Hybrid Test Surrogates for Flexible Composite Pipe Service Life from Sparse Experiments and Models

Abstract

The transition toward more resilient and adaptable infrastructure in the energy sector has precipitated a massive increase in the deployment of flexible composite pipes. These structures are subjected to complex, multi-axial loading and harsh environmental conditions that induce coupled degradation mechanisms over time. Traditional service-life screening relies heavily on purely empirical accelerated life testing or highly computationally intensive physical simulations, both of which struggle to capture the full spectrum of long-term failure modes accurately. This paper introduces a comprehensive framework for developing hybrid test surrogates, which integrate sparse physical testing data with advanced physics-informed digital models to predict the service life of flexible composite pipes more efficiently. By leveraging a bidirectional data flow between physical test benches and computational models, the proposed hybrid surrogate minimizes the time required for life screening while maximizing predictive fidelity. Extensive methodological detailing is provided regarding data acquisition, feature extraction, and the harmonization of physical and virtual test modules. The findings demonstrate that hybrid surrogate architectures significantly reduce the computational and temporal overhead compared to conventional physical screening methodologies. Furthermore, the results indicate enhanced precision in identifying critical failure thresholds, such as matrix micro-cracking and fiber-matrix debonding. This research provides a foundational blueprint for deploying hybrid surrogates in the certification and lifecycle management of non-metallic piping systems.