<p>Non-pneumatic structures significantly enhance tire safety performance, with fatigue life being a crucial indicator for non-pneumatic tires (NPTs). The spoke structure, subjected to complex cyclic loads during rolling, is particularly prone to fatigue failure due to its repeated stress-strain cycles; however, calculating honeycomb structure NPT of fatigue life remains challenging due to the nonlinear material behavior and multi-physics coupling effects. This paper addresses this by proposing a fatigue life calculation method based on tearing energy, which accounts for the energy dissipation mechanism in elastomeric materials. Utilizing finite element numerical simulation, we systematically established relationships between fatigue life and key evaluation indicators (maximum stress, principal logarithmic strain, strain energy density, and strain energy density gradient) under varying loading conditions. Power function-based prediction models for each indicator were then constructed through statistical regression analysis. A comparative analysis of model fitting and prediction accuracy revealed that all four models achieved high fitting accuracy, with the strain energy density gradient-based model exhibiting the highest prediction accuracy, especially for predicting crack initiation and propagation stages. This study provides a theoretical foundation for NPT development and lifespan design improvements, enabling more reliable fatigue resistance optimization in engineering applications.</p>