Dynamic-stabilization-based linear schemes for the Allen–Cahn equation with degenerate mobility: MBP and energy stability
Hongfei Fu, Dianming Hou, Zhonghua Qiao, Bingyin Zhang · 2025
Read Dynamic-stabilization-based linear schemes for the Allen–Cahn equation with degenerate mobility: MBP and energy stability by Hongfei Fu, Dianming Hou, Zhonghua Qiao, Bingyin Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-16447.txt.
Introduction
Hongfei Fu, Dianming Hou, Zhonghua Qiao, and Bingyin Zhang design number schemes in a 2025 paper. The schemes solve the Allen-Cahn equation, a rule for how two mixed states split. Mobility, how fast the mix can move, may fade to zero in spots here. Past schemes can break the cap on values or lose energy control in that case.
Their schemes use a shifting brace that holds values in bounds at all times. Energy stays in check too, even when mobility fades to zero. They prove tight error caps for both a first-order and a second-order scheme. The fast scheme solves just one linear system per time step. Sample runs match the proofs and show clean splits.
Catalog record
| CREATOR | Hongfei Fu, Dianming Hou, Zhonghua Qiao, Bingyin Zhang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 66 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-16447
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Dynamic-stabilization-based linear schemes for the Allen–Cahn equation with degenerate mobility: MBP and energy stability Thanks: The first author was supported in part by the NSFC grant 12131014, the NSF of Shandong Province grant ZR2024MA023, the Fundamental Research Funds for the Central Universities grant 202264006 and the OUC Scientific Research Program for Young Talented Professionals. Thanks: The second author was supported in part by the NSFC grants 12001248 and 12571387, and the NSF of Jiangsu Province grant BK20201020. Thanks: The third author was supported in part by the Hong Kong Research Grants Council GRF grant 15305624, and CAS AMSS-PolyU Joint Laboratory of Applied Mathematics (Grant no. JLFS/P-501/24). Thanks: The fourth author was supported in part by the Fundamental Research Funds for the Central Universities under grant 202461103. Hongfei Fu, Dianming Hou, Zhonghua Qiao, Bingyin Zhang Abstract . In this paper, we investigate linear first- and second-order numerical schemes for the Allen–Cahn equation with a general (possibly degenerate) mobility. Compared with existing numerical methods, our schemes employ a novel dynamic stabilization approach that guarantees unconditional preservation of the maximum bound principle (MBP) and energy stability. A key advance is that the discrete energy stability remains valid even in the presence of degenerate mobility-a property we refer to as mobility robustness. Rigorous maximum-norm error estimates are also established. In particular, for the second-order scheme, we introduce a new prediction strategy with a cut-off preprocessing procedure on the extrapolation solution, and only one linear system needs to be solved per time level. Representative numerical examples are provided to validate the theoretical findings and performance of the proposed schemes. 2020 Mathematics Subject Classification Primary 35K55, 65M06, 65M12, 65M15, 65M50 1. Introduction The classic Allen–Cahn equation, first introduced by Allen and Cahn in 1979 [ 1 ] , was designed to simulate the motion of anti-phase boundaries in crystalline solids. Since its inception, it has been extensively used to model a wide array of phenomena, such as interface dynamics in multiphase fluids [ 24 , 38 , 44 ] , mean curvature flow [ 12 , 20 ] , image segmentation [ 2 , 30 ] , and various applications in materials science. In numerous practical situations, the two phases present in the system (such as the components of a binary alloy) exhibit distinct physical or chemical characteristics, which naturally leads to the diffusion mobility being dependent on the phase variable [ 7 ] . Furthermore, in specific cases, accurately modeling the phase separation process necessitates that the mobility vanishes in pure phases [ 13 , 35 ] . This study focuses on the following Allen–Cahn equation with a general (potentially degenerate) mobility: (1.1) \displaystyle\phi_{t}=-\mathcal{M}(\phi)\mu, \displaystyle t>0,\ \mathbf{x}\in\Omega, (1.2) \displaystyle\mu=-\varepsilon^{2}\Delta\phi-f(\phi), \displaystyle t>0,\ \mathbf{x}\in\Omega, subject to the initial condition \phi(\mathbf{x},0)=\phi_{\text{init }}(\mathbf{x}) for any \mathbf{x}\in\bar{\Omega} , together with either periodic or homogeneous Neumann boundary conditions. Here, \Omega\subset\mathbb{R}^{d} (d=1,2,3) is an open, bounded Lipschitz domain. In ( 1.1 )–( 1.2 ), \phi(\mathbf{x},t) represents the relative concentration difference between the two phases, 0<\varepsilon\ll 1 is a parameter related to the interfacial thickness, and f(\phi)=-F^{\prime}(\phi) with F(\phi)=\frac{1}{4}(1-\phi^{2})^{2} being the standard double-well potential. The mobility function \mathcal{M}(\phi)\geq 0 is allowed to be nonlinear and possibly degenerate. One commonly used degenerate mobility function in materials science is as follows: (1.3) \mathcal{M}(\phi)=(1-\phi^{2})^{m}\quad\text{with}\ m>0, while other forms of mobility can be found in [ 7 , 8 , 9 , 33 , 41 ] or in subsection 4.4 . The Allen–Cahn equation ( 1.1 )–( 1.2 ) satisfies following energy dissipation law (1.4) \displaystyle\frac{d}{dt}E[\phi]=-\int_{\Omega}\mathcal{M}(\phi)\mu^{2}d\mathbf{x}\leq 0, where the free energy E[\phi] is defined by (1.5) \displaystyle E[\phi]=\int_{\Omega}\left(\frac{\varepsilon^{2}}{2}|\nabla\phi(\mathbf{x})|^{2}+F(\phi(\mathbf{x}))\right)d\mathbf{x}. This shows that the free energy E[\phi] decreases monotonically over time. Moreover, the Allen–Cahn equation also satisfies the maximum bound principle (MBP), that is, if the initial data satisfy |\phi_{\text{init }}(\mathbf{x})|\leq 1 , then the solution remains bounded as |\phi(\mathbf{x},t)|\leq 1 for all t>0 ; see [ 35 ] for further details. It is noteworthy that the energy dissipation law and the MBP are two fundamental properties of the Allen-Cahn equation. Therefore, it is crucial for numerical schemes to preserve such physical features at the discrete level, particularly the MBP, since its violation may cause the mobility (e.g., ( 1.3 )) to become negative, leading to ill-posedness and nonphysical numerical solutions. In the past decade, significant progress has been made in developing numerical schemes for Allen–Cahn type equations that ensure the preservation of the discrete MBP and energy stability, particularly for cases with constant mobility. Tang et al. [ 40 ] introduced a first-order linear stabilized scheme that was both energy-dissipative and MBP-preserving. To achieve second-order accuracy in time while maintaining (modified) energy dissipation and MBP, nonlinear variable-step second-order backward differentiation formula (BDF2) schemes [ 28 ] and stabilized Crank-Nicolson/Adams-Bashforth schemes [ 19 ] have been explored. High-order MBP-preserving time integration schemes, reaching up to third- and fourth-order, have been developed for semilinear parabolic equations using the (stabilized) integrating factor Runge-Kutta (IFRK) method [ 21 , 27 , 46 ] . More recently, Liao et al. [ 29 ] proposed corrected IFRK methods using two classes of difference corrections, the telescopic correction and nonlinear-term translation correction, which preserve both steady-state solutions and the original energy dissipation law. Li et al. [ 26 ] developed high-order in time numerical methods for parabolic equations, enforcing the discrete MBP at nodal points through a cut-off operation applied to the lumped mass finite element solution, although energy stability was not addressed. This cut-off approach was subsequently combined with the scalar auxiliary variable (SAV) approach [ 36 , 37 ] , resulting in (modified) energy-stable and MBP-preserving linear schemes [ 43 ] . Further advancements include the use of Lagrange multiplier techniques for positivity- and bound-preserving discretizations of parabolic equations [ 5 , 6 ] , offering a fresh perspective on the practical cut-off method. Additionally, Ju et al. [ 22 ] integrated the SAV method with the stabilized exponential time differencing (ETD) approach [ 10 , 11 ] to design first- and second-order linear schemes that unconditionally preserve the discrete MBP and a modified energy dissipation law for the Allen–Cahn equation with constant mobility. The presence of nonlinear and potentially degenerate variable mobility significantly complicates the construction and analysis of numerical schemes for the generalized Allen–Cahn equation ( 1.1 )–( 1.2 ), compared to the classical Allen–Cahn equation with constant mobility. In [ 35 ] , a stabilized, MBP-preserving first-order linear scheme was developed for the generalized Allen–Cahn equation with an advection term. To circumvent the challenges posed by potential mobility degeneracy, Cai et al. [ 4 ] assumed the existence of a positive constant \mathcal{M}_{*} such that (1.6) \mathcal{M}(s)\geq\mathcal{M}_{*}>0,\quad\forall s\in\mathbb{R}, which excludes degenerate mobility functions like ( 1.3 ) from practical applications. Under this assumption, they proposed stabilized first-order ETD and second-order ETD Runge–Kutta schemes that unconditionally preserve the discrete MBP. Recently, Hou et al. [ 15 ] introduced a linear variable-step second-order scheme for the Allen–Cahn equation with general mobility, combining a prediction-correction approach with stabilization. This scheme employs a classical first-order method to predict a locally second-order and MBP-preserving solution, followed by a stabilized BDF2-type correction step. As a result, two variable-coefficient Poisson-type systems are solved per time step. Using the kernel recombination technique [ 28 , 32 ] ,