Heat Exchangers and Mixing 1

[EdCich]

Please entertain this somewhat "philosophical" question, which I ponder. Briefly: will a heat exchanger always improve mixing between constituents in a media? Specifically, assume I am designing a heat exchanger to lower the temperature of air or combustion products.....say from 700 to 500F. The heat is removed by a tubular heat exchanger, with water internal to tubes. The air or combustion products flow in cross-flow over the tubes; this flow is turbulent entering where the heat exchanger will be, but may or may not be turbulent depending on the heat exchanger design. In addition to removng heat from the air or combustion products, I want to improve the mixing of a trace constituent - maybe CO, unburned hydrcarbons, or other combustion byproducts for eventual clean-up. Will simply passing the combustion products and trace constituents thru the heat exchanger ALWAYS improve their distribution, and lend to a uniform concentration? It seems to me that depending on the spacing and orientation of the tubes, and whether I use fins or not, mixing could actually be retarded or inhibited, especially if I induce laminar flow. I recognize that I can take proactive steps to increase turbulence - such as staggering the tubes, or purposely inducing a recirculation zone, that can improve mixing. But it's not obvious to me that a conventional heat exchanger design will necessarily improve the mixing. I know that I should utilize a cold flow model, or CFD, but i'm not yet equipped to do so.

Any thoughts on whether I need to take proactive steps to improve the mixing as described? I know this question is general and vague, but I'm just trying to find a design strategy at present. Does anyone know of data or experience where heat exchangers can retard mixing between trace constituents?

Thanks!

 

[25362]

Phylosophically speaking, gaseous molecules move -on the average- at thermal speeds much higher than the bulk mass speed of the mixture. What tells you that a turbulent or a laminar flow r[é]gime would affect interdiffusion time or rate ?

 

[rjw57]

I have first hand experience with the tube side of shell and tube designs that do nothing to improve mixing. Unfortunately, they operate in a laminar flow regime. Twisted tape inserts or shear type mixing element inserts (Sulzer or Kenics) would dramatically improve mixing, of course at the expense of pressure drop. Primarily, two things help mixing: shear and time. This is true for either side of the exchanger. Sorry though, since you are asking to do the mixing on the shell side of a T&S exchanger, I am not very familiar with what you can do on that side in terms of enhancement other than baffling.

Also, is the reason you are asking this that you are injecting trace constituents? Why would you think that stratification or "unmixing" would occur? Large density or viscosity differences? As mentioned before (unless you have large property variations generally occurring in liquids only) mixing will only increase for gaseous components given enough time. I would think that combustion products tend to be well mixed by the very nature of the process. Why do you not think they will be?

Bob

 

[rjw57]

As a follow-up to my last post, I realized that there may be the possibility of condensation of certain combustion species (I have no idea what you are combusting in this case). This could certainly cause variations in species concentration throughout the exchanger which would be best resolved by post cooler mixing.

 

[shanwson]

I am looking for technical information on heat exchagers used for pool heating. Any idea of a website or location to get information?

 

[abeltio]

See Process Heat Transfer by Donald Q. Kern
Shell side Reynolds number is almost always way into the turbulent zone... unless the viscosity of the fluid is so high that prevents it.

in PHT there are dozens of worked out examples...




saludos.
a.