Most heat transfer data can be correlated using very simple equations in an Excel spreadsheet. I have several examples on RG. There is a very interesting table in the CRC Handbook of Chemistry and Physics. It lists every dimensionless number (Reynolds, Froude, Nusselt, etc.) and which properties or parameters it includes, along with the exponents (like rho^2, sigma^4). There's also a list of the properties and every dimensionless number they appear in.
Dudley J Benton You are right, but I am concern about the experimental data correlation. For instance, how Dittus-Boelter Equation come in this way? If it is done with excel, how can I do it?
The Dittus-Boelter correlation is Nu=a*Re^b*Pr^c, which is easily done with Excel by creating extra columns ln(Nu), ln(Re), and ln(Pr), then finding ln(a), b, and c with LINEST. Often the correlations are not quite so simple as log linear, yet we expect some form. A good example of this is generator losses, which are a fan of curves at various power factors, yet will reduce to a single curve with a transformation, loss=f(net/PF^b). We can still use the LINEST function, coupled with the Solver, as illustrated in the attached spreadsheet. You can modify this spreadsheet to accommodate your data and have more control over the form of the correlation than simply log linear. The Solver can adjust multiple parameters so as to minimize the residual (difference between data and correlation). You will also need various properties to calculate dimensionless parameters. I often put these on a separate tab so that I have them in the same spreadsheet. I also attached a spreadsheet that analyzes a shell-and-tube heat exchanger, has properties on another tab, and does an iterative calculation to determine the areas and heat transfer coefficients. The iterative calculations use the bisection search technique to solve the nonlinear equations.