Linear and non-linear SVM prediction for fresh properties and compressive strength of high volume fly ash self-compacting concrete Mohammad Azimi-Pour ⇑ , Hamid Eskandari-Naddaf ⇑ , Amir Pakzad Department of Civil Engineering, Hakim Sabzevari University, Sabzevar, Iran highlights Properties of HVF-SCC predicted from VVF-SCC data by using SVM method. Appropriate inputs was used and kernel coefficients calculated by grid search. New models were compared by previous methods and validated by experimental results. article info Article history: Received 28 May 2019 Received in revised form 18 August 2019 Accepted 15 September 2019 Keywords: Support vector machine (SVM) Self-compacting concrete (SCC) Fresh properties Compressive strength High volume fly-ash (HVF) Various volume fly-ash (VVF) abstract Support vector machines (SVMs) have recently been used to model the properties of low volume fly ash self-compacting concrete (LVF-SCC) by means of kernel functions to minimize the experimental work. Appropriate linear and non-linear SVM models with different kernels (linear, polynomial, radial basis and sigmoid) were proposed in this paper to predict the fresh properties and compressive strength of high volume fly ash SCC (HVF-SCC) from various volume fly ash SCC (VVF-SCC). Since most available data contain relative low volume fly ash, new SVM models were trained on VVF-SCC mixture proportions adapted from literature, and the prediction results were compared to those of previous models for HVF-SCC and LVF-SCC. Moreover, an experimental plan containing six different SCC mixtures was estab- lished and the proposed SVM models were validated by experimental results, including compressive strength, L-box, slump, U-box, and V-funnel. Results showed that new SVM models provide better out- comes if considering appropriate input vector and wide range data to obtain the proper kernel function coefficient to predict the various properties of HVF-SCC. Among the kernel functions, prediction results of the SVM – RBF model were more accurate compared to other kernels. Ó 2019 Elsevier Ltd. All rights reserved. 1. Introduction Self-compacting concrete (SCC) has special properties, such as high flow-ability under its own weight, consolidation around rein- forcement without the need for vibration, constructability, better pump ability and being economical [1–4]. These exclusive proper- ties make SCC a useful material in many members of a project using rebar congestion [5,6]. Generally, characteristics of SCC are defined by fresh properties related to workability or rheological parameters [7–9] which are assessed with tests based on workabil- ity, such as slump, U-box, orimet, L-box and V-funnel [10–14]. On the other hand, SCC mixtures contain coarse and fine aggregate, fly ash, high-range water reducers (HRWR) and powder materials in which the amount of each parameter has a significant effect on fresh and hardened properties of SCC [15]. Usually, using low volume fly ash (LVF) around 15–25% as replacement of cement in SCC improves workability and reduces the required HRWR which is applicable for any SCC structures [16]. Also, it has been proved that using higher volume of fly ash (HVF) increases the slump of the concrete mixture and reduces cracking of concrete due to lower heat of hydration [17,18]. Accordingly, SCC with high volume fly ash is used in road construc- tion, marine structure, dams, foot bridges and precast thin struc- tures [19–21]. Since in some of these structures the permeability is an important factor to increase durability, micro silica can be used to improve the hardened properties and durability of SCCs due to pozzolanic reactivity and filler role [22–24]. Due to the advantages of SCC in the construction industry and considering that treatment of this kind of concrete is complex, pre- diction of the fresh properties and compressive strength of SCC, https://doi.org/10.1016/j.conbuildmat.2019.117021 0950-0618/Ó 2019 Elsevier Ltd. All rights reserved. ⇑ Corresponding authors. E-mail addresses: Mohammadhsu@gmail.com (M. Azimi-Pour), Hamidiisc@ yahoo.com, h.eskandari@hsu.ac.ir (H. Eskandari-Naddaf). Construction and Building Materials 230 (2020) 117021 Contents lists available at ScienceDirect Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat