karamanos et al.
[9] and focused on the ultimate capacity of 90 deg steel elbows
under pressurized bending (in-plane or out-of-plane). The study
was based on simulation of elbows with nonlinear elastic–plastic
shell finite elements, supported by the experimental results. The
numerical results have been compared with experimental measurements from a 90 deg elbow reported by Gresnigt et al. [13,14].
A parametric study has also been conducted to investigate the
pressurized bending response of three 90 deg elbows (D/t ¼ 90,
55, and 20). The ultimate moment capacity of the elbows and the
Fig. 7 Elastic flexibility of pipe elbow specimen (D/t 5 53.3 and R/D 5 3) under in-plane closing bending moments; comparison between numerical and experimental results (values of
pressure p are in MPa)
041203-4 / Vol. 138, AUGUST 2016
Transactions of the ASME
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
corresponding failure modes has been identified, with emphasis
on buckling. Furthermore, the effects of internal pressure on the
ultimate bending resistance have been extensively investigated
and discussed, and special attention has been given on the out-ofplane bending response, in terms of the ultimate bending moment
and the corresponding failure mode.
It is interesting to note that the majority of the work published
on the ultimate bending moment capacity of pipe bends refers to
in-plane bending moments, whereas the out-of-plane bending
capacity has received significantly less attention. The out-of-plane
bending experiments reported by Gresnigt et al. [13,14], Greenstreet [16], and Hilsenkopf et al. [17] indicated that, under this
type of loading, 90 deg elbows are capable of undergoing significant inelastic deformation, before collapse, and that the ultimate
moment capacity is affected by the presence of internal pressure.
The same observations have been noticed in relevant numerical
investigations [7,21,22].
The response of a 60-deg elbow subjected to in-plane closing
and opening bending moments is shown in Fig. 8, in terms of the
corresponding moment–rotation diagrams. The experimental
setup is schematically shown in Fig. 9, where bending moments
are applied through special devices at the end sections of the
straight pipe segments, as reported by Karamanos et al. [9]. The
deformed shapes of the elbows are shown in Fig. 10. The pipe
elbows have diameter equal to 261 mm, thickness equal to 2.9
mm, material, and bend radius R ¼ 772 mm, corresponding to R/D
ratio equal to 3. In the case of closing bending moments, the
response is characterized by excessive ovalization, with crosssectional flattening perpendicular to the plane of bending. Failure
of the pipe bend occurs due to the development of excessive flattening and the development of high strains at the “flank” of the
central elbow cross section, as shown in Fig. 10(a). On the other
hand, the response under opening moments is quite different; significant ovalization develops, which flattens the pipe cross section
in the direction of the bending plane (“reverse” ovalization), and
the failure mode of the elbow is local buckling at the central cross
section of the elbow, at a location between the flank and the extrados, as shown in Fig. 10(b). The results from the numerical simulations with shell elements, shown in Fig. 8 [9] and Fig. 11, are in
very good agreement with the experimental results in terms of the
corresponding moment–rotation diagrams and the corresponding
deformed shapes.
The effect of internal pressure on pipe elbow response is
depicted in Fig. 12 for a 30-deg elbow with diameter equal to 160
mm, thickness 2.9 mm, and bend radius of 480 mm, pressurized to
8.77 MPa (60% of the nominal yield pressure, py ¼ 2ry t=D, where
dy is the material yield stress) reported by Karamanos et al. [9].
The presence of internal pressure has a positive effect on the
bending capacity, allowing for significant deformation (rotational)
capacity without fracture, i.e., with “no loss of containment,” provided that the steel material has adequate ductility.
Fig. 9 Schematic representation of elbow experiments [9]
The response of a thin-walled 90-deg elbow under in-plane
bending is shown in Figs. 13(a) and 13(b) for closing and opening
bending moments, respectively. The pipe elbows have diameter
D ¼ 270 mm, thickness t ¼ 3 mm (D/t ¼ 90), and bend radius
R ¼ 480 mm. The results are obtained from the numerical simulations that employ shell elements. In Fig. 13, moment–rotation diagrams are reported and show a dramatic difference between
closing and opening bending [7], also noted in the elbow results
of Fig. 8. A first observation refers to the bending moment
capacity under zero internal pressure; the ultimate closing
moment is less than 20% of the full-plastic bending moment of
the pipe cross section MP ¼ ry D2 t. Most of the deformation
occurs at the two flank locations, due to cross-sectional flattening,
associated with the development of high local strains in the hoop
direction, and the elbow fails due to excessive ovalization, shown
in Fig. 14, a mode also observed in Fig. 10(a). On the other hand,
for the case of opening bending moments the critical moment is
about 50% of the full-plastic moment. At that stage, the elbow
exhibits local buckling at the central section, similar to the one
depicted in Fig. 10(b).
The effect of internal pressure has a significant effect on bending response for the 90-deg thin-walled pipe bend under consideration, as shown in Fig. 13. Apart from the increase of bending
stiffness, in the first stages of loading, the corresponding bending
moment is also increased; this beneficial effect of internal pressure is more pronounced in the case of closing bending moments.
The response of a thick-walled 90-deg elbow under in-plane
bending is shown in Figs. 15(a) and 15(b) for closing and opening
bending moments, obtained numerically [7]. The pipe has diameter D ¼ 165 mm, thickness t ¼ 8.25 mm, and bend radius
R ¼ 480 mm. This response has similarities with the response of
the thin-walled pipe elbow examined previously. Under closing
bending moments, cross-sectional ovalization governs the
Fig. 8 Moment–rotation diagrams for nonpressurized 60 deg elbows (D/t 5 90); comparison between test data and numerical
results [9]
Journal of Pressure Vessel Technology
AUGUST 2016, Vol. 138 / 041203-5
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 10 Deformed shapes of 60 deg elbows (D/t 5 90); (a) flattened configuration under closing moments and (b) and (c) buckled shape under opening moments [9]
Fig. 11 Finite-element simulation of 60 deg elbow deformation
under in-plane closing bending (D/t 5 90): (a) ovalized shape
under in-plane bending moments and (b) buckled shape under
opening bending [9]
Fig. 12 Final configuration of a pressurized 30 deg elbow
specimen (D/t 5 55) subjected to opening bending moments
response, leading to a limit moment, after which bending moment
decreases. However, the ovalization shape, depicted in Fig. 16, is
characterized by a “smoother” bending deformation at the flank
locations, when compared with the corresponding shape of the
thin-walled pipe elbow in Fig. 14. The effects of internal pressure
on the pipe bending response under closing moments is also characterized by an increase of bending moment capacity with
increasing internal pressure, as shown in Fig. 15(a). This response
has the same trends as the ones shown in Fig. 13(a) for the thinwalled pipe (D/t ¼ 90). On the other hand, a reduction of bending
moment capacity is observed for increased internal pressure levels
in the case of opening bending moments (Fig. 15(b)); in this case,
due to the low value of the D/t ratio associated with very small
ovalization, geometric effects are small and the elbow response is
governed by plasticity. Therefore, the increase of internal pressure
level results in early yielding and a decrease of the bending
moment capacity.
The response of a pipe bend under out-of-plane bending is an
important aspect of pipe analysis, but has received less attention
in the literature. Figure 17(a) shows the nonlinear response of a
90-deg elbow, subjected to out-of-plane bending, obtained both
experimentally with the setup of Fig. 6 and numerically, using
finite-element simulations, as described in Karamanos et al. [7].
Note that cross-sectional ovalization is measured at the central
section of the elbow, both experimentally and numerically, at an
oblique direction of 45 deg with respect to the plane of bending
shown in Fig. 17(b), where maximum ovalization occurs (see Fig.
3(c)). The shape of the deformed elbow subjected to out-of-plane
bending is verified by the finite-element simulation depicted in
Fig. 17(c).
The response of a thin-walled 90 deg elbow with D/t ratio equal
to 90 under severe out-of-plane bending moments is shown in
Fig. 13 Response of a thin-walled 90 deg elbow (D/t 5 90) under in-plane opening bending, for three levels of internal pressure: (a) moment–rotation diagram and (b) ovalization–rotation diagram
041203-6 / Vol. 138, AUGUST 2016
Transactions of the ASME
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 14 Thin-walled 90 deg elbow (D/t 5 90): deformed cross-sectional shapes and spread of plastic deformation under closing bending moments for zero pressure
Fig. 15 Response of 90 deg elbows under in-plane closing bending moments and three levels of internal pressure: (a) thinwalled elbow (D/t 5 90) and (b) thick-walled elbow (D/t 5 20)
Fig. 16 Deformed thick-walled elbow (D/t 5 20): cross-sectional shape and spread of plastic deformation under in-plane closing bending moments for zero pressure
Fig. 18, for three levels of internal pressure (0%, 20%, and 40%
of the yield pressure), and the corresponding pipe elbow shapes
are shown in Fig. 19. The results in Fig. 18 indicate that the ultimate bending moment in the absence of internal pressure is well
below the fully plastic moment of the pipe cross section, and the
corresponding failure mode is local buckling at the intrados of the
Journal of Pressure Vessel Technology
pipe elbow, as shown in Fig. 19(a) (three-dimensional view) and
in Fig. 20(a) (cross-sectional view). The shape of the buckled
elbow is characterized by small wavelength wrinkles in a 45-deg
direction with respect to the plane of bending. One should notice
that the application of out-of-plane bending moment, result in
elbow torsion; a simple analysis of principal stresses in torsion,
AUGUST 2016, Vol. 138 / 041203-7
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 17 Out-of-plane loading of an elbow of moderate thickness (D/t 5 53): (a) force Q versus cross-sectional flattening at central cross section and (b) deformed shape of specimen obtained experimentally and numerically
Fig. 18 Response of a thin-walled 90 deg elbow (D/t 5 90)
under out-of-plane bending moments and three levels of internal pressure (0%, 20%, and 40% of py)
explains the oblique orientation of the wrinkles (Fig. 21). In the
presence of internal pressure, the out-of-plane bending moment
capacity of the thin-walled pipe elbow is significantly increased.
Furthermore, increase of internal pressure prevents the distortion
of pipe cross section, resulting in a “bulging” mode of buckling,
depicted in Fig. 19(c) for a pressure level equal to 40% of the
yield pressure.
The effect of pipe diameter-to-thickness ratio on pipe elbow
under out-of-plane bending is shown in Fig. 22 for zero pressure
and pressure 20% of yield. The comparison for zero pressure in
Fig. 22(a) shows that thick-walled pipe elbows (D/t ¼ 20) may
reach a maximum bending moment higher than 70% of the plastic
moment, whereas the bending moment capacity of thinner elbows
(D/t ¼ 55 and 90) is significantly lower. In the presence of internal
pressure, those differences are less pronounced, as shown in
Fig. 22(b).
Pipe elbows are also used in offshore pipeline systems, such as
“Christmas trees,” risers and manifolds [24]. In deep offshore
applications, pipes are quite thick to resist high levels of external
pressure. However, the combination of external pressure and
bending loading may result in pipe failure. Despite the fact that
the effects of those combined loading conditions on structural
response and stability of straight pipe segments have been extensively investigated [25,26], mainly motivated by the installation
process of deep water pipelines, the corresponding behavior of
offshore pipe elbows has received very little attention. Motivated
by offshore pipeline applications, Bruschi et al. [27] presented an
investigation of the mechanical behavior of 32-in. diameter pipe
elbows with thickness equal to 1.1 in., D/t ¼ 30 and bend ratio
R/D ¼ 5, without accounting for the effects of external pressure.
To the author’s knowledge, the work presented by Pappa et al.
[28] has been the only attempt to investigate external pressure
effects on onshore elbows, using numerical solution similar to the
one reported by Karamanos et al. [7,9]. Figures 23 and 24 show
the decrease of bending capacity of elbows, in the presence of
external pressure, for both closing and opening bending moments.
4
Behavior Under Strong Cyclic Loading
Piping and pipeline systems are often subjected to strong cyclic
loading, associated with repeated excursions of pipe material in
the inelastic range, leading to fatigue damage. Pipe elbows have
been identified as critical locations of those piping and pipeline
Fig. 19 Buckled shapes of thin-walled 90 deg elbow (D/t 5 90) under out-of-plane bending moments for (a) zero pressure, (b)
pressure level 20% of py, and (c) pressure level 40% of py
041203-8 / Vol. 138, AUGUST 2016
Transactions of the ASME
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 20 Deformed configuration of central cross section under out-of-plane bending: (a) thin-walled elbow (D/t 5 90) and (b)
thick-walled elbow (D/t 5 20)
Fig. 21 Schematic representation of the state of stress on an
arbitrary location at the “intrados” of the curved pipe portion
due to out-of-plane torque.
systems, where fracture and loss of containment may occur due to
low-cycle fatigue fracture. Under those cyclic loading conditions,
the elbow may exhibit significant accumulation of plastic strain
(often referred to as “ratcheting”), which eventually may lead to
failure due to plastic collapse.
A significant part of the research associated with strong cyclic
loading on elbows has been motivated by the seismic design and
analysis of piping systems [29–31]. Extensive experimental work
on the ratcheting behavior of pressurized 2-in. carbon and stainless steel pipe elbows has been reported by Yahiaoui et al. [32],
under an “increasing input displacement amplitude” loading. This
work was continued by Yahiaoui et al. [33] for out-of-plane bending, whereas Moreton et al. [34] attempted to predict analytically
the ratcheting rate and ratcheting initiation. Slagis [35] reported
an EPRI/NRC experimental testing program on carbon/stainless
steel pipe elbows, through a shaking-table apparatus, for both
component tests and piping system tests. Extensive experimental
work was presented by Fujita et al. [36], through a series of material tests, pipe component tests, and piping system tests (bent
pipes, tees, and straight pipes).
DeGrassi et al. [37] performed seismic time-history finiteelement analysis of piping system for simulating ratcheting, using
the bilinear, multilinear, and Chaboche models in ANSYS. Balan
and Redektop [38] simulated the response of elbow specimen
under cyclic bending and internal pressure with bilinear plasticity
model in the finite-element code ADINA. More recently, Rahman
and Hassan [39] presented an extensive analytical work on cyclic
Fig. 22 Response of elbows with D/t ratio equal to 20, 55, and 90, subjected to out-of-plane bending moments: (a) zero internal pressure and (b) pressure 20% of py
Journal of Pressure Vessel Technology
AUGUST 2016, Vol. 138 / 041203-9
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 23 Moment–rotation diagram of elbows under in-plane closing bending in the presence of external pressure up to 40%
py (D/t equal to 55 and 20)
behavior of steel elbows, supported by three experiments on 2-in.
SCH10 pipes, aiming at determining the capabilities of several
cyclic plasticity models in predicting the ratcheting rate. All the
above works demonstrated that when steel elbows are subjected to
strong repeated loading, they present failure associated with material degradation or cyclic creep. In many instances, the elbow
cross section distorted or bulged with increasing number of
cycles.
In the course of a large European research program
(2009–2012), with acronym INDUSE [40], extensive research has
been presented aimed at investigating the structural safety of
industrial equipment structures and components under seismic
loading with emphasis on process piping and elbows. More specifically, a series of experiments on pipe elbows under cyclic
in-plane bending has been conducted, supported by numerical
simulations. A first publication from this research effort, which
included only numerical results, was reported recently [41], followed by a series of publications that reported both full-scale tests
and numerical simulations [42,43]. It is important to notice that
all the available data on cyclic loading of pipe elbows refer to
in-plane bending loading, whereas there exist little information on
test data and numerical simulations for cyclic loading under outof-plane bending conditions.
Table 1 summarizes the experimental results reported in Varelis
et al. [42,43]; 8-in. diameter SCH40 long-radius steel pipe hot
bends have been tested with nominal outer diameter and thickness
equal to D ¼ 219.1 mm and t ¼ 8.18 mm, respectively, and bend
radius of the elbow equal to R ¼ 305 mm under strong cyclic
in-plane bending. The elbow configuration is shown in Fig. 25(a);
it is composed by a 90-deg elbow fitting, attached to two straight
pipe segments. The material of the specimens is P355N, according
to EN 10216 standard, which is equivalent to API 5L X52 steel
grade. Loading is imposed through the cyclic displacement of
moving support supports, with amplitude Dl, as shown in
Fig. 25(b), causing in-plane bending under repeated closing and
opening conditions. A constant amplitude loading pattern has
been applied on the specimens shown in Table 1, with the value
of Dl ranged from 6 25 mm to 6 300 mm, with the exception of
test no. 8, where an increasing amplitude displacement is
imposed. During the experiments, local strains have been measured, denoted as DeH;exp .
In those tests, both nonpressurized and pressurized failure
occurred in the form of a crack directed along the pipe axis at the
flank location at the central cross section of the pipe elbow, as
shown in Fig. 26(a). The longitudinal direction of the crack indicates that fracture is due to excessive and repeated local strains in
the hoop direction, due to cross-sectional ovalization, clearly
shown in Fig. 26(b). This deformation is compatible with the
deformed shapes of Figs. 10(a), 11(a), and 14.
In Fig. 25(b), a numerical finite-element model that simulates
the response of the pipe elbow under consideration is depicted
[43]. The numerical results are compared with experimental
results in Fig. 26 in terms of the deformed shape, in Fig. 27 in
terms of the load–displacement diagram, and in Table 1 in terms
of the local strain variation in the hoop direction DeH . In this table,
the strain variation in the longitudinal direction DeL is also shown,
and its value is generally lower than the value of DeH .
It is important to note that in obtaining reliable numerical
results from the finite-element simulations, the choice of the constitutive model is of primary importance. The constitutive model
should be capable of describing accurately the response of steel
material under cyclic loading conditions, and in particular: (a) the
plastic plateau upon initial yielding, (b) the Bauschinger effect
upon reverse loading, and (c) most importantly, the phenomenon
Fig. 24 Moment–rotation diagram of elbows under in-plane opening bending in the presence of external pressure up to 40%
py (D/t 5 55 and 20)
041203-10 / Vol. 138, AUGUST 2016
Transactions of the ASME
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Table 1 Level strain variation in cyclically loaded elbow specimens [42,43]
Test no.
1
2
3
4
5
6
7
8
9
10
11
12
13
Dl (mm)
Number of cycles, Nf
Pressure, P (MPa)
DeH;exp (%)
DeH (%)
DeL (%)
625
670
6100
6150
6200
6250
6300
Increasing amplitude
6200
6300
6200
6300
6200
13,160
444
171
61
28
17
10
16
26
10
27
10
22
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.2
3.2
7.0
7.0
12.0
0.33
1.23
—
2.61
—
3.84
4.02
0.33
1.25
1.59
2.55
2.77
3.75
4.03
N/a
2.89
2.39
2.69
2.25
0.63
0.04
0.14
0.14
0.16
0.16
0.18
0.30
3.01
—
—
1.94
—
0.23
0.97
0.39
0.98
1.49
Fig. 25 In-plane bending loading of 8-in. diameter elbows: (a) experimental setup and (b)
finite-element model for numerical simulations [42,43]
of material ratcheting, i.e., the gradual accumulation of plastic
deformation of the material, under constant-amplitude cyclic loading. Ratcheting has been observed in several experiments on
cyclically loaded elbows [35,37,39,44], characterized by a
bi-axial state of strain and constitutes a demanding simulation
problem. For a thorough presentation of the challenges associated
with bi-axial ratcheting, the reader is referred to the recent works
of Islam et al. [45] and Hassan and Rahman [46].
Using the finite-element model shown in Fig. 25(b), Varelis
and Karamanos [43] simulation of ratcheting in cyclically loaded
elbows has been presented, using the Tseng and Lee [47] cyclic
plasticity model, which is based on “two-surface” plasticity concept. The numerical results in Fig. 28 indicated the accumulation
plastic strain in the hoop direction, for the case of a nonpressurized elbow, comparing very well with experimental data. In addition, numerical calculations for the effects of pressure on hoop
and longitudinal strain ratcheting at the critical location are shown
in Fig. 29.
5
Behavior of Buried Pipeline Bends
Steel transmission pipelines, in areas of significant geohazard
areas (i.e., areas of significant seismic activity or areas of potential
landslide risk) are subjected to severe permanent ground-induced
actions, which may threaten the integrity of the buried pipeline.
Pipeline ground-induced deformations are caused by tectonic fault
action, liquefaction-induced lateral spreading, landslide movement, or soil subsidence, and may lead to pipeline failure, in the
form of either local buckling in the form of a pipe wall wrinkling
or pipe wall fracture due to excessive tensile strain.
This scientific area is receiving rapidly increasing attention during the last few years, considering that numerous large-diameter
Fig. 26 Failure of 8-in. diameter SCH 40 elbows: (a) fatigue crack at elbow flank; (b) cross-sectional ovalization, and (c) finiteelement simulation results
Journal of Pressure Vessel Technology
AUGUST 2016, Vol. 138 / 041203-11
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 30 Schematic representation of a buried pipeline bend
subjected to axial force
Fig. 27 Load–displacement diagram for 8-in. diameter SCH 40
(pressure 3.2 MPa and Dl 5 6200 mm); comparison of test and
numerical finite-element results [43]
Fig. 28 Accumulation of hoop strain (ratcheting) at the elbow
critical location (flank); comparison between experimental and
numerical results [43]
hydrocarbon pipelines are being constructed or planned for construction in geohazards areas. In the course of an efficient pipeline
design and analysis under severe ground-induced actions, the
deformation capacity of pipeline bends can be of particular importance. Several attempts have been published recently on the
response of buried pipelines subjected to permanent ground deformations. However, all those works referred to straight pipeline
segments and the effects of elbows have not been examined. To
the author’s knowledge, the only work on soil–pipe interaction
that referred to pipe elbows is reported in the paper of Yoshizaki
and Sakanoue [48], but it focuses on lateral soil–structure interaction. In any case, the mechanical behavior of buried pipeline
elbows is an open research issue, which requires more attention.
The main feature of the response of buried pipeline elbows is
the interaction of the deforming pipe with the surrounding soil.
Figure 30 shows schematically a buried pipeline bend, subjected
to axial tension in one end, while been infinitely long at the other
end. Figure 31(a) shows a finite-element model that represents the
above physical problem. The model is similar to the models presented in Refs. [49–51] and employs shell elements for modeling
the pipeline, solid elements for modeling the surrounding soil, and
friction contact conditions for the soil–pipe interface. The steel
pipeline under consideration has a 36 in. diameter, a thickness of
3=8 in., and material grade X65 according to API 5 L. The elbow is
a 30-deg hot bend (a ¼ 30o), with bend radius parameter R/D
equal to 5. The pipeline is pressurized at a level of 37.8 bar, which
is 56% of the maximum design pressure. Two cohesive soils
(clay) are considered for the surrounding ground conditions; clay
I is a soft-to-firm cohesive soil with cohesion 50 kPa and Young’s
modulus 25,000 kPa, whereas clay II is a stiff cohesive soil with
Fig. 29 Pressure effects on the accumulation of longitudinal and hoop strain (ratcheting) at the elbow critical location (flank),
obtained from numerical simulations [43]
041203-12 / Vol. 138, AUGUST 2016
Transactions of the ASME
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 31 Coupled response of soil–pipeline system for a 30-deg bend subjected to axial (pullout) force; finite-element simulation
cohesion 200 kPa, Young’s modulus 100,000 kPa. The pipe is subjected to an axial force F at the right end (Fig. 31(a)), and it is
considered to be infinitely long at the left end. The latter condition
is enforced by the use of special-purpose nonlinear spring elements, which account for pipeline continuity to an infinite length,
described in detail by Vazouras et al. [51].
Figure 32 shows the response of the coupled soil–pipeline system, subjected to axial pull-out force F, for the 30 deg elbow. The
load–displacement path is shown in Figs. 32(a) and 32(b) for the
two soil conditions under consideration (clay I and clay II),
together with the corresponding response of an infinitely long
straight pipe and the response of two other models, for a 60-deg
elbow, and a 90-deg elbow, respectively. The deformed configuration of the soil–pipe system is shown in Figs. 31(b) and 31(c). The
results in Fig. 32 indicate that for the soil conditions under consideration, the response of buried pipeline elbows is more flexible
than the response of a straight pipe with the same cross-sectional
and material properties. This flexibility increases with the value of
the bend angle, whereas stiffer soil conditions (clay II) result in
stiffer response. Currently, using the above numerical tools, extensive research is being conducted, toward (a) examining elbow
integrity under severe ground-induced deformations and, more
importantly, (b) investigating the use of the above flexibility property for mitigating fault crossing effects on buried pipelines.
arguments have motivated the development and use of specialpurpose elements, often referred to as “elbow,” “pipe,” or “tube”
elements, as alternatives to shell elements. Those elements have
several advantages over shell elements: their computational efficiency in terms of modeling and execution time, their convenience
in applying boundary conditions at several cross sections and
6 Special-Purpose Elements for the Numerical
Modeling of Pipe Elbows
In the course of piping and pipeline analysis, the numerical
tools should be capable of describing accurately the significant
cross-sectional deformation (distortion or “ovalization”) of the
elbow cross section. Regular beam elements with circular cross
section are inadequate to predict such a response, because they
cannot describe the effects of pressure and—most importantly—
the effects of cross-sectional ovalization on the mechanical
response. On the other hand, the use of shell elements to discretize
long segments of pipelines or piping systems is often computationally expensive for the purposes of pipeline design process,
despite the rigorousness of such an approach. The above
Journal of Pressure Vessel Technology
Fig. 32 Pull-out force–displacement diagram for buried pipeline bends embedded in cohesive soil conditions: (a) soft-tofirm clay (clay I) and (b) stiff clay (clay II)
AUGUST 2016, Vol. 138 / 041203-13
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
kinematic constraints restraining in-plane and warping degrees of
freedom, and the fact that the results from such special-purpose
elements are more easily interpreted.
These special-purpose elements combine longitudinal (beamtype) deformation of the tube axis with cross-sectional deformation of the cylindrical tube wall. The first attempt to combine
those two deformation modes in a simple and efficient tube element was described in the papers by Bathe and Almeida [52,53].
The element is based on the classical von Karman solution of
elbow bending and uses a simplified version of cylindrical linear
theory for inextensional cross-sectional in-plane (no-warping)
deformation, which is discretized through a series of trigonometric functions. The work of Bathe and Almeida [52,53] was the first
to distinguish longitudinal (beam-type) deformation from crosssectional ovalization and couple them in a simple and efficient
manner in the finite-element formulation. Militello and Huespe
[54] proposed a further improvement of the above element considering including warping deformation, but keeping the inextensionality condition, using Hermite polynomials. In a more recent
paper, Yan et al. [55] proposed an “enhanced pipe elbow
element,” which further improves the above concepts and capabilities. Their element allowed for warping deformation and
accounted for a certain degree of cross-sectional extensionality
and for nonsymmetric cross-sectional deformation. A specialpurpose elbow element, incorporated in finite-element program
ABAQUS, has been developed for the elastic–plastic analysis of initially straight and bent tubes under pressure and structural loads,
described in section 3.9.1 of ABAQUS Theory Manual [56]. The
element is based on the Koiter–Sanders linear shell kinematics
and on a discrete Kirchhoff concept, imposed through a penalty
formulation. Cross-sectional warping is also included and the corresponding deformation parameters are discretized through the
use of trigonometric functions up to the sixth degree. Finally,
special-purpose “tube” elements have also been developed by the
author to investigate the nonlinear and buckling analysis of steel
tubulars under structural loads and pressure, motivated by the
structural behavior of tubular members in deep offshore platforms
[57]. The elements process three nodes on the tube axis, and
cross-sectional deformation is described at each node on the basis
of classical ring theory [58], extended to include warping. Crosssectional deformations (in-plane and warping) at every node are
discretized using trigonometric functions, while the crosssectional deformation parameters at each node as well as the nodal
displacements and rotations are interpolated through the use of
appropriate Lagrangian polynomials. Using an expansion up to
the 16th degree, those elements, incorporated in nonlinear analysis
framework, have provided excellent results for the response of
straight pipes and pipe bends, compared with experimental data,
and in particular, the simulation of local buckling.
Inclusion of pressure effects (internal or external) in those
special-purpose elements is of particular importance for the accurate prediction of pressurized pipe elbows. Pressure is not a regular surface load; it is a distributed follower load, always normal to
the pipe surface and, therefore, further adjustments are required
on the stiffness matrix of the pipe. Among other contributions,
aimed at describing those effects on the pipe stiffness matrix, one
may notice the paper by Hibbit [59], which has been used extensively in several commercial programs.
The capabilities of such elements in predicting pipe elbow
behavior is shown in Fig. 8, where the results from the “tube element,” introduced by Karamanos and Tassoulas [57], compare
very well with shell element calculations and the experimental
results. In addition to the moment–rotation diagram, the tube element analysis is capable of predicting local buckling and describing postbuckling behavior, similar to the one depicted in
Fig. 10(b) or 11(b). For more information on this issue, the reader
is referred to the paper by Karamanos et al. [9]. These results
demonstrate that special-purpose elements, if properly used, can
simulate the behavior of piping systems and pipelines with a good
level of accuracy.
041203-14 / Vol. 138, AUGUST 2016
7
Conclusions
Pipe bends (elbows) are critical components for the structural
integrity of piping systems and pipelines. Because of their
geometry, under structural loading, they are more flexible in comparison with straight pipes having the same cross-sectional and
material properties, exhibiting significantly higher stresses and
deformations.
Under extreme loading conditions, pipe elbows may undergo
substantial cross-sectional ovalization. In-plane closing bending
moments result in failure due to flattening in a direction perpendicular to the plane of bending. The corresponding bending
moment capacity is quite low with respect to the fully plastic
moment of the pipe cross section. In the case of opening bending
moments, the response is characterized by reverse ovalization,
resulting in higher bending moment capacity and failure occurs in
the form of local buckling. The response under out-of-plane bending moments is characterized by cross-sectional ovalization in an
oblique direction with respect to the plane of bending. The effect
of internal pressure is generally positive, due to its stabilizing
effect that prevents both ovalization and local buckling. However,
for the particular case of thick-walled elbows subjected to opening
bending moments, those geometric effects are minimal, plasticity
governs the response and the presence of internal pressure reduces
the bending capacity. Finally, numerical simulations on rather
thick-walled pipe elbows, candidates for offshore pipeline applications, indicate that the presence of external pressure reduces the
bending moment capacity.
Pipe elbows subjected to severe cyclic loading may fail due to
low-cycle fatigue. Based on a series of experiments on elbows
under in-plane bending, supported by numerical simulations, fracture occurs at the elbow flank due to cyclic “folding” of the area
because of ovalization. The level of internal pressure did not modify the mode of failure and did not affect significantly the number
of cycles to failure. Ratcheting of local strains at critical locations
has also been detected both experimentally and numerically. It is
also noticed that there is a lack of tests on elbows under out-ofplane bending cyclic loading.
The mechanical response of buried pipeline elbows is a topic
that requires further investigation. Their response is characterized
by soil–pipeline interaction and soil conditions have a significant
effect on pipeline elbow deformation. Numerical results for cohesive soil conditions, obtained through a rigorous finite-element
model, demonstrate that the response of pipe elbows is more flexible than the response of straight pipes with the same crosssectional and material properties, and this flexibility increases
with the value of the bend angle.
Finally, a short note on special-purpose elements for modeling
pipes and elbows is offered. Those elements have been developed
as alternatives to shell elements, combine longitudinal deformation with cross-sectional ovalization and, if properly used, they
are capable of describing the mechanical response of elbows
under pressure and structural loading with a good level of
accuracy.
Acknowledgment
A significant part of the work presented in this paper has been
supported during the period 2009–2014 by the Research Fund for
Coal and Steel (RFCS), under research grants RFSR-CT-200900022 (program INDUSE) and RFSR-CT-2011-00027 (program
GIPIPE). The author would like to thank in particular ir. Arnold
M. (Nol) Gresnigt, Associate Professor at the TU Delft, The Netherlands, Dr. George E. Varelis, Senior Engineer at the PDL Solutions (Europe) Ltd, Hexhan, UK, Dr. Polynikis Vazouras,
Research Associate at the University of Thessaly, and Mrs. Patricia Pappa, Research Assistant at the University of Thessaly.
Finally, the continuous support of the Seismic Engineering Technical Committee of ASME PVPD since 2004 is greatly
appreciated.
Transactions of the ASME
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
References
[1] American Society of Mechanical Engineers, 2007, Factory-Made Wrought
Buttwelding Fittings, B16.9, ASME, New York.
[2] American Society of Mechanical Engineers, 2010, Power Piping, B31.1,
ASME, New York.
[3] American Society of Mechanical Engineers, 2010, Process Piping, B31.3,
ASME, New York.
[4] Comite Europeen de Normalisation, 2002, “Metallic Industrial Piping-Part 3:
Design and Calculation,” Standard EN 13480-3, Brussels, Belgium.
[5] Von Karman, T., 1911, “Uber die Formuanderung dunnwandiger Rohre,” Z.
Ver. Deutcher Ing., 55, pp. 1889–1895 (in German).
[6] Rodabaugh, E. C., and George, H. H., 1957, “Effect of Internal Pressure on the
Flexibility and Stress Intensification Factors of Curved Pipe or Welding
Elbows,” Trans. ASME, 79, pp. 939–948.
[7] Karamanos, S. A., Tsouvalas, D., and Gresnigt, A. M., 2006, “Ultimate Bending
Capacity and Buckling of Pressurized 90 Deg Steel Elbows,” ASME J. Pressure
Vessel Technol., 128(3), pp. 348–356.
[8] Shalaby, M. A., and Younan, M. Y. A., 1998, “Limit Loads for Pipe Elbows
With Internal Pressure Under In-Plane Closing Bending Moments,” ASME J.
Pressure Vessel Technol., 120(1), pp. 35–42.
[9] Karamanos, S. A., Giakoumatos, E., and Gresnigt, A. M., 2003, “Nonlinear
Response and Failure of Steel Elbows Under In-Plane Bending and Pressure,”
ASME J. Pressure Vessel Technol., 125(4), pp. 393–402.
[10] Sobel, L. H., and Newman, S. Z., 1980, “Comparison of Experimental and Simplified Analytical Results for the In-Plane Plastic Bending and Buckling of an
Elbow,” ASME J. Pressure Vessel Technol., 102(4), pp. 400–409.
[11] Sobel, L. H., and Newman, S. Z., 1986, “Simplified, Detailed and Isochronous
Analysis and Test Results for the In-Plane Elastic–Plastic and Creep Behavior
of an Elbow,” ASME J. Pressure Vessel Technol., 108(3), pp. 297–304.
[12] Dhalla, A. K., 1987, “Collapse Characteristics of a Thin-Walled Elbow,”
ASME J. Pressure Vessel Technol., 109(4), pp. 394–401.
[13] Gresnigt, A. M., 1985, “Preofresultaten van Proeven op Gladde Bochten en
Vergelijking Daarvan met de in OPL 85-333 Gegeven Rekenregels,” [in
Dutch], Institute for Construction Materials and Structures, TNO-IBBC, Report
No. OPL 85-334, Delft, The Netherlands.
[14] Gresnigt, A. M., 1986, “Plastic Design of Buried Steel Pipelines in Settlement
Areas,” Heron, 31(4), pp. 3–113.
[15] Gresnigt, A. M., and van Foeken, R. J., 1995, “Strength and Deformation
Capacity of Bends in Pipelines,” Int. J. Offshore Polar Eng., 5(4), pp. 294–307.
[16] Greenstreet, W. L., 1978, “Experimental Study of Plastic Responses of Pipe
Elbows,” ORNL/NUREG-24 Report, Contract No. W-7405-eng-26.
[17] Hilsenkopf, P., Boneh, B., and Sollogoub, P., 1988, “Experimental Study of
Behavior and Functional Capability of Ferritic Steel Elbows and Austenitic
Stainless Steel Thin-Walled Elbows,” Int. J. Pressure Vessels Piping, 33(2),
pp. 111–128.
[18] Suzuki, N., and Nasu, M., 1989, “Non-Linear Analysis of Welded Elbows Subjected to In-Plane Bending,” Comput. Struct., 32(3/4), pp. 871–881.
[19] Tan, Y., Matzen, V. C., and Yu, L. X., 2002, “Correlation of Test and FEA
Results for the Nonlinear Behavior of Straight Pipes and Elbows,” ASME J.
Pressure Vessel Technol., 124(4), pp. 465–475.
[20] Shalaby, M. A., and Younan, M. Y. A., 1999, “Effect of Internal Pressure on
Elastic–Plastic Behavior of Pipe Elbows Under In-Plane Opening Bending
Moments,” ASME J. Pressure Vessel Technol., 121(4), pp. 400–405.
[21] Mourad, H. M., and Younan, M. Y. A., 2001, “Nonlinear Analysis of Pipe
Bends Subjected to Out-of-Plane Moment Loading and Internal Pressure,”
ASME J. Pressure Vessel Technol., 123(2), pp. 253–258.
[22] Mourad, H. M., and Younan, M. Y. A., 2002, “Limit-Load Analysis of Pipe
Bends Under Out-of-Plane Moment Loading and Internal Pressure,” ASME
J. Pressure Vessel Technol., 124(1), pp. 32–37.
[23] Chattopadhyay, J., Nathani, D. K., Dutta, B. K., and Kushwaha, H. S., 2000,
“Closed-Form Collapse Moment Equations of Elbows Under Combined Internal Pressure and In-Plane Bending Moment,” ASME J. Pressure Vessel Technol., 122(4), pp. 431–436.
[24] Kathayat, T. S., Hill, R. T., Goyal, R. K., and Dhar, S., 2012, “Behaviour of Mechanical Properties of High Wall Thickness and Grade in Hot Induction Bending,” Offshore Technology Conference, Houston, TX, Paper No. OTC 22906.
[25] Corona, E., and Kyriakides, S., 1988, “On the Collapse of Inelastic Tubes Under
Combined Bending and Pressure,” Int. J. Solids Struct., 24(5), pp. 505–535.
[26] Karamanos, S. A., and Tassoulas, J. L., 1991, “Stability of Inelastic Tubes Under
External Pressure and Bending,” ASCE J. Eng. Mech., 117(12), pp. 2845–2861.
[27] Bruschi, R., Barbas, S., Mandke, J., Bartolini, L. M., and Vitali, L., 2006,
“Deformation Capacity of Induction Bends,” ASME Paper No. OMAE 200692241.
[28] Pappa, P., Tsouvalas, D., Karamanos, S. A., and Houliara, S., 2008, “Bending
Behavior of Pressurized Induction Bends,” ASME Paper No. OMAE2008-57358.
[29] Fujiwaka, T., Endou, R., Furukawa, S., Ono, S., and Oketani, K., 1999, “Study
on Strength of Piping Components Under Elastic–Plastic Behavior Due to Seismic Loading,” Pressure Vessels and Piping Conference, PVP-Vol 387, Seismic
Engineering, pp. 19–26.
[30] Suzuki, K., 2006, “Earthquake Damage to Industrial Facilities and Development of Seismic Vibration Control Technology: Based on Experience From the
1995 Kobe Earthquake,” J. Disaster Res., 1(2), pp. 177–188.
Journal of Pressure Vessel Technology
[31] Paolacci, F., Reza, M. S., and Bursi, O. S., 2011, “Seismic Design Criteria of
Refinery Piping Systems,” COMPDYN 2011, Conference on Computational
Methods in Structural Dynamics and Earthquake Engineering, Corfu, Greece.
[32] Yahiaoui, K., Moffat, D. G., and Moreton, D. N., 1996, “Response and Cyclic
Strain Accumulation of Pressurized Piping Elbows Under Dynamic in-Plane
Bending,” J. Strain Anal. Eng. Des., 31(2), pp. 135–151.
[33] Yahiaoui, K., Moreton, D. N., and Moffat, D. G., 1996, “Response and Cyclic
Strain Accumulation of Pressurized Piping Elbows Under Dynamic Out-ofPlane Bending,” J. Strain Anal. Eng. Des., 31(2), pp. 153–166.
[34] Moreton, D. N., Yahiaoui, K., and Moffat, D. G., 1996, “Onset of Ratchetting
in Pressurised Piping Elbows Subjected to in-Plane Bending Moments,” Int. J.
Pressure Vessels Piping, 68(1), pp. 73–79.
[35] Slagis, G. C., 1998, “Experimental Data on Seismic Response of Piping
Components,” ASME J. Pressure Vessel Technol., 120(4), pp. 449–455.
[36] Fujita, K., Shiraki, K., Kitaide, K., and Nakamura, T., 1978, “Vibration Damage
Experiment of the Curved Pipes (Elbows) Relating to Seismic
Ultimate Strength,” Trans. Jpn. Soc. Mech. Eng., 44(386), pp. 3437–3445 (in
Japanese).
[37] DeGrassi, G., Hofmayer, C., Murphy, A., Suzuki, K., and Namita, Y., 2003,
“BNL Nonlinear Pre-Test Seismic Analysis for the NUPEC Ultimate Strength
Piping Test Program,” Transactions of SMiRT 17 Conference, Prague, Czech
Republic, Paper No. K15-1.
[38] Balan, C., and Redekop, D., 2005, “The Effect of Bidirectional Loading on
Fatigue Assessment of Pressurized Piping Elbows With Local Thinned Areas,”
Int. J. Pressure Vessels Piping, 82(3), pp. 235–242.
[39] Rahman, S. M., and Hassan, T., 2009, “Simulation of Ratcheting Responses of
Elbow Piping Components,” ASME Paper No. PVP2009-77819.
[40] Pappa, P., Varelis, G. E., Vathi, M., Perdikaris, P. C., Karamanos, S. A., Ferino,
J., Lucci, A., Mecozzi, E., Demofonti, G., Gresnigt, A. M., Dijkstra, G. J., S
Reza, S., Kumar, A. Paolacci, F., Bursi, O. S., Kopp, M., Pinkawa, M., Wieschollek, M. Hoffmeister, B., Stamou, A., Diamanti, K., Papatheocharis, T.,
Botsis, C., Chandrinos, I., and Doukas, I., 2013, Structural Safety of Industrial
Steel Tanks, Pressure Vessels and Piping Systems Under Seismic Loading,
INDUSE RFCS project, Final Report, Brussels
[41] Varelis, G. E., Pappa, P., and Karamanos, S. A., 2011, “Finite Element Analysis
of Industrial Steel Elbows Under Strong Cyclic Loading,” ASME Paper No.
PVP2011-57260.
[42] Varelis, G. E., Karamanos, S. A., and Gresnigt, A. M., 2013, “Steel Elbow
Response Under Strong Cyclic Loading,” ASME J. Pressure Vessel Technol.,
135(1), p. 011207.
[43] Varelis, G. E., and Karamanos, S. A., 2015, “Low-Cycle Fatigue of Pressurized
Steel Elbows Under In-Plane Bending,” ASME J. Pressure Vessel Technol.,
137(1), p. 011401.
[44] Fenton, M., and Hassan, T., 2014, “Low-Cycle Fatigue Failure Responses of
Short and Long Radius Elbows,” ASME Paper No. PVP2014-28805.
[45] Islam, N., Fenton, M., and Hassan, T., 2015, “Long and Short Radius Elbow
Experiments and Evaluation of Advanced Constitutive Models to Simulate the
Responses,” ASME Paper No. PVP2015-45688.
[46] Hassan, T., and Rahman, S. M., 2015, “Constitutive Models in Simulating
Low-Cycle Fatigue and Ratcheting Responses of Elbow,” ASME J. Pressure
Vessel Technol., 137(3), p. 031002.
[47] Tseng, N. T., and Lee, G. C., 1983, “Simple Plasticity Model of the TwoSurface Type,” ASCE J. Eng. Mech., 109(3), pp. 795–810.
[48] Yoshizaki, K., and Sakanoue, T., 2004, “Analytical Study on Soil-Pipeline
Interaction Due to Large Ground Deformation,” 13th World Conference on
Earthquake Engineering, Vancouver, Canada, Aug. 1–6, Paper No. 1402.
[49] Vazouras, P., Karamanos, S. A., and Dakoulas, P., 2010, “Finite Element Analysis of Buried Steel Pipelines Under Strike-Slip Fault Displacements,” Soil
Dyn. Earthquake Eng., 30(11), pp. 1361–1376.
[50] Vazouras, P., Karamanos, S. A., and Dakoulas, P., 2012, “Mechanical
Behaviour of Buried Steel Pipelines Under Strike-Slip and Normal/Reverse
Fault Displacements,” Soil Dyn. Earthquake Eng., 41, pp. 164–180.
[51] Vazouras, P., Dakoulas, P., and Karamanos, S. A., 2015, “Pipe-Soil Interaction
and Pipeline Performance Under Strike-Slip Fault Movements,” Soil Dyn.
Earthquake Eng., 72, pp. 48–65.
[52] Bathe, K.-J., and Almeida, C. A., 1980, “A Simple and Effective Pipe Elbow
Element–Linear Analysis,” ASME J. Appl. Mech., 47(1), pp. 93–100.
[53] Bathe, K.-J., and Almeida, C. A., 1982, “A Simple and Effective Pipe Elbow
Element–Interaction Effects,” ASME J. Appl. Mech., 49(1), pp. 165–171.
[54] Millittelo, C., and Huespe, A. E., 1988, “A Displacement-Based Pipe Elbow
Element,” Comput. Struct., 29(2), pp. 339–343.
[55] Yan, A. M., Jospin, R. J., and Nguyen, D. H., 1999, “An Enhanced Pipe Elbow
Element–Application in Plastic Limit Analysis of Pipe Structures,” Int. J.
Numer. Methods Eng., 46(3), pp. 409–431.
[56] Simulia, 2011, Abaqus Theory Guide, ABAQUS Software, version 6.12, Dassault Systèmes Simulia Corporation, Providence, RI.
[57] Karamanos, S. A., and Tassoulas, J. L., 1996, “Tubular Members I:
Stability Analysis and Preliminary Results,” ASCE J. Eng. Mech., 122(1), pp.
64–71.
[58] Brush, D. O., and Almroth, B. O., 1975, Buckling of Bars, Plates and Shells,
McGraw Hill, New York.
[59] Hibbit, H. D., 1979, “Some Follower Forces and Load Stiffness,” Int. J. Numer.
Methods Eng., 14(6), pp. 937–941.
AUGUST 2016, Vol. 138 / 041203-15
Downloaded From: http://pressurevesseltech.asmedigitalcollection.asme.org/ on 11/15/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use