Evaluations of concrete mix stability for below-grade applications

Drilled shafts are a commonly utilized foundation type for bridge construction, especially for bridges involving large loads or difficult ground conditions. Proper construction is required to ensure satisfactory performance of the shafts, and concrete placement is among the most critical aspects of drilled shaft construction. Concrete has evolved to continuously be effective in the construction of modern drilled shafts as they continue to grow larger. Previous issues with concrete used in drilled shafts included flowability and ensuring the concrete was uniformly spread; these issues have been resolved with the introduction of high performance self-consolidating concrete but has introduced new problems with bleed and segregation. Concrete segregation describes the separation of concrete components causing a non-uniform mix. The term 'bleed' is a type of segregation but is specifically used to describe the separation of water from the concrete mix and can be seen, in varying amounts, rising to the surface of certain concrete pours. Occurrences of bleed and segregation can compromise the integrity of a drilled shaft. The research described in this report was initiated with the objective of quantifying bleed and segregation by addressing three questions: what is the best method to predict and quantify bleed and segregation of a concrete mix, what characteristics of a mix design cause the most bleed and segregation, and what are the consequences of bleed and segregation. A better understanding of concrete bleed and segregation predictors and effects can lead to more reliable drilled shaft concrete mix designs. To analyze these objectives, ten different concrete mixes were used to test and create scaled drilled shafts measuring 4 feet in diameter and 42 inches tall. Each concrete mix was run through a series of quality assurance/control (QA/QC) tests to then compare to data collected while the test shafts were being cured, which includes the total volume of bleed water produced, the bleed water flow rate, peak temperature in the concrete and pore pressures developed in the concrete. When comparing these bleed water test results with QA/QC tests and isolated mix design characteristics, conclusions can be made about the relationship between them. No current criterion exists that characterize what amount of bleed water is considered "excessive", but for this test program 10 percent of the total water in the test shaft was deemed excessive. Conclusions relating to QA/QC tests predicting bleed water: The ASTM Static Bleed test produced the most consistent results for predicting the amount of bleed a concrete mix would produce on a larger scale, with a cutoff quantifier of 1 percent bleed water in the test being used to distinguish a bleed prone batch. The Slump, Slump Flow, and Bauer Filter Press tests provide some ability to discriminate between bleed prone concrete mixes, with some exceptions with false positives and negatives. Findings from this research suggest that any mix with values greater than a 10-inch slump, 22-inch slump flow, or 40mL from the 5-minute Bauer test will likely produced excessive bleed. Although these tests are not the most reliable predictors, they are the most practical QA/QC test for concrete bleed at this time. This concludes that the biggest contributor to bleed water within a specific mix is the concrete's relationship with set-up versus time; the only QA/QC test that account for this relationship are the static bleed test and Bauer Filter Press. Conclusions relation to concrete mix design characteristics: A strong relationship exists between peak temperature, time to peak temperature, and the amount of bleed water produced. Both peak temperature and time to peak temperature are consistent predictors of how much bleed water is being produced in the concrete mix. Bleed water produced increases as time to peak temperature increases, and peak temperature decreases. All test mixes that exceed a time to peak temperature greater than 1100 minutes (18.3 hours) produced at least 10 percent bleed water by volume. The biggest variable effecting this cure time is the retarding admixture: the more retardant added, the more bleed water is produced. With one exception, all test mixes with 20 oz/CY or more retarder exceeded 10 percent bleed water. Another relationship contributing to cure time and peak temperature lies within the water/cement ratio (w/cm), which shows that the higher the w/cm ratio, the lower the peak temperature. Generally, aggregate moisture corrected w/cm ratios greater than 0.50 produces excessive bleed. More study is needed to confirm this finding, but mixes with fly ash contents exceeding 150 lb/CY (with concurrent reduction of Portland Cement content) all produced excessive bleed, which suggest that fly ash content should be carefully considered. The quantity of bleed water produced from test mixes decreases for mixes with higher coarse aggregate contents and lesser fine aggregate contents. This is contrary to current practices with self-consolidating concrete that use a higher fine aggregate content than traditional mixes. The consequences of excessive bleed are still unknown. Bleed channels or other compromising anomalies relating to bleed were searched for via sawing through the test shafts post curing however, no conclusive findings can be drawn from these specimens at this time. More study is necessary for finding relationships between excessive bleed and concrete structural stability.