COMMENTARY TO Discussion of the paper titled ‘‘Design of clay/cement mixtures for extruded building products by Khelifi et al.’’ Adnan C¸ olak • Turgay Cos¸gun Received: 12 October 2012 / Accepted: 15 December 2012 / Published online: 8 January 2013 Ó RILEM 2013 Abstract In this study, the paper of Khelifi et al. is discussed. The discussion is about the mathematical models. Keywords Cement  Clay  Concrete  Compressive strength Khelifi et al. [1] obtain high compressive strength values for the cement-clay mixtures. This result is in agreement with the experimental data obtained by Kawamura and Kasai [2]. They also use mathematical models to estimate the compressive strengths of these products. Firstly, I would like to thank to Khelifi et al. due to their comments on my mathematical model. However, in the discussion of this model, they reach an erroneous conclusion. This is due to inaccurate selection of magnitude of the adjustable parameters in the model. There is also a problem in defining the mathematical models. For example, the formulation of Neville model is not given in the study. Here, the mathematical models in question are redefined and the calculation procedure for clay–cement systems is given in detail. Neville model [3, 4] for the compressive strength of Portland cement concrete is given by r ¼ k w=c ð1Þ where ‘‘r’’ represents the compressive strength of Portland cement concrete, ‘‘k’’ is a constant and ‘‘w/c’’ is the water to cement ratio. The results of this hyperbolic relation are confirmed by Alexander and Ivanusec [5] and by Kakizaki et al. [6]. On the other hand, C¸ olak [7] developed the following equation for the compressive strength of Portland cement concrete. r ¼ a w c b þ k w c   n ð2Þ where ‘‘a’’, ‘‘b’’, ‘‘k’’ and ‘‘n’’ are the adjustable parameters. The magnitudes of these parameters change depending on the characteristics of the strength-w/c curve seen in Fig. 1.C¸ olak [7] has shown that Eq. (2) gives a maximum compressive strength at a calculated (w/c) m value which is very close to the non- evaporable water content by mass of cement. Due to these characteristics of Eq. (2), the accurate estimates can be made for the compressive strengths of Portland cement systems in a wide range of w/c ratio. Despite this fact, Khelifi et al. claim that C¸ olak model [Eq. (2)] does not match the experimental data for low w/c ratios. This comment refers to the article available at doi:10.1617/s11527-012-9949-4. A. C¸ olak (&)  T. Cos¸gun Department of Civil Engineering, Faculty of Engineering, I ˙ stanbul University, 34320 Avcılar, Istanbul, Turkey e-mail: adnan@istanbul.edu.tr Materials and Structures (2013) 46:513–515 DOI 10.1617/s11527-012-0002-4