We have been given to understand in the past that calcium chloride eats the cement that binds the aggregate in much the same way battery acid can eat a hole in our trousers. Last summer and fall we put in a plaza area of about 3,000 square feet of exposed aggregate concrete using pebbles up to two to five inches in diameter with an oval shaped face and a cross section which was a very flat ellipse. These were laid in a bed of mortar made with one part colored cement, 2 1/2 parts of medium sand, and an admixture used to improve bond to the very smooth pebbles. The stones were generally laid point to point in the mortar and exposed as deeply as could be done while leaving enough mortar around them to lock them down. The surface was wiped with a sack to eliminate slickness and two coats of a transparent acrylic coating were applied over it. The maintenance superintendent applied chemical deicers other than sodium chloride. Drainage or runoff was impossible and the chemicals ate straight down. A testing laboratory hired by the owner reported 0.10 percent calcium chloride in the hard areas of mortar and 0.63 percent in the deteriorated areas. None had been used in the mix. It is our contention that the cement was eaten up by the calcium ch loride. Are there more comprehensive test records of the effects of both calcium chloride and sodium chloride?
Verbeck and Klieger, in Research Bulletin 83 of the Portland Cement Association, have reported that at the concentrations in which the salts cause most deicer damage during freezing and thawing—say two to four percent—sodium chloride and calcium chloride do roughly the same amount of damage, but calcium chloride is slightly less damaging, if anything, than sodium chloride. Furthermore, the effects of both calcium chloride and sodium chloride on deicer scaling are primarily physical rather than chemical. By far the most effective way to improve the resistance of concrete to scaling from deicers is to entrain a sufficient amount of air in the concrete. Changes in the mix or kinds of materials used in the mix make some difference but their effect is very minor compared to that of entrained air. In another study, reported by Klieger in Research Bulletin 82 of the Portland Cement Association, the length of the curing period was found to affect the resistance to deicer scaling of both air entrained and non-air entrained concrete. A common modern recommendation, based partly on this work and partly on later field experience, is that concrete should be cured according to specifications, then allowed to become completely dry at least once and to attain an age of at least three months before the first application of deicer is permitted. Though linseed oil, toppings, or some other coatings can protect against deicer scaling, air entrained concrete should be used wherever possible as the first line of defense. If the chemical deicers you mention included other materials than calcium chloride there is a possibility that there was chemical attack. Ammonium chloride, ammonium nitrate and many other ammonium salts act as acids to attack concrete and dissolve the cement. If ammonium chloride were used it would probably show up as calcium chloride during chemical analysis of the mortar. Owners should be emphatically warned about the highly damaging effects of ammonium salts.