GRC connection tests

This page explain the procedure performed the evaluate the mechanical performance up to failure of a bespoke connection between a steel bracket and a Glass Reinforced Concrete (GRC) panel.

The mechanical resistance of the connection between a Glass Reinforced Concrete (GRC) panel and the steel bracket has been studied. In particular, to evaluate the performance of the connection, a series of 9 specimens, of the same geometry and composition, have been tested in the Engineering Laboratory at the University of Cambridge (UK). Both the axial (pull) strength and the transversal (shear) strength with respect to the connector has been investigated.

Figure 1: 3D of the connection 

Instrumentation

During all the tests the vertical displacement at the mid-span of the external top surface of the beam has been recorded using a displacement transducer, see Figure 2; while a load cell was measuring the applied load. The load was applied using a manually controlled hydraulic jack (Figure 3).

Figure 2: Detail of the displacement Transducer

Data Acquisition

The test control and the data acquisition system were linked together through a National Instrument data acquisition systems, NI PXIe-1071, and LabView software. This software is capable of synchronised data acquisition among different channels, in this case the displacements and the force signals.

Figure 3: Test Equipment

Test Setup

The specimens were tested in the Structural Research Laboratory of CUED, University of Cambridge.  For all the samples the GRC beam is clamped at the two ends, using two steel beams, and subjected to a load force applied directly to the pin connector. Figure 4(left) show the application of the imposed load through a connector for the pull test and using a plate attached to the beam for the shear tests Figure 4 (right).

Figure 4: Pull Test (left) and shear Test (right) load application

Three different set-up have been considered:

1. Setup 1: a pull test of the beam with span of 270mm (4 samples)
2. Setup 2: a shear test of the beam with span of 270mm (3 samples)
3. Setup 3: a pull-out test of the beam with span of 170mm (3 samples)

The first two configurations aim to provide the strength of the connection when both the pull-out effect and the bending effect are acting on the beam, to estimate the behaviour of the connection in a configuration similar to the real installation (according with the available samples); while the third configuration aims to provide a pure pull-out strength of the connector. In order to do that, the bending effect was reduced by considering a smaller span of the beam; in particular, setup 3 is then designed to allow a full GRC cone extraction and with a chosen span of 170mm. For the shear test, differently than the other two setups, the beam was rotated of 90 degrees around its longitudinal axis, in so the load was applied as shear force for the bolt. 

Results and Analysis:

Prior to testing, the behaviour and stress distribution of the connection under axial (pull test) and transversal (shear test) force has been analysed using Finite Element (FE) modelling. Figure 5 shows the results of the analysis and in particular the areas where the stress concentrates and the directions in which the crack could propagate in the GRC.

Figure 5: FE model of the tested connection

Pull test with 270mm span

Figure 6 shows the displacement and the corresponding recorded load for the Pull test of three samples with testing span of 270mm. 

Figure 6: Applied load vs. Displacement for the Pull test with span of the sample beam of 270mm

The behaviour of all the beams can be summarized into two different stages. First, the linear elastic behaviour of the connections, up to about 30-40kN, and then the elastoplastic behaviour with reduced stiffness up the failure. The recoded differences in terms of displacements are explained due to the different cone failure mechanisms which could not be recorded by the transducer was located, as they appears in different locations.

The figure also include the maximum design load expected for the connection. It can be clearly see that in all the tests the load capacity of the connection due to a pulling force is at least three times bigger than the design load (F_sd = 10.4kN).

If compared with the theoretical strength, in all tests the experimental maximum values are lower than the expected failure and this can be explained due to the combination of local pull-out effect and the bending effect. In particular, in all the tests an initial transversal crack at the middle span of the n the tensile side was observed. This is a usual failure for beams under bending with tensile strength of these materials is typically lower than the compressive strength.

In the pull test, further to the crack due to bending failure, a cone failure was identified on the external surface at the end of the test. In particular it was possible to measure the diameter of the cone being about 160mm (see Figure 7).

Figure 7: Cone failure for the Pull test

Shear test with 270mm span

Figure 8 shows the displacement and the corresponding recorded load for the Shear test of three samples with testing span of 270mm (Test setup 2).  

Figure 8: Applied load vs. Displacement for the Shear test with span of the sample beam of 270mm

As for the previous tests the initial elastic behaviour is followed by aplastic deformation. In this case the minimum recorded value of transition between the two is around 38kN (Shear_270_0). This value is two times the design load expected for the connection (red line in the figure). The negative displacements of the dashed line (Shear_270_1) can be explained only due to an inaccuracy in the positioning of the transducer, as consequence the displacements for this case cannot be taken into account while the value for the strength are acceptable.

 Figure 9 show several steps of the Shear test. In particular after the experiment it was possible to identify a crack starting from the top of the inner plate and propagating with an angle of about 36°deg (see Figure 10).

Figure 9: Different step of the Shear test

Figure 10: Cone failure for the Shear test

Pull test with 170mm span

Figure 11 shows the displacement and the corresponding recorded load for the Shear test of three samples with testing span of 170mm (Test setup 3). 

Figure 11: Applied load vs. Displacement for the Pull test with span of the sample beam of 170mm

The result of the tests for both the axial (pull) strength and the transversal (shear) strength, as well as the expected design loads are summarised and compared in Table 3 showing a minimum safety factor of 2.05.

Table 3: Summary table