During this roundtable experts discuss the latest technologies and emerging trends impacting metrology end users across the testing, production and maintenance domains of aerospace.
Starting with an overview of quality inspection, attendees will be able to interact with panel members from 4D Technology, InnovMetric Software, Spirit Aerosystems and the University of Sheffield’s Advanced Manufacturing Research Centre.
The topics covered include robotics, AI and machine learning, data storage and use, visual inspection, digital twins, IoT and many other advances in areas such as R&D, manufacturing and maintenance.
- Jean-Sébastien Blais, Global Expert, Enterprise Solution Deployment, Innovmetric (JS)Jean-Sébastien accompanies industrial manufacturing organizations to digitally connect their 3D measurement data to their product engineering and manufacturing processes to significantly reduce time-to-market and fabrication costs.
- Dr Erik Novak, Vice President and General Manager,4D Technology, a subsidiary of Onto Innovation (EN)Dr Erik Novak is Vice President and General Manager at 4D Technology, where he was worked since 2013. He has been developing instrumentation for precision metrology for more than 24 years in applications such as semiconductor, optics, aerospace, automotive, photovoltaics, and medical devices.
- Rob Hewison, Technical Manager for Composites, University of Sheffield AMRC (RH)With 30 years of experience in the composite industry, Rob has worked on manufacturing, tooling and design projects in sectors from Formula One to Defence and Spacecraft, with metrology and NDT being critical throughout his career.
- Tom O’Hare, Technical Fellow NDT Materials, Processes and Test, Spirit Aerosystems (TH)Tom is a Spirit AeroSystems Technical Fellow, has managed materials science, non-destructive evaluation programs and the industrialisation of advanced diagnostic inspection platforms for multiple component types
Transcript:
What trends are there in 3D measurement for aerospace?
EN: Component tolerances are becoming more critical. Companies want to optimize efficiencies because of worries about global warming and a need to make parts last longer. They are really pushing the boundaries of capabilities. And that means they need to quantify everything from edges to contact points. Also being able to do rapid CAD comparisons, using programs like PolyWorks and do advanced lifeing models have really brought 3D into the realm of being a necessary technology.
JS: It is all about efficiency. When you have parts that are immediately fitting well together in a large assembly as complex as an aircraft you are gaining efficiency by avoiding certain problems. Also, with international manufacturing, where now you would have a fuselage that is manufactured in a plant that would be in Europe, and being assembled on an aircraft in North America. Being able to predict how well an assembly will match together using 3D metrology data is a great advantage.
RH: One of the biggest challenges is making that data available as quickly as possible as well, to enable us to react to the data, and to enable downstream processes to accept parts from 3D data. There’s a lot of work to do updating all the systems so we can that data fast enough to make it useful almost in real time.
EN: That’s a great point Rob and one of the reasons that we are often asked to move metrology from the lab to the shop floor. Manufacturers want feedback within minutes, rather than days or weeks.
TH: When a manufacturing process is very complex you can have a lot of deviation from the models. You will see a process drift. And if you can do inspection and metrology on the shop floor as a part of the manufacturing process, you can make changes not quite on the fly, but you will understand where you’re having issues and what you can do, so that it can be adapted or changed before or as it gets to your final line of assembly, wherever that may be in the world
How has the role of 3D metrology data changed?
JS: 3D measurement data will help you understand where you’re at with your part. But data needs to be available to the engineer. Availability and access is as important as capturing the data. Who can access it and how fast can you make a decision by looking at that 3D measurement data are key questions.
How is in-process inspection changing manufacturing?
RH: When there is a need to produce things faster as there is in many aerospace companies now, you need to quickly assess the suitability of a part for passing on to the next process and do that in a much more integrated way. In-process inspection enables you to keep non-conforming parts or parts that could non-conform out of the system. More importantly, it enables you to do that as soon as possible, so you’re not costing yourself significant time and money on parts that in the end aren’t going to be used.
How can education be improved to help modern measurement technology?
RH: Particularly in metrology it’s important to understand the fundamentals and what you need to do to measure a part. Rather than focusing on one or two specific processes, education needs to look at those principles of fundamental alignment, repeatability of measurement and how it works.
Without the fundamental understanding, you can still make some pretty significant errors, no matter how accurate your equipment is or how powerful your software.
TH: A lot of our processes until recently would have been grandfathered – where you learn from your peer or the OEM who supplies the equipment. But we’re starting to use more automated equipment, so we are having to improve the education of the people who use it.
There’s an issue around age, in that people of a certain age are not as quick to understand new technology. Younger people tend to take on new techniques or automation of processes faster. And I do believe that even manufacturing apprentices now should do some mechatronics and automation in their education at an early stage.
How are you using automation and robotics in inspection?
TH: Automation is really important and the key is to understand what the best processes are and where best to put them – at what part of a process and at what stage? Technically it’s too late to use a fully automated measurement system at the end. It may stop you from sending a non-conforming part out, but the capture mechanism should have been used earlier in the process.
EN: We see a lot of people targeting the most inefficient parts of a process for automation. In one instance I know of, the inspection process went from more than a week and a half to under five minutes. Just by removing manual inspection steps.
Five years ago when we talked to a lot of people about automated inspection they were using CMMs and not much else. Now you even see automated inspection being used by MROs. Another reason for an accelerating uptake of automation is the labor shortage from Covid-19.
How is the automation of measurement helping advance composite manufacturing?
RH: Recently we’ve started to look at much larger structures and you can’t inspect those without automation. These processes include optical processing or edge scanning of plys for example. You have two main challenges, the actual positioning of a black-on-black ply, and when you automate it, you have to scan its edge and decide if it’s positioned and oriented correctly. Both are significant challenges.
One approach is for a system to flag an area it is not comfortable with for checking, instead of an instant fail. Our processes are well controlled, so by focusing on the areas that aren’t good you save time. Automation is all about time and consistency.
TH: I agree. Value-added inspection is done in areas that are not conforming. The parts that conform, consistently and quickly measured, can be easily automated. It’s where you know your problems arise that the system or software can recognize and flag that problem for human intervention that automation adds value. After a flag it’s fine for a review to be long. But it does greatly speed up the process.
RH: This is where AI and assessment of data and the reliability of data becomes critical. It allows you to be more predictive in the beginning rather than reacting to the issues. You can narrow down to the real key areas that we believe give us problems. And for that to happen you need the most reliable hardware and software available to give you data that you can trust.
What kind of wider pressures are impacting aerospace metrology and inspection?
TH: Supply chain shortages are causing issues worldwide across all the major primary OEMs and the supply chain. Even things like nuts and bolts, carbon fiber and resins have been difficult to get.
When you are components short it can compromise your inspection or metrology plans, because you can’t deliver a fully conforming product to the next stage or you can’t measure the product appropriately.
Another effect of Covid-19 is that we’ve lost people in the industry and some have retired and we’ve not yet got replacements. We’ve been looking at hybrid working to help with that.
How do you account for environmental variations when implementing in-process inspection on the shop floor?
EN: That is a challenge sometimes and it works both ways. There are compromises that happen by moving metrology to the shop floor. For instance, a handheld gauge may have micrometers of resolution, whereas some of the 3D microscopes in the lab have nanometers of resolution.
You have to do correlation studies, there has to be validation of all the methodology. The other challenge is ease of use. For example, we don’t rely on an operator to know when a part is in focus, we color the live video green when the measurement is ready to go. Green means go and you hit the button. They’re not making judgment calls based on 20 years of using a microscope or because they were trained at a university.
Understanding which specifications are important, doing correlative measurements, doing some gauge studies, and then accounting for the operational differences is key.
RH: Underpinning this is that when we design parts we must apply the right tolerances in the right places and that we don’t just blanket increase the tolerances. When you’re looking at design for manufacturing it’s very important that if you’re going to automate inspection on the shop floor, you do the inspection that’s needed and not blanket cover it with inspection that won’t help you to produce a better part.
How do you think inspection technologies are going to change in the near future?
JS: I foresee an increase in the demand for inspection results and 3D measurement data. Their availability impacts how well you can make engineering decisions, do root cause analysis and improve your whole manufacturing process.
RH: In the near future it’s going to be about applying new technologies in a measured way, if you can excuse the pun. As Jean says we will get much better at using data. As we use AI and predictive tools, we will adapt what we do to measure the things that matter more.
EN: Also, we are going to see a push for not just shopfloor or near line measurements, but measurement within the process equipment. As manufacturing goes more additive and automated, as you’re machining it or laying down the next layer, you’re measuring it and getting feedback instantaneously.
TH: That’s exactly right. We are going to end up in a non-contact world where we will be using non-contact metrology right in-line with where we are manufacturing components and parts.
