The Role of Magnesium in Concrete Deterioration, HR-355, 1994

Concretes with service lives of less than 15 years and those with lives greater than 40 years were studied with petrographic microscope, scanning electron microscope, and electron microprobe to determine why these two groups of concrete exhibit such different degrees of durability under highway conditions. Coarse aggregate used in both types of concrete were from dolomite rock, but investigation revealed that dolomite aggregate in the two groups of concretes were much different in several respects. The poorly-performing aggregate is fine-grained, has numerous euhedral and subhedral dolomite rhombohedra, and has relatively high porosity. Aggregate from durable concrete is coarse-grained, with tightly interlocked crystal fabric, anhedral dolomite boundaries, and low porosity. Aggregate in short service life concrete was found to have undergone pervasive chemical reactions with the cement which produced reaction rims on the boundaries of coarse aggregate particles and in the cement region adjacent to aggregate boundaries. Textural and porosity differences are believed to be chiefly responsible for different service lives of the two groups of concrete. The basic reaction that has occurred in the short service life concretes between coarse aggregate and cement is an alkali-dolomite reaction. In the reaction dolomite from the aggregate reacts with hydroxide ions from the cement to free magnesium ions and carbonate ions, and the magnesium ions precipitate as brucite, Mg(OH)2. Simultaneously with this reaction, a second reaction occurs in which product carbonate ions react with portlandite from the cement to form calcite and hydroxide ions. Crystal growth pressures of newly formed brucite and calcite together with other processes, e.g. hydration state changes of magnesium chloride hydrates, lead to expansion of the concretes with resultant rapid deterioration. According to this model, magnesium from any source, either from reacting dolomite or from magnesium road deicers, has a major role in highway concrete deterioration. Consequently, magnesium deicers should be used with caution, and long-term testing of the effects of magnesium deicers on highway concrete should be implemented to determine their effects on durability.

Time-Temperature Strength-Reaction Product Relationships in Lime-Bentonite-Water Mixtures, HR-111, 1965

The interrelation of curing time, curing temperature, strength, and reactions in lime-bentonite-water mixtures was examined. Samples were molded at constant density and moisture content and then cured for periods of from 1 to 56 days at constant temperatures that ranged from 5C to 60C. After the appropriate curing time the samples were tested for unconfined compressive strength. The broken samples were then analyzed by x-ray diffractometer and spectrophotometer to determine the identity of the reaction products present after each curing period. It was found that the strength gain of lime-clay mixtures cured at different temperatures is due to different phases of the complex reaction, lime & clay to CSH(gel) to CSH(II) to CSH(I) to tobermorite. The farther the reaction proceeds, the higher the strength. There was also evidence of lattice substitutions in the structure of the calcium silicate hydrates at curing temperatures of 50C and higher. No consistent relationship between time, temperature, strength, and the S/A ration of reaction products existed, but in order to achieve high strengths the apparent C/S ration had to be less than two. The curing temperature had an effect on the strength developed by a given amount of reacted silica in the cured lime-clay mixture, but at a given curing temperature the cured sample that had the largest amount of reacted silica gave the highest strength. Evidence was found to indicate that during the clay reaction some calcium is indeed adsorbed onto the clay structure rather than entering into a pozzolanic reaction. Finally, it was determined that it is possible to determine the amount of silica and alumina in lime-clay reaction products by spectrophotometric analysis with sufficient accuracy for comparison purposes. The spectrophotometric analysis techniques used during the investigation were simple and were not time consuming.