Constitutive Modeling of Porous Hydrate-Crowned Liquid Films for Dynamic Interfacial Area and Film Thickness Prediction in Gas Pipelines

Abstract

We propose a constitutive modeling framework that replaces conventional sharp-interface film thickness correlations with a dynamic, multi-scale system for predicting gas-water interfacial area and liquid film evolution in deepwater gas pipelines. The core novelty lies in explicitly coupling hydrate nucleation and crystal growth to interfacial rheology, thereby capturing the feedback between microstructure and macroscopic film behavior. The proposed system integrates four interconnected modules: a neural network emulator trained on interfacial Langmuir trough experiments to output viscoelastic moduli and yield stress, a discrete element method crust morphology simulator that maps crystal packing porosity to mechanical strength, a porous hydrate crust constitutive law expressing effective viscosity as a function of porosity and shear rate, and a modified thin-film lubrication solver that incorporates these rheological closures. In this framework, the gas-water interface is no longer treated as a sharp discontinuity with constant tension; instead, it behaves as a viscoplastic material whose properties evolve with subcooling, film Reynolds number, and local porosity. The porosity itself is governed by a dynamic population balance that tracks hydrate mass deposition from the gas phase. We solve the coupled lubrication and porosity equations using a fully implicit finite-volume method, with the neural network and discrete element simulations providing closure relations at each time step. Our approach enables transient prediction of film thinning, crust formation, and interfacial area evolution under stratified, annular, or slug flow regimes. The primary contributions are the introduction of a porosity-dependent yield stress derived from percolation theory, the replacement of static film correlations with a causally coupled rheological film model, and the demonstration of direct interchange between microscale crystal packing and macroscale film dynamics. This framework significantly improves fidelity in hydrate risk assessment for subsea pipelines.

Keywords