Testing the waters - HMS Astute, a unique nuclear submarine
Mike Wilson and Dr Stuart Hambling, metallurgists at BAE Systems Maritime-Submarines, discuss the techniques that were developed during the construction of the Royal Navy’s most sophisticated and powerful nuclear submarine ever to set sail from British waters.
She is 97 metres long, weighs 7,400 tonnes and must withstand pressure equivalent to 400 family saloon cars. She can carry 38 Tomahawk Land Attack cruise missiles and strike at targets up to 1,000km inland with pinpoint accuracy. Her sonar system has the processing power of 2,000 laptop computers and is so sensitive it can hear other vessels 5,000km away. HMS Astute is no ordinary vessel.
Built by BAE Systems Submarines at its yard in Barrow-in-Furness, UK, HMS Astute is the UK’s largest and most powerful attack submarine. Clad in highdensity rubber tiles, she uses state-of-the-art noise and vibration suppression techniques to maintain her stealth, making her virtually undetectable. One of the Royal Navy’s newest nuclear submarines, she is one of the most sophisticated and powerful vessels ever built, and her complexity is often compared to that of a space shuttle.
The Rolls-Royce pressurised water nuclear reactor that drives the propulsion system will last the 25-year life of the vessel without the need for refuelling, enabling her to travel more than 800km in a day and deploy anywhere in the world within two weeks. The submarine is completely self-sufficient, generating her own oxygen and fresh water, and patrols are only limited by the amount of food that can be carried and the endurance of the crew.
HMS Astute’s fabricated pressure hull (the inner hull that holds the difference between outside and inside pressure) must endure significant levels of water pressure at deep operating depths via its thick, high-strength steel structure. As such, weld integrity of the joints between the steel plates is vital. To ensure that the submarine’s hull plates were welded without defect, the joints were submitted to a series of rigorous tests.
The choice of welding consumable for the hull circumferential butt welds was critical. One way to ensure the required level of quality is to produce representative test welds from any proposed new weld metal consumables, which can then be subjected to volumetric non-destructive and full-destructive examination. Destructive tests include macroexamination, tensile, charpy impact, hardness surveys, flawed bulge explosion and crack-tip opening displacement.
The quality control (QC) hardness surveys of the double-sided, full penetration butt welds comprise a series of three hardness tests that are run in parallel. The indents cover the weld caps at 2mm below the surface on side 1, side 2 and through the root of the joint, and are set to incorporate the parent metal heat-affected zones (HAZ) and weld. A hardness force of 10kgf is applied to an 800-grit finish transversesection- macro of the full weld section. These tests are traditionally carried out manually, making them time-consuming and costly to conduct.
When the time came to procure a new hardness tester, it was decided to try to automate the process as much as possible. The magnitude of the welds associated with submarine construction was beyond the capability of most of the available equipment. Limiting aspects were either the image capture capacity or the physical X–Y movement of the mechanised stage. The metallurgists decided that an integrated micro-hardness capability would be additionally beneficial, and selected a sophisticated micro-macro 0.2–30kgf hardness-testing system that incorporated automatic indentation, motorised XYZ stage, multiple lenses and sample scanning software (right). The installed system also uses sophisticated software to facilitate stitching of multiple images, accommodating the microscopic assessments of complete welds.
The survey process was automated using templates created from software by global materials and testing company Zwick Roell, which allowed the metallurgists to generate detailed hardness surveys. The use of micro-hardness loads over larger and larger areas uncovered teething issues associated with the sample preparation techniques. Because of the lower loads, a micro-surface finish was necessary to facilitate the automatic indentation recognition system, requiring the sample to be completely flat and parallel over the entire area of interest.
Having developed the necessary sample preparation techniques, the hardness surveys gradually evolved in size from several tens of points (incorporating small areas of HAZ) until the complete weld could be covered – involving more than 25,000 individual points to date. This was achieved using an indentation spacing of 250μm alongside a hardness force of 1kgf. Loading and spacing were carefully chosen to maximise the density of the testing, while still retaining sufficient resolution of the hardness results to differentiate subtle microstructural differences within the fabricated joint. The metallurgists were then able to produce visual maps of hardness values from the acquired data to distinguish the marginal variations in hardness and corresponding mechanical properties across the full welded section. The post-hardness testing analysis involved the development of standardised colour palettes and a high level of resolution detail, and the extent of area coverage BAE Systems’ metallurgists were able to achieve is a world first – a technique that could not have been developed without the contribution of principal metallurgist Andrew Bell.
To support the development of pressure hull steels for the next generation of submarines, the metallurgists compared submerged arc tee-butt-weld joints manufactured with various welding parameters. The specimens were manufactured from NQ1 and similar high-strength steel plates, with the aim of establishing any potential effects of parent plate dilution into the weld metal that might arise, as a result of the weld preparation geometry of the double-sided tees. The study focused on comparing the root run and the proceeding weld runs in their order of sequence.
In some cases, the parent abutting plates exhibited minor levels of through-thickness plate segregation in parallel with the centre line. Interestingly, the weld metal appeared to have no discernable localised regions of increased hardness associated with dilution of the parent plate. However, the micro-hardness survey exhibited hardness banding that was attributed to individual weld runs. At the set level of resolution, the details observed within the root of the weld were sufficient to highlight hardness bands directly associated within successive columnar and recrystallised equi-axed regions within the weld metal. The peak hardness associated with each weld run was clearly associated with the highest hardness regions within the HAZ. Also noted were small differences in HAZ width and hardness range associated with varied welding parameters from automated welding practices. Additionally, the survey analysis highlighted elevated hardness regions commonly associated within the cap weld bead and, to a lesser extent, the toe weld bead runs.
Post-hardness testing analysis came into its own when the surveys were viewed in combinations of 2D/3D plan and side elevations, which gave the metallurgists complete understanding of the hardness contour visualisations. When coupled with gridded-map overlays, individual areas were easily identified for SEM element mapping analysis, using the hardness indent coordinates as a location guide.
Such a detailed picture of the variations associated with the complete welded joint can prove invaluable for current and future submarine pressure hull and structural material design. The versatile research instrument can also be integrated with further analysis techniques, such as SEM-based analysis. This will enable, for example, more precise targeting for the fatigue cracks associated with the CTOD test pieces to locations of interest within welds, with respect to both the local microstructure and the mechanical properties.
Although large-area comparative hardness mapping is in its infancy, it is already of key importance where changes in weld technique, joint design and welding consumables have been proposed. To complement the localised hardness QC techniques, a much more comprehensive database of the fracture toughness characteristics of welds is now being generated that enhances the safety case for any new design.
Mike Wilson: firstname.lastname@example.org
Dr Stuart Hambling: email@example.com