Cracks and flaws are typical failure modes in mechanical structures such as oil and gas pipeline networks, buildings, aircraft fuselages, and spacecraft. Early detection of these failures is critical for safety, timely replacement of damaged or corroded parts, accident prevention, and resource and cost savings. This requires low-cost, easy-to-implement, non-destructive, and reliable testing. For this purpose, we developed an embedded system in a Zumo 32U4 mobile robot for the magnetic detection of cracks in metal plates. Two AVR microcontrollers were used. One controls the magnetic sensing, signal processing, and data acquisition, and the other controls the robot’s displacement. A graphical user interface (GUI) was also implemented. Low-cost commercial magnetoresistive sensors (Sparkfun MAG3110) were used to monitor the variations of the 3D magnetic field around the cracks. Our detection method is based on the magnetic memory method, where local magnetic distortions are expected around a crack in a metallic material. Such distortions are due to the breaking of the symmetry of the crystal in the region where cracks appear. Experimental tests were carried out on a steel plate with machine-induced cracks of depth and width variations. The magnetic field changes associated with these cracks were analyzed and compared. This crack detection method could be used for structural health monitoring of mechanical infrastructure.
Read full abstract