PSI - Issue 37

Alexey Tatarinov et al. / Procedia Structural Integrity 37 (2022) 453–461 Alexey Tatarinov et al./ Structural Integrity Procedia 00 (2019) 000 – 000

455

3

Table 1. Properties of conditionally “weak” W and “solid” S layers of concrete .

Grades of concrete quality in layers

S

W1 1:4

W2 1:7

W3

Cement-sand ratio

1:3

1:12 2200

Ultrasound velocity, m/s Predicted strength, MPa

4400

3900

3100

36

25

14

8

The specimens for building the grid for the mathematical model were 2-layer rectangular prisms 160x40x40 mm with the upper surface layer of weak concrete W and bottom solid layer S (Fig.2). The thickness of W layer Th W had 7 gradations: 0, 5, 12, 20, 30, 35 and 40 mm. At 0 and 40 mm the specimens were entirely solid concrete S and weak concrete W, correspondingly. Thus, the total number of specimens for the model was 21 (7 grades of Th W and 3 grades of W quality W1-W3). In addition, control specimens with intermediate values of Th W were made for model testing.

Fig.2. Design of 2-layer concrete specimens with conditionally “weak” surface layer W and “solid” substrate S with gradual variation of thickness of W layer Th W and experimental line of specimens with different Th W . 2.2. Ultrasonic data acquisition Ultrasonic propagation signals were acquired by a pair of ultrasonic transducers in the surface transmission mode by the surface profiling of the specimens from the side of layer W. One of the transducers (emitter) was placed immovably at one end of the specimen, while another transducer (receiver) travelled along the specimen’s surface a distance from 20 mm to 120 mm with a step of 5 mm. Ultrasonic signals were recorded at each step of scanning forming a 2D spatiotemporal waveform profile composed of 21 signal or a digital data matrix, where the coordinates are ultrasonic time and the distance between the transducers and the values are the signal amplitudes. 2D spatiotemporal waveform profiles were obtained separately at two ultrasonic frequencies, 50 kHz and 100 kHz to have different depths of propagation of Rayleigh surface waves proportionally to the ultrasonic wavelength. 2D profiles at 50 and 100 kHz served as the raw material for mathematical pattern recognition. To accelerate, automate and unify the acquisition of ultrasonic profiles, a mechatronic motor-driven scanner was designed and manufactured (Fig.3).

Fig.3. Ultrasonic data acquisition setup with mechatronic scanner.