Sie sind auf Seite 1von 14

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269194250

New prediction method for wave-in-deck loads on exposed piers/jetties

Conference Paper · April 2007


DOI: 10.1142/9789812709554_0376

CITATIONS READS

3 172

3 authors, including:

William Allsop
The University of Edinburgh
178 PUBLICATIONS   1,442 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Blocsnet View project

FloodSite View project

All content following this page was uploaded by William Allsop on 26 August 2015.

The user has requested enhancement of the downloaded file.


NEW PREDICTION METHOD FOR WAVE-IN-DECK LOADS ON
EXPOSED PIERS / JETTIES / BRIDGES

William Allsop1, Giovanni Cuomo2 and Matteo Tirindelli3


This paper extends and improves methods to predict hydraulic loadings on deck and
beam elements of open pile piers / jetties exposed to wave action. This work is based on
hydraulic modelling under the UK “Exposed Jetties” project, but draws on further analy-
sis of the experimental data from that study. The paper describes briefly the measure-
ments of wave-induced loads on deck and beam elements in physical model tests, de-
scribes how those data have been refined and re-analysed to improve prediction methods.
The methods discussed here are therefore intended to improve on the original “Exposed
Jetties” predictions for wave-in-deck loads to be used in analysis of jetties / piers (and
highway bridges) exposed to large waves.

INTRODUCTION
Over the last 10-15 years there has been increased demand for development
of large single use industrial terminals (especially those for Liquid Natural Gas
(LNG), and Liquid Petroleum Gas (LPG), but also for grain, cement, oil or
chemicals). These often require deep water for larger vessels, but do not neces-
sarily need shelter to the approach trestles carrying the delivery lines. These
terminals may be required in remote locations where there is no wave shelter, no
existing infrastructure and the construction of new protective breakwaters for the
facility may not be cost effective. In many instances therefore, the jetties and
approach trestles have been constructed without breakwater protection.
Elsewhere, other structures may suffer wave loads on beam or deck ele-
ments. Examples include construction phase falsework hanging below the main

Section A - A Section A - A Section A - A

B B B
Section B - B Section B - B Section B - B

A A A A A A

B B B

Fig. 1 Structural configurations tested during the Exposed Jetties project.

1
Technical Director, HR Wallingford, Howbery Park, Wallingford, OX10 8BA, UK,
w.allsop@hrwallingford.co.uk
2
University of Rome, “RomaTre”, Department of Civil Engineering, Via Vito Volterra 62,
00146 Rome, Italy, cuomo@uniroma3.it
3
Coast & Harbor Engineering, 155 Montgomery Street, Suite 608, San Francisco, CA
94104 USA, matteo@coastharboreng.com
1
2

structure, and highway bridges in hurricane or cyclone-prone areas.


PREVIOUS WORK
Under the Exposed Jetties research project, a general method for prediction
of wave-induced loads on exposed jetties was developed by McConnell et al.
(2003, 2004) using empirical analysis of results from physical model tests at HR
Wallingford (Tirindelli et al. 2002, Allsop & Cuomo, 2004, among others). The
intention of the methods presented was to provide designers with safe and easily
applicable predictive equations for quasi-static (pulsating) wave loads on beams
and decks hit by wave action.
Three different structural configurations were used in these model tests (Fig.
1), namely flat deck (FD), Panels (P or 2d) and No-panels (NP or 3d), chosen by
the steering group to represent a wide range of structure alternatives. The main
tests used the NP configuration, but when strong 3-dimensional effects were no-
ticed, particularly for the downward or inundation loading, side panels (P) were
added to control the sideways flow effects. Both of these configurations featured
beams projecting downward below the general deck level. The third configura-
tion inverted the deck to explore (uplift) loadings on a platform with a flat un-
derside.
The simplified general method and coefficients described in the Exposed
Jetties guidelines does however suffer from having to include variations in the
loading process, these different structural configurations, the relative position of
the element along the jetty, and wave conditions / deck clearance.
A further problem was that some measurements from the beam and deck
measurement elements had been corrupted by high-frequency vibrations of the
measurement element, not sufficiently removed in the initial analysis, see dis-
cussion by Cuomo et al (2003).

FURTHER ANALYSIS OF WAVE LOADINGS


Following on from the
original Exposed Jetties pro-
ject, an improved method for
the analysis of non-stationary
time-history loads was devel-
oped by Cuomo et al. (2003)
and Cuomo (2005) based on
wavelet transform. This
transform was used to filter
out corruption in some of the
recorded signals from dy-
namic responses of the meas-
uring instruments (Fig. 2).
The new analysis did how- Fig. 2 Signal filtering by wavelet analysis (NB
ever retain the key transient model units).
signal characteristics beyond
3

the capabilities of Fourier methods. The application of wavelet transform meth-


ods allowed generation of a revised data set, in turn allowing further analysis of
these revised data. In general, changes to the wave loads were relatively small,
but in a few instances peak values were revised downwards.

NEW PREDICTION FORMULAE


Slowly-varying loads
Confirming previous results by Tirindelli et al. (2003), the dominant vari-
ables for the loading process were again found to be the incident significant
wave height (Hs), the maximum crest elevation (ηmax), and the clearance of the
element above the s.w.l. (cl). Revised prediction formulae (and coefficients)
were developed using the following dimensionless variables:
Fqs
Fqs* = (1)
ρgH S A
η max − cl
η* = (2)
d
where: A = face area of element normal to direction of wave attack; d = local
water depth.
A relatively linear growth of forces Fqs* with η* is generally recognisable
for all the structural configuration and elements for both vertical and horizontal
loads. The following prediction formula is therefore proposed for the prediction
of the quasi-static load Fqs:
( )
Fqs = a ⋅η * + b ⋅ ρgH S A (3)
Empirical coefficients a and b have been evaluated by fitting data from the
physical model tests. Further analysis comprehended distinguishing among data
from tests run with different structural configurations and wave loads on differ-
ent structural elements and at different locations along the structure. This al-
lowed specializing Equation 3 to better fit each sub-data sets, reducing scatter
around prediction and capturing different features of the loading process. Values
of coefficients a and b are listed in Table 1.
Effects of element position along the structure on loading are illustrated in
Fig. 3, showing vertical force time-history recorded simultaneously on the ex-
ternal and the internal deck elements during physical model test run with the no-
panel (NP) configuration. Different importance of green water overtop-
ping/inundation on internal and external elements is evident.
Variations to loadings are further demonstrated in Fig. 4 showing dimen-
sionless vertical loads on different elements for different structural configura-
tions and position along the structure.
4

Table 1 Coefficients a and b for revised prediction method, vertical (up-lift) and
horizontal (positive) wave loads

Parameter Direction Element Position Config. a b


Force Horizontal Beam External All 1.19 0.43
Force Horizontal Beam Internal NP 0.56 0.75
Force Vertical Beam External FD 1.74 0.14
Force Vertical Beam External P 0.71 0.57
Force Vertical Beam External NP 1.10 0.46
Force Vertical Beam Internal NP 1.36 0.46
Force Vertical Deck External FD 2.31 0.05
Force Vertical Deck External P 1.23 0.51
Force Vertical Deck External NP 1.57 0.52
Force Vertical Deck Internal FD 0.83 0.13
Force Vertical Deck Internal P 0.58 0.19
Force Vertical Deck Internal NP 1.57 0.73

2
External
Internal
1.5

1
FV / ( ρ g Hs A )

0.5

-0.5

-1
502.2 502.4 502.6 502.8 503 503.2 503.4 503.6
t / Tm

Fig. 3 Different effect of inundation on external and internal deck elements

Impact loads
Once quasi-static loads Fqs have been evaluated, the impulsive force Fimp
acting on the element can be evaluated as follows:

Fimp = a Fqs+ (4)

where a is again an empirical coefficients from data fitting.


5

External Beam - P External Beam - NP


1.2 1.2
1 1
F / ( ρ g H s A)

F / ( ρ g H s A)
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0.4 0 0.2 0.4 0 0.2 0.4
External Deck - P External Deck - NP
1.2 1.2
1 1
F / ( ρ g Hs A)
F / ( ρ g H s A)

0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0.4 0 0.2 0.4 0 0.2 0.4
Internal Deck - P Internal Deck - NP
1.2 1.2
1 1
F / ( ρ g Hs A)
F / ( ρ g H s A)

0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0.4 0 0.2 0.4 0 0.2 0.4
(η max - cl )/d ( ηmax - cl)/d

Fig. 4 Effect of different structural configurations on dimensionless uplift force,


from top to bottom: external beam; external and internal deck elements

Details of the new prediction method are given in Cuomo et al. (2007). In
the following, the procedure for the evaluation of the design wave-in-deck loads
by the new method is first briefly summarized, then compared with existing
methods and finally applied to example case studies.
6

Ext Deck - NP
Ext Deck - P
Ext Deck - FD
Int Deck - NP
FV,qs+ Measured [N]

80 Int Deck - P
Int Deck - FD
60

40
20

0
0 20 40 60 80
FV,qs+ Predicted [N]
Ext Beam - NP
Ext Beam - P
FV,qs+ Measured [N]

20 Ext Beam - FD
Int Beam - NP

10

0
0 5 10 15 20 25
FV,qs+ Predicted [N]
Pa - NP
Fig. 5 Comparison of measured vertical (uplift, top graph) and horizontal (seaward, bot-
tom graph) forces on deck and beams with prediction by Equation 3

Outline of the new method


Evaluation of design loads by the new method proceeds through the follow-
ing steps:
1. evaluate the design wave height and period along the structure;
2. evaluate the maximum crest elevation ηmax along the structure as a
function of the local water depth d, Hs and Tp;
3. evaluate the wetted area A corresponding to:
a. the incident wave kinematics: ηmax, Tp;
b. the geometrical characteristics of the structure;
c. the type of wave load (horizontal/vertical);
4. evaluate dimensionless parameters Fqs* and η* as defined in equa-
tions 1 and 2 above;
5. evaluate Fqs+ and Fqs- using equations 3 above;
6. evaluate Fimp using equation 4 above

When compared to the revised data set (Fig. 5), predictions by Equation 3
with appropriate coefficients (see Table 1) compare satisfactorily with meas-
7

urements, especially when considering that this approach still uses a single
rather simple relation to predict loads on significantly different elements or posi-
tions on the jetty.
Horizontal (seaward) - External Beams
2250
Guidelines
2000 Present method

1750
Predicted [Pa]

1500

1250

1000

750

500
500 750 1000 1250 1500 1750 2000
Measured [Pa]
Fig. 6 Predictions of uplift forces on an external beam element using the Exposed Jetties
guidelines (open circles) and new method (dots).

COMPARISON WITH EXISTING METHODS


Improvements over the “Exposed Jetties” predictions include both the re-
moval of high-frequency corruption from some signals, but also better resolution
of the effects of structural configurations where those data had been ‘lumped to-
gether’ in the original analysis.
Uplift - Internal Deck
100
Guidelines
90
Present method
80

70
Predicted [N]

60

50

40

30

20

10

0
0 10 20 30 40 50 60 70 80 90 100
Measured [N]

Fig. 7 Predictions of uplift forces on an internal deck element using the Exposed Jetties guide-
lines (open circles) and new method (dots).
8

These improvements are illustrated in Figs. 6 and 7 where predictions using


the new method and coefficients give rather better agreement than the “Exposed
Jetties” formulae and coefficients. For the external beam in Fig 4, the revised
method gives high effective pressures (NB – model values) than given by the
“Exposed Jetties” method.
For up-lift forces on an internal deck element in Fig 7, the previous guide-
lines appear to show a rather ‘flat’ response, under-predicting at one end of the
range and over-predicting at the other end.

Fig. 8 US highway bridge at Gulport showing major damage from wave-in-deck loads
during hurricane Katrina.

CASE STUDY EXAMPLES


An interesting and unusual application of these prediction methods is to the
cases of a number of US highway bridges damaged during hurricane Katrina. A
significant number of these bridges support highways (e.g. US10 and I90) run-
ning along the Florida, Alabama, Mississipi, Louisiana coastlines. In almost all
instances, unrestrained or lightly restrained decks were lifted off their pile caps
and dropped into the water, Figs. 8 and 9.

Fig. 9 US highway bridge I90 at Biloxi showing major damage along lower spans with
less or no damage for elevated spans
9

7
Water level (mD) or significant wave height (m)

1 Water level, mD

Wave heights, Hs, m


0
08:00 08:30 09:00 09:30 10:00 10:30 11:00 11:30 12:00
Time, hrs

Fig. 10 Water levels and wave heights for I90 at Biloxi (courtesy Dr Jim Chen, Prof
Scott Douglass, USA)

No measurements of wave conditions or water levels at these bridges were


available, but as part of a co-operation with University of South Alabama and
Federal Highways, time series of water levels and wave conditions were gener-
ated by Dr Jim Chen of USA for the Interstate 90 Bridge at Biloxi (Fig. 9) for
the peak of Katrina, see Fig 10.
These hindcast water levels and wave conditions were used to calculate
wave forces on idealised deck elements on I90 at spans 44 and 100. These repre-
sent one of the spans (span 44) that were damaged during Katrina, and one at a
slightly higher level that was not displaced (span 100).
A set of comparison calculations were made for uplift forces on deck ele-
ments at span 44 and span 100 where the soffitt level increased from +4.8mD to
+8.1mD. The projected water levels varied from +3.63mD at 08:00, up to +6.55
mD at 10:30, then falling back away. The deck was taken as 7.67m x 15.85m x
0.15m, giving a projected (dry) weight of 472 kN, but cast into the deck were 6
support beams each of 188 kN and two edge beams of 38 kN, giving a total
(dry) restraining force of 2360kN.
Using the “Exposed Jetties” methods, uplift forces calculated for the 4 hours
around the highest water levels are shown in Fig 11, as is the dry weight of
2360kN. The uplift load calculations use equations / coefficients for the two dif-
ferent methods. In this instance, the agreement at maximum load is very good,
although the two methods differ for small loads (low values of ηmax - cl). Both
methods show that for span 100, wave up-lift loads on the deck would just have
reached the dry weight, but not have exceeded it by any significant margin.
Conversely, the dry weight for those spans that were lifted was exceeded ap-
proximately an hour before the peak of the storm, allowing ample time for the
10

4000

3500

3000
Up-lift force, kN

2500

2000

1500

1000
Span 44, Exp Jetties
500 Span 44, New method
Span 100, Exp. Jetties
Span 100, New method
Deck weight
0
08:00 08:30 09:00 09:30 10:00 10:30 11:00 11:30 12:00
Time, hrs

Fig. 11 Uplift load for I90 at Biloxi compared with self weight of deck + beams.

sideways loads to overcome any limited restraint and displace the deck / beam
sets.
An additional example of the cruise terminal piers at Cozumel Island is dis-
cussed by Bardi et al (2007). The piers were struck by Hurricane Wilma in 2005
and one of them was completely destroyed, see Fig. 12.

Fig. 12 – Cozumel cruise terminal before and after Hurricane Wilma

Wave loads on the pier’s unsupported, exposed deck sections were evalu-
ated using three different methods: Kaplan (Kaplan et al, 1995), the method of
11

this paper, and a VOF numerical model. The method described here was as-
sessed as providing the most conservative, although reasonable, estimates of
wave uplift loads, see Fig. 13.

40,000
VOF
Newimpact
Cuomo Method Impact Kaplan
30,000
Uplift Load (N/m2)

20,000

10,000

0
18 20 22 24
-10,000
Cuomo
Newdownward
Method Downward
-20,000
Time (s)
Fig. 13 Evaluation of uplift force on pier deck at Cozumel
(adapted from Bardi et al, 2007)

CONCLUSIONS
A re-analysis of wave-in-deck loads recorded under the HR Wallingford’s
Exposed Jetties project by means of wavelet transform has allowed a new inter-
pretation of the data set and the development of an improved prediction method.
The new method (Cuomo et al. 2007) allows accounting for the variability
of loading with type and location of element within the structure as well as for
the configuration of the main structure.
When compared with prediction by Exposed Jetties formulae the new me-
thod reduces bias and scatter and improves resolution.
The new method has been successfully applied to explain failure of bridges
and piers in hurricanes Katrina and Wilma.

ACKNOWLEDGMENTS
Support by Universities of Rome TRE and Bologna, HR Wallingford, Coast
and Harbor Engineering and the Marie Curie programme of the EU are grate-
fully acknowledged. Research data used in this analysis were supported by DTI
under contracts PECD 7/6/263 & 312. Second author studentship is founded by
Japanese Society for the Promotion of Science. The authors wish to thank John
Bardi (BERGER ABAM) for providing details on Cozumel case study, and Fed-
eral Highways / University of South Alabama for data relating to the I90 bridge.
12

REFERENCES
Bardi J., Ostbo B., Fenical S. & Tirindelli M. (2007) Cozumel's International
Cruise Terminal - Hurricane Wilma Recovery and Reconstruction, paper
accepted for Ports 2007, ASCE
Bentiba R., Cuomo G., Allsop N.W.H. & Bunn N.P., 2004. Probability of occur-
rence of wave loading on jetty deck elements. Proc. ICCE 2004, Lisbon.
Cuomo (2005) “Dynamics of wave-induced loads and their effects on coastal
structures” PhD Dissertation, University of Roma TRE, Italy.
Cuomo G., Tirindelli M. & Allsop N.W.H. (2007) Wave-in-deck loads on ex-
posed jetties. Coastal Engineering, Elsevier, in press 2007.
Cuomo G., Allsop N.W.H. & McConnell K., 2003. Dynamic Wave Loads on
Coastal Structures: Analysis of Impulsive and Pulsating Wave Loads.
Coastal Structures, Portland.
Kaplan P., Murray J.J. & Yu W.C., 1995. Theoretical analysis of wave impact
forces on platform deck structures. Offshore Technology, OMAE Copenha-
gen, Volume 1-A.
McConnell, K. J., N. W. H. Allsop, & I. C. Cruickshank. (2004) Piers, jetties
and related structures exposed to waves London, UK: Thomas Telford.
McConnell, K.J., N.W.H. Allsop, G. Cuomo, & I.C Cruickshank. (2003) “New
guidance for wave forces on jetties in exposed locations." Proc. COPEDEC
VI, Colombo, Sri Lanka. 20pp.
Rienecker M.M. & Fenton J.D., 1981. A Fourier approximation method for
steady water waves. Journal of Fluid Mech., 104:119-131.
Tirindelli M., Cuomo G., Allsop N.W.H. & Lamberti A. (2003) “Wave-in-deck
forces on jetties and related structures” Proc. ISOPE 2003, Honolulu, Ha-
waii, 3, pp. 823-830
Tirindelli M., McConnell K., Allsop N.W.H. & Cuomo G., 2002. Exposed Jet-
ties: Inconsistencies and Gaps in Design Methods for Wave-Induced Forces.
Proc. 28th ICCE (ASCE), Cardiff: 1684-1696.
13

KEYWORDS – ICCE 2006

NEW PREDICTION METHOD FOR WAVE-IN-DECK LOADS ON EX-


POSED PIERS / JETTIES”1619

Wave-in-deck loads
Design of coastal structures
Wave-structure interaction
Suspended deck structures
Jetty
Piers
Coastal bridges

View publication stats

Das könnte Ihnen auch gefallen