Open Channel Flows in Star-CCM+
I am trying to model an Open Channel flow (free surface flow) in a canal with a trapezoidal area. The flow follows a gradual slope, then a short steep slope of 1:2.5, followed by a gradual slope again.
The requirement is to obtain the increase in velocity at the bottom of the steep slope section as it enters high supercriticality, the minimum depth at the same section, and the subsequent hydraulic jump height and location from slope base, and validate it with the results from online empirical Open channel/Hydraulic jump calculators.
In CCM+, I have defined a field function to fix the water level at the inlet (and used the same for initialization). I have tried innumerable strategies (combined inlet and seperate inlets) with different boundary conditions (velocity inlet/mass flow inlet;pressure outlet/split outlet) with no luck in obtaining the expected results. The velocity and Hydraulic jump is too low.
-Turbulent -> K-e
-Gravity in -y dir (flow in +x)
-Reference location set in +y where air is situated (above water-air interface).
Phases: Water & Air both at constant density
-Combined Inlet: Mass flow inlet [Magnitude = (Given Vol. Flow rate of water x rho-water) + (rho-air x A-air x 0.5m/s)]
-Outlet: Pressure Outlet [Magnitude = (rho-water x g x expected Free surface level of water) + (rho-air x g x remaining height)]
-Top-Opening: Flow Split Outlet (Atmospheric press)
-Turbulence specificatios for Inlets and Outlets:
TI = 1% (assumed)
Turb length scale (max height of water level)
-Time step parameters:
Time step = 0.05s
No of Iterations = 10
Max physical time = 200 secs based on given problem
Max no of iterations = 200/0.05*10=40000
-Canal Wall: Roughness Height = 6mm based on given Mannings roughness coefficient of 0.018. (calculated using obtained formula); Default S & B.
Please have a look at the problem set up and sugest any change in strategy (or anything missing) to obtain a higher velocity and hydraulic jump height.
As you might very well know, it is very critical that you get the grid right in cases where you are running a turbulent simulation. I had worked on a similar sort of a problem quite some time ago. Please send the grid you used by email to firstname.lastname@example.org.
I shall have a look and get back to you if I have any comments on the same.
Re: Open Channel Flow in Star CCM+
Thanks for your response. Unfortunately I cannot share the mesh due to client confidentiality. I can, however, provide you some details about the mesh:
3D Hexa Symmetry - 200,000 elements for a 400m x 20m x 5m domain (approx)
1st cell distance near canal wall = 40mm
Aligned structured flow maintained
Hope this is enough to determine where the problem is.
What is the Y+ value you are getting for a first cell distance of 40mm?
Sorry for the late reply...
The y+ is pretty bad, around 500 to 5000 for the first cell layer. If I make my mesh any finer, the cells go way beyond what our resources can handle, as I am keeping a nominal cell thickness is 100mm to capture the air-water interface to an acceptable level. As I mentioned in my original post, the domain is huge, around 400 m length and 5 m wide and 20m high, but the water level varies upto 5m high, above which the mesh is very coarse.
Considering the roughness height is around 6 to 10mm, will the 40mm 1st cell thickness make such a difference?
Of late I have been playing with the outlet pressure and Turbulence parameters. I have also now split the inlets and outlets:
Boundary: Type - Magnitude - Turb Intensity - Viscosity Ratio - Vol Fraction (H2O,Air)
Inlet_Water: Mass flow inlet - given mass flow - 1% - 100 - (1,0)
Inlet_Air: Velocity Inlet - 1.7m/s - 1% - 100 - (0,1)
Outlet_Air: Pressure Outlet - Atmospheric - 1% - 100 - (0,1)
Outlet_Water: Pressure Outlet - ????? - 100 - (0.75, 0.25) approx
I am not sure what the pressure at the Outlet_Water should be: Less than atm (as per Hydrostatic pressure), more than atm (as per Bernoullis eqn applied at the low depth/high velocity start and higher depth/lower velocity end of the Hydraulic jump).
Thanks for your time, Nepo.
|All times are GMT -4. The time now is 10:40.|