The development of wire and arc additive manufacturing (WAAM) is being driven by the need for increased manufacturing efficiency of engineering structures. Its ability to produce very near net shape preforms without the need for complex tooling, moulds or dies offers potential for significant cost and lead time reductions, increased material efficiency, improved component performance and reduction of inventory and logistics costs by local, on-demand manufacture.
First patented in 1920, WAAM is probably the oldest, outwardly simplest, but least talked about of the range of additive manufacturing (AM) processes. Using wire as feedstock, the basic process has been used to perform local repairs on damaged or worn components, and to manufacture round components and pressure vessels for decades. However, the advent of high quality computer aided design and manufacturing (CAD/CAM) software has made AM in general possible, with WAAM in particular being an area of significant development. With a resolution of approximately 1mm and deposition rate between 1 and 10kg/hour or more (depending on arc source), the operating window of WAAM is between, and complementary to, highly accurate but slower laser-based systems and less accurate high-deposition-rate multi-arc plasma and electron beam systems.
The potential range of WAAM systems is vast, but most fall into one of two types: robotic or machine tool-based. At the time of writing, some commercial machine tool and robotic WAAM systems are available, these are leading the market for integrated systems and include some very capable manipulation systems and CAD/CAM software. However, almost any three-axis manipulator or robot arm and an arc welding power source can be combined to make an entry level WAAM system. Similarly, there are different types of power source available and to some extent the material in use will drive the arc deposition process selected. For example, titanium alloys are usually deposited with more stable TIG or plasma transferred arc, whilst most other materials are deposited with MIG/MAG equipment. The emerging range of low-heat-input MIG/MAG systems are proving particularly suitable for WAAM. Figure 1 shows one of the systems used for WAAM at TWI; this is an industry standard robotic welding setup which is also used for AM projects. The adaptions for AM on this system include modification of the turntable for endless rotation, modified control software, increased thermal management and robust wear parts in the power source to cope with long arc-on durations.
Machine tool-based systems into which the deposition equipment has been integrated have additional potential to allow the combination of AM and subtractive/cutting (SM) processes in a layer-by-layer manner, allowing features to be created and finish machined that would not otherwise be possible. There are laser/powder-based hybrid AM/SM machines available; development of hybrid WAAM/SM systems is underway and it is only a matter of time before a system is brought to market.
WAAM-based process variables and control
Although described above as ‘outwardly simple’, WAAM is not a simple process to use. Whilst the main controllable variables are the same as robotic welding, AM is a different process. All the process variables combine to produce deposit bead geometry, and it is manipulation of this bead that results in the desired component shape. Unfortunately and unlike welding, bead geometry is affected by more than just the deposition parameters; the residual heat as the part is built results in a continuously changing thermal field that must be accounted for if a deposited layer is to be accurate and free from defects.
As parts become more complex, the programmed path of each layer becomes significantly more so. It is rarely possible to strike an arc at the beginning of a layer and extinguish it at the end. Most layers consist of several ‘sub-shapes’ which are programmed and deposited separately but joined together. Figure 2 shows an example of a relatively simple part, which is made up from ten sub-shapes in four different configurations (i.e. L, T, angled T and five-legged X).
Although offline programming software for WAAM is becoming available, the success of the process can be reliant upon the skill of the operator. Although the degree of software capability is developing rapidly, until the ability to break a component down into sub-shapes, decide the order in which they are to be built, consider thermal field, residual stress and distortion, assign appropriate deposition parameters and compile a part programme is completely automated, A highly skilled programmer / operator is required. Despite the difficulties, examples of some very complex parts have been presented in the public domain.
Materials and deposit properties
As a generalisation, if a material is available as a welding wire, it can be used to manufacture parts by WAAM. TWI has deposited materials including carbon and low alloy steels, stainless steel, nickel-based alloys, titanium alloys and aluminium alloys. For many of the materials, the deposit properties are similar but not exactly those expected from conventional weld metal in a joint. The notable exceptions to this are precipitation strengthening aluminium alloys and αβ titanium alloy Ti-6Al-4V.
Al-Mg alloys can be strengthened significantly by work hardening after deposition. The strength of heat-treatable alloys will be increased by solution treatment and ageing, but due to the absence of stretching to create nucleation points for precipitates, they are unlikely to be fully equivalent to the peak strength of a T6 tempered wrought material. Aluminium alloy deposits can suffer from porosity, but it can be minimised by the use of special deposition arc waveforms, high-quality consumables and welding best practice for preparation and handling of all materials.
Deposition of Ti-6Al-4V will result in a strongly columnar prior-β grain structure which has anisotropic tensile properties. However, it has been shown that introduction of cold work as each layer is deposited leads to recrystallisation of the grain structure, resulting in isotropic tensile properties approaching those of wrought product.
Figure 3 shows macro sections of Ti-6Al-4V, AA 4043 and IN718 WAAM walls.
WAAM has significant potential for cost and lead time reduction for medium-to-large engineering components of medium complexity. Careful design for WAAM can enable some topological optimisation and careful selection of wire feedstock can make added material optimisation and multi-material components possible. If AM is combined with a machining platform, it becomes possible to create some otherwise impossible shapes.
WAAM is not a net-shape or fully automated process at this time; until fully capable commercial AMCAD/CAM software becomes available, the part model must be interpreted and the manufacturing process manually prepared, so some operator skill is required for successful part build.
The surface finish (waviness) of WAAM usually means the part must be finish-machined to achieve geometrical or surface finish requirements. However the envelope of material to be removed can be as little as 1mm; this does not increase with component size, so material efficiency actually increases as parts get larger.
TWI has developed extensive knowledge of WAAM over several years' work in generic and contract research.
If you would like to discuss this topic, or for more information, please contact us.