PSI - Issue 11

Giorgio Frunzio et al. / Procedia Structural Integrity 11 (2018) 153–160 Prof. Ing. Giorgio Frunzio, Ing. Luciana Di Gennaro 00 (2018) 000–000

159

7

4. Structural Retrofit

The case of a particular deck composed of four chestnut beams (principal frame) and one cross-beam placed on the intrados of the decking, with the materials characteristics described in the previous paragraphs has shown below. The calculation analyzed the behaviour of the wooden chestnut decking before and after the structural restoration. In the pre-restoration model the secondary frame is only thought to be made up of the poorly connected panconcelle using nails posed without pre-drilling hole (in accordance with the findings). In the post-restoration model, the secondary frame is thought to be made up of double crossed planks in solid chestnut wood (restoration using the wood-wood technique). In order to better describe the pre-intervention behaviour, the connection between the elements has been considered both with none stiffness and finite stiffness. A calculation model based on finite element modelling software has been adopted. Two models have been developed for each condition to assess (in accordance with the provisions of Eurocode 5) the stresses with the combination of the ultimate limit state loads (SLU), and the deformations with the combination of loads at the operating limit state (SLE). At first was analysed the pre-restoration condition considering two different hypothesis, a structure model with none stiffness connection and another structure model with finite stiffness connection. The first hypothesis considers the total absence of connection between the main and secondary frame. In this configuration the two elements constituting the composed section slide perfectly each other and therefore the bending stiffness of the section is given by the formula: ( ) 0 = ∑ = 1 1 2 ∙ ( 1 ∙ 1 ∙ ℎ 13 + 2 ∙ 2 ∙ ℎ 23 ) (20) Note that in this condition there is no dependence on the K parameter so the value of EJ 0 is the same for SLU and SLE. Moreover, the distribution of the moment will be exclusive function of the bending stiffnesses of the two sections. Following are the diagrams relating to this condition: The second hypothesis considers the elements poorly connected by nails posed without pre-drilling hole. The content of equation (1) is therefore valid except for the K ser value of equation (10) which is replaced by the following: = 1 , 5 ∙ 0 , 8 30 (21) The post-restoration condition considers the structure after restoration with the wood-wood technique with metal cylindrical shank pins connectors, thus all the equations described in paragraph 3 are valid. A load test was made on the wooden chestnut decking of the present study in order to test the structure in the post-restoration condition. The test check the structure both in terms of resistance and elastic response, subjecting it to the maximum stresses in accord with the working loads. The test, as a consequence, provided both a valid reference on the behaviour of the main supporting elements (beams) and on their insertion in the vertical masonry structures (supports). A vertical load was placed on the floor along the testing beam direction . Five rod comparators located along the barycentric axis of the wooden beam, allowed the movements reading. The maximum test stress was reached through increments of about 1,00 kN/sqm. During the loading phase the structure was left under constant load for a sufficient period of time to allow a comple deformation. The load removal was also gradual and steps were taken to read the movements even after the complete discharge of the structure to ascertain the possible presence of permanent deformations. At the end of the test and net of the supports lowering, the maximum deformation found in the barycentric axis was 0,96 mm, significantly below the values obtained with the calculation. This discrepancy is attributable to the screed contribution that was not considered in the calculation for the benefit of safety. 5. Load Test