Scheduling Optimization of Multiple CFD for Reducing Time in Ship Performance Analysis

Recently, performance of the ship propulsion is very important due to energy efficiency and greenhouse gas emissions. Therefore, computational Fluid Dynamics (CFD) is frequently used during the design phase to improve performance. Three CFD analyses have to be performed sequentially to calculate the ship's performance: resistance test, propeller open water test, and self-propulsion test. However, due to the limited number of licenses, it takes much time to finish CFD analysis of several ships simultaneously. In this study, we used a constraint satisfaction problem (CSP) to schedule the simulations sequentially and set a cumulative constraint for each simulation to constrain the number of simulations that can be processed simultaneously. The CP-SAT solver was used to minimize the time taken for CFD simulations. The optimized simulation times were then visualized using Gannt charts to easily compare the results of scheduling simulations in vessel number order with the optimization technique. We could see that the analysis time scheduled by the optimization technique was significantly reduced, which resulted in a significant improvement in work efficiency. We also compared the total analysis time to derive the optimal number of licenses.