Performance analysis of load–strain models for circular columns confined with FRP composites Pedro Faustino Marques a,b,⇑ , Carlos Chastre b a Escola Superior de Tecnologia do Barreiro – Instituto Politécnico de Setúbal, Portugal b Department of Civil Engineering, FCT/UNIC, Universidade Nova de Lisboa, Portugal article info Article history: Available online 22 May 2012 Keywords: Analysis oriented model Circular RC columns Axial confinement modelling Dilation behaviour Design oriented model FRP abstract The use of FRP composites for the confinement of concrete has become an important aspect to consider on strengthening of concrete columns. It is important therefore that accurate modelling tools are avail- able for the design of this system considering, not only the peak values of load and strain, but also the complete stress–strain behaviour. A wide group of authors have proposed several models specific for FRP-confined concrete based either on theoretical assumptions (analysis-oriented-models – AOMs) or on mathematical calibration from testing results (design-oriented-models – DOMs). This article carries out the implementation and analysis of nine existing models for circular concrete columns in view of axi- ally tested reinforced concrete columns confined with CFRP with three different diameters: 150; 250 and 400 mm. The global shape of curves, peak compressive load, stress–strain relation, axial-to-lateral rela- tion and dilation response were studied to conclude which models’ curves were closer to tests. Quanti- fication of errors in face of the testing results was carried out for the most important parameters – ultimate load, strain and lateral stress – as well as for other curve parameters. Some models are accurate in predicting the peak load, though only few can accurately predict the load–strain and dilation behaviour. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction The study of confining materials such as FRP for concrete columns has been the aim of several authors’ research in order to enhance these elements’ strength and ductility. Considering the importance of design calculations for new structures or the strengthening of exist- ing ones, the modelling prediction of the performance of circular con- crete columns subjected to axial compression is proposed by different authors as regards stress–strain behaviour. The existing types of confining models are based on different premises and quantification of the properties of the materials and structural systems involved. Consequently, the approximation of these models’ data against the real behaviour of tests is different for each model and with different influence on the several proper- ties that characterise and quantify the performance of confined concrete columns. This evidence will be shown ahead in this article. As in other structural systems, such as slabs and beams, the strengthening effect can be active or passive. In the specific case of actively-confined columns, stress state is laterally applied and externally controlled [20], being the lateral expansion restrained, while the axial stress increases. Several of the existing confining models are based on this principle. On the other hand, in passive confinement the confining stress of the strengthening is activated by the lateral expansion of the concrete core. In practice this is the behaviour of concrete cores confined with steel or FRP strips or jackets, though with distinct performance between these two materials expressed through the dilation properties and explained in Section 4.4 [15,16,24,26]. The existing confining models under axial compression may be divided into two groups [9]: (i) analysis-oriented models (AOM) and (ii) design-oriented models (DOM). In the first group most of these theoretical models are based on stress–strain curves of con- fined concrete obtained from active confinement curves by use of an incremental numerical process [11,15,16,26,4,27,10]. The sec- ond group includes models that are based on passive confinement and in which peak/ultimate stress and strain are first determined being then the non-linear behaviour is mathematically calibrated with experimental data [8,24,28,23,9,14,3]. The discussion around advantages or disadvantages between AOM and DOM in modelling the confinement under axial compres- sion is still open, although some authors clearly consider AOM more accurate and DOM easier to implement due to their direct use in design calculations [9,7]. 0263-8223/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compstruct.2012.04.036 ⇑ Corresponding author at: Escola Superior de Tecnologia do Barreiro – Instituto Politécnico de Setúbal, Portugal. E-mail addresses: pedro.marques@estbarreiro.ips.pt (P.F. Marques), chastre @fct.unl.pt (C. Chastre). Composite Structures 94 (2012) 3115–3131 Contents lists available at SciVerse ScienceDirect Composite Structures journal homepage: www.elsevier.com/locate/compstruct