I am performing a simulation of a round jet of 1in diameter within fluent. I am trying to compare the k-e standard and realisable models. The realisable model will converge but the solution is not valid as the momentum distribution is not constant and the flow is not self-similar. The flow is modelled axisymetrically about the centre of the jet, I have tried both 1st and 2nd order discretisation, tried decreasing the relaxation factors. The convergence criteria I am using are 1e-06. I was wondering if anyone had any ideas why the solution is not valid and is there anything I can try to make it so? I have had no problems with the standard k-e model. thx

the boundary conditions are above jet is a wall (west side), top is pressure inlet, end (east side) is a preessure outlet, jet centreline is an axis, the fluid is air, the jet is velocity inlet with velocity of 56.2 m/s and turb intensity of 0.58. Fluent default values have been used in values not mentioned.

From what I have been reading on this site my mesh may not be detailed enough. The geometry is 5in vertical at jet rising to 25in at 110in horizontally from jet. I have 11.4k nodes in the mesh with 12 across the jet inlet. Aspect ratio is ~10.

By convergence criteria I meant residuals.

try to use Coupled Solver