I executed simulation for a heat exchanger (shell and tube) . but i have a question that if i dont set the reference values and leave them as default values and as you know default temperature of reference is 288.15k in fluent. and with this default refrence values i am getting incorrect heat transfer cofficients at tube and shell sides hence incorrect nusselt numbers. how to specify refrence values in this case to get correct inner(tube side) and outer(shell side) heat transfer cofficients in fluent?
My 2 inlet thermal boundary conditions are:
Cold Fluid enters shell side at 33degreecelsius=306K
Hot Fluid enters in tube side at 11degreescelsius=284K
Hi. In my opinion calculation of heat transfer coefficient of Nusselt number as reported in numerical commercial software is not a good practice. The reference parameter values is the main handicap. I prefer to obtain the values from the fluxes on surfaces using my own definitions
In a Fluent simulation, the software does not care what your are solving. It calculates the heat transfer coefficient (HTC) by q"/(Tw-Tref). Both q" and Tw are calculated by Fluent but Tref is simply given by the user. In a boundary layer flow, this usually works.
However in an internal flow, for you to obtain an engineering HTC, you should integrate q" all over the internal surface, A, to get the whole heat transfer rate followed by dividing the integrated q" with A to have q"mean.
Then, the h can be defined as q"mean/dTmean. You may set dTmean as the log mean temperature difference (LMTD). For the definition of LMTD, refer to any heat transfer textbook. To evaluate the LMTD, for both the cold and hot fluids, you have to the exit temperature from the fluent results.
More accurately, you can define dTmean as the integrated average all over the surface. For each cross-section normal to the flow, you may have a mean temperature difference and you will be able to obtain an integrated average all over the surface to set the Tmean.
Ruben Borrajo can you give me idea how you obtain by your own definitions.
Sun Kyoung Kim i got you but LMTD method is used for calculating overall heat transfer cofficeint that is fi9 i have no problem with that. the problem is with the individual inner heat transfer coffiecient of tube side hi and outer heat transfer cofficient ho. i am getting these hi and ho surface heat transfer cofficients in fluent but these are incorrect. My confusion is that if it were a single tube flow i could have been used a refrence value. but now i have two flows on the shell side and tube side now how to define refrence values for both of them i dont know because in fluent you can give only 1 refrence value.
Sun Kyoung Kim because refrence values in internal flows are the free stream temperature values. and i have two flows as i mentioned so there must be different refrence values for hot and cold fluid. how to define them?
First of all, create two separate fluid subdomains SD1 SD2 for tube and shell. Then, you will be able to set the reference values differently. If this is difficult for you, run the post-processing twice with the different reference temperatures.
A free stream can be defined only where there is no shear stress, no pressure gradient and no temperature gradient. Because this flow is inherently pressure-driven, there is no such zone. So, this HTC based on a constant does not have much meaning. It is just another expression of the wall heat flux.
There is no strict definition in the HTC, which just depends on the engineering convention of the corresponding flow. However, we have been using, q"=HTC*(Tw-BMT), for two-dimensional cases to evaluate local HTC.
So, the LOCAL spatial HTC should be based on the local bulk mean temperature (BMT), which is tricky to define in a three-dimensional flow path. To be loyal to the convention, you should be able to calculate the local BMT at a given location. Of course, the location-wise BMT should be calculated by the UDF in Fluent according to your own definition of the BMT. For example, you can set a cross-section normal to the surface at a given point on the wall. Then, you might able to calculate the mean temperature over the cross-section.
Suppose you perform a temperature measurement experiment to evaluate the HTC. You might try to measure the wall temperature at a location together with the fluid temperature that corresponds to the wall. Think about what the fluid temperature will be. Most likely, it will be a temperature averaged over certain cross-section having the measurement point on the wall.
I think you need calculate using for example the LMTD afected by the correction factor according to the type of flow, (countercorrent, parallel, number of pass and so on) the heat transfered can be determined from the fluid properties, inlet and outlet temperatures and mass flow. Then the global heat transfer coefficient can be calculated. For each single fluid you can average the film heat transfer coefficient because you know the heat transferred, the inlet and outlet fluid temperature and the wall temperatyre. If you want local values will be neccesary determine the bulk temperature in each cross section normal to the main flow.
- Badges
- Science topic