Fact-checked by Grok 2 weeks ago

MacCormack method

The MacCormack method is a explicit finite-difference predictor-corrector scheme widely employed in (CFD) for numerically solving partial differential equations, such as the Euler or Navier-Stokes equations governing compressible, inviscid or viscous flows. Introduced by Robert W. MacCormack in 1969 while at , it advances solutions through time using a forward-difference predictor step (typically forward in time and backward in space) to generate an intermediate estimate, followed by a backward-difference corrector step (forward in space) that averages the original and predicted values for refinement, thereby achieving second-order accuracy in both space and time. The method's formulation preserves the conservation form of the governing equations when implemented accordingly, making it suitable for capturing discontinuities like shock waves, though artificial is often added to suppress oscillations near shocks. For the simple linear equation \frac{\partial u}{\partial t} + a \frac{\partial u}{\partial x} = 0, the predictor step computes an intermediate value \tilde{u}_i^{n+1} = u_i^n - \frac{a \Delta t}{\Delta x} (u_i^n - u_{i-1}^n), while the corrector yields u_i^{n+1} = \frac{1}{2} \left[ u_i^n + \tilde{u}_i^{n+1} - \frac{a \Delta t}{\Delta x} (\tilde{u}_{i+1}^{n+1} - \tilde{u}_i^{n+1}) \right], demonstrating its extension to nonlinear systems via flux differencing. Stability of the scheme is conditional, governed by the Courant-Friedrichs-Lewy (CFL) criterion, which requires the Courant number C = \frac{|u| + a}{\Delta x / \Delta t} \leq 1 (where u is and a is speed), ensuring information does not propagate beyond adjacent grid points per time step; violations lead to . Its dissipative nature, similar to the Lax-Wendroff method but with greater flexibility in differencing directions, makes it effective for time-marching toward steady-state solutions in applications like quasi-one-dimensional flows (with errors of 0.3–3.3% against analytical results on coarse s) and supersonic layers. Historically prominent for about 15 years following its debut—applied to impact cratering and blunt-body —the MacCormack method's simplicity and ease of programming on early computers established it as a for explicit schemes, though it has since been supplanted by higher-order, more robust methods like TVD or ENO schemes for complex geometries and long-time integrations. Variants, including implicit formulations and extensions to , continue to inform modern CFD practices.

Introduction

Overview

The MacCormack method is an explicit, second-order accurate scheme designed for solving laws. It advances solutions in time using a predictor-corrector structure, where an intermediate prediction step is refined by a correction based on differenced values to enhance accuracy. This approach ensures second-order precision in both space and time while maintaining computational efficiency. The method's explicit formulation makes it well-suited for nonlinear partial differential equations, including the Euler equations governing inviscid compressible flows in . It applies forward and backward differences alternately in the predictor and corrector phases to propagate information stably. For , the Courant-Friedrichs-Lewy (CFL) number must remain below 1. Among its primary advantages are simplicity in formulation and ease of implementation in code, rendering it accessible for a wide range of simulations. Additionally, it effectively captures shocks inherent to systems without excessive numerical dissipation, balancing resolution and robustness.

Historical Development

The MacCormack method, a second-order accurate explicit finite-difference scheme, was introduced by Robert W. MacCormack in 1969 while he was a research scientist at . Developed as a predictor-corrector approach to solve time-dependent partial differential equations, it emerged from efforts to numerically simulate complex compressible flows using early computational resources. MacCormack's innovation built on prior schemes like Lax-Wendroff but simplified the process by avoiding explicit evaluations, making it more practical for implementation on the limited computers of the era. This development occurred amid NASA's intensive research into supersonic and hypersonic , driven by challenges in re-entry and planetary entry vehicles. At Ames, which led early (CFD) initiatives, scientists sought reliable methods to model shock waves, boundary layers, and viscous effects in high-speed flows without relying solely on experiments. MacCormack's method addressed these needs by providing an efficient tool for the Euler and Navier-Stokes equations, initially applied to axisymmetric hypervelocity cratering problems relevant to meteoroid protection for . The initial publication appeared as AIAA Paper 69-354, presented at the AIAA , where it demonstrated second-order accuracy in both time and space for compressible viscous flows. MacCormack's contributions gained formal recognition in 1981 when he received NASA's Medal for Exceptional Scientific Achievement for advancing CFD techniques. By then, the method had become a in the field, influencing the evolution of numerical algorithms for through the 1970s. Its simplicity and robustness facilitated widespread adoption, paving the way for modern high-resolution schemes and methods in simulating and supersonic flows. Seminal applications at , such as shock-boundary layer interactions, underscored its role in establishing CFD as a vital complement to physical testing in design.

Mathematical Formulation

Governing Equations

The MacCormack method is designed to numerically solve systems of partial equations in form, particularly those modeling wave propagation phenomena in . The general form of such laws in one is given by \frac{\partial \mathbf{U}}{\partial t} + \frac{\partial \mathbf{F}(\mathbf{U})}{\partial x} = 0, where \mathbf{U}(x,t) is the of conserved variables, and \mathbf{F}(\mathbf{U}) is the corresponding that depends on \mathbf{U}. This form ensures that the numerical scheme preserves the physical principles when integrated over control volumes, making it suitable for capturing discontinuities like shocks. A application of the MacCormack method is to the one-dimensional Euler equations governing inviscid , which express the , momentum, and total : \frac{\partial}{\partial t} \begin{pmatrix} \rho \\ \rho u \\ E \end{pmatrix} + \frac{\partial}{\partial x} \begin{pmatrix} \rho u \\ \rho u^2 + p \\ u(E + p) \end{pmatrix} = 0. Here, \rho denotes , u is the , E = \frac{1}{2} \rho u^2 + \rho e is the total energy per volume with e the specific , and p = (\gamma - 1) \rho e is the assuming an with constant ratio of specific heats \gamma. These equations arise in scenarios such as shock waves and supersonic flows, where is negligible. The character of these equations stems from their real , which define curves along which information propagates at finite speeds corresponding to physical wave , such as material and speed. This wave-like behavior implies that perturbations or discontinuities travel along these characteristics without instantaneous influence across the domain, often leading to the formation of fronts that require , non-oscillatory numerical approximations—either upwind or centered schemes with controlled . While the MacCormack method originated as a scheme applied directly to values, it aligns naturally with finite volume interpretations by integrating the conservation laws over cells to compute cell-averaged updates, thereby maintaining integral conservation properties.

Discretization Approach

The MacCormack method discretizes partial differential equations, such as the governing conservation laws, using approximations on a uniform spatial . The one-dimensional spatial domain is partitioned into evenly spaced points x_i = i \Delta x for i = 0, 1, \dots, N, where \Delta x is the constant grid spacing, while time is advanced in discrete levels t^n = n \Delta t, with \Delta t denoting the time step size. The numerical solution at these grid points is represented by U_i^n, which approximates the exact solution U(x_i, t^n). This structured setup facilitates straightforward implementation of difference operators while capturing wave propagation in systems. Spatial derivatives in the discretized equations are approximated using one-sided s to leverage the predictor-corrector structure for enhanced accuracy and . The forward , biased to the right, is defined as \delta^+ F_i = \frac{F_{i+1} - F_i}{\Delta x}, providing a second-order accurate to the \frac{\partial F}{\partial x} at x_i using values from the forward direction. Conversely, the backward , biased to the left, is given by \delta^- F_i = \frac{F_i - F_{i-1}}{\Delta x}, which similarly approximates the using values from the backward direction. These one-sided operators are integral to the method's ability to handle nonlinear fluxes in form without introducing excessive numerical dissipation. By applying these operators alternately in the temporal evolution, the discretization transforms the continuous problem into a sequence of algebraic relations solvable explicitly, ensuring the method's suitability for time-dependent simulations of compressible flows.

Algorithm Description

Predictor Phase

The predictor phase of the MacCormack method constitutes the initial step in this explicit, second-order scheme for solving hyperbolic systems of conservation laws, such as the Euler equations in . It generates an intermediate approximation of the vector at the next time level by employing one-sided backward spatial differencing on the fluxes computed from the current time level values. This step serves to estimate the temporal evolution of the conserved variables across the computational grid, providing a provisional that captures propagation in an upwind manner for positive speeds. The choice of backward or forward differencing is selected to align with the propagation direction of characteristics for improved . For a one-dimensional of the form \partial U / \partial t + \partial F(U) / \partial x = 0, where U is the of conserved variables and F(U) is the , the predictor formula at grid point i and time level n is given by \tilde{U}_i^{n+1} = U_i^n - \frac{\Delta t}{\Delta x} \left( F_i^n - F_{i-1}^n \right), with \Delta t and \Delta x denoting the time step and spatial spacing, respectively. If source terms are present, \partial U / \partial t + \partial F(U) / \partial x = S(U), add \Delta t \, S(U_i^n) to the right-hand side. Here, the fluxes F_i^n = F(U_i^n) and F_{i-1}^n = F(U_{i-1}^n) are evaluated directly at the grid points using the known solution values from time level n, accommodating nonlinear functions inherent to systems like equations. This backward differencing approximates the spatial derivative \partial F / \partial x as (F_i^n - F_{i-1}^n)/\Delta x, yielding first-order accuracy in this phase alone, which is later refined in the subsequent corrector step. In the context of multidimensional problems or more complex geometries, the predictor phase extends analogously by applying backward differences along each spatial direction to the respective flux components, while intermediate fluxes based on the predicted \tilde{U} values may be computed to facilitate corrections, though the core estimation remains rooted in the upwind provisional solution. This approach ensures computational efficiency and simplicity, as the fluxes are handled without requiring Riemann solvers, making it particularly suitable for nonlinear problems where direct evaluation at grid points preserves the method's explicit nature.

Corrector Phase

The corrector phase of the MacCormack method refines the intermediate solution obtained from the predictor phase by applying a forward to the spatial derivatives, thereby compensating for the backward-biased errors introduced in the initial estimate. This step evaluates the flux function using the predicted values and incorporates a forward differencing to the solution vector. For the \partial \mathbf{U}/\partial t + \partial \mathbf{F}(\mathbf{U})/\partial x = 0, the final solution at the new time level is obtained by \mathbf{U}_i^{n+1} = \frac{1}{2} \left[ \mathbf{U}_i^n + \tilde{\mathbf{U}}_i^{n+1} - \frac{\Delta t}{\Delta x} \left( \mathbf{F}_{i+1}^{n+1} - \mathbf{F}_i^{n+1} \right) \right], where \mathbf{F}^{n+1} = \mathbf{F}(\tilde{\mathbf{U}}^{n+1}) denotes the flux evaluated at the predicted intermediate state \tilde{\mathbf{U}}^{n+1}, \Delta t is the time step, and \Delta x is the spatial grid spacing. This formulation combines the original solution, the predicted intermediate state, and the forward flux difference, effectively balancing the one-sided approximations to mitigate leading-order truncation errors from the predictor step. The forward differencing corrects these errors by introducing an opposing bias that, upon averaging, yields a more symmetric and accurate representation of the spatial derivatives. When source terms are present in the governing , such as \partial \mathbf{U}/\partial t + \partial \mathbf{F}(\mathbf{U})/\partial x = \mathbf{S}(\mathbf{U}), they are incorporated by averaging the source evaluated at the current and predicted states: add \frac{\Delta t}{2} \left[ \mathbf{S}(\mathbf{U}_i^n) + \mathbf{S}(\tilde{\mathbf{U}}_i^{n+1}) \right] to the right-hand side of the corrector formula, ensuring consistency with the overall scheme. This treatment maintains the method's explicit nature while accounting for additional physical effects like reaction or body forces.

Theoretical Analysis

Stability Conditions

The stability of the MacCormack method applied to linear hyperbolic equations, such as the equation, is determined through analysis, which examines the amplification factor of modes in the numerical solution. This analysis reveals that the scheme is conditionally stable, requiring the Courant-Friedrichs-Lewy (CFL) number to satisfy \left| \lambda \frac{\Delta t}{\Delta x} \right| \leq 1, where \lambda is the advection velocity, \Delta t is the time step, and \Delta x is the spatial step. Violation of this condition leads to unbounded growth of high-frequency modes, resulting in numerical . For nonlinear hyperbolic systems, the stability criterion extends the linear case by basing the CFL number on the maximum wave speed across the computational domain, again requiring \left| \lambda_{\max} \frac{\Delta t}{\Delta x} \right| \leq 1, where \lambda_{\max} is the largest absolute eigenvalue of the system's matrix. This local approach provides a practical guideline, though rigorous nonlinear proofs are challenging due to variable coefficients and source terms; empirical adjustments often impose stricter limits, such as CFL < 0.5, to prevent instability in regions of steep gradients. A notable limitation of the MacCormack method is its tendency to produce non-physical oscillations near discontinuities like shocks. This issue stems from the central differencing in the corrector step and lack of inherent upwinding, amplifying dispersive errors that manifest as post-shock oscillations. Such behavior can degrade solution quality unless mitigated by artificial viscosity or flux limiters. The CFL restriction fundamentally impacts computational efficiency, as it mandates small time steps proportional to the spatial resolution and maximum wave speed, increasing the total number of iterations and thus the overall runtime for time-dependent simulations. In practice, this constraint limits the method's applicability to problems with moderate Courant numbers, often requiring adaptive time stepping or implicit variants to enhance efficiency without sacrificing stability.

Accuracy and Error Analysis

The MacCormack method is a second-order accurate finite difference scheme for solving hyperbolic partial differential equations, with a local truncation error of O(\Delta t^3 + \Delta x^3), leading to global accuracy of O(\Delta t^2 + \Delta x^2) under a constant . This order is established through Taylor series expansions applied to the predictor and corrector steps: in the forward (predictor) phase, the one-sided difference approximation introduces errors involving third-order time and space derivatives, while the backward (corrector) phase averages these with the original values, canceling the leading first- and second-order terms to yield the higher-order accuracy. Analysis of the modified equation reveals that the leading error terms in the MacCormack scheme arise from even-order spatial derivatives, which contribute to numerical dissipation (analogous to artificial viscosity that damps high-frequency modes), and odd-order derivatives, which introduce numerical dispersion (causing phase errors and wave speed distortions). These effects are quantified via von Neumann stability analysis, where the amplification factor's imaginary part reflects dispersion and the real part dissipation, with coefficients depending on the CFL number. In comparison to first-order upwind schemes like Lax-Friedrichs, which suffer from excessive numerical viscosity due to dominant second-order dissipative terms (O(\Delta x) accuracy), the MacCormack method provides significantly reduced dissipation, enabling better preservation of smooth gradients and fewer grid points per wavelength (typically 6-8 versus 10+ for Lax-Friedrichs). However, its linear formulation lacks monotonicity-preserving properties, leading to non-physical oscillations near under-resolved discontinuities such as shocks, where dispersion errors amplify Gibbs-like phenomena without additional limiters.

Example and Implementation

Linear Advection Problem

The linear equation serves as a fundamental test case for numerical methods solving hyperbolic partial differential equations, given by \frac{\partial u}{\partial t} + a \frac{\partial u}{\partial x} = 0, where a > 0 is the constant advection speed and u = u(x, t) represents the advected quantity. This equation models the transport of a at constant velocity without or in the continuous case, with the exact being a pure of the u(x, 0). To apply the MacCormack method, the domain is discretized on a uniform grid with spatial step \Delta x and time step \Delta t, where the Courant number is \lambda = a \Delta t / \Delta x. The predictor phase uses a forward-time forward-space differencing : \tilde{u}_i^{n+1} = u_i^n - \lambda (u_{i+1}^n - u_i^n). This provides an intermediate approximation at the next time level. The corrector phase then employs forward-time backward-space differencing on the predicted values, followed by averaging with the original values: u_i^{n+1} = \frac{1}{2} \left[ u_i^n + \tilde{u}_i^{n+1} - \lambda (\tilde{u}_i^{n+1} - \tilde{u}_{i-1}^{n+1}) \right]. This two-step process yields the updated solution at grid point i and time level n+1. For smooth solutions, the MacCormack method applied to this equation achieves exact second-order accuracy in both space and time, with a local truncation error of O(\Delta t^3 + \Delta x^3). This equivalence to the Lax-Wendroff scheme in the linear case ensures low numerical dissipation and dispersion errors for well-resolved waves. Numerical experiments with periodic initial conditions, such as a Gaussian bell profile on a periodic domain, demonstrate the method's ability to preserve the solution shape while advecting it at the correct speed. For instance, as the grid resolution increases (e.g., from 50 to 400 points), the error converges at second order, with minimal dispersion evident in the smooth translation of the profile over multiple periods, though slight oscillations may appear near the peaks for coarser grids.

Numerical Implementation Notes

Implementing the MacCormack method in one dimension typically involves a straightforward explicit scheme on a uniform grid, suitable for solving hyperbolic conservation laws such as the linear . The algorithm proceeds in a time-marching loop, where each step consists of a predictor using forward spatial differencing followed by a corrector using backward spatial differencing. Below is a outline for a 1D implementation, assuming a scalar \partial_t u + \partial_x f(u) = 0 discretized on N interior grid points with spacing \Delta x:
Initialize u[0:N+1] with initial conditions and boundary values
Set t = 0
While t < T_final:
    # Predictor step (forward differencing)
    for i = 1 to N:
        u_pred[i] = u[i] - (dt / dx) * (f(u[i+1]) - f(u[i]))
    
    # Corrector step (backward differencing on predicted values)
    for i = 1 to N:
        u[i] = 0.5 * (u[i] + u_pred[i] - (dt / dx) * (f(u_pred[i]) - f(u_pred[i-1])))
    
    # Update boundaries (see below)
    UpdateBoundaries(u)
    
    t = t + dt
This structure ensures second-order accuracy in space and time while remaining simple to code. Boundary conditions must be carefully handled to maintain stability and accuracy, particularly since the forward and backward sweeps require values outside the domain. For inflow boundaries (e.g., left side at i=0), specify the incoming characteristic variables directly from physical conditions, such as fixed values for subsonic inflow. For outflow boundaries (e.g., right side at i=N+1), use linear extrapolation of interior values to avoid reflections. Ghost cells can facilitate this: extend the grid with one or two extra points on each side, setting ghost values via one-sided differences or extrapolation for the predictor (forward sweep) and corrector (backward sweep). For example, in the linear advection case with constant speed a > 0, the left ghost cell u{{grok:render&&&type=render_inline_citation&&&citation_id=0&&&citation_type=wikipedia}} is set to the inflow value, while the right ghost cell u[N+1] is extrapolated as u[N+1] = u[N] + (u[N] - u[N-1]). This approach ensures consistent differencing without modifying the interior loop. The time step \Delta t is chosen to satisfy the Courant-Friedrichs-Lewy (CFL) condition for stability, given by \Delta t = \mathrm{CFL} \cdot \Delta x / |a|_{\max}, where |a|_{\max} is the maximum wave speed (e.g., velocity or sound speed) across the domain, and the CFL number is typically set to 0.9 to balance stability and efficiency while avoiding oscillations near 1.0. In practice, compute \Delta t dynamically at each step based on current solution values to adapt to varying speeds. Due to its explicit nature, the MacCormack method has a computational cost of O(N) operations per time step for N grid points, primarily from the two sweeps over the in predictor and corrector phases. This linear scaling supports efficient on modern , such as using array operations in languages like or , making it suitable for moderate grid sizes without parallelization overhead.

Applications and Variants

Use in Computational Fluid Dynamics

The MacCormack method has been widely applied to solve the one-dimensional and two-dimensional Euler equations in , particularly for simulating inviscid compressible flows involving discontinuities such as shocks and expansion fans. These equations, which describe the , momentum, and energy, are discretized using the method's predictor-corrector approach to capture wave propagation accurately in Riemann problems, where initial discontinuities evolve into complex flow structures. For instance, in one-dimensional simulations, the method resolves the interaction of shock waves and waves effectively, as demonstrated in tests with calorically perfect gases and dissociating , maintaining stability under CFL conditions less than 1. In practical applications from the 1970s, the MacCormack method was integrated into codes for simulating high-speed flows, including problems and supersonic flows. Similarly, in two-dimensional supersonic simulations, the method has been employed to predict flow acceleration to 1.5, including shocks and Prandtl-Meyer expansions over ramps or contoured walls, providing reliable results for propulsion system analysis. These examples highlight its role in early CFD for problems, where explicit time-marching allowed efficient computation on limited hardware of the era. A key advantage of the MacCormack method in early CFD was its ability to handle discontinuities with relatively low numerical dissipation compared to the Lax-Wendroff scheme, preserving sharper shock profiles while introducing manageable dispersive oscillations that could be controlled. This made it preferable for simulations requiring resolution of fine-scale features without excessive smearing, contributing to its adoption in standard toolsets during the 1970s and 1980s. However, in multidimensional applications, the method exhibits sensitivity to alignment, particularly when strong shocks are not parallel to grid lines, leading to increased oscillations or inaccuracies that necessitate careful design. To mitigate these issues and stabilize solutions near discontinuities, artificial terms are often added, enhancing monotonicity but introducing tunable parameters.

Extensions and Modifications

One significant extension of the MacCormack method addresses its conditional by integrating it with the backward-forward compensation and correction (BFECC) , rendering it unconditionally stable while preserving second-order accuracy in space and time. This modification, proposed by Selle et al. in 2007, reformulates the MacCormack scheme to explicitly use BFECC-style estimation for correcting forward errors, allowing larger time steps without instability in applications like smoke simulation and . To mitigate excessive numerical in low-Mach number flows, a generalized MacCormack scheme was developed post-2010, incorporating limiters and preconditioning to scale appropriately. Gallagher et al. (2017) introduced this variant within a dual-time framework, splitting into convective and acoustic components, applying central differencing with JST-type selectively to terms, and using modified eigenvalues for preconditioning, which suppresses oscillations and improves accuracy for unsteady low-Mach flows like vortex . Beyond traditional , the MacCormack method has been adapted for simulating incorporating infiltration processes, particularly in hydrological overland flow modeling since 2013. For instance, Ngon et al. (2019) extended the scheme to handle terms for spatially infiltration and microtopography using fractional steps and upwind biasing, enabling accurate prediction of discontinuous flows over wettable and dry beds in 1D and 2D settings. Similarly, for viscous flows, adaptations involve scaling to balance explicit treatment of viscous terms at high Re, as in MacCormack's 1981 method for compressible viscous equations, which uses implicit analogs for across Re regimes without excessive . Multidimensional extensions of the MacCormack method typically employ dimension-by-dimension application or operator splitting to efficiently handle higher dimensions without full tensor operations. MacCormack's original time-splitting approach (1972) decomposes the multidimensional problem into sequential one-dimensional solves along each coordinate, allowing larger time steps and maintaining second-order accuracy, as further refined in three-level explicit schemes for convection-diffusion equations. Hybrid schemes combining the MacCormack method with (TVD) techniques enhance monotonicity preservation near shocks. Liang et al. (2007) developed a TVD-MacCormack for routing, incorporating flux limiters to eliminate oscillations while retaining the predictor-corrector structure, proving effective for rapidly varying flows with hydraulic jumps. More recently, implicit formulations of the MacCormack method have been applied to supersonic turbulent jet flows (2020) and phenomena in pipelines (2024).

References

  1. [1]
    None
    Below is a merged summary of MacCormack's Method based on all the provided segments. To retain all the detailed information in a dense and organized manner, I will use a table in CSV format for key details, followed by a narrative summary that integrates additional context and explanations. This approach ensures comprehensive coverage while maintaining clarity and conciseness.
  2. [2]
    [PDF] AIAA 81-0110R A Numerical Method for Solvingthe Equations of ...
    This paper describes one such development. A new method for solving these equations has been devised that I) is second-order.Missing: PDEs | Show results with:PDEs
  3. [3]
  4. [4]
  5. [5]
    The effect of viscosity in hypervelocity impact cratering
    The effect of viscosity in hypervelocity impact cratering. R. MACCORMACK. R ... 28 April 1969 - 30 April 1969. Cincinnati,OH,U.S.A.. https://doi.org ...
  6. [6]
    Landmarks and new frontiers of computational fluid dynamics
    Jan 31, 2019 · The vision of a systematic CFD development for aerospace community was crystalized by NASA Ames Research Center in the later 1960's, and ...
  7. [7]
    Numerical Computation of Compressible and Viscous Flow
    Jul 22, 2014 · Numerical Computation of Compressible and Viscous Flow is written for those who want to calculate compressible and viscous flow past aerodynamic bodies.
  8. [8]
    Three decades of accomplishments in computational fluid dynamics
    Early milestones. In 1969, MacCormack [26] published his first of many landmark numerical algorithm ... The theoretical division of NASA contributed many ...
  9. [9]
    [PDF] A Study of Numerical Methods for Hyperbolic Conservation Laws ...
    In order to avoid oscillations near discontinuities, MacCormack's method can be modified by adding a flux-correction step motivated by the theory of TVD methods ...Missing: scheme | Show results with:scheme
  10. [10]
    [PDF] Finite Volume Analysis with the MacCormack Method Applied to ...
    As metal forming flow is governed by Conservations Differential Equations, this work proposes to use the MacCormack method to solve the governing equations of ...
  11. [11]
    [PDF] Euler equations
    Euler equations. For incompressible flow the inviscid 1D Euler equations decouple to: ρt + uρx = 0 ut + px ρ. = 0 et + uex = 0. The 3D Euler equations are given ...
  12. [12]
    [PDF] Fast Euler Solver for Steady, One-Dimensional Flows
    A numerical technique to solve the Euler equations for steady, one-dimensional flows is presented. The technique is essentially implicit, but is structured ...
  13. [13]
    [PDF] Chapter 7: Hyperbolic equations - UC Davis Mathematics
    Hyperbolic PDEs model waves, like acoustic, elastic, electromagnetic, or gravitational waves. They have finite domains of influence and dependence.
  14. [14]
    [PDF] An Unconditionally Stable MacCormack Method
    Jun 7, 2007 · Abstract. The back and forth error compensation and correction (BFECC) method advects the solution forward and then backward in time.
  15. [15]
    [PDF] Chapter 3. Finite Difference Methods for Hyperbolic Equations 3.1 ...
    See Section 2.3. MacCormack (an example of two-step predictor-corrector method). Predictor: 1 *. 1. ( ) n n n n i i i i u u u. u c t x. +. + −. = − ∆. ∆.Missing: formula | Show results with:formula
  16. [16]
    [PDF] NUMERICAL SOLUTION OF PARTIAL DIFFERENTIAL EQUATIONS ...
    Jul 4, 2023 · This course covers numerical solutions of partial differential equations, including introduction, separation of variables, Fourier series, ...
  17. [17]
    [PDF] The advection equation
    ⇒ Von Neumann stability analysis: insert ρ(x, tn) = G n ρoe ikx ... • order of accuracy of MacCormack method: ⇒ L. MC. ∆t. = ∆t. 2. 6 v. ∆x. 2. ∆t2 − v.
  18. [18]
    [PDF] A MacCormack Method for Complete Shallow Water Equations with ...
    Mar 20, 2019 · The MacCormack method is used to solve 1D shallow water equations with source terms, using fractional steps to treat friction and upwind ...
  19. [19]
    High resolution numerical schemes for solving kinematic wave ...
    Nov 27, 2014 · The MacCormack scheme with a dissipative interface is free of oscillation but with considerable diffusions. The Godunov-type schemes are ...
  20. [20]
    Stability analysis and convergence rate of a two-step predictor ...
    This paper deals with a two-step explicit predictor-corrector approach so-called the two-step MacCormack formulation, for solving the one-dimensional ...
  21. [21]
    [PDF] On Increasing the Accuracy of MacCormack Schemes for ...
    In this work, the linear wave propagation characteristics of MacCormack-type schemes are investigated, and methods for greatly improving their performance are ...
  22. [22]
    [PDF] 3.3. Phase and Amplitude Errors of 1-D Advection Equation
    The combined effect of dissipation and dispersion is often called diffusion. To isolate these errors, we derive the Modified Equation, which is the PDE that is ...
  23. [23]
    [PDF] N94" 3672 - NASA Technical Reports Server (NTRS)
    UMPIRE employs an explicit MacCormaek algorithm with dissipation introduced via Roe's flux-difference split upwind method.Missing: expansions | Show results with:expansions
  24. [24]
    [PDF] Development of Implicit Methods in CFD NASA Ames Research ...
    Jan 28, 2010 · Explicit schemes are very useful and schemes such as MacCormack's explicit algorithm [4] were used extensively in the 1970's - 1980's. The extra.
  25. [25]
    [PDF] Computational Fluid Dynamics Modeling of a Supersonic Nozzle ...
    The propulsion system with the MacCormack method model of the nozzles has a slightly more significant response to the steps up and down in fan speed causing a ...
  26. [26]
    None
    Below is a merged summary of the MacCormack Method and related topics from LeVeque's "Numerical Methods for Conservation Laws," combining all information from the provided segments into a concise and dense format. To maximize detail retention, I’ll use a table in CSV format for key aspects (e.g., method details, comparisons, dissipation, and shock handling), followed by a narrative summary with additional context and URLs. This ensures all information is preserved while maintaining clarity.
  27. [27]
    [PDF] DEVELOPMENT OF COMPUTATIONAL METHODS FOR HEAVY ...
    MacCormack. 35 introduced an implicit line relaxation method ... scheme when the grid lines are not aligned with strong shock waves, 11 it seems that the.
  28. [28]
    An Unconditionally Stable MacCormack Method
    Nov 8, 2007 · In this paper, we rewrite the MacCormack method to illustrate that it estimates the error in the same exact fashion as BFECC.
  29. [29]
  30. [30]
    [PDF] Numerical Solution of Compressible Viscous Flows at High ...
    17) MacCormack, R. W.: A numerical method for solving the equations of compressible viscous flow. AIAA Paper 81-0110, St. Louis, Missouri. (Jan. 1981). (8) ...
  31. [31]
    A boundary-fitted numerical model for flood routing with shock ...
    This paper presents a highly efficient TVD–MacCormack scheme, which is straight forward and simple to apply, as no characteristic transformation is needed in ...