PSI - Issue 47
Ezio Cadoni et al. / Procedia Structural Integrity 47 (2023) 331–336
334
4
Author name / Structural Integrity Procedia 00 (2023) 000–000
a) b) Fig. 3. Stress versus time curves obtained for di ff erent initial pre-stress: a) UHPC; b) UHPFRC.
a) c) Fig. 4. Stress versus time curves obtained for di ff erent sample length of UHPC and UHP(FR)C: a) 30 mm; b) 45 mm; c) 60 mm. b)
4. Numerical results
Dynamic compression tests were simulated using a material model named DAMP (Damage Propagation model) (Riganti and Cadoni (2020)). The application of the DAMP model enables the numerical reconstruction of 3D-MHB dynamic compression tests as well as shear (Cadoni et al. (2020)) and tensile experiments (Cadoni and Forni (2016)). The entire 3D-MHB apparatus was modelled by using Ls-Dyna explicit solver and hexahedral elements measuring 1 mm for specimen and 5 mm for bar. The experiments were simulated by integrating the constitutive equations of the material model with the boundary set up conditions for the UHPC and UHP(FR)C material (lengths 30, 45, and 60 mm). The damage propagation material model DAMP and its version dedicated to fibre reinforced concrete DAMP-FRC material model are implemented in Ls-Dyna by user defined subroutine. Fig. 5 depicts the damage evolution of UHP(FR)C (a, b, c, d) and UHPC (e, f, g, h) specimens (indicating 1 for fully damaged and 0 for intact material) showing the suitability of the DAMP model in reproducing 3D-MHB tests. The DAMP material model can accurately simulate the structural response of materials with unique mechanical properties under a range of stress states, from static to dynamic loading conditions, while taking into account the specimen size. Simultaneous reconstruction has proven to be di ffi cult with classical concrete material models (Riganti and Cadoni (2020)). Fig. 6 depicts the damage evolution of UHPC sample 30 mm long and pictures of the failure process. Fig. 7 compares the stress versus time curves obtained experimentally and numerically for both 30, 45 and 60 mm specimens. In some experiments, the loading wave was not monotonic due to the characteristics of the experimental apparatus. In these cases, the numerically reconstructed tests still accurately reproduce the input and output signals obtained experimentally. The output signal adapts realistically to various input loading conditions and specimen lengths. The good agreement between experimental and numerical results is due to the fact that the DAMP material model evaluates the interactions between damage propagation, loading wave, and specimen size. Another significant di ff erence from
Made with FlippingBook Annual report maker