Feasibility study into the use of digital image correlation for creep strain monitoring of fossil power plant welds

• Main Author: Cardenas, Nicolas
• Other Authors: Knutsen, Robert D
• Format: Dissertation
• Language: English
• Published: University of Cape Town 2019
• Subjects: Materials Science
• Online Access: http://hdl.handle.net/11427/29595

The life span of high temperature power plant pipework is principally a function of material creep damage - an irreversible plastic deformation of the material when subjected to temperatures and loads beyond a certain threshold. Within Eskom, the South African parastatal power utility, creep damage is primarily quantified by way of Metallographic Replication (replicas). This is a quasi NonDestructive Examination (NDE) technique that looks at the microstructure of the sample in question. Although well-known and used extensively, replicas, as with any technology, have their shortcomings. Extracting of replicas and their subsequent analysis are manual processes that inherently suffer from subjectivity. Furthermore, storage and archiving of vast quantities of physical replica slides for future reference is cumbersome - a challenge that digitisation can address. The aforementioned vulnerability to analysis subjectivity and benefits of digitisation are areas which a technology known as Digital Image Correlation (DIC) - a non-contact, full field, deformation measurement technique - can potentially address. Some research has been done on using DIC for power plant creep measurement; however literature quantifying its performance in this specific application is scarce. This study thus looks into setting up a DIC system optimised for measuring strain in an area of the pipework welds known as the Heat Affected Zone (HAZ) - the weakest part of the weldment. The achievable accuracy is established and the major parameters that affect DIC accuracy are investigated, elucidating the trade-offs between optimising each. In addition, two scenarios exist for the acquisition of DIC measurement data from a plant: when the plant is operating (online); or when the plant is shut down for maintenance (offline). The encumbrance of imaging a hot surface makes the former scenario the more demanding, and was thus investigated. This data was subsequently used to substantiate whether DIC has the potential to be used online (i.e. at elevated temperatures) or is limited to use during shut downs.