[vzdh]

Hello colleagues!

I used Flow-3D to calculate the head losses in a pipe with a diaphragm (orifice) and got strange results. I have found that the piezometric head losses obtained with Flow-3D are lower than theoretically expected or measured experimentally.

For example: a pipe with a diameter of 10 m (D), a diaphragm with a diameter of 7.5 m (d). Volume flow rate in the pipe is 1950 m3/s. I measure the pressure on the pipe lines at characteristic points located at a distance of -1D from the diaphragm and +5D from the diaphragm (along the flow vector and assuming the diaphragm coordinate is 0).
Under these conditions, I should get a piezometric head loss of about 80 m (according to Borda's formula and experiment). In Flow-3D I get a loss of about 60 m.

The result is not affected by the boundary conditions, mesh size, turbulence model, or even the fluid viscosity. The numbers turn out to be a little different, but conceptually the result is the same.