Following the ITTC (2017) recommended procedures of the ship model resistance tests and uncertainty analysis, the error sources in the test process were analyzed. The influence of these error sources on the test results are determined according to the propagation principle of uncertainty. The error components are divided into type A uncertainty for random error during the model tests and type B uncertainty for other error sources. Repeat-run tests were carried out on a single-screw benchmark ship model equipped with and without rudder at 7 speeds. According to the guidelines of model tests, these test results are converted to a unified nominal water temperature at 15°C. The analysis results show that: 1) the average resistance results of the ship model without rudder are lower than the average resistance of ship model with rudder, the differences are between 0.499% to 0.929%. 2) At 95% confidence level(k=2), when the ship is without rudder, the expanded uncertainty of the average resistance at different speeds is between 0.316% and 0.554%. When the ship is with rudder, the expanded uncertainty of the average resistance at different speeds is between 0.311% and 0.508%. 3) Among the uncertainty components, the type A uncertainty of the model tests and the type B uncertainty of the carriage speed account for a large proportion of the uncertainty component. Finally, this paper further puts forward the measures to reduce the uncertainty of the model tests.
The accuracy of the ship model test has a direct and important impact on the performance prediction of the full-scale ship, each towing tank member is also taking measures to improve the test accuracy, such as improving the instrument measurement accuracy, increasing the period of measurement time, improving the standardization and so on.
Uncertainty analysis is a method to quantify errors, which is based on the theoretical true value and measures the probability level of the test results within a certain range of errors(Coleman, 1999; Coleman, 2018). In the uncertainty analysis steps of ship model test, the test process should be combed firstly. Then the influence degree of each variable, namely sensitivity, is determined based on the DRE (Data Reduction Equation). Finally, the uncertainty of the variable and its sensitivity are combined, and the uncertainty of ship model test can be calculated.