Low-Cost Wireless Structural Health Monitoring of Bridges · Komarizadehasl S, Lozano F, Lozano-Galant JA et al.

4. Real Structure Test and Results

This section presents a field test carried out on a short-span bridge for comparing the measured eigenfrequencies of LARA with those of a high-precision sensor.

In the first place, the footbridge is shown in Figure 10a. As shown, it is connected to an elevator box and from the other side, it is located on an abutment. Figure 10b,c show the plan and section of the bridge, respectively.

After mounting the sensors to the bridge, they were connected to the rest of the monitoring components (power bank, USB 4G dongle, Arduino, and Raspberry Pi) manually. It is important to highlight that to enable its communication throughout the Internet, the Raspberry Pi was previously configured with a WiFi hotspot. In fact, after its assemblage on site the Raspberry Pi was used to initiate the data acquisition process remotely, using the Virtual Network Computing (VNC) software [73]. The acquired data was first collected on the memory card of the Raspberry Pi. Then, when the data acquisition process finished, the obtained data were moved to a computer using the VNC software. Figure 11a presents the diagram of LARA setup on the bridge. The location of the commercial dynamic sensor (HI-INC) and LARA on the mid-span of the bridge is shown in Figure 11b.

It is also important to mention that the cables shown in Figure 11b are not for connecting LARA or HI-INC to a laptop. Both systems were controlled remotely. These cables are either for power-source connection or for connecting the Arduino and the Raspberry Pi.

The outputs of the eigenfrequency analysis of LARA are shown in Figure 12a–c for x, y, and z axis, respectively.

Analysis of Figure 12 shows that the bridge is not significantly excited about its x-axis (longitudinal direction). It is appeared that the bridge is excited very well about y and z axes.

Table 4 presents results of the eigenfrequency analysis of LARA and the used high-precision inclinometer dynamic (HI-INC). This table also presented the difference percentage of the measured frequencies with LARA from those of the HI-INC sensor.

Comparing the measured eigenfrequencies of LARA with those of HI-INC shows a maximum error of 1.28%.