PSI - Issue 78

Dalila Rossi et al. / Procedia Structural Integrity 78 (2026) 98–104

102

(a) (b) Fig. 3. Baseline measurements taken on sample 2. (a) Gage length of 10 cm. (b) Gage length of 14 cm.

In Fig. 4, a comprehensive overview of all the test results is provided. They illustrate the concrete stress estimated from the flat-jack pressure measurements in accordance with Equation (1), f m , versus the stress applied to the concrete slab, σ c . The results obtained with the 25 cm flat-jack show a marked variability in the test outcomes. In two out of three tests, the stress in the concrete is underestimated, only the test on the third sample (3M25) returns values consistent with expectations. In contrast, all the three tests performed with 33 cm flat-jack provide results that are in good agreement with the applied stress on the concrete slab. As a summary of the test’s consistency, Table 1 reports the relative error E r and the coefficient of variation CoV at each pressure level for each test with the 25 cm and 33 cm flat-jacks. Relative errors around 30% indicate that the measurement with the 25 cm flat-jack is not very accurate, as it deviates significantly from the reference value. Additionally, a CoV of 20% suggests that the measurements are quite variable, and the measurement is not sufficiently repeatable. In contrast, the 33 cm flat-jack shows better accuracy, with relative errors around 10%, indicating that the measurement is closer to the reference value. Moreover, the 6%-7% CoV indicates good repeatability and suggests that the measurements are consistent.

(a) (b) Fig. 4. Calculated and applied stress using the (a) 25 cm flat-jack; (b) 33 cm flat-jack.

Table 1. Relative errors E r and coefficients of variation CoV for the tests conducted. 25 cm flat-jack

33 cm flat-jack

σ C [MPa]

E r (1M25) E r (2M25) E r (3M25)

CoV E r (1M33)

E r (2M33)

E r (3M33)

CoV

2 3

-34% 13% 10% 20% -10% 1%

14% 6% 17% 6%

-34% 36% -2% 21% -2%

9%

4

-40% 36% -3% 22% 3%

2%

13% 7%