The new model is validated mostly against SSPA's experiments, but is also validated against a selection of Vantorre's and Renilson's measurements. Below, the validation is discussed according to the influence of different variables.

　

4.1 Influence of ship speed (Froude number Fn)

　

 Fig.2 and Fig.3 show the sway force coefficient CY and yaw moment coefficient CN varying with Fn at four different h/T and at y/B = -0.6 for a vertical bank on the ship's port side. The predictions by the model are compared with the corresponding experimental data. It is clear that the mathematical model gives very good agreement with the measurement.

　

　

　

 Fig.4 shows an example of the predicted CY compared with experimental data of vertical banks from Renilson [7][8] and Vantorre [6], while Fig.5 shows the comparison of CN for the same case. As is seen, the agreement between the prediction by the model for vertical banks and the measurement is generally satisfactory, except for CN at h/T=1.51 and y/B=-1.0, where the moment measured by Renilson is much larger than the predicted one. In fact, doubt was raised for this test case when studying ref. [7].

　

　

　

 The above experiment results show a fact that the sway force and yaw moment vary with ship speed in a more complex manner. If they were proportional to the speed squared, then CY and CN in relation to Fn would have been horizontal lines. This is obviously not always the case, especially at very shallow water (h/T<1.2), the sway force and the yaw moment seem to be related to speed to a power higher than 2.

　

4.2 Influence of water depth h/T

　

 The influence of water depth ζ4 on CY and CN can be seen in Fig.6 and Fig.7 respectively, where both the measured and the predicted data are compared for three cases: a case with SSPA's experiment, a case with Renilson's [8] and Vantorre's [6] measurement data respectively. The cases have different speeds (Fn) and different distances y/B. The general tendency revealed by experiments has been well captured by the mathematical model.

　

　

　

 It is evident from the plots that the water depth plays an important role in bank effects. From deep (h/T>3) to shallow water (1.2<h/T<3), the sway force CY is a suction and increases its magnitude with decrease of h/T to a maximum suction, then decreases rapidly with decreasing h/T. Approximately at a value h/T≈1.10〜1.15, CY reduces to zero in Fig.6. With further decrease of h/T towards extreme shallow water (h/T<1.10), CY becomes a repulsive force and increases with decreasing h/T. The moment is generally small from deep (h/T>3) to shallow water (1.2<h/T<3). But at very shallow water (h/T<1.10), CN increases drastically as h/T decreases in Fig.7.