PSI - Issue 72

Andrey Yu. Fedorov et al. / Procedia Structural Integrity 72 (2025) 453–457

456

For experimental testing of the considered defect detection algorithm, experiments were performed on a prismatic cement sample with dimensions of 400 × 70 × 70 mm. To create a cement sample, 400 grade Portland cement and river sand were used in a ratio of 1 : 5. During the sample manufacturing process, an optical fiber was embedded at a distance of 20 mm from the lower surface with 5 Bragg grating sensors, each 5 mm long and located at a distance of 25 mm from each other. The strain was recorded when the sample was loaded according to the three-point bending scheme shown in Fig. 3a. In the experiments performed, a transverse cut served as the defect. The cut was 2 mm wide and ranged from 5 mm to 15 mm in depth. A photo of the sample with the defect is shown in Fig. 3b. Strain measurements at different loads were performed in a sample without a cut and in samples with cuts of 5, 9, 12 and 15 mm in depth.

a)

b)

Fig. 3. Three-point bending loading scheme of cement sample (a), photograph of a sample with a defect (b).

The obtained experimental results were compared with the results of numerical modeling by the finite element method. The mechanical characteristics of the material (E = 9.2 GPa, ν = 0.18) were selected based on one of the loading variants of the sample without a cut. Fig. 4a shows the results of strain measurement by FOS and the strain distribution along the optical fiber obtained by the finite element method in a sample without a cut under loadings of 100, 150, 210, 310, 350 and 400 N. Fig. 4b shows similar results in samples without cuts and with cuts at a load of 400 N. Table 1 shows the relative values of strains at different loads and defect depths h.

Fig. 4. Strain distribution along the length of the sample without defect under loads 100 , 150 , 200 , 310 , 350 , 400 N (a); at a load of 400 N in samples with defects (cuts) of di ff erent depths (b).