Physics-Based ML & Predictive Digital Twins for Submarines

Project summary

From empirical testing to certification by analysis.

The proposed project stems from funded European and UKRI-wide proposals in which digital technologies for multiscale design are adopted. The particular task of this study is to start the implementation of a fully physics-based synthetic environment to support the overall design of nuclear submarine systems.

The overall aim is to introduce a predictive digital engineering framework integrating three complementary scientific pillars:

1. Integrated Computational Materials Engineering (ICME) for multiscale modelling of materials and manufacturing processes.
2. Physics-of-Failure reliability modelling enabling probabilistic fatigue and damage-tolerance assessment.
3. AI/ML-assisted digital engineering environments coupling synthetic simulation, structural modelling and predictive digital twins.

This integrated approach will enable earlier reliability assessment, reduced reliance on large-scale physical testing and improved predictive capability for certification-relevant structural behaviour.

Aims and objectives

The ambition is to demonstrate that predictive digital engineering can transform the development of nuclear systems from a predominantly empirical, test-driven process to a simulation-supported certification paradigm. Specific objectives include:

• Reduction in the number of required physical tests in the certification test pyramid.
• Improved accuracy of fatigue life prediction compared with conventional design approaches.
• Validated coupling between operating-conditions simulation, structural modelling and digital twin prediction.
• Demonstration of a predictive digital twin capable of supporting virtual structural assessment at component level.
• Quantified reduction in structural mass while satisfying the same load and reliability requirements.
• Quantified improvement in predicted service life and reduction in premature replacement risk.

Alignment to STAND-UP impact targets