Quality control, consistency and product-focused, fit-for-purpose selections of application have been key to the development of additive manufacturing solutions within GKN Aerospace Engine Systems.
The term Additive Manufacturing (AM) covers a suite of different processes all following the principle of creating components layer by layer, enabling the production of complex structures directly from 3-D CAD models. Additive manufacturing (AM) is a disruptive, revolutionary technology that will impact global manufacturing and can be applied right across GKN Aerospace’s product portfolio. AM is one of GKN Aerospace’s strategic priority technologies and the business is leveraging the skills and expertise across all divisions to offer a fully integrated solution.
Most people in the aerospace industry will by now have seen demonstrations of AM and the almost magical way in which the most complex shapes can be built up, layer by wafer-thin layer, by a computer-controlled welding source melting metal with incredible precision.
But what does AM mean for a company such as GKN Aerospace?
The potential benefits are considerable and enable designers to think in radically different ways when developing components. But making use of AM requires extreme levels of precision and control. Different metals and alloys behave differently when they are molten and when they solidify; that can require changes to the manufacturing process.
On the engines side, the company’s AM journey started around 2001 with development of wire-based deposition using TIG and Nd-Yag lasers as heat sources. This development was very exploratory and mainly focused on developing the process itself. Titanium 6-4 was chosen as the preferred material in order to restrict the scope of work. The ambition level was high and a target of building metre-sized structures was discussed.
Very soon after the company’s initial trials a dedicated initiative started with the aim of developing a nickel-based wire additive for a major spacecraft component. This was the next generation of main engine nozzles of the European Ariane space launcher system.
“The development team has been working very hard to develop robust parameters and deposition techniques that now gives us the opportunity to tailor the solution adopted for the Ariane 6 hardware to our customer,” says Henrik Runnemalm director of research and technology at GKN Aerospace Engine Systems. “The component is roughly two metres high and 1.5 metres in diameter, so it’s quite a substantial part. The learning from building that component has generated fundamental knowledge of building large amounts of material and that, of course, has been used to expand into other types of components using the same, or similar, materials.”
Creating larger components by AM brought its own set of challenges, particularly in stability terms. In large parts “the melt-cool area is not really stable”, explains Robert Reimers, manager of the research and technology group leading the manufacturing process development. Overcoming this problem required a more robust control system to keep the manufacturing process constant.
Peter Jonsson, programme lead for the additive manufacturing technology portfolio, has been part of this journey from the start and admits there have been ups and downs. Jonsson recalls situations 10 years ago when he was happy just producing a geometry of defined shape. Today, it’s all about quality, consistency of results and cost efficiency in the process down-select.
Over the years, GKN Aerospace Engine Systems has improved its AM control systems to monitor the AM process, so it can be adjusted in real time. This gives a more automated process, which has always been one of the company’s main targets.
“One major area, where we learned the hard way, is that using a TIG-based approach” – where a source of heat melts a wire and the molten metal is deposited on to a substrate – “was not optimal for the application we were looking for and we therefore had to change from TIG to using lasers,” says Runnemalm. “That has allowed us much better control of deformation that will appear.”
The AM development programme has focused on all aspects of the process, from laser to material interactions, material characterisation, process control, machine control and specification, health and safety aspects, human machine interactions and process control and automation system.
“What has been key to our understanding has been the building of an agile data acquisition, analysis and feedback system,” says Jonsson. “This system has been developed between our welding and additive teams and is now the backbone of everything we do. Data is essential and analysis capability is key to success.”
“We have spent a great deal of energy in generating data on what is really going on in the AM process,” adds Runnemalm. Having a control system with the built-in sensors that can constantly track what is happening in the deposition has been a key element in helping to create materials that can be certified as ‘fit for flight’. This ‘agile data acquisition system’ can have sensors added or taken away in order to provide the information that is needed for specific material applications.
In parallel to the physical work, a simulation capability to predict the behaviours of additives has been developed. Simulation of temperature, deformation stress build-up and shielding gas flow behaviour, as well as material formation, is necessary to predict the behaviour of the additive process.
“It’s a staggered approach,” says Runnemalm. “You take one step forward in the real world and another step forward in the virtual, or simulated, world and by adding those two together you can prove the development of the simulation tool. The benefit then being that you can use the simulation tool to predict a multitude of possible scenarios in the physical world.
“For example, you can use it to predict what deformation there will be in a component if you deposit with a certain parameter. That’s very attractive to learn ‘up front’, before we actually do the deposition.
“We have the benefit of years of development in predicting welding that we have now expanded into this area. The main challenge we have overcome in this area is to bring the computational techniques to the industrial level in this field. The ability to predict full-scale engine components in welding with hours of physical welding has brought us to an understanding on when to select pre-defined strategies to run successful simulations with known accuracy levels in the models.”
There is a ‘triangle’ of aspects involved in developing the relevant computational techniques. To start, you need to be able to translate the additive manufacturing process into a computer program. Deep knowledge of the parameters of the various materials being used is also essential, as is advanced ‘number-crunching’.
One aspect that complicates these calculations, says Reimers, is that, unlike most simulations where the amount of material involved in a calculation remains constant, the very nature of additive manufacturing means that more material is being added as the process continues: “What you’re analyzing is constantly changing in mass,” he says. “That’s a challenge; traditional finite element tools don’t operate like that.”
In 2008 a major decision was made to introduce Laser Metal Deposition – Wire (LMD-W) into Rolls-Royce’s Trent XWB program as a preferred solution for the Intermediate Pressure Compressor Casing (ICC) where a number of features (bosses) in Titanium were required on a large fabrication. The design, material and process development and certification continued until 2012 when the product received certification. Currently GKN Aerospace is producing this hardware at the rate of five to six completed examples leaving the production plant in Trollhättan, Sweden, every week.
Since this introduction GKN Aerospace’s main focus has been given to developing a second-generation control system and advancing the process window for the LMD-W process. That allows more advanced geometrical features to be developed.
Another AM method, Laser Metal Deposition – Powder (LMD-P) that began to be developed in 2005-2006, reached the point in 2013 where it could generate the qualification logic, material data and process parameters of introducing the LMD-P process for Inco718 parts. The first production machine that is applying the AM methods was successfully installed in production in 2016.
As the AM process has matured, GKN Aerospace Engine Systems is increasingly looking for new ways in which it can use the method. The company increasingly seeks to design components in a way that makes them candidates for AM. “Additive manufacturing will definitely be one of our key processes to build new components,” says Runnemalm.
GKN Aerospace Engine Systems always likes to have a suite of new technologies from which it can select the most appropriate for a task, he adds. Developing AM together with the accompanying simulation techniques adds one new element to this selection matrix of techniques: “We believe this will open new windows of opportunity for products that have not been possible to achieve before.”
For example, AM allows the company to integrate several functionalities into a single component. That major advance allows designers to think in new ways when creating a new component.
GKN Aerospace has gone through a number of material development programmes to qualify the material coming out of the process. This is critical to understand the allowables that can be used in the design process. Tight collaboration with the design and analysis teams has led to a number of successful certifications of products that are now flying.
The number of products being produced by AM is growing exponentially. Reimers explains that GKN Aerospace’s ambition has been to make sure that it develops robust processes with the aim of putting the results on products where they make a difference: “In many ways, we can build ‘anything’, but it always has to win its way on to an engine.”