If you were to ask an engine development team what issues were keeping them awake at night, their answers would probably include the challenge of down-sizing and down-speeding of product. While fast running engines can produce more power, typically harnessing turbochargers and other technologies, the additional load this places on them requires engine castings to operate at levels previously unseen.
However, once designers can accurately predict how engine blocks or heads will act under extremes of heat or stress, they can begin to optimise the shape, weight and performance of each casting. This is where computerised tomography (CT), or CT scanning, is already making a fundamental difference to the design and development of modern engines.
The power of big data capture
Real time inspection is widely used in the engineering and medical sectors. However, the adoption of more powerful scanners to capture very large amounts of data in an engineering context is somewhat of a novelty. At the Grainger and Worrall facility in Wolverhampton, we typically gather between two to three gigabytes of data per casting which can necessitate scanning sessions of up to three hours for a complex cylinder head, but it delivers powerful output.
The data creates the foundation for a virtual model, enabling us to check for microscopic defects such as porosities or other visual flaws. The real challenge however, is extracting more detail than what is obvious from the data set. While high level dimensional analysis is becoming more common in precision casting circles, we are using big data to identify typical signs of defects, which can include gas bubbles, metal mould and other clues to underlying issues.
Our experience in the motorsport community has been a powerful catalyst in the development of this more in-depth application of CT scanning. In an environment where a millisecond’s improvement on the track could mean the difference between success or failure, the value of this type of data in identifying and avoiding potential problems, becomes paramount.
In the rare instance of castings failure (in service), the data remains an important record of the supply condition of the part. Any sub-forensic analysis can be focused on very exacting locations, size (>0.2mm) and position within the casting. Such highly detailed analysis of defects can be useful for avoiding problems in the future.
This approach is also proving highly effective in manufacturing situations where tight timescales don’t allow engineers to react, due to shorter new product development cycles. In the world of modern casting, there is no time available to just react to observed occurrences during the process – one has to look presently at all available evidence as a basis for anticipated design or manufacturing challenges.
‘Right first time’ – taking track technology to the road
The ‘right first time’ ethos prevalent in motorsport has influenced a forensic approach to casting and this methodology is now gaining more traction among small-series, high-performance car makers. The speed of scanning, coupled with the detailed analysis such large amounts of data can produce, enables engineers to more accurately predict a part’s performance under increasingly extreme conditions. Key measures, such as dimensional stability, levels of integrity, surface related interactions and behaviour of the casting, are no longer mysterious outcomes, but predicable and manageable aspects of the process.
Through our deployment of CT scanning we have the confidence to move quickly through the engine development process – as long as the analyses are consistent with expectation. Where we encounter unexpected outcomes such as core flotation or metal mould reactions, it may not be a cause for immediate concern but it would indicate a need for further investigation. With the use of advanced CT scanning and big data, this granular-level quality analysis can be done in a rapid and non-destructive way.
Increasing efficiencies in engine development cycles
Castings engineers can often spend a lot of time examining the analytical output of CT scanning but the customer will often require something different, and will typically the see most value in the 3D visualisation of the part. CT scanning allows us to compile and share three dimensional scans in real time to provide a complete and absolute statement of conformity. This validation process, which we pioneered with F1 and other global motorsport race teams, enables us to quickly share information and avoid costly delays in resolving castings development issues.
Traditionally, part approval for a new high performance sports car would take several weeks to resolve but CT technology allows users to identify challenges and overcome them within days, rather than weeks, thus significantly reducing the timescales required for part approval.
Looking back to move forward
CT scanning is also playing a key role in retrospective analysis – an emerging area of non-destructive testing that is helping to realise the dream of zero defects in castings. In the rare cases where customer engineers experience issues that merit investigation, retrospective analysis can be used to determine whether issues belong to the casting or not.
Whatever the issue, this new technology enables vital retrospective analysis of named parts. The sensitivity of the technology enables us to track and trace defects that cannot be identified using sectional processes, such as incipient cracks and defects, which start out as tiny clues that would go unseen by conventional, destructive analysis. The resolution of this technology achieves sub 0.2mm visibility and produces an excellent baseline for analysis.
With our use of CT scanning helping to bring this niche manufacturing process into the light, the days of casting being seen as a dark art have long gone. Not only can we assess, measure and check the external shape of a casting, but also its internal geometries and material quality.
The views of the writer do not necessarily represent the views of the Institution.