Flow Over Backward Facing Step

Challenge under the course Advanced CFD for IC Engine Applications-Skill Lync

Submitted by: D. Sai Kamal

Software’s used: CONVERGE v2.3, Paraview

Geometry:

Setup:

Initial Conditions being Temperature of 300K and atmospheric pressure of 101325 Pa with air as species with default mass fractions.

Time Step Info for all cases:

1e-006	dtstart
1e-008	dt
2.0	dt_max
1e-006	dt_min

Case 1:

Grid Size: 5e-04 (0.0005 m)

Figure 1 Grid size with fixed embedding near the walls

The Figure 1 shows good grid size with fixed embedding present near the walls for smoother results.

Observed Development of Pressure:

Figure 2 Pressure conditions at time of 2sec

An Inflow pressure of 110325 Pa and Outlet desired of 101325 Pa causes the flow and are maximum at the inlet and outlet respectively. Due to back step, the pressure build up at the step is less influenced due to change in the contour.

Figure 3 Pressure conditions at time of 4sec

After a time of 4 sec, the pressure build is reaching the stabilised conditions with observed minimum pressure the central part of the step and the maximum are uniformly laid at that position like some shock waves at regular intervals.

Figure 4 Pressure Condition from time of 4 sec to 15 sec

After 4 sec of the simulation, the pressure profile is almost the same throughout the end as it has reached the stabilised state.

Observed Development of Velocity:

Figure 5 Velocity condition at time of 2 sec

The velocity values were high at the inlet and the due to step, the fluid tends to exhibit lower velocity at the middle zone of upper wall. The velocities are unstable as it is turbulence modelled and is yet to come out of it and become stable.

Figure 6 Velocity condition at time of 4 sec

During stabilizing i.e. at 4 sec, the profile develops and as it not yet reached steady state completely, at the rear face or outlet, the values still did not show a uniform velocity near the walls.

Figure 7 Velocity condition after time of 5 sec till 15 sec

For a completely steady state achieved flow, the velocity of the fluid remains same all throughout near the upper wall and very low or zero velocity at the bottom wall after the backward step.

Figure 8 Velocity Glyph vectors showing the flow direction

Case 2:

For same initial conditions and geometry, Base Grid Size of 3e-4 (0.0003 m) the system did not support and hanged. Though I waited for 35min, there is no response or output files.

Case 3:

A base mesh size of 1.5e-4 resulted in displaying a warning message that the cell count is too much as the grid size is very small for the geometry. Also it is supported by the Cygwin terminal showing an error requiring a memory of 17GB solely for running the simulation and by itself aborted the run command. So the case of 1.2e-4 or 1e-4 is not even good to attempt.