Keywords: timber structures, timber connections, composite structures, composite timber concrete structures, finite element method, computational analysis.

The effectiveness of the composite behaviour on timber-concrete structures is highly dependent on the efficiency of the joints. However, so far, their mechanical behaviour is not well known. This paper presents 3D non-linear FEM models that have been developed to predict the mechanical behaviour of timber-concrete joints made with dowel-type-fasteners. Two different models are presented, the first one intended to simulate embedding tests on timber, and the second one to simulate the shear tests on the joints.

The first model was used to estimate the necessary input data for the joint model based on embedding experimental results, while the second one simulates the shear tests. Both models consider non linear isotropic behaviour for concrete and steel and non-linear orthotropic behaviour for timber. The models also take into account the different types of friction that occurs during the test, such as: friction between dowel and timber, and friction between timber and concrete. These phenomena are simulated using contact bodies.

The models presented here have two different types of boundary conditions: internal, which are constituted by the interactions between the different materials; and external, which are constituted by the load conditions, the support conditions and the symmetry conditions. The boundary conditions are usually rather simple to identify but very difficult to simulate correctly and their correctly simulation might increase significantly the model complexity and thus increase significantly the risk of numerical instability and the time required for the calculations.

Two types of external boundary conditions were considered: that due to the supports and another due to symmetry considerations. The supports constrains were used to simulate the support conditions and the load application in the experimental tests. In the model the supports were simulated considering prescribed values of zero displacements for the three translations in all the nodes that correspond to support areas in the experimental tests (the nodes in the timber elements base). Loads were introduced by increasing controlled displacements on the top of the concrete element, as done in the experimental tests.

The models used for the current study were solid 3D models. The bricks of the model were made with eight node elements from MSC Marc software package, element 7 MARC (2003) [2].

In order to evaluate the accuracy of the model, it was decided to compare them with data from experimental shear tests performed by Lopes, Cruz and Dias (2003) [1]. The fastener used was a dowel type fastener made of smooth steel bars 10mm diameter. The concrete was intended to be classified as C30/35 (strength class according to EC2 [4]). 150mm cubes gave a mean compression strength of 40 MPa. The timber was glue-laminated spruce produced with timber from the strength class C18. The values for the timber properties were estimated from EN338 (1995) [3], considering timber from strength class C18.

1

Lopes, S., Cruz, H., Dias, A. "Trial tests on timber-concrete connections", 2nd ISBPDR - CIB - Lisbon, 2003.

2

MARC "Analysis Research Corporation, Manual", version 2003 USA, 2003.

3

CEN, "EN 338 - Structural Timber - Strength Classes", CEN 1995, 1995.

4

CEN, "prEN 1992-1-1: Eurocode 2: Design of Concrete Structures - Part 1: General Rules and Rules for Buildings", European Committee for Standardisation, Brussels, 2002

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £135 +P&P)