A model for simulation was built up, based on the hydrodynamic derivatives derived from PMM test and the line in Fig.12 expresses the spiral characteristics predicted by the model. The simulated values are in good agreement with the measured ones and the model is used to evaluate the overshoot angles in various zigzag manoeuvre. This is the prototype model for simulation and hereafter called the original one.

　

 In order to investigate the course keeping ability by the same method as mentioned for conventional ships, the other models that have the different loop width are necessary for simulations.

　

 Another model was built up after the oblique captive model test using the model ship that is the original one with center skeg shown in Fig.11. The hydrodynamic derivatives for drift angle were obtained from the test and only changed from the original model. This model is like the original ship with a center skeg as shown in Fig.11. Exactly speaking, center skeg affects the other hydrodynamic derivatives on turning but they are neglected in this case. The spiral characteristics predicted by simulation are shown in Fig.13. The loop width is decreased by the effect of the center skeg but turning ability at large pod angle is not so much changed. This is hereafter called a model with center skeg.

　

　

 The third model was built up after the only modification of the strut area of pod system to 1.5 times as large as that of the model with center skeg. The loop width is 8.4deg and narrowest in the three models. It has almost same turning ability with the other models at large pod angle. This is hereafter called a model with center skeg and large strut.

　

 It is clear that center skeg and the increase of strut area is effective to decrease the loop width. Therefore the ship equipped with two pod propulsion systems is advantageous to obtain the larger lateral area than with a single propulsion system.

　

3.2 The overshoot angle in modified Z manoeuvre

　

 The zigzag manoeuvres were simulated by three models that have the different loop widths. The results are shown in Fig.14-16. The horizontal axis is the loop width of the spiral characteristics and the vertical axis is the overshoot angle. 1^st and 2^nd overshoot angles in modified zigzag manoeuvre are respectively shown in Fig.14 and Fig.15. The combination of pod angles and heading ones are same with the cases of section 2.2 for a ship with a conventional propulsion system. In addition, 1^st overshoot angle for 10deg/10deg Z manoeuvre is shown in Fig.14 and 2^nd overshoot angle in Fig.15. Moreover 1^st overshoot angle for 20deg/20deg Z manoeuvre is shown in Fig.16.

　

 It is shown from these figures that the overshoot angles increase in proportion to increasing the loop width. It is moreover clear that the overshoot angles for loop width in modified zigzag manoeuvre change less than in 10deg/10deg and 20deg/20deg zigzag manoeuvre and that their differences and values show the most remarkable for loop width in 10deg/10deg and 20deg/20deg zigzag manoeuvre.

　

　

　

　

 It is concluded from the facts that the overshoot angles in 10deg/10deg and 20deg/20deg zigzag manoeuvres are the better item to evaluate clearly the difference of course keeping and yaw-checking ability of a ship.

　

 It is said that modified zigzag manoeuvre was originally developed in order to find the indices of K and T by the small helm angle for a ship that has a large overshoot angle or that cannot be conducted zigzag manoeuvre test [7]. Ship motion in modified zigzag manoeuvre is smaller than in zigzag manoeuvre at same helmed angle because the turning motion is smaller in modified one. Therefore the results in Fig.5-7 and Fig.14-16 are reasonable. On the other hand, the fact that a ship cannot have overshoot angle in 10deg/10deg Z manoeuvre means that the ship has a wide loop width in spiral characteristics like 20deg and the ship has a bad course keeping ability.

　

3.3 The effect of strut area to course keeping and yaw-checking ability

　

 As well as for a ship with a conventional propulsion system, the effect of pod strut area to course keeping and yaw-checking ability was investigated for a ship with a pod propulsion system by means of simulations using a model with center skeg, the pod strut area of which was purely modified.

　

 The results are shown in Fig.17. The horizontal axis expresses the percent for prototype strut area and the vertical axis expresses 1^st and 2^nd overshoot angles in 10deg/10deg Z manoeuvre and 1^st overshoot angles in 20deg/20deg Z manoeuvre. All Z manoeuvre began at pod angle to starboard side. In the figure, the marks express the simulated results and the lines express the criteria of course keeping and yaw-checking ability in the resolution A.751(18).

　

 It is found from the figure that overshoot angles decrease in proportion to the increase of strut area and that the ship can almost comply with the resolution if she has center skeg and 2.5 times of a pod strut area as large as the prototype.

　

 It is moreover shown that the criterion of 2^nd overshoot angle is the severest in the criteria on course keeping and yaw-checking ability. In this case, the ship has 6deg loop width.