FD生成的PMM数据在初步设计阶段进行评估英文文献和中文翻译
Comparison between the computed and measured forces and moments showed a quite good agreement. Formal verification and validation, taking numerical and experimental uncertainties into account, was made for a selected condition. The result showed that the X-force is validated, while the Y-force and the yaw moment are not. For the simulated manoeuvres, the results also look promising, since reasonable good agreement was found when comparing the simulated turning circle and zigzag manoeuvres, obtained with measured and computed input data. Introduction Today manoeuvring simulators are used to find the trajectory of the ship in arbitrary manoeuvres. During the ship design phase, focus is mainly on the requirements to the IMO standard manoeuvres (10/10 and 20/20 zigzag and 35 turning circle), which the new ship must comply with. The simulator models are based on integral quantities and the idea is to solve the equations of motion together with large measured sets of hydrodynamic force/moment data. The major advantage of the simulators compared to CFD is their ability to simulate general manoeuvres in real time. But, since they rely on usage of experimental PMM data, which is costly and time consuming to produce, they are usually used late in the design phase, after the final hull form is known and the scale model can be built. Consequently, the manoeuvring characteristics of the ship are evaluated late in the design process. Using CFD instead of the experiment for the data generation would not require the scale model and the simulation could therefore be done earlier in the design process. It is well known that CFD requires much CPU time when it comes to producing hydrodynamic force/moment data by sweeping through large test matrices. But, taking into account how much extra useful information the CFD results provide about the flow field, this is acceptable. Maneuvers have traditionally been simulated by empirical methods, for instance as done in Agdrup (2008) or by simulators based on mathematical models using experimental rotating arm or PMM data, for instance as done in Otzen (2008). The application of CFD in connection with maneuvering has shown promising results in the past. However, much of the work performed so far has primarily focused on inpidual PMM conditions like static drift and static rudder configurations, e.g. Simonsen and Stern (2003), Simonsen and Stern (2005) and Simonsen et. al (2006) or pure yaw, e.g. Simonsen and Stern (2008). The SIMMAN2008 manoeuvring workshop, showed the same tendency, except for very few cases where CFD was used to predict the final manoeuvres directly. Only one participant, Hochbaum et. al (2008), tried the approach used in this present paper, i.e. CFD based PMM. The approach was somewhat simplified, since effects of free surface and dynamic sinkage and trim were not taken into account, but even so quite good results were obtained for the final manoeuvres.
Based on above, the objective of the current work is to be able to perform standard deep water IMO maneuvering simulations in the early design phase purely based on computed input data, i.e. without using experimental data. This also includes investigations on how the numerical PMM test can be reduced to minimize the computational effort, while still producing sufficient information to give reliable maneuvering simulations. Approach All the experimental and numerical activities in the present work are focused on the appended KCS containership, which, according to the 24th Maneuvering Committee ITTC Report, is appointed as one of the benchmark ships within the maneuvering community. A full standard test matrix for the PMM test is defined and the PMM test is conducted in FORCE Technology’s towing tank in Denmark. Based on this full data set, a mathematical model for the simulator is made and the IMO maneuvers are simulated. A reduction of the full test matrix is then made by removing points from the matrix and corresponding maneuvering simulations are performed to study the influence on the maneuvers, which will be quantified by overshoot angles and turning circle diameters. For each of the simulations the resulting maneuvers will be compared to the initial reference case obtained with the full matrix to quantify the changes. Based on this work, the test matrix for the computational work will be chosen. With the reduced test matrix, CFD computations will be made for all the conditions and the resulting X and Y forces and the yaw moment for each condition will be compared with data from the measurement as a point to point comparison. The computations are performed with the RANS solver Star-CCM+. Finally, all the computed force and moment results will be included in the simulator model in order to rerun the maneuvering simulations based on the numerical data input. The simulated maneuvers will be compared with the results from the previous simulations based on the measured data with full and reduced test matrix. The structure of the paper is as follows. First the model test is presented. This is followed by a presentation of the CFD work. In this connection, point to point comparison of all measured and computed quantities will be made. Next part covers the maneuvering simulations, i.e. the applied mathematical model and simulations based on measured PMM data and the reduction of test matrix. Finally, simulations will be made where the EFD data from the static conditions in the reduced test matrix are replaced with data computed with CFD. The result based on EFD and CFD will be compared to conclude about how well the CFD based input data performs. Data generation based on model testing The model testing is carried out in FORCE Technology’s towing tank with a scale model of the KCS container ship using a scale of 1:53.667. The towing tank is 240m long, 12m wide and 5.5m deep. Figure 1. PMM model test setup at FORCE Technology. The main particulars and propeller data for the ship are shown in Table 1. The approach speed for the manoeuvres is U0=1.701 m/s (24.0 knots full scale) corresponding to a Froude number of 0.26. The test is described in detail in Larsen (2012), so only a brief summary is given here. Table 1: Hull and propeller data. Lpp [m] 4.3671 B [m] 0.6114 T [m] 0.2051 S