Abstract
This project aims to develop an automated workflow between various Python packages and the WECSim time-domain numerical simulation tool in MATLAB to efficiently evaluate the performance of WECs and use its time-domain outputs to inform subsequent design iterations. iProTech’s Pitching Inertial Pump (PIP) WEC is used as the test WEC and is parameterized into a series of independent variables (design variables) and dependent variables that can be used to create a mesh of the WEC, calculate its hydrodynamic coefficients, and simulated in a time-domain environment. This process is wrapped into an optimization function to automatically calculate time-domain results and use those results to inform how the optimizer should adjust design variables. Various single variable optimization studies are performed to determine the relative influence that each variable has on the power output of the device. A complete design optimization of the entire PIP WEC was not included as this involves more advanced optimization algorithms and better software organization and architecture.
The PIP WEC uses an internal water coil to generate electricity through pressure differences. The outer hull can be parameterized into three adjoined circles: a nose, a bottom, and a stern, and the internal water coil can also be parameterized into variables such as the diameter of the water coil, or the stiffness and damping of the numerical PTO system. Each of these design variables is optimized to maximize the power output of the device over a range of regular wave periods (running irregular wave simulations increases the computation time dramatically). The diameter of the internal water coil proved to be the most influential, as the moment of inertia of the coil and the hull have a direct correlation to the relative pitch motion of the device (power output). The bottom circle’s radius has the highest influence on power output among the geometric design variables. The remaining design variables have less influence on the device, mostly due to the initial design description of the PIP WEC.
Two out of the three main objectives of the project were met. The automation of the workflow between Python packages and WEC-Sim in MATLAB was completed and refined to where each creation and simulation of a design took on the order of 1-3 minutes. A sensitivity study of how different design variables affect the power output of the WEC was completed. A full system optimization, however, was not completed due to the need to integrate other more advanced optimizers and a reorganization of the design tool’s software.
Regardless, this work set up the start of a design tool that can incorporate the effects included in a timedomain environment. Expanding on this tool can promote future design optimization studies of WECs that include the time-domain. For the PIP WEC, the developers can focus on certain design variables over others to continue optimizing their design. Continued work should involve the reorganization of the design tool’s software to better integrate other optimizers and post-analysis scripts to fully optimize the design of any WEC of any function.