Learning Forum

[Lucjan Nastalek]

Hi everyone, 

I'm currently using Fluent 2023 R2.

I'm attemptin to simulate a conjugate heat transfer problem for a WEM of a small turbomachine with a water jacket cooling.

There is a water flow of less than 10 lpm through the jacket, inlet temperature around 350K.

The attached picture depicts the surface heat flux around the water inlet. There are no heat sources defined in the adjacent solid nor the fluid, but I get a huge heat rejection in this small sliver of inlet port (the depth of the blue region is ca. 10 mm) of about 25kW. The heat goes out of the water and into the steel casing, decreasing the water temperature from 350 K to about 300 K. 

I was hoping to hear some ideas as to what may have gone wrong. It is either a bug in the software, or something that I am overlooking. The port is not its separate surface, but rather a part of the surface collection (defined as a named selection in Discovery) that includes the entire surface of the steel casing in contact with the water.

The wall between the fluid-solid is defined by the Fluent's internal 1:1 interface (shadow zone), and shows in the "Thermal" definition as "coupled". I am using a Pseudo Time-step formulation, so I considered that perhaps the time-step factor in the solid is too low, but changing that did nothing to that heat flux prediction: the steel casing absorbs that amount of heat, and I've no idea where that heat is going! The casing remains at its room temperature (RT) condition. 

Could this be a turbulence model? I don't have a y+ <= 1 in that region, and I'm using the k-omega SST model. Could this be mis-calculating the heat transfer coefficients (HTCs) so badly?

 

I'd appreciate any pointers. In the meantime, I'll re-run it with GEKO for example, see if it helps.

 

Thanks,

LN

 

P.S. is it possible to specify different turbulence models for different fluid domains? If so, how? I only know of the option to make a zone "laminar".

[Lucjan Nastalek]

 

– Update: here’s another glimpse on that very weird behavior at the inlet port. 

The picture shows solid (on the outside) and liquid (on the inside) separated by this weirdly high heat flux values. 

 

Neither GEKO, nor the 1e6 time scale factor helped. I probably shouldn’t have even expected them to.

 

But then, this is the cross-sectional view of the temperature:

So I'm thinking - perhaps there is an accidental heat flux or fixed temperature switched on somewhere...

But no:

 

Any suggestions would be appreciated.

 

Thanks,

LN