Can ice damage building masonry? Masonry elements that look upward, that face the sky, are often in bad condition, and ice may play a part. We can damage masonry elements in the lab by subjecting them to frost dilatation testing, which completely surrounds a water-saturated sample with cold temperature, and drives ice inward. How about unidirectional cooling on a vertical masonry wall? The hydraulic model of ice growth, described earlier, says no—damage should not be expected. That’s great, and encourages us to adopt energy measures in existing buildings. But…how complete is the hydraulic model? Are there no exceptions? We may begin looking at one early one—frost heaving—in the work of Stephen Taber.

Prof. Taber was a geologist from the University of South Carolina—an odd place for the birth of research into frost heaving. Taber was at home in the field, everywhere from the Caribbean to Alaska. He was made significant contributions to our understanding of permafrost, and Aalaska’s Quaternary history in the 1930s. In 1942 the U. S. Army’s construction of the Alcan Highway was a disaster due to ice-rich permafrost. The job of writing an army manual for northern highway construction in 1943 fell not to Taber, who was then largely forgotten in permafrost studies, but to Siemon Muller because of his fluency in Russian, providing access to those sources. A frustrating struggle for influence marked Taber’s later career.

Taber’s name is remembered now for his laboratory work in 1929. Old research often shows appreciation for craftsmanship in experimental apparatus, and narrates the lab sequence as if the reader was looking over the shoulder of someone peering for the first time into an unexplored area. Old research often captures the thrill of discovery, such as here. Figure 1. shows a version of his setup, intended to measure frost heave lifting a weight. Cooling was from above. Clay, sand and soil was placed in cardboard cylinders (former milk containers according to the author), and these were waterproofed with paraffin or shellac. Ample water was provided to the sample through a hole in the bottom, and the sample was placed in saturated sand.

Taber challenged the hydraulic model of freezing in soil. That model claims that the volume expansion of water into ice causes the unfrozen water to flow away from the growing ice front, and if the water flow is blocked, then the ice mass will displace, and heave. Instead, he showed the formation of an ice segregation or “lens.” See Figure 2. This represents the outcome of one of his initial studies. 

The sample contained two masses of clay, with a one-inch layer of sand in the middle. The lower mass of clay was saturated with water by capillary action, which was broken by the sand. Cooling continued to the point where the interface between the sand and the lower saturated clay reached freezing. Then water, fed by the capillary clay, began to freeze. There is a slight depression of freezing temperature in capillaries, so the saturated clay would continue to feed the growing crystal, provided temperature is maintained at around 0°C right at that interface, and provided the cooling is sufficient to remove all the heat of fusion generated by the ice formation. The resulting ice segregation is about one inch thick. The upper clay and the sand layer were lifted upward by that amount. Apparently the temperature was sufficiently well controlled that icing did not continue down into the lower clay mass. 

Here is Taber’s description: 

A growing ice crystal is in contact with a thin film of water similar to the adsorbed layer that forms on other solids. As molecules are removed from the film and attached to the crystal, they are replaced by others from the surrounding water. When an ice crystal grows in a direction in which growth is opposed by a solid body, such as a clay particle, the pressure is exerted through this thin film which separates them. The film probably consists of little more than a single layer of molecules, for it is not expelled by pressure. After the available water has been exhausted, the film may be frozen, but it does not freeze easily. Cohesion is greater between the molecules in the film and between these molecules and the ice than it is between water molecules that are not close to ice crystals. 

Taber does not explain how his containers accommodated that upward movement. He notes this limitation in his experiments, but he was not forthcoming on whether paraffin was a sufficient lubricant to the container sides. (I retain a concern for the strength of attachment of ice crystals to the containers in which they are found. More on this later.)

It is odd that he finds that the thin liquid layer, which feeds the growing ice crystal, may be as thin as one layer of molecules. He fails to take into account the roughness of the clay surface, and of the ice crystal. The surface roughness for normal glass may be 400 nanometers, which leaves room for about 3000 water molecules. This odd idealized characterization of solid surfaces seems to continue in subsequent research.

The paper continues his investigations into factors influencing frost heaving. 

Unidirectional cooling may become locally omnidirectional. 

… if ice crystals in growing downward come in contact with the top of a large soil particle and begin to surround it so that the temperature at the top of the particle drops below freezing, then the temperature of the bottom of that particle will reach the freezing point before the water with which it is in contact. Therefore, freezing takes place in part outward from the surfaces of large mineral particles that are in contact with water, and not merely through the downward growth of ice crystals as in pure water. 

If colloids are present, they may hamper crystal growth.

Other things being equal, the height to which water rises varies inversely as the diameter of the capillaries; in extremely fine material it maybe 3m or more, but most very fine soils have a high percentage of colloids. the effect of which is to decrease permeability and prevent water from rising as high as it otherwise would. 

Crystal growth can lift with considerable strength. 

In experiments a pressure of over 14 kg!cm2 has been obtained through growth of crystals to form ice segregations 

When Taber speaks of pressure, he is describing a force that is applied by the growing solid onto an external surface. The pressure is applied externally. Subsequent researchers used the term “pressure” to describe a condition internal to the solid—a use which shocks the thinking of an engineer attributes stress, not pressure, to solids. More on this later. 

Ice lensing also occurred with nitro-benzene and benzene, both of which decrease in volume with freezing.

He summarizes with this: 

The chief factors controlling segregation and excessive heaving are: size of soil particle, amount of water available, size and percentage of voids, and rate of cooling. Segregation occurs readily if the particle diameter is less than a micron, and under favorable conditions where particles are somewhat larger. High water-content favors segregation, and additional water may be drawn from the water-table to form very thick ice layers. Water occupying very small soil voids does not freeze readily and may be undercooled in the immediate vicinity of ice crystals. Rapid cooling when the temperature is near the freezing point prevents segregation in some soils, but has little effect if the soil particles are sufficiently small. 

This paper serves as the starting point for a large body of literature on crystal formation in porous media. Ten years prior to his work on frost heaving, he did laboratory work on the growth of salt crystals. Salt and ice have this in common—the total volume before crystallization is smaller than the total volume (liquid plus solid) once crystals are formed. The concern is raised with both, whether this expansion upon crystallization can do damage to the containing matrix. Taber was able to break a glass container by lowering the temperature of the saturated salt solution that it contained. See Figure 3.

Growing salt crystals can break their containers. Can ice do this within masonry materials? Stay tuned.

There are no equations in any of Taber’s papers. This helps make them accessible to a lay audience. But demonstrating such a curious phenomenon invited subsequent researchers to explain what happens in terms of chemical and physical thermodynamics. Here, in this blog, the effort will be made to bring this later literature within the grasp of a—face it—equation-phobic reading audience.

References

Black, Patrick B and M. J. Hardenberg. 1991. Historical perspectives in frost heave research: The early works of S. Taber and G. Beskow. U.S. Army Corps of Engineers Cold Regions Research and Engineering Laboratory. Special Report 91-23.

Nelson, F. E. and H. M. French. (no date). Stephen Taber and the development of North American cryostratigraphy and periglacial geomorphology. https://onlinelibrary.wiley.com/doi/am-pdf/10.1002/ppp.2096

Taber, S. 1929. Frost heaving. The Journal of Geology , Jul. – Aug., 1929, Vol. 37, No. 5 (Jul. – Aug., 1929), pp. 428-461 

Taber, S. 1916. The growth of crystals under external pressure. American Journal of Science. https://zenodo.org/records/1850100

Taber, S. 1917. Pressure phenomena accompanying the growth of crystals. Proceedings of the National Academy of Sciences. Pp. 297-302. https://archive.org/details/jstor-83678