Implicit Sub-stepping Scheme for Critical State Soil Models Publication type: Journal Article
Hoang-Giang Bui, Jelena Niníc, Günther Meschke, Hoang-Giang Bui, Günther Meschke · 2025
Read Implicit Sub-stepping Scheme for Critical State Soil Models Publication type: Journal Article by Hoang-Giang Bui, Jelena Niníc, Günther Meschke, Hoang-Giang Bui, Günther Meschke (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-2504-17476.txt.
Introduction
Hoang Giang Bui, Jelena Ninic, and Günther Meschke built a consistent tangent operator in 2025 for a sub-stepping method in soil modeling. When a soil model integrates stress in big load steps, the usual implicit Euler step can fail to converge on the local nonlinear problem. Splitting the step into sub-steps rescues convergence, but makes the tangent operator, the derivative the global solver needs, messy.
The paper develops that operator for two widely used critical state models, the modified Cam-Clay model and a unified clay and sand model. With it, the global Newton iteration regains its fast quadratic convergence. The method is aimed at problems with large steps, such as tunnel simulation. The catalog title appends a publication type and its creator list repeats names; the head governs. Readers get the math that makes sub-stepping practical in big simulations.
Catalog record
| CREATOR | Hoang-Giang Bui, Jelena Niníc, Günther Meschke, Hoang-Giang Bui, Günther Meschke |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-17476
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Implicit Sub-stepping Scheme for Critical State Soil Models Publication type: Journal Article Hoang-Giang Bui, Jelena Niníc, Günther Meschke, Hoang-Giang Bui, Günther Meschke Abstract The stress integration of critical soil model is usually based on implicit Euler algorithm, where the stress predictor is corrected by employing a return mapping algorithm. In the case of large load step, the solution of local nonlinear system to compute the plastic multiplier may not be attained. To overcome this problem, a sub-stepping scheme shall be used to improve the convergence of the local nonlinear system solution strategy. Nevertheless, the complexity of the tangent operator of the sub-stepping scheme is high. This complicates the use of Newton-Raphson algorithm to obtain global quadratic convergence. In this paper, a formulation for consistent tangent operator is developed for implicit sub-stepping integration for the modified Cam-Clay model and unified Clay and Sand model. This formulation is highly efficient and can be used with problem involving large load step, such as tunnel simulation. keywords Modified Cam-Clay, CASM, sub-stepping scheme, implicit integration † † corresponding: Hoang Giang Bui, University of Birmingham, UK. † † citation: ), , , . † † copyright: , ( † † year: 2020 † † ctitle: Implicit Integration with Sub-stepping Scheme for Critical State Models † † cjournal: International Journal for Numerical Methods in Engineering † † cvolume: xxxx;xx:xx–xx BUI et al 1 Introduction The introduction of critical soil models marks an important advance in computational geomechanics. Based on elastoplastic theory, the model is first proposed by Roscoe and coworkers [ 1 , 2 ] as Cam-Clay model (CC), and subsequently improved as modified Cam-Clay model [ 3 ] . The modified Cam-Clay model (MCC) is shown to sufficiently predict main attributes of the constitutive behavior of normally consolidated soft cohesive soils ( Gens_Potts:88 ) , although deficiencies were reported when modeling highly overconsolidated soils . Later, Yu Yu:98 generalized the standard Cam-Clay model to allow modeling the behavior of sand. The unified model, denoted as CASM Yu:98 ; Yu_Khong_Wang:07 , has the ability to resolve some drawbacks of the standard Cam-Clay models, such as the overestimation of the strength on the “dry side” or the behavior of loose sand in undrained tests, while limiting the number of parameters to seven, which is an advantage to adopt the model both in academic research and practical applications. To apply the critical soil models in numerical simulations, an integration algorithm is required. In the context of implicit methods, the backward Euler algorithm is typically used. This is frequently resorted to the return mapping algorithm Simo_Hughes:98 ; Simo_Taylor:86 , which will ensure quadratic convergence of the Newton-Raphson algorithm if certain conditions are satisfied. The advantage of the implicit method is its robustness, second-order accuracy and able to use with large load step. Nevertheless, the implicit algorithm typically requires the high-order derivatives of yield functions and plastic flow rule. The return mapping algorithm has been successfully applied to Cam-Clay models in Borja:90 and, being extended to the finite strain regime, in Simo_Meschke:93 . Further applications of the return map algorithm to extensions of the Cam Clay model have been suggested in Borja_Lin_Montans:01 ; Borja_Choo:16 ; Borja:04 . In contrast to implicit methods, explicit methods, based on the forward Euler integration, requires only the first order approximation of the yield surface and plastic flow rule and it is considered to be more straightforward to implement. Because only first order accuracy is attained, the load step must be sufficiently small to reduce the accumulated error. The efficiency and robustness of explicit algorithms has been substantially improved by using sub-stepping algorithms Sloan:87 ; Sloan_Abbo_Sheng:01 ; Sheng_Sloan:01 ; Zhao_et_al:05 . An automatic error control for sub-stepping algorithm is proposed in Abbo_Sloan:96 . It is well known from analysis of convergence properties of stability of the Newton algorithm, from applications reported in the literature Crisfield:91 ; Borja:13 , and also experienced by the authors, that the computational robustness of implicit integration schemes applied to elastoplastic models, and, in particular, to critical state models, is limited, and that no convergence of the local iterative scheme at the integration point level is attained when the load step is large. Therefore, it is crucial to sub-divide the load into smaller increments to ensure the existence of local admissible stress state and the convergence of the local iteration, to compute the internal variables and the plastic increment. Because the sub-load increment can be adjusted, this scheme is highly effective for problems, in which the loading path is difficult to control. An example of this class of problems are excavation problems, which are a priori stress control problems if no arc length methods are applied. A consistent sub-stepping scheme has been proposed for elastoplastic models in ( Perez-Fouget_Rodriguez-Ferran_Huerta:01 ) . In this paper the trial stress predictor uses a linear extrapolation based on a (constant) elastic tangent, which is typically used for elastoplastic material models. However, this scheme is not optimal for critical state soil models, because the stress-strain rule in the elastic region of the critical state model is nonlinear Roscoe_Burland:68 . In Section 2 of this paper, a sub-stepping algorithm for the implicit integration algorithm for critical state soil models, more specifically for the Modified Cam Clay model and the Clay and Sand model, is proposed, which makes use of the more accurate nonlinear approximation of hydrostatic pressure. The effectiveness of the implicit sub-stepping scheme is demonstrated by selected numerical examples presented in Section 3 . The findings from these analyses are concluded in Section 4 . 2 Implicit integration of the critical state soil models In this section, the basic governing equations of the Modified Cam-Clay (MCC) ( Section 2.1 ) model and the Clay and Soil model ( Section 2.2 ) are summarized before an implicit integration algorithm incorporating sub-stepping is elaborated for both models in subsections Section 2.4 . 2.1 Brief summary of the modified Cam-Clay model The Modified Cam-Clay (MCC) model Schofield_Wroth:68 ; Roscoe_Schofield_Wroth:58 is characterized by a yield surface formulated in the p-q plane by: f^{\text{MCC}}(p,q,p_{c})=\left(\dfrac{q}{M}\right)^{2}+p(p-p_{c}), (1) by the associative plastic potential g^{\text{MCC}}(p,q,p_{c})=f^{\text{MCC}}(p,q,p_{c}) and by the nonlinear stress-strain relation in the elastic domain: \dot{p}=K\dot{\varepsilon}^{e}_{v}\,,\qquad\dot{\mathbf{s}}=2\mu\dot{\boldsymbol{\varepsilon}}^{e}_{d}. (2) The shear modulus and the pressure-dependent compression modulus K are given as K=\dfrac{1+e}{\kappa}p\,,\qquad\mu=Kr\,,\qquad r=\dfrac{3}{2}\dfrac{1-2\nu}{1+\nu}, (3) and the hardening rule is specified as \dot{p}_{c}=\theta p_{c}\dot{\varepsilon}^{p}_{v}\,,\qquad\theta=\dfrac{1+e}{\lambda-\kappa}. (4) In Equations 1 , 2 , 3 and 4 , M,e,\lambda,\kappa,\nu are material constants and have following meanings: M is the slope of the critical state line, e is the void ratio, \lambda is the compression index, \kappa is the swelling index and \nu is the Poisson ratio. For the details on the material parameters and their roles in reflecting the characteristics of soil, the reader is referred to Wood:90 . p denotes the hydrostatic pressure and q is the deviatoric invariant q=\sqrt{3/2\,\bf{s}:\bf{s}} , with \mathbf{s}=\boldsymbol{\sigma}-p\mathbf{I} as the deviatoric stress and p = \dfrac{1}{3}\text{tr}(\boldsymbol{\sigma}) . The subscript (\bullet)_{v} denotes the volumetric component and (\text{\textbullet})_{d} denotes the deviatoric component of the strain. The superscript (\bullet)^{e} denotes the elastic component and (\bullet)^{p} denotes the plastic component of the strain. The void ratio e is considered not to change significantly during a load step and is kept fixed within one increment . 2.2 Brief summary of the