We have learned to live with drying shrinkage, one of the scourges of concrete, and have learned to accommodate it, compensate for it, control its magnitude, and dissipate stresses associated with it.

Shinkage-compensating cements first were invented by Alexander Klein as Type K around 1960, with other versions developed soon after now known as Types “N” and “S.” Since the innovation of these cements, there have been periodic successes of their use in floor slab and pavement construction. When the cements are used properly and the concrete mixes designed accurately, and with joint spacings larger than 120 feet, projects have proven successful.

Experience has cautioned about unpredicted problems when these cements were developed initially. Specifications for these cements are in ASTM C 845, “Standard Specification for Expansive Hydraulic Cement,” while concrete and reinforcing steel designs are included in ACI 233, “Standard Practice for Shrinkage-Compensating Concrete.”

The basic premise upon which the system works is an early and controlled slight concrete expansion resulting from the expansive component known as kleinite (essentially anhydrous calcium sulfoaluminate) reacting with water and calcium to form ettringite (the primary hydration product of the expansive component of the Type K cement that causes the slight expansion). Reinforcing steel is put in tension by the expansion, which compresses the concrete and counters cracking as long as the concrete remains in compression.

Slabs on grade 6 inches thick with joints spaced of 120 feet apart were constructed using Type K cement. Within a month or two, the slabs cracked extensively and curled, except for two sections.

Laboratory investigations were completed on cores from the cracked and curled slabs and the two noncracked slabs. Petrographic examinations revealed there is conformance to materials requirements; hydration of the portland and kleinite cements are normal; ettringite is present; and the position of reinforcing bars in the slabs is correct.

Chemical analyses indicated sulfate contents are at typical levels when Type K cements are used. Everything about the Type K concrete system appeared normal and appropriate for the shrinkage-compensating cementitious system. As a final analytical evaluation, infrared analyses of solution extracts of the concretes were completed.

An abnormally large amount of triethanolamine, a component in some water-reducing admixtures, was found in slabs that cracked, and a normal amount in slabs that did not crack. This indicated an excessive amount of an admixture containing triethanolamine had been added to concrete represented by the cracked slabs.

U.S. Ramachandran reports in “Hydration of Cement—Role of Triethanolamine” that triethanolamine accelerates the formation of ettringite. Based upon that information and the laboratory finding of relatively large amounts of triethanolamine in the cracked and curled concrete, the conclusion postulated was that excessive amounts of triethanolamine triggered early formation of ettringite before the concrete hardened sufficiently so the concrete expansion needed to make the system work did not occur.

Consequently, the concrete was not restrained, was never in compression, and acted as normal concrete with unsuitably large-spaced joints—it cracked and curled as would be anticipated because of the large joint spacings. Erlin provides a detailed description of the laboratory investigation in “The Magic of Investigative Petrography: The Practical Basis for Resolving Concrete Problems.”

Next month find out how cherry-picking analytical data can be dangerous if some of it is slighted in explanations and interpretation, and speculation contravenes the “truth.”

Bernard Erlin is president of The Erlin Co. (TEC), Latrobe, Pa., and has been involved with all aspects of concrete for 52 years.

William Hime was a principal with Wiss, Janney, Elstner Associates and began working as a chemist at PCA 58 years ago.