[shaman]

My furnance model have superheater coils running superheated steam.
The potential heat transfer is from flue gas to solid coil then to steam.
My model is too big so i divide into 2 portion.

[shaman]

Portion 1 has flue and coil tube external wall.
Portion 2 has coiltube and steam circuit only.
I plan to use wall heat flux from portion 2 as input to portion 1.

[shaman]

To capture heat flux from portion 2.. the wall is adiabatic. The corresponding result than used as in flux for portion 2...
Is this right....

[ghorrocks]

That does not sound right. I would expect the tubes to be at a defined temperature in the flue gas simulation. It should cool the flue gas down as it passes over it.

[shaman]

Hi glen.. long time no hear... u r right that the tube should have a defined temp.... but then again i do not know what the tube temp as it is dependent on the heat transfer within the steam carrying tube.
Thus the intention is to capture the heat flux unto the tube wall as thermal bc for the seperate tube with steam circuit...

[ghorrocks]

There will be no heat flux if you make the tube adiabatic.

I suspect that the flue gas is very hot and the tube temperature will be very cold relative. So a guessed tube temperature will probably be good enough to start with as the total temperature difference between the gas and tube will not change. Also note there is likely to be significant radiative heat transfer to the tube as well, not just convective.

[shaman]

For now its just the tube outer dia wall only as the tube is not modeled... tube with internal steam circuit in another model.
Plan is to extract heat flux from this ongoing flue model as BC for the 2nd heat exchange tiub+steam model

[shaman]

Do i still use guessed temp at tube od wall to obtain heat flux...

[urosgrivc]

As I understand you dont want heat flux but HTC (probably beter to set t-bulk constant in cfx)
and transfer HTC contour to your second simulation. HTC doesent change much with temperature of the tube so you can guess the temperature as it wont make any diference for the htc contour.

heat flux is dependant on the wall temperature so you can not guess it. (it is like guessing the result of the simulation)

+heat flux [W/mˇ2]
+HTC [W/m^2 K] this K is what will come handy in second simulation
ΔT: difference in temperature between the solid surface and surrounding fluid area, K.


Is it not posible to make all in one simulation-> CHT simulation?
How big is your second portion realy

[shaman]

Hi uros... i did consider htc .. but htc require surrounding temp which is quite high in the without coil run since no heat transfer unto coil via steam.
Also htc might not capture radiated heat which is main heat source in furnance.

[shaman]

U mention t-bulk.. what value shud i use...

[shaman]

It really is not possible as the steam curcuit is about 36mil cells including near wall to better capture htc..s
I am runnibg iawps-97 steam.. so memory very demanding close to 110gb

[urosgrivc]

T-bulk walue doesent mater as you would enter it again in second simulation, it is far field temperature, I mentioned this as if you dont set this HTCcontour transfer gets more complicated including bulk fluid temperature contours and it is yust easier this way.
In second simulation heat flux threw pipe would be calculated with respect to (pipe temperature (constant T-bulk) and HTCvalue)

It woud be nice if radiative koeficient is avaliable you woul be able to transfer both but I am not sure about this as I have only done one way HTC transfers.

[shaman]

But pipe temperature is also a function of heat transfer via coil internal superheated steam..
So supposing my pipe starts of at 321c and increases due to extrrnal htc... is it correvt tobset tbulk as 321 in first analysis and then again 321 as tube temperature with mapped htc...

[urosgrivc]

+(But pipe temperature is also a function of heat transfer) exactly that is why you use HTC.
->if temperature of the tube will be smaller than outside temperature (which is not T-bulk!), heat will flow into the tube wals,
and if temperature of the tube is higher than outside temperature, heat fill flow out from the tube so equilibrium will be reached and steady state will be obtained

+Yes T-bulk seems ok you will have to specify it in second simulation as a far field temperature

as T-bulk is set it is a constant, so you enter a number (for far field temp in second simulation) (it doesent mater what number it is (search somewhere here on the forum what it does))
if it would not have been set you would have to specify a (fluid bulk temperature >contour<) in second simulation. what is simpler? number vs contour
You wil already need a contour of HTC.

+one way htc contour transfer:
I think that best practice is to export it from the firs simulation in cfx post via BCdataexport>htc contour>creates a csv. file
than in second simulation you inport it under Initialize profile data and stick it on the wals of the tube.

I am still not sure what to do with radiation as I havent done it before.
But the proces should be identical to the htc one, as radiative heat tranfer coefiicient exists also, just that equation will
probably include (pipe temperature and furnace temperature also)

[shaman]

Thnks uros... in fact i have prepared the tbulk in expert parameter as 321.... and also as pipe outer wall only as temp value too and solving... wil check the variance between them

[shaman]

Heat transfer htc and temp mapping should be formality.....