Brewer Stadium, Appalachian State University, Boone, NC, was constructed in 1962 using steam-cured precast lightweight concrete channel sections for the bleacher seats. Spalling of the bleachers was observed within a few years. By 1995, approximately 375 of 860 precast elements exhibited sufficiently severe cracking and delamination that reconstruction of the stadium was the most economical alternative.
The destructive process causing the cracking and delamination of the 375 units is DEF. The (fibrous, crystalline) ettringite occurs in the concrete intergrown in the paste, in veins and cracks throughout the mortar, filling voids in the mortar, and as rims in the paste around coarse- and fine-aggregate particles. Also observed, in the same spatial relationships as the crystalline ettringite, and commonly intergrown with it, was a clear, gel-like material with low-order, mottled birefringence. This appears to be the same proto-ettringite material as described by Mielenz et al. (1995).
Visibly undistressed members exhibit incipient forms of the same DEF features that are well developed in the distressed members. Sulfate analyses show no difference between distressed and undistressed members.
The physical and chemical circumstances of the failure conform to the model presented by Heinz and Ludwig, whereby cements capable of promoting DEF meet the sulfate limitations of ASTM Specification for Portland Cement (C 150), and DEF occurs due to excessive temperature reached during heat curing of concretes made with cements of a particular sulfate to alumina ratio.
These results support the view that revisions to current standards for portland cement are unnecessary, and that what is necessary is adherence to maximum temperature guidelines during production of heat-cured concrete products.
Petrographic evidence supports the view that expansion due to DEF is caused by swelling of a proto-ettringite gel—by water absorption, similar to alkali-silica gel—that forms in the paste and that later, following expansion and crack formation, crystallizes in the cracks and peripheral separations, forming the familiar veins and rims of fibrous ettringite.
This mechanism speaks to the origin of the expansive force in the general case of sulfate attack, and to why the formation of the delicate, fibrous, ettringite crystals in the cracks is not the mechanism that exerted sufficient force to cause the cracks in which the ettringite occurs.