A CAGD plus CFD integrated optimization model for design of AUVs

The design of 'Autonomous Underwater Vehicles (AUVs)' is mission specific and it is getting more and more complex with time because of higher demands of range, endurance, payload, operation flexibility, navigational capabilities for deep and restricted water depths, energy efficiency and special mission requirements as placed by the user agencies. These requirements of design are primarily conflicting in nature and hence they can be efficiently satisfied or at least a best design suiting a set of requirements can be computed with the application of optimization techniques. At present, the design process for AUVs is dominated by ad-hoc' approaches or empirical formulations. Recent developments in computational science and engineering, optimization techniques and 'Computational Fluid Dynamics (CFD)' have the capability to enable the designer to compute better estimates of the drag and other parameters that are important in design. This premise has motivated the present work and in this paper we present a design optimization framework for design of an AUV. In the proposed model, 'Computer Aided Geometric Definition (CAGD)' is integrated with the CFD along with the optimization framework for optimizing a given parameter (i.e. resistance). In our work, CFD is used in place of empirical formulations for the estimation of drag as with CFD one can accurately estimate the drag and thus increasing the fidelity of the optimization model. Furthermore, the use of CFD by integrating it with CAGD and optimization method allows the study of parametric hull form generations and analysis. With the genetic algorithm driven optimization technique, CFD simulation, the automatic generation of geometry based on the design parameters, automatic generation of mesh and automatic analysis of fluid flow, the objective function is computed and optimized efficiently and the results are presented in this work. In this paper, one of the most popular genetic algorithm, 'Non-dominated Sorting Genetic Algorithm NSGA-II' is implemented with MATLAB*(TM) and the CFD analysis is implemented with SHIPFLOW**(TM). And, in the design process GA and CFD are integrated and used to optimize the design variables for minimization of an objective function (i.e. viscous resistance) and finally, we present a design example motivated by the real world applications and show that the integration of NSGA-II with CFD and CAGD is effective for AUV hull form optimization. Furthermore, the effectiveness of the design variables considered for the optimization process on the design of the low drag hull forms for the design of AUVs is critically examined in this work.