7 replies to this topic

[Art Montemayor]

Eliot:

You can also find the methods to calculate the viscosity of gas or liquid mixtures in:

"The Properties of Gases and Liquids"
by Robert Reid and Thomas Sherwood
McGraw-Hill Book Company

The Wilke method is considered the best known and one of the most accurate methods for estimating the viscosity of gas mixtures.

It is reported that there is presently no reliable method for estimating the viscosity of liquid mixtures.

[siretb]

For gas mixtures at low pressure, not containing hydrogen, the following simple formula (not as good as Wilke & al) can be used
Mu(mixture)=Sigma(yi Mui sqrt(Mi))/Sigma(yi sqrt(Mi))
M is molecular weight and y the vapor mole fraction.

[Art Montemayor]

Eliot:

The Wilke viscosity estimation method for gas mixtures is reported as having been extensively tested and yielding an average deviation of less than 1%. However, the Wilke method is also reported as yielding mixture viscosities greater than experimental for mixtures containing Hydrogen or Helium with a heavier gas; the mixture (H2 + N2) in point. For this mixture, predicted values may be as much as 50% greater than those found experimentally.

The Herning and Zipperer Estimation Method, another option, yields results within 3% - but starts to fall into inaccuracy when applied to Hydrogen-rich mixtures.

You may find it a problem dealing with H2.

[siretb]

1) the units for viscosity are whatever you want. centipoise for instance.
2) I use the equation I posted for gas mixtures, works well, but I have no hydrogen. I have no specific experience about H2, I can only quote what is in Perry's handbook, 5th Ed.
Equation has been well tested for hydrocarbon mixtures...... for small H2 concentrations, errors are small, but may reach 10% for h2 rich mixtures. If H2 content is greater than 25% Wilkes equation is preferable ##Perry, 5 ed, page 3-249##

[JMW]

There are some methods for liquids that base the calculation on the known viscosity of the components. This is a non-linear relationship. e.g. the Louis and Refutas Methods.
Accuracy? A good question. I don't know the answer to that.
However, if you know the molecular weight of the product, this is related to the viscosity.
Try this link:
http://www.pslc.ws/macrog/vis.htm
and I suppose you could use ASTM D2502-92(1996) backwards i.e. compute the viscosity from the molecular weight. I'd be interested in any comments on that.