Finite Element Method: Principles, Applications and Accuracy

Concept of FEM and Procedure

The Finite Element Method (FEM) is a numerical technique used to solve complex engineering problems by dividing a large system into smaller, simpler parts called finite elements.

Instead of solving the whole structure at once, FEM solves each small element and combines the results to get an approximate solution.

Procedure of FEM

  1. Discretization (Meshing): Divide the entire domain into small elements.
  2. Selection of Element Type: Choose suitable elements (1D, 2D, 3D).
  3. Derivation of Element Equations: Form equations for each element using physical laws.
  4. Assembly: Combine all element equations into a global system.
  5. Apply Boundary Conditions: Apply constraints and loads.
  6. Solution: Solve equations to find unknown values (displacement, temperature, etc.).
  7. Post-Processing: Interpret results like stress, strain, and deformation.

Advantages and Disadvantages of FEM

Advantages

  • Can solve complex shapes and geometries.
  • Suitable for non-uniform materials.
  • Handles complex boundary conditions.
  • Gives accurate approximate results.
  • Widely used in computer software (e.g., ANSYS).

Disadvantages

  • Requires high computational power.
  • Results are approximate, not exact.
  • Needs skilled knowledge.
  • Meshing errors can affect accuracy.
  • Time-consuming for large problems.

Engineering Applications of FEM

FEM is used in many engineering fields:

  • Mechanical Engineering: Stress analysis, vibration analysis, heat transfer.
  • Civil Engineering: Structural analysis of buildings, bridges, and dams.
  • Aerospace Engineering: Aircraft wing design, fatigue analysis.
  • Automobile Engineering: Crash analysis, engine component design.
  • Electrical Engineering: Electromagnetic field analysis.
  • Biomedical Engineering: Bone stress analysis, implants design.

Difference Between FDM and FEM

Finite Difference Method (FDM)Finite Element Method (FEM)
FDM gives values only at nodal points.FEM gives values at any point (including nodes).
Interpolation is required for other points.Uses shape functions for interpolation.
Requires a large number of nodes for accuracy.Requires fewer nodes for good accuracy.
Suitable for simple problems.Suitable for complex problems.
Difficult to handle curved or irregular boundaries.Can easily handle curved and irregular boundaries.
Uses point-wise approximation.Uses piece-wise approximation.
Continuity only at nodes.Continuity at nodes and along element boundaries.

P and H Formulations in FEM

FEM is an approximate numerical method, so some errors are always present. The accuracy depends on:

  • Discretization (meshing)
  • Shape function (interpolation function)

To improve accuracy, the following methods are used:

1. h-version (h-refinement)

Here, h represents the size of the element. Accuracy is improved by reducing element size.

  • Mesh is refined into smaller elements.
  • Number of elements increases.
  • Shape function order remains constant.
  • Widely used in practice.

👉 In short: Smaller elements → better accuracy.

2. p-version (p-refinement)

Here, p represents the degree of the polynomial.

  • Polynomial order is increased.
  • Mesh (number of elements) remains the same.
  • Accuracy improves without remeshing.
  • Additional nodes may be introduced.

👉 In short: Higher polynomial order → better accuracy.

3. r-version (Rearrangement Method)

Nodes are repositioned without changing mesh size.

  • Number of elements remains the same.
  • Polynomial degree remains the same.
  • Nodes are shifted to important regions.
  • Improves accuracy efficiently.

👉 In short: Better node placement → better results.

bWBRCjMBNgAAAABJRU5ErkJggg==