Issue 48

V. M. G. Gomes et alii, Frattura ed Integrità Strutturale, 48 (2019) 304-317; DOI: 10.3221/IGF-ESIS.48.30 308 fixed on cover plates. The slip factor values, slip µ , and the characteristic values, c µ , of the slip factors that accounts for 5% of occurrence probability, for a confidence interval of 75%, are obtained by Eqns. (1) and (2): slip slip bolt shear planes preload F µ = n ×n ×F (1) c slip AVG µ µ =µ -2.05×S (2) where the slip F is the load applied by machine test leading to slip, bolt n is the number of bolts, shear planes n is the number of shear planes (equal to 2), preload F is the preload applied to the connection, slip AVG µ is the average value and µ S the standard deviation of the slip factors obtained for all the tests. In accordance to EN 1090-2 standard, preload force may be applied from different methods: torque method; combined method; HRC method or direct tension indicator method. In this study, torque method was considered, and the calibration of the preload level was performed. The torque clamping, clamping M , is related with bolt nominal diameter bolt d , preload force and the calibration coefficient C as shown in Eqn. (3). Before performing the monotonic static and slip tests, the calibration of the preload level for two preloads was measured in order to obtain the calibration coefficient C , using a dedicated load cell. The results confirmed a coefficient C equal to 0.2. In addition, tensile tests on samples of the steel grades used in the bolted joints were also performed to evaluate the differences on their mechanical behaviours. Tensile tests were performed with extensometer with gauge length equal to 50 mm. clamping bolt preload M =C×d ×F (3) Figure 3 : Geometry of the specimens for slip tests (left), test setup with LVDTs (centre) and scheme of the test setup (right) according EN1090 standard. N UMERICAL MODEL DESCRIPTION he finite element method was used to simulate the behaviour of the bolted connections. The main goal of this study was to determine the ultimate strength of the joints and then compare it with the average ultimate loads obtained by experimental tests. ANSYS 18.2 was used in this study and the components of the joints were modelled with hexahedral meshes composed by 20-noded solid elements (SOLID186). Contact surface elements (CONTA174 and TARGE170) composed by 8 nodes were also used to model the friction/contact behaviour between sliding faces of the plates and fastening. PRETS179 elements were inserted on middle of bolt body to apply the bolt pretension. The used numerical models are detailed below. T