Analysis of the bearing response test for polymer matrix composite laminates: bearing stiffness measurement and simulation B. Vangrimde a , R. Boukhili b, * a Department of Metallurgy and Materials Engineering, Katholieke Universiteit Leuven, Kasteelpark Arenberg 44, 3001 Leuven, Belgium b Mechanical Engineering Department, Centre for Applied Research on Polymers, Ecole Polytechnique de Montr eal, Station Centre-Ville, Box 6079, Montr eal, Que., Canada H3C 3A7 Abstract This paper is the first part of a project that aims to investigate the mechanical and fracture behaviour of bolted joints in general purpose glass fibre-reinforced polyesters (GRP). In the present study a procedure is set up to measure the bearing stiffness of a GRP laminate in a single-bolt double lap joint. With a three-dimensional finite element model it is shown that the bolt and fixture de- formations affect the stiffness results. Hence the experimental displacement data were corrected before calculating the coupon bearing stiffness. The coupon bearing stiffness was also simulated by a two-dimensional finite element model. Provided that bolt– hole clearance, material non-linearity and bolt–hole friction are taken into account, good agreement is observed with experimental data. Bearing strain and bearing stiffness are based on the bearing deformation of the coupon, not on the hole elongation. This makes the stiffness data useful for design and allows an easy installation of the displacement measurement devices. Ó 2002 Pub- lished by Elsevier Science Ltd. Keywords: Bolted joints; Bearing deformation; Bearing stiffness; GRP; Finite element method 1. Introduction When testing bolted joints made with composite ma- terials, the interpretation of the results depends heavily upon the method used to generate the load–displace- ment response. While the load measurement is made directly via a load cell and can be trusted, this is not exactly the case for the displacement, which is measured mainly with linear variable differential transformers (LVDTs) or extensometers. In fact, the problem is not with the LVDTs or extensometers but rather with the choice of the reference points for the displacement measurement. Obviously, the deformations that are ac- tually measured depend on the reference points taken for the displacement transducers. Since extraneous dis- placements in the experimental set-up cannot be avoided and since the stress field around the hole of a bolted joint is not uniform, an arbitrary choice of these refer- ences points makes comparison between different data sources tedious. In addition, the obtained data may be of little help for design purposes. The answer to such ambiguity is usually addressed by the development of standard test methods or may result from a consensus between researchers. A literature survey shows that Crews [1] and, more recently, Kallmeyer and Stephens [2] have managed to measure the elongation of the bearing hole. This is accomplished by using a stiff wire that passes through the hole via a slot in the washers, placed between the coupon and the fixture plates. However, this method does not seem attractive, mainly because it is difficult to set up. Consequently, most au- thors choose to measure the displacement at a fixed distance beside [3] or beneath the hole [4–7]. The ASTM D5961 standard test method for bearing response of polymer matrix composite laminates [8] suggests the use of the latter approach but without specifying the exact location of the reference points. When measuring at a certain distance from the hole, the measured displace- ment will include hole deformation as well as plate de- formation [9]. This is problematic since, in the ASTM standard, the bearing strain is based on the hole defor- mation but no data reduction procedure is worked out to separate the hole deformation from the plate defor- mation. Actually, it is not the hole elongation which interests the designer but rather the flexibility of the joint. Indeed, * Corresponding author. Tel.: +1-514-340-4757; fax: +1-514-340- 5867. E-mail address: rachid.boukhili@polymtl.ca (R. Boukhili). 0263-8223/02/$ - see front matter Ó 2002 Published by Elsevier Science Ltd. PII:S0263-8223(02)00020-X Composite Structures 56 (2002) 359–374 www.elsevier.com/locate/compstruct