by the axial restraint makes torsion tests particularly suitable for validation of the 3D lattice model and other numerical models in general. In this paper an overview is given of the results of torsion tests on concrete. They form the basis for validation of the 3D lattice model. EXPERIMENTAL DETAILS Tests were conducted on concrete cylinders with diameter D=34 mm and height H=68 mm. Both un-notched and notched cylinders were used. The notch, at half height of the cylinder, has a depth d 0 =8mm. The specimens were drilled from concrete blocks, casted in plywood moulds. The aggregates used for the mixture have maximum size d a =2 mm. Portland CEM I 32.5R cement was used, with water/cement ratio 0.4. During pouring of the mixture and for other 30 seconds after complete filling, the moulds were vibrated on a small vibrating table. Next, they were covered with plastic sheets. After two days, the concrete blocks were demoulded and placed under water. The cylinders, drilled from the concrete blocks after 28 days, were kept in a climate box, at 200 temperature and 90% relative humidity, till the moment of testing. The tests were conducted with a servo-hydraulic INSTRON 8874. This machine has rotational and vertical loading axes, which can be coupled or de-coupled. The maximum loads which can be reached are 100 Nm and 10 kN, respectively. The feed-back signal during the tests is the rotation, measured with a rotational potentiometer connected to a wire attached to the upper loading platen, as shown in Figure 1. Different tests with different axial restraint were conducted. The cases considered are: zero axial force, zero axial displacement and constant axial compressive force, corresponding to a nominal stress of -4 MPa. An hour before testing the specimen was glued to the loading platens. After hardening of the glue, the axial load was applied first, whereafter the rotational load was applied at the constant speed of 0.002 deg/sec. During the tests the axial displacement (stroke), the axial load, the rotation and the torque were recorded. Figure 1: Set-up for measurements of the rotational angle EXPERIMENTAL RESULTS For each loading condition, three tests were conducted. The average of the peak torque and of the axial load, with the corresponding standard deviations, are reported in Table 1. There, the tests are named after the applied loading condition: d2f, d2dandd2c refer to tests on cylinders without notch under zero axial force, zero axial displacement and constant axial compression, respectively. In the same way, d2nf, d2nd and d2nc refer to tests on notched cylinders. In Figure 2 and Figure 3, some of the obtained experimental results are reported, together with the crack patterns. The torque versus rotation curves are represented with a continuous line, while the stroke (LVDT in the hydraulic actuator) or the axial load (depending if the experiment was conducted under constant load or fixed displacement) are represented with a dashed line. In the insets in the diagrams, the crack patterns are those photographed after the end of the test, when the upper loading platen was driven far from the bottom platen. The other crack patterns shown in these figures were photographed from different orientations
RkJQdWJsaXNoZXIy MjM0NDE=