Press Room - CFD improves traffic safety on waterways

CFD improves traffic safety on waterways

Transverse flows from lateral water discharges into waterways, for example those caused by storm water outfalls, may cause passing ships to drift. In order to provide traffic safety, such effects need to be restricted so that outfall structures create well-balanced flow fields of low intensity.

Until now, the common practice for developing outfall structures has been to carry out very time- and cost-consuming physical model tests, but in future numerical simulations could help to reduce this overhead. An outfall structure has been developed by the University of Hannover using physical model tests; it will be put into practice this year in the Main Danube Canal in the city of Bamberg (Germany). In order to judge the capabilities of numerical simulations in this application, the structure has also been studied using STAR-CD.

Figure 1 shows the 36m-wide outfall structure including features such as the pressure pipe, overfall weir and submerged wall, as well as the topographic situation 200 metres up- and downstream, which were all transposed into the numerical simulation. A mesh of about 80,000 cells was generated, with edge lengths between 0.05m and 3m. The water/air free surface was modelled using the VOF method and the standard high Reynolds-number k-e turbulence model was used. The calculation time was about 20 hours on a SGI Origin 200 Workstation.

Figure 2 shows the weir overfall in the outfall structure for the initial numerical simulation. It is abundantly clear that the well-balanced weir outfall seen in the model tests is similarly reproduced, in that the discrepancy between calculated and measured overfall height never exceeds 6%. A similarly low divergence between physical model test and numerical simulation can be recognised when comparing the calculated and measured flow field within and in the vicinity of the outfall structure at half-water depth. This is illustrated in Figure 3, which shows particularly good agreement in both the flow velocities and directions in the outfall structure itself.

For similar investigations in the future, the use of more numerical simulations is recommended in order to reduce the amount of very time-consuming and costly physical model tests. Note, however, that physical model tests cannot yet be replaced completely, since although the current simulation methodology makes it possible to develop an overfall structure from a purely fluid-mechanical point of view, it does not take into account the influence of ship traffic on the waterway. This is because the interaction of a moving object and its environment is not presently implemented. Research is under way to close this gap.