Consistent Dilatancy-Based Constitutive Model of Rock Salt for Geo-Energy Storage Applications: An Open-Source Numerical Implementation

Accurate numerical simulation of salt cavern performance during underground geo‑energy storage of hydrogen, natural gas, compressed air, and CO₂ requires the integration of precise design criteria into a rigorous mechanical constitutive framework. Ideally, this framework should be consistent, describing all mechanical loading (either short-term or long-term) deformations with a single parameter set. Due to methodological disconnects between experimental derivation of design criteria and constitutive laws, only a limited number of consistent constitutive models exist in the literature. The RTL2020 model exemplifies such a mechanical constitutive law by incorporating dilatancy both as a feature in the constitutive model and as a design criterion, enabling accurate prediction of short-term or long-term mechanical deformations, including volumetric strain, with a single parameter set. The novelty of the RTL2020 model lies in the dilatancy-induced volumetric strain, which shifts the focus of design criteria from merely surface deformations to encompass volumetric changes at the cavern wall, critical for precisely assessing salt cavern integrity and performance. This study presents a novel numerical implementation strategy for the RTL2020 model within an open-source numerical simulator, and captures the effectiveness of the dilatancy-based design criteria. The implementation approach employs an iterative stress–strain update algorithm that combines the Newton Raphson method with an adaptation of the classical implicit integration scheme (elastic predictor–inelastic corrector). This formulation ensures unconditional stability, rapid convergence, and high numerical accuracy at low computational cost. Validation against experimental data demonstrates the model’s ability to reproduce key rock salt behaviours, including the negligible influence of mean pressure on axial and deviatoric strain, and the strong dependence of volumetric strain on mean pressure, highlighting the robustness of our implementation. Ongoing application of the model will demonstrate its capability of handling parametric analysis of field-scale hydrogen storage conditions in single and multi cavern systems. Future investigations will integrate the RTL2020 model into fully coupled multiphysics frameworks (thermo-mechanical or hydro-mechanical) to simulate the more complex conditions characteristic of underground hydrogen storage. As an open-source implementation, the model provides the geomechanics community  with a broadly accessible and reliable consistent mechanical constitutive framework for salt cavern design in underground geo‑energy storage applications.