Pearson, Neil; Eddyfi Technologies; United Kingdom
Boat, M.A.; Eddyfi; United Kingdom
Mason, J.S.D.; Eddyfi; United Kingdom
Pearson, N.R.; Eddyfi; United Kingdom
Session: Modeling and data processing Electromagnetic Techniques 3
Time: 14:30 - 14:50
Magnetic ﬂux leakage (MFL) continues to be a widely used approach to detect defects caused by corrosion in a wide range of applications encompassing steel products and components. The principal merit of MFL is that large areas can be covered relatively quickly making it beneﬁcial for example in the inspection of asset components that are costly to expose and large in area; a good example is the ﬂoor of above ground storage tanks. However, the MFL signal is usually reported to have a complex relationship with the geometry of defects. In some cases, this relationship might result in a many-to-one mapping, meaning that a given MFL signal might arise from more than one defect geometry.
This potential ambiguity is addressed in  through a novel frequency response (FR) approach and was used to find the bandwidth of MFL under a given controlled condition. The investigation showed that any defect shape with components outside the lower and upper spatial frequencies of MFL cannot be accurately represented leading to the conclusion that MFL is band limited and cannot capture spatial geometries outside of its band limits.
The influence of sensor height was described in , initiating the investigation of other parameters that can influence the bandwidth of MFL. This paper extends to include a range of defect depths whereas until now, investigations have been confined to defects with a depth of 50%. Results show circumstances when MFL signals from different geometries can have the same amplitudes resulting in ambiguity.
 N. R. Pearson, M. A. Boat and J. S. D. Mason. Bandwidth of MFL in steel plate inspection. 19th World Conference on Non-Destructive Testing, June 2016.
 N. R. Pearson, M. A. Boat and J. S. D. Mason. Factors that influence the Bandwidth of MFL. 56th BINDT 2016, September 2017.