[Kimican]

Hi guys,

I have a project about a Wells turbine.. I plot the area weighted avg total pressure at outlet to decide iterations are enough. When i start the calculation first 30-35 iterations are normal. From now pressure suddenly jumps then well known trouble comes "reversed flow."

I think that solution diverges. I do not get any dicergence warning. There is a serious problem. This is certain.

What is your opinion? Thx..

[mprinkey]

Originally Posted by Kimican

Hi guys,

I have a project about a Wells turbine.. I plot the area weighted avg total pressure at outlet to decide iterations are enough. When i start the calculation first 30-35 iterations are normal. From now pressure suddenly jumps then well known trouble comes "reversed flow."

I think that solution diverges. I do not get any dicergence warning. There is a serious problem. This is certain.

What is your opinion? Thx..

Reversed flow is not (necessarily) an indication of divergence. It may however indicate that you have flow dynamics in the neighborhood of a pressure outlet. Reversed flow happens with the outlet pressure is higher than the computed pressure in the neighboring cell and the outlet starts acting like an inlet, which upsets the mass balance. You never want to have recirculation or other significant flow dynamics happening near an outlet condition. Try extending the domain at the outlet condition by adding an "exhaust pipe" and applying the pressure condition at the end of that pipe and see what happens.

[Kimican]

Ty for reply

I have extended the outlet and i have detected the elements where the turbulance viscosity ratio reaches the 1e5. Then i have made their quality very well.. But i could not get rid of reversed flow completely.. E.g my iterations are normal for 50 iterations then i get reversed flow for 15-20 iterations then normal then reversed flow then then then.... Is this a problem or normal? Can this situation effect my results? Ty..