Forward, Backward, and Central Difference Schemes

Forward, Backward, and Central Difference Schemes

Introduction to Finite Differences

In CFD, partial differential equations (PDEs) governing fluid flow and heat transfer are approximated using finite difference methods. This involves replacing continuous variables and derivatives with discrete values at grid points, converting a calculus problem into an algebraic one.

Uniform Grid Concept

In a uniform grid, spacing \( \Delta x \) and \( \Delta y \) is constant. We denote the value of a function \( u \) at grid point \( (i\Delta x, j\Delta y) \) as \( u_{i,j} \).

Forward Difference Approximation

  • Formula: \[ \left(\frac{\partial u}{\partial x}\right)_{i,j} \approx \frac{u_{i+1,j} - u_{i,j}}{\Delta x} \]
  • Truncation Error: \( O(\Delta x) \)

Backward Difference Approximation

  • Formula: \[ \left(\frac{\partial u}{\partial x}\right)_{i,j} \approx \frac{u_{i,j} - u_{i-1,j}}{\Delta x} \]
  • Truncation Error: \( O(\Delta x) \)

Central Difference Approximation (First Derivative)

  • Formula: \[ \left(\frac{\partial u}{\partial x}\right)_{i,j} \approx \frac{u_{i+1,j} - u_{i-1,j}}{2\Delta x} \]
  • Truncation Error: \( O(\Delta x^2) \)

Central Difference Approximation (Second Derivative)

  • Formula: \[ \left(\frac{\partial^2 u}{\partial x^2}\right)_{i,j} \approx \frac{u_{i+1,j} - 2u_{i,j} + u_{i-1,j}}{(\Delta x)^2} \]
  • Truncation Error: \( O(\Delta x^2) \)

Higher-Order Central Difference Schemes

Fourth-Order Five-Point Formula:

\[ \left(\frac{\partial^2 u}{\partial x^2}\right)_{i,j} \approx \frac{-u_{i+2,j} + 16u_{i+1,j} - 30u_{i,j} + 16u_{i-1,j} - u_{i-2,j}}{12(\Delta x)^2} \]

Truncation Error: \( O(\Delta x^4) \)

Sixth-Order Seven-Point Formula:

\[ \left(\frac{\partial^2 u}{\partial x^2}\right)_{i,j} \approx \frac{-u_{i+3,j} + 12u_{i+2,j} - 39u_{i+1,j} + 56u_{i,j} - 39u_{i-1,j} + 12u_{i-2,j} - u_{i-3,j}}{360(\Delta x)^2} \]

Truncation Error: \( O(\Delta x^6) \)

Fourth-Order Compact Three-Point Formula (Implicit):

\[ u''_{i,j} + \frac{1}{6} u''_{i+1,j} + \frac{1}{6} u''_{i-1,j} = \frac{u_{i+1,j} - 2u_{i,j} + u_{i-1,j}}{(\Delta x)^2} \]

Solving this implicit relation yields higher accuracy with a compact stencil.

Key Concepts and Properties

  • Truncation Error (T.E.): Difference between the actual derivative and its discrete approximation due to Taylor series truncation.
  • Order of Accuracy: If truncation error is proportional to \( (\Delta x)^n \), the method is said to be of order \( n \).
  • Consistency: A scheme is consistent if T.E. vanishes as \( \Delta x \to 0 \).
  • Difference Molecule: A visual representation of the grid stencil used (e.g., i−1, i, i+1 for central differences).
Conclusion: Central, forward, and backward difference schemes form the foundation of numerical methods in CFD. Higher-order schemes provide improved accuracy but often require more complex handling near boundaries.