Magnetic measurements for the assessment of creep damage and detection of aberrant material in power station steels
J W Wilson, A J Peyton, D J Allen, A Shibli and Y Hasegawa
Creep cavitation damage in pressurised components operating at elevated temperatures can rapidly develop into cracking and failure. Consequently, early detection of creep damage at the micron-scale creep cavitation stage is a major goal in in-service inspection. Currently, the only proven NDT technique able to achieve this outcome is surface replication, which is slow, cumbersome and limited to surface inspection. Recent experimental trials organised by Eurpean Technology Development (ETD) Ltd on a welded pressure vessel tested at high temperature showed encouraging results with electromagnetic (EM) sensor technology developed by the University of Manchester and used at inspection outages to scan welds. The results, analysed by Impact PowerTech, showed a clear correlation with creep cavitation formed in the near-surface region of the vessel.
The follow-up project is described here. Interrupted laboratory uniaxial cross-weld creep testing of a Grade 91 weldment was carried out by NST, using a rectangular creep specimen geometry to enable surface scanning with an improved sensor system. Tests were interrupted for inspection at a range of creep life fractions from 6% to 70%. The results confirmed that both magnetic Barkhausen noise (MBN) and magnetic permeability showed consistent correlations with life fraction. A simple qualitative model of magnetic behaviour can provide a plausible explanation of these observations. Comparison with metallographic techniques indicated that EM becomes sensitive to creep damage before it reaches a size that can be detected by optical metallography. Potentially, therefore, EM could be a fast, simple and effective tool for creep damage detection.
Trials on real martensitic steel plant components have also shown that the EM sensor can identify ‘aberrant’ mismanufactured items with weak, non-conforming ferritic microstructures, thus enabling action to be taken to avoid their premature failure in service. Further work is planned to develop the sensor for full-scale component inspection, improve its capabilities on irregular surfaces, confirm its ability to quantify damage and to bring the technology into practical application on high-temperature plant.
The follow-up project is described here. Interrupted laboratory uniaxial cross-weld creep testing of a Grade 91 weldment was carried out by NST, using a rectangular creep specimen geometry to enable surface scanning with an improved sensor system. Tests were interrupted for inspection at a range of creep life fractions from 6% to 70%. The results confirmed that both magnetic Barkhausen noise (MBN) and magnetic permeability showed consistent correlations with life fraction. A simple qualitative model of magnetic behaviour can provide a plausible explanation of these observations. Comparison with metallographic techniques indicated that EM becomes sensitive to creep damage before it reaches a size that can be detected by optical metallography. Potentially, therefore, EM could be a fast, simple and effective tool for creep damage detection.
Trials on real martensitic steel plant components have also shown that the EM sensor can identify ‘aberrant’ mismanufactured items with weak, non-conforming ferritic microstructures, thus enabling action to be taken to avoid their premature failure in service. Further work is planned to develop the sensor for full-scale component inspection, improve its capabilities on irregular surfaces, confirm its ability to quantify damage and to bring the technology into practical application on high-temperature plant.
