i have simulated (in fluent) the heat transfer for turbulent flow of air in a tube of which surface is exposed to constant heat flux of 8000 w/m2. for two conditions 1) plain tube only 2) tube with twisted tape insert
simulation details are described below:
steady state analysis
inlet air velocity, temperature= 4 m/s, 300k
tube wall heat flux= 8000 w/m2
tube length and internal diameter= 900 mm and 26.6 mm
viscosity model= sst k-w
twist ratio of insert= 5
for plain tube i gained outlet temperature of air = 536k
tube surface average temperature= 702k
for tube with insert outlet temperature= 544 k
tube surface average temperature= 725 k
so i gained increased outlet and tube wall temperature for flow with insert then without insert for same flow rate and heat flux. is it ok? my confusion is since heat flux is same for both case rise in tempearture if reduce convection heat transfer cofficient for insert then plain tube as both cases have equal heat transfer from pipe wall. need your opinion, thanks
hi Yu,
i am not clear if you are saying temperature at outlet should be same for with or without insert for same flow conditions. however i found a discrepancy. in fluent solution i found heat addition from pipe wall is 601 W for both case but when i calculated heat addtion using Q=m*Cp*(Tout-Tin) it is always higher. heat balance do not occur. i do not use any heat generation unit. is it caused by poor grid size? i checked for grid sensitivity using 1 mm, 2mm, 3mm maximum face size. from literature review i also observed that many authors use constant heat flux where many use constant temperature at surface. what is practical approach and what should be the results for both of them?
Hi Yu,
thanks for your answer. you are right. i used area average for temperature at outlet. but now i used mass flow average for temperature in outlet in both cases and the outlet temperature is same for both cases with heat balance. however still there is two problem remaining. how to calculate surface temperature of wall and bulk fluid temperature? i used area average for surface temperature and arithmetic mean of inlet and outlet temperature to calculate bulk temperature before. is it ok? now, using costume field function for your above mentioned procedure (in your answer) then evaluation by surface integral gives new temperature of bulk fluid. say, inlet temperature was 300 k and outlet temperature was 446 k of which arithmetic mean gives 373 k bulk temperature. this procedure is followed by many reputed authors. however from surface integral of whole fluid volume (steps 1 and 2 of your answer and then ratio) gives answer 379 k. if bulk temperature is not exactly arithmetic mean i assume? here i attached two images. insert domain was boolean substract to replicate adiabatic wall.
- Badges
- Science topic