Discretization is crucial in finite element analysis (FEA) because it involves breaking down a continuous domain into smaller, manageable subdomains, known as elements. This process allows complex problems to be addressed using simpler, individual components. When performing discretization, factors such as the element size, shape, and type must be considered, as they affect both accuracy and computational cost. Mesh refinement, particularly in areas of high gradients, is often necessary to ensure accurate results. An efficient discretization strategy balances these factors to optimize the analysis while maintaining computational feasibility .

Adaptive meshing improves FEM results by dynamically refining the mesh in regions with high gradients or complexities, such as stress concentration zones or where geometric complexities arise. This refinement increases the number of elements around critical areas, enhancing the accuracy and convergence of the solution without significantly increasing the overall computational load. By concentrating computational resources where they are needed most, adaptive meshing balances accuracy and efficiency, ensuring more reliable simulation outcomes .

The main steps in the Finite Element Method (FEM) include: 1) Problem Definition, which involves defining the physical problem, including geometry, material properties, and boundary conditions. This step sets the groundwork for a structured approach to the problem. 2) Discretization, which divides the domain into smaller, finite elements through a process called meshing, allowing complex geometries to be handled. 3) Element Equations, wherein equations for each element, such as the stiffness matrix for structural problems, are derived to capture the physical behavior of the elements. 4) Assembly, which combines these equations to form a global system of equations representing the entire model. 5) Application of Boundary Conditions, wherein the global system is modified to reflect constraints, ensuring the solution's physical validity. 6) Solution, where the system of equations is solved numerically to find unknowns like displacements or temperatures. 7) Post-Processing, which involves analyzing and visualizing results for insights into aspects such as stress and heat flux .

In FEM, boundary conditions (BCs) are essential in defining how the system interacts with its environment and are categorized as Dirichlet, Neumann, and Mixed BCs. Dirichlet BCs, or essential conditions, directly specify the value of the solution at certain boundaries, such as fixed displacements. Neumann BCs specify the derivative of the solution, such as a specific force applied. Mixed BCs combine both Dirichlet and Neumann conditions. The methods for applying BCs include the Penalty Method, which enforces constraints by introducing large stiffness values, and Lagrange Multipliers, which adds additional variables for constraints. Correctly applied BCs are critical for ensuring the physical accuracy and stability of the FEM solution .

The global stiffness matrix in FEM is assembled by combining the stiffness matrices of individual elements, representing the whole system's response to applied loads. Its key properties include symmetry ([K]=[K]T), which ensures equilibrium in forces, positive definiteness, which guarantees solution stability and non-negativity of potential energy, and sparsity, where many elements are zero, reflecting that most nodes interact only with their neighbors. These properties are critical as they affect the computational efficiency and stability of the numerical solution in FEM analysis .

To effectively address problems involving discontinuities or singularities in FEM, special adaptations such as enriched finite elements or adaptive meshing techniques are required. Enriched finite elements augment standard elements with additional functions that can capture singular behavior accurately, like stress concentration near cracks. Adaptive meshing dynamically refines the mesh near discontinuities to achieve higher resolution and accuracy without excessive computational cost. These adaptations help overcome FEM’s limitations in dealing with sudden changes in geometry or material properties within a modeled domain .

The stiffness matrix in a bar element captures the relationship between the nodal displacements and external forces acting on the element. It is mathematically represented in a matrix form for a bar element with length L, cross-sectional area A, and Young’s modulus E, as: [k] = (EA/L) [1 -1; -1 1], where the entries determine how displacement at each node contributes to forces at other nodes. This matrix is crucial in determining how the element deforms under loading and in assembling the global stiffness matrix for larger systems .

Mesh quality factors significantly influence the accuracy of FEM analyses. The aspect ratio describes the relative proportions of an element's dimensions, and a poor aspect ratio can lead to errors in the solution by misrepresenting how loads and stresses distribute. Skewness refers to how much an element's shape deviates from ideal shapes like equilateral triangles or squares; high skewness can cause numerical inaccuracies. Element distortion, which indicates how much an element has been stretched or compressed from its original shape, can lead to inaccurate results and convergence issues. Proper attention to these factors ensures more reliable simulations and reduced computational errors .

The evolution of computational power and software has significantly expanded the application scope of FEM. Originally developed in the 1940s and 1950s for structural analysis, advancements in computational technologies have enabled FEM to handle increasingly complex problems across various fields such as mechanical, aerospace, biomedical, and electrical engineering. Enhanced computational capabilities have allowed for the simulation of larger and more intricate models with finer meshing, improving accuracy and broadening the range of applications, including biomechanics and geotechnical engineering .

The Finite Element Method (FEM) offers several advantages in engineering applications, such as the ability to handle complex geometries and boundary conditions, and applicability to various engineering problems like structural analysis and fluid dynamics. It provides accurate results, given proper meshing and modeling efforts. However, FEM has limitations, including being computationally expensive for large-scale problems and needing significant expertise in model setup and interpretation. Its accuracy is heavily dependent on the quality of the mesh, and it faces challenges in solving problems with discontinuities or singularities without special techniques .