Scholarlink

<p>Continental rifting involves extensive deformation of the lithosphere and mantle thermomechanical convection, driving continental extension and oceanic crust formation. However, accurately capturing rapid viscosity changes associated with nonlinear power-law rheology and strain softening is essential. Although nonlinear iterations effectively resolve viscosity variations during early rifting, continuous nonlinear iteration significantly increases computational cost. To mitigate computational cost while preserving numerical stability, we implemented a partial nonlinear iteration strategy, applying Picard iterations exclusively during the early simulation stage. We investigated the influence of different solving strategies, NI (No Iteration), PI (Partial Iteration), and FI (Full Iteration) on surface topography, surface heat flux, and nonlinear mantle viscosity evolution in continental rifting and seafloor spreading models. Although the NI solver requires fewer computational resources, numerical errors accumulated during progressive time-stepping induce structural asymmetry in subsurface properties and surface observables. The 1% PI solver showed improved stability relative to the NI solver, exhibiting slight asymmetry in strain-rate and viscosity at 10 Myr. Even when nonlinear iterations were limited to only the initial 5% and 10% of total simulation time (5% and 10% PI solvers), surface heat flux and internal physical properties, including temperature, viscosity, strain, and strain-rate, closely matched those obtained with the FI solver. The 10% PI solver produced surface topography and mantle viscosity profiles along the rift-axis most similar to those of the FI solver. Finally, we confirmed that the proposed partial nonlinear iteration strategy effectively maintains structural stability and symmetry in three-dimensional rifting models, highlighting its suitability for advanced three-dimensional geodynamic simulations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Partial Nonlinear Iteration During Early Continental Rifting for Symmetric Surface Topography and Heat Flux

  • 期刊论文

  • 发表日期：

    2025年9月7日

查看全文
  • Jieun Chang,
  • Byung-Dal So

摘要

Continental rifting involves extensive deformation of the lithosphere and mantle thermomechanical convection, driving continental extension and oceanic crust formation. However, accurately capturing rapid viscosity changes associated with nonlinear power-law rheology and strain softening is essential. Although nonlinear iterations effectively resolve viscosity variations during early rifting, continuous nonlinear iteration significantly increases computational cost. To mitigate computational cost while preserving numerical stability, we implemented a partial nonlinear iteration strategy, applying Picard iterations exclusively during the early simulation stage. We investigated the influence of different solving strategies, NI (No Iteration), PI (Partial Iteration), and FI (Full Iteration) on surface topography, surface heat flux, and nonlinear mantle viscosity evolution in continental rifting and seafloor spreading models. Although the NI solver requires fewer computational resources, numerical errors accumulated during progressive time-stepping induce structural asymmetry in subsurface properties and surface observables. The 1% PI solver showed improved stability relative to the NI solver, exhibiting slight asymmetry in strain-rate and viscosity at 10 Myr. Even when nonlinear iterations were limited to only the initial 5% and 10% of total simulation time (5% and 10% PI solvers), surface heat flux and internal physical properties, including temperature, viscosity, strain, and strain-rate, closely matched those obtained with the FI solver. The 10% PI solver produced surface topography and mantle viscosity profiles along the rift-axis most similar to those of the FI solver. Finally, we confirmed that the proposed partial nonlinear iteration strategy effectively maintains structural stability and symmetry in three-dimensional rifting models, highlighting its suitability for advanced three-dimensional geodynamic simulations.