PSI - Issue 37

Karina Buka-Vaivade et al. / Procedia Structural Integrity 37 (2022) 563–569 Karina Buka-Vaivade et al. / Structural Integrity Procedia 00 (2019) 000 – 000

566

4

Fig. 3. The adopted schemes for numerical modelling of the good quality glued areas in timber-concrete connection (a) defect areas consist of many identical areas equal to 10% of all connection surface area; (b) defect area consist of one non-glued area.

The study of the effect of a low-quality glued connection on the load-bearing capacity of TCC element subjected to the flexure consist of two parts. In the first part of the study the maximum shear stresses is determined for TCC models with good-quality timber-concrete connection equal to the 90%, 75%, 50% and 25% of the total connection area, in such way, that the length of poor-quality glued area is 10% of the total sample length. In the second part of the study, calculations have been made to determine the maximum shear stresses at 20%, 30% and 40% defect areas from the total timber-concrete connection area, in such a way that a poorly bonded area forms one large area with a length from 20 to 40% of the total length of the sample. The adopted schemes of the qualitative glued areas in connection between timber and concrete layers are shown on Fig.3. 3. Results and discussions The comparison of the behaviours of the TCC specimens produced by classical dry method and proposed method with granite chips obtained by three-point bending test, numerical and analytical calculations is shown on Fig. 4 (a). The dependence of TCC with granite chips is shown as average values of two specimens. As it can be seen, the TCC without granite chips develops much larger displacements than those obtained in the calculations, at the load level of 30 kN the experimentally determined displacements are equal to 22.8 mm, what is more than 52% larger than those calculated with the 3D non-linear damage model. At the same time, proposed production method of rigid connection provides full composite action of the specimen and additional local strengthening. The local strengthening effect at the stage when nonlinear behaviour begins to develop in TCC can be explained by granite chips as concrete aggregate better strength than strength of the cement paste. The specimens produced by proposed method with granite chips showed both higher load-bearing capacity, as well as smaller maximal vertical displacements. The average destructive load of this specimens was 41 kN, then for TCC produced by classical dry method destructive load was only 32 kN. Maximal displacements at the load level of 30 kN for TCC with granite chips is smaller than 6 mm or one three hundredth of specimen span. As it can be seen on Fig. 4 (a), experimentally obtained behaviour of TCC panel produced by classical dry method is characterized by less stiffness in comparison with analytically and numerically obtained. The reason of this is poor quality of the obtained glued connection by classical dry production method. On Fig. 4 (b) is shown timber-concrete composite panel produced by classical dry method after collapse of concrete layer. It can be clearly seen, that the qualitive connection, where connection was working well, was formed less than on 30% of total connection surface.