Backward flow step using converge with different mesh sizes

Flow over a backward facing step

Backward facing step is a very common domain that is used to test the effectiveness of the software since we have lot of experimental data to compare the results. This domain actually take less computational power to give the results as it is a very simple geometry.

In this task, I have created a geometry and case set up using the converge studio and ran the simulation in the cygwin command software. The obtained results are then post processed using the Paraview software to form the contours and the plots are developed using the converge plotting tool.

As per the instruction, I have created 3 different base mesh sizes, they are,

case1= dx = 2e-3m, dy = 2e-3m, dz = 2e-3m

case2= dx = 1.5e-3m, dy = 1.5e-3, dz = 1.5e-3m

case3= dx = 1.0e-3m, dy = 1.0e-3m, dz = 1.0e-3m

simulations ran for 15000 cycles each. This is steady state solver we are not using time steps.

all the simulation plots, contours and meshes are below for all cases

Contours from the paraview software after post processing the results are as follows

Mesh

case1

case2

case3

Meshes become finer and finer as the the cell size decreases from the case1-3. We can notice that the mesh is even more finer towards the wall of the domain this is due to specification of mesh refinement which helps in capturing the better details of the gradients across the walls.

Pressure contours

case1

case2

case3

We can see that the pressure is high at the outlet and comparetively low at the inlet even though the initial values of pessure are high at the inlet and low at the outlet. This phenomenon is due to the large area after the step which decreased the pressure. As the cell size decreased we can find the more details in the pressure of the flow.

Velocity contours

all the contours are rescaled automatically

case1

case2

case3

Just like the pressure profile the flow has better details towards the high cell density mesh domain. Since this is pressured based steady state solver velocity is derived from the pressure profile using the PISO solver. Velocity of the flow is high through out the throat area. However in the large area after the back step region we see that the flow still has high speed upto some distance this is due to the fact that there are eddies in that region which helped the flow that is comin out of the throat region sustain the same velocity upto some distance. You will get clear idea by the below contour of velocity vector field.

Plots from converge studio

bound_id_1=inlet

bound_id_2=outlet

Total cell count

case1

case2

case3

Depending on the cell size in the mesh of the domain there is increase in the total number of cells from case1-case3.

Mass flow rate

case1

case2

case3

Mass flow rate is very high at the outlet and very low at the inlet this is due to the area of the inlet and outlet. The eddies formation as well as reverse flow due to eddies after the throat section lead to negative massflow rate from the inlet. We can observe plots becoming more curvee as we move from cae1-3. it is result of the more number of cells.

Total pressure

case1

case2

case3

Total pressure denotes static pressure+dyamic pressure we can see it is even at inlet and varied at outlet for almost 5000 cycles. This variations at the outlet is due to the pressures at between the inlet and outlet.

Static pressure

case1

case2

case3

When ever we are refering to the term static pressure we mean pressure. Static pressure is the pressure that a fluid exerts when it is not moving. As there is no resistance at the inlet the static pressure kept on decreasing and finally reached a steady state after 6000 cycles where as the pressure at outlet remained even throughout the simulation because it is the boundary condition.

case1

case2

case3

From the velocity plot we can notice that the avg velocity of the inlet kept on decreasing where as the outlet increased exponentiatly. This is due to formation of eddies in the domain which resulted in the reverse flow of fluid. we can observe in the above contour of velocity vector field. sudden increase in area in other direction made the flow to decrease the velocity in flow direction in order to fill up the empty space in the domain. large variations in the flow evened before 6000 cycles.