Proceedings of the ASME 2016 35th International Conference on Ocean, Offshore and Arctic Engineering

OMAE2016
June 19-24, 2016, Busan, South Korea

OMAE2016-54238

DNV GL-ST-F101 COMBINED LOADING CRITERION -

BACKGROUND AND DERIVATION

Leif Collberg Erik Levold

DNV GL Statoil
Høvik, Norway Trondheim, Norway

ABSTRACT S Effective force, the externally applied force

The local buckling criterion for combined loading in DNV = N − pi ⋅ Ai + pe ⋅ Ae
GL-ST-F101 (DNV-OS-F101) [1] was introduced in its present
Sp Axial plastic capacity
format in the 2007 revision. The formulation is based on a
=fy(D-t)πt
plasticity theory, modified for strain hardening and, finally,
t Thickness
finite element calculations. Derivation of this formula has not
been published before and is given in the paper. Further, there is
some confusion in the industry with respect to the use of
effective force in the formulation which will be explained. This Greek
paper is supported by a paper that compares this formulation αc Strain hardening correction
with tests and other criteria, Fyrileiv et al [2]. Δp difference in internal pressure compared to during laying
σ stress

NOMENCLATURE Subscripts
b bursting
c compressive
Definitions
E Equivalent (von Mises) stress
Effective force The externally applied force, see S
h hoop
Local buckling “Buckling mode confined to a short length
l longitudinal
of the pipeline causing gross changes of the
m mean/middle
cross section; collapse, localized wall
p plastic
wrinkling and kinking are examples
r radial
thereof”. (DNV-OS-F101)
t tension
Pipe wall force, The integrated axial stress at a cross section,
y yield
see S. Also referred to as True force.
Specified Minimum Yield stress, fy as per DNV-OS-F101
definition
INTRODUCTION
Symbols
The DNV pipeline code was a pioneer introducing the limit
A Cross sectional area
state based design for pipelines in 1996 [3]. The concept
As Cross sectional steel area,
implies that the object to be designed is checked versus relevant
=(D-t)πt
failure modes and was not new as such, as it had been used for
D Nominal outside diameter
structural design of other objects for many years. Not only the
Mp Plastic moment capacity
limit state concept was new to pipeline design but also the use
= Dm ⋅ t ⋅ f y
2
of reliability based safety factors. The calibration of the safety
N Pipe wall force, or true force factors had been enabled through the SUPERB project where
qh Hoop stress to yield ratio

1 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo

 statistical basis for pipeline properties was established, Jiao et considered are denoted standard X65 and seamless
al [4], [5]. X65.
- Internal pressure ratio, qh, equal to 0.0, 0.2, 0.4, 0.6
Even though “failure mode” sounds quite defined, it can be and 0.8.
argued that only two failure modes exists; fracture and - Axial load ratio N/Ny between 0.240 and -0.587.
instability. These can be further refined into more specific limit
states and a general overview of the most relevant ones are now The total number of simulations is 120: 80 with the X65
tabled in DNV-OS-F101 [1]. The most important limit state for seamless material and 40 with the standard X65 material. After
pipelines is the pressure containment limit state. Another very the development of the combined loading criterion DNV GL
important limit state is the local buckling limit state. This has got access to project data for a total of 185 FE analyses
describes gross deformation of the cross section and are ranging from X52-X70, with Ramberg-Osgood as well as
described for; External pressure only (collapse), Propagating Luder plataou type material and D/t from 15-48. The fit has
buckling and combined loading criterion. The latter describes been confirmed also for these tests.
the moment or strain capacity in the presence of axial load,
pressure and moment. The combined loading criterion will The local buckling – combined loading criterion is referred to
often be the critical limit state when determining the as a failure mode, an instability failure mode. But what is
intervention work related to global buckling of pipelines in failure? Vitali [10] propose 10 different failure modes for
accordance with DNV-RP-F110 [6]. This can be applied to pressurized pipelines exposed to internal pressure. For the
uneven seabed or in combination with trawling where the range of pipes analyzed the curvature and moment capacities
Safebuck Guideline [7], [8] not applies. This constitutes a were always governed by the peak moment. Hence, the other
strong motivation to reduce excessive conservatism in this limit failure modes will occur at higher curvatures than the peak
state. Even though it was developed in the Hotpipe project [9], bending moment curvature and are of no interest if the peak
[10], the derivation has not been previously published. bending moment is defined as the failure mode, see Figure 1.

In the derivation of the combined loading criterion, the Figure 1 further illustrates that internal over pressure increase
following steps were performed. the deformation capacity (i.e. curvature) even if the moment
- Validate a FE model versus tests capacity is reduced.
- Use the FE model to establish a database of results
- Develop an analytical basis for the failure mode
- Adjust the analytical basis with the FE-database
results to derive the final criterion.

These will all be touched upon on in the paper while the main
focus will be on the analytical basis, adjusted for FE-results.
The comparison with the FE results is given in Fyrileiv [2].

BENDING MOMENT CAPACITY

Full scale tests

A literature review identified only one published test with

enough details to validate a FE-model, Mohareb [11]. The
Hotpipe project therefore performed four full scale tests with
internal over pressure, Vitali [12]. The Hotpipe four tests
consisted of two geometries and each of them with and without
a girth weld. Figure 1 Illustration of interacting failure modes as
function of bending moment and no pressure (solid
line) with pressure (dashed line)
FE-calculations
As can be seen in Figure 1 the presence of internal pressure
may lead to significant strains/deformation capacity which may
With the validated FE-model, a set of pipes were analyzed:
be defined as unacceptable prior to reaching the peak bending
- Pipe diameter of 0.508m.
moment. This needs to be controlled by a fracture limit state
- Outer diameter to thickness ratio (D/t) equal to 20, 30,
criterion.
40 and 60.
- X65 steel material with minimum specified yield
One aspect to keep in mind is that none of the FE-calculations
stress of 450MPa, Young modulus of 207000MPa and
simulated the girth weld and this is discussed later in the paper.
two different hardening coefficients. The two materials

2 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo

 ANALYTICAL BASIS
A limit state criterion should preferably be based on an
analytical basis and then simplified and adjusted to test results.
A fully plastified cross section has been selected as such an
analytical basis for further adjustments to FE-results and
outlined in the following.
For a straight pipe the maximum moment point is caused by
yielding combined with ovalisation type of cross section
deformation (that reduces the section modulus).
A more complete introduction to formulation of local buckling
limit state based on linear elastic-perfect plastic material
behavior, but with no stress variation in the radial direction is
given by Kirkemo [13][14].
The different analytical formulation may have different
advantages but “the proof of the pudding” is really how well
these fit with test results.
For the sake of the completeness of this paper, the full Figure 2 Illustration of interacting axial load,
development has been presented here, for more details bending moment and pressure.
reference is given to the above mentioned papers.

ELASTIC-PERFECT PLASTIC ANALYTICAL At = γ ⋅ Dm ⋅ t Eq. 4

FORMULATION Ac = (π − γ ) ⋅ Dm ⋅ t Eq. 5

Axial stress N = At ⋅ σ y ,t + Ac ⋅ σ y ,c = Dm ⋅ t ⋅ (γ ⋅ σ y ,t + (π − γ ) ⋅ σ y ,c ) Eq. 6

The principal stresses in a pipeline include; N − π ⋅ σ y ,c ⋅ D m ⋅ t N − Aσ ⋅ σ y ,c
γ= = ⋅π
D m ⋅ t ⋅ (σ y ,t − σ y ,c ) Aσ ⋅ (σ y ,t − σ y ,c )
Circumferential stresses due to internal pressure; Axial stress Eq. 7
due to internal pressure, effective axial force and bending
moment; and Radial stresses due to internal pressure. In Substituting the compressive and tensile yield stress with
addition shear stresses may be present due to variation in the von Mises expression above it reads:
bending moment, but is neglected in the further development.
The von Mises stress reads:
 2 
σh 3  σ h  
N − Aσ ⋅  − f y 1− ⋅
4  f y  
Eq. 1
σ e = σ h2 + σ l 2 + σ r 2 − σ h ⋅σ l − σ l ⋅σ r − σ r ⋅σ h
 2
γ=   ⋅π =
Assuming that the radial stress is much less than the other 2
3 σ 
stresses, this simplifies the von Mises formulation to: Aσ ⋅ 2 ⋅ f y 1 − ⋅  h 
σ e = σ h + σ l − σ h ⋅σ l
2 2 Eq. 2 4  f y 
Solved for the resulting yield stress in axial direction and   Eq. 8
 
replacing the von Mises stress with the yield stress, fy, it reads:
 σh 
N − Aσ ⋅
Eq. 3
π
 + 1 ⋅
2
σh 3 2 σ 3 σ  = 2
σ l,y = ± σ e − ⋅σ h = h ± f y ⋅ 1 − ⋅  h    2
2

4  f y 
2
2 4 2  3  σ h  
 Aσ ⋅ f y ⋅ 1 − 4 ⋅  f  
  y 
Bending stress The two terms in the numerator is the definition of the
In the following, a fully plastified cross section is effective axial force, S! By introduction of this effective force,
considered, see Figure 2. The compressive and tension yield the expression will be simplified. By also introducing the
stress will be given by the equation above. Assume that one plastic axial capacity, Sp, the formulation reads:
part yields in compression and one in tension, the
corresponding cross sectional properties then becomes:

3 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo

   bending moment capacity is zero, it does not give any axial
  capacity. This will be discussed later in this paper.
  Putting in the description of γ gives:
π
 + 1 ⋅
S
γ=
  2
Eq. 9   
2
 S  
 3  σ h     
⋅ − ⋅ 2

 S 1
   3  σ h  Sπ π
 + 1  =
π

4  fy 
 M = M π ⋅ 1− ⋅ ⋅ Sin
4  f y    
2  2
 1 − 3 ⋅  σ h   
The resulting formulation is therefore expressed in terms of
effective axial force even if it was derived based on true wall
 
4  fy    
stress/force as it was based on von Mises formulation.   
Expressed in terms of effective axial force makes the
formulation simpler and it is often also more natural to know   Eq. 14
 S p 
the effective force rather than the true wall force. Hence, this 2
 ⋅ 
becomes an “apparent error” that erroneously has been 3 s  Sp 2
= M p ⋅ 1− ⋅ h  ⋅ Cos  
proposed to be “corrected” in the literature, Aynbinder [15].
4  f y    
2

 1− 3 ⋅ s h  
For calculation of the bending moment, the distance to the
center axis reads:
 4  f y  
 
π The second term below the square root sign can be
D D
∫γ 2m ⋅ Cosφ ⋅ t ⋅ 2m ⋅ dφ expressed as:
⋅ [Sinφ ]γ
Dm π
yc = =  
2
(π − γ ) ⋅ Dm ⋅ t 2 ⋅ (π − γ ) 2  
3  σ h 
2
Eq. 10 ∆ p  ∆p 
2 =  =  Eq. 15
4  f y   2⋅t 2   pb 
⋅ Sin(π − γ )  ∆ −t ⋅ fy ⋅
Dm Dm
=− ⋅ Sinγ = − 
2 ⋅ (π − γ ) 2 ⋅ (π − γ )  3
γ This gives finally the simple expression of the theoretically
Dm D
∫ 2
⋅ Cosφ ⋅ t ⋅ m ⋅ dφ
2 D
correct expression of the plastic bending moment capacity
around the center axis in presence of pressure and axial load as:
yt = 0
= m ⋅ Sinγ Eq. 11
γ ⋅ Dm ⋅ t 2 ⋅γ
 
2  S ⋅p 
 Sp 2 
2
The resulting bending moment then reads:  ∆p 
M = At ⋅ σ y ,t ⋅ yt − Ac ⋅ σ y ,c ⋅ yc = M = M p ⋅ 1 −   ⋅ Cos   Eq. 16
 pb    ∆p  
2

 Sinγ Sin(π − γ )   1 −   
= Dm ⋅ t ⋅  γ ⋅ σ y ,t − (π − γ ) ⋅ σ y ,c =
2
  pb  
 2 ⋅γ 2 ⋅ (π − γ ) 
This is the formulation that was given in DNV rules [3]
1  3 2
Eq. 12 (expressed in terms of the pipe wall force N and not the
= Dm ⋅ t ⋅ ⋅  2 ⋅ f y − σ h  ⋅ Sin(γ )
2 2
effective axial force, S).
2  4 
Introducing the plastic bending moment Mp, it reads
2 Expansion
3 σ 
M = M p ⋅ 1 −  h  ⋅ Sin(γ )
By Taylor expansion of the cosine term it reads
Eq. 13
4  f y 
With γ equal to π/2 and no pressure, i.e. pure bending
moment, the bending moment becomes equal to the plastic
bending moment which is right. Another reflection to keep in
mind is that this is a derivation of the bending moment
capacity, and how the axial force and pressure influences the
bending moment capacity. It is therefore not an interaction
formulation between bending moment, pressure and axial force.
This is obvious as a γ equal to 0 or π only gives that the

4 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo

   
2
 stress. The amount of strain hardening will in addition to the
  p S   general form of the stress-strain curve also depend on the actual
2   ⋅  
 ∆p  strain level.
M = M p ⋅ 1 −    1  2 Sp
⋅ 1 − ⋅  + ...  = In a bending moment-curvature plot, the bending moment
2   
 pb 
2
 
  1 −  ∆p    peak point will increase with decreasing D/t, i.e. the maximum
  p    bending moment capacity will occur for a higher strain, the
   b   Eq. 17 lower the D/t is. Hence, the amount of strain hardening will be
 p S 
2
 higher the lower the D/t.
    By introduction of a strain hardening factor, αc, as well as
 ⋅ 
2
 ∆p  1  2 S  factor k1 and αp, and comparing this capacity equation to FE
= M p ⋅  1 −   − ⋅ 
p 
+ ...  results, the effect of material strain hardening, shortcomings
  pb  2  ∆p 
2 
 1 −    due to the Taylor expansion and effect of increasing ovality
  pb   during bending deformation could be included.
 
Truncating additional terms, it can be modified as;
2
2 2
 S 
2  M  S   
2
∆p 
2
 ∆p   ∆p  1  p  
2
M    
⋅ 1 −   ≤ 1 −   − ⋅   ⋅   α ⋅ M + k1 ⋅  α ⋅ S   +  α p ⋅ α ⋅ p  ≤ 1
Eq. 18 Eq. 23
S 
Mp  pb   pb  2  2   p  c p  c p   c b 

2 The comparison with the FE-results led to the following

 S 
2 2
 ∆p   ∆p  1  p 
2
M
⋅ 1 −   +   + ⋅   ⋅  ≤ 1 Eq. 19 recommendations.
S 
Mp  pb   pb  2  2   p
This is essentially the formulation that was in DNV-OS- 0.5 D / t 2 < 15
F101:2000 [16] with some additional terms for strain hardening  60 − D / t 
discussed later in this paper. β = 0.5 ⋅   15 < D / t 2 < 60 Eq. 24
Further elaboration of this equation yields:  45 
0 60 < D / t 2
1 p 
2

⋅ 
α c = (1 − β ) + β ⋅ u
2 f
 S 
2
M 2 2  ∆p 
⋅   ≤ 1 − 
Eq. 25
+  fy
2 S  Eq. 20
Mp  ∆p   p   b
p
1 −   α p = 1− β Eq. 26
 pb  Hence, the characteristic moment capacity is multiplied by
2 these correction factors to take benefit of strain hardening.
  Figure 3 shows that αc allows higher moment for pipes that can
 1 p 
2

 ⋅   sustain higher strains, i.e. pipes with lower D/t’s and for
 S    ∆p 
2 2

 M + 2 2 ⋅    +   ≤ 1 Eq. 21
materials with lower yield to strength ratios.
Mp S  
 pb 
2
 ∆p   p
 1 −   
  pb  
 
This give the fundamental form of the local buckling
criterion reflected in DNV-OS-F101: 2007 [17].
2
 M  S    ∆p 
2 2
    
 M + k1 ⋅  S   +  p  ≤ 1
Eq. 22
 p  p   b

ADJUSTMENTS TO NUMERICAL DATA

Benefit of strain hardening

The above equation is a slight simplification of the plastic Figure 3 Benefit of strain hardening (0.84 and
bending moment capacity around the center axis in presence of 0.90)
pressure and axial load assuming an elastic-perfect plastic Note that the αp factor does not imply an increased
material behavior. In a real pipeline, the material will have allowable internal over pressure (this is limited by the hoop
some strain hardening, giving higher stress than the initial yield

5 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo

 stress criterion) but that the detrimental effect of internal over DISCUSSION
pressure on the bending moment capacity is lower.
Axial load validity
Other The plastic axial capacity can be derived by formulating
The factor k1 in Eq. 22 is the factor from the Taylor the following equation
expansion and the simplification (separating the pressure from N = A ⋅σ + A ⋅σ
y t y ,t c y ,c
Eq. 27
the axial force). At this stage it is proposed to denote this as a
constant equal to unity and look into when selecting the safety An angle of 0 degrees, will give the compressive plastic
factors. axial capacity and an angle of pi will give the tension plastic
capacity. The capacity then becomes:
Derivation summary  2
σ  ∆p 
The different stages in the derivation of the final equation N y = ∆m ⋅ t ⋅ p  ± f y ⋅ 1 − 
h
 
pb  
Eq. 28
is shown in Figure 4. The Equation 23 allows a higher moment 2 
than equation 16 and 21 as intended.  
Again, introducing the concept of effective force, S, and
the axial plastic capacity, Sp, simplifies the expression to:
2
 ∆p 
S y = ± S p ⋅ 1 −   Eq. 29
 pb 
Squaring both sides it eventually becomes:
2
 S y   ∆p 
2

  +  =1 Eq. 30
S   p 
 p  b
If Eq. 30 is solved for the effective axial load in presence
of no moment, the formulation will be similar except for that
the axial force fraction is in power of 4 instead of in power of 2.
The magnitude of the effective axial force in a pipeline is
normally small and close to buckles or bends is close to zero. A
restrained pipeline has a compressive effective axial force
corresponding to a hoop stress of about 20%, i.e. less than 20%
Figure 4 Comparison of the different derivation
of the plastic capacity. Temperature loads may increase the
formulas
compressive force further. The maximum compressive force is
further limited by the global buckling capacity, again limited by
All curves are plotted as the moment versus the plastic
the lateral resistance, which mostly will occur at lower values
(uniaxial) moment capacity on the left hand axis and the hoop
except for buried pipelines. Buried pipelines on the other hand
stress versus the yield stress on the lower axis.
are designed to stay in place and will have negligible moment.
Without internal over pressure all equations gives similar
The axial capacity may then be determined from e.g.
values with some minor benefit for lower D/t and higher strain
Suzuki[18] as recommended in DNV-RP-F110 [5].
hardening.
Due to the above considerations, a limitation of the axial
The brown curve, giving the cosine formulation in eq. 16 is
capacity is required and was introduced in DNV-OS-F101:2012
completely covered by the truncated Taylor formulation from
[19] to avoid unintended application of the formula as:
Eq. 21 showing that the truncation has a negligible effect.
The difference between the previous two curves and the
final formulation is the influence of strain hardening and D/t. It Sd
shows that a higher strain hardening has some beneficial effect, ≤ 0.4 Eq. 31
as expected and that the effect of D/t is more pronounced
αc ⋅ S p
giving higher utilization for lower D/t as intended. For a ratio of the effective axial force to the plastic
Note that the allowable range on qh is less than 0.84 (safety capacity less than 0.4, the difference between Eq. 30 and Eq. 22
class normal) and that no safety factors are included. solved for axial capacity in presence of no moment will be in
the order of 6% and the limitation not very restrictive to
pipeline applications.
For axial load beyond this value, reference is made to
DNV-OS-F201 [20] in which the local buckling formulation is
derived from a different basis.

6 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo

 The same limitation is not required for the pressure as this CLOSURE
also has to comply with the pressure containment limit state The paper has illustrated the derivation of the local
that then will govern. Hence, it is not allowed to derive an buckling formulation – combined loading with internal over
allowable pressure or dominant axial force from this pressure in DNV-OS-F101. The publication will also hopefully
formulation! lead to avoidance of erroneous interpretations of the same in
the future.
Geometric validity
The FE-results were based on ideal pipes, without girth ACKNOWLEDGMENTS
welds simulated, ranging from 0 to hoop stress above 90% of The authors would like to acknowledge all the people that
yield and therefore covers the whole pressure application range. have contributed to the development during the last 20 years
Ghodsi et al [21] discovered that the presence of girth weld leading to the current formulation. The main contributors have
could be quite detrimental on the capacity and reductions to been Roberto Bruschi, Luigino Vitali and Enrico Torsiletti in
40% have been shown. This reduction primarily affects the Snamprogetti (now Saipem) and Ralf Peek in Shell. However,
curvature/strain capacity, as reflected in DNV-OS-F101 none of these are expected to complain if we put our DNV
displacement controlled local buckling capacity, and not on the colleague during the last 20 years, Kim Mørk first on the list of
bending moment capacity. The HOTPIPE project [9] performed those to be acknowledged as the architect behind the
four tests with internal pressure, two geometries, two with girth formulation. The authors have mainly put his work into public
welds, that confirmed a significant drop in curvature capacity domain.
results but only limited influence on the moment capacity. The
physical explanation to this reduction is not fully understood REFERENCES
and is probably due to a combination of effects. One such effect [1] DNV-OS-F101:2013, “Submarine Pipeline Systems”,
could be geometrical imperfections triggering the buckling DNV GL Høvik, Norway
capacity. Such an imperfection is likely to be more pronounced [2] Fyrileiv, O., Collberg, L. (2016) “DNV-OS-F101
for thinner pipes, i.e. higher D/t. This is further supported by Combined loading criterion - Range of application and
that the curvature reduction shown by Ghodsi, [21] was larger comparison with other criteria”. 35th International
for higher D/t’s (the tests were from D/t=60 and upwards). Conference on Ocean, Offshore and Arctic Engineering,
Even though geometries were analysed up to D/t=60 a OMAE2016-54741
limitation on geometry has been drawn at D/t=45 to account for [3] Det Norske Veritas. (1996). “DNV Rules for Submarine
these uncertainties from the girth weld. This is discussed in Pipeline Systems”.
section 13 of DNV-OS-F101 where it is stated that applications [4] Jiao, G., Sotberg, T., Bruschi, R., & Igland, R. T. (1997).
above 45 have to be supported by FE-analyses including effects “The SUPERB Project: Linepipe statistical properties and
that may be caused by the girth weld. implications in design of offshore pipelines” 16th
International Conference on Offshore Mechanics and
Characteristic internal pressure Arctic Engineering (ss. v 5, p 45-55). Yokohama:
A “design load” is for most application a conservatively American Society of Mechanical Engineers.
high load that is used for design purposes. This is not the case [5] Jiao, G., Mørk, K. J., Bruschi, R., & Sotberg, T. (1997).
for “design pressure” or MAOP which is allowed to be The SUPERB project: Reliability based design procedures
exceeded by 10%. DNV GL has therefore adopted the local and limit state design criteria for offshore pipelines. 16th
incidental pressure as a characteristic pressure in structural International Conference on Offshore Mechanics and
design. The incidental pressure is defined as the 100-year return Arctic Engineering (ss. v 5, p 57-69). ASME.
pressure and is defined at a certain elevation while the local [6] DNV-RP-F110:2007. “Global Buckling of Submarine
incidental pressure is adjusted for the column weight of the Pipelines - Structural Design due to High Temperature”.
internal media to the relevant elevation. Det Norske Veritas Høvik, Norway
Finite Element analyses shows a slight increase in bending [7] Bruton D., Carr M.,”Buckling the trend” Offshore
moment capacity for increasing internal over pressure up to a Engineer, March 2010.
certain pressure dependent on the D/t, beyond which it [8] Matheson I., “Buckling under pressure” World Pipelines,
decreases again. The derived local buckling equation does not Febrary 2014
account for this extra capacity. One reason for not allowing this [9] Mørk K. et al. (1999): “The HOTPIPE Project – Design
slight increased capacity is the less strong correlation between Guidelines for High Temperature/Pressure Pipelines”,
pressure and bending moment. How to select a pressure at Proc. 9th Int. Offshore and Polar Engineering Conference,
which a bending moment capacity should be determined? By 1999
the adopted formulation a high pressure will always be [10] Vitali, L., Bruschi, R., Mørk, K. J., Levold, E., & Verley,
conservative. R. (1999). “Hotpipe Project - capacity of pipes subject to
internal pressure, axial force and bending moment”. 9th
International Offshore and Polar Engineering Conference.
Brest, France: ISOPE.

7 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo

 [11] Mohareb M. et al. (1994): "Deformational Behaviour of
Line Pipe", Structural Engineering Report No. 202,
Department of Civil Engineering, University of Alberta,
Edmonton, Alberta, Canada T6G 2G7.
[12] Vitali, L., Bartolini, L., Askheim, D., Peek, R., & Levold,
E. (2005). HOTPIPE JI Project: Experimental test and FE-
analyses. 24th International Conference on Offshore
Mechanics and Arctic Engineering (OMAE 2005) (ss.
OMAE2005-67526). Halkidik: American Society of
Mechanical Engineers.
[13] Kirkemo, F., & Holden, H. (2001). Burst and Gross Plastic
Deformation Limit State Equations for Pipelines; Part 1 -
Theory. International Offshore and Polar Engineering
Conference. Stavanger: The International Society of
Offshore and Polar Engineers.
[14] Kirkemo, F., & Holden, H. (2001). Burst and Gross Plastic
Deformation Limit State Equations for Pipes: Part 2 --
Application. International Offshore and Polar Engineering
Conference. Stavanger: The International Society of
Offshore and Polar Engineers.
[15] Aynbinder, A. (2008, September 1st). “Proposal to modify
a subsea pipeline standard“ Offshore – Magazine article
[16] DNV-OS-F101:2000 “Submarine Pipeline Systems”. Det
Norske Veritas Høvik, Norway
[17] DNV-OS-F101: 2007 “Submarine Pipeline Systems”. Det
Norske Veritas Høvik, Norway
[18] Suzuki, N., Toyoda, M. (2002). “Seismic Loadings on
Buried Pipelines and Deformability of High Strength Line
Pipes”. Pipe Dreamer's Conference. Yokohama, Japan.
[19] DNV-OS-F101:2012 “Submarine Pipeline Systems”. Det
Norske Veritas Høvik, Norway
[20] DNV-OS-F201. (2001). Dynamic Risers. Høvik, Norway:
Det Norske Veritas.
[21] Ghodsi, N. Y., Kulak, G. L., & Murray, D. W. (1994).
«Behaviour of Girth Welded Line-pipe”. University of
Alberta, Civil Engineering. Structural Engineering Report
No. 23.

8 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89858/ on 01/31/2017 Terms of Use: http://www.asme.org/abo