Simulations with Autopilot Control; Since the measured controls applied to the ship in the modeled channel did not reproduce the measured ship behavior, the ship transit was modeled using a line-following autopilot with corrections for deviation of the desired sailing line. This simulation was performed for the Base, 1000-ft, and 500-ft channels to determine if the channel modeling detail would impact the results. It was also modeled for bank moments at 1 and 3 times theoretical. Figures 14 and 15 show the results of the Base and 500-ft channels with bank multipliers of 3. It can be seen that there is a significant (10 deg) port rudder required to bring the ship into control at the start and then starboard rudder of the same magnitude was required at the location where a change in the port bank was noted above, just below markers 59-60. Beyond that the ship was easily control led with less than 5 degrees of rudder and maintained very close proximity and heading to the channel. The 500-ft channel model showed considerably more variation, as would be expected and developed a deviation from the channel centerline of 75 ft(23 m) at two locations. The rudder required was generally less than 10 degrees; however, a 35-degree rudder was used below markers 67-68.

　

　

　

 Increasing the bank moments generally increased the rudder control pattern by the magnitude of the multiplier with the pattern remaining similar; i.e., the rudder required for the modeling with bank moment multipliers of 1.0 required only about one-third of the rudder magnitude to control the ship.