Mechanical damage, caused by externally applied forces, can deform the cylindrical shape of a pipe, scrape away metal, change material properties, and leave residual stresses and plastic strains. Mechanical damage is especially detrimental when it leaves the potential for delayed failure. In-line inspection tools both must detect mechanical damage and characterize parameters such as microstructure changes, residual stresses and extent of removed metal. These are the parameters that determine whether a defect is a candidate for delayed failure.
Magnetic flux leakage (MFL) is commonly used for nondestructive detection and sizing of corrosion and metal-loss defects in pipelines, and it is sometimes cited as useful for detecting mechanical damage. Flux leakage from mechanical damage results from geometric and magnetic changes. The geometric part of the flux leakage signal is caused by denting, metal-loss, and wall thinning. The magnetic part of the signal is caused by cold work, plastic deformation and residual stress.
At high magnetization levels, MFL signals are due mostly to defect geometry, such as metal-loss length, depth, and shape. At lower magnetization levels, the signals are caused by both geometric and magnetic deformation. To determine the signal due to magnetic deformation only, the geometric MFL signal obtained at high magnetization levels can be scaled and subtracted from the mixed MFL signal obtained at a low magnetization level. In this way, the MFL signal can be decoupled into its geometric and magnetic components.
Using the decoupled magnetic signal, information on cold work, plastic deformation, and residual stresses can be gleaned. This paper shows that scaling of high magnetization level signals is possible and differences between signals at multiple magnetization levels provide unique information about the nature of mechanical damage defects. Representative results from hundreds of mechanical damage defect examined are discussed. Included in the analysis are the effects of pipe material and inspection variables, such as pressure and inspection speed. Prospects for tool design and application for in-line inspection are also presented.
This work was supported by Department of Transportation, Office of Pipeline Safety.