The objective of this project was to compute the bistatic scattering amplitudes for unexploded ordnance (UXO) targets for use in the Target-in-the-Environment Response (TIER) model to simulate an experimental scenario for detection, localization and identification of these targets. The mathematical details and the rationale for the geometries used in computing the bistatic scattering amplitudes have been presented before.

This project summarizes the development and application of a physics-based numerical modeling framework for simulating acoustic scattering from underwater UXO. The work focuses on computing high-fidelity bistatic acoustic scattering amplitudes for realistic UXO geometries and validating those results through a series of physically meaningful reconstructions. The resulting data products support research and development efforts aimed at producing high fidelity target response to sonar.

The modeling approach is based on a coupled finite element–boundary element (CFEBE) formulation, which accurately captures the structural response of complex targets. To ensure broad applicability of the scattering database across different environments, far-field scattering responses were computed in free space for a large ensemble of source and receiver locations distributed over the surface of a sphere with a 10m radius. The resulting bistatic scattering data were then provided to the TIER model, which incorporates environmental effects into the target response.

Because bistatic scattering amplitudes are difficult to validate directly, a major emphasis of this work was placed on finding ways for their validation and physical interpretation. Several postprocessing techniques were developed to reconstruct physically meaningful acoustic quantities from the bistatic data. These included interpolation-based reconstruction of the acoustic color, projection of the scattered field onto coordinate planes, and reconstruction of time-domain pressure responses. In each case, reconstructed results were shown to be consistent with independently computed reference solutions, providing strong evidence that the bistatic data preserve the essential physical characteristics of the scattering process.

To enable efficient simulation of large and numerically challenging targets, the CFEBE framework was enhanced by incorporating a clustering technique with the aim of reducing the pressure degrees of freedom. These improvements significantly increased computational robustness and efficiency, particularly for large-caliber ordnance. The enhanced capabilities were implemented in a modern, extensible software environment, providing a foundation for continued development and broader application.

In addition to supporting physics-based modeling and validation, the generated scattering data can be adapted for use in data-driven classification studies. With its capability to exactly integrate the complex ocean environment in scattering computations, the CFEBE can also be used as a benchmarking tool.

Overall, the results of this work demonstrate that advanced numerical modeling of acoustic scattering has reached a level of maturity where simulated data can closely approximate the fidelity of experimental measurements. This capability offers a practical pathway for supplementing or partially replacing costly and time-consuming field experiments, enabling more rapid exploration of target and environmental variability. Beyond UXO-related applications, the methods and tools developed here are broadly applicable to problems in underwater acoustics, structural–acoustic interaction, subsurface sensing, and wave propagation, supporting future research and technology development across multiple scientific and engineering domains.

The results of this effort demonstrate that physics-based numerical modeling of acoustic propagation and scattering has advanced to a level where simulated data can closely approach the fidelity of experimentally measured data. This represents a meaningful shift in how future research and development activities may be conducted. As model accuracy continues to improve, high-quality synthetic data can be used to supplement, guide, or partially replace costly and time-intensive experimental campaigns. This capability enables rapid exploration of parameter space, supports hypothesis testing prior to field deployment, and reduces overall program risk while accelerating technology maturation.

Advances from this work have direct implications for the development of next-generation systems for the automatic detection, classification, and localization of UXO. Accurate simulation tools allow detection and classification algorithms to be developed, trained, and evaluated under controlled and repeatable conditions that are difficult or impractical to achieve experimentally. By reducing dependence on large-scale field experiments, this approach offers a cost-effective pathway for advancing mission-critical technologies while preserving experimental resources for validation and operational testing. (Project Completion - 2026)