Additive Manufacturing Experts Talk Quality

Talk of the promise of additive manufacturing and 3D printing often centers on aspects of cost, speed, advanced designs, clinical outcomes. With any technology used in our highly-regulated, patient-facing industry, quality—of the process, of the machine, of the finished product, of the knowledge overseeing the entire process—must be an important part of the conversation.

Device companies implementing or exploring the use of additive manufacturing must consider how quality applies to the technology. It’s a question that must be answered whether you’re utilizing additive in-house or outsourcing.

A panel comprising OEMs, suppliers and a surgeon—all at the forefront of additive—convened at OMTEC 2016 to discuss, among other things, the quality of the technology and the advancements that will make additive a widely-used manufacturing process in orthopaedics.

In the next five to ten years, additive manufacturing will be able to produce any orthopaedic part made through traditional manufacturing, machine suppliers said. OEMs were more cautious.

Here, we present an excerpt of the conversation. Our thought in presenting the transcript is that it allows you to better understand the technical and complex nature of the topic.


 Andy Christensen web
Andy Christensen, Additive Manufacturing Veteran

Andy Christensen, Additive Manufacturing Veteran: Let’s talk about tech.  Quality is important, and everyone in this room knows how important it is to this industry. Thinking about in-process monitoring as a way to check another box and know something about quality, can you talk about how you see such tools being used moving forward?

   Tim Van Cleynenbreugel web
Tim Van Cleynenbreugel, Ph.D., 3D Systems

Tim Van Cleynenbreugel, Ph.D., Director of Business Development, Healthcare, 3D Systems: In-process monitoring has a big advantage. Obviously it will not replace a thorough validation of your machine combination, but having thorough in-process monitoring gives a full documentation of what happens during a build.

If anything goes wrong, you will be aware of the problems and able to solve them. Tying that into a feedback loop where you do, for example, a melt pool control or melt pool analysis and then tie that back to changing or varying your laser power and laser speed, that can help get higher accuracy parts, a smoother surface. Whether that’s a necessity, I’m not sure, because we have shown that without it we’re consistently able to produce good parts that meet all of the requirements, but it will help to make the process even more stable.

 Magnus Rene web
Magnus René, CEO, Arcam

Magnus René, CEO, Arcam: Layer monitoring is underestimated. I think it’s fantastic. If you compare it to casting, which I think is a fair comparison; you will never know what is in the center of a casting, but here you have this new technology through which you can actually know, after the part is built, exactly what is inside the part. That brings ease of mind to the user or the manufacturer. It also reduces the need and the cost for validation after the part is built. If you know exactly what is inside the part; you don’t need to x-ray it or CT scan it or use dye penetrate. You can actually pass the part without validation, which will cut costs from the manufacturing.

Christensen: Can you touch on the burden of validation and inspection, and what this might mean for the future?

   Kumar web
Mukesh Kumar, Ph.D., Zimmer Biomet

Mukesh Kumar, Ph.D., Director, Advanced Process Technology, Zimmer Biomet: I welcome in-process inspection. If you can see what is happening with your build while it is building, that is a great thing. The question I have for machine manufacturers—and everybody in the room—is, there is a huge amount of data that will be coming out and it will definitely have some blips that something went haywire…what do you do? How does one figure out, based on this information that is coming out is, whether this a good part or a bad part? Until we start answering those questions, just having this data doesn’t do much good for us. There has to be guidance or validation of this sensor system that helps us makes this decision. Just the data doesn’t do anything.

René: Absolutely. You need to validate the validation tools.

Kumar: Exactly.


 james burn web
P. James Burn, MB, CH.B, FRACS, orthopaedic surgeon


P. James Burn, MB, CH.B, FRACS, orthopaedic surgeon: Magnus, if you were to grow a rod of material at the same time as you are building the parts and then you have a layer issue, would it not come out in the rod, which is going to be a test part that you could analyze?

René: Not necessarily, because you could have an issue at a certain position of the build chamber which wouldn’t hit the rod, but would hit the part. You would find it with the layer inspection.

What I was going to add is that you can see not only the discrepancies in the melting process, but you can actually see discrepancies in the material. You have an opportunity to look with XPS (x-ray photoelectron spectroscopy) to see exactly what kind of material is melted and at each point, and you can actually track not only how the material is melted, but also what the material is.

Christensen: I’m assuming as we move forward that you’ll see more areas encroaching on higher fatigue sensitivity. Earlier [Naomi Murray of Stryker] touched on in-process monitoring, and that being a key part of convincing folks [about higher fatigue applications]. What else is there? Is it design driven, is it bulk material driven that’s holding us back from moving more quickly into fatigue-sensitive areas?

René: It’s more of a perceived risk than a real risk. We have technologies today that are used for making high fatigue parts for airplanes, so why wouldn’t it work for hip stems? There are already high fatigue parts like hip stems made today.

Van Cleynenbreugel: With the high purity machines now available on the market, the bulk mechanical properties of the material are not a concern. Obviously, in order to get good fatigue strength you need smooth surfaces, and that might conflict with a desire to have an orthopaedic implant that has a rough surface for tissue adhesion and bone ingrowth. I imagine that we will move in the direction, where you need high fatigue strength you would do machining or polishing to bring down that surface roughness and maximize the fatigue strength.

Kumar: I have asked surgeons what it would take for us to convince them that we can start doing fatigue prone implants. They always ask: Is it as strong as what I use today? Can you vouch for that? Until we achieve that level of just-as-good raw material and fatigue, most surgeons will shy away from it. 

Christensen: Talking about verification versus validation and how companies have approached that for things like cleaning residual powder or cleaning residual materials, it would seem to be much like the counterparts in traditional machining and manufacturing. But we brought up CT scanning earlier. What are your thoughts on verification versus validation for 3D printing of metals?

Kumar: From the legacy Biomet side, we have a lot of experience machining a highly porous structure and then cleaning it. What we have learned from that, we’ve applied to 3D-printed parts. I think there is a solution depending on what you’re trying to do, what is your porous structure, what is your cleaning material. You have to understand that, and then there is work toward cleaning.

René: We as machine manufactures need to do a better job when it comes to characterization of the machines. We have machines that are very complex, where the outcome depends not only on the machine and how the machine is set up, but also the way that you orient the parts. For the industry to develop methods to characterize the machines, I think that is going to be important moving forward, because that will make it easier for our customers to set up jobs and also validate the jobs.

Van Cleynenbreugel: When we start volume manufacturing projects with our customer, whenever we install a new machine or a new material, we always do a thorough validation of the combination of the specific machine, the material and scan parameters so that it is validated to make sure that the bulk material properties that come out of the machine meet the requirements. Once we work on a specific project, we log the whole process flow starting from the manufacturing but also the annealing step, post machining steps, cleaning steps and so on. That is validated as a whole on the specific product, and that is typically done in very close collaboration with the customer.

René: Exactly. That is how it is done today. I’m talking about making this more generic, because today it really is the characterization part of validation.