Dynamic Load Inversion Method of Ship Body Based on Influence Coefficient Matrix

Abstract

Ships sail in complex marine environments, the load is complex and uncertain. In the actual hull strength analysis, hull health monitoring and diagnosis, structural optimization design and other issues, it is very important to determine or estimate the dynamic load which applied on the hull structure. Due to the complexity and uncertainty of the working environment of the ship, many dynamic loads such as slamming loads applied on the bow of the hull are impossible or difficult to be measured directly. However, the structural response (strain, displacement, acceleration, etc.) of the corresponding part and other system characteristic parameters such as natural frequency, vibration mode, damping ratio, etc. can be obtained by installing sensors. Using the dynamic response of the structure, combined with the load identification model, the dynamic load on the hull can be calculated reversely. The time domain method and the frequency domain method are currently the most commonly used methods on load inversion. The time domain method is very suitable for the inversion identification problem of non-stationary loads and impact loads. The solution is intuitive and easy to use. In this paper, on based the time domain method, for the non-transient dynamic load, the displacement and time orthogonal function basis functions are loaded on the structure as the load function, and the response of the structure is obtained. The influence coefficient matrix is established by the response, and then the inversion load function is fitted; For transient dynamic loads such as slamming load, unit pulse load is applied on the structure to obtain the response of the structure. Then the corresponding data are filtered to overcome the ill-conditioned matrix problem. The matrix of the influence coefficient is established by the response, and the dynamic load is obtained. The two methods of dynamic load inversion based on the influence coefficient matrix are verified by numerical examples. The error is within a reasonable range, which proves that the inversion method is feasible.