Skip to content

INterpolated-Two-D(E)-Navier-Stokes-Equations- an attempt to implement higher order interpolations and add unsteady terms to a FVM discretized Navier-stokes solver.

RSuryaNarayan/INT--E--NSE

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

INT--E--NSE

Introduction

INT(E)NSE is (an intended pun and) acronym for IN-terpolated Two-D(E)-Navier-Stokes-Equations. This is an unsteady Navier-Stokes solver I developed during my free time. Interpolations for the face-velocities are done using a blended second-order and first-order accurate scheme to be able to capture the vortices. Pressure-velocity coupling is achieved using SIMPLE algorithm and the time-stepping is first order accurate following backward finite differencing discretization. I am looking for someone who can help me modularize this code and break it down into smaller files. Currently everything is organized using comments(!). Will soon be adding mesh-clustering for the wall and include the energy equation. If you are interested to upgrade/contribute to this solver either by:

  1. Modularizing this code into smaller files to make it easy to perform bug-fixes
  2. Extend to 3-D
  3. Higher-order interpolations
  4. Higher-order time stepping schemes
  5. Alternative pressure-velocity coupling schemes
  6. Add turbulence model
  7. Accelerate code performance using suitable paralellization Please contact me through mail/submit a pull request. Curerntly the code is written entirely in MATLAB. The test cases for which I have got results are shown below

Why INTENSE

So the motive behind developing INTENSE is contained in its name-to grasp the concepts of CFD in the most intense way! Ok jokes apart, it is intended as a code to help people starting CFD to better understand what happens inside the black-box when they push a button in any commerical software.To better gain insights into the development of a CFD code. The entire code is thoroughly commented and includes relavant one-liner explanations wherever required. Now I know I could have done a better job using a Jupyter notebook, but I started out coding this as a hobby and before I knew it had over 300 lines! Plus this does away lesser memory. But with solver getting upgraded with new features like mesh-clustering, energy equation, and maybe turbulence, I might like extending this to a more generic CFD code written in MATLAB. However, the intent remains the same-INTENSE-ly understand CFD!

How to use INTENSE

You can either use this for your direct simulation by tweaking the mesh-parameters and flow-variables in the beginning and hit run in MATLAB (which is highly unlikely as there are standard solvers out there that do a better job!) or you can use this to understand how CFD codes are built by going thru the code and the comments. I advise you to tweak some parameters in the code like the blending factor or the number of iterations for the various sub-solvers to understand the importance of each loop and their effect on the convergence. Oh yea, this has a residual monitor too. so that will show you if your results have converged.Comment or uncomment some lines and look at how they effect the solution.

Some geeky details

INTENSE solves the unsteady version of the incompressible Navier stokes equations together with the continuity equation using the numerical Finite Volume Method (FVM). I have enabled three major problems as benchmarks/test cases. The first being the the standard fully-developed channel flow, the second a pressure-driven channel flow (with pressure boundary conditions) and the lid-driven cavity. The Semi-Implicit method for Pressure-Linked Equation (SIMPLE) is used to couple the pressure and velocity terms. The solver uses a staggered mesh with ghost-cells in the boundary and a blended first and second-order accurate interpolation scheme is used to obtain face-velocities on the staggered-mesh. Time-stepping is done using a first-order accurate finite difference scheme. In addition there's a provision to block cells out-giving us results for the flows around bluff bodies.

About

INterpolated-Two-D(E)-Navier-Stokes-Equations- an attempt to implement higher order interpolations and add unsteady terms to a FVM discretized Navier-stokes solver.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages