Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
0Activity
0 of .
Results for:
No results containing your search query
P. 1
Probabilistic optimisation of revetment on dikes along Frisian coast

Probabilistic optimisation of revetment on dikes along Frisian coast

Ratings: (0)|Views: 0|Likes:
Published by Patricia Dillon
paper 42 For centuries the dikes in the Netherlands have been protected against wave attack by revetments constructed from pitched blocks. Based on experience a size was selected and, if a severe storm damaged the revetment, heavier blocks were applied. Recent research has however shown that the design storm surge would cause severe damage to the protection of the dike. A new design philosophy is developed
that assesses the function of the revetment in the entire dike and optimises the thickness probabilistically.
paper 42 For centuries the dikes in the Netherlands have been protected against wave attack by revetments constructed from pitched blocks. Based on experience a size was selected and, if a severe storm damaged the revetment, heavier blocks were applied. Recent research has however shown that the design storm surge would cause severe damage to the protection of the dike. A new design philosophy is developed
that assesses the function of the revetment in the entire dike and optimises the thickness probabilistically.

More info:

Published by: Patricia Dillon on Apr 12, 2014
Copyright:Traditional Copyright: All rights reserved

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

04/12/2014

pdf

text

original

 
1
 
INTRODUCTIONFor centuries the dikes in the Netherlands have been protected against wave attack by revetmentsconstructed from pitched blocks. In the old daysthese blocks took the form of relatively roundedrock (Vilvoordse steen, granite boulders) or basalthexagones. Presently, rectangular concrete blocksare frequently used, but basalt is also still applieddue to the high specific mass and the interesting vis-ual appearance.Until recently the dimension of the blocks waschosen empirically. Based on experience a size wasselected and if a severe storm damaged the revet-ment heavier blocks were applied. When physicalmodels verified with scale model tests were devel-oped, it appeared that revetments were able to with-stand e.g. a 1/10 year storm, but that the designstorm surge would cause severe damage to the pro-tection of the dike. Consequently, it is decided to re- place the revetments on the Dutch dikes.A new design philosophy has to be developedthat centres on the function of the revetment in thetask of the entire dike to protect the hinterland frominundation. In this paper a first effort is made to op-timise the revetment size in this framework.2
 
DESCRIPTION OF THE CASEThe Frisian coast is situated in the Northern part of the Netherlands, along the Wadden Sea.Study area
Figure 1: The location of the study area
The probabilistic optimisation of the revetment on the dikes along theFrisian coast
M. Hussaarts, J.K. Vrijling, P.H.A.J.M. van Gelder 
Subfaculty of Civil Engineering, Delft University of Technology, The Netherlands
H. de Looff 
 Ministry of Transport, Public Works and Water Management, Road and Hydraulic Engineering Division, The Netherlands
C. Blonk 
Simtech BV, The Netherlands
ABSTRACT: For centuries the dikes in the Netherlands have been protected against wave attack by re-vetments constructed from pitched blocks. Based on experience a size was selected and, if a severe stormdamaged the revetment, heavier blocks were applied. Recent research has however shown that the designstorm surge would cause severe damage to the protection of the dike. A new design philosophy is developedthat assesses the function of the revetment in the entire dike and optimises the thickness probabilistically.
 
 
The storm surge level HW along the Frisian coastat the location is given by a Gumbel distributionwith parameter values
 A
= 2.91m and
 B
 = 0.36m. Thedesign water level with an exceedance frequency of 2.5
10
-4
 per year equals NAP+5.89mThe significant wave height in front of the dike is based on 22 storms, hindcasted with the wavegrowth model SWAN. The conditional mean of thesignificant wave height
 H 
 s
 is described by a linear function of the storm surge level:
 EHHWHW 
 s
|.
g
=
0338The conditional distribution of
 H 
 s
 is assumed to be Normal with the standard deviation equal to0.064mThe SWAN-calculations show that the wavesteepness
 s
0p
, defined on the peak period and thedeep water wave length, is independent of the sig-nificant wave height. The steepness is described by anormal distribution with a mean of 0.036 and a stan-dard deviation of 0.004.Applying the relations given above the significantwave accompanying the design water level of  NAP+5.89m is approximately equal to 2m with asteepness of 0.036 and a peak period of 6s.3
 
THE STABILITY OF THE REVETMENTThe stability of a revetment block is derived con-sidering the equilibrium of the water pressure under the revetment during wave run down and the weightof the block. The resulting formula reads:
 H  D
 s p
=
3
0
cos
αξ
In which:
 H 
 s
:Significant wave height;
:Relative density of the blocks;
 D
:Thickness of the blocks;
α
:Angle of the slope of the revetment;
ξ
:Iribarren parameter.A definition sketch is given in figure 1.
Figure 2: Definition sketch of revetment stability
In the analysis of the block stability the frictionforces between the critical block and its neighboursis neglected, as well as the loss of water pressure dueto leakage through the joints between the blocks.The derived formula must therefore be conservative.A comparison with scale model results confirmsthis. From a statistical analysis of 7 representativetests with irregular waves it appears that the factor 3should be replaced by a factor
 M 
 with a mean valueof 4.06 and a standard deviation of 0.7. The distri- bution type cannot be inferred from the 7 data points but a normal distribution is not rejected. Rewritingthe previous formula in basic variables results in:
 H  D Ms
 s p
=
02
cossin
αα
It should be noted that from the scale model test performed with regular waves, a far higher value of 
 M 
 could be inferred. If one assumes that
 H 
 =
 H 
 s
 afactor
 M 
 with a mean value of 10.9 and a standarddeviation of 5.4 results. A difference that should befurther studied.For a revetment with a slope of 1:3 attacked by awave field with a significant wave height of 2m anda steepness of 0.036 a revetment thickness of
 D
 =0.55m is required according to the formula. This es-timate does however not account for the uncertain-ties in the various variables.
 
 
4
 
RELIABILITY ANALYSIS OF THEREVETMENT UNDER WAVE ATTACK The reliability of the revetment under wave attack is analysed using the following reliability function,that follows directly from the relations given above:
 ZD Ms
 s p
=
sincos
αα
02
And after substitution of the relation gouverning
 H 
 s
:
 ZDHWfH  Ms
 s p
= +
0338
02
.sincos
b g
 αα
The following distributions and parameter valueshave been used in the analysis to account for the un-certainties
Table 1: Distribution type and parameter values of the basicvariablesVari-ableDistr. typeAB
 µ σ
 HW 
Gumbel2.910.362.050.30
 s
0p
 Normal0.000.000.0360.004
 fHs
 Normal0.000.000.000.20cot
 
α
 Normal0.000.000.330.01
 M 
 Normal0.000.004.060.698
 Normal0.000.001.620.02
 D
 Normal0.000.000.70.02
Probabilistic calculations give a relation betweenthe thickness of the revetment blocks and the prob-ability of failure.
Table 2Probability of failure of the revetmentRevetment Thickness
 D
 (m)Failure probability (1/year)0.65
10
-3
0.78
10
-4
0.85
10
-4
0.91
10
-4
Failure of the revetment does however not imme-diately imply failure of the flood defence and inun-dation.5
 
THE SEQUENCE OF EVENTS AFTER REVETMENT FAILUREAfter revetment failure the clay layer behind therevetment is exposed to wave attack. Although clayhas considerable resistance to the attack by cleanwater, the number of hours that it can withstand at-tack by waves armed with sand and gravel (origi-nating from sandy shoals and the filter layer under the revetment respectively) is quite limited.If the clay layer is damaged the erosion process progresses. Here the erosion process of the body of the dike is viewed as a dune erosion process as mod-elled by the DUROSTA-model. A breach in the dikeis formed if the erosion destroys the cross section.The event tree (figure 2) depicts the sequence of events leading to inundation of the polder.For high values of the storm surge level the prob-ability of failure of the dike given revetment failureis approximately equal to 0.35. Applying this valuefor the conditional probability of a breach leads tothe following refinement of table 2.
Table 3Probability of failure of the revetment and of inunda-tionRevetmentThickness
 D
(m)Failure probability(1/year)Probability of inun-dation (1/year)0.65
10
-3
1.7
10
-3
0.78
10
-4
2.8
10
-4
0.85
10
-4
1.7
10
-4
0.91
10
-4
3.5
10
-5

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->