What causes crazing?
G. H. Tattersall, Senior Lecturer at the University of Sheffield, has recently reported to Concrete magazine (published by the Cement and Concrete Association in England) that the most important climatic factor is the relative humidity of the drying period in a wetting and drying cycle. Conditions become more severe as the relative humidity becomes lower. For this reason crazing is most likely to occur in northern climates in the spring, when periods of rain are followed by periods of low relative humidity. It is least likely to occur in November when the relative humidity tends to remain high. A surface layer rich in cement will experience greater stress than one not so rich and therefore will be more subject to crazing. For this reason steel-troweled surfaces are more susceptible to crazing, whereas surfaces from which the cement-rich portion has been removed by etching or mechanical treatment are less so. He points out that anything that reduces the gradient of moisture concentration will reduce the stress and the tendency to craze. Extremely porous materials (deficient in fines) cannot support a concentration gradient and do not craze. On the other hand materials that contain 15 percent or more particles passing the Number 100 sieve or materials that have been well treated with a waterproofer may be sufficiently impermeable to prevent a moisture gradient from becoming established and these also remain free of crazing. It is the materials in between that are most susceptible. It appears that such factors as carbonation, temperature changes, and age at which the concrete is first exposed have little to do with the matter.