Initializing…
Home
🌍 Seismic
Live Quakes
   PHK Seismic Hub
πŸ“š Knowledge
Blogs Publications
πŸ‘€ About Me
About CV Projects Contact
EN فا
Seismic Lexicon / Computational & Data Methods / Finite Element Method
💻

Finite Element Method

FEM
⭐ Featured Term

Definition

The Finite Element Method (FEM) is a numerical technique for solving partial differential equations by dividing a complex domain into smaller, simpler subdomains called finite elements. It is the dominant computational tool in structural engineering for analyzing stresses, deformations, and dynamic response of structures under arbitrary loading.

Detailed Explanation

The Finite Element Method (FEM) is a numerical technique for solving boundary value problems governed by partial differential equations. The method divides a continuous domain — a structural component, a soil mass, a fluid region — into a mesh of smaller, non-overlapping subdomains called finite elements. Within each element, the unknown field (displacement, temperature, pressure) is approximated by interpolation functions (shape functions) that relate the values at discrete nodes. This discretization converts the governing partial differential equations into a system of algebraic equations — typically written in matrix form as K·u = F for static problems or M·ü + C·u̇ + K·u = F(t) for dynamic problems — which can be solved numerically. The method was introduced in the 1950s for aircraft structural analysis and has since become the dominant computational tool in structural, mechanical, civil, and aerospace engineering.

FEM is the foundation of modern structural analysis and earthquake engineering. In static analysis, it computes stresses, strains, and deformations under gravity, wind, or other loads. In dynamic analysis, it computes natural frequencies and mode shapes (eigenvalue analysis), response to time-varying loads (time-history analysis), and response spectra (response spectrum analysis). In nonlinear analysis, it captures material nonlinearity (yielding, cracking, crushing), geometric nonlinearity (large deformations, P-delta effects), and boundary nonlinearity (contact, friction, soil-structure interaction). Modern earthquake engineering relies on FEM for performance-based design, where the goal is to predict the deformation and damage of a structure under specific ground motions. Commercial software such as SAP2000, ETABS, ABAQUS, ANSYS, OpenSees, and Perform-3D implement FEM and are standard tools in engineering practice. OpenSees (Open System for Earthquake Engineering Simulation) is particularly important in the earthquake engineering research community, providing an open-source platform for advanced nonlinear analysis.

FEM is a powerful but demanding tool that requires careful application. The quality of a finite element analysis depends critically on the mesh β€” its refinement, its element type (beam, shell, solid), and its ability to capture the expected response. Too coarse a mesh produces inaccurate results; too fine a mesh is computationally expensive. The choice of constitutive model β€” linear elastic, elastoplastic, damage, or more advanced β€” determines the fidelity with which material behavior is captured. Boundary conditions must be realistic; incorrect assumptions about supports or soil-structure interaction can dominate the response. Solution algorithms (direct, iterative, implicit, explicit) affect convergence, stability, and computational cost. Verification and validation β€” comparing results against analytical solutions, experimental data, or benchmark problems β€” is essential for confidence in the results. Modern FEM practice increasingly uses model updating, probabilistic methods, and machine learning to reduce uncertainty and to accelerate analysis. In Iran, FEM is widely used in structural and geotechnical engineering practice, with SAP2000, ETABS, ABAQUS, and OpenSees being common tools, particularly for high-rise buildings, bridges, dams, and critical facilities.

Formula

K · u = F (static) | M · ü + C · u̇ + K · u = F(t) (dynamic)
Share this term: X LinkedIn Telegram