[MLeong]

Hi everyone,

I've got a heat transfer problem that I'm not sure I've set up correctly. The problem is heat emitted from an incineration stack. The temperature and flow rate of the exhaust is measured at 820 degC at the mouth of the stack. the velocity of gas exiting the stack is small at about 0.7m/s, however max wind speeds modelled go up to 8m/s.

I initially tried creating the model with the exhaust inlet boundary set at as the mouth of the stack. I am however finding that the temperature in the immediate cell drops immediately from 820 to less than half within one cell. I am not sure if this is physically correct. I've then extended the model to include the region within the stacks itself, keeping the inner walls at 820 degC, expecting the column of hot gas in the stack to be about 820 degC at steady-state. I am not finding this at all and am somewhat puzzled - am I doing this correctly?

If there is a better way to model this, do please tell!

Best regards and thanks in advance,
Melvin

[ghorrocks]

If the hot gas velocity is 0.7m/s and the wind is 8m/s I would expect there to be lots of mixing at the top of the stack and the cold outside air probably penetrates the top of the stack some distance. I would also not expect any noticeable plume as it is probably highly mixed with outside air before it leaves the stack.

What this means is you will need to model the top section of the stack. The flow in this region is going to be complicated so a finer mesh than you probably expect is likely to be required.

I bet if you do this things will become clearer.

[MLeong]

Glenn,

thanks for the quick response. The mesh resolution goes down to about 0.2m length scale and the stack diameter is about 2.1m. I suppose I can go finer than that, I'll try this and get back to you.

Would you say however that, when modelling the stack exhaust mouth at the top as the inlet into the domain (static temperature at 820degC), we should expect a larger region with a high temperature perhaps at least 500 degC and above? Despite having a mesh resolution of about 0.2m length scale (over a 2.1m diameter mouth) and a low wind speed of 0.2m/s, I am still seeing the heat dispersed to less than half over a relatively small radius. I will inspect my model again to see if I have made any errors in the setup, failing which I will go for an even finer mesh. Will keep you posted. Thanks for the advice!

[MLeong]

Hi Glenn,

Sorry I made a mistake on the exhaust flows. It's about 0.07m/s (one order of magnitude lower!) at 800 degC, and that is measured at the tip of the exhaust (dia. 2m). I've refined it down to a 0.1m grid size with inflationary layers. I've not modelled the inside of the stacks itself this time, given the information on the inlet boundary.

Results are pretty much the same with heat dispersed very quickly, in fact 800degC goes down to less than 100degC in less about 1m. I guess with a high wind speed and such a low exhaust flow the heat gets dispersed very quickly.

What do you think?

[ghorrocks]

It's about 0.07m/s

Well then it will be even more mixed. This makes it even more important to model the top section of stack.

I've refined it down to a 0.1m grid size with inflationary layers.

This is still VERY coarse. I would be surprised if it is accurate at all at that coarseness - and it also explains why it has a large jump in temperature. The elements are too large to resolve any of the smaller scale features which will be happening. You will certainly need more mesh than that.

Yes, the temperature will dissipate quickly but unless you resolve the mixing process your results are rubbish.