Cartesian Flow Solver

An in-house 2D Cartesian flow solver is developed using Semi-Implicit Method for Pressure Linked Equation (SIMPLE)  Algorithm. The objective is to study flow patterns and analyse velocity variation in a Lid-driven cavity with various Aspect ratios (height/width). This analysis helps to predict the efficacy of mixing process which has application in industrial paint preparation. The simulations were carried out for flow with Reynolds number (lid_velocity*Length/kinematic_viscosity) = 100, 500 and 1000. Top wall was being moved at a constant velocity of 1 m/s. 

Details of the Project :

Timeline - July to November 2017

Supervisor - Dr. Arun Mahalingam, Professor, Department of Mechanical Engineering, National Institute of Technology, Karnataka.

Main Objectives of the study - 

1. To develop a efficient fluid solver which can handle high Reynolds Number (Re) flows without any stability issues.

2. To understand which parameters can lead to optimised mixing of particles in channel through fluid flow visualisation.

Details of study  - 

The details of the domain used and boundary conditions applied are given below.

Significant Outcomes - 

Following conclusions can be drawn from the results that we have obtained:

1. It is observed that at high Re, thinning of wall boundary occurs and there is much more movement of fluid in the near wall region as compared to lower Re.

2. The relaxation given to the pressure correction term has to be high at higher Re and the number of grids along x and y direction should also be increased.

3. The variation of aspect ratio produces different kind of secondary vortices and in some cases, these secondary vortices become highly pronounced.

4. It is observed from the graph of height v/s velocity that the variation of velocity is almost linear with height in the mid section of the domain (especially at higher Re)

Fig. Comparison of X Velocity at vertical centerline of Lid driven cavity (Ar = 1) with the results available in literature for various Reynolds numbers.