Академический Документы
Профессиональный Документы
Культура Документы
IPC2010-31622
Olav Fyrileiv
Det Norske Veritas,
Veritasveien 1,
N-1322 Høvik, Norway
ABSTRACT were raised and some experimental data presented that claimed
Free span assessment has more and more become an to prove that the way the internal pressure was handled in the
important part of modern pipeline design. The reason for this is DNVRPF105 “Free Spanning Pipelines” (2006) [2] is wrong.
partly that the remaining hydrocarbon reservoirs are located in The intention of this paper is to show how the internal
more challenging places, e.g. with very uneven seabed. Another pressure influences on the structural response of free spans, and
explanation is that the pipeline design codes a few decades ago that the DNV codes and standard nonlinear FE software, e.g.
did not allow for vibrating free spans, while the modern, state Abaqus, handle this effect in an adequate manner.
oftheart pipeline codes, such as DNVOSF101 “Submarine
Pipeline Systems” (2007) [1] and its Recommended Practices, Keywords: Pipelines, Effective axial force, Pressure effects,
opens for long spans that are allowed to vibrate as long as the Pipeline codes.
structural integrity is ensured. By opening for longer free spans
significant seabed intervention costs associated with trenching, INTRODUCTION
rock dumping and supporting spans by other means are saved. As the easiest accessible offshore petroleum reservoirs
One of the governing parameters to ensure the structural have been discovered and developed over the last five decades,
integrity of free spans is the natural frequency of the span. This the trend in the offshore industry points towards deeper waters
is a parameter that the designer can to some degree control by and harsher environment. This trend also calls for increased use
means of moderate seabed intervention, e.g. span support. of subsea installations instead of platforms and use of pipelines
Since the natural frequency of the span together with the water to transport the hydrocarbons to shore or into a pipeline grid.
flow velocity normal to the span determine the vibrations and The pipelines may be installed at a very rough, uneven seabed
the cyclic loading it is of vital importance to be able to estimate causing a high number of free spans that can be difficult and/or
a realistic value of this frequency. very costly to intervene.
The natural frequency is influenced by several effects. One Free span design is challenging and important – some
of them is the effect of the internal pressure. This may represent decades ago only short spans were allowed, no vibrations –
a challenge since the effect of the pressure is the opposite of now, DNVOSF101 “Submarine Pipeline Systems” (2007) [1]
what one instantaneously thinks is correct. Quite recently some and the associated DNVRPF105 (2006) [2] open for long
discussion about the effect of internal pressure on free spans spans and allowing vibration as long as fatigue is controlled. In
1 Draft Copy
shallow waters stress cycling due to direct wave force may As will be shown later in this paper, the effective axial
cause fatigue failure while in deeper waters VIV may do the force governs the structural response of the pipeline in an
same. overall perspective, influencing on lateral buckling, upheaval
VIV is governed by the socalled reduced velocity, an buckling, anchor forces, end expansion and natural frequencies
expression found by setting the shedding frequency given by of free spans. For this reason, it of utmost importance to
the wellknown Strouhal’s number equal to the natural understand its effects and be able to estimate it accurately in
vibration frequency of the span. In this way the natural order to end up with a safe and reliable design.
frequency becomes one of the most important parameters along Some nonlinear finite element analysis results will be
with the flow velocity and the outer diameter of the pipe. given and compared with recent experimental results. These
The structural response parameters of a free span, such as cases will prove that recent claims about the way pressure
natural frequencies and associated modal stress, are often based effects are treated by the industry and in the DNV codes are
on simplified beam theory expressions. The response is based on a wrong interpretation of the experimental results.
significantly influenced by factors such as the soil stiffness, the Further it will be shown that the standard way of analysing
axial force and the static deflection of the span. For this reason, pipelines using standard pipe elements will yield realistic
ordinary beam theory expressions will often fail to estimate the estimates for the structural response as long as materials,
structural response with a required accuracy for free span boundary conditions and loads are represented realistically.
assessment.
One alternative to beam theory expressions is clearly using PRESSURE EFFECTS AND EFFECTIVE AXIAL
nonlinear finite element (FE) methods. Provided that the FORCE
modeling of span support conditions and other details are done The effective axial force is explained in detail in many
properly, this method provides reliable and accurate results. papers, e.g. Sparks [7] and Fyrileiv and Collberg [11]. The
The method also provides large flexibility in modeling different effect of the external pressure is most easily understood by
aspects like span geometry, span support etc, see Fyrileiv et al. considering the law of Archimedes:
[3] and Kristiansen et al. [4]. However, FE analysis is time
consuming and not very efficient if a large number of spans are “The effect of the water pressure on a submerged body is
to be screened. an upward directed force equal in size to the weight of the
According to the Recommended Practice DNVRPF105 water displaced by the body”.
“Free Spanning Pipelines” (2006) [2], the internal pressure will
have a negative effect on the natural frequency. This is Now, consider a section of a pipeline exposed to external
normally accounted for by the socalled effective axial force, pressure as illustrated in fig. 1. For clarity the only sectional
and as this force increases towards the critical global buckling force included is the axial force, N, the socalled true wall force
value, the natural frequency will asymptotical move towards found by integrating the steel stress over the steel crosssection
zero. Even though the effective axial force has been used in area.
pipeline codes for several decades, see for example DNV’76
(1976) [5], it is still misunderstood and misinterpreted when it
N
comes to the effect of pressures. The effect of the internal
pressure is treated by for example Palmer and Baldry [6] and pe
Sparks [7] who seemed to introduce the term effective axial
force. N
However, some years ago a discussion about this concept = pe
peAe
was initiated and doubt on how to treat the internal pressure
raised, ref. Galgoul et al. [8, 9]. Recently, Massa et al. [10]
+
conducted an experiment in order to prove that the frequency is
governed by the true axial force rather than the effective axial
force. The main challenge with the internal and external
pressures is that the effect of these is often the opposite of what
one instantaneously thinks is correct. Therefore, these effects
have been sources to misunderstandings and wrong designs.
Figure 1 - Equivalent physical systems – external pressure
2 Draft Copy
As shown in the figure, the section with the axial force and R
v q
the external pressure, pe, (left figure) can be replaced by a
section where the external pressure acts over a closed surface
M+dM
and gives the resulting force equal to the weight of the
displaced water, the buoyancy of the pipe section (middle M N
figure), and an axial force equal to N + peAe. Considering the
effect of the external pressure in the way as shown in fig. 1 N Q+dQ
Q
does not change the physics or add any forces to the pipe
section. However, it significantly simplifies the calculation as
the alternative would be to integrate the pressure over the x
double curved pipe surface. dx
In this way it becomes clear that the effect of the external Figure 2 – Forces acting on an infinitesimal length of a pipe.
and internal pressures may be accounted for by the socalled
effective axial force:
By considering equilibrium of lateral forces and the
S N p i Ai p e Ae (1) bending moment acting on the pipe section shown in fig. 6, the
dynamic equation of motion for the pipe section exposed to an
axial tensile force, N, and a distributed lateral force, q, reads:
3 Draft Copy
and the fundamental frequency:
Looking at the governing differential equation it is obvious (6)
that most of the terms remain the same; the effects of the
EI (N piAi)
bending and axial force do not change. Note that the axial
force, N, is strictly speaking the socalled true axial force given
by the integrated axial stresses over the steel crosssection and
contains terms from residual lay tension, temperature expansion f0 C1 4 1C2
effects, Poisson effects from hoop stresses and any axial
expansion/sliding effects.
The external and internal pressure will have an indirect
meL PE
effect on this equation through the hoop expansion and the
Poisson’s effect on the true axial force. However, as shown by In the same way the effect of the external pressure can be
Galgoul et al. [8], the effect of the internal pressure can also be incorporated. It is seen that by using the definition of the
seen as a lateral force, q. Consider the infinitesimal length of effective axial force:
the pipe as shown in fig. 6 which is deformed with a radius R =
S N p i Ai p e Ae (7)
∂2v/∂x2. Due to the deformation, the pipe fibres at the
compression side become shorter than the ones at the tensile both the governing differential equation and the expression for
side. The relative shortening becomes (Rr∙sin)/R where is the frequency are simplified.
defined in fig. 7. Of course, the same expressions would have been deducted
if the concept of effective axial forces has been applied instead
of the axial force, N, from the beginning. Then the cumbersome
R=1/d2v/dx2
integration of doublecurves surfaces would have been avoided.
The effect of the pressures can also be explained by a more
physical model. During vibration of the span the pipe is being
bent, the pipe surface at the compression side becomes smaller
q than the surface at the tensile side. Thus, the internal pressure
will cause a resultant force acting away from the static/mean
dynamic position. Hence, the internal pressure will act in the
same way as increased mass or increased compressive axial
force and the frequency will drop.
Solving the differential equation and considering just the
pi
static part, i.e. neglecting the dynamic, inertia term, will give a
solution where the span buckles when the effective axial force
has reached a certain critical value depending on the boundary
Figure 3 – Deformed pipe cross section with internal conditions and any significant static span deflection. This
shows that there is a clear relationship between how the internal
pressure and with a bending radius R.
(and external) pressure affects both global buckling and the
natural frequencies of a span.
The net effect of the internal pressure will be similar to a
It is also quite obvious that the natural frequency will
distributed load, q. The integrated effect becomes:
decrease as the internal pressure increases. Finally, as the
2
R ri sin (4) 2 v effective axial force approaches the critical buckling load, the
q pi
0
R
ri sin d p i Ai / R p i Ai
x 2 frequency approaches zero, 0. This occurs for the theoretical
case with pinnedpinned support conditions. In reality, a
where ri denotes the internal pipe radius. The governing pipeline span will experience a gradual deflection or sagging
equation of motion now becomes: that will give raise to nonlinear effects not accounted for in the
(5) expressions above. By this reason, the natural frequency of free
4v 2v 2v
EI ( N pi Ai ) me 0 spans should be estimated by use of other tools, e.g. nonlinear
4 2 2
x x t FE analysis, when the effective axial force reaches a certain
4 Draft Copy
value and nonlinear effects become significant, as
recommended in the DNVRPF105 (2006) [2].
For most real spans, the span deflection will gradually
increase as the effective axial force tends towards the critical
buckling value as correctly stated by Palmer and Kaye [14] and
Galgoul et al [8]. This deflection will release some of the axial
force and may also cause the span to vanish as the span height
is limited. Anyway, as the effective axial force increases in
compression and approaches the theoretical buckling limit, the
pipeline response becomes complicated and highly nonlinear.
Therefore, expressions based on linear beam theory can not be
applied and must be replaced by for example nonlinear FE
analysis.
Due to this, the range of applicability for the expressions Figure 4 - Experimental setup used by Massa et al. [10].
based on beam theory in DNV RPF105 (2006) [2] is limited
to short and moderate span lengths L/D < 140, moderate
deflections /D < 2.5 and moderate geometrical stiffness effects Based on the experimental results and FE analysis
C2Seff/PE < 0.5 This is explained in more detail in the paper by performed for this single test case, Massa et al. suggested the
Fyrileiv and Mørk [13]. following modified expression for the natural frequency of a
span:
EXPERIMENTAL RESULTS
Massa et al. [10] conducted some experiments where a EI NL2
2
f 0 B1 1 2 B2 (8)
free span was supported in a test rig and instrumented, see fig. 2 me L4 EI D
Figure 4. Then the internal pressure was gradually increased,
and the natural frequency measured. Massa et al. support in Here N is the true axial force (resultant of the pipe wall stress).
general the concept of effective axial force being a simplified The stiffness contribution from concrete coating is not
solution to global buckling and some other cumbersome included, neither is the effective span length as this equation is
problems but claim that that the concept has its limitations as assumed valid for a pinnedpinned case. The boundary
well. One of these is the application to the structural response condition coefficients B1 and B2 are set to 0.906 and 1.580
of free spanning pipelines and the natural frequency in respectively.
particular. Based on the test observations they claim that the Comparing this with the corresponding expression in
pipe wall stresses (the true axial force) will govern the natural DNVRPF105 (2006) [2] shows that the main difference is the
frequencies for the span. Thus, increased internal pressure will change from effective axial force to true axial force together
cause pipe wall tension and increased frequency. This is the with the boundary condition coefficient of the sagging term.
direct opposite effect than predicted by DNV RPF105 (2006) Figure 5 gives the main results of Massa et al.’s
[2]. comparison between their experimental results, their FE ana
lysis and their use of the expression for the frequency in DNV
RPF105 (2006) [2] (same as Eq. (2)). Three observations can
be made looking at this figure:
The analytical expression and the FE solution agree for
zero pressure whereas they move in opposite directions
with increasing internal pressure. However, note that the
calculations according to DNV-RP-F105 are probably
incorrect. This claim will be explained later.
The equation from DNV-RP-F105 predicts zero frequency
(or infinite period) when buckling occurs at approximately
20 MPa internal pressure which is in accordance with also
Massa et al.’s understanding.
5 Draft Copy
The experiment is approximately 60% off at zero pressure Figure 6 – Comparison of natural frequencies
and shows poor agreement with both the analytical given by FE analysis and by DNV-RP-F105
approaches for increasing internal pressure. with experimental values by Massa et al.
All curves spread out significantly and the experimental [10], long span, L/D = 151.
data does not really show that the DNV-RP- F105 (2006) [2]
approach using the effective axial force is wrong. As the tested span is classified as a quite long span it is
expected that 2nd order stiffening effect from the span deflection
(sagging term) will be significant for the natural frequency in
the vertical plane. As the internal pressure increases, the span
will release the potential build up of effective axial force in
compression by deflecting even more. The effect on the
frequency will be that the significance of the effective axial
force term that tend to decrease the frequency more or less
vanish and the effect of the sagging term that tend to increase
the frequency increases.
To check the effect of the internal pressure without any
significant influence from the sagging term, a moderate length
Figure 5 – Comparison of natural frequencies span of L/D = 50 has also been analysed. All input parameters
made by Massa et al. [10]. and boundary conditions were kept the same as for the long
span case of L/D = 151. The results are shown in Figure 7 with
In order to investigate why the results reported by Massa et
frequencies from FE analysis and DNVRPF105 (2006) [2].
al. [10] do not fit with the estimates from DNVRPF105, the
As seen the results compare very well, and the natural
test pipe was later analysed by DNV using a FE model with
frequency decreases as the internal pressure increases. For this
shell elements in the commercial FE program Abaqus.
span length the sagging term is negligible, and the effect of the
The results of the FE analysis for different internal
internal pressure is stemming from the effective axial force and
pressures are shown in Figure 6. As seen the frequency matches
the buckling term.
more or less perfectly the frequencies measured by Massa et al.
This is in contrast to the FE results shown by Massa et al.
which showed a more rapid increase with the pressure. Also the
frequencies predicted by the DNVRPF105 increases with the
pressure increase in DNV’s calculations, also shown in Figure
6. However, as the range of applicability for both span length
(L/D = 151 > 140), static deflection or sag (d/D < 2.5) and axial
force (Seff/Pcr > 0.5) are exceeded; only the start value with
zero internal pressure is regarded as reliable. This value does
however match the experimental value very well in contrast to
the comparison showed in Figure 5 made by Massa et al.
The expression for the natural frequency as given in DNV
RPF105, Eq. (2) consists of three different terms, the bending
stiffness contribution, the contribution from the effective axial
force, due to nonlinear effects or buckling, and the
contribution from the deformed shape, the arc stiffening effect.
6 Draft Copy
Figure 8 – Relative importance of different Figure 9 – Comparison of natural frequencies
terms in expression for natural frequency given by FE analysis, by DNV-RP-F105 and by
according to DNV-RP-F105, L/D = 151. modified expression by Massa et al. [10] for
medium span length, L/D = 50.
These three terms are plotted with relative importance in
Figure 8 for the tested span of L/D = 151. Here it is seen that
the sagging totally dominates the frequency when the internal DNV-RP-F105 SPAN RESPONSE QUANTITIES
pressure is increased. The direct contribution from the effective The following approximate response quantities based on
axial force is more or less negligible. However, as the internal the effective span length concept are given in the new DNV
pressure increases, the effective axial force causes the sagging RPF105 (2002) [2].
to increase and thereby the frequency to increase. This means The expression for the lowest natural frequency reads:
that the indirect effect of the increase in internal pressure is the (9)
EI S eff
2
opposite of the expected effect from the effective axial force f 0 C1 1 CSF 1 C 2 C3
me L4eff PE D
term and dominates the change in frequency. Please, note
however that these relative importance numbers should be
where is the static/steadystate deflection at span midpoint.
taken as indications rather than exact numbers for this case as
Hence, the geometrical stiffening effect due to sagging is
the range of applicability for Eq. (2) is exceeded. It is also of
accounted for by the C3 term. For the horizontal direction C3
interest to check the proposed expression for the frequency for
normally equals zero. Note that PE is now based on effective
another case than the one used in the experiment. Using a
span length, i.e. PE=2EI/Leff2
moderate span length more typical for most free spanning
The modal (dynamic) stress for one diameter maximum
pipeline cases with L/D of 50, some interesting results are
deflection amplitude is derived from the mode shape and is
revealed, see Figure 9. Here the FE analysis is assumed to give
given by:
reliable results as no experimental data exists. The frequencies
predicted by DNVRPF105 is very close to the FE results and D ( Ds t ) E (10)
shows a decreasing trend as the internal pressure increases.
AIL / CF C 4 1 CSF
L2eff
This is as expected as the sagging term for this case is
negligible. The modified expression by Massa et al. shows the where Ds is the steel pipe diameter and D is the outer pipe
opposite effect. Starting below the value of the FE analysis, it diameter (including any coating). t is the steel pipe wall
shows an increasing trend with the internal pressure as claimed thickness and C4 is a boundary condition coefficient.
by the originators of this modified expression. This is simply The static bending moment may be estimated by:
not correct. (11)
q L2eff
M static C5
S
1 C 2 eff
PE
where q represents the loading, i.e. the submerged weight of the
pipe in the vertical (crossflow) direction or the drag loading in
7 Draft Copy
the horizontal (inline) direction. This expression is based on The approximate response quantities specified in this
the static moment of a span with an approximate correction due section may be applied for free span assessment provided:
to the effective axial force. Such a correction is discussed by Conservative assumptions are applied with respect to span
Timoshenko & Gere [15] and is found to give an error less than lengths, soil stiffness and effective axial force.;
The span is a single span on a relatively flat seabed, i.e. the
2% for C2Seff/PE < 0.6. Limiting the effective axial force in this
span shoulders are almost horizontal and at the same level;
way means that other sources such as the effective axial force The symmetrical mode shape dominates the dynamic
level and the soil stiffness will be the dominating sources of response (normally relevant for the vertical, cross-flow
uncertainty and not this approximation itself. response only). Here the following limits apply:
In case the static deflection is not given by direct L/Ds < 140
/D < 2.5
measurement (survey) or estimated by more accurate analytical
Note that these are not absolute limits; the shift in cross-
tools, it may be estimated by: flow response from the symmetrical to the unsymmetrical
(12) mode will depend on the sagging and the
q L4eff 1 levelling/inclination of the span shoulders. In cases where
C6 a shift in the cross-flow response is considered as likely,
EI (1 CSF ) S
1 C 2 eff the structural response of the span should be assessed by
PE using FE analysis including all important aspects;
Bar buckling is not influencing on the response, i.e.
C2Seff/PE > -0.5;
The boundary condition coefficients C1 to C6 are given in
Table 1 for different support conditions. The present
INDUSTRY PRACTICE
formulation for the effective span length implies that the values
The industry practice for the last decades when designing a
for single spans on elastic supports (seabed) approach the fixed
new pipeline with free spans or assessing free spans during
fixed values, as the effective span length becomes equal to the
operation of the pipeline can be summarised as use of:
real span length (very long free spans and very high soil
Simplified response quantities; and
stiffness).
Non-linear finite element analysis.
In case of single spans on relatively even seabed and with
Table 1 Boundary conditions coefficients moderate operational temperature and internal pressure,
Pinned Fixed Single span on seabed simplified response quantities e.g. as given by DNVRPF105
Pinned 2) Fixed 3) can be an efficient and quite accurate way of assessing the
spans.
C 1.57 3.56 3.56
In case of uneven seabed, multispanning sections, high
1
temperature and/or high internal pressure (HTHP conditions)
C 1.00 0.25 0.25
and very long spans, more accurate methods than the
2
expressions based on beam theory are required. The industry
C 0.8 1) 0.2 1) 0.4 1)
practice is then to apply common nonlinear FE analysis
3
software such as Abaqus and Ansys. The span frequency is
C 4.93 14.1 Shoulder: 14.1(L/Leff)2
hard to estimate for several reasons like boundary conditions
4 Midspan: 8.6
(where does the span start and end, what is the soil stiffness
C 1/8 1/12 Shoulder: etc), effective axial force and effect of geometrical shape
1
5
(sagging). Therefore, the most reliable analytical/numerical
18( Leff / L) 2 6 estimate of the frequency is found by applying nonlinear FE
Midspan: 1/24 analysis. Then the survey results may be utilised as input to the
C 5/384 1/384 1/384 analysis and/or as data to check the realism in the FE results.
6
FE analysis of pipeline on uneven seabed is normally
based on beam/pipe elements. All the most significant effects
Note that different effective span lengths apply in the with respect to span frequency are accounted for such as:
horizontal and vertical directions and for the static and dynamic Axial and bending stiffness
response quantities in case of different soil stiffness. Boundary conditions like soil stiffness and resistance
8 Draft Copy
Axial force – release by sagging into spans and lateral In Figure 11 below an example of a pipeline installed on
buckling uneven seabed is shown. Due to the internal pressure and
Geometrical shape – 2nd order stiffness effects
temperature, the pipeline will lift off the seabed at high points
Added hydrodynamic mass
and buckle sideways. The effect on the free span lengths and
their fatigue lives in this area are strongly influenced by the
In additional the effect of different operational conditions
buckling response.
like aslaid (empty), hydro test, operational and design
Figure 12 shows the fatigue lives as functions of span
conditions may be simulated.
length for the case without buckling and full effective axial
As spans are supported by the seabed with a certain soil
force and the case after release of axial force by buckling. The
stiffness the exact boundary conditions are hard to determine.
fatigue assessment is based on the SN approach. It can easily
Uneven seabed and multispanning span section further
be seen that the maximum acceptable span length (given a
complicates the boundary conditions when applying
design life of 50 years) increases significantly from around 35m
approximate expressions for the frequency et cetera. An
to 72m when the release of effective axial force due to lateral
example of this is shown in Figure 9, where the lower part
buckling is accounted for.
illustrates the pipeline and seabed with its spans and the upper
part shows the three most relevant vertical (crossflow)
vibration modes.
9 Draft Copy
force the nonlinear FE method provides a tool to realistically Configurations Subjected to Vortex Induced Vibrations”,
model and analyse all important effects. ISOPE’98, Montreal, Canada, May 2429, 1998.
Present industry practice is to use nonlinear finite element [5] DNV’76, “Rules for the Design, Construction and
tools based on beam/pipe elements. The geometric stiffness of Inspection of Submarine Pipelines and Pipeline Risers”,
these elements is normally based on the effective axial force Det Norske Veritas January 1976.
concept and should therefore give a reliable and quite accurate [6] Palmer, A.C. and Baldry, J.A.S., “InService Buckling of
estimate of the pipeline response. Heated Pipelines”, Journal of Petroleum Technology, pp.
12831284, November 1974.
CONCLUSIONS [7] Sparks, C.P., “The Influence of Tension, Pressure and
The following comments can be used to summarise the Weight on Pipe and Riser Deformations and Stresses”,
discussion within this paper: Proceedings of OMAE, 1983.
The expression for the natural frequency as given in the [8] Galgoul, N.S., de Barros, J.C.P. and Ferreira, R.P., “The
DNV code is correct and gives good estimates as long as Interaction of Free Span and Lateral Buckling Problems”,
used within its range of application. Proceedings of IPC 2004, Calgary, Canada, 2004.
Short to moderate spans will show a drop in the natural [9] Galgoul, N.S., Massa, A.L.L. and Claro, C.A., “A
frequency as the internal pressure increases, while long Discussion on How Internal Pressure is Treated in
spans may show an increase in the natural frequency as the Offshore Pipeline Design”, Proceedings of IPC 2004,
internal pressure increases. This is due to the stiffening Calgary, Canada, 2004.
effect of increased deflection (arc effect). [10] Massa, A. L. L. , Galgoul, N. S., Guevara Jr., N. O.,
Using realistic values of the effective axial force and the Fernandes, A. C. , Coehlo, F. M., and Neto, S. F. da
static deflection cause the DNVRPF105 (2006) Silva., “The influence of internal pressure on pipeline
expression for the natural frequency to fit very well with natural frequency”. Proceedings of OMAE2009, Hawaii,
the frequencies measured by Massa et al. [8]. USA, 2009.
The stiffening effect of the increased static deflection due [11] Fyrileiv, O. and Collberg, L., “ Influence of
to increased internal pressure was wrongly interpreted as Pressure in Pipeline Design Effective Axial Force”,
the effect of the true axial force by Massa et al. Proceedings of OMAE2005, Halkidiki, Greece, 2005.
In many cases, and especially in case of uneven seabed, [12] Clough, R.W. and Penzien, J., “Dynamics of Structures”,
global buckling and/or long and multispanning sections, McGrawHill, 1975.
use of nonlinear FE analysis is highly recommended to [13] Fyrileiv, O. and Mørk, K., “Structural Response of
account for variations in effective axial force and Pipeline Free Spans based on Beam Theory”, Proceedings
geometrical stiffness effects. of OMAE2002, Oslo, Norway, 2002.
A common misinterpretation of DNVRPF105 is that the [14] Palmer, A.C. and Kaye, D., “Rational Assessment Criteria
limitations in range of application for the response for Pipeline Spans”, OPT, Offshore Pipeline Technology
quantities are valid for the code in general. This is not conference, 1991.
correct. There is no known limitation in the span length as [15] Timoshenko, S.P. and Gere, J.M., “Theory of Elastic
long as the span frequency and modal stress are estimated Stability”, McGrawHill 1961.
within acceptable accuracy e.g. by nonlinear FE tools.
REFERENCES
[1] DNVOSF101, DNV Offshore Standard “Submarine
Pipeline Systems”, Det Norske Veritas, 2007.
[2] DNVRPF105, DNV Recommended Practice, ”Free
Spanning Pipelines”, Det Norske Veritas, 2006.
[3] Fyrileiv, O., Mørk, K.J., Kathrud, K., & Sortland, Lars
“Free Span Assessment of the Zeepipe IIA Pipeline”,
OMAE’98, Lisboa, Portugal, July 69, 1998.
[4] Kristiansen, N.Ø., Tørnes, K., Nystrøm, P.R. and
Damsleth, P., “Structural Modelling of MultiSpan Pipe
10 Draft Copy