Journal of Testing and Evaluation

doi:10.1520/JTE20180595 / Vol. 48 / No. 1 / 2020 / available online at www.astm.org

Zhang Zechao,1 Yu Jinghai,2 Chen Zhihua,1,3 Huang Jinchao,1 Wang Zhe,1

and Liu Hongbo4

Experimental Study on the Buckling

of Subsea Pipe-in-Pipe Systems

Reference
Z. Zechao, Y. Jinghai, C. Zhihua, H. Jinchao, W. Zhe, and L. Hongbo, “Experimental Study on the
Buckling of Subsea Pipe-in-Pipe Systems,” Journal of Testing and Evaluation 48, no. 1 (January/
February 2020): 409–427. https://doi.org/10.1520/JTE20180595

ABSTRACT
Manuscript received August 20, Global buckling often occurs in high-temperature and high-pressure subsea pipe-in-pipe (PIP)
2018; accepted for publication systems. However, only a few studies, especially experimental studies, have been conducted
November 26, 2018; published
online March 1, 2019.
on the lateral buckling of PIP systems. An experimental study on the lateral buckling of a PIP
system was carried out in this article. The experimental device for the global buckling of oil
1
Department of Civil Engineering,
temperature-loaded PIP system was designed, and the lateral buckling experiment was carried
Tianjin University, No. 92, Weijin
Rd., Tianjin 300072, People’s out for nine groups of PIP systems. The working mechanism of the PIP system was obtained
Republic of China, https:// through the experiment, and the critical point for pipeline buckling was determined according
orcid.org/0000-0002-9768- to the sum of the axial forces of the outer and inner pipes. The experimental results indicated
6664 (Z.Z.)
that the lateral buckling response were related to the wavelength l0, whereas the effect of
2
Tianjin University Research amplitude Vm was relatively small.
Institute of Architectural Design
and Urban Planning, No. 92, Weijin
Rd., Tianjin 300072, People’s Keywords
Republic of China
pipe-in-pipe, imperfection, global buckling, critical buckling axial force, experimental study
3
Key Laboratory of Coast Civil
Structure and Safety, Ministry of
Education, Tianjin University, Introduction
Tianjin 300072, People’s Republic
of China
The global buckling mode is mainly caused by large axial compressive forces, which are
4
Department of Civil Engineering, produced by constrained expansion resulting from the thermal and inner pressure loads in
Tianjin University, No. 92, Weijin the pipeline.1 Inappropriately controlled global buckling can lead to a series of severe con-
Rd., Tianjin 300072, People’s
sequences, such as pipeline cracking and buckling propagation.
Republic of China (Corresponding
author), e-mail: [email protected] In terms of the theoretical study on single-layer pipelines, Hobbs solved the problem of
the global buckling axial force of the pipeline and presented an analytical solution of infinite
horizontal buckling mode; results showed that the subsea naked laid pipeline can probably go
through horizontal and vertical global buckling. However, restricted by soil mass, the buried

Copyright © 2019 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959 409
410 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

pipeline or pipeline inside the groove generally only goes through vertical buckling.2 Ju and Kyriakides organized and
summarized three morphotypes of the initial imperfection of the pipeline under vertical buckling and introduced
them into the model of Hobbs. Their analysis results indicated that the pipeline was only sensitive to the amplitude
and morphology of initial imperfection but not to friction before vertical buckling; however, the situation was the
opposite in the post buckling stage.3 Miles and Calladine studied the horizontal buckling problems of a slender
body with initial imperfection and computed its post-buckling morphology using the finite element (FE) method.4
Feng et al. used the distributed fiber optic sensor to study the global buckling features of a pipeline.5 Maltby and
Calladine implemented a simplification based on the achievements of Tvergaard and Needleman and used the fun-
damental differential equation of elastic beam and imported nonlinear pipe-soil relation to study the vertical buckling
problems of a subsea pipeline.6 Just like in the traditional pipeline, the multilayer subsea pipeline also faces the global
buckling problem. However, the axial force computation of the multilayer pipeline is much more complicated than
that of the single-layer pipeline under the effects of temperature and pressure, because the inner and outer pipes of the
multilayer pipeline work as a whole. Goplen et al.7 also pointed out that the focus of the analysis and application of
multilayer pipeline should be on regarding it as one whole system but not as two connected pipelines.
In the theoretical study on the global buckling of PIP system, Bokaian simplified the jointing element of the
dual-layer pipeline and inferred the analytical formula of temperature stress of the dual-layer pipeline.8 Based on
the computational conclusion drawn by Bokaian, Zhao, Duan, and Pang presented the analytical expressions of
the axial stress and circumferential stress of the dual-layer pipeline system.9 Wang et al. proposed the computa-
tional formula of the critical bucking force of a vertical global buckling in a PIP system based on dimensional
analysis and numerical simulation.10 Liu et al. studied the global buckling mechanism of a subsea pipeline by
establishing an FE model.11 Wang, Chen, and Liu studied the upheaval buckling with the FE method and showed
that imperfections, along with clearance (the gap between the centralizer and outer pipe) and stiffness ratio of the
inner and outer pipe, influence buckling response and buckling force.12 Alrsai, Karampour, and Albermani
showed the buckling mechanisms in PIP systems with thin and moderately thin carrier pipes with a diam-
eter-to-thickness ratio in the range 26–40 are investigated using two-dimensional analytical and three-
dimensional nonlinear (material and geometry) FE models.13 Present studies on dual-layer pipelines mostly focus
on theoretical analysis, and only a few studies have attempted to identify the collaborative working mechanism
between the outer and inner pipes and the change laws of axial force through experimental studies.
Experimental studies about the methods of loading are mainly divided into mechanical loading and thermal
oil loading, in which the temperature loading mode needs a more complicated loading control device, but temper-
ature rising and declining processes can be fully taken into consideration. Several main experimental studies over
the recent 20 years are reviewed here. For example, Allan studied the global buckling behaviors of elastic steel bar
on rigid foundation and verified the defect sensitivity of this type of buckling.14 Che and Duan used a small-size
model to study the global buckling of a dual-layer pipeline during mechanical loading and obtained the axial
force–displacement relation and critical axial force in buckling process.15 Albrmani, Khalilpasha, and Karampour
used circular pipes with all kinds of initial imperfections as specimens and established an in-pipe axial force in a
temperature-rising manner; their experimental results revealed some deficiencies of the original theory.16 Maltby
and Calladine conducted an experiment on the vertical buckling problem of a buried pipeline, and their exper-
imental results reverified the imperfection sensitivity of the global buckling problem of the pipeline. They were
able to observe the “combinational buckling mode” (i.e., the simultaneous existence of vertical buckling
and horizontal buckling), which was not previously described by a theoretical solution.17 The American
Petroleum Institute carried out an experimental study on the vertical buckling of a buried pipeline and grooved
pipeline and also studied the vertical buckling properties of the buried pipeline under different soil conditions and
imperfection modes.18 Gong and Li presented the experimental study and numerical simulation of the buckle
propagation scenarios for PIP systems in the quasi-static steady-state conditions, the closed-form expressions for
the propagation pressures (PPS) from uniform ring collapse models, a more accurate empirical formula for the
propagation pressure of an outer pipe with a solid rod insert of the same diameter as the inner pipe, and a more
reasonable empirical formula for the propagation pressure of PIP systems is proposed.19 Based on the hypothesis

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 411

of small rotation angle and linear elasticity, the ordinary differential equations of horizontal and vertical global
buckling of an ideal straight pipe with friction are obtained by using the bending moment balance equation of the
pipeline.20 Tvergaard and Needleman carried out an experimental study on global buckling stimulated by a sin-
gle-layer pipeline sleeper method and distributed buoyancy method; they also studied the influences of factors
such as different sleeper depths and buoyancy distribution on the global buckling of the pipeline.21 On the basis of
the summary presented previously, most studies on the dual-layer pipeline are theoretical works, and relatively
few experimental studies have been carried out. Therefore, conducting an experimental study of the oil temper-
ature loading of a PIP system is necessary in order to determine the global buckling development and change
mechanism under different lengths and amplitudes of horizontal initial imperfection.
The remaining parts of the article are arranged into sections. The “Experimental Schemes” section describes
the global buckling experimental device of the PIP system under oil temperature loading and experimental
scheme, including the element design, loading scheme, and the monitoring point arrangement. The
“Analysis of the Experimental Results” section analyzes the experimental results of global buckling, including
critical buckling force analysis, displacement analysis, and strain analysis. “Summary and Conclusions” draws
the study’s conclusions.

Experimental Schemes
EXPERIMENTAL OBJECTIVES
An oil temperature-controlled loading device was used to complete the buckling experiments on nine PIP re-
duced-scale models and to investigate the following:

(1) The influence laws of the wavelength and amplitude of initial imperfection on the global buckling proper-
ties of the PIP system, and
(2) The global buckling mode of the PIP structure, along with the collaborative working mechanism between
the inner and outer pipes.

SPECIMEN DESIGN
With a reference to the existing experimental study results and performances of the experimental equipment,
6061-T6 aluminum alloy circular pipes were selected to simulate the pipeline,22 and the material properties were
listed in Table 1. The pipeline length was 9,000 mm, and the horizontal initial imperfection was set at the midspan
position, as shown in figure 1. The thickness values of both outer and inner pipes were 3 mm, and their outer
diameters were 36 and 62 mm, respectively, as shown in figure 2. The centering loops, which were welded on the
inner pipe, were set every other 500 mm along the axial direction, and aluminum silicate cloth was used fill the gap
and serve as the thermal insulation material between the inner and outer pipes, as shown in figure 3. The hori-
zontal initial imperfection was simulated using the triangular sine curve, and the different wavelengths and am-
plitudes of initial imperfection were considered. According to the imperfection mode based on the sine function
given by Maltby and Calladine,6 we have
 
πx
l = V m sin (1)
l0

where l0 is the wavelength of horizontal initial imperfection of the pipeline, l is the horizontal initial imperfection
of the pipeline, and Vm is the maximum amplitude of the horizontal initial imperfection.

TABLE 1
Material properties

Alloy Designation Young’s Modulus, GPa Poisson Ratio Thermal Expansion Coefficient, mm/°C

6061-T6 69.82 0.3 1.88le − 5

Journal of Testing and Evaluation

412 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

The initial imperfection on the inner pipes was same as that on the outer pipes in one group of experiment,
and the initial imperfection were processed by numerical control pipeline bender to make sure that the initial
imperfections were perfect sine functions as equation (1), and the pipelines were eccentric on the machine.
As shown in previous studies,23 the global buckling of pipeline occurs at a small part of full-length range,
the minimum l0 is L/5, so the minimum l0 is 2,000 mm, and in order to avoid the effect on buckling and
post-buckling of PIP systems by initial imperfections, the minimum Vm was selected as 20 mm. So different
wavelengths and amplitudes of initial imperfection were considered, and nine experimental models were
designed, which were divided into three groups according to imperfection wavelength l0. The model parameters
are presented in Table 2.

LOADING SCHEME
An oil temperature-controlled loading device was used for loading, and this device mainly included a hot-oil
circulating heating control device, an experimental sand tank whose properties are listed in Table 3, and a fixing
device at pipe fitting end, as shown in figure 4. Mono-temperature loading mode was adopted, air temperature
was set as T0, the temperature of pipeline was increased until buckling, and a loading grade was set every other

FIG. 1 Top view of the experimental models.

Camped Supported Constrains

FIG. 2
Sectional view of the
experimental models of
PIP systems.

FIG. 3 Longitudinal profile of the PIP systems.

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 413

TABLE 2
Initial imperfection designs of the experimental models

Group Element No. Wave Length, l0, mm Amplitude, Vm, mm

The first group S1-1000-20 1,000 20

S2-1000-30 1,000 30
S3-1000-40 1,000 40
The second group S4-2000-20 2,000 20
S5-2000-30 2,000 30
S6-2000-40 2,000 40
The third group S7-4000-40 4,000 40
S8-4000-60 4,000 60
S9-4000-80 4,000 80

TABLE 3
The soil type and properties

Type Effective Bulk Density of Soil, γ, kN/m3 Internal Friction Angle of Soil, φ, ° Cohesive Force of Soil, C, kPa

Sand 7.8 18 1.8

10°C. Once the temperature was increased to one design level, and the deformation became stable, the data of the
strain gauge and displacement were acquired, after which loading was continued until buckling.24 Figure 4E
express the logically of the whole loading process on the PIP global buckling, and six major processes were built.

MEASURING POINT ARRANGEMENT

Normal- and high-temperature strain gauges were used in the experiment to measure pipeline strain. A wire-
drawing displacement meter was used to monitor the lateral horizontal displacement of the pipeline. Twelve
strain gauges were symmetrically arranged at the inner and outer arcing sides of the inner pipe. The measuring
points were densely arranged at the initial imperfections in the pipeline arcing region and the initial imperfection
position, as shown in figure 5. Nine strain gauges were symmetrically arranged at the inner and outer arcing sides
of the inner pipe, as shown in figure 5A and 5B. Nine displacement meters were evenly distributed along the axial
direction of the outer pipe, as shown in figure 5C. Before the experiment, the strain acquisition instrument and
displacement meters were calibrated to reduce instrument error. The sampling rate of the sensors was 10 Hz, the
sensitivity rate was 1.0, and the model of the static acquisition system was DM-YB1820, which was produced by
the manufacturer Dan Mo technology company in China.
Figure 6 presents a detailed drawing of the structure at the pipeline end where the outer pipe is connected
with the fixing device at the end through the adapter ring. As can be seen, the inner pipe is screwed down with an
oil supply hole, and neither the inner nor outer pipes could move along the axial direction, although both can
rotate freely.

Analysis of the Experimental Results

GLOBAL BUCKLING CRITICAL TEMPERATURE AND BUCKLING DISPLACEMENT ANALYSIS
Analysis of the Critical Temperature Change and the Buckling Displacement in the First
Group
Experiments on the pipelines in the first group (l0 = 1,000 mm) were carried out. The state comparison before and
after the experiment is shown in figure 7.
The critical temperature change and critical displacement of at the middle points of the first group are shown
in Table 4.

Journal of Testing and Evaluation

414 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

FIG. 4 Experimental devices: (A) hot-oil circulating heating control device, (B) experimental sand tank, (C) fixing device
at the pipe fitting end, (D) layout plan of the experimental loading device, and (E) logic of the loading process.

Critical Temperature Change

According to the Table 4, at the middle points of the specimens under buckling, the critical temperature
changes of the S1-1000-20, S1-1000-30, and S1-1000-40 pipelines are 56.6°C, 55.6°C, and 52.6°C, respectively.
The critical temperature change of the pipelines decreased from 56.6°C to 52.6°C as imperfection Vm increased.

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 415

FIG. 5 Arrangement plan of measuring points: (A) arrangement plan of the strain gauges of the inner pipe along the
direction of length (cm), (B) arrangement plan of the strain gauges of the outer pipe along the direction of length
(cm), and (C) arrangement plan of the displacement meters along the direction of length.

FIG. 6 Structure at the pipeline end.

In S1-1000-20, taken as the standard, the imperfection amplitude Vm increased from 2 cm by 50 % and 100 %,
respectively, and the temperature change reduced by 1.7 % and 7.1 %, respectively. The decreased tempera-
ture change was not that obvious; thus, it could be seen that for the pipeline with a small wavelength less than
L/525 only increasing its imperfect amplitude had an insignificant effect on lowering the critical temperature
change.

Journal of Testing and Evaluation

416 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

FIG. 7
States of the first group
before and after the
experiment: (A) the
S1-1000-20 pipeline
before and after the
experiment: S1-1000-20
before the experiment
(left) and S1-1000-20
after buckling (right);
(B) the S2-1000-30
pipeline before and after
buckling: S2-1000-30
before the experiment
(left) and S2-1000-30
after buckling; and
(C) the S3-1000-40
pipeline before and S1-1000-20 before the experiment S1-1000-20 buckling

after the experiment: (A)

S3-1000-40 before the
experiment (left) and
S3-1000-40 after
buckling (right).

S2-1000-30 before the experiment S2-1000-30 after buckling

(B)

S3-1000-40 before the experiment S3-1000-40 after buckling

(C)

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 417

TABLE 4
Experimental result of the first group

Specimen No. Critical Temperature Change, °C Critical Displacement, mm

S1-1000-20 56.6 −108.3

S2-1000-30 55.6 115.3
S3-1000-40 52.6 110.6

Pipeline Displacement
The displacement directions of the S2-1000-30 and S3-1000-40 pipelines were both imperfection directions before
and after buckling; the buckling displacements at the pipeline middle points were 115.3 and 110.6 mm, respec-
tively. The findings indicate that when imperfection length was certain and as amplitude increased, pipeline
buckling displacement would be reduced; however, the reduction amplitude was not obvious, and the lateral
displacement was only reduced by 4.2 %. This indicated that during the pipeline buckling process, the effect
induced by amplitude increase was not very obvious. In the testing of the S1-1000-20 pipeline, reverse displace-
ment occurred during its buckling process. The pipeline experienced buckling when displacement reached
−108.3 mm. The reason for the reverse displacement of the pipeline was in the process of thermal loading,26
because of the accumulation of deformation at the initial imperfections, where starting and ending points formed
bending deformation, and then the axial force of the pipeline formed component opposite to the direction of
initial imperfections, which leads to the sudden reverse displacement of the pipeline.

Analysis of the Critical Temperature Change and Buckling Displacement in the Second
Group
Experiments on the pipelines in the first group (l0 = 2,000 mm) were carried out. The state comparison before and
after the experiment is shown in figure 8.
The critical temperature change and critical displacement of at the middle points of the second group are
shown in Table 5.

Critical Temperature Change

As shown in the Table 5 at the middle points of the specimens under buckling, the critical temperature change of
S4-2000-20, S5-2000-30, and S6-2000-60 are 45.2°C, 43.5°C, and 40.2°C, respectively. The critical temperature
change of the pipeline decreased from 45.2°C to 40.2°C as imperfection Vm increased. In the S1-1000-20 pipeline,
taken as the standard, the imperfection amplitude Vm increased from 20 mm by 50 % and 100 %, respectively, and
the temperature change reduced by 3.9 % and 12.4 %, respectively. The temperature change was not that obvious;
thus, for the pipeline with a small wavelength less than L/5, only increasing its imperfect amplitude had an insig-
nificant effect on lowering the critical temperature change. Through a comparison of the test member bars, the
difference values of the critical temperature change of pipelines with the same imperfection amplitude (Vm = 20,
30, and 40 mm) were 11.4°C, 12.1°C, and 12.4°C, respectively, after l0 increased from 1,000 to 2,000 mm. Thus, the
critical temperature change decreased by 25.2 %, 27.8 %, and 30.1 %, respectively. Therefore, pipeline length l0 had
a significant effect on lowering critical temperature change.

Pipeline Displacement
The displacement directions of the S4-2000-20, S5-2000-30, and S6-2000-60 pipelines were all imperfection
directions before and after buckling. The buckling displacements at their middle points were 78.5, 70.2, and
48.7 mm, respectively. Based on the analysis of displacements at the pipeline middle points, both the
S4-2000-20 and S5-2000-30 pipelines went through small displacements before buckling. The middle point dis-
placement of the S6-2000-40 pipeline before buckling was not obvious and then suddenly deformed and reached
buckling once the critical temperature change was reached.

Journal of Testing and Evaluation

418 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

FIG. 8 States of the second group before and after the experiment. (A) The S4-2000-20 pipeline before and after the
experiment: S4-2000-20 before the experiment (left) and S4-2000-20 after buckling (right). (B) The S5-2000-30
pipeline before and after the experiment: S5-2000-30 before the experiment (left) and S5-2000-30 after buckling
(right). (C) The S6-2000-40 pipeline before and after the experiment: S6-2000-40 before the experiment (left)
and S6-2000-40 after buckling (right).

S4-2000-20 before the experiment S4-2000-20 after buckling

(A)

S5-2000-30 before the experiment S5-2000-30 after buckling

(B)

S6-2000-40 before the experiment S6-2000-40 after buckling

(C)

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 419

TABLE 5
Experimental result of the second group

Specimen No. Critical Temperature Change, °C Critical Displacement, mm

S4-2000-20 45.2 78.5

S5-2000-30 43.5 70.2
S6-2000-40 40.2 48.7

Analysis of the Critical Temperature Change and Buckling Displacement in the Third
Group
Experiments on the pipelines in the first group (l0 = 4,000 mm) were carried out. The state comparison before and
after the experiment is shown in the figure 9.
The critical temperature change and critical displacement of at the middle points of the third group are
shown in Table 6.

Critical Temperature Change

According to the displacement diagram at the middle points of the specimens under buckling, the critical temper-
ature change of the S7-4000-40, S8-4000-60, and S9-4000-80 pipelines was 35.7°C, 32.2°C, and 22.5°C,
respectively.

Pipeline Displacement
The displacement directions of the S7-4000-40, S8-4000-60, and S9-4000-80 pipelines were all imperfection di-
rections before and after buckling. The buckling displacements at their middle points were 18.4, 11.2, and 8.4 mm,
respectively. Before global buckling, the displacement of the pipelines was not obvious, and at the stage of buck-
ling, the displacement was changed quickly.

Analysis of All Specimens

The post-buckling modes from S1-1000-10 to S9-4000-80 are shown in figures 7–9. For the first group of pipe-
lines with l0 = 1,000 mm, the pipelines present symmetric deformation during the loading process, except
S1-1000-10, because of the accumulation of deformations to midpoint of pipeline initial imperfection. For
the second group of pipelines with l0 = 2,000 mm, because of the defects caused by the transportation process,
the lateral distribution of S4-2000-20 at loading process was nonsymmetric, but when the global buckling oc-
curred, the pipeline presented regular distribution. For the third group of pipelines with l0 = 4,000 mm, all the
pipelines present symmetric deformation buckling modes after the buckling, and buckling can occur more easily.
Through a comparison of the experimental results, the effect of the wavelength cannot only reduce critical buck-
ling temperature of the pipeline but can also easily stimulate symmetric deformation buckling modes during the
buckling process. The probability of stimulating the first type of buckling mode can reach 90 %.
Through a comparison in figure 10, it can be seen that under certain l0 and different vm parameters, the
difference values of the critical temperature change of the pipelines were 4.0°C, 5.0°C, and 13.2°C, respectively.
When vlm0 = λ = 2, the S1-1000-20, S6-2000-40, and S9-4000-80 pipelines were selected. The difference values of
the critical temperature difference were 16.4°C and 17.9°C, respectively, indicating that l0 exerted the main effect
that induced the changes in the critical buckling temperature of the pipelines.
Two sides at the pipeline middle point in the loading phase did not go through symmetrical displacement
when l0 = 4,000 mm, indicating that the pipeline was more sensitive to initial imperfection as wavelength l0
increased. For the S7-4000-040 pipeline, symmetrical displacement occurred in the measuring points on the pipe-
line before buckling at two sides of its middle point towards the imperfection direction. The pipeline went through
buckling after the critical point was reached; subsequently, the displacement quantity near the oil inlet was
obviously larger than at the outlet. The displacement difference was more obvious at the pipeline imperfection

Journal of Testing and Evaluation

420 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

FIG. 9 States of the second group before and after the experiment. (A) the S7-4000-40 pipeline before and after the
experiment: S7-4000-40 before the experiment (left) and S7-4000-40 after buckling (right). (B) the S8-4000-60
pipeline before and after the experiment: S8-4000-60 before the experiment (left) and S8-4000-60 after
buckling (right). (C) the S9-4000-80 pipeline before and after the experiment: S9-4000-80 before the
experiment (left) and S9-4000-80 after buckling (right).

S7-4000-40 before the experiment S7-4000-40 after buckling

(A)

S8-4000-60 before the experiment S8-4000-60 after buckling

(B)

S9-4000-80 before the experiment S9-4000-80 after buckling

(C)

TABLE 6
Experimental result of the third group

Specimen No. Critical Temperature Change, °C Critical Displacement, mm

S7-4000-40 35.7 22.3

S8-4000-60 32.2 20.3
S9-4000-80 22.5 18.4

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 421

FIG. 10
Variation trend diagram
of the critical 55
l0 = 1 m
temperature change of
50 l0 = 2 m
the specimens.

Temperature Difference ( )
l0 = 4 m
45

40

35

30

25

20
2 3 4 5 6 7 8
Amplitude Vm cm

segment, and the maximum difference between the No. 4 and No. 6 displacement meters was 16.2 mm. For the
S8-4000-060 pipeline, negative displacement happened to measuring point 9 in the initial loading phase, and the
maximum displacement difference between measuring points 4 and 6 after buckling was 15.1 mm. For the S9-
4000-080 pipeline, its displacement change laws were similar to those of S7-4000-040. Negative displacement
happened to the S8-4000-060 pipeline at measuring point 9. When the pipeline was being processed, negative
bending happened nearby measuring point 9, which was about 20 mm away along the axis. Therefore, one im-
perfection at this position would experience lateral displacement during the heating process with oil supply. For
the S7-4000-040 and S9-4000-080 pipelines, displacement close to oil inlet of the pipeline was obviously higher
than at the outlet, and it was more obvious at the imperfection. This was because raw oil was transported under
the rising room temperature in order to guarantee the experimental reliability during the oil flow process.27 Oil
temperature rising speed was controlled at 1°C/min, and oil transportation speed was 1 m/s. Pipeline imperfection
went through displacement first at the oil inlet. When l0 = 4,000 mm, there was a 2°C temperature change at two
imperfection ends in the pipeline because the imperfection was long. Therefore, this would result in asymmetrical
displacement, indicating that the pipeline was more sensitive to imperfection during the buckling process as l0
increased.

AXIAL STRESS AND FORCE ANALYSIS

The resultant force can be obtained from the following formula28:

F cr = F i + F o (2)

where Fcr is the resultant force of the PIP system, Fi is the force of inner pipe, and Fo is the force of outer pipe.
Total strain, which mainly consists of axial strain and bending strain, and strain values can be solved through
the following equations:

ε1 + ε2 = εA (3)

ε3 + ε4 = εB (4)

where,

ε1 = −ε3 , ε2 = ε4 (5)

Journal of Testing and Evaluation

422 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

Therefore,

ε2 = ε4 = ðεA + εB Þ=2 (6)

ε1 = −ε3 = ðεA − εB Þ=2 (7)

F i = Ai σ i = Ai Eε1i (8)

F o = Ao σ o = Ao Eε1o (9)

F cr = F i + F o = Ai Eε1i + Ao Eε1o (10)

where ε1, ε2, ε3, and ε4 are the strain value in the pipeline measuring points, Ai is the area of inner pipe, Ao is the
area of outer pipe, σi is the stress of the inner pipe, and σo is the stress of the outer pipe.
Figure 11 showed the axial strains and force of different specimens during the buckling process. During the
loading process of the S1-1000-20 and S2-1000-30 pipelines, both axial strains were symmetrically distributed at
two sides of their middle points. Strain values at the rightmost starting side of the imperfection in the S3-1000-40
pipeline changed slowly, which was related to the initial processing imperfection of the pipeline. For the pipeline
with wavelength of l0 = 2,000 mm, the strain values presented symmetrical wave-shaped distribution at two sides
of the pipeline middle point, and the strain at the initial imperfection of the pipeline showed the greatest change
during the loading process. For the pipeline with the wavelength l0 = 4,000 mm, the strain value was the largest at
the end and presented a linear decrease along the axial direction of the pipeline. The strain value was the smallest
at the starting point and end point of the imperfection. The strain values in the imperfection were basically
identical, indicating that the axial forces borne by imperfection segments of the pipelines with the l0 =
4,000 mm wavelength were basically identical.
From equations (2)–(10), the axial stress of the pipeline during buckling can be obtained. The axial stress
distribution in the pipeline is consistent with the axial strain.29 By analyzing the global buckling process of
the nine specimens, it was found that the axial stress on the section of initial imperfection increases rapidly before
buckling, especially at the initial stage of loading, while the axial stress on the section of initial imperfection
decreases and the end of pipeline increases continuously after buckling. This is mainly because at the initial
loading stage, the PIPs were sensitive to the initial imperfection, so the global buckling would first happen
at the section of initial imperfection, which means the axial force significantly increased. However, after the global
buckling, the PIPs were mainly bent, so the axial force value at the initial imperfections was basically unchanged,
and the bending phenomenon was more obvious at the end because of the restriction where the axial force would
continue to increase, resulting in the stress distribution near the initial imperfections appeared wavy shape.

When Amplitude Vm Remained Unchanged

As shown in the Table 7, comparative analysis of the three specimens S3-1000-40, S5-2000-40, and S7-4000-40
(Vm = 40 mm) was conducted, and their wavelengths l0 were 1,000, 2,000, and 4,000 mm, respectively. The value
of l0 = 1,000 mm was taken as the standard, and the wavelengths of S5 -2000-40 and S7-4000-40 increased by
100 % and 300 %, respectively, whereas their critical buckling forces decreased by 43.8 % and 53.7 %, respectively.
Therefore, when amplitude was unchanged, the changing wavelength could obviously reduce the critical axial
force of buckling.

When Wavelength l0 Remained Unchanged

After comparison of the specimens with l0 = 1,000, 2,000, and 4,000 mm, respectively, their differences in the
maximum critical axial forces were 351.8, 249.3, and 146.6 N, respectively, which decreased by 28.1 %, 32.7 %, and
35.1 %, respectively, when compared with the critical axial forces under the minimum amplitudes of the spec-
imens. This indicated that when wavelength amplitude Vm was unchanged, the increase of wavelength l0 had an
obvious effect on reducing the critical buckling axial force.

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 423

FIG. 11 Axial strains and force of different specimens during the buckling process: (A) Axial strains at measuring points of
the S1-1000-20 pipeline during the buckling process (left) and axial force of the inner and outer pipes under the
middle point buckling in the S1-1000-20 pipeline (right); (B) axial strains at measuring points of the S2-1000-30
pipeline during the buckling process (left) and axial force of the inner and outer pipes under the middle point
buckling in the S2-1000-30 pipeline (right); (C) axial strains at measuring points of the S3-1000-40 pipeline during
the buckling process (left) and axial force of the inner and outer pipes under the middle point buckling in the S3-
1000-40 pipeline (right); (D) axial strains at measuring points of the S4-2000-20 pipeline during the buckling
process (left) and axial force of the inner and outer pipes under middle point buckling in the S4-2000-20 pipeline
(right); (E) axial strains at measuring points of the S5-2000-30 pipeline during the buckling process (left) and axial
force of the inner and outer pipes under the middle point buckling in the S5-2000-30 pipeline (right); (F) axial
strains at measuring points of the S6-2000-40 pipeline during the buckling process (left) and axial force of the
inner and outer pipes under the middle point buckling in the S6-2000-40 pipeline (right); (G) axial strains at
measuring points of the S7-4000-40 during the buckling process (left) and axial force of the inner and outer pipes
under the middle point buckling in the S7-4000-40 pipeline (right); (H) axial strains at measuring points of S8-
4000-60 during buckling process (left) and axial force of inner and outer pipes under middle point buckling in S8-
4000-60 (right); (I) axial strains at measuring points of the S9-4000-80 pipeline during the buckling process
(left) and axial force of the inner and outer pipes under the middle point buckling in the S9-4000-80 pipeline
(right).

×102 ×103
16 10 1.0 Inner pipe
20 Outer pipe
14 30 0.5 Resultant force
40
12
Axial Force (N)

50 0.0
56
10 60
–0.5
8
–1.0
6
–1.5
4
–2.0
2
–2.5
0
0 1 2 3 4 5 6 7 8 9 –10 0 10 20 30 40 50 60 70 80
PIP Length Positions (m) Temperature Change
(A)
×102 ×102
16 5.0 10 Inner pipe
15.0 Outer pipe
14 30.0 Resultant force
5
Axial Force (N)

40.0
12 50.0
55.0 0
10 60.0
–5
8

6 –10

4 –15
2
20

0 1 2 3 4 5 6 7 8 9 0 10 20 30 40 50 60 70
PIP Length Positions (m) Temperature Change
(B)

Journal of Testing and Evaluation

424 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

FIG. 11 Continued
×102 ×102
Inner pipe
18 10 5 Outer pipe
20 Resultant force
16
30
14 40

Axial Force (N)

50
0
12
52.6
10 60
–5
8
6
–10
4
2
–15
0
0 1 2 3 4 5 6 7 8 9 0 10 20 30 40 50 60
PIP Length Positions (m) Temperature Difference
(C)
2
×102 10 ×10
Inner pipe
10
Outer pipe
5 Resultant force
Axial Force (N)
8
0
6
–5
4
–10
2
–15
0
–20
0 1 2 3 4 5 6 7 8 9 0 10 20 30 40 50 60
PIP Length Positions (m) Temperature Difference (䉝 )
(D)
×102 ×102
10 Inner pipe
2
Outer pipe
Resultant force
8 0
Axial Force (N)

×102
6
–2

4
–4

2
–6
0
–8
0 1 2 3 4 5 6 7 8 9 0 10 20 30 40 50
PIP Length Positions (m) Temperature Difference (䉝 )
(E)

Summary and Conclusions

On the basis of the experimental study on the global buckling of the PIP system, the effects of different imper-
fection wavelengths and amplitudes on the global buckling of the PIP system were analyzed. The main conclu-
sions are as follows.

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 425

FIG. 11 Continued
×102 ×102
Inner pipe
10 5 Outer pipe
Resultant force
8

Axial Force (N)

0
6

–5
4

2 –10

0
–15
0 1 2 3 4 5 6 7 8 9 0 10 20 30 40
PIP Length Positions (m) Temperature Difference
(F)
×102 ×102
8 8 Inner pipe
6 Outer pipe
7 Resultant force
4
Axial Force (N)
6 2
5 0
–2
4
–4
3
–6
2 –8
1 –10
–12
0
0 1 2 3 4 5 6 7 8 9 0 5 10 15 20 25 30 35 40 45
PIP Length Positions (m)
(G)
2
×10 ×102
8 10 Inner pipe
Outer pipe
7
Resultant force
6 5
Axial Force (N)

5
0
4

3 –5
2

1 –10

0
–15
0 1 2 3 4 5 6 7 8 9 0 5 10 15 20 25 30 35 40 45
PIP Length Positions (m)
(H)

The global buckling change and development laws of the PIP system under initial imperfections in terms of
critical buckling force, critical temperature difference, and critical displacement were obtained by conducting
experiments on the PIP system. The collaborative action mechanism between the inner and outer pipes in
the PIP system containing lateral initial imperfection was as follows: axial force and lateral deformation were
generated in the inner pipe under compression as oil temperature increased. Such force drove the common

Journal of Testing and Evaluation

426 ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS

FIG. 11 Continued
×102
×102
6 10
Inner pipe
Outer pipe
Resultant force
5
5

Axial Force (N)

4
0
3

2 –5

1
–10
0
–15
0 1 2 3 4 5 6 7 8 9 0 5 10 15 20 25 30 35
PIP Length Positions (m)
(I)

TABLE 7
Critical axial force values of pipelines with different parameters

Specimen No. Critical Axial Force, N Specimen No. Critical Axial Force, N Specimen No. Critical Axial Force, N

S1-1000-20 −1,251.8 N S3-2000-20 −760.2 N S7-4000-40 −417.2 N

S2-1000-30 −943.4 N S4-2000-30 −550.1 N S8-4000-60 −333.3 N
S3-1000-40 −899.6 N S5-2000-40 −510.9 N S9-4000-80 −270.6 N

deformation of the outer pipe under the effect of centering loop, and the outer pipe was mainly affected by tensile
force.
The influence factor of the critical buckling of the PIP system was mainly related to the imperfection wave-
length l0. The critical buckling temperature decreased by about 50 % every time l0 doubled, but Vm had a small
effect. Analysis results indicated that the greater the wavelength l0, the more evenly the lateral displacement
developed, so an increase in l0 exerted a protective effect on the pipeline safety.
For pipelines with small l0, the strains at the imperfection segments during the buckling process were not
even but presented a symmetrical wave-shaped distribution. Strain changes at the imperfection segments were
relatively identical when l0 was large. Moreover, the critical buckling axial force reduced by 31.1 % every time l0
doubled on average. Hence, during the oil transportation process of the pipelines, global pipeline buckling could
be stimulated very early so as to ensure the safety of the PIP system. The active control method on the global
buckling should be studied in the future by experimenting.

ACKNOWLEDGMENTS
The authors are grateful for the support provided by the National Basic Research Program of China under
Grant No. 2014CB046801.

References
1. Z. Wang, Z. Chen, H. Liu, and Z. Zhang, “Numerical Study on Lateral Buckling of Pipelines with Imperfection and
Sleeper,” Applied Ocean Research 68 (October 2017): 103–113. https://doi.org/10.1016/j.apor.2017.08.010
2. R. Liu and S. Yan, “Brief History of Upheaval Buckling Studies for Subsea Buried Pipeline,” Journal of Pipeline Systems
Engineering and Practice 4, no. 3 (August 2013): 170–183. https://doi.org/10.1061/(ASCE)PS.1949-1204.0000127

Journal of Testing and Evaluation

ZECHAO ET AL. ON PIPE-IN-PIPE (PIP) SYSTEMS 427

3. R. E. Hobbs, “In-Service Buckling of Heated Pipelines,” Journal of Transportation Engineering 110, no. 2 (March 1984):
175–189. https://doi.org/10.1061/(ASCE)0733-947X(1984)110:2(175)
4. D. J. Miles and C. R. Calladine, “Lateral Thermal Buckling of Pipe-Lines on the Sea Bed,” Journal of Applied Mechanics
66, no. 4 (December 1999): 891–897. https://doi.org/10.1115/1.2791794
5. X. Feng, W. Wu, X. Li, X. Zhang, and J. Zhou, “Experimental Investigations on Detecting Lateral Buckling for Subsea
Pipelines with Distributed Fiber Optic Sensors,” Smart Structures and Systems 15, no. 2 (February 2015): 245–258. https://
doi.org/10.12989/sss.2015.15.2.245
6. T. C. Maltby and C. R. Calladine, “An Investigation into Upheaval Buckling of Buried Pipelines—ii. Theory and Analysis
of Experimental Observations,” International Journal of Mechanical Sciences 37, no. 9 (September 1995): 965–983. https://
doi.org/10.1016/0020-7403(95)00005-I
7. S Goplen, O. Fyrileiv, J. P. Grandal, and L. Borsheim, “Global Buckling of Pipe-In-Pipe: Structural Response and
Design Criteria,” in 30th International Conference on Ocean, Offshore and Arctic Engineering (New York: ASME,
2011), 777–784.
8. A. Bokaian, “Thermal Expansion of Pipe-In-Pipe Systems,” Marine Structures 17, no. 6 (October 2004): 475–500. https://
doi.org/10.1016/j.marstruc.2004.12.002
9. T. Zhao, M. Duan, and X. Pang, “An Approach of Lateral Buckling Analysis of Ht Non-Compliant Pip System,” The
Ocean Engineering 26, no. 3 (2008): 65–69.
10. Z. Wang, Z. Huachen, H. Liu, and Y. Bu, “Static and Dynamic Analysis on Upheaval Buckling of Unburied Subsea
Pipelines,” Ocean Engineering 104 (August 2015): 249–256. https://doi.org/10.1016/j.oceaneng.2015.05.019
11. R. Liu, X. Hao, X. Wu, and S. Yan, “Numerical Studies on Global Buckling of Subsea Pipelines,” Ocean Engineering 78
(March 2014): 62–72. https://doi.org/10.1016/j.oceaneng.2013.12.018
12. Z. Wang, Z. Chen, and H. Liu, “Numerical Study on Upheaval Buckling of Pipe-In-Pipe Systems with Full Contact
Imperfections,” Engineering Structures 99 (September 2015): 264–271. https://doi.org/10.1016/j.engstruct.2015.04.055
13. M. Alrsai, H. Karampour, and F. Albermani, “Numerical Study and Parametric Analysis of the Propagation Buckling
Behaviour of Subsea Pipe-in-Pipe Systems,” Thin-Walled Structures 125 (April 2018): 119–128. https://doi.org/10.1016/
j.tws.2018.01.019
14. T. Allan, “One-Way Buckling of a Compressed Strip under Lateral Loading,” Journal of Mechanical Engineering Science
10, no. 2 (April 2006): 175–181. https://doi.org/10.1243/JMES_JOUR_1968_010_025_02
15. X. Che and M. Duan, “Experimental Study and Numerical Simulation of Global Buckling of Pipe-in-Pipe Systems,”
Applied Mathematics and Mechanics 35, no. 2 (2014): 188–201.
16. F. Albrmani, H. Khalilpasha, and H. Karampour, “Propagation Buckling in Deep Sub-Sea Pipelines,” Engineering
Structures 33, no. 9 (2011). https://doi.org/10.1016/j.engstruct.2011.04.026
17. C. de Oliveira Cardoso and R. F. Solano, “Performed of Triggers to Control Thermal Buckling of Subsea Pipelines Using
Reduced Scale Model,” Proceedings of the 25th International Ocean and Polar Engineering Conference (Mountain View,
CA: International Society of Offshore and Polar Engineers, 2015): ISOPE-I-15-445.
18. American Petroleum Institute, Recommended Practice for Planning, Designing and Constructing Fixed Offshore
Platforms—Working Stress Design, RP 2A-WSD (Washington, DC: American Petroleum Institute, 2000).
19. S. Gong and G. Li, “Buckle Propagation of Pipe-in-Pipe Systems under External Pressure,” Engineering Structures
84 (February 2015): 207–222. https://doi.org/10.1016/j.engstruct.2014.11.032
20. R. E. Hobbs, “Pipeline Buckling caused by Axial Loads,” Journal of Constructional Steel Research 1, no. 2 (January 1981):
2–10. https://doi.org/10.1016/0143-974X(81)90027-4
21. V. Tvergaard and A. Needleman, “On Localized Thermal Track Buckling,” International Journal of Mechanical Sciences
23, no. 10 (1981): 577–587. https://doi.org/10.1016/0020-7403(81)90038-2
22. Z. Hong, R. Liu, W. Liu, and S. Yan, “Study on Lateral Buckling Characteristics of a Submarine Pipeline with a Single Arch
Symmetric Initial Imperfection,” Ocean Engineering 108 (November 2015): 21–32. https://doi.org/10.1016/j.oceaneng.
2015.07.049
23. X. Zeng, M. Duan, and X. Che, “Critical Upheaval Buckling Forces of Imperfect Pipelines,” Applied Ocean Research 45
(March 2014): 33–39. https://doi.org/10.1016/j.apor.2014.01.001
24. Z. Hong, R. Liu, W. Liu, and S. Yan, “A Lateral Global Buckling Failure Envelope for a High Temperature and High
Pressure (Ht/Hp) Submarine Pipeline,” Applied Ocean Research 51 (June 2015): 117–128. https://doi.org/10.1016/j.apor.
2015.02.008
25. Z. Wang, K. J. Ma, and Z. Chen, “Overview of Global Buckling of Deep-Sea Pipeline,” Journal of Tianjin University
(Science and Technology) 47, no. 7 (July 2014): 17–23.
26. L. Wang and R. Shi, “Numerical Simulation for the Dynamic Process of Pipeline Buckling,” The Ocean Engineering 33,
no. 5 (2015): 73–80.
27. L. P. Guo, R. Liu, and S. W. Yan, “Low-Order Lateral Buckling Analysis of Submarine Pipeline under Thermal Stress,” in
New Frontiers in Engineering Geology and the Environment (Heidelberg: Springer, 2013), 191–194.
28. Z. Wang, Z. Chen, and H. Liu, “On Lateral Buckling of Subsea Pipe-in-Pipe Systems,” International Journal of Steel
Structures 15, no. 4 (December 2015): 881–892. https://doi.org/10.1007/s13296-015-1209-3
29. V. Tvergaard and A. Needleman, “On the Localization of Buckling Patterns,” Journal of Applied Mechanics 47, no. 3
(September 1980): 613–619. https://doi.org/10.1115/1.3153742

Journal of Testing and Evaluation