PSI - Issue 64

Aeneas Paul et al. / Procedia Structural Integrity 64 (2024) 1287–1294 Aeneas Paul et al. / Structural Integrity Procedia 00 (2019) 000 – 000 7 and the corresponding slip are determined: s = 2.31 mm and ̅ b,max = 14.44 N/mm² for NSC without prestress, s = 1.69 mm and ̅ b,max = 17.24 N/mm² for NSC with prestress and s = 1.30 mm and ̅ b,max = 24.87 N/mm² for HPC without prestress. The comparison of the results shows that additional prestress increases the maximum bond stress by 19 %, while the slip decreases by 27 % (see Fig. 5b). It should be noted that P 0 is not fully transferred to the concrete, as the bond length of 5 cm is below the common transfer lengths l pt (Hegger et al. (2007)), (Sanio and Mark (2020)) for prestressed concrete. The use of an HPC instead of an NSC results in an increased bond stress by 72 % while the slip decreases by 43 %, which indicates a stiffer bond compared to the NSC. 1293

Fig. 5. Results of pull-out tests: a) NSC without prestress b) NSC with prestress c) HPC without prestress in form of slip-bond stress and slip force diagrams.

4.4. Comparison of measurements and simulations A significant difference in the magnitude as well as in the distribution of the strains is observed between the concretes. The shorter length of impact can be explained by the better bond conditions, as already indicated by the pull-out tests ( ̅ b,max : 72 % higher in HPC) and are supported by the measurements on the tendons ( l pt : 44 % shorter in HPC). The transfer lengths l pt obtained from the concrete measurements (NSC: ≈50 cm, HPC: ≈35 cm) exceed the predicted ones fitting the tendon measurements (NSC: 39 cm, HPC: 22 cm) and yield a more distributed strain field. The overall smaller surface strains detected on HPC can be explained by larger transversal stresses. According to the stress-strain relation for isotropic, linear-elastic materials, an increase of σ zz (and analogously σ yy ) comes along with a decrease of ε xx , due to the Poisson ratio ν (Bower (2009)). The FE simulations support that even not considering the Hoyer effect, larger transversal stresses develop when a load of the same size is applied over a shorter length, which also reduces ε xx . 5. Conclusions Two prestressed beams were experimentally investigated to detect tendon breaks by DFOS measurements on the concrete surface. Breaks at a depth of 0.15 m could be clearly localized and identified from altering concrete strains. The bond behavior plays a key role in the re-anchoring process and the measurement results. While the emerging strain fields are qualitatively similar for NSC and HPC, the strain decreases and the lengths of impact are individual. The better the bond is the smaller the length and the smaller the change of strains. This highlights the potential of DFOS for permanent monitoring of prestressed structures. However, it must be noted that the recorded strains are fairly small and therefore sensitive to external or white noise. The method needs further elaboration to include and distinguish other influential factors like the maximum depth, at which a DFOS-based monitoring system is able to reliably detect tendon breaks.