Desktop simulation was used in formulating the "pre-liminary set of candidate slow speed maneuvers". It is quite effective to uncover issues and refine the tests, e.g., the choreography of test and the visualization of results. A cautionary note, simulation is only as good as its model. Therefore, simulation trials cannot uncover all potential issues until these candidate tests are checked out in real world.

　

 Based on the maneuvers suggested in Table 2, Jakobsen [34] conducted a series of computer simulation tests using four ship models as summarized in Table 4. The models were selected for a spread of block coefficients (Cb) from 0.56 to 0.83. The intention was to identify a pattern or trend. All ships are single screw ships with a right-handed propeller. The Mariner is the only ship equipped with a steam turbine, all the other simulated ships have slow speed diesel engine. The following section presents the results and discussions of several interesting tests.

　

　

6.1 Back & Fill - Fill First (B&F-FF);

　

 Back & Fill maneuvers are operations conducted in a constrained space. Since the lateral force developed by the propeller during backing plays a significant role in these maneuvers, these maneuvers need to be tested in both starboard and port directions. When the propeller lateral force is in line with the rudder force, the vessel changes its heading more effectively. For the B&F-FF maneuver, filling to starboard is therefore more effective than filling to port, as illustrated in Figure 2a and 2b for the VLCC.

　

 By the same token, backing to port is more effective than backing to starboard for the B&F-BF maneuver, as illustrated in Figure 3a and 3b. These results explain the preferred maneuvering strategies of pilots when dealing with single-screw, right-handed screw vessels in a tight area. Figure 4 illustrates the results of the same test procedures as in Figure 3(a), but for deep water. For a normal turn, the differences of Advance and Transfer at 45° heading change between deep water and shallow water are usually not that big. When the heading change reaches 900 and 180°, then the difference becomes significant. The tighter maneuver in deep water is expected.

　

 Note that during the period when the propeller lateral force is prominent, the rudder is not effective. Consequently, a moderate change in the modeling of propeller lateral force, rudder force, or hull force can have significant impact on the vessel's behavior. This reinforces the need for proper modeling of all these components and their interactions in all four quadrants during slow speed operations.

　

 Note that B&F tests are integrated tests of several component tests. Thus, for example, if the performance of B&F-BF is not satisfactory, a necessary trouble-shooting step is to look at the astern turn performance, i.e., a good simulation of astern turn is the prerequisite of good simulation of B&F-BF.

　

　

　

　

　

　

　

　

　

6.2 Minimum Effective Rudder (MER) Tests

　

 The MER test proposed by Fuller [35] for directionally unstable ships is like a truncated spiral test identifying an angle that is comparable to one half of the spiral instability loop width. The motivation of this test is to l) provide a pragmatic way of assessing ship directional instability (identified by the Japanese pilots as a key cause of maneuvering difficulties), see Yamazaki [36], 2) to convey the information as the minimum counter rudder required to stop a swing at various vessel speeds. For extremely unstable /difficult to control vesels, a MER table can provide valuable warning to the pilot even before he starts the transit.

　

 The desired generic outcome is a table as illustrated in Table 5 describing the MER at each bell in deep, 1.5 and 1.2 depth/draft shallow waters. Note that the sea trial of Esso Osaka [37] and other research efforts have shown that an unstable vessel in deep water can become further unstable around 1.5 depth/draft, and then becomes stable in shallower water (1.2 depth/draft). The deep-water results in Table 5 are based on the information in Table 6, which logs the VLCC MER test results in a nondimensionalized format.