Specialist metrology solutions provider, INSPHERE’s chief operating officer, Craig Davey examines the entire gamut of measurement challenges for composites applications and how to solve them.
All modern composites manufacturing – whether in aerospace, automotive or other sectors – involves a diverse range of processes. Robust, reliable measurements both in-line and for finished components, is enormously important to certify parts and to control manufacturing variation. However, the nature of composite materials presents some very real measurement challenges.
INSPHERE has a long history of measuring composites in production and research environments. In this article we discuss the things that make composites ‘special’ from a metrology perspective, and how we have tackled specific challenges to develop reliable measurement processes, and indeed to use metrology to enhance process control in composites manufacturing.
INSPHERE is a specialist metrology solutions provider; our team of experts help our customers to navigate the complex landscape of metrology to develop robust reliable, highly-automated measurement solutions. We also deliver training in metrology fundamentals and offer consultancy services to develop practical applications of metrology in both research and production settings.
Composites components have unique properties that don’t always lend themselves to off-the-shelf measurement solutions.
Bigger and bigger
In the aerospace sector, some 15m components are manufactured as single parts requiring huge mould tools and machining centres, all of which are reliant on very tight aerospace tolerances. In wind power generation, we measure 80m composite blades to maintain tight dimensional control. These applications are pushing the limits of metrology, and indeed physics!
Non-contact measurement often depends on reflected or back-scattered light to record data. Composites, with glass or carbon fibres in a resin matrix, present challenges of high absorption, varying contrast and unpredictable reflection or transmission through resins, all of which lead to ‘noisy’ or unreliable pick-up of the true surface condition. In some situations, surface treatment is possible (e.g. white powder spraying), but with uncured parts this is not permissible; very careful optimisation of laser parameters is necessary, in addition to post-processing data to filter out noise and capability assessments to ensure results are valid.
As measurements are very frequently taken ‘in-process’, parts can rarely be measured in an ideal laboratory environment. Thermal influences affect measurement data and in composites, the situation is often complicated by the interactions of non-homogenous materials with different expansion coefficients. Temperature shifts can also be huge – we have measured composite parts and tooling at over 200°C. Strategies for managing these situations range from straightforward scaling of datasets to complex finite element analysis, but the key is always to understand the nature of the effect well enough to ensure the measurement is valid and fit for purpose.
Some of the features that need to be dimensionally assessed in composites are ‘defects’ that should not be present. Often, they cannot be adequately quantified using conventional tolerancing and it can be necessary, for example, to collect dense point-cloud datasets of parts and to develop bespoke methods of comparing parts with one another to identify regions of interest.
Suitable methods to assess defects cannot easily be generalised; they depend on teamwork between composites manufacturing specialists and metrology expertise. For example, we recently worked with hydrodynamics experts at Ben Ainslie Racing to develop a software tool to analyse critical underwater surfaces of their racing boat based on point-cloud data.
Measure the part or process?
In-process measurement can reduce or even eliminate the need for expensive, time-consuming finished part measurement. However, this saving can only be ‘earned’ through excellent data-analytics to prove that dimensional control of parts is being maintained.
We are experts in understanding the sources of manufacturing variation and how to apply dimensional measurement in-process. For example, we monitor material deposition end-effectors to avoid process drift, and to apply corrections where needed.
It is often vital to maintain datum definitions through a complex sequence of manufacturing processes (e.g. material deposition, trimming, transfer to mould tool, de-mould operations, machining and assembly operations). Detailed input from metrology expertise is essential. We have worked with Airbus, Rolls-Royce, GKN and others to develop robust datum pick-ups.
Statistical assessments of accuracy and repeatability can be difficult to achieve when production rates are small, for example, with composite fan cases or large wing parts. We have developed our own unique tool-kit to ensure robust data assessment methods are employed in all settings, making the best use of limited data to identify ‘signals’ of manufacturing process drift or component defects.
Composites manufacturing is a young field and there are still big wins to be found through innovative R&D programmes. This is very much in evidence at the National Composites Centre (NCC), where the INSPHERE team has been providing embedded support since the centre opened in 2011.
We work closely with NCC member companies to ensure that the complex metrology requirements are met. This ranges from helping to build and commission new manufacturing cells, to developing numerous bespoke measurement approaches to complement new manufacturing techniques.
INSPHERE do not only work with customers on their projects – we have our own major internal R&D programme too. We have developed an automated inspection cell for large components, and we also have technology for in-line monitoring of robots that are ideally suited to material deposition applications.
Embracing Industry 4.0
Composites manufacturing is well-positioned to embrace Industry 4.0 (i4.0) principles to drive automation and improve control of process variation through the intelligent use of data. At INSPHERE, we understand that high-quality, in-process metrology data is the key that will unlock the potential of smart manufacturing. Our experience in the field of composites is paying dividends in helping customers to achieve the efficient, stable processes that they need.
Integrated multi-sensor networks show great promise to guide complex manufacturing decisions and, in future, to deliver closed-loop autonomous control of processes.
Setting up a digital environment to simulate a full manufacturing process (digital twinning) is an important concept in i4.0. Digital twins are far more valuable if they are an accurate representation of reality, and robust metrology data is vital. We have frequently helped customers to understand the ‘built-condition’ of their manufacturing environments to allow accurate simulation and offline programming.
i4.0 promises to deliver greater levels of automation, but automation can mean many things and the end goal may need to be reached by taking small steps. For example, we work with many customers to create software tools to automate their metrology data analysis, applying standard approaches to solve complex tasks and eliminate human errors.
In composites manufacturing, we know the sticky problems and we have developed many neat solutions. We work with our customers to develop in-house solutions to measurement challenges and integration of metrology into new or existing manufacturing processes. Of course, different customers have different needs, and we always try to tailor our offerings accordingly.
We deliver regular training courses in metrology fundamentals, good practice use of measurement instruments, and application-based training for specific challenges.
Not all customers want or need to invest in in-house skills or expensive metrology equipment, and we offer comprehensive measurement services for these clients.
Our active programme of fundamental research in metrology means we are always happy to discuss clients’ individual needs as we are interested in developing new solutions where real industrial needs exist.