pipette
ESEspañol

Positivity-preserving scalar auxiliary variable schemes for gradient flows via a quadratic reformulation

Qiong-Ao Huang, Zhi-Hao Liu, Ying-Wei Wang, Li-Na Yan

Preprint

In the authors' words

The scalar auxiliary variable (SAV) method replaces the nonlinear part of the free energy by a positive scalar , thereby yielding linear, unconditionally energy-stable schemes for gradient flows. At the discrete level, however, the standard backward Euler and Crank--Nicolson discretizations provide no guarantee that the computed remains positive, an inconsistency with the continuous definition that contradicts the square-root ansatz and may compromise long-time robustness. Although the SAV method has been widely applied, this subtle but consequential issue has received little attention. We first characterize this failure quantitatively by deriving a sharp criterion and a sufficient condition on the time step size, and construct an explicit counterexample showing that sign loss occurs for parameters of practical relevance. Rather than modifying the definition of as in existing positivity-preserving variants, we retain the square-root form and reformulate the discrete evolution from to , which converts the scalar equation into a convex quadratic with a strictly negative constant term, always yielding a unique positive root. For the Crank--Nicolson scheme, the product-form discretization preserves this quadratic structure, while conventional alternatives do not. The resulting schemes incur the same computational cost as the original SAV method and are proved unconditionally energy-stable. Numerical experiments for the Cahn--Hilliard equation confirm the predicted positivity, energy stability, and convergence rates.

Main resultThe abstract does not state a limitation.

Appeared: Friday, October 2. arXiv. Preprint, not yet peer-reviewed.