In one of the buildings analyzed by NCREP the possibility of performing destructive and medium-destructive tests on the floors came up because, due to architectonical incompatibilities, the general project dictated its demolition. The adoption of destructive tests is naturally to avoid, because it doesn't allow the tested elements to be maintained. However, this opportunity of testing a set of floors that would be demolished ended up being extremely important, since it allowed the evaluation of parameters and characteristics that will surely be helpful in similar structural rehabilitations.
The surveyed building, built in 1916, has an area of 12,0 x 30,0 m2 and presents floors with 0,20 m diameter beams, spaced 0,60 m and with 6,0 m span. Above the beams there is a 2,5 cm thick floorboard. Performing the transversal link, there are 0,10 m diameter bars, spaced 2,0 m.
Besides a campaign of non destructive tests carried out with the Pilodyn and the seismographs, among other instruments, a load test was then performed together with a very particular test for the evaluation of the beam-wall friction, using a hydraulic actuator.
The campaign of tests, in particular the load test and the frequencies evaluation, had the main goal of assessing the global floor and the single beams behaviour when submitted to a monotonic loading history. The setup was also prepared to allow evaluating the effectiveness of the transversal bar system and the floorboard contribution to the stiffness and resistance of the floor.
5.2 Load test
The strength capacity and the stiffness of the floors are strongly affected by factors such as the moisture content and the malfunctioning or degradation of its structural elements, beams and transversal bars, and of the floorboard. The load test was carried with the main purpose of evaluating the structural behaviour of the floor and the contribution of its single elements to its global resistance and stiffness, Figure 13. In order to better understand and to be more conclusive about the floors behaviour, the test was done on a 3,5 m wide strip (corresponding to an assembly of 7 beams) that was isolated from the rest of the floor. It was chosen a well preserved floor area (beams, transversal bars and floorboard).
With the purpose of estimating the efficacy of the transversal bars and the floorboard on the distribution of the loads, only the central beam was loaded. Three reservoirs with a capacity of 600 kg each, making up a total load of 1,80 tons were used. A flow measurement instrument was installed at the entrance of the reservoirs to control and register the volume of water i.e., the load installed on the floor. During the loading process, the vertical displacements of all the 7 floor beams were monitored using LVDT's. The data, load and displacements, was acquired continuously and analysed using an on-line homemade software.
Having in mind the objectives referred previously, the tests were done under three conditions: (a) with the original floor, i.e. with transversal bars and floorboard; (b) without the transversal bars i.e., only with the beams and the floorboard; (c) on an isolated beam. Therefore, the load test, which is normally non destructive, ended up to have a destructive character. In order to proceed to the sequence (a) to (c), the test remained within the elastic behaviour range. However, with the intention of complementing the obtained information, a load test will be done at the Laboratory of Earthquake and Structural Engineering (LESE) to evaluate the ultimate behaviour of the isolated beams.
The analysis of the results, in particular of the load-displacement diagrams, and for the applied load range, shows a linear behaviour, Figure 14. From the results obtained it was possible to estimate the wood modulus of elasticity (around 12 GPa), verify the increase of
Figure 14. Load-displacement diagram for two isolated beams loaded simultaneously (the load refers to the total load applied).
the floor stiffness conferred by the floorboard and the poor effectiveness of the transverse bars to distribute the applied loads when compared to the floorboard. Other conclusions and analyses from the test will be published in the future.
The data from the seismographs were particularly helpful in the evaluation of the relative importance of floor elements. In particular, several readings were made. Firstly, the floor with the original width was analyzed (30,0 m); Secondly, the measurements were repeated after the cut that isolated the 3,5 m strip from the rest of the floor, (final area of 6,0 x 3,5 m2). Afterwards, measurements were done without the transversal bars and, finally, on an isolated beam, Figure 15. The comparison between the results of these four situations, with the main natural frequency changing from 7,0 Hz to 10,0 Hz, allowed evaluating the importance of the floorboard, transversal bars and beams, to the floor structural behaviour and, simultaneously, estimating the medium values of the wood modulus of elasticity. It was also verified that the values of the natural frequencies measured in-situ were within the range of theoretical values calculated according to EC5 (1998).
Apart from these conclusions, the measurements using the seismographs can be used to calibrate numerical models to simulate such type of timber floors on old buildings. When needed, the simulation of the floors can be done considering the beams and the transversal bars as they exist on the floor, and adopting
auxiliary bars, with small inertia, to simulate the floorboard, linking the beams. Another possible model, that tends to approach the floorboard real effect over the whole structure, consists on considering low thickness slab elements to simulate the floorboard. These elements rest on the beams and transversal bars, but aren't connected to the resistant walls, as it is usually observed in buildings (Neves, 2004).
To evaluate the resistance of the friction connections between the timber beams and the masonry walls, a destructive test was carried out. A segment of beam was cut of to install a hydraulic actuator that would act along the beam axis. In order to maintain the two segments of the tested beam in the original position, it was conceived a metallic structure with pulleys to support it (not allowing vertical movements but free horizontal displacements), Figure 16. To set the actuator on the beam, metallic corner cupboards fastened by M12 bolts were used. The test consisted on applying pull out and in movements of the beam in relation to the wall. Displacement transducers were installed at the extremities of the beam to evaluate the load-displacement diagram of the beam-wall connection. With the purpose of verifying any possible horizontal movement of the wall during the test, transducers were installed on the walls. The results permitted to conclude about the limited efficacy of the connection between the two elements, traduced by a friction force of about 2,0 kN.
Was this article helpful?