Week 9 - FVM Literature Review

1. What is Finite Volume Method(FVM)?

Ans. The Finite Volume Method (FVM) is a numerical technique used in Computational Fluid Dynamics (CFD) to solve partial differential equations (PDEs) that govern fluid flow and other physical phenomena. It discretizes the domain into a collection of finite control volumes or cells, with each cell representing a small portion of the computational domain. The FVM focuses on the integral form of the governing equations and employs a conservation law approach.

The Finite Volume Method offers several advantages, including conservation of mass, energy, and other quantities at the discrete level. It is particularly suitable for complex geometries and unstructured grids. By dividing the domain into control volumes, it allows for a flexible representation of the flow variables within each volume. However, it does require careful treatment of the numerical schemes, boundary conditions, and the selection of appropriate control volumes to ensure accurate and stable solutions.

2. Write down the major differences between FDM & FVM.

Ans. Both FDM and FVM have their advantages and limitations, and their suitability depends on the specific problem and computational requirements. FVM is generally considered more versatile for complex geometries and unstructured grids, while FDM is often preferred for simpler geometries and structured grids due to its ease of implementation and numerical efficiency. Following are the major differeneces between FDM and FVM -

1. Control Volume Approach: FVM is based on a control volume approach, where the computational domain is divided into discrete control volumes. In contrast, FDM is based on a grid-based approach, where the domain is divided into a grid of points or cells.

2. Conservation Form: FVM is inherently a conservative method as it focuses on the integral form of the governing equations. It ensures conservation of mass, momentum, and energy within each control volume. On the other hand, FDM can be both conservative and non-conservative, depending on how the equations are discretized.

3. Integral and Differential Equations: FVM directly integrates the governing equations over control volumes to obtain algebraic equations, while FDM discretizes the equations by approximating the derivatives using finite differences. FDM operates on a pointwise basis, whereas FVM operates on a cell-based basis.

4. Grid Types: FVM can handle both structured and unstructured grids. It allows for flexible grid generation and adaptation, making it suitable for complex geometries. FDM, however, is primarily designed for structured grids, where the points are regularly spaced in a fixed pattern.

5. Numerical Fluxes: In FVM, the fluxes across the cell faces are evaluated using numerical interpolation schemes to estimate the values at the faces. This allows for greater flexibility in capturing flow features at cell boundaries. FDM, on the other hand, approximates the derivatives at each grid point using finite differences, without explicitly considering the fluxes at cell interfaces.

6. Treatment of Boundary Conditions: FVM treats the boundary conditions by applying them directly to the control volumes at the domain boundaries. This allows for a natural incorporation of different types of boundary conditions. FDM often requires additional techniques, such as ghost cells or extrapolation, to handle boundary conditions.

3. Describe the need for interpolation schemes and flux limiters in FVM using the 1D linear heat conduction equation as an example.

Ans. In the Finite Volume Method, interpolation schemes and flux limiters play crucial roles in accurately capturing the behavior of the solution across control volume interfaces. Consider the 1D linear heat conduction equation as an example:

`(delT)/(delt) = del/(delx)(k(delT)/(delx))`

where T is the temperature, t is time, x is the spatial coordinate, and k is the thermal conductivity.

Interpolation Schemes: In FVM, the control volumes divide the domain into discrete regions. To evaluate the fluxes across the control volume interfaces, interpolation schemes are used to estimate the values of the solution variables at these interfaces. In the context of the heat conduction equation, we need to interpolate the temperature T from the cell centers to the control volume faces.

The interpolation scheme is crucial as it affects the accuracy and stability of the numerical solution. An inadequate interpolation scheme may introduce excessive numerical diffusion or oscillations, leading to inaccurate results.

Flux Limiters: Flux limiters are employed to control the propagation of numerical errors, particularly in regions with steep gradients or discontinuities. They are used to limit the values of the numerical fluxes across control volume interfaces, ensuring that the solution remains bounded and oscillation-free.

In the case of the above heat conduction equation, a flux limiter can be used to control the gradients of the temperature. For instance, the Minmod limiter is a commonly used flux limiter. It compares the slopes of neighboring cells and selects the smallest magnitude to limit the flux. This effectively reduces spurious oscillations and prevents unphysical solutions.By applying flux limiters, the FVM can maintain the stability and accuracy of the solution in the presence of sharp gradients or discontinuities, ensuring a more realistic representation of the temperature distribution.

In summary, interpolation schemes and flux limiters are essential in FVM to accurately estimate the values of solution variables at control volume interfaces and control numerical errors. In the context of the 1D linear heat conduction equation, interpolation schemes help estimate the temperature values at control volume faces, while flux limiters control the fluxes to maintain stability and prevent oscillations in regions with steep gradients or discontinuities.