A common question that pops up amongst us overclockers is 'What size captube do I need?', or even worse: 'I have captube XXX; how long does it need to be?'
To answer those questions, there are tables with information, such as the Cubigel captube sizing table.

I tried to pour that table with some interpolation into a nice easy-to-use piece of software. I normalised the values in those tables to 1mm I.D. captube, and plotted the data to see what function could be used for interpolation. Then I noticed that there are large jumps in the plot, so it it no good.

Further research pointed me to a model developed by Wolff et al. (1995). They used the Buckingham pi theorem to predict mass flow. According to them, less than 1% of the experimental results was more than 5% off from the predicted values. Now, that's close enough. And estimating refrigeration capacity from mass flow is no big deal once you know parameters like subcooling and evaporation temperature. Thus, it seems like a good model to pour into a piece of software.

Wolff's model only uses a parameter which I cannot find in refrigerant datasheets: the surface tension.

Who can provide this parameter for various refrigerants?