Hello, I want to simulate a jet by using FLUENT. But I am sure about the best method for solving a subsonic M0.75 single flow jet (Cold/Hot). Usually k-e method is used, but I don't know really if using the default constants is enough or it is needed to change some. I will be very thanful if you share your knowledge with me on this problem. Thanks Mahdi

Hi,

As you know the k-epsilon method is kinda a work horse for turbulent flow calculations - especially if you are not that much interested in Reynods stress components. I used to prefer RSM model since it gives all the components ( nine of them) directly without going through the 2/3 rule. To me RSM works just fine especially in boundary layer analysis specifically 3D BL - we had a pretty good agreement with test data. As for the Jet flow, well if you are interested in subsonic/lower end of transonic jet flow...k-e should be ok. It has a better convergence behaviour/stability and gives you a quick picture of what the flow would look like...and then you can further twik/refine by changing the parameters. Here there is no general rule that fits all flows. The people at Fluent made experiments on water and air and came up with those constants and it worked for most flows but you can change them as neccesary.

regards

Melaku Habte

[ahasan1]

Dear Friends,
I think this is the proper thread to place my question. I am using RSM, k-epsilon Standard, k-omega SST, Transition SST models in FLUENT to simulate single, round (axisymmetric), isothermal, turbulent jet. However, it seems like my simulation results never match with theory or experimental results in the literature. Can anyone suggest me why it is not matching? The jet is issuing (jet velocity=51 m/s) to open air. I am using 'velocity inlet' for nozzle exit boundary condition, 'axis' for jet centerline, 'pressure outlet' at top and right end of the domain, 'wall' boundary condition at side-west wall. I have tried changing (increasing/decreasing) mesh sizes however it did not work. I have also tried steady cases and unsteady (time step size=0.001s, iterations/time step=20, all residual fall below 10^-7) cases for the cases. Please suggest me what to do?


Thanks.

regards-
AHASAN

Originally Posted by Melaku Habte
;118582

Hi,

As you know the k-epsilon method is kinda a work horse for turbulent flow calculations - especially if you are not that much interested in Reynods stress components. I used to prefer RSM model since it gives all the components ( nine of them) directly without going through the 2/3 rule. To me RSM works just fine especially in boundary layer analysis specifically 3D BL - we had a pretty good agreement with test data. As for the Jet flow, well if you are interested in subsonic/lower end of transonic jet flow...k-e should be ok. It has a better convergence behaviour/stability and gives you a quick picture of what the flow would look like...and then you can further twik/refine by changing the parameters. Here there is no general rule that fits all flows. The people at Fluent made experiments on water and air and came up with those constants and it worked for most flows but you can change them as neccesary.

regards

Melaku Habte