A matter of form

An old technique from the shipyards could be making a hi-tech comeback in aerospace. Laser forming, based around the same principles as the venerable technique of flame bending, could be used to ensure that complex assemblies of welded parts all fit together easily, according to engineers at

.

Flame bending, explained lead researcher Andrew Moore, has been used by engineers since the 19th century. It works by creating stresses within a single piece of metal by careful application of heat. ‘If you heat up a small area of metal rapidly, you’ll make the hot area expand, and that yields against the cooler material and the whole piece bends,’ he said. ‘The key to it is forcing the rapid expansion in a confined space, and the temperature gradient that sets up the stresses.’

The Heriot-Watt team, in the Applied Optics and Photonics Group under Prof Duncan Hand, is bringing the technique up to date. In place of the flame are high-powered lasers, with their ability to apply intense and closely-defined heat to small areas. The goals are also different, said Moore. Rather than bending a piece of metal into the specific shape needed for a project, the team is looking at correcting bends imparted by welding so that assemblies regain their specified shape.

Laser forming is not a new technique, said Moore, but until now it has been a process of trial and error. ‘If you wanted to produce a particular shape, the process of deciding exactly where to shine the laser, how you’d apply the heat and so on, was very much empirical. There were no systems available to control it. What we want to do is to make it a non-empirical system, which a non-skilled person could operate, which would allow you to make any shape you wanted.’

The first thing the team did was to simplify how laser energy was applied to their sample metal sheets. ‘Previously, people have tried radial lines and concentric circles — anything that might get the metal to bend into the right shape. We decided that we would only use a grid system. So we project the laser into an orthogonal chessboard grid, and also in lines at 45° to that.’

Varying the amount of heat applied to the metal is an essential part of the process, and there are several ways this can be done; by altering the laser intensity or power, for example, or by controlling the length of time the laser is on the material. The Heriot-Watt team keeps the power constant, using a 2.7kW CO2 laser with metal-cutting optics, and alters the energy input by controlling the speed the laser scans across the metal sheet. ‘The slower the laser moves, the more energy goes in,’ said Moore. ‘The clever part is to determine exactly how fast the laser has to move in which areas to give you the shape you want.’

The key to this lies in three factors. First, a calibration test is carried out on a sample of the material to be formed, determining how much energy will produce how much of a bend. Then, a computer modelling system is used to advise how much stress, and in which positions, will be needed to take the test piece from its initial shape to its target shape. The final factor is that the process is iterative — the final shape is built up over the course of several cycles of laser scanning and heating.

‘The iterative factor was something that was originally specified by the project’s first sponsors, BAE Systems and Rolls-Royce,’ explained Moore. The aerospace sector is interested in laser forming because the titanium alloys needed for high-temperature applications are extremely expensive, both in terms of the cost of the material and the processes needed to make and mould them.

Techniques like laser forming will reduce the amount of material wastage and allow complex components to be made faster. However, titanium alloys are often very springy, said Moore, which means that it is extremely difficult to make large changes in their shape in a single action. An iterative process allows the engineers to make a small change, reassess the stresses needed and try again.

‘It also means that we don’t have to worry about the residual stresses that might exist in the individual pieces of metal, or grain structures that could make it bend faster in one direction than another, for example,’ said Moore. ‘Our initial calibration test can be on any sample of the material, not the actual workpiece.’

Another thing that distinguishes the Heriot-Watt work from previous laser forming projects is the result. ‘Before, people have really only tried to bend sheets of metal in one direction,’ said Moore. ‘But because our system allows us to factor in membrane stress — that’s caused by the shrinking or shortening of the inside surface of the bent sheet — we can make 3D, doubly-curved shapes.

‘An example of this is dished or bowl forms, the sort of shapes that otherwise might involve cutting the material to form.’

The main application of the technique isn’t so much in making new shapes as in restoring parts to their specified shape, however. ‘We’re now working with Airbus UK and Airbus Germany, and their main concern is with parts which have already been welded. This in itself is a process which sets up stresses, and tends to cause distortion. We’ve just achieved our first success in taking a piece of distorted, friction-stir welded metal, and returning it to a flat state.’

The reason Airbus is so interested in this is that it will allow welded structures to be corrected so they fit together perfectly, without any concerns about how they might have distorted during manufacture. Wings, for example, are built up from many curved sheets reinforced by welded strips of metal called stringers, and welding these can alter the shape of the sheet unpredictably. Laser forming could quickly and easily correct these deformations, irrespective of the structure’s initial shape. ‘There’s no real difference what shape you start with, because the system looks at start and end points and iterates towards what you want,’ said Moore.

The next phase of the research is to look at metal components whose thickness varies. ‘Some will have recesses or other features, and we need to know how to cope with those as well as others, especially things like turbine blades, which change thickness along their length,’ said Moore.

And the research may have other applications, too. ‘Because our calibration methods allow us to perform in-process control without needing lengthy finite element analysis, there’s also interest in applying the technique to laser welding itself, to try and prevent stresses building up and causing the deformation.’