Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
RECOMMENDED PRACTICE
DNV-RP-F105
© Det Norske Veritas. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including
photocopying and recording, without the prior written consent of Det Norske Veritas.
If any person suffers loss or damage which is proved to have been caused by any negligent act or omission of Det Norske Veritas, then Det Norske Veritas shall pay compensation to such person
for his proved direct loss or damage. However, the compensation shall not exceed an amount equal to ten times the fee charged for the service in question, provided that the maximum compen-
sation shall never exceed USD 2 million.
In this provision "Det Norske Veritas" shall mean the Foundation Det Norske Veritas as well as all its subsidiaries, directors, officers, employees, agents and any other acting on behalf of Det
Norske Veritas.
Recommended Practice DNV-RP-F105, March 2002
Page 3
CONTENTS
• acceptance criteria.
1 General
1.3.2 Free spans can be caused by:
• seabed unevenness
1.1 Introduction
• change of seabed topology (e.g. scouring, sand waves)
• artificial supports/rock berms etc.
1.1.1 The present document considers free spanning
pipelines subjected to combined wave and current loading. 1.3.3 The following environmental flow conditions are
The premises for the document are based on technical de- described in this document:
velopment within pipeline free span technology in recent • steady flow due to current;
R&D projects, as well as design experience from recent
• oscillatory flow due to waves; and
and ongoing projects, i.e.
• combined flow due to current and waves.
• DNV Guideline 14, see Mørk & Fyrileiv (1998)
• The sections regarding Geotechnical Conditions and The flow regimes are discussed in section 1.6.
part of the hydrodynamic model are based on the re-
search performed in the GUDESP project, see Tura et 1.3.4 There are no limitations to span length and span
al., (1994). gap with respect to application of this Recommended Prac -
• The sections regarding Free Span Analysis and in-line tice.
Vortex Induced Vibrations (VIV) fatigue analyses are
based on the published results from the MULTISPAN The basic cross-flow VIV Response model is, however,
project, see Mørk et al., (1997). based on single mode response.
• Numerical study based on CFD simulations for vibra-
tions of a pipeline in the vicinity of a trench, per- In case several potential vibration modes can become ac-
formed by Statoil, DHI & DNV, see Hansen et al, tive at a given flow velocity, the mode associated with the
2001. largest contribution to the fatigue damage shall be applied.
• Further, recent R&D and design experience e.g. from Unless otherwise documented the damage contribution for
Åsgard Transport, ZEEPIPE, TOGI and TROLL OIL any modes should relate to the same critical (weld) loca-
pipeline projects are implemented, see Fyrileiv & tion.
Mørk (1998).
The basic principles applied in this document are in 1.3.5 The free span analysis may be based on approxi-
agreement with most recognised rules and reflect state-of- mate response expressions or a refined FE approach de-
the-art industry practice and latest research. pending on the free span classification, see section 6.2.
This document includes a brief introduction of the basic The following cases are considered:
hydrodynamic phenomena, principles and parameters. For • single spans
a thorough introduction see e.g. Sumer & Fredsøe, (1997)
• spans interacting with adjacent/side spans.
and Blevins (1994). The stress ranges and natural frequencies should normally
be obtained from an FE-approach. Requirements to the
1.2 Objective structural modelling and free span analyses are given in
section 6.
1.2.1 The objective of this document is to provide ra-
tional design criteria and guidance for assessment of pipe- 1.3.6 The following models are considered:
line free spans subjected to combined wave and current • Response Models (RM)
loading. • Force Models (FM)
An amplitude response model is applicable when the
1.3 Scope and Application vibration of the free span is dominated by vortex induced
resonance phenomena. A force model may be used when
1.3.1 Detailed design criteria are specified for Ultimate the free span response can be found through application of
Limit State (ULS) and Fatigue Limit State (FLS) due to in- calibrated hydrodynamic loads. The selection of an
line and cross-flow Vortex Induced Vibrations (VIV) and appropriate model may be based on the prevailing flow
direct wave loading. regimes, see section 1.6.
The following topics are considered: 1.3.7 The fatigue criterion is limited to stress cycles
• methodologies for free span analysis; within the elastic range. Low cycle fatigue due to elasto-
• requirements for structural modelling; plastic behaviour is considered outside the scope of this
document.
• geotechnical conditions ;
• environmental conditions & loads;
• requirements for fatigue analysis;
• response and direct wave force analysis models; and
Recommended Practice DNV-RP-F105, March 2002
Page 5
1.3.8 Fatigue loads due to trawl interaction, cyclic 1.3.9 The main aspects of a free span assessment to-
loads during installation or pressure variations are not con- gether with key parameters and main results are illustrated
sidered herein but must be considered as a part of the inte- in the figure below.
grated fatigue damage assessment.
Components
Screening
Fatigue
ULS
1.5.1 An overview of typical free span characteristics is length. The ranges indicated for the normalised free span
given in the table below as a function of the free span length in terms of (L/D) are tentative and given for illus-
tration only.
Note that α=0 correspond to pure oscillatory flow due to waves and α =1 corresponds to pure (steady) current flow.
The flow regimes are illustrated in Figure 1-2.
Recommended Practice DNV-RP-F105, March 2002
Page 7
6
1.8.4 Marginal Fatigue Capacity is defined as the fa-
tigue capacity (life) with respect to one seastate defined by
5 its significant wave height, peak period and direction.
current dominated flow
4
Flow Velocity(Uc+U w)
wave dominated flow α=0.8 1.8.5 Response Model is a model where the structural
3 response due to VIV is determined by hydrodynamical
2 parameters.
α=0.5
1
1.8.6 Span Length is defined as the length where a con-
0 tinuous gap exists, i.e. as the visual span length.
α=0.0
-1
time
-2 1.9 Abbreviations
(e/D) seabed gap ratio Ntr true steel wall axial force
f0 in-line (f0,in) or cross-flow (f0,cr) natural fre - Nc soil bearing capacity parameter
quency (determined at no flow around the Nq soil bearing capacity parameter
pipe) Nγ soil bearing capacity parameter
fsn concrete construction strength pe external pressure
fs vortex shedding frequency (Strouhal pi internal pressure
U Pi probability of occurrence for i’th stress cycle
frequency) = S t q deflection load per unit length
D
FL lateral pipe-soil contact force
FV vertical pipe-soil contact force PE Euler load = (1+CSF)π2 EI/Leff 2
fv dominating vibration frequency Ra axial soil reaction
fw wave frequency Rc current reduction factor
F() distribution function RD reduction factor from wave direction and
g gravity spreading
gc correction function due to steady current Rv vertical soil reaction
gD drag force term RIθ reduction factor from turbulence and flow
gI inertia force term direction
G shear modulus of soil or incomplete comple - Rk reduction factor from damping
mentary Gamma function UD
Re Reynolds number D=
G(ω) frequency transfer function from wave eleva- ν
tion to flow velocity s spreading parameter
Heff effective lay tension S stress range, i.e. double stress amplitude
HS significant wave height Ssw stress at intersection between two SN-curves
h water depth, i.e. distance from the mean sea Seff effective axial force
level to the pipe Sηη wave spectral density
I moment of inertia SSS stress spectra
Ic turbulence intensity over 30 minutes SUU wave velocity spectra at pipe level
ip plasticity index, cohesive soils su undrained shear strength, cohesive soils
k Wave number St Strouhal number
kc soil parameter or empirical constant for con- t pipe wall thickness or time
crete stiffening Texposure load exposure time
kp peak factor Tlife fatigue design life capacity
ks soil coefficient Tp peak period
kw normalisation constant Tu mean zero up-crossing period of oscillating
K soil stiffness flow
KL lateral (horizontal) dynamic soil stiffness Tw wave period
KV vertical dynamic soil stiffness U current velocity
(k/D) pipe roughness Uc current velocity normal to the pipe
U Us significant wave velocity
KC Keulegan Carpenter number = w
fw D Uw significant wave induced flow velocity normal
4 πm e ζ T to the pipe, corrected for wave direction and
KS stability parameter = spreading
ρD 2 v vertical soil settlement (pipe embedment)
K0 coefficient of earth pressure at rest U + Uw
k1 soil stiffness VR reduced velocity = c
f0D
k2 soil stiffness
w wave energy spreading function
L free span length, (apparent, visual)
y lateral pipe displacement
La length of adjacent span
z height above seabed or in-line pipe displace-
Leff effective span length
ment
Ls span length with vortex shedding loads
zD height to the mid pipe
Lsh length of span shoulder
zm macro roughness parameter
me effective mass per unit length
z0 sea-bottom roughness
m fatigue exponent
zr reference (measurement) height
m(s) mass per unit length including structural
mass, added mass and mass of internal fluid
Mn spectral moments of order n
ni number of stress cycles for stress block i
N number of independent events in a return
period
Ni number of cycles to failure for stress block i
Recommended Practice DNV-RP-F105, March 2002
Page 9
2.3.3 The in-line natural frequency f0,IL must fulfil: where η is the allowable fatigue damage ratio, Tlife the
fatigue design life capacity and Texposure the life or load
f 0, IL Uc ,100year L / D γ IL exposure time.
> ⋅ 1 − ⋅
γf VRIL,onset ⋅ D 250 α
2.4.2 The fatigue damage assessment is based on the
Where accumulation law by Palmgren-Miner:
n
γf Safety factor on the natural frequency, see Dfat = ∑ i
Ni
2.6
γIL Screening factor for inline, see 2.6 Where
Current flow ratio=
α Uc ,100year D fat Accumulated fatigue damage.
max ;0.6
Uw ,1year + Uc ,100year ni Total number of stress cycles corresponding to the
(mid-wall) stress range Si
D Outer pipe diameter incl. coating N Number of cycles to failure at stress range Si
L Free span length Σ Implies summation over all stress fluctuations
Uc,100year 100 year return period value for the cur- in the design life
rent velocity at the pipe level, see 3
Uw,1year Significant 1 year return period value for 2.4.3 The number of cycles to failure at stress range S
the wave induced flow velocity at the pipe is defined by the SN curve of the form:
level corresponding to the annual signifi-
cant wave height Hs,1year, see 3 a ⋅ S − m1 S > S sw
VRIL, onset In-line onset value for the reduced veloc- N = 1 −m
ity, see 4. a 2 ⋅ S 2 S ≤ S sw
>
U w ,1year + U c,100year 3 (a2; m2 )
2.4.1 The fatigue criterion can be formulated as: Figure 2-2 Typical two-slope SN curve.
η⋅ Tlife ≥ Texposure
2.4.4 The SN-curves may be determined from: Where PHS , TP , θ is the probability of occurrence of each
• Dedicated laboratory test data,
individual sea-state, e.g. the probability of occurrence re-
• Accepted fracture mechanics theory, or
flected by the cell in a scatter diagram. The in-line fatigue
• DNV-RP-C203 “Fatigue Strength Analysis of Off-
life capacity is conservatively taken as the minimum ca-
shore Steel Structures”.
pacity (i.e., maximum damage) from VIV (RM) or direct
The SN-curve must be applicable for the material,
wave loads (FM) in each sea state.
construction detail, location of the intial defect (crack
initiation point) and corrosive environment. The basic
The fatigue life is the minimum of the in -line and the
principles in DNV-RP-C203 apply.
cross-flow fatigue lives.
2.4.5 The fatigue life capacity, Tlife , can be formally
2.4.9 The following marginal fatigue life capacities are
expressed as:
evaluated for (all) sea states characterised by (Hs , Tp , θ)
1
Tlife = RM , IL
THs Marginal fatigue capacity against in-
f v ⋅ S im ⋅ Pi , Tp , θ line VIV and cross-flow induced in-
∑
line motion in a single sea-state (Hs , Tp ,
a θ) integrated over long term pdf for the
Where current, see section 4.2.2.
RM , CF Marginal fatigue capacity against
THs, Tp, θ
Pi Probability of occurrence for the “i”th stress cross-flow VIV in a sea-state (Hs , Tp , θ)
cycle integrated over long term pdf for the
current, see section 4.2.1.
2.4.6 The concept adopted for the fatigue analysis ap- FM , IL Marginal fatigue capacity against direct
THs, Tp, θ
plies to both response models and force models. The stress wave actions in a single sea-state char-
ranges to be used may be determined by: acterised by (Hs , Tp , θ) using mean
• a response model, see section 4 value of current, see section 5.2.2.
• a force model, see section 5.
2.4.7 The following approach is recommended: 2.4.10 Unless otherwise documented, the following
• The fatigue damage is evaluated independently in assumptions apply:
each sea-state, i.e., the fatigue damage in each cell of • The current and wave induced flow components at the
a scatter diagram in terms of (Hs , Tp , θ) times the pipe level are statistically independent.
probability of occurrence for the individual sea state. • The current and wave-induced flow are assumed co-
• In each sea-state (Hs , Tp , θ) is transformed into (Uw , linear. This implies that the directional probability of
Tu , θw ) at the pipe level as described in section 3.3. occurrence data for either waves or current (the most
• The sea state is represented by a significant short-term conservative with respect to fatigue damage) must be
flow induced velocity amplitude Uw with mean zero used for both waves and current.
up-crossing period Tu , i.e. by a train of regular wave
induced flow velocities with amplitudes equal to Uw 2.5 ULS Criterion
and period Tu . The effect of irregularity will reduce
the number of large amplitudes. Irregularity may be 2.5.1 For the local buckling check reference is made to
accounted for provided it is properly documented. the combined loading – load controlled condition criterion
• Integration over the long-term current velocity distri- in DNV-OS-F101, section 5, D500. Static and dynamic
bution for the combined wave and current flow is per- bending moment, axial force and pressure shall be ac-
formed in each sea-state. counted for.
2.4.8 The total fatigue life capacity in the in -line and 2.5.2 For extreme wave conditions, which can be as-
cross-flow direction is established by integrating over all sumed to cause large deformations on the shoulders, de-
sea-states, i.e. tailed analyses of the soil stiffness at the shoulders may be
−1 required. In lieu of detailed documentation, the boundary
PHs,T p,θ conditions for the free span should be assumed as pinned-
= ∑ ∑∑
( )
IL
Tlife
θ H T min T RM ,IL ; T FM,IL pinned (only valid for the Force Model calculations).
S P Hs,T p,θ Hs,T p,θ
−1 2.5.3 The maximum dynamic bending moment due to
PHs,T p,θ VIV and/or direct wave action may be found from the dy-
CF
Tlife
= ∑ ∑∑ RM ,CF
θ H T T namic stresses:
S P Hs,T p,θ
M E = σ dyn
2 ⋅I
Ds − t
Recommended Practice DNV-RP-F105, March 2002
Page 13
• The return period stress range, e.g. 100 year, for in- D fat = Texp osure ⋅ ∑ ≤η
line VIV, Sin , defined in section 4.3. a
• The stress from 50% of the cross-flow induced VIV γf, γon , γk and γs denote partial safety factors for the natural
motion. All parameters are defined in section 4.4. frequency, onset of VIV, stability parameter and stress
range respectively. The set of partial safety factor to be
2.5.6 The maximum dynamic stress σFM,max from direct applied are specified in the table below for the individual
wave loading may be calculated using a “design storm” safety classes.
approach using:
σFM,max= kp σs Table 2-2 Safety factors for fatigue
Safety Class
Safety Factor
where kp is the peak factor given by: Low Normal High
η 1.0 0.5 0.25
γs 1.051) (1.0)
k p = 2 ln (f v ∆T ) +
0.577
2 ln (f v ∆T ) γf 1.201) (1.15)
γk 1.30
∆T is the storm duration equal to 3 hour and fv is given in γon 1.10
section 5.2. kp may conservatively be taken equal to 4. 1) This safety factor is intended to be used in design when detailed
data about span length, gap etc is not known. If a span is assessed
σs is the standard deviation of the stress (amplitude) re - in-service with updated and measured span data, the safety factor in
sponse calculated from a time domain or frequency do- brackets may be used.
main analysis, see section 5.
Comments:
2.6.3 The reliability of the pipeline against local buck-
• η apply to both Response Model and Force Model
ling (ULS criterion) is ensured by use of the safety class
• γs is to be multiplied to the stress (S γS ) concept as implemented by use of safety factors according
• γf applies to the natural frequency (fo /γf) to DNV-OS-F101, section 5 D500 alternatively section 12.
• γon applies to onset values for in-line and cross-flow
VIV (VR,on/γon )
• γk applies to the stability parameter (KS/γk )
• For ULS, the calculation of load effects is to be per-
formed without safety factors (γS = γ f = γk = γon = 1.0),
see also section 2.6.3.
Recommended Practice DNV-RP-F105, March 2002
Page 15
3.1 General
3.2.2 For water depths greater than 100 m, the ocean
currents can be characterised in terms of the driving and
3.1.1 The objective of the present section is to provide
steering agents:
guidance on:
• the driving agents are tidal forces, pressure gradients
• the long term current velocity distribution,
due to surface elevation or density changes, wind and
• short-term and long-term description of wave induced
storm surge forces.
flow velocity amplitude and period of oscillating flow
• the steering agents are topography and the rotation of
at the pipe level, and
the earth.
• return period values.
The modelling should account adequately for all agents.
3.1.2 The environmental data to be used in the assess-
3.2.3 The flow can be divided into two zones:
ment of the long-term distributions shall be representative
• an Outer Zone far from the seabed where the mean
for the particular geographical location of the pipeline free
current velocity and turbulence vary only slightly in
span.
the horizontal direction.
3.1.3 The flow conditions due to current and wave ac- • an Inner Zone where the mean current velocity and
tion at the pipe level govern the response of free spanning turbulence show significant variations in the horizon-
pipelines. The principles and methods as described in tal direction and the current speed and direction is a
Classification Note CN 30.5 may be used in addition to function of the local sea bed geometry.
this document as a basis when establishing the environ-
mental load conditions.
3.2.4 The outer zone is located approximately one local
3.1.4 The environmental data must be collected from seabed form height above the seabed crest. In case of a flat
seabed, the outer zone is located approximately at height
periods that are representative for the long-term variation
of the wave and current climate, respectively. In case of (3600 z0 ) where z0 is the bottom roughness, see Table 3-1.
less reliable or limited number of wave and current data,
the statistical uncertainty should be assessed and, if sig- 3.2.5 Current measurements (current meter) should be
nificant, included in the analysis. made in the outer zone outside the boundary layer at a
level 1-2 seabed form heights above the crest. For large-
3.1.5 Preferably, the environmental load conditions scale currents, such as wind driven and tidal currents, the
choice of measurement positions may be based on the
should be established near the pipeline using measurement
data of acceptable quality and duration. The wave and cur- variations in the bottom topography assuming that the cur-
rent characteristics must be transferred (extrapolated) to rent is geo-strophic, i.e., mainly running parallel to the
large-scale bottom contours.
the free span level and location using appropriate conser-
vative assumptions.
Over smooth hills, flow separation occurs when the hill
slope exceeds about 20o . Current data from measurements
3.1.6 The following environmental description may be
in the boundary layer over irregular bed forms are of little
applied:
practical value when extrapolating current values to other
• Directional information, i.e., flow characteristic versus
locations.
sector probability;
• Omnidirectional statistics may be used if the flow is
uniformly distributed. 3.2.6 In the inner zone the current velocity profile is
If no such information is available, the flow should be approximately logarithmic in areas where flow separation
assumed to act perpendicular to the axis of the pipeline at does not occur:
all times. (ln(z ) − ln(z 0 ) )
U ( z) = R c ⋅ U ( z r )
(ln (z r ) − ln(z 0 ) )
3.2 Current conditions
where:
3.2.1 The steady current flow at the free span level may Rc reduction factor, see 3.4.1.
be a compound of: z elevation above the seabed
• tidal current; zr reference measurement height (in the outer zone)
z0 bottom roughness parameter to be taken from
• wind induced current;
Table 3-1:
• storm surge induced current, and
• density driven current.
Table 3-1 Seabed roughness 3.2.10 For ULS, 1 min average values should be applied.
The 1 minute average values may be established from 10
Seabed Roughness z0 (m) or 30 min average values as follows:
Silt ≈ 5 10-6 (1 + 1.9 ⋅ I c ) ⋅ U10min
U1min =
fine sand ≈ 1 10-5 (1 + 2. 3 ⋅ I c ) ⋅ U30 min
Medium sand ≈ 4 10-5
where Ic is the turbulence intensity defined below.
coarse sand ≈ 1 10-4
Gravel ≈ 3 10-4 3.2.11 The turbulence intensity, Ic, is defined by:
Pebble ≈ 2 10 -3 σ
Ic = c
Cobble ≈ 1 10-2 Uc
Boulder ≈ 4 10-2 where σc is the standard deviation of the velocity fluctua-
tions and Uc is the 10 min or 30 min average (mean) veloc-
3.2.7 If no detailed analyses are performed, the mean ity (1 Hz sampling rate).
current values at the free span location may assume the
values at the nearest suitable measurement point. The flow
(and macro-roughness) is normally 3D and transformation 3.2.12 If no other information is available, the turbu-
lence intensity should be taken as 5%. Experience indi-
of current characteristics should account for the local bot-
cates that the turbulence intensity for macro -roughness
tom topography e.g. guided by numerical simulations.
areas is 20-40% higher than the intensity over a flat seabed
with the same small-scale seabed roughness. The turbu-
3.2.8 For conditions where the mean current is spread
lence intensities in a rough seabed area to be applied for
over a small sector (e.g. tide-dominated current) and the
in-line fatigue assessment may conservatively be taken as
flow condition can be assumed to be bi-dimensional, the
typical turbulence intensities over a flat bottom (at the
following model may be applied in transforming the mean
same height) with similar small-scale seabed roughness.
current locally. It is assumed that the current velocity U(zr )
in the outer zone is known, see Figure 3-1. The velocity
profile U(z*) at a location near the measuring point (with 3.2.13 Detailed turbulence measurements, if deemed
essential, should be made at 1 m and 3 m above the sea-
zr*>zr) may be approximated by:
bed. High frequency turbulence (with periods lower than 1
(ln (z *) − ln(z m ) ) minute) and low frequency turbulence must be distin-
U (z*) = U ( z r )
(ln (z )− ln(z ))
guished.
*
r m
3.3 Short-term wave conditions
The “macro-roughness” parameter zm is given by:
0.25
3.3.4 Both spectra are describing wind sea conditions 0.20
that are reasonable for the most severe seastates. However, 0.15 0.5
moderate and low sea states, not dominated by limited γ=5.0 T n=(h/g)
0.10
γ=1.0
fetch, are often composed of both wind-sea and swell. A 0.05
two peak (bi-modal) spectrum should be considered to 0.00
account for swell if considered important. Further details 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
may be found in Classification Note No. 30.5. Tn /Tp
given by:
1
ω ⋅ cosh( k ⋅ ( D + e) )
G( ω) = 0.9
sinh(k ⋅ h) 0.8 0.5
Tn =(h/g)
Where h is the water depth and k is the wave number es- 0.7
tablished by iteration from the transcendental equation: 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Tn /Tp
ω2 ⋅ h
kh = coth(k ⋅ h ) Figure 3-3 Mean zero up-crossing period of oscillating
g flow at pipe level, Tu
γ=5.0
0.5 conservative value in the range 2-8 shall be selected.
0.4
1
0.3 0.9
0.2 0.8
Reduction factor RD
0.7
0.1 0.5
Tn =(h/g) 0.6
0 s=2
0.5 s=4
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Tn/Tp 0.4 s=6
s=1000
0.3
Figure 3-4 Bandwidth parameter for flow velocity at 0.2
sin( θrel)
pipe level, ε 0.1
0
0 20 40 60 80 100 120 140 160 180
Relative pipeline-wave direction, θrel
s
π Γ1 + 3.5.2 The directional (i.e. versus θ) or omni-directional
k coss (β) β < 1 2
w (β) = w 2 ; kw = current data can be specified as follows:
π 1 s
0 else Γ + • A Histogram in terms of (Uc, θ) versus probability of
2 2 occurrence
Recommended Practice DNV-RP-F105, March 2002
Page 19
∑ (H s − µ) ⋅ PHS
3
assumed on the versatile form
HS
δ =
Tp = C T (H s )α T σ3
Where 6 ≤ CT ≤ 8 and 0.3 ≤ α T ≤ 0.5 are location spe-
cific Where PHs is the discrete occurrence probability. The same
• A long term probability density function (pdf) principle apply for current histograms.
The corresponding Return Period Values (RPV) for 1,
10 and 100 year are established from section 3.6. 3.6.4 The return period value for to be used for direc-
tional data is taken as the maximum projected flow veloc-
• Based on Return Period Values ity, i.e.
The corresponding Weibull distribution is established ( )
max x c, i ⋅ R D (θ rel , i , s ) / R D ( θ rel = 0, s )
i =1.. n
from 3.6.2 using 3 equations (xc for 1, 10 and 100
year) with 3 unknowns (α, β and γ). This is, in princi- where RD is a reduction factor defined by 3.4.3, θrel,i is
ple, always feasible but engineering judgement ap- the relative direction between the pipeline direction
plied in defining return period values can correspond and the flow direction for direction i. For current flow
to an unphysical Weibull pdf. s>8.0 may be applied.
1
T Hs, T p, θ =
(f )dF
4.1.4 The reduced velocity, VR, is in the general case RM, CF
∞
⋅ S CF
m
with combined current and wave induced flow, defined as:
∫
v
Uc
0 a
U + Uw
VR = c where
f0D
SCF Cross-flow stress range defined in 4.4
where
fv Vibration frequency; see 4.2.3.
f0 Natural frequency for a given vibration mode a Fatigue constant, depending on the relevant
Uc Mean current velocity normal to the pipe; see stress range, see 2.4.3
section 3.4. m Fatigue exponent, depending on the relevant
Uw Significant wave induced flow velocity; see stress range, see 2.4.3
section 3.4.
D Outer pipe diameter
The integral ∫ (...) dFu c indicates integration over the long-
term dis tribution for the current velocity represented by a
Weibull distribution, or histogram.
Recommended Practice DNV-RP-F105, March 2002
Page 21
4.2.2 For the in-line direction, the marginal fatigue 4.3.4 (A y /D) is defined as the maximum in-line VIV
capacity against VIV in a single sea-state characterised by response amplitude (normalised with D) as a function of
(Hs , Tp , θ) is taken as: VR and KS, see Figure 4-1 for illustration. The corre -
sponding standard deviation may be obtained as (A Y
1 /D)/√2.
, Tp ,θ =
RM , IL
THs m
S A
0.20
f ⋅ max S IL; CF IL
2 ACF
0.18
∞ v Riθ,1 = 1.0
Ksd =0.00
A IL Stress due to unit diameter in-line mode shape 0.06 Ksd =1.00
AY
4.4 Cross-flow Response Model
D
IL A Y ,1
VR 1, ;
D
4.4.1 Cross-flow VIV are affected by several parame-
IL A Y,2
Inline VIV Amplitude
V R ,2 ;
D
ters, such as the reduced velocity VR, the Keulegan-
Carpenter number, KC, the current flow velocity ratio, α,
the stability parameter, KS, the seabed gap ratio, (e/D), the
Strouhal number, St and the pipe roughness, (k/D), among
others. Note that Reynolds number, Re, is not explicit in
(V ) the model.
)
IL
;0
R ,onset
VRIL,end ;0
Reduced Velocity
4.4.2 For steady current dominated flow situations,
onset of cross-flow VIV of significant amplitude occurs
Figure 4 -2 Response Model generation principle. typically at a value of VR between 3.0 and 5.0, whereas
maximum vibration levels occurs at a value of VR between
5 and 7. For pipes with low specific mass, wave dominated
4.3.7 The reductions RIθ,1(Ic,θrel) and RIθ,2 (Ic) accounts flow situations or span scenarios with a low gap ratio,
for the effect of the turbulence intensity and angle of at- cross-flow vibration may be initiated for VR between 2 and
tack (in radians) for the flow, see Figure 4-3. 3.
π
R Iθ,1 = 1 − π 2 − 2 ⋅θ rel ( I c − 0.03 )
4.4.3 The cross-flow VIV induced stress range SCF due
0 ≤ R Iθ,1 ≤ 1 to a combined current and wave flow is assessed using the
2
(I − 0.03) following Response Model:
R Iθ, 2 = 1. 0 − c 0 ≤ R Iθ, 2 ≤ 1
0. 17
S CF = 2 ⋅ A CF ⋅ ( A Z / D) ⋅ R k ⋅ γ s
1
Where
0.9
Ri θ, 1 θrel=60ο
0.8 A CF
Unit stress amplitude (stress due to unit diame -
0.7 ter cross-flow mode shape deflection);
0.6 Rk Amplitude reduction factor due to damp-
0.5 ο
ing
Ri θ, 1 θrel=45
0.4
γs Safety factor to be multiplied on the stress
ο
range
0.3 Ri θ,1 θrel =30
The cross-flow VIV amplitude (A Z/D) in combined current
0.2
R iθ ,2 and wave flow conditions may be taken from Figure 4-4.
0.1 ο
R iθ ,1 θrel=0 all angles
The figure provides characteristic maximum values. The
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
corresponding standard deviation may be obtained as (A Z
Turbulence Intensity, Ic /D)/√2.
1.5
Figure 4-3 Reduction function wrt turbulence intensity 1.4 α > 0.8 ; all KC ψproxi,onset = 1.0
1.3 ψ trench,onset = 1.0
Cross-Flow VIVAmplitude (AZ/D)
0.4
4.4.4 The amplitude response (A Z/D) as a function of α 4.4.8 ψα,onset is a correction factor accounting for the
and KC can be constructed from: current wave ratio:
3 ⋅ ψproxi,onset ⋅ ψmass,onset ⋅ ψα ,onset ⋅ ψtrench,onset α
VRCF,onset = 1+ for α < 0.5
γon ψ α, onset = 3
VRCF,1 =5 1.167 else
9 A
VRCF,2 = VRCF, end − ⋅ Z,1
1.3 D
4.4.9 ψtrench,onset is a correction factor accounting for the
VRCF,end = 16 effect of a pipe located in/over a trench:
1.3 α f 0.8 all KC
A Z,1 0.7 KC p 10 ∆
= ψ trench,onset = 1 + 0.5
D 0.7 + 0.01 ⋅ (KC − 10 ) α ≤ 0.8 10 ≤ KC ≤ 30 D
0.9 KC f 30 where ∆/D denotes a relative trench depth given by:
A Z, 2 A
= Z,1 ∆ 1. 25d − e
D D =
D D
∆
where 0 ≤ ≤ 1
D
C F A Z,1 C F A Z ,2
V ; V ; The trench depth d is to be taken at a width equal to 3
R ,1 D R , 2 D
outer diameters. ∆/D = 0 corresponds to a flat seabed or a
Cross-Flow VIVAmplitude
pipe located in excess of D/4 above the trench, i.e. the pipe
is not affected by the presence of the trench, see Figure
4-6. The restriction ∆/D < 1.0 is applied in order to limit
the relative trench depth
3D
(V CF
R ,onset ;0.1 )
VRCF,end ;0 )
(2.0;0.0 ) D
Reduced Velocity
The bandwidth parameter ε is defined by: In lieu of more detailed data, λmax may be taken as:
A
M 22 λ max = IL λ1
ε = 1− D
M 0M4
where A IL is given by 6.8.4.
The process (spectrum) is narrow-banded for ε → 0 and 5.2.8 The linearisation constant b is given by:
broad banded for ε → 1 (in practice the process may be
considered broad-banded for ε larger than 0.6). Uc
b = 2.11 ⋅ σ u ⋅ g c ( )
σu
5.2.6 The n th response spectral moment is given by:
∞ where Uc is the mean current and σu =Uw /2 is the standard
M n = ∫ ω n S SS ( ω) dω deviation of the wave induced flow velocity. g c(•) is a cor-
0 rection function accounting for the effect of a steady cur-
rent given by:
Where SSS(ω) is the one-sided stress response spectral
density function given by: 1
g c ( x) = 2π ϕ(x ) + x ⋅ Φ(x )− ;
( )
SSS (ω) = R 2D ⋅ b 2 g 2D + ω2 g 2I ⋅ G 2 (ω) ⋅S ηη (ω) ×
1
− x2
2
ϕ(x ) =
1 2
λ2max 2π
e
m 2e (
(ω20 − ω ) + (2ζ T ω0 ω)
2 2 2
) Φ ( x) = ∫
x
−∞
ϕ( x) dx
5.2.7 λmax is an equivalent stress factor given by: In such cases, the short term fatigue capacity against direct
wave actions in a single sea-state characterised by (Hs , Tp ,
(D s − t )E ∂ 2 φ1 θ) may be estimated as follows: (cf. 2.4.5)
λ max = (1 + CSF ) λ 1 max
L ∂x
2 2
THFM
S , TP ,θ
= a ⋅ S− m Tu
φ1 (x)) 1st mode shape
E Young’s modulus where S is the quasi-static stress range response from a
CSF Concrete stiffness factor; see 6.3.5 direct regular wave load at pipe level using Morison’s
Ds Outer steel pipe diameter equation. Tu is the mean zero upcrossing period in 3.3.6.
t Pipe wall thickness
L Length of mode shape 5.4 Force Coefficients
λ1 Mode shape weighting factor given by:
L
5.4.1 The force P(x,t) per unit length of a pipe free span
∫ φ1 ( x)dx is represented by the Morison’s equation. Assuming that
λ1 = 0
L the velocity of the structure is not negligible compared
∫ φ1 ( x) dx
2
with the water particle velocity Morison’s equation reads:
0
π
λ1 is typically in the order of 1.3 P ( x, t) = g D (U − y& ) U − y& + g I U& − C a ρ D 2 &y&
4
Recommended Practice DNV-RP-F105, March 2002
Page 27
2.0
1.8
C M = C M , 0 ⋅ ψ k ,CM ⋅ ψ proxi ,CM ⋅ ψ trench,CM
1.6
Drag Coefficient CD
1.4 5.4.11 CM,0 is the basic inertia coefficient for a free con-
1.2
α crete coated pipe taken as, see Figure 5-2:
1.0 0.0
5 ⋅ (2 − f (α ) )
0.1
0.2
0.8
C M, 0 = f ( α) +
0.3
ψ k, CD = 1 .0
(KC + 5)
0.4
0.6
ψ proxi, CD = 1.0 0.5
2.0
α 5.4.12 ψ k ,CM is a correction factor accounting for the
1.8
0.0
1.6 0.1 pipe roughness.
Inertia Coefficient CM
The curves are, in strict terms, only valid for the vertical
6 Structural Analysis (cross-flow) dynamic response but may also be used for
assessment of the horizontal (in-line) response.
6.1.1 The following tasks are normally required for 6.3.1 The structural behaviour of the pipeline shall be
assessment of free spans: evaluated by modelling the pipeline, seabed and relevant
• structural modelling artificial supports and performing static and dynamic
• modelling of pipe-soil interaction analyses. This section presents requirements for the struc-
• load modelling tural modelling.
• a static analysis to obtain the static configuration of
the pipeline Soil-pipe interactions are treated in section 7.
• an eigenvalue analysis which provides natural fre-
quencies and corresponding modal shapes for the in- 6.3.2 A realistic characterisation of the cross-sectional
line and cross-flow vibrations of the free spans behaviour of a pipeline can be based on the following as-
• a response analysis using a response model or a force sumptions:
model in order to obtain the stress ranges from envi- • the pipe cross-sections remain circular and plane;
ronmental actions. • the stresses may be assumed constant across the pipe-
wall thickness;
6.2 Morphological classification • load effect calculation is normally to be performed
using nominal un-corroded cross section values;
6.2.1 The objective of the morphological classification • the application of this document is limited to elastic
is to define whether the free span is isolated or interacting. response, hence plasticity models and effects of two-
The morphological classification determines the degree of dimensional state of stress (axial and hoop) on bend-
complexity required of the free span analysis: ing stiffness need not be considered.
• two or more consecutive free spans are considered to
be isolated (i.e. single span) if the static and dynamic 6.3.3 The effect of coating is generally limited to in-
behaviour are unaffected by neighbouring spans; creasing submerged weight, drag forces, added mass or
• a sequence of free spans is interacting (i.e. multi- buoyancy. The positive effect on the stiffness and strength,
spanning) if the static and dynamic behaviour is af- see 6.3.5, is normally to be disregarded. If the contribution
fected by the presence of neighbouring spans. If the of the coating to the structural response is considered sig-
free span is interacting, more than one span must be nificant, appropriate models shall be used.
included in the pipe/seabed model.
6.3.4 Non-homogeneity of the bending stiffness along
6.2.2 The morphological classification should in gen- the pipe, due to discontinuities of the coating across field
eral be determined based on detailed static and dynamic joints or other effects, may imply strain concentrations that
analyses. The classification may be useful for evaluation shall be taken into account.
of scour induced free spans and in application of approxi-
mate response quantities. 6.3.5 The stiffening effect of concrete coating may be
accounted for by:
In lieu of detailed data, Figure 6-1 may be used to classify 0. 75
EI
the spans into isolated or interacting dependent on the soil CSF = k c conc
types and span and support lengths. EI steel
1.1
Interacting sand
where CSF denotes the stiffness of concrete coating rela-
a/L
1.0 tive to the steel pipe stiffness and (1+CSF) is the stress
Relative length of adjaent span, L
length is 0.5-1.0m and the concrete coating thickness does 6.3.12 ULS conditions may require a more refined pipe-
not exceed 150mm. soil modelling than the linearized eigen-value analysis due
to sliding at the span supports.
6.3.6 In lieu of detailed data, it is conservative to as-
sume that a girth weld is present in the most heavily
6.4 Functional Loads
loaded cross-section. This is also a basis for the concrete
stiffening effect given above. 6.4.1 The functional loads which shall be considered
are:
6.3.7 The cross-sectional bending stiffness of the con- • weight of the pipe and internal fluid;
crete coating, EIconc, is the initial, uncracked stiffness. • external and internal fluid pressure;
Young’s modulus for concrete may be taken as: • thermal expansion and contraction, and
• residual installation forces.
E conc = 10000 ⋅ fcn 0.3
6.4.2 Response calculations must account for the rele -
where fcn is the construction strength of the concrete. Both vant sequence of load application, if important.
Econc and fcn are to be in N/mm2 .
6.5 Static analysis
6.3.8 The boundary conditions applied at the ends of
the pipeline section modelled shall be able to simulate the 6.5.1 The static configuration is to be determined for
pipe-soil interaction and the continuity of the pipeline. the following conditions if relevant:
Sufficient lengths of the pipeline at both sides of the span • as-laid condition;
must be included in the model to account for the effects of • flooded condition;
side spans, if relevant. • pressure test condition, and
• operating condition.
6.3.9 The element length to be used in a finite element
model is dictated by the accuracy required. If the stress 6.5.2 The static analysis should normally account for
ranges are to be derived from the mode shapes, see 6.8.4, non-linear effects such as:
the accuracy of the stress ranges becomes strongly affected • large displacements (geometric non-linearity);
by the element length, especially at the span shoulders. • soil non-linear response, and
• loading sequence.
Ideally the maximum element length should be found by
reducing the length until the results (natural frequencies 6.5.3 The stiffness of the pipeline consists of material
and stresses) converge towards constant values. In practice stiffness and geometrical stiffness. The geometrical stiff-
this may be difficult to perform and, as guidance, the ele- ness is governed by the effective axial force, Seff. This
ment length should be in the order of the outer diameter of force is equal to the true steel wall axial force, Ntr, with
the pipeline (1D). However, higher order modes and/or corrections for the effect of external and internal pressures:
short spans (L/Ds < 30) may require shorter elements.
S eff = N tr − p i A i + p e A e
6.3.10 Structural damping is due to internal friction
forces of the pipe material and depends on the strain level Where
and associated deflections. If no information is available, a Ntr “True” steel wall axial force
structural modal damping ratio of pi Internal pressure
pe External pressure
ζstr = 0.005 Ai Internal pipe (bore) cross section area
can be assumed. If concrete coating is present, the sliding Ae External pipe (incl. Coating) cross section area
at the interface between concrete and corrosion coating
may further increase the damping to typically 0.01-0.02. The effective axial force in a span is difficult to estimate
due to uncertainties in operational temperature and pres-
sure, residual lay tension and axial force relaxation by
6.3.11 It is recommended to verify the finite element sagging, axial sliding (feed-in), lateral buckling, multi-
modelling and the post-processing by comparing the re- spanning and significant seabed unevenness. All these
sults from the finite element analysis with the approximate effects should be considered and taken into account if
response quantities of section 6.8 for a single span with relevant. Then the most reliable method to estimate the
zero effective axial force and L/Ds = 60. The in-line and effective axial force is use of non-linear FE analysis.
cross-flow natural frequencies and stress ranges shall show
similar values within ±5%. As boundary values, the effective axial force for a
completely unrestrained (axially) pipe becomes:
S eff = 0
Recommended Practice DNV-RP-F105, March 2002
Page 31
while for a totally restrained pipe the following effective approximate response quantities may be applied under
axial force apply (if pipe considered thin-walled): certain limitations, see section 6.8.
S eff = H eff − ∆ pi Ai (1 − 2ν) − As E∆Tα e 6.6.2 Using an FE-approach, the following comments
apply:
Where • the eigenvalue analysis shall account for the static
Heff Effective lay tension equilibrium configuration;
∆p i Internal pressure difference relative to laying, • in the eigenvalue analysis, a consistent linearisation of
see DNV-OS-F101 the problem must be made;
As Pipe steel cross section area • the pipe-soil linearisation should be validated;
∆T Temperature difference relative to laying • the effect of geometric non-linearity on the dynamic
αe Temperature expansion coefficient, may be response should be assessed;
temperature dependent • the span support may be assumed invariant during
Vortex-Induced Vibrations (VIV) and motions from
direct wave loading.
6.5.4 In this document, the static environmental loads
are confined to those from near bottom current. If the load 6.6.3 For analysis of a pipeline stretch with several
is much smaller than the vertical functional loads, then it spans and especially with interacting spans, special care
may be disregarded in the analysis. However, for light must be paid to the determination of the eigenvalues and
pipes or long span lengths it should be considered if rele- associated eigenvectors. This is due to the potential occur-
vant. rence of very close eigenvalues, especially with respect to
the identification of correct eigenvectors.
6.5.5 Load history effects such as the lay tension and
submerged weight during installation will influence the 6.7 Added Mass
static deflection and stresses which are mainly determined
by the submerged weight and effective axial force in the 6.7.1 The added mass may be considered as
phase considered.
( )
C a ( e / D) = C M α = 0; KC ↓ 0; e / D − 1
Furthermore, the span geometry such as inclination of the
span shoulders will have a significant influence on the The ↓ symbolise KC approaching 0. Note that the effect of
static stresses and deflection. For this reason, the static pipe roughness and trench is not accounted for.
response should be based on survey results (measured de- According to section 5.4.10, Ca becomes:
flections) and/or FE analysis if considered as critical for
the span assessment.
1. 6
0.68 + for e/D < 0.8
Ca = (1 + 5 ⋅ ( e / D) )
6.5.6 In addition to the static penetration due to the 1 for e/D ≥ 0.8
submerged weight of the pipeline, the penetration may
increase due to effects from laying, erosional processes
and self-burial. where e/D is the span gap ratio. This expression applies for
both smooth and rough pipe surfaces.
6.6 Eigen-value analysis
6.8 Approximate response quantities
6.6.1 The aim of eigen-value analyses is to calculate
the natural frequencies and corresponding stress due to 6.8.1 The approximate response quantities specified in
associated mode shapes. The analysis is normally complex this section may be applied for free span assessment pro-
and depends on: vided:
• the temporal classification (scour or uneveness in- • Conservative assumptions are applied with respect to
duced free span); span lengths, soil stiffness and effective axial force.;
• the morphological classification (single or multispan); • The span is a single span on a relatively flat seabed,
• the pipeline condition (i.e. as-laid, water-filled, pres- i.e. the span shoulders are almost horizontal and at the
sure test and operation); same level;
• the pipe and soil properties; • The symmetrical mode shape dominates the dynamic
• the effective axial force and the initial deflected shape response (normally relevant for the vertical, cross-
after laying; flow response only). Here the following limits apply:
• the loading history and axial displacement (“feed-in”) L/Ds < 140
of the pipe. δ/D < 2.5
In general, it is recommended that the response quantities Note that these are not absolute limits; the shift in
be assessed using non-linear FE-analyses conducted over cross-flow response from the symmetrical to the un-
an appropriate stretch of the pipeline. However, symmetrical mode will depend on the sagging and the
6.8.3 The effective mass, me, is defined by 6.8.7 In case the static deflection is not given by direct
measurement (survey) or estimated by accurate analytical
∫ m (s) φ2 (s )ds
tools, it may be estimated by:
me = L
∫ φ ( s) ds
2
q ⋅ L4eff 1
L δ = C6
EI ⋅ (1 + CSF ) S
where φ(s) is the assumed mode shape satisfying the 1 + C 2 ⋅ eff
boundary conditions and m(s) is the mass per unit length PE
including structural mass, added mass and mass of internal where C6 is a boundary condition coefficient.
fluid.
Note that Leff shall be calculated using the static soil stiff-
6.8.4 The unit diameter stress amplitude (stress due to ness in the Leff/L calculation.
unit outer diameter mode shape deflection) may be calcu-
lated by: Note that, due to historical effects and the local seabed
A IL / CF = D
κEI D s = 1 D E D κ geometry, there is a large uncertainty associated with this
s
I 2 2 simplified expression, see 6.5.5.
where Ds is the steel pipe diameter and D is the outer pipe
6.8.8 The coefficients C1 to C6 are given in the table
diameter (including any coating). κ is the curvature of the below for different boundary conditions. For multi-
assumed mode shape satisfying the boundary conditions
spanning scenarios the choice of coefficient should be
mode shape:
supported by dedicated FE-analyses.
Recommended Practice DNV-RP-F105, March 2002
Page 33
7.2.5 For sands with low content of fines, the frictional Note that the sand type is identified by the value of the
component of the axial and lateral resistance is propor- friction angle ϕs (Table 7-1), and the clay type is identified
tional with the vertical force at any time. For clays, the by the value of the undrained shear strength s u (Table 7-2).
frictional component is proportional with the undrained
shear strength. For pipes supported by rock, values for the modal soil
damping ratios may be taken as for dense sand.
7.2.6 Where linear soil stiffness has to be defined for
the eigenvalue analysis, the soil stiffness should be s e- Table 7-4 Modal soil damping ratios (in %) for clay.
lected considering the actual soil resistance and the ampli-
Horizontal (in- Vertical (cross-
tude of the oscillations.
Clay type line) direction flow) direction
7.2.7 The soil stiffness for vertical loading should be L/D L/D
evaluated differently for static and dynamic analyses. The <40 100 >160 <40 100 >160
static soil response will be governed mainly by the maxi- Very soft - Soft 3.0 2.0 1.0 3.0 2.0 1.0
mum reaction, including some cyclic effects. Dynamic Firm – Stiff 2.0 1.4 0.8 1.2 1.0 0.8
stiffness will be characterised mainly by the unloading/re-
loading situation. Very stiff - Hard 1.4 1.0 0.6 0.7 0.6 0.5
Alternatively, the modal soil damping ratio, ζsoil , may be 7.3.2 The bearing capacity factors Nc, Nq and Nγ versus
taken from Table 7-3 or Table 7-4, in which L denotes the
the internal friction angle ϕs may be calculated from the
length of the free span and D is the outer diameter of the following formulas:
pipeline. Interpolation is allowed.
ϕ
N q = exp( π tan ϕ s ) tan 2 ( 45 + s )
Table 7-3 Modal soil damping ratios (in %) for sand. 2
N c = ( N q − 1) cot ϕs
Horizontal (in-line) Vertical (cross-flow)
Sand direction direction N γ = 1.5( N q − 1) tan ϕs
type L/D L/D
<40 100 >160 <40 100 >160
For clayey soils the friction angle is set equal to 0°, i.e. Nq
Loose 3.0 2.0 1.0 2.0 1.4 0.8 = 1.0 and Nc = 5.14.
Medium 1.5 1.5 1.5 1.2 1.0 0.8
Dense 1.5 1.5 1.5 1.2 1.0 0.8
Figure 7-1 Bearing capacity factors Nc, Nq and Nγ ver- 7.3.8 For a detailed determination of KL , the following
sus the internal friction angle ϕs expression may be applied:
K L = 0.76 ⋅ G ⋅ (1 + ν )
7.3.3 The maximum static, axial soil reaction per unit
length may be taken as: which is based on elastic halfspace theory for a rectangular
foundation under assumption of a pipe length that equals
Ra = Rv µa − sandy soils 10 times the contact width between pipe and soil. Pois-
Ra = min{Rv µa, bτmax} − clayey soils son’s ratio ν is given in Table 7-1 and Table 7-2.
Where
7.3.9 For conditions with small-amplitude deforma -
Rv Vertical static soil reaction given by 7.3.1 tions, the shear modulus for the soil can be taken as:
µa Axial friction coefficient
B Given in 7.3.1 2000 ⋅ (3 − es )2
σs
OCR i 1. 3i p
1 + es
for sand
[kN/m2 ]
= 1 − p + ip in % G=
2.61 200 200
kc
1300 ⋅ (3 − e s ) σ (OCR )k s
2
for clay
s
Soil shear strength: 1 + es
2
0. 5(1 − k c ) Rv where
τmax = s u2 −
b σs Effective mean stress (in units of kPa), see
7.3.10.
OCR Over-consolidation ratio
es Void ratio
7.3.4 The static vertical stiffness is a secant stiffness
ks Coefficient, taken from Figure 7-2
representative for penetration conditions such as during
installation and erosion and during development of free
spans.
0.5
The static vertical stiffness KV,S is defined as KV,S=RV/v,
where RV is the static vertical soil reaction per unit length 0.4
of pipe and v is the vertical penetration of the pipe re-
quired to mobilise this reaction. Unless effects of pipelay 0.3
ks
KV=∆FV/∆δV, where ∆FV is the incremental vertical force Figure 7-2 k s versus plasticity index, i p
between pipe and soil per unit length of pipe, and ∆δV is
Recommended Practice DNV-RP-F105, March 2002
Page 37
This expression for G, which gives lower-bound values for 7.3.11 The procedure in Sections 7.3.7-7.3.10 leads to
the initial shear modulus, has been calibrated to measure- values of the dynamic stiffness KV and KL , which can be
ments from free pipeline spans. It is based on an expres- considered as lower-bound values for initial small-strain
sion for the initial shear modulus, formulated by Hardin stiffness at either end of the free span, but which are not
and Drnevich (1970) and based on experimental results for adjusted for possible non-linear soil behaviour at larger
a broad range of soil types. strains.
penetration of 20-40% of the diameter and Lsh /L as given Table 7-9 Equivalent stiffness, k 2 ,
in Table 7-7 and Table 7-8. for span supports on sand
v/D k2 (kN/m/m)
The simplified expressions for KV and KL , with values of 0.00 0
CV and CL given in Table 7-7 and Table 7-8, are valid for
0.25 19
dynamic conditions with small-amplitude deformations.
0.35 28
0.50 44
7.3.14 For free spans supported by sand, the lateral dy-
1.00 105
namic stiffness KL should be calculated under an assump-
tion of loose sand properties in order to properly account
for effects of erosion. For span supports on clay, the equivalent stiffness k2 can
be calculated as
−0 .4
s
1 .3
7.3.15 For extreme conditions, which can be assumed to v
cause large deformations on the shoulders, a smaller spring k 2 = 8.26 ⋅ s u ⋅ u ⋅
stiffness than that associated with small-strain conditions Dγ soil D
apply. For vertical loading, this smaller spring stiffness is in which
typically of the same order of magnitude as the static ver- γsoil Total unit weight of soil (=γsoil ’+γwater)
tical stiffness KV,S. For lateral loading, reference is made γsoil ’ Submerged unit weight of water
to Section 7.3.16. Unit weight of water (=10 kN/m3 )
γwater
v Vertical pipe penetration at span shoulder
7.3.16 In case of lateral loads large enough to cause slid-
ing at the span supports, the following bilinear force-
For span supports on sand, the maximum lateral resistance
displacement curve, which covers both small-strain condi-
per unit length of pipe is
tions and extreme displacement conditions, may be applied 1. 25
at span supports on sand and clay: v
FL,max = µ LFV + 5. 0 ⋅ γ soil'⋅D2 ⋅
D
k1 ⋅ y for FL < µ LFV
µ F For span supports on clay, the maximum lateral resistance
FL =
µ F + k2 ⋅ ( y − L V ) for µLFV ≤ FL < FL, max per unit length of pipe is
L V k1 − 0. 4
s
1 .3
v
FL,max = µ LFV + 4.13 ⋅ D ⋅ s u ⋅ u ⋅
in which Dγ soil D
y Lateral displacement of pipe on shoulder
FL Lateral force per unit length of pipe at dis- 7.3.17 The axial dynamic soil stiffness is usually not
placement y important. However, when long free spans are considered,
FL,max Maximum lateral resistance per unit length it is important to include an axial soil-support model with
of pipe friction and stiffness. If no information is available about
FV Vertical contact force per unit length of the axial dynamic soil stiffness, it may be taken as equal to
pipe on shoulder the lateral dynamic soil stiffness KL as described above.
µL Lateral friction coefficient
k1 Initial stiffness up to mobilisation of full 7.4 Artificial supports
friction
k2 Equivalent stiffness for deformations past
mobilisation of full friction 7.4.1 Gravel sleepers can be modelled by modifying
the seabed profile, considering the rock dump support
Unless data indicate otherwise, µL =0.6 may be applied for shape and applying appropriate stiffness and damping
characteristics.
span supports on sand and µL =0.2 for span supports on
clay.
7.4.2 The purpose of mechanical supports is generally
For supports on sand, the initial stiffness k1 may be taken to impose locally a pipeline configuration in the vertical
and/or transverse directions. Such supports can be mo d-
as equal to the lateral dynamic stiffness KL for loose sand.
For supports on clay, the initial stiffness k1 may be taken elled by concentrated springs having a defined stiffness,
as equal to the lateral dynamic stiffness KL for the clay taking into account the soil deformation beneath the sup-
port and disregarding the damping effect.
type in question.