Академический Документы
Профессиональный Документы
Культура Документы
O v e T. G u d m e s t a d a & Geir M o e b
ABSTRACT
745
746 O. T. Gudmestad, G. Moe
cients. Measurement programmes to obtain full scale global force data simulta-
neously with wave and current data are furthermore recommended.
© 1996 Elsevier Science Limited.
1 INTRODUCTION
TABLE 1
Parameters for Calculation of Deterministic Hydrodynamic Loading
TABLE 2
20th vs 19th Edition of API RP 2A Wave Force Procedures 2' 3,4 and Gulf of Mexico Meto-
cean Criteria 17
first step (called long term statistics) the highest significant wave height and
its associated period are predicted from field data, usually based on averages
of 20 rain periods, registered at 3 hour intervals. The use of individual storms
may be a better strategy, however, since these may be assumed to be statis-
tically independent events. 15'21 In the second step (short term statistics), the
expected amplitude al of the highest wave for such an extreme seastate is
estimated, assuming linearity, so that the higher peaks will be Rayleigh
distributed. For predictions of surface geometry, second order terms may be
added, increasing the peaks and decreasing the depths of the troughs, relative
to the linear estimate, but the particle velocities seem to be better predicted
on basis of the linear amplitude al. An adjustment factor to account for
wave directionality is a useful concept and is included in the API recipe
which, however, also recommends the use of a regular nonlinear design
wave, e.g., a Stokes 5th order wave. An approach that may prove more
advantageous is to predict the wave form and the particle kinematics from
the statistically based new-wave. 2°' 12
API recommends the following drag and inertia values for unshielded
circular cylinders:
Smoot]~ cylinders: Ca = 0.65, C,, = 1.6
Rough cylinders: Ca = 1.05, C,n = 1.2.
These values are said to be appropriate for
- - the case of a steady current with negligible waves;
or
may be used for the global dynamic analysis, provided the linearization of
the drag force can be justified.
The hydrodynamic coefficients developed for use with individual determi-
nistic waves can, according to API, also be used for random wave analysis
(either time or frequency domain) of fixed platforms by using:
--significant wave height
and
--spectral peak period
to calculate K, the Keulegan-Carpenter Number. 3'4
3 H Y D R O D Y N A M I C C O E F F I C I E N T S R E C O M M E N D E D BY D N V
The DNV rules 6' 7 represent North Sea Design Practice for the calculation of
hydrodynamic loads on offshore truss structures. These suggest tentative
values of Cm for different cross-sectional shapes. For circular cylinders the
value amounts to 2-0. It is in particular noted that DNV call for use of the
selected Cm value in 'Conjunction with the acceleration of water particles as
calculated using an appropriate wave theory' (see also Ref. 23).
The selection of Ca values should, according to Ref. 6, take into account
the variation of Ca as a function of:
- - Reynold's number, Re;
--Keulegan-Carpenter number, Kc;
- - r o u g h n e s s number kr/D where kr is the effective roughness height
and D the diameter of member, kr/D = 10 -2 in the absence of more
reliable data for marine growth);
- - v a r i a t i o n of cross-sectional geometry.
Tentative values for the drag coefficient for a circular cylinder of varying
roughness in steady flow are shown in Fig. 1, while tentative values in the
supercritical regime in steady flow for some in-service marine roughnesses
are given in Fig. 2. Note that the following values of surface roughness k,
could be used in the determination of the drag coefficient:
k, (metres)
Steel, new uncoated 5 x 10-5
Steel, painted 5 x 10-6
Steel, highly rusted 3 x 10-3
Concrete 3 x 10-3
Marine growth 5 x 10-3-5 x 10-2
Calculationof hydrodynamicloads 751
1.4
1.2
1.0
o= 0.s
~ 0.6
[kdDxl0-3f 3.1 V ~"¢, ~ " °-"°"
~ 0.4 L L
I 2/"z.
,(
1.4 , , .°.."
0.2
~ I I I II I1 I SIm~ltil(kr/D"~';)
I , , 1 I I I
105 106
Reynold's number (Re)
Fig. 1. The drag of sand-roughened cylindersin steady uniform flow.
1.4
: : .,"
1/20 , ~
1.5
1/100 . . . . , N ~
.~ 1.0
k/D< 1 / 1 0 0 0 0 ~ "'°",,. .................
(smooth) ~
U
0.5
Note that this reduction should not apply to the design o f individual
members.
Reference 7, furthermore, allows for the introduction o f group effects
(shielding and blockage etc.). It should also be noted that the effect o f
marine growth on appurtenances such as anodes, etc. should be considered
when selecting effective diameters and drag coefficients. Standard industry
practice :is to increase the drag coefficient by 7-10% to account for anodes.
F r o m the above it can be concluded that the drag coefficients recom-
mended by D N V in general are consistent with the API recommended prac-
tices, except for the acceptance o f using a Cd value of 0.7 for rough members
in global deterministic wave analysis. If this value is used together with
group effects (blockage and shielding) and the best estimate o f wave and
current kinematics based on joint probability calculations, the global force
for a slender offshore structure m a y be substantially less than found by using
the recent API recipe. 3'4 Note, however, that it is not North Sea Design
Practice 1:o consider joint probability o f waves and currents.
A brief comparative analysis o f the loading on a riser platform located in
82 m water depth in the North Sea has been carried out using the input values
given in ".Fable 3. The results o f this analysis are presented in Table 4.
4 N P D ' s A P P R O A C H TO S E L E C T I O N O F H Y D R O D Y N A M I C
COEFFICIENTS
TABLE 3
Input Values for Comparison Analysis
TABLE 4
Sensitivity Analysis for Calculation of Loads on a North Sea Jacket Riser Platform
The preferred codes are those which use consistent values for all para-
meters in the load recipe. To use conservative estimates of wave plus
current kinematics in combination with low drag factors for deterministic
analysis is not considered consistent. In order to assess the safety margins
that result when using the latter approach, a full calibration is necessary
for any new concept and any new application (waterdepth, etc.). API, on
the other hand, suggest an approach which is considered to be consistent.
This facilitates application to new concepts and new conditions and also
makes code modification easier and is believed to be a good argument for
selecting the new API RP 2A recommendations 3 as the most relevant code
for estimating hydrodynamic loads on truss structures.
6 ASSESSMENT OF U N C E R T A I N T I E S IN ESTIMATE OR
H Y D R O D Y N A M I C COEFFICIENTS
In view of the very large difference between the API recommended hydro-
dynamic drag coefficient for rough members (Cd = 1-05) and the standard
North Sea deterministic global design practice ( C d = 0 . 7 for rough
members), key attention should be focused upon resolving the major differ-
ences between the input to the load calculation recipes.
In probabilistic analyses using state-of-the-art values for all parameters
together with best estimate of parameter uncertainties, the uncertainty in the
API reconamended hydrodynamic coefficients could be established by revi-
siting data from relevant model tests. For the drag coefficient a variation of
less than 10% is expected.
It is concluded that the new API recipe 3' 4 for the calculation of hydrodynamic
loads on truss structures uses state-of-the-art values for hydrodynamic coeffi-
cients as well as for other parameters influencing the load recipe. North Sea
Design Practice for global deterministic analysis 7 combines a low drag factor
with conservative estimates of wave kinematics and current.
It is strongly recommended that the most relevant values for all the para-
meters influencing the estimation of loads on offshore truss structures are
used. Following this recommendation it will be necessary to carry out careful
full scale measurements of global force data simultaneously with wave and
current measurements to identify the design (10 -2 per year) loading as well
as the wave plus current wave kinematics level which should be used in
conjunction with proper hydrodynamic coefficients. It is recommended that
North Sea practice be changed to incorporate such measurements and most
relevant hydrodynamic coefficients. Dramatic increases in design loads are
not expected, although the total loads at certain locations may increase
above the design values in current use.
Any attempt to utilise the low drag coefficient normally used in the North
Sea Design Practice in combination with the other parameters of the API
recipe, 3 i.e., wave kinematics factor, low current value, current blockage and
shielding, i.e., parameters from different recipes, may result in an under-
estimation of design loads potentially leading to unsafe structures, and this
must by all means be avoided.
ACKNOWLEDGEMENTS
REFERENCES