BURIAL OF INSTRUMENTED CYLINDERS IN THE SINGAPORE STRAIT Michael D. Richardson a, Karanam A.P. Roopsekhar b, Eng Soon Chan b, Edward F. Braithwaite a, Monty Spearman a, Paul A. Elmore a, and Shen Linwei b. a Naval Research Laboratory, Stennis Space Center, MS 39529-5004 USA b Physical Oceanography Research Laboratory, Tropical Marine Sciences Institute, National University of Singapore, Singapore Michael D. Richardson, Marine Geosciences Division, Naval Research Laboratory, Stennis Space Center, MS 39529-5004, USA; FAX 1-228-688-5457; E-Mail mike.richardson@nrlssc.navy.mil Abstract: Experiments on the burial of four instrumented cylinders (2-m length; 53 cm diameter) by current-induced scour and by migrating sand dunes were conducted in the Singapore Strait in 2009-2010. Instrumented cylinders were deployed in the eastern Singapore Strait for 6 months in 12-m water depth beginning in September 2009. Modeled and measured semidiurnal tides were nearly identical with maximum currents speeds of 0.87 m/s and a maximum tidal range of near 3- m. Measured wave heights never exceeded 40 cm. Based on sediment mean grain size, water depth, and hydrodynamic conditions the cylinders should have partially buried, primarily the result of current-induced scour. However, the cylinders were not buried and small scour pits that developed around the cylinders decreased the surface area of the cylinders covered with sediment from about 26-28% to 19-22% in 45 days. One of the cylinders was repositioned, perhaps a result of fishing activity, and the current-induced scour process was repeated. Data from pressure sensors suggest the cylinders did not move (settle) relative to the mean sediment-water interface. Except for the fishing incident, sensors in the cylinders measured little change in cylinder roll, pitch, or heading. The difference in the measured and modeled burial is probably the result of higher than predicted values of the Shields parameter as a result of 15-30% silt and clay in the sediment. Sediments at the second site, a narrow channel north of Sentosa Island, were gravellysand with less than 2% silt and clay. Migrating sand dunes were responsible for the complete burial of one of two cylinders deployed at this site and the strong tidal currents caused several episodes of change in roll, pitch and heading in the other. This behavior is typical of scour burial where the cylinder pitches and then rolls into scour pits created by enhanced sediment transport from the turbulent flow. Keywords: current-induced scour, burial, sediment transport
1. INTRODUCTION Naval mines have been used in littoral warfare for over 200 years. They provide a cheap and effective way to significantly alter naval operations. Bottom mines in shallow water are particularly difficult to detect and classify when they are partially or wholly buried. The Office of Naval Research and Naval Research Laboratory executed an experimental and modelling program to determine when and where bottom mines are likely to bury [1]. As part of that program, acoustic instrumented cylinders were developed to measure burial in shallow sandy environments [2]. Wave-induced scour burial experiments were conducted in the northeastern Gulf of Mexico, off Martha s Vineyard, and near Hawaii. These experiments provided environmental and burial data used to develop and validate physics-based burial models [3]. In this paper we present a preliminary evaluation of burial of those cylinders in the Singapore Strait. The joint Naval Research Laboratory and National University of Singapore experiments were conducted in 2009-2010 at two sites where scour from tidally-induced currents was expected to be the dominant burial mechanism. The instrumented cylinders are mine-like shaped, blunt-end, bronze cylinders with a 0.53- m diameter and a 2.03-m length. They were designed to measure the actual burial and the environmental processes responsible for that burial. The cylinders utilize acoustic transducers to measure burial and scour, localized flow rates, and sediment size and concentration in the water column. In addition, the cylinders contain sensors for measuring orientation, bottom pressure fluctuations, water temperature, and motion resulting from sediment liquefaction or rolling into scour pits. Hydrophones are used to measure acoustic energy impinging on the cylinder s surface from search and classification sonar. Data from pressure sensors provide a look at local conditions of the environment, such as significant wave height, wave period, and tides. In areas where time series of tidal heights are available, averaged pressure time series from the cylinders can be used to calculate an estimated burial with respect to the sediment water interface. The acoustic sensors are used to estimate percent of the cylinder surface covered with sediment. 2. DESCRIPTION OF THE SINGAPORE EXPERIMENTS On 1 September 2009 four instrumented cylinders and an instrumented platform were deployed in 12-m water depth in the eastern part of the Singapore Strait near the Changi International Airport. The instrumented platform was recovered on the 12 January 2010 and the instrumented cylinders on 19 April 2010. A second experiment was conducted in the channel north of Sentosa Island (21 April - 10 August 2010). Only two cylinders were deployed at 6-m water depth. Burial conditions of the cylinders were occasionally monitored by divers during both experiments. Hydrodynamic conditions were modelled using the Tropical Marine Hydrodynamic Model (TMH) developed by the Tropical Marine Sciences Institute of the National University of Singapore [4]. The instrumented cylinders provided time series of orientation (roll, pitch and heading) and burial calculated as surface area of the cylinder covered and height of the cylinder above the surrounding seafloor. The hydrodynamic forcing (tidal currents and waves) was measured using sensors on the instrumented platform and pressure sensors within the cylinder. These time series were compared to modelled time series and both were used to predict burial.
3. ENVIRONMENTAL MEASUREMENTS 3.1 Hydrodynamic conditions Hydrodynamic conditions were measured using an upward-looking Acoustic Doppler Current Profiler (ADCP) and a downward-looking acoustic Pulse-Coherent Acoustic Doppler Profiler (PC-ADP) both attached to the instrumented platform, a bottom-mounted quadruped. Time series of depth-averaged current velocity, tidal elevation, near-bottom current velocity and significant wave height and period were collected between 2 September 2009 and 12 January 2010. Time series of the total suspended sediment concentration and bed-stress were calculated from data measured using the PC-ADP. The tidal range was between 1-3 m with the greatest tidal range occurring during spring rather than neap tide. Significant wave heights in this protected area averaged 14 cm and only rarely exceeded 35 cm. Wave period ranged between 4.0 s and 10.0 s (mean 5.1 s). Mean depth-averaged tidal currents peaked at 0.88 m/s (mean 0.27 m/s) with slightly higher speeds on the ebb or easterly flow. Although the measured and modelled (based on the TMH model) mean tidal range and mean depth-averaged tidal current were nearly identical, the measured values of peak tidal currents were up to 25% greater than the modelled tidal currents. The highest concentrations of near-bottom suspended sediments coincided with the periods with the highest values of depth-averaged bottom current velocity (or values of shear stress). The values of total sediment concentration were thus highest during the ebb or easterly flow. The mean values of sediment concentration did not vary more than a few mg/l (20-28 mg/l) over individual tidal cycles. Significant wave heights were well below that needed to scour sediments from around the cylinders and analysis will therefore concentrate on scour caused by the tidal currents. The threshold depth-averaged current velocity (0.44 to 0.47 m/s) necessary to initiate sediment motion [5] was often exceeded during both ebb and flow conditions. The threshold bed stress (0.19 to 0.27 N/m) needed to initiate sediment motion was only exceeded during the ebb or periods of easterly tidal flow. The tidal fluctuations and time series values of significant wave height and period calculated from the pressure sensors on the instrumented cylinders were similar to the same time series collected from the ADCP. The wave data, not repeated here, suggest that wave-induced scour around the instrumented cylinders was not significant enough to induce burial for the entire 7½ moths of deployment in the eastern part of the Singapore Strait. The bottom-mounted quadrupod was not used in the second deployment and current data were only collected during short observational trips with a hull-mounted ADCP. The maximum tidal range was greater than 3 meters measured at a nearby tidal station and using the pressure sensors on AIM. Significant wave heights calculated from pressure series from sensors in the cylinders were all less than 10 cm which are much below that required to scour sediment from around the cylinders. Any burial must therefore be caused by current-induced scour or by migrating sand dunes. It is unfortunate that no time series (measured or modelled) of depth-averaged tidal current exist for the site. It can be assumed from diver observations of scour and fill around the cylinders and from the 3-m tidal range in this narrow (200-300 m wide) 2 km channel that strong tidal currents are present.
3.2 Surface sediment description Sediment samples were collected by divers within a meter of the location of each cylinder. Grain size distribution was determined by dry-sieving the gravel and sand fractions and by pipette methods for assaying the silt and clay fractions. Average grain density was determined from dried samples with a pycnometer. Sediments collected from the eastern end of the Singapore Strait contained a diverse mixture of gravel (mean 19%), sand (58%), silt (18%), and clay (3%). Average grain density was 2672 kg m -3.Median grain size in the upper 16 cm of sediment ranged from 0.26 to 0.53 mm. The median grain size of only the gravel- and sand-size particles collected from this site was 0.6 mm which can be classified as coarse sand. Sediments collected from the Sentosa Channel were poorly-sorted gravelly-sand with little variation with depth below the sediment surface or between the two sites. Median grain size (0.87 to 0.97 mm) were equivalent to coarse sand with roughly 75% sand-size particles and 25% gravel-size particles. Average grain density was 2770 kg m -3. 4. MOVEMENT AND BURIAL OF THE INSTRUMENTED CYLINDERS 5.1 Eastern Singapore Strait Various sensors within and on the surface of the cylinders were used to measure the movement and burial of the cylinders. Not all of the sensors functioned for the entire 7½ month deployment period. For two of the cylinders (AIM1 and AIM3), continuous burial and orientation data were only collected between 2 September and 10 October 2009. Both AIM2 and AIM3 provided orientation data for most of the deployment period, but burial data from the acoustic sensors in AIM2 ended on 20 October 2009 and 31 December 2009 for AIM4. In spite of these difficulties, a general picture of burial of these cylinders can be derived (Fig 1). AIM1 sensors provided continuous data for cylinder orientation and surface area exposed/covered between Day Dates 245-283 (2 September to 10 October 2009). Except for a minor change in orientation 12 hours after deployment the cylinder did not move. The heading remained 205, roll +4.5, and pitch angle -0.3. The surface area of the mine covered with sediment decreased from 28% to 22%. Sporadic orientation data collected after 10 October 2009 suggests a change in heading (205 to 214 ), roll (+5 to +22 ) and pitch (-0.5 to +1.5 ) between 18 and 24 December 2009. Minor changes in orientation continued until 18 April 2010 when the cylinder was recovered. The final cylinder orientation was 212 heading, +2.5 pitch and +34 roll. AIM2 sensors provided continuous orientation data for the entire deployment period and continuous acoustic data for surface area covered until 20 October 2009. The heading (range, 201 to 204 ) and pitch (range, +0.5 to -1.5 ) of the cylinder varied little over the 7½ month deployment period. The roll angle changed from -40 to -32 from the beginning of the experiment until 22 December 2009. After that date the roll angle was unchanged. The surface area of the mine covered with sediment decreased from 27% at the beginning of the experiment to 19% on 20 October 2009.
AIM3 sensors provided continuous orientation and burial data from 2 September until 10 October 2009. The heading (359 ), pitch (-1.8 ), and roll (1 to 2 ) varied very little over that period. The surface area of the mine covered with sediment decreased from 27% at the beginning of the experiment to 21% on 10 October 2009. Fig. 1: Measured cylinder burial expressed as percent surface area covered. Time expressed as Days after 3 September 2009. AIM4 sensors provided continuous orientation data until 28 March 2010 and continuous burial data until 31 December 2009. An abrupt change in cylinder orientation and burial state occurred within a 20 minute period on 22 October 2009. The cylinder changed heading from 28 to 290, pitch from -0.15 to +0.4, and roll from +2 to +3.5. The surface area of the cylinder covered with sediment also abruptly changed from 21% to 28%. This abrupt change in the exposed state on the cylinder over a 20 minute period is not likely the result of scour burial, but instead caused by some sort of man-generated event which repositioned the cylinder. Prior to the event on 20 October 2009, the cylinder recorded very little movement with less than 0.2 change in roll, pitch or heading. After the event, changes in roll, pitch and heading were less than 1.5. The surface area of the mine covered with sediment decreased from 26% at the beginning of the experiment to 21% on 22 October 2009. The cylinder was reset to 28% burial by the man-induced repositioning. Surface area covered then decreased to 22% by 6 December 2009. A slight increase in burial from 22% to 24% was recorded between 17-31 December 2009. Changes in the height of the cylinders above the mean ambient seafloor were estimated based on a comparison of tidal height time series measured at the mine surface (six sensors record pressure fluctuations at the surface of each cylinder) and time series of tidal height measured at fixed points. The fixed point tidal heights were measured at a nearby tidal station and using the ADCP mounted on the instrumented platform. Preliminary analyses suggest no change in height of the cylinders relative to the sediment surface. Reduction in percent of the surface area of the cylinders covered with sediment is therefore probably related to local scour around the cylinders rather than burial relative to the sediment surface.
5.2 Sentosa Island Only two instrumented cylinders were deployed at 6-m water depth in the channel north of Sentosa Island (21 April 10 August 2010). The acoustic sensors that record percent surface covered/exposed were operational on AIM3 until 20 June. Orientation and pressure sensors recorded data on AIM2 until 10 August and on AIM3 until 6 June. The divers noticed that AIM2 was fully buried, and repositioned the cylinder on 24 June. AIM3 exhibited considerable variations in orientation between 21 April and 16 June 2010. Within 15 hours of the initial deployment (heading 180, roll angle 0, pitch +1 ), the cylinder pitched to +5 and back to +1 and rolled to -38 with little change in heading. Eleven days later, AIM3 pitched between +1 and -3, rolled back to +5 and changed heading (180 to 192 ). This process was repeated during May 15-17, and again on 30 May. The total range of pitch angle was 8 ; the range of roll angle was 135 ; and the heading changed from 180 to 212. This behavior is typical of scour burial where the cylinder first pitches then rolls into the surrounding scour pit and changes heading where the long axis of the cylinder is parallel with the current. Roughly 30% of the mine surface was covered with sediment over the month long period that the mine acoustic sensors operated. The percent of the cylinder buried (sensors covered) did not vary substantially over that period in spite of the cylinder rolling more than 90 degrees. The scalloped effects of scour and infilling were most pronounced when tidal amplitudes were at their greatest during Spring Tides. The scour and fill events only occur diurnally in spite of the semidiurnal tides. Maximum scour depths (increase in distance of the transducer face to sediment surface) were 10-15 cm. After the full 111 day experiment diver observations found the cylinder to be buried 20-30%. This is in general agreement with the acoustic observations during the first month of deployment. The presence of relict coral may have restricted the final cylinder burial. Except for when the divers repositioned AIM2, the cylinder did not change orientation as much as AIM3. During the first three days after deployment the cylinder changed roll angle from -2 to -27.5 and back to -3.5. During that same period the heading changed from 215 to 205 and the pitch angle from -2 to -7. Except for a minor change in orientation on 16 June, the cylinder did not move until the diver repositioned AIM2 on 24 June. Divers observed a fully buried mine (actually the mine was buried about 30 cm below the sediment surface) before the mine was moved. This would suggest that the cylinder was buried by a migrating sand dune. After repositioning, the cylinder went through a few periods of change in pitch (range, -9 to -14 ) and roll (range, -74 to -84 ) without a significant change in heading (175 to 180 ). The initial pitch suggested that the cylinder was deployed on the face of a sand dune. There was little cylinder movement after 21 July. Final diver observations show the cylinder buried 90% on one end and 20% at the other end. Again this suggests burial by migrating sand dunes. 5. CYLINDER BURIAL PREDICTION Predictions of burial relative to the ambient level of the seabed were made using the implementation of the HR Wallingford scour model described in Trembanis et al. [3]. It is assumed that potential burial was caused by current-induced scour as the values of
significant wave height were well below the threshold needed for sediment transport. Two time series of depth-averaged tidal currents are used: the predictions from the Tropical Marine Hydrodynamic Model (TMH) and in situ measurements based from the ADCP mounted on the quadrupod. The time series of depth-averaged tidal currents calculated from the ADCP were not continuous, and where gaps existed, the time series was supplemented with data from the TMH model. The grain size distribution varied between cylinder locations, and the following values of medium grain size were used for making the burial calculations (D 50 = 0.533 mm, 0.352 mm, 0.265 mm, 0.476 mm) for AIMs 1-4 respectively. The semidiurnal peaks of the in situ measured values of depth-averaged current velocity were higher than peak values predicted using the TMH model (Figs. 2 and 3). The result is greater predicted burial based on the in situ measured depth-averaged current velocity. In either case, the percent surface area buried (28-40%) between days 40-50 is greater than that measured using the cylinders (19-22%). Predicted surface area buried is calculated from the depth of burial assuming a level sediment surface. This accounts for some of the difference between measured and predicted burial. However, burial models predict a 20-34% burial relative to the height of the cylinders below the mean seafloor. Based on data collected using the cylinders, this burial did not occur. Fig. 2: Percent burial of the four cylinders based on depth averaged current velocity from the TMH model. Dates are days after 3 September 2009. The bottom four lines represent percent height of the cylinders buried below the seafloor; the top four lines represent percent surface area of the cylinders covered. Experiments in laboratory flumes have shown that the addition of cohesive material to sandy sediments increases the erosion threshold and decreases the rate of sediment erosion when critical shear stress in exceeded [6, 7]. The effect of the addition of cohesive material to the threshold and rate of erosion enters the prediction of burial [3] via its effects of the hydraulic roughness on the critical bed stress. It would therefore appear that median grain size is an insufficient parameter to predict scour burial where a significant percent cohesive mud is mixed with an otherwise cohesionless mixture of sand and gravel
Fig.3: Percent burial of the four cylinders based on depth averaged current velocity from using the bottom mounted ADCP. Dates are days after 3 September 2009. The bottom four lines represent percent height of the cylinders buried below the seafloor; the top four lines represent percent surface area of the cylinders covered.. 6. ACKNOWLEDGEMENTS This work was support by the US Office of Naval Research (Brian Almquist, Program officer) and the Singapore Defense Science and Technology Agency (Lim Cheah Siang, Program officer). REFERENCES [1] Wilkens R.H. and M.D. Richardson. Special Issue on Mine Burial Processes. IEEE Journal of Oceanic Engineering 32(1):1-283, 2007. [2] Bradley, J., S. Griffin, M. Thiele, M.D. Richardson and P.D. Thorne. An acoustic instrumented mine for studying subsequent burial. In: Wilkens R.H and M.D. Richardson (Editors) Mine Burial Processes. IEEE Journal of Oceanic Engineering 32(1):64-77, 2007. [3] Trembanis, A.C., C.T. Friedrichs, M.D. Richardson, P. Traykoviski, P.A. Howd, P.A. Elmore and T.F. Wever. 2007. Predicting seabed burial of cylinders by wave-induced scour: Application to the sandy inner shelf off Florida and Massachusetts. In: Wilkens R.H and M.D. Richardson (Editors) Mine Burial Processes. IEEE Journal of Oceanic Engineering 32(1):167-183. [4] (http://www.porl.nus.edu.sg/main/research/hydro-model) [5] Whitehouse, R. Scour at Marine Structures: A Manual for Practical Applications. Thomas Telford, London, 198 pps., 1998. [6] Mitchener, H., H. Torfs and R. Whitehouse. Erosion of mud/sand mixtures. Coastal Engineering 29: 1-25, 1996. [7] Panagiotopoulos, I., G. Voulgaris and M.B. Collins. The influence of clay on the threshold of motion of fine sandy beds. Coastal Engineering 32: 19-43, 1997.