There is major concern about the problem of delayed ettringite formation (DEF) in the Nation¿s infrastructure. However, there are also many controversies about the subject including the geographical extent of DEF, the severity of damage and its relationships with possible causative factors. Moreover, the relationship between DEF and alkali-silicate reaction (ASR) remains a subject of dispute. Unfortunately, there are very data points consisting of samples from actual concrete bridges that could be used to resolve these issues. Therefore, to overcome this lack of data there is an urgent need to collect more samples from actual concrete structures for analysis of DEF using a consistent and systematic approach that would enable rigorous statistical analyses In a recent study funded by the Maryland State Highway Administration (MDSHA), a pilot program for conducting a field survey of DEF was developed (see attached PDF file). A sample of 16 (ultimately 28) bridges was selected to give a uniform geographical distribution across the Maryland State Bridge districts. At each bridge several cores were drilled for subsequent examination of fracture surfaces by Scanning Electron Microscope (SEM) with energy dispersive X-ray diffraction. In most bridges, ettringite was detected, but alkali-silica reaction (ASR) gel was detected only very rarely. In more heavily damaged locations, the occurrence of ettringite crystals was more frequent, appearing in the rims around aggregates as well as in air voids. Also, the morphology of the ettringite crystals appeared to be more lamellar than acicular. The implications are that DEF is widespread geographically, map cracking is not diagnostic only for ASR, and the onset of DEF may be associated with a change in ettringite crystal morphology. Based on these preliminary findings, it is proposed to carry out a more extensive survey of concrete bridges in a number of states using the approach developed in the pilot study.

1. Establish a data base of well characterized concrete samples from 100 representative concrete bridges across North America. 2. Develop a statistically significant analysis of the geographical extent of DEF, the severity of damage and relationships with possible causative factors.

1. Develop standard methodology for selecting candidate bridges for sampling using the National Bridge Inventory and state-level PONTIS databases. 2. Develop standard guidelines for drilling cores from the bridges and conducting petrographic analysis of the cores. 3. Develop standard guidelines for nondestructive characterization of damage at concrete bridges including image analysis of close-range photogrammetry, impact-echo and potassium autoradiography. 4. Conduct destructive sampling of selected bridges with associated nondestructive testing. 5. Analyze the cores collected in the field using the standard petrographic analysis protocol. 6. Analyze the data for statistical relationships concerning the geographic distribution (ubiquity) and severity of DEF-related damage. Also analyze the data for correlations with various possible causative factors such as age, mix proportions, curing conditions etc. 7. Present findings in articles for technical journals and in presentations at professional society meetings. 8. Prepare a final technical report including recommendations for predicting DEF occurrences in existing bridges and preventing it in new concrete bridges.

Concerning the organization of the pooled fund study, FHWA is listed as the Lead Agency for the purpose of this solicitation. However, once the study gets underway, it would be preferable for a state DOT to volunteer to take the lead. Also, for reasons of economy of scale and data consistency, it would be desirable to contract with a single organization to set up a central laboratory for the petrographic analyses of the cores.