PSI - Issue 48

Evgeniia Georgievskaia / Procedia Structural Integrity 48 (2023) 310–317 Georgievskaia / Structural Integrity Procedia 00 (2019) 000 – 000

313

4

3. LSS comprehensive inspection In order to determine the crack formation causes in the HU supporting structures, comprehensive inspection was performed, which included:  equipment operating modes analysis;  results of earlier tests and measurements analysis;  non-destructive testing of LSS base metal, weld joints and fasteners (studs and nuts);  structure strength and durability computational studies using methods of mathematical 3D-modeling. LSS non-destructive testing was carried out on a stopped unit as part of its planned overhaul. The following non destructive inspection methods were used to identify cracks in LSS welds:  visual and measuring control (VMC) to identify and fix visible defects and damage;  dye-penetrant flaw detection (PFD) to detect surface cracks not visible or those in suspect areas;  ultrasonic flaw detection (UFD) to detect hidden metal discontinuities;  monitoring using magnetic coercive force (MCF) to determine the actual microdamage number in the LSS metal structure. VMC conducted for 100% welds in accessible places. PFD and UFD selectively conducted in the most loaded LSS weld areas or suspect areas fr om the visual control results. The LSS most loaded welds’ position was determined based on computational 3D modeling and MCF results. The MCF was conducted in base metal and near LSS welds. The main problems of effective crack identification in LSS welds are:  significant weld length (total weld length is more than 1,400 m per unit);  crack closure after stopping the unit;  welds being not always high-quality in the actual structure prevent thorough control area cleaning to carry out PFD and UFD The comprehensive LSS inspection also included LSS base metal ultrasonic thickness measurements and LSS The results of VMC on the LSS of the surveyed unit revealed numerous defects of angle welds in the form of single pores, overlaps, undercuts, weld leg irregularity and asymmetry. Such defects are likely to have occurred either during manufacturing period, or during scheduled repairs, notably quite a while ago. These defects do not prevent further HU operation, therefore can be classified as permissible. In addition to the tolerable defects listed above, two long cracks were found in two different locations on two different LSS spokes (Fig. 3a) PFD confirmed two previously detected cracks by visual means. The cracks were classified as dangerous defects and repaired during the maintenance using the technology, agreed upon with the manufacturer. No other cracks preventing further operation were identified. The UFD use in this case was not very effective for the following reasons:  the operational load stress distribution in the weld metal causes the development of fatigue cracks extending from the weld surface; such cracks are easily identified by VMC or PFD;  crack development is fast if there are significant internal or subsurface weld defects (for example, in case of violation of welding technology); as a result, in the early lifetime stage the crack emerges to the surface, where it is easy to detect by other methods;  slight internal defects outside the stress concentration areas do not normally cause dangerous crack appearance and may remain undeveloped throughout the HU lifetime; fasteners testing by various methods. 4. Non-destructive testing results