Target Strength (TS) is one of the critical design parameters used to evaluate the susceptibility of naval ships, which is directly related to their survivability. Numerous studies have been conducted to analyze and estimate TS, and software tools dev...
Target Strength (TS) is one of the critical design parameters used to evaluate the susceptibility of naval ships, which is directly related to their survivability. Numerous studies have been conducted to analyze and estimate TS, and software tools developed based on these studies provide highly accurate solutions even for arbitrarily shaped underwater vehicles. Recently, with the increase in platform size and the advancement of stealth technologies, the computational burden associated with TS analysis has been significantly growing. In response, there has been increasing interest in improving the computational efficiency of TS analysis by leveraging the rapid development of GPU technology optimized for high-speed processing.
In this study, we propose a computationally efficient methodology for TS analysis using a GPU-based image-based Physical Optics (PO) approach combined with a ray-tracing method. In particular, we suggest a near-field TS analysis method based on ray tracing and apply it to enhance the overall TS analysis procedure. To verify the feasibility of the proposed method and examine its applicability to various underwater targets, we developed a GPU-based TS analysis software package that includes a user interface (UI). This software was applied to perform numerical analyses on simple geometries such as plates, spheres, and cylinders, as well as on standardized submarine target models. The results were compared and validated against those obtained using a combined method of Geometric Optics (GO) and Physical Optics (PO).
In addition, this study presents TS analysis procedures for targets with attached target strength absorbing materials (TSAMs) using the proposed methodology. The validity of this procedures was confirmed through numerical simulations of standard target models. The results demonstrate that the proposed GPU-based method significantly improves computation speed compared to conventional CPU-based TS analysis, while maintaining comparable accuracy. The proposed methodology is expected to enable reliable and rapid TS analysis across a wide range of simulation scenarios and analysis conditions.