--- title: "Foundation of FEM; choosing a numerical method" subject: "Computational Techniques in Civil Engineering" chapter: 1 key: "semester_8/computational_techniques_in_civil_engineering:fem-foundation-large-deformation" source: https://www.bctengineer.com/q/semester_8/computational_techniques_in_civil_engineering/fem-foundation-large-deformation site: BCT Engineering --- # Foundation of FEM; choosing a numerical method Computational Techniques in Civil Engineering · Chapter 1: Introduction - Asked in: 2070 Magh, 8 marks - Video: [11 differences between linear and nonlinear FEA](https://www.youtube.com/watch?v=MNSccPRJph4) — The Modeling Guy - Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14 - Page: https://www.bctengineer.com/q/semester_8/computational_techniques_in_civil_engineering/fem-foundation-large-deformation - Whole chapter: https://www.bctengineer.com/md/semester_8/computational_techniques_in_civil_engineering/chapter-1 ## Question Explain the foundation of the finite element method. Why is this method less appropriate for large deformation problems? How do you choose a numerical method for different problems? Illustrate with examples. ## Short answer FEM is founded on potential energy stationarity, using piecewise shape functions to form [K]{U}={F}. It is less suitable for large deformation because Lagrangian mesh distortion causes element tangling and accuracy loss. Method choice depends on geometry and material continuity; e.g., FEM for structural stress, DEM for granular flow. ## Answer #### Foundation of the finite element method FEM rests on the idea that a complex continuum can be approximated by a collection of simple pieces (elements) whose behaviour is known. Its mathematical basis is: 1. **Variational / energy principle:** the correct displacement field makes the total potential energy $\Pi=U-W$ stationary (minimum), or equivalently satisfies the virtual work equation. Weighted residual (Galerkin) methods can also be used. 2. **Piecewise approximation:** displacement in each element is $u=[N]\{d\}$ with shape functions that are continuous across elements. 3. **Element equations:** from $\{\varepsilon\}=[B]\{d\}$ and $\{\sigma\}=[D]\{\varepsilon\}$, $[k]=\int [B]^T[D][B]\,dV$. 4. **Assembly and solution:** compatibility at nodes and equilibrium give $[K]\{U\}=\{F\}$. #### Why FEM is less suitable for large deformation - Elements are attached to the material (Lagrangian); under very large strain the mesh distorts, elements become tangled or negative in volume, and accuracy falls. - Small-strain formulation $[B]$ and constant $[K]$ assume small displacement; for large deformation the stiffness changes with geometry and needs non-linear incremental solution and re-meshing. - Material separation, fracture and flow are not represented by continuous elements. Methods such as DEM, SPH or ALE formulations are better. #### Choosing a numerical method | Problem | Suitable method | |---|---| | Stress in frame, plate, dam | FEM | | Seepage, unsteady channel/pipe flow on a grid | FDM | | Infinite/semi-infinite soil, tunnel, crack | BEM | | Granular soil, jointed rock, rockfall | DEM | | Dam-break, debris flow, large deformation | SPH / mesh-free | Selection depends on geometry, material behaviour (linear or non-linear, continuum or discontinuum), boundary conditions, required accuracy and computing resources. ## Check yourself The answer's key points as multiple-choice questions, with the correct option and why. 1. What principle underlies FEM? A. Total potential energy is stationary (minimum) for the correct displacement field B. Maximum entropy C. Superposition of fundamental solutions only D. Maximum stiffness Answer: A. Variational/energy principle or virtual work gives the element equations. (Key point: Variational basis) 2. Why is FEM less suitable for very large deformation? A. Shape functions are discontinuous B. Matrix [K] becomes full C. FEM cannot use [B] D. Mesh attached to material distorts, elements tangle and accuracy falls Answer: D. Lagrangian meshes distort under very large strain. (Key point: Large deformation) 3. Which methods are better for large deformation? A. Finite difference only B. Analytical methods C. DEM, SPH or ALE formulations D. Boundary element for non-linear problems Answer: C. Methods without attached meshes handle large deformation better. (Key point: Alternative methods) 4. Which method suits an infinite soil or tunnel problem? A. FDM B. BEM C. DEM D. FEM always Answer: B. BEM only meshes the boundary and handles infinite domains. (Key point: Selection) ## Related in Computational Techniques in Civil Engineering - [Techniques of solving problems; FEM and FDM](https://www.bctengineer.com/question.md?key=semester_8%2Fcomputational_techniques_in_civil_engineering%3Afem-fdm-working-techniques) — Computational Techniques in Civil Engineering, chapter 1 - [Types of problems solved by FEM](https://www.bctengineer.com/question.md?key=semester_8%2Fcomputational_techniques_in_civil_engineering%3Afem-problem-types) — Computational Techniques in Civil Engineering, chapter 1 - [Basic steps in finite element analysis](https://www.bctengineer.com/question.md?key=semester_8%2Fcomputational_techniques_in_civil_engineering%3Afem-basic-steps) — Computational Techniques in Civil Engineering, chapter 1 - [Concepts of FEM and FDM with advantages](https://www.bctengineer.com/question.md?key=semester_8%2Fcomputational_techniques_in_civil_engineering%3Afem-fdm-concepts-advantages) — Computational Techniques in Civil Engineering, chapter 1 - [Necessity of computational techniques; FEM algorithm](https://www.bctengineer.com/question.md?key=semester_8%2Fcomputational_techniques_in_civil_engineering%3Anecessity-computational-fem-algorithm) — Computational Techniques in Civil Engineering, chapter 1 - [Software used for FEM and FDM](https://www.bctengineer.com/question.md?key=semester_8%2Fcomputational_techniques_in_civil_engineering%3Asoftware-fem-fdm) — Computational Techniques in Civil Engineering, chapter 1 ## Related in other subjects - [Steps and advantages of finite element method](https://www.bctengineer.com/question.md?key=semester_8%2Ffinite_element_method%3Afem-steps-advantages) — Finite Element Method, chapter 1 - [Foundation classification and problems](https://www.bctengineer.com/question.md?key=semester_6%2Fbuilding_technology%3Afoundation-class-problems) — Building Technology, chapter 2, asked 2× - [Compare FDM, FVM and FEM](https://www.bctengineer.com/question.md?key=semester_7%2Fapplied_computational_fluid_dynamics%3Afdm-fvm-fem-comparison) — Applied Computational Fluid Dynamics, chapter 3 - [FEM steps and geometric modelling](https://www.bctengineer.com/question.md?key=semester_7%2Fmachine_design_ii%3Afem-steps-geometric-modelling) — Machine Design II, chapter 1