Shaping up - shaped metal deposition
Automated shaped metal deposition is moving forward. Dr Rosemary Gault, Project Manager at the Advanced Manufacturing Research Centre at the University of Sheffield, UK, reports
Shaped metal deposition (SMD) is a manufacturing technique which builds up near-net components from welded wire. It can deliver substantial savings in time, cost and environmental impact for manufacturing companies in aerospace and other sectors.
A typical part for SMD consists of a complex hollow shape, often with external bosses for locating and fixturing purposes, such as engine casings or wear components. These parts are traditionally made by one of, or a combination of, three basic processes:
• Machining the entire part from a solid block. This has high costs in terms of capital investment, material, energy and time, and can make it difficult to modify the design once the tooling is in place.
• Machining from a near-net shape forging or casting. This needs less material and machining, but still consumes a large amount of energy and requires a long lead-time for tooling the initial casting or forging. More complex shapes might not be possible by this method, and finished parts have different material properties.
• A powder-based additive process. Powders can be hazardous to work with, and finished parts will again have different material properties. While this technique is suitable for mass production of small, detailed articles, it may not be appropriate for large components.
The SMD process can eliminate many of these problems. Building the part out of layers of weld material means there is little waste, and machining and tooling are minimised. The additive nature of the process means that small design changes can be made by further machining or deposition after the basic component has been produced. Items up to one cubic metre can be built in the current chamber.
The technique was created and patented by Rolls-Royce, UK, and is now being developed by the Rapid Production of Large Aerospace Components (Rapolac) project led by The University of Sheffield’s Advanced Manufacturing Research Centre, UK.
The system consists of a robot arm with a tungsten inert gas welding head and manipulator, housed inside a sealed chamber. Before welding begins, the chamber is flooded with inert argon gas. A robot then starts to build up the part by layering the wire weld material, with the bead width varying from four to 11mm.
Although the robot arm is computer-controlled, the welding is managed manually by a skilled technician, who monitors the welding head via a CCTV camera within the chamber, and manipulates the process by altering the wire feed speed, welding current and track speed. The feed speed has to be constantly adjusted so that the wire enters the molten pool in a uniform manner and the height of the deposited bead remains in line with that expected from the robotic motion program.
Before the SMD technology can be commercialised, the welding process also has to be automated. This is one of the core aims of the Rapolac project, and it is a complex problem. It requires a thorough understanding of the welding process. Important factors include heat transfer and weld models, as well as SMD parameters such as wire feeds and speeds, clamping points, substrate material and its flexure. Automation also requires complete robot models, including stiffness and vibration effects and how these change as the cell temperature increases.
The Rapolac team is working to improve understanding of the microstructural and mechanical properties of the manufactured parts, and how these change with different control strategies. The finished parts will have different characteristics to those produced by established methods, and the group is developing a database of material properties to identify and demonstrate applications where SMD is most suitable.
The main focus of the Rapolac project is on Ti-6Al-4V, one of the most commonly used titanium alloys in aerospace. Other metals are also being investigated, such as the ultra high strength low alloy steel 300M, the stainless steel 308S93, and the nickel base alloy IN713. Preliminary results showing the mechanical properties for these materials are summarised in the table (previous page).
The Ti-6Al-4V SMD components are fully dense and exhibit large elongated prior ? grains, grown across the deposition layers towards the tungsten inert gas head. The microstructure mainly consists of a phase lamellae in a ? matrix, often forming a ‘basket weave’ Widmanstätten morphology that continues to the surface. There is no ?-case and some structures indicate a martensite.
According to instrumental gas analysis, the SMD components show oxygen and nitrogen concentrations similar to the wire used, indicating that there is no contamination during deposition.
The ultimate tensile strength varies between 936-1,014MPa, depending on the orientation and location of the specimens. The material is twice as ductile when perpendicular to deposition layers rather than parallel. The hardness and Young’s modulus are similar to conventional Ti-6Al-4V with low oxygen content. The mechanical properties are competitive with components from other additive layer manufacturing techniques, as well as cast and, to some extent, wrought materials.
The prototype SMD cell is operated by The University of Sheffield in conjunction with a local SME called Footprint Tools. The Katholieke Universität Leuven in Belgium is investigating the material properties and microstructure of the SMD parts, and weld modelling is being carried out by INTEC, part of the Universidad Nacional del Litoral, Argentina.
Robot modelling and control is being looked at by SAMTECH in Belgium and the University of Catania in Sicily, Italy. The Catania team has developed an automated control for a simplified SMD cell, and is now working closely with the UK collaborators to integrate this with the production cell.
Project partner DIAD, an Italian SME specialising in environmental research, has shown that SMD has the potential to greatly reduce the environmental impact of manufacturing activities. For a typical cylindrical part, SMD uses less energy and oil and produces less CO2 and NOx than machining from solid or from a casting or forging. This advantage increases with the complexity of the part.
The next goal is to integrate the robot controller and develop a business plan for use, giving information on the expected cost savings and material properties for components built using this technology.
Further information: Rapolac