Section: New Results

Sensor and hardware based research

Cracks detection in pavement by a distributed fiber optic sensing technology

Participant : Xavier Chapeleau.

This paper presents the feasibility of damage detection in asphalt pavements by embedded fiber optics as a new non-destructive inspection technique. The distributed fiber optic sensing technology based on the Rayleigh scattering was used in this study. The main advantage of this technique is that it allows to measure strains over a long length of fiber optic with a high spatial resolution, less than 1 cm. By comparing strain profiles measured at different times, an attempt was made to link strain changes with the appearance of damage (cracking) in the pavement. This non-destructive method was evaluated on accelerated pavement testing facility, in a bituminous pavement. In our experimentation, the optical fibers were placed near the bottom of the asphalt layer. The application of 728 000 heavy vehicle loads (65 kN dual wheel loads) was simulated in the experiment. Optical fiber measurements were made at regular intervals and surface cracking of the pavement was surveyed. After some traffic, a significant increase of strains was detected by the optical fibers at different points in the pavement structure, before any damage was visible. Later, cracking developed in the zones where the strain profiles were modified, thus indicating a clear relationship between the increased strains and crack initiation. These first tests demonstrate that distributed fiber optic sensors based on Rayleigh scattering can be used to detect crack initiation and propagation in pavements, by monitoring strain profiles in the bituminous layers [17].

Wireless sensors and GPS synchronization

Participants : Vincent Le Cam, David Pallier.

Most of recent development in WSN domain focused on energy (saving or harvesting), on wireless protocols, on embedded algorithms. But it is a fact that, most of monitoring applications need samples to be time-stamped. According to the application, the wished time resolution could be up to one second for automation monitoring, one millisecond for vibration, one microsecond for acoustic monitoring, one nanosecond for electricity or light propagation... The consequence for a Wireless network of electronic nodes is that, by nature, no common signal could physically provide a synchronization top. But, as each electronic device, a wireless sensor time-base uses a timer incremented by a quartz whose initial value is theoretical up to some p.p.m. and whose period drift on time because of age, temperature,... Two kind of solutions could be regarded : a synchronization signal provided by the wireless protocol itself; an absolute synchronization from a referential source such as: GPS, Frankfurt clock, Galileo,... In the first way, it will be demonstrated the poor accuracy and the need of energy such a mechanism offers. In the second way, the article will details how a deterministic (Universal Time), accurate and resilient algorithm has been implemented. The article also provides specific results of application on acoustic monitoring system and electricity propagation where the accuracy of a WSN has reached up to 10 nanosecond UT. Consequence on energy consumption of this algorithm are given with a description of future works to improve the energy balance while keeping the device sober and synchronized [33].