The good agreement in Fig. 2 and 3 between computations and model experiments leads to a reasonable question, what would be changed if the comparison is not at the resistance test-state but at the self-propulsion test-state. With other words, what kind of influence would the propeller have on the sinkage and trim? To get a reasonable answer, Fig. 4 compares the measured sinkage (Graph a) and trim (Graph b) at a representative water-depth for the inland vessel at the towing state (+) and self-propulsion state (x) by using a conventional ducted propeller. As expected, the mean sinkage (more or less at the midship section) seams almost to be independent of the propeller action. However, the trim is influenced by the propeller action, which would tendentiously lead to a bow-up trim, as observed in Graph b. The influence of the propeller action on the ship's squat is shown in Fig. 4 (c). Since the relative small contribution of the whole trim portion to the squat, the effect of the propeller-induced trim can be neglected for ships with conventional propellers.

　

　

Coming now to the subject seagoing ship, Fig. 5 compares the measured (+) and computed (o) squat in shallow water of three depths as well as the estimation from the empirical formula from Barrass (1979). Like for the inland vessel, a similar agreement between computations and measurements was achieved for the subject seagoing ship, demonstrating the relative robustness of the computer program BEShiWa for predicting the squat of ships in shallow water. In contrast to the computation, the empirical formula yielded only good results at a lower speed and in water of a moderate depth.