Portland blast-furnace cement is normally manufactured by grinding together clinker, granulated slag and gypsum. In his review of slag cements for the 1960 Symposium, Kraemer dealt, inter alia, with the then current knowledge concerning separate grinding of clinker and slag. He did not deal specifically with blending of the two products in the factory or, with the more novel process of blending at construction site. In 1948-1952 the author carried out laboratory and pilot plant studies on separate grinding, and blending in the concrete mixing process. He established important advantages for this procedure compared to integral grinding. Separate grinding on an industrial scale was started in South Africa in 1953; although some difficulties were experienced initially by the users due to the novelty of the idea and also because the construction industry was not familiar with portland blast-furnace cements, the demand for the product, and for separately ground slag, has grown impressively. Of the slag more than half was separately ground. Advantages of separate grinding: 1. The grinding process is more efficient; 2. product quality is under better control; 3. ground slag has much better storage properties than portland cement; 4. proportions of portland cement and slag is at discretion of user; 5. workability can be significantly improved or water/cement ratio reduced. Of special technical and economic interest is the freedom to choose the proportions of slag in the final cement; portland cement and slag have different properties and by this choice optimum combination for any job can be decided by the user; since slag costs less to produce, this can also yield material economies. There is therefore a case for all slag intended for use in hydraulic cements to be ground separately and blended with portland cement at construction site. Modern methods of mix design, proportioning and control make this readily possible.
A specific gravity method was proposed for the determination of surfacing binder content. The principal features of this ntethod are: (a) to reduce the time taken before results on the mixture become available, and (b) to minimize the capital outlay on apparatus to carry out field plant control tests.