Digital Alloys’ Guide to Metal Additive Manufacturing - Part 9
Directed Energy Deposition (DED)
June 10, 2019
Directed Energy Deposition (DED) is a category of metal additive manufacturing (AM) that utilizes robotic welding processes to print at high deposition rates but with relatively low resolution. DED systems use an electric arc, plasma, laser or electron beam to melt metal feedstock (wire or powder) into a molten deposit pool. The DED process is typically used for prototyping, low volume production of large, simple parts, and feature addition and repair.
This post provides an overview of DED, digs deeper into its advantages and disadvantages, and highlights applications for which it is best suited. Providers of DED printers are highlighted in the chart below.
Directed Energy Deposition (DED) definition - ASTM (F2792):
“An additive manufacturing process in which focused thermal energy is used to fuse materials by melting as they are being deposited”
How Directed Energy Deposition (DED) Works
Wire DED Diagram:
Powder DED Diagram:
Four key elements of DED systems influence design, part quality, and economics:
1) The Core Printing Process
As with any manufacturing method, understanding the core process provides insight into its fundamental strengths and challenges. In DED, the core process involves how the heat source and feedstock are directed and controlled, and the energy and melt dynamics.
There are four different heat sources utilized in DED: electric arc, plasma, laser and electron beam. It is called “directed energy deposition” because the heat source is directed at the feedstock, at or near the point of deposition. The feedstock, either wire or powder, is then fed into the path of the heat source which melts it, causing the feedstock to drip or spray into a comparatively large melt pool. By controlling the motion of the heat source and material feed, the melt pool is directed around a toolpath where it eventually freezes into a solid metal bead. In wire-based DED, this bead is much wider than the input feedstock. The melting mechanisms are the same as conventional welding processes and require a large amount of energy to maintain the melt pool and successfully bond the deposited material to the part.
Slow Motion: Wire DED (Wire Arc Additive Manufacturing):
Slow Motion: Powder DED (Laser Powder Deposition):
The high energy input has important design and metallurgical implications. To borrow a welding term, there is a “heat affected zone” around the melt pool which is subjected to large thermal gradients that cause residual stresses which can lead to part distortion. These stresses, coupled with the cyclical nature of the thermal process, can adversely affect the grain structure and strength of the printed metal. The residual stresses in DED can be so severe that sometimes the print must be interrupted and stress relieved. This involves monitoring the print, stopping it when distortion surpasses an acceptable limit, allowing it to cool, and then moving the build (which can be very large and heavy) to a furnace to perform a lengthy heat treatment. All of these steps need to be completed before the part can be returned and realigned in the printer to continue the build.
The speed of the printing process directly affects the throughput and economics. In DED processes, the print speed is correlated to the resolution, measured by the width of the deposition. Powder DED processes tend to be higher resolution and thus have lower print speeds than wire-based DED. Below is a chart representing various DED processes and their approximate print speed vs resolution.
Data gathered from printer OEMs, case studies and specifications found online
2) Geometric Capability
Pushing a large liquid melt pool to produce a part layer does not allow much ability to do overhangs or complex internal geometries. Some DED processes can achieve overhang angles by tilting the print bed, but this approach requires advanced hardware and software, and produces only simple shapes of uniform wall thickness (e.g. a hollow sphere or curved tube).
Design of DED parts must also consider the residual stresses induced by the process. A large XY footprint or lack of part symmetry can exacerbate this issue. Toolpaths can be optimized to reduce thermal gradients and residual stresses, but this requires advanced software and simulation expertise.
3) Environment and Power Handling
DED processes create molten pools of metal that require special environments to prevent oxidation and fire. This is especially important for reactive metals like Titanium. DED systems print in a vacuum or inert gas chamber, or locally shield the molten metal with inert gas. There are many pros and cons to each of these solutions. To highlight a few:
Vacuum – This provides the highest quality environment but requires a heavy, expensive, reinforced chamber to withstand vacuum forces (see system pictured above). Also, the chemical composition of the metal can be compromised if the low-pressure environment causes elements in the melt pool to evaporate (as an example, aluminum evaporates when printing Ti6Al4V in a vacuum).
Inert Chamber –Purging the print environment with an inert gas creates a high-quality environment, but requires an enclosed system (chamber). This is the most common approach. Purging a large chamber with inert gas (usually Argon) can be cost and time prohibitive.
Local Shielding – This method directs inert gas directly at the melt pool through fluidic equipment attached to the print head. Local shielding provides less atmosphere purity and consistency but may eliminate the need for an expensive enclosed environment. Local shielding is typically used for very large parts where quality compromises may be acceptable and enclosed environments are prohibitively expensive.
Most DED processes use high heat loads which require large, expensive power supplies. The heat source usually requires active cooling, further increasing power requirements. The large power supplies used in DED systems contribute significantly to machine cost, and expensive electrical infrastructure is often needed to run them.
4) Build Size
DED build envelopes range in size from a 150 mm cube to multiple meters on each dimension. Some DED printer OEMs custom-build extra-large printers for specific applications. The print envelope is related to the resolution and speed of the equipment - the largest printers produce simple, very low-resolution parts. It is also important to note that larger parts and systems are more susceptible to issues with the residual stresses referenced earlier.
DED systems utilize either wire or powder as the feedstock. Most systems use commercial off the shelf (COTS) materials developed for welding or powder metallurgy. This has advantages in material selection as well as availability, quality and price. The wire typically ranges from 1-3 mm in diameter (the deposition width is usually many multiples of the wire diameter). Powder particle sizes are similar to those used in powder metallurgy processes, 50-150 micron.
Most DED processes are capable of processing a wide range of materials. Virtually any metal that is weldable can be printed with DED. Every system OEM has a unique, validated list of compatible materials that they advertise. Some are open systems that allow for experimentation. Below is an example of the materials advertised by Sciaky for their Electron Beam Wire DED (EBAM™) systems:
The chart below illustrates how power-based and wire-based DED processes compare to other metal AM technologies in printing, material, and post-processing costs for an example Titanium part production program. You can read more on this study in our post: Economics of Metal Additive Manufacturing.
DED’s high print speeds and low-cost feedstocks (compared to other metal AM processes) result in low printing costs, especially for larger parts produced with wire-based systems. However, since most DED processes are low-resolution with limited geometric capability, there can be significantly more material deposited than the final part volume. This is referred to as a near-net-shape process, similar to castings and forgings. The ratio of the volume of material deposited to that of the final part is called buy-to-fly. Low-resolution wire DED processes can have a buy-to-fly ratio greater than 2:1 which results in expensive material waste and excess finish machining. While the powder used in DED is more expensive than wire, the powder processes are higher resolution and therefore can get closer to the net-shape, reducing material waste and finish machining costs.
DED machine prices range from $200k for small R&D systems (e.g. Optomec) to over $2M for large industrial printers (e.g. Sciaky). For DED printers with relatively high utilization rates, the machine cost is a minor component of overall part economics. The majority of cost in these scenarios is material use (including near-net-shape expansion) and post-processing.
In DED, usually, only one part is printed at a time. DED printing of small parts is not cost-effective due to the long setup times and other fixed costs associated with each build. For powder-based DED processes, the economics fit best for part sizes that range between the size of a tennis ball and a beach ball. Wire-based DED processes are cost-effective for parts from the size of a beach ball up to multiple meters in length.
Industries & Applications
The strengths and weaknesses of DED determine the applications where it is most useful. These fall into three high-level categories: near-net-shape parts, feature addition, and repair:
There are three primary drivers for choosing metal AM over conventional manufacturing: improved product designs, time savings, and production cost reduction. (please see our blog post on The Business Value of Metal AM). Since DED has limited resolution and geometric capability, the parts that it can produce are similar to what can be made by conventional machining. Therefore, uses cases for DED are restricted to applications where conventional manufacturing is very slow or expensive. One example is the production of machined parts from metals that are expensive or hard to cut. (Please see our post on the Comparison of Metal AM and CNC Machining). DED’s limitations result in it being employed for applications like brackets, enclosures, ribs, tanks, etc. These geometries tend to be low-complexity, but are slow and expensive to machine from billet, cast, or forge at low volumes. An example is the part below - the first FAA-certified DED part flying on a commercial plane (Boeing 787). You can see from the design that it would be time consuming and wasteful to machine this titanium part from billet (a standard process for aerospace brackets).
Using DED for Near-net-shape parts has mostly been confined to the aerospace and defense, energy, marine, and industrial industries. DED’s fixed cost structure, post-processing requirements, and low resolution haven’t made sense for the smaller, higher volume applications that are typical in other industries.
DED can print onto existing parts and substrates. This use case was first introduced as a robotic cladding process to add layers of material to surfaces. Since then, advances in multi-axis robotics and software tools have allowed more complex shapes to be built onto part surfaces. Feature addition can be valuable if the printed feature is expensive to produce conventionally.
In the same way that features can be added to existing surfaces, some DED processes allow multiple metals to be combined by switching feedstock during a print. For multi-metal applications, there are important technical considerations around design, bonding dissimilar metals and ensuring quality. Most manufacturers agree there is a lot more development needed to utilize this capability in production.
Manual welding has been used to repair metal parts and tools for almost a century. DED automates this operation to improve the control and repeatability of the operation, especially important for complex and precision topologies.
DED processes have two key advantages for metal AM: print speed and material cost. The ability to rapidly print using commodity powders and wires has many manufacturers excited about DED’s economic potential. However, DED’s speed comes at the sacrifice of resolution. High speed, low-resolution 3D printing fits applications where extensive finish machining of the printed parts is acceptable or even expected.
A key application of DED printing is the near-net-shape production to reduce the cost and waste of machining high buy-to-fly parts from billet. DED also has the potential to replace many slow, expensive, low-volume castings and forgings. Lastly, the market for DED in feature addition and repair applications is still small but promising.
Increasing adoption in these applications will require large continued investments in education and development. The growth and long-term success of DED technologies also depend on: maturation of the equipment, better pre-process and in-process solutions for DED’s high heat loads and residual stresses, and reduction in material waste and machining time through improvement in resolution (without sacrificing print speed).
As with any manufacturing tool, the user must comprehensively understand its strengths and weaknesses to identify and validate the best applications. DED has a multitude of valuable applications that many manufacturers can take advantage of with the technology available today.
If you are not already on our mailing list and found this post useful, please subscribe with your work email below to receive updates on future posts in our Guide to Metal Additive Manufacturing:
We would like to acknowledge and thank the following individuals that contributed to this post:
Yash Bandari - Oak Ridge National Laboratory
Ronnie Wilson - Oak Ridge National Laboratory
Markus Pieger - Trumpf
Melanie Lang - Formalloy
John O’Hara - Sciaky
Nick Mayer - Norsk Titanium
Please check out other posts in our blog series,
Learn about the technology behind our process,
VP Business Development
Digital Alloys is committed to providing the technology and expertise manufacturers need to use metal additive manufacturing in production — enabling them to save time, shrink costs, and produce valuable new product.