Cement and Concrete Research 29 (1999) 1313–1321 0008-8846/99/$ – see front matter © 1999 Elsevier Science Ltd. All rights reserved. PII: S0008-8846(99)00154-4 Alkali-activated slag mortars Mechanical strength behaviour A. Fernández-Jiménez a, *, J.G. Palomo b , F. Puertas a a Instituto Ciencias de la Construcción Eduardo Torroja (CSIC), 28033 Madrid, Spain b E.U. Arquitectura Técnica (UPM), Av. Juan de Herrera No. 6, 28040, Spain Received 17 September 1998; accepted 12 July 1999 Abstract The objective of the present work is to know the joint influence of a series of factors (specific surface of the slag, curing temperature, activator concentration, and the nature of the alkaline activator) on the development of mechanical strengths in alkaline-activated slag ce- ment mortars. To reach this aim, a factorial experimental design was carried out (a complete 2 3  3 1 design) for every age studied (3 to 180 days). Through the variance analysis, the most significant factor on the response turned out to be the alkaline activator nature. The activator used, Na 2 SiO 3  nH 2 O + NaOH, was the factor that gave the highest mechanical strengths in all tests. The next most statistically significant factor was the activator concentration, followed by curing temperature, and, finally, the specific surface of the slag. The equa- tions of the model describing the mechanical behaviour for flexural and compressive strengths and their relationships for each age studied were established © 1999 Elsevier Science Ltd. All rights reserved. Keywords: Granulated blast furnace slag; Alkali-activated cement; Mechanical properties; Modeling 1. Introduction The industrial manufacturing process of cements based on the alkaline activation of blast furnace slag (AAS) started in Ukraine between 1960 and 1964 [1]. The utilisa- tion of these types of cements has solved an important eco- logical problem: the use of an industrial subproduct. The use of these cements also presents economical advantages due to the lower energy cost of their production, signifi- cantly lower than that of Portland cement. These AAS ce- ments present some technological advantages over ordinary Portland cements. These are: the development of earlier and higher mechanical strengths, lower hydration heat, better re- sistance to chemical attack, better behaviour upon carbon- ation, higher resistance of the aggregate-matrix interface, better behaviour to freeze-thaw cycles, among others [2–4]. However, they also present some problems such as rapid setting, high shrinkage, subsequent formation of micro- cracks, the possibility of expansive reactions occurring be- cause of alkali-aggregate reactions, and higher formation of salt efflorescences. Extensive investigations have been published about the main factors affecting the development of mechanical strengths of AAS mortars and concretes [5]. These factors include slag specific surface, curing temperature, activator concentration, and alkaline activator nature. In all the stud- ies, the influence of these factors are independently studied. For this reason, the objective of the present work is to learn the joint influence of a series of factors (specific surface of the slag, curing temperature, activator concentration, and nature of alkaline activator) on the development of mechani- cal strengths of alkaline-activated slag cement mortars. Like- wise, mathematical models are established, describing the mechanical behaviour of the mortars as a function of the vari- ables considered at different ages studied. 2. Methods 2.1. Materials The chemical composition of Spanish blast furnace slags used in this work is shown in Table 1. The mortar was pre- pared according to the Spanish standard UNE-80-101-88. The size of the prisms was 4  4  16 cm. The aggregate/slag ra- tio used was 2/1 and the alkaline solution/slag ratio was 0.51 for the 450 m 2 /kg slag and 0.61 for the 900 m 2 /kg slag. The samples were cured at 25 and 45°C for 20 h, then they were kept at ambient temperature in a humid chamber (98%) until the testing day. Flexural and compressive mechanical strengths were measured at 3, 7, 28, 90, and 180 days. * Corresponding author. Tel.: +34-1-302-04-40; fax: +34-1-302-60-47. E-mail address: pesfjs8@ietcc.csic.es (A. Fernández-Jiménez)