Submitted to General Relativity and Gravitation (2025) — Preprint available: arXiv:2512.11885

This project studies a phenomenological Einstein-Cartan ekpyrotic model in which a steep scalar potential and an effective Weyssenhoff spin fluid produce a nonsingular, torsion-supported bounce in a homogeneous, nearly FLRW background. The goal is not to present a complete cyclic cosmology, but to analyze whether this class of bounce backgrounds can remain dynamically viable once homogeneous shear and curvature are included.
To do that, I extend the Copeland-Liddle-Wands scalar-fluid system to a six-dimensional phase space including shear, curvature, and spin-torsion, derive the full Jacobian, and study the fixed-point structure of the model. I then scan the softening parameters of the scalar potential to identify a finite “bounce basin” where trajectories remain nonsingular, and I supplement the linear analysis with maximal Lyapunov exponent calculations, finding no indication of chaotic behavior within the homogeneous parameter ranges explored.
The project is explicitly limited to homogeneous backgrounds. It does not yet address perturbations, observational fitting, entropy accumulation across cycles, or the full inhomogeneous BKL problem, all of which are left for future work.