RP C103 PDF
RP C103 PDF
RP C103 PDF
DNV-RP-C103
COLUMN-STABILISED UNITS
FEBRUARY 2005
Det Norske Veritas. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including pho-
tocopying 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-C103, February 2005
Changes Page 3
Main changes The text relating to 'tank pressures' (sec.3.8) has been
amended. Formulations have been simplified and clari-
Text and definitions have generally been co-ordinated fied.
with the 2004 revisions of the relevant structural standards The text relating to support of mooring equipment
(e.g. references, terminology, definitions, lay-out of text, (sec.6.1) has been updated and clarified.
etc.)
dfg
CONTENTS
Table 1-1 indicates which limit states are usually considered in The applied environmental conditions should be stated in the
the various design conditions. design basis/design brief.
Table 1-1 Design conditions and limit states Typical environmental loads to be considered in the structural
design of a column-stabilised unit are:
Instal- Operat- Surviv- Transit Acci- Dam-
lation ing al dental aged
wave loads, including variable pressure, inertia, wave
ULS x x x x "run-up", and slamming loads
FLS (x) x (x) (x) wind loads
ALS x x current loads
Limiting design criteria for transfer from one mode of opera- snow and ice loads.
tion (draughts) to another mode of operation shall be clearly
established and documented. Different modes of operation or Due consideration should be made to site specific environmen-
phases during the life of a column-stabilised unit may be deci- tal phenomena such as hurricanes, cyclones etc.
sive for the design. For each of the relevant design conditions Design for worldwide operation shall be based on Classifica-
the design criteria shall include relevant consideration of the tion Note 30.5 with use of the scatter diagram for the North At-
following items: lantic as given in Classification Note 30.5 Table 3.3. See also
2.2 and DNV-OS-C101 Sec.3 E200, DNV-OS-C103 Sec.4
intact condition, structural response and strength (ULS, B100, DNV-OS-C103 Sec.5 B500 and DNV-OS-C103 Ap-
FLS, ALS) pendix B.
damaged condition, structural response and strength
(ALS) Design for specific region or site shall be based on specified
air gap (ULS) environmental data for the area(s) the unit shall operate, see
compartmentation and stability requirements (intact DNV-OS-C101 Sec.3 E and DNV-OS-C103 Appendix A and
(ULS) and damaged (ALS)). DNV-OS-C103 Appendix B. The parameters describing the
environmental conditions shall be based on observations from
1.4 Abbreviations or in the vicinity of the relevant location and on general knowl-
edge about the environmental conditions in the area. Classifi-
ALS Accidental Limit States cation Note 30.5 may be used as guidance for determination of
the environmental loads.
DFF Design Fatigue Factor
DNV Det Norske Veritas 2.2 Environmental conditions
FLS Fatigue Limit States
2.2.1 General
LRFD Load and Resistance Factor Design
RAO Response Amplitude Operator The most significant environmental loads for the hulls of col-
umn-stabilised units are normally those induced by waves. In
RP Recommended Practice order to establish the characteristic response, the characteris-
SCF Stress Concentration Factor tics of waves have to be described in detail. This description
ULS Ultimate Limit States may either be based on deterministic design wave methods or
on stochastic methods applying wave energy spectra.
WSD Working Stress Design
If a design condition is limited to a certain range of environ-
mental conditions, then these limitations are applied in the
evaluation of the environmental loads for that design condi-
2. Environmental Conditions and Loads tion, rather than the actual environmental data for the site or re-
gion. Care should be taken to choose the most unfavourable
2.1 Introduction combination of environmental conditions from the specified
The suitability of a column-stabilised unit is dependent on the range.
environmental conditions in the area of the intended operation. The description of waves is related to the method chosen for
A drilling unit may be intended for worldwide operation or op- the response analysis, see Chapters 3 and 4.
eration in a specific region. A production unit may be planned
to operate at a specific site. Such a site may be harsh environ- More details for wave, wind and current conditions are given
ment or benign waters. in Classification Note 30.5.
Hence the environmental conditions and environmental loads 2.2.2 Regular wave parameters
depend on the area where the unit is intended to operate. A col-
umn-stabilised unit is normally designed for one of the follow- Deterministic methods are used when the sea state is represent-
ing conditions: ed by regular waves defined by the parameters:
There is also a limitation of wave steepness. Wave steepness is waves of extreme height or extreme steepness need to be con-
defined by: sidered.
For design purposes the maximum wave height Hmax corre-
2 H sponding to the 90% percentile in the extreme value distribu-
S = -------------- tion is used:
2
gT
When H100 = 32 m, then the wave height and period combina- H max 2.12H s
tions on the steepness limit are given by:
2
0.22T for T 6s The steepness of a specific sea state is defined by:
H = T
2
2 Hs
----------------------------------------------- for T > 6s S s = ----------------
4.5 + 0.02 ( T 2 36 ) g Tz
2
where T is in seconds and H is in metres. The sea steepness need not be taken greater than the 100 year
The design wave data are represented by the maximum wave sea steepness for unrestricted service (worldwide operation),
height as well as the maximum wave steepness. The wave which normally may be taken as:
lengths are selected which are the most critical to the structure 1-
----- for T z 6s
or structure part to be investigated. 10
2.2.3 Irregular wave parameters S s = -----
1-
for T z 12s
15
Stochastic analysis methods are used when a representation of
the irregular nature of the sea is essential. A specific sea state Linear interpolation for 6s < T z < 12s
is then described by a wave energy spectrum, which is charac-
terised by the following parameters: Then the significant wave height and period combination on
the steepness limit are given by:
significant wave height, Hs
average zero-up-crossing period, Tz 0.156T 2 for T z 6s
z
The design and analyses for ULS and FLS shall, for worldwide
H s = 0.206T z 2 0.0086T z 3 for 6s <T z <12s
operation, be based the scatter diagrams for the North Atlantic
given in Classification Note 30.5 Table 3.3. 0.104T 2 for T z 12s
An appropriate type of wave spectrum should be used. z
with a homogenous wave climate. Other results may be ob- 2.2.7 Wave energy spreading function
tained for areas where the climate is characterised by long pe- In stochastic wave load analysis the effect of wave short-crest-
riods with calm weather interrupted by heavy storms of short edness may easily be included by introducing a wave energy
duration. spreading function. The Pierson-Moskowitz wave spectrum
together with a cos4 wave spreading function should be uti-
lised in the fatigue analyses (FLS) of column-stabilised units,
see DNV-OS-C103 Sec.5. is angle between direction of ele-
mentary wave trains and the main direction of the short-crested
wave system.
For the extreme wave analysis (ULS) the wave spreading func-
tion should be according to the designer's specification or
cos10, see Classification Note 30.5 3.2.
2.3 Wave loads
2.3.1 Worldwide operation
The wave loads acting on column-stabilised units intended for
operation worldwide without any restriction, should be ana-
lysed by use of a diffraction model, which takes into account
the reflection of waves. For preliminary design, Morison's
equation may be applied, together with a contingency factor,
see 2.3.3.
Figure 2-1
Wave height exceedance diagram 2.3.2 Site specific harsh environment
The wave loads acting on column-stabilised units intended for
2.2.5 Wave theory operation in a site specific harsh environment area, should be
analysed by use of a diffraction model. For preliminary design,
When the individual waves have been defined, wave particle a simplified Morison's equation may be applied, together with
motions may be calculated by use of an appropriate wave the- a contingency factor, see 2.3.3.
ory. The linear (Airy) theory is normally sufficiently accurate
for column-stabilised units. It may be applied in shallow water, 2.3.3 Benign waters
intermediate depth and deep water ranges, whereas other wave
theories are often applicable over a more restricted range of The wave loads acting on twin pontoon column-stabilised
depth. Furthermore, the linear approach has been found to be units intended for operation in benign waters and following
satisfactory for column-stabilised units even when there are normal class survey intervals, may be calculated with a simpli-
quite major departures from the small wave height assumption. fied model based on Morison's equation.
In connection with stochastic response analysis, linear (Airy) When a Morison model is utilised, a contingency factor of 1.3
theory should always be used. for ULS and 1.1 for FLS shall be applied, see also DNV-OS-
C103 Sec.4 B107, DNV-OS-C103 Sec.5 B403 and DNV-OS-
2.2.6 Long term description of the sea C103 Appendix B.
Long term statistics are associated with non-stationary proc- Design for benign waters shall be based on specific environ-
esses occurring over a period of months and years, whereas mental data for the area the unit shall operate. Due considera-
short term statistics relate to the stationary processes in periods tion shall be given to environmental loads caused by swell and
over only a few hours. In forming a long term statistical de- currents.
scription of the sea a suitable statistical model providing a joint
probability distribution of wave height and wave period is re- 2.3.4 Morison equation
quired. The Morison equation is given in Classification Note 30.5 6.1,
Long term data for wave conditions are commonly given in the including inertia and drag forces. This equation is valid for
form of a scatter diagram for significant wave height and wave slender structural elements, for example wave length/diameter
zero-up-crossing period or peak period. The North Atlantic ratio above 5.
scatter diagram, described in DNV Classification Note 30.5, It is important when calculating added mass for the pontoon
shall be applied for worldwide operation (see 2.1). Site specific that interaction effects between column and pontoon, end ef-
scatter diagrams may be used for restricted operation. FLS re- fects, and the effect of rounded cross sectional corners are tak-
sponse calculations may be directly based on the sea-states en into account.
represented in the scatter diagram, or a representative conden-
sation into a smaller number of states. ULS calculations should 2.3.5 Linearisation of the drag force
be based on a joint distribution function fitted to the wave data,
to take account of the possibility of more unfavourable waves Non-linear, hydrodynamic drag forces acting on braces should
than have been recorded in the observed data set. The three pa- be accounted for. The drag force on submerged braces may be
rameter Weibull distribution is commonly used to describe the linearised by means of recognised methods. It is normally ac-
marginal distribution of significant wave heights, and a condi- ceptable to linearise the drag force to account for the correct
tional log-normal distribution is often used to describe the force under the maximum water particle speed due to the
wave periods. waves, calculated in accordance with linear wave theory, or
conservatively taken as 5 m/s. The wave loads acting on the
The FLS response analysis should cover the range of probabil- pontoons are normally dominated by hydrodynamic inertia
ity levels from 10-1 to 10-4, for exceedance of stress ranges in forces, so the linearisation of the drag term for the pontoons is
the platform lifetime. less critical.
The ULS response analysis should cover the range from 10-1 It is advisable to check that the modelling of drag loads pro-
to 10-8, for exceedance of load effect maxima in the platform vides a damping effect that leads to realistic heave response at
lifetime. resonance.
2.3.6 Asymmetry factor in air gap calculation Hydrostatic sea pressure for local analyses and scantlings of
Generally a wave asymmetry factor of 1.2 should be applied in tanks are given in 3.8.
the air gap calculations unless model tests are available. In this 3.4 Variable functional loads (Q)
case the air gap shall be calibrated against the model tests.
Variable functional loads are described/defined in DNV-OS-
Calculations for sufficient air gap is further referred to in 6.3. C101 Sec.3 D and DNV-OS-C103 Sec.3 D.
Variations in operational mass distributions, especially in the
pontoons with maximum and minimum ballast, shall be ade-
3. Design Loads quately accounted for as part of the global load effects for the
structural design.
3.1 Introduction Tank pressures for local analyses and scantlings are given in
As described in DNV-OS-C101 and DNV-OS-C103, the fol- 3.8.
lowing load categories are relevant for column-stabilised units: 3.5 Environmental loads (E)
permanent loads (G) 3.5.1 General
variable functional loads (Q)
environmental loads (E) Environmental loads are in general terms given in DNV-OS-
accidental loads (A) C101 Sec.3 E and DNV-OS-C101 Sec.3 F and in DNV-OS-
deformation loads (D). C103 Sec.3 E.
Wave loads are given in Chapter 2.
Characteristic loads are reference values of loads to be used in Practical information regarding environmental loads is given
the determination of load effects. The characteristic load is in the Classification Note 30.5.
normally based upon a defined fractile in the upper end of the
distribution function for the load. Note that the characteristic Hydrodynamic sea pressure and vertical accelerations form an
loads may differ for the different limit states and design condi- integrated part of tank pressures and sea pressures for local
tions. analyses and scantlings of tanks as given in 3.8.
The basis for the selection of characteristic loads for the differ- 3.5.2 Current loads
ent load categories (G, Q, E, A, D), limit states (ULS, FLS,
ALS) and design conditions are given in DNV-OS-C101 Current loads may normally be calculated from the drag term
Sec.3. in the Morison equation. The variation in current profile with
water depth may be determined in accordance with Classifica-
A design load is obtained by multiplying the characteristic load tion Note 30.5.
by a load factor. A design load effect is the most unfavourable
combined load effect derived from design loads. The global response from the current loads on a column-stabi-
lised unit is negligible compared with the response from wave
3.2 Loads to be applied in global and local models loads for consideration of overall hull strength. The main con-
tribution of current is to the reaction forces of anchor lines and
Analytical models shall adequately describe the relevant prop- thrusters, see 3.5.5. Current loads need therefore normally not
erties of loads, load effects, stiffness, and displacement, and be considered for global and local structural analyses of col-
shall satisfactorily account for the local and system effects of umn-stabilised units.
time dependency, damping, and inertia.
It is normally not practical, in design analysis of column-stabi- 3.5.3 Wind loads
lised units, to include all relevant loads (both global and local) The horizontal wind force on a column-stabilised unit for a
in a single model. Generally, a single model would not contain maximum sustained wind speed will cause the unit to heel,
sufficient detail to establish local responses to the required ac- with a maximum heel angle of the unit of order 10-15. Such
curacy, or to include consideration of all relevant loads and heel angle would give a considerable sideway gravitational
combinations of loads. Assessment of single model solutions deck load component.
is further discussed in 4.8.2. In practice, however, sustained heel of the platform due to
It is often more practical, and efficient, to analyse different wind is kept less than about 3 by re-ballasting and by anchor
load effects utilising a number of appropriate models and su- forces, and it is considered too conservative to add the maxi-
perimpose the responses from one model (global) with the re- mum theoretical wind heeling effect to the 100 year wave forc-
sponses from another model (local) in order to assess the total es. Hence, in the global analysis of the platform wind forces
utilisation of the structure. may normally be neglected.
The modelling guidance given in 3.8.6, Chapter 4, and Chapter 3.5.4 Pressure height for sea pressure
5 can be considered as a proposed use of different models in
accordance with an acceptable analytical procedure. The pro- Sea pressure (static and dynamic) shall include the effect of
cedures described are not intended to restrict a designer to a wave height and relative motion of the platform.
designated methodology when an alternative methodology To be consistent with the analysis of global wave forces, the ef-
provides for an acceptable degree of accuracy, and includes all fect of reduced wave particle motion with increasing depth
relevant load effects. Further, the modelling procedures and may be included (Smith effect).
guidance provided are intended for establishing responses to Due to the requirement for positive air gap, see DNV-OS-
an acceptable level of accuracy for final design purposes. C103 Sec.4 D100, the pressure height may be calculated as:
For preliminary design, simplified models may be used in or-
der to more efficiently establish the design responses, and to
achieve a simple overview of how the structure responds to the k ( h0 a )
h = h 0 1 ----- 1 e = C h
a
design loads. w 0
h0
3.3 Permanent loads (G)
Permanent loads are described/defined in DNV-OS-C101 h0 = distance from load point to underside of lowest deck
Sec.3 C and DNV-OS-C103 Sec.3 C. a = wave amplitude
For tanks where the air pipe may be filled during filling oper- ps = g0Cw(TE zb) (kN/m2) 0
ations, the following additional internal design pressure condi- pe = g0Cw(DD zb) (kN/m2) for zb
tions shall be considered: pe = g0Cw(DD TE) (kN/m2) for zb <
TE = extreme operational draught (m) measured vertically
from the moulded baseline (B.l.) to the assigned load
2 waterline
p d = ( g 0 h op + p dyn ) f, G, Q ( kN/m )
Cw = reduction factor due to wave particle motion (Smith ef-
fect, see 3.5.4) Cw = 0.9 unless otherwise documented
DD = vertical distance (m) from the moulded baseline to the
pdyn = pressure (kN/m2 ) due to flow through pipes, minimum underside of the deck structure. (The largest relative
25 kN/m2 distance from the moulded baseline to the wave crest
may replace DD if this is proved smaller.)
This internal pressure need not to be combined with extreme zb = vertical distance (m) from the moulded base line to the
environmental loads, because they it is not likely that the tank load point
filling operation will occur together with extreme waves. Nor- ps = static sea pressure
mally only static global response need to be considered. pe = dynamic (environmental) sea pressure.
Parameters for tank pressures are illustrated in Figure 3-1. Re- The load factors are given in DNV-OS-C103 Sec.4 Table A1.
fer also to tank pressures in damaged compartments specified
Parameters for sea pressures are illustrated in Figure 3-2.
in the ALS heeled condition, see 3.8.5.
The Smith effect (Cw = 0.9) shall only be applied for loading
The load points for plate fields, for stiffeners, and for girders conditions including extreme wave conditions, i.e. ULS a) and
are defined in DNV-OS-C103 Sec.3 B100. b) loading conditions.
Relevant combinations of tank and sea pressures, with combi-
nations of maximum/minimum pressures, are specified and
pdyn discussed in 3.8.6 and 5.3.
Top of airpipe
3.8.4 Sea pressure at wave trough, ULS
The equations for pe and ps in 3.8.3 correspond to the maxi-
mum sea pressure (wave crest elevation).
In combination with the maximum tank pressures, the external
sea pressure up to the lowest waterline at wave trough may be
h op applied in the design for the external plate field boundaries.
Such external sea pressure may be taken up to half the pontoon
pdyn height, see Figure 3-2. The design sea pressure up to the lowest
Maximum waterline is regarded as permanent load and may be taken as:
filling height Waterline
hp
p d = g 0 ----- z b f, G, Q
2
(kN/m ) for z b < h p 2
2
h op
B.l.
Figure 3-1
Parameters for tank pressures
3.8.5 Sea pressures, ALS heeled condition 3.8.6 Combination of tank and sea pressures
For ALS design with 17 heeled condition as given in DNV- Local structural models should be created in order to evaluate
OS-C103 Sec.6 F, the design sea pressure can be expressed as: responses of the structure to various sea and tank pressures.
Examples of considerations that should be evaluated in con-
2 nection with the load cases of local pressures acting on pon-
p d, ALS = g 0 h 17 f, A (kN/m ) toon and column sections of a column-stabilised unit are given
below:
h17 = vertical distance (m) from the load point to the damaged The intention of the local model is to simulate the local
heeled condition still water line after accidental flood- structural response for the most unfavourable combination
of relevant local loads. Relevant combinations of internal
ing (maximum heel 17o, effect of submersion included, (tank) and external (sea) pressures for tanks should be con-
see also 5.2.7 item 2 and DNV-OS-C301 Ch.2 Sec.1 sidered for both the intact and damage load conditions.
E400). If cross-section arrangements change along the length of
f,A= load factor in the ALS is 1.0, see DNV-OS-C101 Sec.2 the structure, several local models may be required in or-
D700. der to fully evaluate local response at all relevant sections.
The sea pressure in heeled condition is illustrated in Figure For tanks separated by internal watertight bulkhead/deck,
3-3. For ALS design, the Smith effect Cw = 1.0. Note that the the internal tank pressure should normally not be consid-
sea pressure for ALS heeled condition shall also be applied as ered to act simultaneously on both sides of the bulkhead/
internal tank pressure on bulkheads and decks surrounding the deck. Combinations with maximum tank pressure from
damaged compartment(s). each of the tanks and zero tank pressure from the neigh-
bouring tank should be considered. Effects of sea pressure
Heeling of the unit after damage flooding shall be accounted response on the internal watertight bulkhead/deck should
for in the assessment of structural strength. The unit shall be be assessed and included when relevant.
designed for environmental condition corresponding to 1 year For some structural elements, for example access/pipe tun-
return period after damage. To simplify the design approach, nel in the pontoon and some girder configurations, the ef-
the environmental loads may be disregarded if the material fac- fect of simultaneous tank pressure in neighbouring tanks
tor is taken as M =1.33, see DNV-OS-C103 Sec.6 F. may be governing for the design. Effects of simultaneous
sea pressure response should be assessed and included
when relevant.
For external structural components (stiffened plates adja-
cent to sea), the maximum external (sea) pressure and the
maximum internal (tank) pressure will normally not act si-
multaneously.
Loads are usually applied in the analysis models at the
girder level and not at the individual stiffener level (typical
global and in some cases local analysis models). In such
cases the local stiffener bending is not included in the
model responses. The stiffener bending response will then
be explicitly included in the buckling code check as lateral
pressure (for plate induced and/or stiffener induced buck-
ling).
For transversely stiffened structures (i.e. girders orientated
in the transverse direction) the local responses extracted
from the local model are normally responses in the struc-
tural transverse direction (see y in Figure 5-5). Shear and
bending responses in girders shall also be considered.
For structural arrangements with continuous, longitudinal
girder arrangements, a longitudinal response will also be
of interest (see x in Figure 5-5).
For structural transverse sections without continuous lon-
gitudinal girder elements, two-dimensional structural
models may be considered as being adequate.
Figure 3-3 For space frame beam models, relevant consideration shall
Sea pressures in heeled condition be given to shear lag effects.
Table 3-1 Design pressure combinations for tank pressure and sea pressure
External or internal
Limit state Desgn pressure combination
structural component
Refer DNV-OS-C401 Ch.2 Sec.4, Testing of Watertightness and Structural Tests.
Note DNV-OS-C103 Sec.3 D307: "In cases where the maximum filling height is less
Tank testing than the height to the top of the air pipe, it shall be ensured that the tank will not be
over-pressured during operation and tank testing conditions".
Imax: pd and E = 0 1)
Internal Emax: pd,ULS and I = 0 2)
ULS
both a) and b) Imax: pd and E = 0 3)
External Emax: pd,ULS and I = 0
e) The maximum response corresponding to the 90% fractile g) Further, by taking into account the probability of different
in the extreme value distribution (short term response): wave spectra, the long term response distribution of the re-
sponse in question to be established, see Figure 4-1 (5-6).
1 The slope of this long term distribution is an important pa-
----
N rameter for the fatigue analysis, see Figure 4-1 (6).
Resp ( max ) = R ln N = s 0.5 ln 1 p = 2.12 s
Scatter diagram for the North Atlantic, applicable for
worldwide operation, is given in Classification Note 30.5
N = number of waves (1080 in a 3 hour storm) Table 3.3.
p = fractile level, (1 p) = probability of exceedance
f) By taking into account the probability for different wave
headings, the maximum response amplitude, 90% percen-
tile, for a specified sea state may be derived.
Figure 4-1
Stochastic stress analysis procedure
The following partial factors shall be applied: The Weibull distribution parameter h for simplified fa-
tigue analysis should have a value h = 1.1 in combination
When Morison model is utilised, a contingency factor of with worldwide operation criteria. Alternatively the
1.2 shall be applied, see DNV-OS-C103 Appendix B, in Weibull distribution parameter h may be calculated based
addition to the relevant load factors specified in DNV-OS- on site specific criteria.
C103 Sec.4 A.
See DNV-OS-C103 Sec.5 and DNV-OS-C103 Appendix
B.
MAX RESPONSE
PARAMETER
(90% FRACTILE)
Figure 4-2
Design wave analysis procedure
4.6 Characteristic global hydrodynamic responses stochastic approach as described in 4.3, see Figure 4-4.
4.6.1 General A typical example of a transfer function for this response is
shown in Figure 4-5.
The following responses will normally be governing for the
global strength of the unit. The responses are illustrated on a This response will normally give the maximum axial force in
typical twin pontoon unit with horizontal braces between the the transverse horizontal braces of a twin pontoon unit. For a
pontoons, and diagonal braces between deck and pontoon, see unit without these braces, this response will give maximum
Figure 4-3. bending moment for the transverse deck structure. In addition
to the axial force, which is a global response, local vertical
Similar responses are also applicable for the design of ring drag and inertia force on the bracing structure should be ac-
pontoon units, applying split force (FS), torsion moment (Mt) counted for.
and shear force (FL) related to both transverse and longitudinal
axes. This to account for responses in both transverse and lon- For a ring pontoon unit, this response will give axial force and
gitudinal pontoons of a ring pontoon unit. bending moment in the pontoons (about pontoon vertical and
transverse axes), with maximum responses both at pontoon
1) Split force between pontoons, FS mid-section and at pontoon end.
2) Torsion moment about a transverse horizontal axis, Mt
3) Longitudinal shear force between the pontoons, FL
4) Longitudinal acceleration of deck mass, aL
5) Transverse acceleration of deck mass, aT
6) Vertical acceleration of deck mass, aV.
Figure 4-3
Characteristic hydrodynamic responses
Figure 4-8
Transfer function for torsion moment (example)
Figure 4-11
Longitudinal racking
Figure 4-12
Transverse acceleration
This shear force between upper and lower hull (deck and pon-
toon) has to pass through the construction in one of two ways,
see Figure 4-13:
a) For units with diagonal braces the shear force will be ex-
perienced as axial force in the braces and shear force with
corresponding bending moments in the columns. The dis-
tribution between these responses is depending on the
stiffness properties. For typical twin pontoon units the re- Figure 4-14
sponse in the columns is rather small and can be neglected Vertical acceleration
compared to the axial response of the diagonal braces.
b) For units without diagonal braces the approach will be
similar as presented in 4.6.5. 4.6.8 Vertical wave bending moment on the pontoon
This response will reach its maximum value at head seas, =
0. The critical wave length will be slightly larger than the pon-
toon length. The wave height may be derived from the steep-
ness relations given in 2.2.
The wave bending moment should be established with the
wave crest at mid pontoon, resulting in a symmetric wave
bending moment, see Figure 4-15.
For pontoons with three or more columns, the wave bending
moment should in addition be established with the wave zero-
crossing point in mid pontoon, giving an asymmetric wave
bending moment of the pontoon.
Figure 4-13
Transverse racking
Figure 4-15
Vertical wave bending moment
given in DNV-OS-C101 Sec.3. Global design load conditions stiffness such that the resulting global responses are appropri-
should be established based on representative variable func- ate for the design. Local analyses of such connections may also
tional load combinations. Limiting global mass distribution be required.
criteria should be established taking into account compliance Three types of global structural model are referred to in DNV-
with the requirements to intact and damage hydrostatic and hy- OS-C103 Appendix B (with Types 3, 4, and 5 as used in Ap-
drodynamic stability. pendix B):
For the pontoons the response is experienced as a bending mo- Type 3: Beam model.
ment due to the buoyancy in the pontoon, relevant both for
A complete three-dimensional structural beam model is
twin pontoon units and ring pontoon units. Effect of different
normally not accepted for global structural analysis except
ballast distribution shall be adequately accounted for as part of in special cases for twin pontoon units in benign waters,
the global load effects for the structural design. Simple frame following normal class survey intervals.
analyses with beam elements representing pontoons and col-
Type 4: Combined shell/beam model.
umns with fixed boundary conditions at deck level may be ap-
plied for assessment of different ballast distribution. For twin pontoon units with braces between pontoons and
Maximum and minimum ballast conditions, representing unfa- deck, such braces may be represented by beam elements.
In such cases, local analyses shall be performed for the
vourable conditions used in the tank load plans, are typically
brace connections to the hull and deck.
given in Figure 4-16. Effects with non-symmetric, diagonally
distributed tank loading conditions should also be assessed. Type 5: Complete shell model.
A complete shell model shall be applied for all ring pon-
toon units, both hull and deck. Plate stiffeners and girders
may be lumped and included as beam elements for correct
stiffness representation in the shell model.
The global structural model usually comprises:
pontoon shell, longitudinal and transverse bulkheads
column shell, decks, bulkheads and trunk walls
main bulkheads, frameworks and decks for the deck struc-
ture (secondary bulkheads and decks which are not taking
part in the global structural capacity should not be mod-
elled)
bracing structure for twin pontoon units.
Analysis with anisotropic plate stiffness (also referred to as
"stressed skin" design philosophy) may in certain cases be ap-
plied for the analyses and design of deck structures. Conditions
for application and design based on this philosophy are given
in 5.2.6.
Examples of global structural models of twin pontoon unit and
ring pontoon unit are given in Figure 4-17 and in Figure 4-18,
respectively. The twin pontoon model is a combined shell/
beam model (Type 4), and the ring pontoon model is a com-
plete shell model (Type 5).
Figure 4-16
Ballast distributions in pontoon
the responses from one model with the responses from another
model in order to assess the total utilisation of the structure.
Single model solution with use of the design wave analysis ap-
proach as described in 4.3 is not possible to combine.
4.8.3 Mass modelling
A representative number of global design load conditions, sim-
ulating the static load distribution for each draught, should be
evaluated in the global model. This may be achieved by the in-
clusion of a "mass model". The mass model may be an inde-
pendent model or may be implicitly included in the structural
(or wave load) model(s).
Usually, only a limited number of load conditions are consid-
ered in the global analysis. Therefore the global model may not
adequately cover all "worst case" global load distributions for
each individual structural element. Procedures shall be estab-
lished to ensure that the most unfavourable load combinations
have been accounted for in the design, see also 3.2.
In respect to global pontoon tank loading arrangements the
maximum range of responses resulting from the most onerous,
relevant, static load conditions shall be established. In order to
assess the maximum range of stresses resulting from variations
Figure 4-18 in pontoon tank loading conditions a simplified model of the
Example of ring pontoon global structural model (Type 5 shell structure may be created. This simplified model may typically
model) be a space frame model of the unit, see 4.7.
4.8.4 Boundary conditions
4.8.2 Assessment of single model solutions
To avoid rigid body motion of a global structural model, at
It is normally not practical to consider all relevant loads (both least 6 degrees of freedom have to be fixed.
global and local) in a single model, due to the following rea-
sons: Fixed boundaries or spring stiffness may be applied depending
on what is the most appropriate for the structure in question.
Single model solutions do not contain sufficient structural The selection of the boundary conditions may be as illustrated
detailing. in Figure 4-19, with the following restraints:
For ULS structural assessment, responses down to the lev- 3 vertical restraints (Z)
el of the stresses in plate fields between stiffeners are nor- 2 transversal horizontal restraints (Y)
mally required. Examples of insufficient structural 1 longitudinal horizontal restraint (X).
detailing may be:
internal structure is not modelled in sufficient detail to When spring stiffness for the vertical restraints are applied, the
establish internal structural response to the degree of total vertical stiffness should be according to the water plane
accuracy required area.
element type, shape or fineness (e.g. mesh size) is in-
sufficient.
Single model solutions do not normally account for the
full range of tank and sea pressure combinations, see 3.8.6.
Examples of effects that may typically not be fully ac-
counted for include:
internal tank pressure up to the maximum design pres-
sure
maximum sea pressures (e.g. by use of a "design
wave" approach) the sea pressure height resulting
from the design wave is not the maximum sea pressure
the section may be subjected to
variations in tank loading across the section of the Figure 4-19
pontoon; for example if the pontoon is sub-divided Boundary conditions
into watertight compartments across its section
load conditions that may not be covered by the global
structural analysis; for example ALS heeled condi-
tion. 5. Local Structural Analyses and Strength
Single model solutions do not normally account for the
Criteria
full range of "global" tank loading conditions, see also 4.7. 5.1 Introduction
Generally, single model solutions that do contain sufficient de- An adequate number of local structural models should be cre-
tail to include consideration of all relevant loads and load com- ated in order to evaluate response of the structure to variations
binations are normally extremely large models, with a very in local loads, for example in order to evaluate different tank
large number of loadcases. It is therefore often the case that it and sea pressure combinations, lay-down loads acting on deck
is more practical, and efficient, to analyse different load effects plate field, support loads of heavy equipment/items, stress con-
utilising a number of appropriate models and superimposing centration details for fatigue assessments, etc.
The model(s) should be sufficiently detailed such that resulting Model level 3 - Stiffener between girder model
responses are obtained to the required degree of accuracy. Sev- The purpose of the local structural analysis with stiffener mod-
eral local models may be required in order to fully evaluate lo- el is to analyse heavy loaded stiffeners and laterally loaded
cal response at all relevant sections. stiffeners, including brackets, subject to relative deformations
Design and scantlings shall be performed on basis of strength between girders/stringers.
criteria referred to in DNV-OS-C103 and DNV-OS-C101, The following typical areas should be given particular atten-
based on relevant combination of global and local load effects tion:
for each individual structural element, including various tank
and sea pressures acting on pontoon and column sections. longitudinal stiffeners between transverse bulkhead and
Guidance for superimposing responses are given in 5.3. the first frame at each side at of the bulkhead (pontoon)
Strength criteria for structural utilisation are referred to in 5.4. vertical stiffeners between horizontal decks and stringer in
column
5.2 Local structural analyses stiffeners at bulkheads in way of pontoon-column and col-
umn-deck intersections
5.2.1 General for twin pontoon units: brace to column connection and
Four typical modelling levels for analyses of a column-stabi- brace to deck connection
lised unit are described. The finite element modelling of the local support areas for support of for example fairleads,
unit or structural component should be carried out in accord- windlass, crane pedestal, drilling derrick, flare tower, liv-
ance with the principles and details given in the following. ing quarters, etc.
Other equivalent modelling procedures may also be applied. effect of different tank and sea pressure combinations, see
3.8.
Examples and guidance for local structural models are given in
sections 5.2.2 to 5.2.7. It should be noted that the model levels Model level 4 - Stress concentration models
might comprise a wide range of different models and sub-lev- For fatigue assessment, fine element mesh models should be
els of models. A local structural model may be included in made for critical stress concentration details, for details not
higher level model or run separately with prescribed boundary sufficiency covered by stress concentration factors (SCF) giv-
deformations or boundary forces. en in recognised standards, see for example DNV-RP-C203.
Model level 1- Global structural model The following typical areas should be given particular atten-
tion:
As part of the global response analysis, relatively coarse ele-
ment mesh model for the entire unit should be used. The over- hot spot stress at the cruciform plate connections in way of
all stiffness of the main load bearing members of the hull and pontoon-column and column-deck intersections
deck shall be reflected in the model. hot spot stress in the welded supports of for example fair-
The intention of the global analysis model should be to enable leads, crane pedestal, flare tower, etc.
the assessment of responses resulting from global loads. hot spot stress at local column/brace connection (twin pon-
toon)
The model should be used for analysing global wave responses hot spot stress at attachments
and still water responses where found relevant. See Chapter 4 details in way of the moonpool
for assessment of structural components, loads and effects to large and small penetrations
be included and considered as part of the global structural re- corners at door openings
sponse analysis. Note the assessment of single model solutions stiffener and girder terminations
in 4.8.2. weld profiling of cruciform joints
The global structural model may be used for redundancy anal- cast insert pieces.
yses (ALS), for example for twin pontoon units with ineffec- The size of the model should be of such extent that the calcu-
tive brace(s), see 5.2.7. lated stresses in the hot spots are not significantly affected by
Model level 2 - Girder model the assumptions made for the boundary conditions.
The purpose of the local structural analysis with girder model Element size for stress concentration analyses is normally to be
is to analyse structural details (e.g. transverse girders of pon- in the order of the plate thickness. Normally, 8-node shell ele-
toon or horizontal stringer of hull column) and loading condi- ments or 20-node solid elements should be used for the analy-
tions (e.g. different combinations of sea and tank pressures) sis. The correlation between different loads such as global
which have not been accounted for in the global analysis. bending, external and internal fluid pressure and acceleration
of the topside should be considered in the fatigue assessment.
The extent of the structural model should be decided based on For further details see DNV-RP-C203 item 2.13.
structural arrangements, loading conditions and method of re- In some cases detailed element mesh models may be necessary
sponse application. Typical girder model may be: for ultimate limit state assessment in order to check maximum
peak stresses and the possibility of repeated yielding.
frame model representing girder/stringer and effective
plating (beam elements) 5.2.2 Pontoons
three dimensional shell/membrane model with one or sev- The pontoon may be divided into the following structural ele-
eral girders/stringers, plating, web frames, major brackets, ments:
and stiffeners (e.g. as bar/beam elements).
pontoon top, sides and bottom
The following typical areas should be given particular atten- longitudinal bulkheads
tion: transverse bulkheads with girders
transverse frames and longitudinal girders.
transverse girders and bulkheads in pontoon
stringers and decks in column Example of a local analysis model of part of pontoon structure
effect of different tank and sea pressure combinations, is given in Figure 5-1.
see 3.8 Simple frame (beam element) or shell models may also be ap-
deck areas with concentrated or distributed loads. plied to for example transverse girders.
Figure 5-1
Part of ring pontoon local analysis model (model level 2)
5.2.3 Columns
The columns may be divided into the following structural ele-
ments:
Figure 5-2
shell plating and stiffeners Part of detailed shell element model of column/brace connection
horizontal stringers
horizontal decks
vertical bulkheads. 5.2.6 Deck structure
Local models may be of similar types as applied to the pon- Figure 5-3 shows a typical part of deck structure, with the deck
toons. divided into the following structural categories:
For local strength analysis in way of fairlead and windlass, the primary girders/bulkheads
design should be based upon the breaking strength of anchor secondary girders
lines, see also 6.1: stiffeners
plate.
local support of windlass
local support of fairlead
load effects on column shell between fairleads and anchor
winches.
The load effects on column shell between fairleads and anchor
winches may be estimated by hand calculations, applying rel-
evant width of the external shell plating of the column as load
carrying area, and combined with the stresses from the global
response analysis.
5.2.4 Pontoon/column and column/deck intersections
Local design of both pontoon/column and column/deck inter-
sections is normally governed by the fatigue criteria, with local
peaks of excessive yielding and buckling.
In particular the major intersections of pontoon/column of a Figure 5-3
Typical part of deck structure
typical ring pontoon unit are sensitive to fatigue fracture, re-
quiring special attention to local analyses and design at the fol-
lowing positions:
Primary girders/bulkheads
centre bulkhead pontoon/column intersection at pontoon The primary girder and the main bulkheads with effective part
upper deck of deck plating/heavy flanges take part of the global stiffness
pontoon outer wall/column intersection at pontoon upper and strength of the deck structure. Such primary girders will be
deck subject to both global and local responses, i.e. the stresses from
pontoon/pontoon intersections. the global analyses model should be superimposed with the
girder bending stresses caused by the local deck loads.
5.2.5 Brace to column connection (twin pontoon units)
The effects of cut-outs shall be considered. Large cut-outs/
Requirements for brace arrangements are given in DNV-OS- openings are normally included in the global analysis model.
C103 Sec.7 A200. The forces in the brace structures shall be Secondary girders
transmitted to the surrounding structure. The connections are
normally determined by fatigue of hot spots and yielding. The secondary girders are supported by the primary girders/
bulkheads.
The following considerations are recommended for the struc-
tural design and local analysis of brace to column connections: In cases where the secondary girders are taking part in the glo-
bal strength of the deck structure, these girders shall be de-
brace forces should be transmitted through sound connec- signed for the combined effect of global and local loads,
tions to bulkheads, decks and shell (circular connections) including girder bending stresses caused by the local deck
loads. Buckling of the panel, comprising girders, stiffeners, 5.2.7 Damaged structure
and plate, shall be considered. Requirements for Accidental Limit States (ALS) and structural
In cases where the secondary girders, plate, and stiffeners are redundancy of slender main load bearing structural elements
not taking part in the global strength, analysis and design based are given in DNV-OS-C103 Sec.6 and DNV-OS-C103 Sec.7.
on the anisotropic plate stiffness may be applied, see below for The damaged condition may be divided in two main groups:
conditions and recommendations.
Plate and stiffeners 1) Structural redundancy
comprising fracture of bracing or joint between bracings
All the various stress components (local and global) should (for twin pontoon units) and fracture of primary girder/
normally be evaluated and the relevant combination of stresses truss element in deck structure.
to be checked against the buckling and yield criteria.
2) Stability and watertight integrity
Fatigue see intact and damage stability requirements including
Fatigue evaluation of the deck structure shall be performed. compartmentation and watertight integrity in DNV-OS-
Such evaluation may be based on a screening process with a C301.
simplified fatigue analysis approach to identify critical areas/ The structural design comprise flooding of tanks and void
details. Special attention should be made to the following: spaces with heel angle after loss of buoyancy not to exceed
17.
areas of the deck where the dynamic stresses are signifi- The unit shall be designed for environmental loads with return
cant period not less than 1 year after damage, see DNV-OS-C101
areas of the deck where the stress concentration factor may Sec.3 B100.
become large, such as at penetrations, cut-outs, door open-
ings, attachments, flare supports, crane supports, etc. Fracture of bracing or joint (for twin pontoon units)
Analyses for static and wave induced loads should be carried
Design based on anisotropic plate stiffness out for damaged cases assuming successive bracings to be in-
Analysis with anisotropic plate stiffness (also referred to as effective. The global analysis model may normally be used.
"stressed skin" design philosophy) may be applied to the anal- Typical damaged case are illustrated in Figure 5-4.
yses and design of deck structures. The philosophy implies If the 100 year return period is used as basis for the analyses,
specific requirements for both the global model as well as the the 1 year responses may taken as:
local model(s) as referred to below.
1h
For the deck structure, the "stressed skin" philosophy may be 1 = 0.77 100
applied to large deck areas in-between primary girders/bulk-
heads. The stressed skin elements will represent plate panels
that only resist shear forces in the global analysis model. This 1 = one year stress response
means that all membrane stresses, both tensile and compres- 100 = 100 year stress response
sion stresses, are ignored in the panels. The purposes of intro- h = Weibull shape parameter; a value of 1.1 may be ap-
ducing stressed skin elements is to let primary girders/ plied together with-worldwide criteria for twin pon-
bulkheads and trusses (including thick deck plates representing toon units if not further documented.
heavy flanges close to the web) have sufficient strength to take Local yield and buckling can be accepted provided it can be
the global loads. The deck plates in-between will be designed demonstrated that excessive forces can be redistributed to oth-
to resist local loads and shear forces from global analysis. er members. Such redistribution may be demonstrated by dif-
Hence the structural design is based on the following basic as- ferent methods:
sumptions:
recalculation with reduced stiffness of elements with plas-
a) Plate panels with stiffeners are only assumed to resist glo- tic behaviour
bal shear stresses in plate and local loads. redistribution by hand calculation of the excessive forces
b) Secondary girders are assumed to resist local loads. obtained as the difference between the analysed forces in
the elastic analysis and the plastic capacity.
c) Shear forces may be redistributed to obtain equal shear
flow over the total panel length.
d) Primary girders/bulkheads/trusses (including heavy flang-
es in decks) carry the normal stresses from global analysis
model. These structures are treated with normal isotropic
material properties in the global analysis, and will take
care of the global strength integrity of the upper hull deck
structure (ULS).
e) Stressed skin elements may be modelled by adjusting the
material matrix for global analysis (ULS). Adjustments
may be performed by using anisotropic material model,
for example: maintaining the shear stiffness and divide the
axial stiffness by 100.
f) Note that fatigue evaluation based on analysis with
stressed skin elements will be non-conservative for the
stressed skin elements (see item e)).
Hence global analysis for fatigue assessment shall be per-
formed with isotropic material model (axial stiffness not Figure 5-4
modified). Example of damaged cases for redundancy analyses
Fracture of primary girder/truss element in the deck structure of the responses for logical load combinations.
The structural arrangement of the deck structure with main When evaluating responses by superimposing stresses result-
girders/truss elements is to be considered with regard to the ing from several different models, consideration shall be given
structural integrity after failure of any primary girder/truss el- to the following:
ement, similar as described above for fracture of bracing or
joint. Loads applied in global and local models as discussed in
3.2.
Fires or explosions may be the critical event for this accidental
scenario. Relevant combination of tank and sea pressures as dis-
cussed and specified in 3.8.6.
Heeled condition after damage flooding Assessment of different global and local model solutions
As part of the damaged stability requirement, see DNV-OS- as discussed in 4.8.2.
C301, the static heel angle is not to exceed 17 in any direction. It should be ensured that responses from design loads are
not included more than once (see for example the effect of
The structural design comprises flooding of tanks and void simple frame analysis models in sections 4.7 and 4.8.3 re-
spaces. The hydrostatic pressure in flooded spaces is the verti- lated to variations in pontoon tank loading conditions).
cal distance between a load point and damaged waterline in
static heeled condition (17), see 3.8.5. Continuous, longitudinal structural elements, for example
stiffened plate fields in the pontoon deck, bottom, sides,
At this 17 heel angle the gravity component parallel to the bulkheads, tunnels etc., located outside areas of global
deck is 0.29 g. This static component and the dynamic effects stress concentrations, may be evaluated utilising linear su-
shall be accounted for in the assessment of structural strength perposition of the individual responses as illustrated in
of deck structure. Local exceedance of the structural resistance Figure 5-5 for a pontoon section.
is acceptable provided redistribution of forces due to yielding, When transverse stress components are taken directly
buckling and fracture is accounted for. from the local structural model ( y (Local Model ) in Figure
In the damaged heeled condition (e.g. ALS condition after a 5-5), the transverse stresses from the global model may
collision event), the unit shall resist the defined environmental normally be neglected.
conditions corresponding to 1-year return period. It is not nor- Stiffener induced buckling failure normally tends to occur
mally considered practicable to analyse the global structure in with lateral pressure on the stiffener side of the plate field.
this damaged, inclined condition with wave loads as the deck Plate induced buckling failure normally tends to occur
structure becomes buoyant, due both to the static angle of in- with lateral pressure on the plate side. Relevant combina-
clination and also due to rigid body motion of the unit itself. tions of buckling code checking should therefore include
The global system of loading and response becomes extremely evaluation of the capacity with relevant lateral pressure
non-linear. Additionally, as soon as the deck structure starts to applied independently to both sides of the plate field.
become buoyant, the global load effects resulting from the in- In order to ensure that local bending stress components re-
clined deck mass rapidly become reduced. Hence, it is normal- sulting from loads acting directly on the stiffeners are in-
ly acceptable not to perform global response analysis with cluded in the buckling code check, the lateral pressure
wave loads for the 17o heeled condition. The effects of envi- should be explicitly included in the capacity check. The
ronmental loads may be accounted for by use of the material capacity checking should include a buckling check with no
factor M =1.33 as given in DNV-OS-C103 Sec.6 F102. lateral pressure in addition to the case with lateral pressure
(unless there is always pressure acting over the stiffened
5.3 Superimposing responses plate field being evaluated).
The simultaneity of the responses resulting from the local and Superimpose local compression stresses from bending of
global analysis models, including various sea and tank pres- deck girders (stiffeners) with global compression stresses
sures, may normally be accounted for by linear superposition in buckling check as described in 5.2.6.
Figure 5-5
Combination of stress components for buckling assessment of an individual stiffened plate field in a typical pontoon section
a) Breaking load of one single mooring line: rate for the life time of a structural element.
Slamming is defined as a situation occurring when a structural
F d, w1 = F B f member (brace) hits the water surface due to relative motion
between the unit and the waves. Two conditions must be ful-
filled to get slamming:
Fd,w1 = design load on windlass (corresponding to one
mooring line) The relative vertical motion must exceed the distance from
FB = characteristic breaking strength of one mooring line the exposed member to the still water surface.
f = 1.25 (load factor, see 6.1.1) The velocity of the member relative to the wave must be
The material factor M is 1.0 in this case. greater than a certain threshold value. This value is nor-
mally defined when the slamming force becomes greater
b) Operational loads from all mooring lines: than the buoyancy force (guidance 3-3.5 m/s).
The design of all structural elements influenced by the moor-
ing loads shall take into account relevant loads (ULS and ALS) By taking into account the probability for exceeding the above
found from the mooring analysis. mentioned conditions, number of slamming and most probable
largest slamming force per unit length during N wave cycles
The static and dynamic contributions to the mooring line forc- may be calculated. The above mentioned slamming criteria
es should be considered for relevant application of load and may be calculated for different sea states (Hs, Tz) and speeds.
material factors according to DNV-OS-C103 Sec.4 (ULS) and It may be possible to specify limiting sea states for when it is
sec. 6 (ALS). necessary to submerge from transit draught or to slow down
forward speed.
6.1.3 Horizontal and vertical design angles for the fair-
lead supports Slamming loads from waves and procedure for fatigue damage
are given in Classification Note 30.5 6.4.
Fairlead with vertical inlet angle and horizontal working angle
is shown in Figure 6-1, with mooring line tension T. 6.3 Air gap
The most critical vertical inlet angle and horizontal working Requirements for sufficient air gap are specified in DNV-OS-
angle shall be considered for the local strength analysis and C103 Sec.4 D100.
design of the fairlead supporting structure:
In the ULS condition, positive clearance between the deck
The vertical design inlet angled should not be taken larger structure and the wave crest, including relative motion and in-
than 10 unless otherwise documented (the support result- teraction effects, should normally be ensured. Localised, neg-
ant force on supporting structure will decrease with in- ative air gap may be considered as being acceptable for
creased). overhanging structures and appendages to the deck structure.
The horizontal design working range (DWR) should be In such cases full account of the wave impact forces is to be
20 larger than the operational working range (2 ), see taken into account in the design. The consequence of wave im-
DNV-OS-E301 Ch.2 Sec.4 L300: pact shall not result in failure of a safety related system (e.g.
DWR = 2 + 20 lifeboat arrangements).
. The wave asymmetry factor in air gap calculations is given in
2.3.6.
It is recommended, in the design phase, to consider operational
aspects, including requirements to inspection and mainte-
nance, which may impose criteria to air gap that exceed mini-
mum requirements.
In the context of DNV-OS-C103 Sec.4 D103, column run-up
load effects are not considered as resulting in negative air-gap
responses.
6.4 Transit condition
6.4.1 General
Figure 6-1 Weather restrictions and criteria related to the transit condition
Fairlead with vertical inlet angle ( ) and horizontal working an- shall include consideration of the following items:
gle ( )
1) Motions and accelerations; see sections 6.4.2 and 6.4.3 be-
low
6.2 Wave slamming 2) Wave slamming, see 6.2
Slamming on bracing structures on typical twin pontoon units 3) Stability during the ballasting sequences, see DNV-OS-
may be a governing criterion for deciding when to submerge C301.
from transit draught. It is normally stated as a design criterion
that significant slamming should be avoided. For design stage 6.4.2 Transit analysis
evaluation of the transit capabilities the following criteria may
be used: To decide the critical sea states when it is necessary to sub-
merge a column-stabilised unit from transit draught, the design
Slamming frequency: The number of slamming impacts should responses as defined in 4.6.1 may be calculated in transit con-
not exceed 1 slam per 20 wave encounters. dition (strip theory or 3-D diffraction theory), and the maxi-
Extreme stress: The extreme slamming induced stress in the mum responses taken equal to the values obtained for the
brace in a certain stationary design period (e.g. 3 hours) should survival/operating conditions in a 100 year storm condition.
satisfy ULS combination b). An example is shown in Figure 6-2. From such chart it may be
Fatigue rate: The fatigue growth rate due to slamming in tran- possible to conclude which responses that may be critical for
sit condition should not be greater than the nominal average the transit condition as well as limiting environmental criteria.
MO
Heave: ---------- = 1.0
a
Roll and Pitch: --------- = 2
a
MO = motion amplitude
a = wave amplitude
= rotation amplitude in radians
Figure 6-2 = wave length
Critical sea states for transit condition (example)
Surge should be corrected due to coupling with pitch. Normal-
ly acceptable multiplication factor for the linear motion of the
Direct calculations of the transit analysis may be an alternative resonance peak is 0.5.
approach, with similar wave load and stress analysis approach Sway and yaw are not so strongly affected by the linearities
as for the operating and survival conditions. and no corrections should be made.