Traceability and costs of powders for additive manufacturing

Author Phil Kilburn is the commercial director at LPW Technologies

During my twenty-plus years working for different organizations in the additive manufacturing (AM) sector, I have regularly been confronted by the operational challenges of the technology, specifically in industrial settings. Don’t get me wrong, I believe in AM and that the benefits for a growing field of industrial applications are real and compelling, but with that said, in real life—on the factory floor—the issues involved with running an AM machine, or a suite of them, are many and varied.

Most are to do with the starting materials—the powders. Issues include sourcing the right powder for the application, powder handling and control, material traceability, contamination and, of course, costs. While the AM hardware systems themselves have improved dramatically over the last decade in terms of their capabilities, output and increased automation, the issues around the powder materials for AM have remained largely unresolved.

The big one is traceability, and here we are talking about traceability of the powder from production right through the supply chain to in-process production, recycling and reuse.

The powder production itself is subject to stringent standards, and is (relatively) easy to certify. But what happens to it during storage, in transit to point of use, when stored again at the AM production facility, used, recycled, reused and so on? At every stage, both environmental—humidity and temperature, for example—and in-process conditions can affect the standard of the powder and, vitally, open up opportunities for contamination of the powder batch(es).

If you then take into consideration that every facility where AM machines are being operated will be different—even if it’s the same multi-site company operating the same processes—being able to trace the powder, and the conditions it has been subjected to, becomes increasingly important.

Addressing traceability makes cost issues apparent. One of the much-vaunted benefits of AM is that it compares favorably with traditional subtractive manufacturing techniques in terms of waste materials and the associated costs. This is true, the waste is indeed less. But it is not zero and it is not negligible. Most industrial users I have worked with write off 5-10 percent of the material costs straight away as ‘lost in process’, and there are costs associated with that. And if you get your powder handling wrong (contamination), it gets much worse.

Traceability
For full material traceability, it is necessary to capture data on the powder at every stage in its life cycle and provide a fool-proof audit trail that demonstrates that it is always within specification. As I mentioned above, traceability is vital across the entire life cycle of the powder. Key stages include: the point of production of the virgin powder; transit to point of use; transfer to the hopper and/or machine; in the AM machine; and afterwards if unused powder is recycled which involves sieving and blending processes; and finally, when the powder is re-stored or reused.

One of the most prevalent, but under-reported risks associated with AM powders is contamination of the material batch. Any contamination can render the powder useless, since the complete batch has to be scrapped with all those associated costs. If the contamination is not picked up prior to a build, the situation is even worse. Not many users are aware that even a single tungsten powder particle in a titanium component could potentially form a stress razor and result in the failure of the component. For rotating parts or components, the failure can be catastrophic. What is more, the potential for contamination can occur at any stage across the material life cycle.

One relatively straightforward proposition for the pre- and post-process material handling stages is a hermetically sealed hopper—used to move and store the powder—that can directly feed the AM system with adaptable connection systems. If the hopper also features built-in sensors that monitor the seal and all environmental conditions, this data can be captured to provide a traceable history of the powdered material.

More complex
The traceability issue becomes infinitely more complex once a batch of material enters the AM system. To understand the implications, let’s look at an arbitrary batch of 100 kilogram (kg) of powdered material for a laser bed system (selective laser melting (SLM)/direct metal laser sintering (DMLS), etc.).

Batch A is split into 2 x 50 kg sub-batches and then split again into 4 x 25 kg. These four sub-batches have had no action taken on them that would affect the quality or chemistry, so if required they could be used or recombined without any detriment. Effectively, they are all still Batch A, or 1st generation, virgin material (Gen A).

Image 1 illustrates the potential outcomes for a batch of powdered material. Once the Gen A powder has been through a machine cycle, there are several outputs, namely:

  • components;
  • test bars/samples;
  • powder samples;
  • powder that has been in the hopper but not in the build envelope and not affected by the build (Gen A); and
  • powder that has been in the build envelope and could have been affected by the build process (Gen B).

Each one of these outputs generates information and, in some cases, test data that requires traceability and linking to the specific build. In addition, they each have a different requirement as to what happens to them next, namely:

  • components require post processing and heat treatment;
  • test bars require heat treatment and testing;
  • powder samples require chemical testing, flow analysis and powder size distribution;
  • Gen A powder could be recombined with original batch (company/application specific); and
  • Gen B powder, if recycled, will need to be sieved, blended and then requalified before being used again.

Furthermore, the powder output from each machine may differ and the components manufactured in each build could also have a different effect on the powder removed from the machine. This data must all be captured to maintain traceability on the history of the powder and the quantities retrieved from each build.

Image 2 shows the complexity of the traceability issues over a very short time. In my experience, many companies are trying to use Excel to track and maintain traceability.

At LPW Technology, these are exactly the issues that we have been working on with PowderSolve, our powder life cycle quality control system. It was also this work that led me to explore the cost issues of in-process powder control.

Costs and the potential for savings
Again, to make my point I am going to go with an arbitrary part and arbitrary volumes, along with some basic assumptions based on production manufacturing. The point, however, is to provide a methodology that can be adopted for real-world applications.

For this example, the requirement is for 450,000 components over 25 years. Fifteen components can be constructed from a single powder batch of 300 kg, with 10 components produced per build. Running 10 machines that fulfil 15 builds per machine, per month.

At average costs of 300 GBP (366.5 USD) per kg, a single batch of powder (300 kg) is 90,000 GBP (110,000 USD). Fifteen builds with this batch produces components with a total weight of 225 kg. This means that 75 kg of powder moves out of specification and becomes unusable and must be scrapped. The costs attributed to the scrap per batch are 22,500 GBP (27,500 USD).

So, let’s stack those costs up:

Description

Input

Scrap

15 builds per month on one machine

90,000 GBP (110,000 USD)

22,500 GBP (27,500 USD)

Monthly requirements for 10 machines

900,000 GBP (1099.6 USD)

225,000 GDP (274,900 USD)

Annual powder requirements for 10 machines


10.8 million GBP (13.2 million USD)

2.7 million GBP (3.3 million USD)

Powder required for the complete project life cycle    

270 million GBP (329.9 million USD)

67.5 million GBP (82.5 million USD)

While it is unlikely that these costs will ever be totally eliminated, I believe it is possible to minimize them by increasing powder usage. By increasing the build per batch by just one, the figures reduce considerably as the total powder requirements reduce by 16.8 million GBP (20.5 million USD), and the scrap material is reduced by 4.21 million GBP (5.1 million USD). Calculated over the lifetime of the project, this reduces the initial cost of the part from 600 GBP (733 USD) to 562.5 GBP (687 USD)—a 6 percent reduction.

In Conclusion
The issues around AM for production are complex and varied, but there are many good and knowledgeable people out there working on resolving them. At LPW Technology, we are playing a small part in that, but progress is coming each and every day and I am really privileged—and happy—to be a part of that progression. 

About Phil Kilburn

Phil joined LPW Technology as commercial director in 2014. Focusing strongly on the needs of the end user, Phil is drawing on his proven commercial and design skills and expertise to ensure that LPW Technology’s innovations in software provision and powder handling anticipate the needs of the AM industry. Working with a highly-skilled team of software and applications engineers, Phil is also developing business opportunities to capitalize on the company’s expertise in alloy development, materials analysis and powder life cycle management solutions.

Phil has worked in the AM industry for over 25 years, and led on design, manufacture and rapid prototyping (RP) of custom implants for a leading global prosthetics manufacturer. He has also developed the medical markets, and later the AM market for metal components in the UK aerospace, medical, oil/gas and automotive sectors, for an international consumer and industrial 3D printing organization.