Three dimensional time domain simulation of ship motions and loads in large amplitude head waves

Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2015
Thesis identifier
  • T13992
Person Identifier (Local)
  • 201099457
Qualification Level
Qualification Name
Department, School or Faculty
  • This PhD thesis presents the development of a practical computational tool named Large Amplitude RESponse (LARes), based on 3D quasi-non-linear time-domain technique, to predict ship motions and loads in large amplitude waves which can be accessible to ship designers. Firstly, a linear 3-D Green source panel code (LARes L1) was developed to perform linear time-domain analysis ship motion and internal load simulations based on the frequency-domain hydrodynamic coefficients which were calculated in the linear PRECAL software. Linear simulations are validated with the linear time-domain PRETTI software results using rectangular barge geometry. The motions, internal loads, global and sectional hydrodynamic forces were agreed well with the linear PRETTI model results in zero and forward speed simulations. Then, non-linear time-domain panel code (LARes L2) was developed in order to predict ship motions and loads in large amplitude waves using the Froude-Krylov nonlinearity level. At each time step, the exact wetted area of the ship surface under the wave profile was calculated and fed in the time-domain motion and load equations while the diffraction and radiation forces were kept as linear. The present program achieved good agreement with the non-linear PRETTI model results both for the barge and S175 container geometries at zero and forward speed conditions in small amplitude waves. Moreover, the S175 container ship results are compared with the available experimental data and agreed well with the experimental results in forward speed case. It has been observed that PRETTI code is over-estimating motion and load responses especially around the resonant frequency due to the surge motion influence in the memory forces evaluations. In the Froude-Krylov nonlinear level predictions, it has been observed that PRETTI diverges from the experimental results when the wave steepness is higher than 0.08 due to the linear radiation and diffraction forces. Based on the same framework, a more advanced nonlinear time-domain panel code (LARes L3) was developed in order to investigate the effects of quasi-non-linear diffraction and radiation forces in large amplitude ship simulations. A new mesh generator was introduced in order to cut and correct the original panels under the still water level in the updated position of the ship after displacements and rotations. The quasi-non-linear diffraction and radiation forces were calculated at the pre-defined position cases and stored in a database. In order to lower the computational cost multi-dimensional integration and interpolation codes were generated. The S-175 containership was tested in 120 different position cases and resulting hydrodynamic coefficients and forces were stored in the database. The results of the LARes L3 model were compared with the available experimental data using the S-175 containership in forward speed. The computed motion responses showed a good agreement with the experimental data. Moreover, three of the developed models are compared with the experiments and their performances were investigated with respect to the increasing wave slope. In addition to that, the effect of the wave length and ship speed in large amplitude waves are investigated in detail. Non-linear behaviors of the codes were compared with the experimental results which showed a good agreement. Finally, the Vertical Shear Force (VSF) and Vertical Bending Moment (VBM) responses were investigated in large amplitude motions. It was observed that, in the validation section, numerical model peak amplitudes showed well agreement with the experimental results, but they were observed to be shifted to the higher frequencies compared to the experimental results. The reason for that was attributed to the longitudinal mass distribution on the ship in the experimental setup which had not been provided in detail in the published experimental results.
Resource Type
Date Created
  • 2015
Former identifier
  • 1219181