Direct rapid manufacturing of metallic parts - the technology of the future? or is it here and now? We offer a UK industry overview By Phil Reeves

Rapid Manufacturing (RM) is the name given to the production of ‘series’ or ‘end-use’ component parts made using Additive Layer Manufacturing (ALM) processes. Traditionally, ALM processes were used to manufacture prototypes and casting patterns. However, recent advances in ALM technologies and materials, now allow us to manufacture parts in polymers, ceramics and metals for a variety of production applications.

How Does Additive Layer Manufacturing Work?
The principle is relatively simple. As opposed to machining, where material is removed from a solid block, or casting where material is melted and forced into a cavity, additive processes work by building-up the required geometry particle-by-particle, layer-by-layer, from the bottom-up.

There are many different mechanisms for both generating a single layer and also for bonding layers together. In some simple systems, layers are cut from sheet material and bonded using adhesives or ultrasonic welding type processes. In others, layers are generated by melting fine powder using a laser or electron beam, and consolidating the new layer onto the previous layer by remelting.

There are over thirty different ALM processes marketed by over forty different companies around the world, but most are focused on polymeric materials and not metallics.

Why is RM becoming so important to the UK Economy?
It is seen by some as one of the most important emerging technologies that will drive the future manufacturing economy. One of the most notable advantages is the potential elimination of tooling. Without the constraints of casting or moulding tools, or machining jigs and fixtures, RM provides manufacturers the ability to produce cost effective batches of one, or the ability to manufacture parts at multiple locations or with multiple product design iterations at no additional cost.

Why is RM different to traditional manufacturing?
RM uses layer-wise manufacturing, so many of the traditional Design for Manufacture (DFM) principles no longer need apply; parts are not made in tool cavities or held within fixtures. Therefore, RM components can be manufactured with no split lines, or with complex internal and re-entrant features.

RM therefore allows for significant part consolidation, reducing manufacturing, assembly and inspection costs. It also allows for the manufacture of topologically optimised components, producing parts that are ‘manufactured-for-design’ as opposed to ‘designed-for-manufacture’.

This can eliminate many secondary manufacturing steps such as internal machining operations or secondary fabrication.

How will RM affect the traditional supply chain?
In principle, RM can reduce or eliminate many stages of the traditional supply chain, which reduces lead times, inventory and supply chain transaction and logistics costs. Moreover, because RM parts are made using additive manufacturing technologies, as opposed to subtractive or formative processes, little if any waste material is generated. This is particularly true of the newer metallic processing technologies, which we will discuss later. Additive manufacturing processes are therefore lean, yet agile, allowing the manufacture of low volume batches of component parts, with little manual intervention.

So who is using RM today?
In recent years, there has been an almost exponential increase in the number of companies using RM across a broad range of industrial sectors. Examples of RM applications include aerospace and automotive components, packaging, medical implants, hearing aid shells and surgical guides, and consumer products as diverse as light shades, furniture and even football boots.

How did we get to where we are today?
The technologies behind RM have been in existence since the mid to late 1980s, when processes with names such as Stereolithography (SLA), Fused Deposition Modelled (FDM) and Selective Laser Sintering (SLS) were introduced as solutions to manufacture prototype parts directly from 3D CAD data. Hence, the term Rapid Prototyping or RP was coined. However, in these early days, the processes produced exclusively polymeric or paper parts.

As the accuracy and repeatability of these early systems improved, parts were then used as patterns for down stream casting processes, such as investment (lost wax) casting, sand casting or vacuum casting of low melting alloys into silicon tools. Hence, Rapid Casting or RC was born.

During the mid 90’s, developments in both ALM systems and materials allowed us to manufacture ‘quasi-metallic’ parts, directly from 3D-CAD data without the need for intermediate Rapid Casting. There processes produced ‘green-state’ parts made of metallic powders held together with either binders or polymers mixed with the metal powder. These were then fired and infiltrated to achieve their ultimate strength. The main limitation with this ‘in-direct’ approach, was that the final part, although resembling a production part had none of the mechanical or metallurgical characteristics of the desired component. Hence, these parts were very seldom used as end use production items. However, some technologies were used to manufacture Rapid Tool (RT) cavities for injection moulding and die casting.

In the late 90’s advances in both laser power and electron beam control technology, allowed companies to develop ALM systems capable of manufacturing parts in ‘real’ engineering metals. With the advent of higher powered solid state lasers, such as yttrium fiber lasers in the early 2000s, ‘direct’ additive manufacturing in engineering grade metals has now become a reality.

The polymeric to metallic RM divide?
Although RM has now become a relatively mainstream, if little exploited, production process for polymeric parts, the process remains in its infancy for metallic parts. This is largely due to a lack of understanding of the technologies available but also the relative immaturity of the technologies on offer. Moreover, most metallic parts are subject to greater stresses, loading and environmental exposure than polymeric parts. Hence, process and materials validation of metallic RM is a far greater consideration to end users than for polymeric parts. Again, this has in many cases slowed the technology implementation. However, there is no doubt that true metallic RM will happen. For example, aerospace companies such as Airbus, Boeing, Rolls Royce, GE, and BAE Systems are all investing in either direct metal technology platforms or research collaborations to implement direct metallic RM into mainstream production.

So what are the processes available for ‘direct’ metallic RM?
Direct Metallic ALM processes fall into three camps, powder bed systems, powder feed systems and sheet consolidation systems. However, process must also be considered as both net shaped and near net shaped. Net shaped parts are within the manufacturing tolerance to the original CAD representation, whereas near net parts will require some form of post process finishing or machining to achieve their ultimate geometric tolerance.

Sheet Consolidation Systems
The concept of sheet consolidation or lamination is not new. For many years, tooling companies in the Far East have manufactured low pressure foam moulding tools by cutting out a tool cavity profile into hundreds of steel sheets, which are then assembled, clamped and bonded to make a laminate tool. More recently this concept has been adopted by US based Solidica Inc. (www.solidica.com) and incorporated into their Ultrasonic Compaction (UC) Form-ationTM Machine tool. It should be noted that this process is not truly additive, as it also requires a significant level of in-processes machining.

UC works by taking sheet material in the form of foil, which is then placed on a build platform. A rotating ultrasonic sonoatrode connected to a transducer is then passed over the foil causing localised bonding of the foil to the platform. A CNC milling head then subtractively machines the profile of the first layer into the foil. The platform is then lowered by one layer thickness and a second layer of foil is positioned over the first. The ultrasonic consolidation processes and machining are then repeated. This cycle is continued until the geometry is complete.

UC has some advantages compared to other metallic ALM processes, as it can be used to bond dissimilar materials, manufacture metal matrix composites and to build parts with embedded optical fibres'. UC is also a net shaped process, which is governed by the accuracy of the machining head within the process cell. However, part complexity is seriously restricted, as the system is incapable of easily processing overhanging or re-entrant features. Moreover, the process is also highly wasteful, as expensive ultrasonic welded metallic foil is machined away as waste.

Powder Feed Systems
Research into metallic feed systems started in the late 1980s using MIG welding torches fitted to multi-axis robots, as a way of building up material onto expensive damaged parts, such as turbine blades or mould tools. In the 1990's welding wire was replaced with blown powder, which although much slower, enabled far better melt-pool control and subsequent part accuracy.

The principle of a powder feed system is to take a jet of metallic powder and literarily fire this into the path of a laser beam. At the point of convergence, the powder is melted. By moving the laser beam and powder feed nozzle over a substrate, using either a 3 or 5-axis CNC machine or robot, layers of material can be deposited onto the substrate at the point of convergence. Powder feed systems are available either as complete machine tool solutions or as modular systems.

Powder feed: Systems
Machine tools (intended in some cases exclusively as repair technology) include the LENS 850-R and 750 systems from Optomec Inc. (www.optomec.com), the Huffman Corporation HP115 and HC-205 systems (www.huffmancorp.com) and the Trumpf DMD 505 System (www.trumpf.de). In each case these technologies produce, near net shaped parts that will require some post process finishing of critical surfaces.

More recently, machine tool company Hermle Innovaris (www.innovaris.de) have launched the Alchemy C40 high material deposition system. Unlike other powder systems, the C40 does not use a laser to melt the powder, but relies on supersonic jetting of the powder onto the substrate causing localised super-plastic deformation and bonding. This supersonic jetting is achieved through the expansion of super-heated steam. The main benefit of the C40 is the very high deposition rate achieved by the process when compared to laser systems. The system also incorporates a multi-axis CNC milling centre, which is used to datum the top of layer following deposition, but also, like the Solidica process, to cut out the exact profile of each layer, resulting in a net shaped part.

Powder feed: solutions
In addition to commercial powder feed systems, a number of companies have taken the approach of selling modular powder feed solutions. Leading German machine tool manufacture Trump (www.trumpf.de) sells a number of modular solutions where a laser head can be integrated with an industrial robot and powder feed systems, to deliver a high flexible and tailored solution. Like the Trumpf DMD system, the Trumpf bespoke solutions also produce near net shaped parts.

Canadian based Accufusion (www.accufusion.com), a spin out company from the Canadian National Research Council also market and sell a modular powder feed laser process called Laser Consolidation. Unlike other power feed systems, Laser Consolidation works by pulsing both the laser beam and powder into an accurately controlled melt pool. The results is a net shaped part that requires little (if any) post process finishing. 

Overview of powder feed systems
Like all technologies, powder feed systems have their advantages and disadvantages. When compared to powder bed systems, powder feed systems are relatively fast in depositing material. They also have larger build enveloped and most significantly they allow the deposition of material onto a substrate, which could be an existing part. Powder deposition systems are therefore often used to add detailed features onto much larger fabricated structures.

The drawback of powder feed systems is that they are often expensive to operate, as they have a large machine foot-print, require a significant volume of inert shielding gas and often produce only net shaped parts which require post-process CNC machining. However for certain applications in the tool making, defence and aerospace sectors, powder feed technologies will have a significant future impact.

Powder Bed system
Powder bed ALM systems are enclosed machine tools very much on the scale of a small 3-axis CNC machine. They incorporate a build chamber into which powder material is deposited onto a build platform. The build platform is allowed to move down in the Z-axis in incremental steps equal to the desired layer thickness. An energy source such as a laser or electron beam is then directed into the build chamber. In the case of a laser beam this is then directed onto the surface of the build platform using scanning mirrors. In the case of an electron beam the beam is directed onto the build platform using magnetic fields.

The process starts by depositing a single layer of powder, typically 0.02mm - 0.2mm onto the build platform. Following pre-heating, the laser or electron beam is then scanned across the surface creating a moving ‘melt pool’. This consolidates the powder material to the both the particles around it, but also the material below. By scanning both the profile and internal cross section of a slice, a layer can be consolidated that represents a slice taken from the original CAD file. After the layer is consolidated, the build platform moves down by a single layer thickness and another layer of loose powder is deposited.

The melting process is then repeated using the scan data for the next layer, before the platform is indexed again and new powder is added. The process is repeated over and over again until the last layer is consolidated. With laser based direct metal systems the whole process takes place in an inert gas atmosphere. However, for the electron beam system the process takes place in a vacuum.

To date, a wide range of production engineering metals have been processed using powder bed ALM systems, these include, stainless steel (17-4, 316L, Ph1), tool steel (20ES - 91RW) and maraging steel (18 Mar 300), cobalt chromium, Inconel 625, titanium (Ti pure, TIAl6V4, Ti Al6 Nb7) and aluminium (Al Si 12 Mg, ALSi12). Research and development is also being conducted to commercialise Inconel 600 and 718, beryllium, copper and Hastelloy.

Powder bed direct metal systems are available from a number of companies, all of which are currently based within Europe. Direct Laser Technologies are available from MCP Tooling Technologies Ltd located is Stone Staffordshire (www.mcp-group.com), Electro Optical Systems (EOS) GmbH, based in Munich Germany (www.eos.info), Concept Laser, part of the Hoffman Tooling Group based in Lichtenfels Germany, and Phenix Systems or Clermont Ferrand in France (www.phenix-systems.com). In all cases laser based powder bed systems are marketed as net shaped production technologies.

In addition, ARCAM AB of Mölndal Sweden (www.arcam.com) manufacture and sell the Electron Beam Melting (EBM) process, which replaces the laser energy source with a high power electron beam. Although significantly faster than laser based systems, the EBM process works in thicker layer with a lower X-Y accuracy. Hence, it is more suited to near net shaped parts.

How do powder bed systems compare?
Each of these technologies has its own unique benefits, including materials flexibility, cost of ownership, the speed of build cycle, cycle time between jobs, build envelope capacity, layer thickness, part accuracy and repeatability, surface finish and metallurgical properties. Some processes build within a heater chamber to reduce residual stresses, whilst other build parts onto strong base plates, which are post process heat treated to remove build stresses. In some processes the parts can literately be broken away from the build plate, whilst in other processes they require removal using wire erosion.

Where are the current applications for direct metallic ALM technologies?
It would be misleading to suggest that all direct ALM technologies are being used for direct part manufacture, as most of the parts produced are still being used as pre-production form, fit and function prototypes. This is very much the case in aerospace where the materials and production process have yet to be fully validated. However, direct metallic RM has already been validated in the medial industry for the manufacture of Orthopaedic Implants, maxiofacial reconstructive implants and surgical cutting guides. Direct metallic RM has also been used in automotive and motorsport applicationsand in the manufacture of complex tooling cavities and inserts.

Economics and reality
Direct metallic ALM is expensive. Machine tools range from £200K to almost £1-million, added to material costs of £80/kg for 316L stainless steel up to £475/kg for titanium 6-4. Given the relatively slow deposition rate of some technologies, the resulting parts can seem disproportionately expensive when compared to cast or even machined parts. However for many users, the geometric complexity that is possible with RM, coupled with the economics and freedom of tool-less manufacture are compelling.

It should be stressed however, that it is not possible to manufacture all geometries using these systems. Most notably machines are limited in size. Moreover, certain geometries can cause problems such with residual stress during the build cycle, which can result in either delaminating or more likely fouling with the powder re-coating system. In truth, operator experience is the only current way of predicting whether a geometry will make a successful build.

Direct Metallic ALM within the UK
Given the nature and configuration of powder feed systems it is simply not possible to put an exact figure on the number of machines within the UK, as many machines have been self assembled using commercial robots and powder feed heads. To date there is only one known UC system at Loughborough University which is used solely for applied research.

However, within the UK, there are currently 26 installed powder bed direct metallic ALM systems. A further 2 MCP SLM systems have recently been sold to a leading research organisation with a global packaging company purchasing an EOSINT M270 (data correct as of Feb 08). Hence, the UK installed based for direct powder bed metallic ALM systems will be close to 30 machines by the end of Q1 2008.

Interestingly, of the currently installed powder bed systems, only five are located ‘behind closed doors’ within leading aerospace, automotive, motor-racing and packaging companies. The remaining 21 machines are located in sub-contract service bureaus, research establishments or Universities, and as such are available in the most part to UK businesses.

The Future
Rapid Manufacturing has been described as ‘an industrial revolution for the digital age’. Given that over half of all direct metallic ALM systems in the UK were installed in the past 12-months, there is some evidence to suggest, that although not yet a full blown revolution, changes are afoot. The true scale of direct metallic ALM will only become clear as the processes are validated into new market sectors and demanded by corporate customers. However, it is not inconceivable to imagine hundreds, if not thousands of machines supporting industry in the future, if the economic balance.

Dr Phil Reeves (PhD Engineering, BEng Hons Manufacturing) is the managing director of Econolyst Ltd

http://www.econolyst.co.uk