Dicalcium silicate in blast-furnace slag: A critical review of the implications for aggregate stability

Abstract

Small amounts of β dicalcium silicate can occur in certain blast-furnace slags and in some cases it may invert expansively to the γ form during cooling, fracturing the surrounding material if sufficient inversion occurs. The consequent fragmentation ('Falling') of the slag may take some days to become fully developed. If the slags are to be used as aggregates, a test is necessary to allow any such weakened batches to be distinguished at an early stage from those which are sound. In 2004 BS 1047, which imposes chemical limits on air-cooled blast-furnace slag, will be replaced by a European Standard with an ultraviolet lamp test for 'Falling' slags. A review of the literature is employed to establish what test is most appropriate so that blast-furnace slag can contribute best to the growing need for secondary aggregates. In the 1960s a belief developed in Britain that β dicalcium silicate that did not invert during the cooling of a slag might do so destructively at some later date - a postulated phenomenon known as 'Late falling'. This belief has dominated attitudes to specifications for blast-furnace slag in Britain (but nowhere else) since then. Examination of the origin of this belief shows it to have been based initially on misinterpretations of earlier work and subsequently on uncritical repetition from one publication to the next. In fact, there is no conclusive evidence that this phenomenon has ever occurred, let alone affected the quality of an aggregate. For rapid identification of weakened slags the ultraviolet test appears suitable. The chemical clauses and the microscope test of BS 1047 have no obvious value in indicating slag quality to customers and their image of mathematical precision is also undeserved.