Physical model tests on new armour block Crablock for breakwaters to come to preliminary design guidance

Abstract

In the design of rubble mound breakwaters, nowadays single layer systems using concrete armour units have become more common practice compared to conventional two layer systems. However, after the introduction of the accropode in eighties, a small number of single layer armour units have been developed over the years; for example core-loc, A-jack, xbloc, accropode-II, cubipod and core-loc II. Recently, a new concrete armour unit called crablock has been invented and applied as one layer system in one damaged breakwater at Al Fujeirah, UAE. In contrast to other existing monolayer units, the shape of this unit is symmetrical which allows placing both in uniform and random pattern. As the crablock unit is still under development, no design guidance exists yet for this concrete armour unit. To use crablock as monolayer system the preliminary design guidance on placement of crablock, stability and wave overtopping are required. This led the present research to investigate the placement pattern, packing density and wave overtopping over slope to come with first design guidance for the application of crablock. It should be mentioned that stability of the crablock against wave attack was also looked at, but that will be reported by Mr. André Broere, an MSc-student at Delft University of Technology. The present research was based on a literature study, small scale dry placement tests and small scale hydraulic tests in a wave flume. Regarding to the review of literature studies on the existing single layer units and crablock, set up of dry placement tests and flume tests have been made for this experimental research. Dry placement tests as well as 2D wave flume tests were carried out at the Fluid Mechanics Laboratory of the Faculty of Civil Engineering and Geosciences at Delft University of Technology, Netherlands. Both placement and hydraulic tests were executed with the use of small units. In total 14 independent placement test series were executed to investigate the placement pattern, placing grid and packing density of crablock. All the tests were performed on a 1:4/3 slope with the use of random and uniform placement in a rectangular as well as in a diamond shaped placing grid. Results of placement tests showed that uniform placement of crablock is achievable with the use of relatively small and smooth under layer in a rectangular placing grid. The performance of regular placement using a conventional under layer with size 1/10th to 1/15th of the size of the armour layer was not so satisfactory. Furthermore, it was found that uniform pattern was hardly reachable in a diamond-shaped grid with conventional under layer. However, irregular placement of crablock was certainly easier to construct and possible to place with higher accuracy compared to uniform placement in a diamond grid. It should be noted that all the tests using a conventional underlayer were performed without the fixation of first row due to the difficulties in placement with model crablock units. If this can be fixated by designing dedicated toe units (both in rotation and location) it may perform better. Still, the large underlayer makes it difficult to place uniformly. Finally, two preferred placing patterns appeared from the placement tests, a regular pattern in a rectangular grid using a relatively small under layer and a random pattern in a diamond grid using a conventional under layer. For the determination of wave overtopping, altogether 14 different test series were performed in a wave flume. In this research, two constant spectral wave steepnesses of 0.02 and 0.04 were tested together with two different orientations of units, two different placing grids and four different packing densities. The preferred placing patterns were constructed in a wave flume on a modelled breakwater cross-section in front of the sloping foreshore of 1:30. Each test series was comprised of number of sub tests for specific wave height and period. In each test series, wave heights and periods were continued to measure until the failure of armour slope. The armour layer was reconstructed prior to start of each test series. The test results of 2D flume tests showed that wave overtopping over crablock slope did not vary much between the different test series with same wave steepness. Nevertheless, it was observed that wave overtopping is little bit higher for longer wave period that means for low steepness compared to short period. Based on test results, it was also found that overtopping behaviour does not really change with the change in packing density and also with different placement pattern of armour layer. Regarding to the comparison of relative overtopping rate over crablock armour between test results and empirical prediction, it was found that that empirical equation with assuming roughness factor of 0.45 underestimate the measured wave overtopping over crablock. However, the comparison between the test results on overtopping percentages and estimation by EurOtop (2007) proved that percentage of waves overtopped over crablock can be well predicted by using empirical formula. Furthermore, the measured wave overtopping over crablock slope was found slightly higher in comparison to CLASH (2004) results on other single layer units. This variation was mainly observed for the test results with low wave steepness. Besides relatively low wave steepness, most of the tests on crablock were performed with relatively longer wave periods in comparison to CLASH (2004) which was also one of the triggering factor for higher overtopping over crablock slope compared to CLASH (2004). Moreover, the use of sloping foreshore (1:30) instead of horizontal one by CLASH (2004) might also influence the overtopping behaviour. The 1:30 slope changed the shape of the waves and the waves at the structure toe showed a clear increase in velocity of the wave crest (near or at breaking) The resulting wave overtopping over crablock slope was also compared with the overtopping over xbloc slope measured by DMC (2003). From the comparison, it was found that wave overtopping over crablock is significantly lower compared to xbloc measurements by DMC (2003). Based on the comparison of wave overtopping over different armour slope with and without Ursell parameter, it was recognised that use of the Ursell parameter may explain wave period differences in some cases, but introduces also unexpected differences.