11th International Conference on Hydrodynamics (Nanyang Singapore)

October 2014




Report of the 11th International Conference on Hydrodynamics (ICHD) held at Nanyang technological university, Singapore October 2014.

This conference began on Monday 20th October. The chairman of the organising committee, Professor Yousheng Wu  welcomed all those present, promising that there were a big number of papers to be presented and that the standard very high.

The first keynote speaker was Prof Odd Faltinsen who spoke on The Hydrodynamics of marine and offshore structures. He stressed the importance of the structures in terms of food transport, energy supplies, and oil spillage. He mentioned the effect of Kelvin Helmholtz instability which can produce springing and whipping of boats. Prediction of these forces is difficult and although in theory they are two dimensional there can be 3D effects. The forward speed of boats is important but often neglected. Slamming of water against a boat structure is important. For offshore engineering nonlinear process need to be predicted such as the ringing and springing effects of offshore waves. The use of a 2D numerical wave tank was mentioned. Prismatic tanks have been built and the results of experiments were shown. The effect of wave sloshing is a non linear problem. Ship design will affect the level of Co2 emissions. Prof Faltinsen also mentioned the use of solar panels and fuel cells in order to reduce the use of fossil fuels. He then described the effects of choosing the best route across the ocean to reduce fuel consumption, also a voluntary speed reduction. Food supplies from the oceans only make up 2% of world production. We need to study the hydrodynamics of fish cages. There have been problems of wave damage and deformation of the net, keeping the fish inside and bio pollution. At Crosing Sognefjord research on submerged floating tunnels is already underway, and may prove more economic and equally as effective as traditional bridges. Finally he mentioned estimates of wave loads in extreme events.

The second keynote talk was by Prof Tao Jiannhua, who spoke about Modelling for eco-environmental characteristics of coastal waters. Prof Jianhua emphasised the importance of coastal areas for Primary Productivity: chlorophyll A is estimated. using two models have been produced. The first is a deterministic one, the second non-deterministic. The Bohai Sea in China (Bohai Bay) was her study area. The exchange of water in the bay as well as the longshore current were both modelled and measured. The results were compared favourably. Another aspect of her studies was the transport of pollutants. The wave current-long shore current model was used, based on a topographic survey of the bay. Numerical experiments were also carried out and the results shown as a video film. Field data were collected on bacterioplankton and benthic bacteria. The discharge of  ammonia and nitrate were obtained from the municipal sewerage works. There was a good comparison of the measurements of local data and model results. Further work involved the use of a hydroinformatics based model, based on probability theory. The results are validated from data collected by the Tianjin Oceanic Environmental centre. Some results of remote sensing of the bay were also shown. Although the distribution of pollution can be modelled by conditional fluid dynamics method the processes are not fully understood.

Dr Qianxi Wang with co-authors gave a thought provoking talk on Bubble dynamics in a compressible liquid. The shockwaves and collapse of bubble caused a pressure of up to 1 Giga bar. The bubble jet was produced at the end of the bubble collapse. The effect on marine structures was considerable: this includes cavitation damage to propellers, turbines, and fluid machinery. The applications of this research include drug and gene delivery in human tissue, such as using ultrasonics to cause the collapse of injected blood particles. Modelling the process of bubble expansion and then collapse has to take account of the oscillation of bubble size. There are large Reynold’s numbers – turbulent flow – and weak vorticity in the boundary layer. The results of computation compare well with observations in the first and second cycles of oscillation.

Paper number 103 by Enzo Marion and co-authors was on the Aero-hydroelastic instabilities on an offshore fixed bottom wind turbine. Enzo emphasised the importance of collecting data on site and the use of second order theory. Recently the effect of steep waves which do not break has been a resonant vibration in the turbine tower which could cause serious damage. Non breaking waves have a complicated structure from linear to non-linear. A 5MW Turbine Model where the structure was 90m high and 126m diameter blades, where the local water depth was 20m, and a wind speed of 33 m/s was produced. The results showed a total tower top displacement, with the higher values with non-linear waves. For breaking waves the resonant vibrations can cause more displacement of the tower.
Paper number 157 described the use of Drilling tenders for offshore work, by Aziz Merchant and co-authors. Different design of tender were described with the Semisubmersible Drilling Tender being the best. The design with six submersible columns was found to work better – more stable than other designs, while variations on the design allow drilling to be carried out in +1000m deep water.

Paper number 205 was on Experimental and numerical analysis of steady state sloshing in a rectangular tank fitted with a slatted screen, by Renza Firoozkoohi and Odd Faltinsen. The effect of the slatted screen was to reduce the sloshing frequencies in the test tank. Measurements of the non-linear effects were made at 1cm from the tank screen. The tank height was 0.23m high. Several designs of slat were tested including coarse, medium, and fine. The results were compared with numerical calculations. Having a coarser grid causes new harmonics to be observed; some are even enhanced. With a solidity ratio of >0.5 the natural frequency of the tank will change leading to secondary resonance.

Paper number 94 was a related paper in that The effects of fuel tank baffles in order to stop sloshing was described by Xuekang Gu. Model studies at a scale of 1:50 were performed in 3 model test tanks. The results showed that at 30% of the tank height there was pitching of the vessel, while higher depths produced lower effects. When the orientation of the baffles was changed, it was found that horizontal obstructions caused severe splash as compared with a smooth tank. A transverse bulkhead reduced the sloshing in the tank.
The start of the second day began with an informative talk on Catastrophic tsunamis and hurricanes in the last decade by Philip Lui who mentioned events in 2004, 2010, 2011, 2014 (tsunamis) and 2005, 2012, and 2013 (hurricanes). The 2004 event in the Indian Ocean caused the death of about 23000 people in 14 countries. The waves were 30m high. In 2005 hurricane Katrina, a category 5 event, turning to a category 3 event in SW Louisiana. Levees failed in New Orleans and many people drowned with water reaching 10-19km inland. The damage was estimated at $108 billion dollars. Prof Liu emphasised the different timescale of these events – 10 minutes in the case of the earthquake, but whose effects took several hours to reach land, as compared with about one week for a hurricane. The 2011 Tohuku earthquake was located 70km east of Oshika peninsula produced waves 40.5m high and travelled 10km inland. Philip then described the generating mechanism of earthquakes with reference to the subduction zone. The plate boundary moves several meters each year but this varies according to the stresses and then rapid movement takes place. The plate will then rebound. The slip velocity, width and length of the fault line are hard to predict. In 1964 there was a vertical displacement of 11.3m. The Sumatra earthquake was similar.

The resulting waves in open water are about 1m high with a wave length of about 250km. Deep wave pressure sensors measure the long wave height. The wave period is about 30 minutes. Towards the coast the wave is huge – like a bore. The water is dark in colour and damage awful. The debris causes more even more damage. In Japan the nuclear power plane was caused by flooding.

Storm waves and surge generation occur with hurricanes. In 2012 Hurricane Sandy produced pressure gradient changes with time. The waves fields were complicated and non-linear. In coastal areas wing waves came frequently.
Modelling the waves involves the use of hurricane, wave, and surge models. Shallow water equations have to be applied according to the timescale of the event – 30 minutes for tsunamis, several hours for storm surges. The detailed forecasting of the waves is difficult. The propagation of the tsunami  is more straightforward. In conclusion he stated that wind-wave models need improvement as did near shore process description. The application of these studies is of vital international importance in order to save lives.
The response of hydrodynamics to Caofeidian Project (China) in 2012,  was described by Jianwen Qi on behalf of his seven co-authors. In Bohai Bay a land reclamation project had been completed. Two islands south of the mainland have been linked together with the mainland. The project was completed in 10 years from 2002 making an extra 310 km2 land area. The Delft 3D model was used to estimate the effects of the project. This model solves unsteady flow equations; it covered an area of 107km x 76km and had 129x185 grid boxes. The model was verified from the data of two waves measuring stations. The effects on the tidal current were shown by a reversal of the flow eastward in the floods tide and westward in the ebb tide. Flow velocities are now higher in front of the foreland and the west side of the islands to the extent of 10-40% and 10% respectively. On the west side velocities are reduced, and also on the west and east branches of the Laolonggou  where sedimentation may take place in the future.

The effects of propeller wash on the distribution of sedimentation around Marina Bay cruise centre in Singapore was well described by Mamta Jain. There was considerable re-suspension in the harbour and the aim was to quantify the effect on the distribution of natural sedimentation. An empirical method was needed to simulate the hydraulics of a propeller jet. The velocity distribution effect of moving ships and the influence of the rudder clearance was also required. Observations were made of the rpm of the propeller and the effects of the rudder under clearance distance. It was found that diameter of the propeller has the least effect on sedimentation.

Field measurements also included water sampling along transects behind the vessel at agreed safe distances. Tug boats affected the results during turning and berthing operations. Prop reverse also had an effect. Concentrations of sediment up to 5000 mg/l were observed. The total traffic of ship movement was estimated using the Australian Identification System. A variety of vessels was studied and six months’ data were gathered to estimated the sediment concentration overall. The hydrodynamics of sediment transport were estimated using DHIMike 21 model.
Over a period of 3 years the marina bed changed by 0.4m and this must be attributed to natural processes as well as ship movement.

Prof Joseph Lee gave an interesting description of The fluid mechanics of chlorine disinfection for the Hong Kong Harbour treatment system. A 10% solution of chlorine was used to kill the bacteria. The discharge of effluent the outlet varied from 10-20 cumecs. A concentration of 10-20 ppm chlorine after mixing was produced, giving a rate of dilution of 10,000. The chlorine concentrate is applied using jets. Sometimes it does work properly. An upward discharging jet of chlorine can help since there will be better mixing. Cross dominated jets can also help but the mixing ratio is only about 10% which is not high enough. Inadequate mixing leads to wastage of chlorine. There are many users for Victoria Harbour such as fish farms, toilet flushing, kayaking, and surfing. To help users know what is going on a beach grading system ahs been produced, from 1 to 5 where 1 = best and 4 = worst with concentrations of E.coli ranging from <24 counts per 100ml to >610 counts per 100 ml. The measure is the geometric mean of five measurements. Now there is real time data for short term decision making by the users of the harbour, although it takes 24 hours to process the data. For bathers the count must be <610 counts per 100ml. The outflow from sewage treatment works has to be modelled. The project for this work is called WATERMAN.  Data driven models are used; field studies and point data are used for calibration. Environmental factors are also measured such as solar radiation, wind speed, temperature, and salinity. The decay rate of E.coli is a function of solar radiation, temperature and salinity. The full range of conditions have to be dealt with in order to predict E.coli levels. A hydrologic model is used which includes bacterial decay modelling and turbulent mixing. Data are also needed on wind direction and tides as well as rainfall which can cause dilution of the E.coli. The level of accuracy of E.coli on 20 beaches ranges from 75-90%. The test was carried out every day for three years.

Prof Fred Stern gave a tour de force on the Recent progress in computational fluid dynamics of naval architecture and ocean engineering.  It was based on nearly a lifetime’s work. CDF offered new insights into design technology. Simulation base design used experiments and computational uncertainty analysis. Model test are only done when necessary because of the high cost. CDF Iowa was developed about 20 years ago: version 4.5 is the main method used today. The prediction of resistance on boats is an important are of research since there is a low level of accuracy at present. He made the point that 90% of all simulations are being done with less than 10 million data points. Propeller design needs improvement which means resolving all of the forces. Naval architecture (ship design) can be improved by getting a grip on all of the forces on a static ship and then adding in motion. The Iowa towing test tank is 3m x 3m and 100m long. In Fred’s opinion you cannot do enough validation of the models. Other modelling done included using the Delft cata-marang, free running course keeping, turbulence modelling, ship to ship interaction, advanced hull forms and the effect of wave slams.
For wave run up on ocean structure the job is to resolve the processes at all scales, down to the single bubble. The computer time for this level of detail is very high. There are vortices: Fred reminded the audience of the Japanese print of breaking waves with their finger like projections (vortices). The design of the optimal ship’s stern was a matter for the test tank although theoretical calculations can also be made. He concluded by saying that the future lay in simulation technology, an approach that industry has yet to adopt.
My contribution was The application of coarse sediment transport hydraulics to the estimation of the peak flow of catastrophic floods. This was based on previously published work on the Cheviot Hills (1893), Boscastle (2004) and  North York Moors (1857) floods where evidence from boulder deposits has been used. The application of the Bagnold equations for unit stream power and critical stream power was suggested and this was a novel approach. The results for the three floods had been supported by hydrograph analysis which included and estimate of how dry the soil was just before the flood. The results for the Cheviot Hill event were somewhat lower than had originally been suggested in 2004 but still nearly 30 cumecs per km2 for a catchment area of about 11 km2. In all cases the flow was subcritical and velocities in the range 3-4 m/s. The importance of a sensible assessment of historic – both recent and old – floods cannot be emphasised too much.

This conference gave me an opportunity to meet people working in the related field of hydrodynamics. What struck me most was the use of both field, test tank, and theoretical approaches to the same problem. The arbiter is, of course, the full scale model or in my case the river itself. One aspect which did cause concern was the apparent lack of estimating the rarity of what was seen as rare events. This could have an effect on project design. For example after the conference there was a tour to the Marina Barrage in downtown Singapore. The Tour Guide was confident that the design level of the barrage, which was 3m above the highest known tide, would be good enough for all conditions in the future. Yet what might happen in this part of the globe with a major offshore earthquake? This, like many other questions remain to be answered in the future.
I am very grateful indeed for the financial support from the Royal meteorological Society which enabled me to attend this conference.

Colin Clark, 24 November 2014