Learning Forum

[Snow1Runner]

Hi everyone!

Recently, I have changed fans in a unit I'm modeling from axial to a centrifugal blower. Essentially, from a top view, the airflow looks like the following:

[Rob]

I'm not overly familiar with Icepak use, but as it has Fluent in the background I wonder if it's material density related. What is the air density, and how does the field temperature differ from the initial condition? n

[Snow1Runner]

Hi Rob, nThank you very much for your answer. I'm using density of dry air at 30C, which is 1.614 kg/m^3. For the above simulation, the temperature of the air in this closed loop is at a max of about 65C but in previous simulations with the axial fans, I expect to see temps hovering around 75-85C. The initial condition is set to 50C ambient and radiation temp as well is set to 50C. nLet me know what you thinkn

[Rob]

It's not a density issue then. The classic error in Fluent is letting density vary with temperature and not giving the gas anywhere to expand to! n

[Satyajeet Padhi]

Could you run an isothermal case with a finer mesh and more iterations? I also noticed you are using a blower object. Blowers usually take a lot of iterations to converge. n

[Rob]

Having checked, the other suggestion is a small bleed port to give the solver somewhere to gain/lose mass as it converges. n

[Snow1Runner]

nThank you both for your suggestions. nI will run an isothermal simulation with a finer mesh and will let it run for 1000+ iterations. If I still see the behavior of pressure heading to negative infinity I will try to apply a small bleed port. I will report back with my findings as soon as those simulations are finished. nThanks again!n

[Snow1Runner]

Array Array nHi Rob and Satya,nI have tried both of your suggestions but sadly haven't reached a converging solution for my simulation. For running the simulation for more iterations, I have done so here and I am experiencing the same behavior. The solution is failing to converge and the continuity equation is slowly diverging (I think that's what it looks like). At the same time, the pressure inside the unit is heading to negative infinity, showing no trend of trying to find stability. Before this run I actually refined even more my mesh. nAnd here is the intake of the fans (note that the air is actually flowing through the top cavity where the monitoring points are located): nAgain, it seems like the exhaust air of the fans is not going anywhere and the fans are just creating a negative vacuum inside the unit. Any thoughts on what could be causing this? Does my diagnosis makes sense? nSomething is certainly going astray somewhere around 100 iterations when the velocity of the monitoring points start diving and the pressure starts heading to negative infinity.Rob, I actually tried adding a small bleed port and it sort of supported my theory. The simulation reached convergence but it is pretty clear that it is not the case. At the location of the bleed port, air is rushing in since the fans have created such a negative pressure inside the unit. This causes the velocity of the monitoring points to rise but it's caused by the air rushing in, not the fans that should already be circulating the air. That's why I believe that the fans are just trying to intake all of the air inside the unit without actually experiencing any airflow across the loop. n