Resolution of Magnetic Intelligent Tool in Case of Stress-corrosion Cracks Detection

The basic problem of main pipelines non-destructive examination is reliability and validity of obtained parameters of flaws (including stress-corrosion cracks). One of the most rapid and efficient NDT techniques for main pipelines is the method of magnetic flux leak (MFL) over defect. In accordance with magnetic method canon the examination reliability (that is probability of crack detection starting from some preset threshold level) is a multiparameter task. The MFL technique is widely used for detection of internal and external stress-corrosion. The instruments implementing MFL method provide magnetization of pipeline wall material up to the technical saturation level; after that the magnetic filed flux leak is registered by magnetic field sensors. Stress-corrosion and corrosion like defects present in the pipeline wall are detected based on information picked up from those sensors. The difficulty in this case is the fact that basic measured parameter is subjected to impacts of numerous factors including material sensitivity towards magnetization, types of sensors and there location, pipeline wall thickness, the gap between pipeline surface and measuring element, etc. The magnetic instruments, as a rule, comprise: magnetic module inducing magnetic field in the pipeline wall; sensors placed between magnets at certain distance from each other. If it is required to evaluate the depth of the stress-corrosion crack the number of sensors and distance between them should be determined. Also it is necessary to evaluate and determine the distance between neighboring stress-corrosion cracks making it possible to differentiate them. In the paper posed is the problem and discussed are approaches of its solution, i.e. how to define the sufficient number of sensors to detect peak signal amplitude that can be correlated to the depth of stress-corrosion crack and how to determine minimal distances between cracks for which the last ones can be differentiated.