On the measurement of enhanced energy storage in composite concentric helical pipes with RSA-DNN verification

This study presents a measurement framework for evaluating the thermal and mechanical response of a concentric helical energy storage unit. The device comprises two coiled pipes: the inner pipe carries a heat-transfer fluid (HTF), and the outer pipe is maintained at the liquidus temperature of a binary solar salt that serves as phase-change material (PCM). The PCM-filled annulus is modeled via the enthalpy–porosity method to capture mushy-zone behavior with temperature-dependent properties for both PCM and HTF. The inner pipe is an Al/AlN metal–matrix composite (MMC); its effective conductivity is computed using the Maxwell–Eucken relation. Transient CFD with energy, momentum, and mass conservation is coupled to a one-way fluid–structure interaction (FSI) step to quantify tube deformation. Model validity is established against published data with a maximum discrepancy of 5.2%. To enable rapid performance estimation, a Reptile Search Algorithm–optimized deep neural network (RSA-DNN) is trained on simulation data to map operating/material inputs to the outlet–inlet temperature rise and to replicate the FSI-informed trends. The surrogate achieves validation R² up to 0.947 and ≤0.5 K absolute error across the operating envelope while preserving expected monotonic trends with mass flow and AlN fraction. As an application, four HTFs are benchmarked: Therminol 62 attains a 48 °C gradient at 200 s with the lowest mass flow (50 g/s), whereas Therminol 66 requires 57.89 g/s. The integration of CFD–FSI modeling with an RSA-optimized DNN surrogate constitutes a novel, data-efficient framework that simultaneously captures thermal and structural responses in PCM-assisted helical systems. This dual focus on thermo-mechanical performance, combined with a metaheuristic-optimized surrogate, goes beyond existing PCM-based exchanger studies. It enables rapid HTF/material screening, reduces reliance on costly simulations, and informs the design of compact, durable, and application-ready energy storage systems.