Roberts – Admixtures and C Ash – A Challenge to Model

Lawrence R. Roberts

Roberts Consulting Group

Acton, MA


Efforts to model cement hydration continue to increase. The majority of published work concentrates on silicate hydration, which makes sense as the resulting C-S-H is the main strength producing phase. However, the majority of problems encountered with silicate hydration in real-world concrete, especially those involving unpredicted behavior with admixtures, and for which better models are needed, appear to involve the aluminates phases in complex ways. Specifically, the balance of aluminate phase activity with available sulfate appears critical to effective action of many admixtures. The potential for aluminate action to control silicate has been known for many years; indeed Lerch’s classic 1946 paper shows clearly, and without the presence of admixtures, the depression of silicate hydration possible in under-sulfated systems.

Today the cement industry is faced with increasing pressure to provide strength development potential with reduced environmental impact – specifically climate change impact based on CO2 emission. This will drive the use of higher and higher levels of supplementary cementitious materials such a natural and waste pozzolans, slag etc., in order to drive down the clinker component of the final concrete. The aluminate activity of these materials can have a marked effect on silicate hydration behavior, leading to some very counter intuitive and difficult to model results, including such anomalous behavior as extended set and severely slowed strength development as temperature rises. The most frequent example of such problems in the author’s experience is in the combination of ASTM C 618 Class C fly ashes, especially at relatively high ratios to clinker, in concrete containing admixtures.

This contribution focuses on several examples of this behavior, taken both from the author’s experience and the literature.  The goal is to bring before the cement hydration modeling community this question – how can we use this odd behavior to challenge the models being developed, and by so doing both advance the technology of modeling, and create models which will help the industry meet today’s and tomorrow’s environmental requirements?



Leave a Reply

You must be logged in to post a comment.