Viscosity and surface tension

[Robert Coyle]

I came across a set of calculations for Viscosity and surface tension from oxide composition. Has anyone used these types of calculations to get useful glaze information?

[Lawrence Weathers]

No, I have not, but I might give it a try.  Could you share a link to these equations?

[Norm Stuart]

I haven't run across prediction of viscosity and surface tension but I'd like to see how it's done.  I find every bit of glaze prediction useful, in combination with test tiles.

I also like the melt tests of raw materials Tony Hansen has posted on various pages of Digitalfire.

The final viscosity of a glaze is highly dependent upon the temperature, so I assume this would be a calculation at a particular final temperature?

Surface tension would be very helpful when layering one glaze over another.

Surface tension which is not enough to cause crawling can cause crawling if given enough time, so time is important too.  The zinc semi-mattes from "Mastering Cone 6 Glazes, which lay flat when cooled quickly or at the 185 F per hour the book recommends, but crawl when subjected to our 6 hour 50 F per hour slow-cool.

There are other important aspects of glaze which are not related to their final melted qualities.

1.) The temperature range over which a glaze melts is also important to the final look. Many glaze breaks are the result of one portion of the glaze melting out of the glaze at a lower temperature.

2.) Also important is whether a glaze with an LOI component, lost as gas, loses this gas before the glaze is melted, or after some or all of the glaze is liquid. This creates an oil spot glaze, or pin holing in a viscous glaze.

In practice I test all Cone 6 glazes in a Cone 04 bisque to try understand what has happened by that point. Some are fully melted, others more an engobe, and some have separated into a liquid fraction and a refractory portion.  Understanding the mechanism behind each glaze look is detective work.

[Robert Coyle]

BaseName    LinExp    Viscosity           Surface Tension
Na2O            3.9              3                 1.5
Li2O            0.67              3                 4.6
K2O            3.31              3                 0.1
MgO            0.26              1                 6.6
CaO            1.48              1                 4.8
SrO            1.59              2                 0
BaO            1.29              2                 3.7
TiO            1.44              0                 3
ZrO            0.99              0                 4.1
Cr2O3            1.7              0                 0
W3O8            0              0                 0
MnO            0.73              0                 4.5
Fe2O3            1.33              0                 4.5
CoO            1.47              0                 4.5
CuO            0.73              0                 0
ZnO            0.94              2                 4.7
Al2O3            0.63              0                 6.2
SiO2            0.35              0                 3.4
B2O3            0.31              3                 0.8
SnO2            0.67              0                 0
PbO            0.83              3                 1.2
P2O5             0.65              0                 0

These are calculated similar to linear COE. the calculated oxide percentages are multiplied by the factor and summed.
Since both viscosity and surface tension are functions of temperature, I don't have any idea of what the units might be.

I dug these out of an old glaze software program. Can't remember which one. It had all the data as delimited text files. I don't think it is around any more.

[Norm Stuart]

It appears a Viscosity of 3 for sodium, potassium and lithium means the glaze runs more than Viscosity of 1 which is Calcium and Magnesium, both low flow oxides.

Magnesium which has such a high surface tension it crawls back onto itself away from the clay has the highest Surface Tension of any oxide at 6.6 which makes sense.  Potassium which flows flat has a Surface Tension of 0.1 which seems right.

It's interesting that Lithium which has an extremely low COE has a Surface Tension of 4.6 compared to 1.5 for Sodium and 0.1 for Potassium.

[Robert Coyle]

Hi Norm

These values do sort of follow what you might expect from where the elements appear on the periodic table.

From what I have read, these values are sort of first approximations. Oxide interactions in silica melts are not simply additive. There is a dependency on what other oxides are in the melt.There are a whole lot of papers published on oxides in lower melting glasses that indicate that this type of calculation is possible. There are also a large body of information of this type on mineral formation. Unfortunately, the latter usually also includes high pressure and the presence of water. 

But even crude, first order estimates might be useful from the standpoint of ultimate melt characteristics, like maybe... yes it will not craze, but will it run off the pot at cone 6 but not cone 5?

Just wondering if anyone else had looked at this.