European fullscale flow research and technology (EFFORT)

No wakefields are available during the design of propellers for inland water vessels. What can be achieved when with use of full-scale CFD calculations the wakefield will be calculated. What consequences will this have for the propeller design? Which improvements for the inland vessel design and propeller design can be expected? To solve these questions full-scale CFD calculations will be carried out for a typical inland water vessel. Propeller designs will be made based on the traditional approach and based on the calculated wakefield, the differences in pressure pulses and cavitation extent will be calculated, to demonstrate the propeller improvements. The aftship shape will be improved to demonstrate the potential gains that can be achieved using this tool.

Hopper dredger operates most of time in shallow water. During this operation the wakefield changes significantly. This case study assesses the changes in the wakefield with use of full scale CFD calculations and will demonstrate the differences in propeller design if shallow water is taken into account.

The objective of this Result was to provide a database with measured flow parameters around the hulls of two models of the vessels measured at full scale. Model tests will be required for determining of scale effects and differences in predictive capabilities of CFD-codes for model and full scales. In this way a complete data set is produced that can answer the main questions in validation of CFD predictions and scaling procedures. The known relations between flow fields around a model and a full-scale ship give an opportunity to calibrate CFD codes. Typically, the designers will be able to perform only model tests to gather data necessary for the improvement of the CFD codes, so the knowledge of the expected scale effects will increase their confidence in the predictive capabilities of the CFD codes. The latest PIV (Particle Image Velocimetry) techniques were used to measure the 3D flow at model scale, laser sheets were used to measure the wave pattern generated by the model, furthermore the traditional resistance, propulsion and nominal wake tests were performed for both ships. Resistance, PIV, wave pattern and wake measurements were justified by the available uncertainty analysis. The basic resistance and propulsion tests were carried out for range of 5 knots (0.5 knot steps). The stock propeller was used in propulsion tests. Streamline test for model with running propellers were carried out using paint method. Longitudinal wave cut measurements were carried out. Complete wave field, even that on and very close to the model, were measured using new laser system and a set of 4 CCD cameras. The whole system was calibrated using 7-hole Pitot tube. The pressure distribution on the model hull surface, especially in the stern region, was measured. Additionally, HSVA has gathered the experimental data of 6 vessels where in the past model scale flow measurements have been carried out. Furthermore a detailed hull form has been given including rudders in IGES format and the main propeller characteristics. For the LDV measurements in the towing tank the three velocity components are available. The database contains the following vessels: - Sydney Express (Lpp 210 m); - St. Michaelis (Tanker, Lpp 174.0 m); - Hamburg Test Case (Lpp 153.7 m); - CVsym (CV with symmetric aft body, Lpp 133.7 m); - Cvasym (CV with asymmetric aft body, Lpp 133.7); - Bulk Carrier (Lpp 194m). The database has been delivered to all partners. The delivery of this database, together with the full-scale database, is an important step in the EFFORT project, since this means that the CFD participants can use for the first time ever the full-scale and the corresponding model-scale reference data for their developments and validation as is done in WP3 and WP4.

Appendage orientation is traditionally carried with use of model-tests. However due to the difference in boundary layer at model scale and full scale, it is very well possible that not the best orientation is found at model-scale. Leading to extra resistance and cavitation problems. This case study will demonstrate the use of full-scale CFD calculations for appendage orientation and compare the results with model scale experiments.

Initially, a tanker was supposed to be computed being similar to the previous cases. However, the construction of passenger vessels is the main task of this shipyard nowadays, and the knowledge of the flow field around such vessels may improve its competitiveness. According to the wishes of the shipyard, it was decided to take a new challenge and to compute a new ship type. Due to the huge amount of work and computations performed in WP 4, only very little time was left for the computations in WP 5. Therefore, the propeller computations with appendages were skipped. The viscous flow around the bare hull of the AKF vessel was computed at full-scale Reynolds numbers. The computations were performed as viscous-free surface and as double-hull computations. The double-hull computations were not exactly double-hull computations as the free surface is slightly deformed at the bow and at the stern. The goal was to use the FINFLO-SHIP code as it is. Difficulties appeared in the viscous-free surface computations due to the wedgelike transom. The wave height had to be kept fixed at the transom causing no deformation of the wave pattern behind the stern, which was an obvious bug in the code. The bug was corrected by Reijo Lehtim¨aki, however, due to the lack of time the free-surface computations could not be completed on the finest grid level. The surface piecing bulbous bow did not cause particular difficulties as in the case of the Uilenspiegel. However, the change from the medium grid to the fine grid will require a manual correction of the grid at the bow, which was not carried out here. The double-hull computations did not cause particular problems for all grid resolutions. The computational times of are too long for engineering purposes where a solution is required within less than 24 hours. However, for scientific purposes the FINFLO-SHIP is suitable due to its accuracy as shown in WP 4. At the present, the FINFLO-SHIP code is further developed for commercial purposes, which should make it user friendly and faster. In the future, the computational time will not be an issue anymore as more processors may be used due to parallel computing and the processor speed will increase still significantly. The bottleneck will remain the manual grid generation. Only an automatic grid generation would reduce the times significantly. However, for structured grid automatic grid generation seems to be no feasible yet. The design itself may be improved by a slight modification of the bulbous bow and the stern. Close to the free surface, the curvature of the bulbous bow should be increased below the bow wave. At the stern the curvature of the hull should be slightly reduced at x = -30 m, where a low-pressure peak is present.

Title: Transom immersion of a dredger In work package 5 of EFFORT several case studies have been carried out for application & demonstration purposes. This is the first case study. Transom immersion is a most common phenomenon for hopper dredgers due to their characteristic fullness together with the deep dredging draught (the operational freeboard of a dredger is only one third of the statutory freeboard). Hopper dredgers have large transom immersion and it is assumed that the effects of it on the total resistance are considerable (transom immersion is important for other ship types too, but to a lesser extend). To avoid the immersion of the transom the buttocks of the stern must be stretched so, that the transom is fully above the waterline, or they must be curved upward. The disadvantage of the first solution is that the ship length will increase with an amount of approximately 10 m! That increase will have a lot of consequences and must be stipulated as unrealistic. The disadvantage of the second solution is the fact that the flow will not follow that curvature and will separate from the hull. Then, the resistance will increase too and when the curvature starts in front of the propeller also vibration problems can be expected when the separated flow will enter the propeller plane. The use of potential flow CFD is normal business nowadays to improve the hull of a dredger, but the stern waves are not predicted very well because of the transom immersion (there is definitely no potential flow anymore). A viscous flow CFD code must be used for this matter. For this reasons a study has been carried out to investigate the use of viscous CFD for this purpose. Unfortunately, the code in this case was not able to deal with a free surface. So, the applicability with respect to the sternwaves was not investigated. Nevertheless, this study remained interesting because a comparison was made between a transom stern and an extended stern. For the flow and resistance matters, the following results can be noted: - The extension of the stern to avoid any transom immersion is significant and can be judged as unrealistic in practice. - The computed wake field is not affected by the simplified approach of an extended stern instead of an immersed transom. - From the propeller plane to the transom further downstream gradually a difference in pressure is noticeable, but rather little. - In general, model scale and full scale show the same trend. - Full-scale computations for the shallow water condition show more influence on the axial flow, but only slightly. - The computed resistance coefficients learned that the right order of magnitude of power increase due to the immersed transom was predicted. This result will lead to further investigations: - Improvement of the stern without extension with the help of viscous CFD. - CFD codes must be developed to a full employable tool to enable prediction of the stern waves.

Universities will use the output to update their teaching material. Since an important change in the role of model testing in ship design is expected to take place in the future, the early involvement of universities educating naval architects is crucial.

Rezultaty

Udostępnij tę stronę

Pobierz