Designing Heat Exchangers for Non-Newtonian Fluids

That would be difficult, getting a higher reynolds number to creating turbulent flow, and whether the product is heat sensitive

-Plate H.X. the delta P would be great.

-Shell end tube possible the same, or accept a lower efficiency.

-double tube or triple tube would be the only way I can think of.

I am only going by my experience with conventional.

You may want to try unconventional ways, such as high pressure.

Just throwing something out there

Thanks for your reply. I dont have any emperical data to start with. I would like to know how to proceed with the designing. how about starting with shell and tube ex. how will design this ex for non - Newtonian fluid.

I think that you will have to resort to computational intensive design by actually solving the fluid flow and heat transfer equations for the heat exchanger types you want to design. The reason is that the reynolds number will be axial location dependent as the viscosity continuous to change with axial location. You should have to attack the design with Numerical Methods.

I don't think thats what the guest wanted to hear.

Especially with no empirical data to start with.

Sorry, My bad!

Thanks for your reply. I dont have any emperical data to start with. I would like to know how to proceed with the designing. how about starting with shell and tube ex. how will design this ex for non - Newtonian fluid.

Alright here is a suggestion, and I am abstracting from the archaic method of Similitude and Dimensional Analysis, but just more rigorously.

Write down the standard Partial Differential (PD) Equations forms of the Momentum and Heat Transfer for the tube side and shell-side. Next non-dimensionalized the equations with the operating design values, and the viscosity, conductivity and other thermophysical properties based on the entry conditions of each fluid. That should give you the set of dimensionless numbers in the equations. Use these in the empirical equations for designing HE for Newtonian fluids to get the starting Heat Exchanger design. You can even improve on this design by using either the geometric or arithmetic mean of the inlet and outlet temperatures of each fluid, and use the resulting values to calculate the viscosity and thermal conductivity of the non-Newtonian Fluids. You will have to iterate against this reference or ideal HE.

Now because you will posit this newtonian fluid based HE as an ideal by which to start the Numerical method iteration, remember to non-dimensionalize the viscosity relative to the values you used to design the "ideal" HE and that means, that the momentum and conductivity Diffusion Terms of the PD Equations must still have the viscosity and conductivity values subject to differentiation by the outer gradient operators: ∂/∂z(μ∂V/∂z) and ∂/∂z(κ∂T/∂z); where V is dimensionless velocity or vorticity depending on your choice of equation form and T is dimensionless temperature, and μ is the dimensionless viscosity and κ is the dimensionless thermal conductivity. Of course, these values change axially (and possibly also radially) for each non-Newtonian fluid because of the non-Newtonian nature of the fluid.

For more informed solution you will now solve the dimensionless PD Equations iterating against the ideal newtonian fluid HE you have designed. In the worst case scenario use the Newtonian fluid based HE as your design.

I hope that I am helpful. Good luck