Diagnosing delayed ettringite formation in concrete structures
Introduction
In a previous paper [1] presented the results of a forensic evaluation conducted on core samples retreived from precast, prestressed concrete bridge girders. These girders were exhibiting extensive cracking, which was fortunately observed before the girders were placed in service, and were then subjected to examination by numerous investigators. It was generally agreed by all who examined samples from these girders that the concrete was undergoing some form of internal chemical attack, however, there was considerable debate as to whether the principal cause of deterioration was alkali–silica reaction (ASR) or delayed ettringite formation (DEF). Scanning electron microscopy (SEM) revealed evidence of reacted siliceous aggregate particles, alkali–silica gel and abundant quantities of ettringite filling the cracks and voids in the concrete. However, whereas there was firm evidence of expansion caused by ASR in the form of reacting aggregate particles with cracks emanating into the surrounding paste, there was little evidence of significant expansion of the paste due to DEF and it was concluded that ASR was the primary cause of distress and was a precursor to the subsequent precipitation of ettringite in the resulting cracks [1]. This type of ettringite formation is frequently referred to as secondary ettringite formation and is the result of the continuous dissolution of ettringite in fine cracks and voids and its reprecipitation to form larger crystals in bigger empty spaces [2], a thermodynamically-driven spontaneous process known as Ostwald ripening. There is no evidence to suggest that this process leads to damage.
In most of the reported “cases” of DEF some other mechanism of deterioration, usually ASR, has also been present and it is difficult to determine the actual contribution made by DEF to the damage or, indeed, whether DEF played a significant role at all in the deterioration of the concrete [1], [3]. Based on the number of confirmed cases of DEF damage in the field, it would appear that the risk of this process causing premature deterioration in real concrete structures is low compared to other more widespread deterioration processes. However, recently Sahu and Thaulow [4] diagnosed DEF as the sole cause of deterioration of precast concrete railroad ties (railway sleepers) in Sweden. This paper presents another such case, i.e. where DEF (in the absence of ASR) has been isolated as the sole contributory cause of premature deterioration in cast-in-place reinforced concrete bridge columns in North America.
Section snippets
Details of bridge columns
The bridge columns investigated in this study support a concrete viaduct in southern U.S.A., which was constructed in the late 1980s. Premature deterioration of the columns in the form of cracking was observed shortly before the concrete reached 10 years of age. Fig. 1 shows photographs of a typical column and the extent of cracking after approximately 15 years service. The actual dimensions of the columns and the extent of the damage varied from one column to the next, the cracking ranged from
Forensic evaluations
The full investigation of this structure involved a comprehensive evaluation of a large number of concrete columns, both above and below ground level, using a range of different tests conducted on core samples supplemented with in situ monitoring of the columns. In this paper, only data related to the forensic evaluation conducted for four of the columns are reported.
Discussion
There have been relatively few documented cases where the process of DEF has been unequivocally established as a cause of damage to concrete structures and in most of these cases other deterioration processes (particularly ASR) have also contributed to the observed distress [1], [3], [7]. Of the three damaged concrete columns investigated in this study, two of them were affected by both DEF and ASR. However, in one column (designated as Column A) DEF appears to have been the sole cause of
Conclusion
Delayed ettringite formation was confirmed as the sole cause of distress in one cast-in-place concrete column and as a contributory cause in two other columns. The action of damaging DEF was readily established by the presence of numerous aggregate particles completely surrounded by ettringite-filled gaps. Small-diameter (50-mm) cores drilled from the concrete showed significant expansion (0.3 to 1.3%) when immersed in limewater at normal laboratory temperature. Laboratory tests showed that
References (9)
- et al.
Delayed ettringite formation in Swedish concrete railroad ties
Cement and Concrete Research
(2004) - et al.
The effect of pozzolans and slag on expansion of mortars cured at elevated temperature: Part 1: Expansive Behaviour
Cement and Concrete Research
(2003) - et al.
Influence of the storage conditions on the dimensional changes of heat-cured mortars
Cement and Concrete Research
(2001) - et al.
Diagnosing the cause of deterioration in prestressed concrete beams—ASR or DEF?