Additive Pioneer Keeps Digging for 3DP Answers

Kim Phelan

Nate Bryant didn’t set out to become an expert in additive manufacturing. In 2013, he was a freshman at University of Northern Iowa (UNI) when the school metalcasting lab installed its 3D printer––only the fourth installation in the country––and he couldn’t wait to get his hands on it. Over the next few years, through continuous use, experimentation, and study, he became one of the lab’s foremost authorities, and word was getting out. Well-tenured professionals from the industry began seeking his advice, and meanwhile, Bryant’s graduate work and research expanded his expertise and contributions to industry’s knowledge base for 3D printed sand molds and cores. Today, he is project engineering manager at UNI’s new smart foundry, a non-teaching research role for the university.
Since 2017, Bryant has authored a dozen AFS-published papers and has conducted both AFS-sponsored and other research on surface finish, machine parameters, smart manufacturing, chromite sand, as well as diverse binders in 3D printed sand applications, with and without coatings. For these accomplishments, he was the 2024 recipient of the AFS Additive Manufacturing Division’s Technical Achievement Award.

Modern Casting met up with Bryant on a video call in mid-May––the conversation went like this:

Modern Casting: You’ve been producing research for over seven years––what do you consider to be some of the most breakthrough outcomes of your work for the metalcasting industry?

Bryant: When we first started to interface with 3D sand printing, we were extremely limited in the materials that were available for it. The supplier of the machine also supplied the raw materials, meaning the binder, and they were pretty exorbitantly priced. I don’t want to say that the manufacturer made a mistake by putting it at a university, but we didn’t necessarily believe them when they told us that you have to use these materials or else it won’t work. 

So, we launched a pretty significant investigation, which looked at regionally available and commercially available materials as alternatives to what the manufacturer was supplying. 

Our findings were actually published in the IJMC in a 2016 paper called “Advancements in Materials for 3D Sand Printing,” and what that showed was, yes, you can use the sand that you’re using in your typical foundry operations––you can use commercially-available furan resin for 3D sand printing. As a result, we reduced the cost of materials by 80% to 90% to produce sand molds and cores using 3D printing. Now the machine was still very expensive, but we reduced that barrier to entry by reducing the cost of additively manufactured molds and cores so that others could essentially use us, the university, as a service bureau to try out this technology. And once they started to see the similarities to manufacturing cores with the cold box or hot box process, that led to accelerated adoption to the point where we are today.

Modern Casting: Weren’t you just a little bit afraid of breaking that $1.5 million machine by using ‘unauthorized’ materials?

Bryant: Luckily, I was a novice at the time … I was 18. I didn’t have the fear of breaking it because I didn’t know any better. 

The material the manufacturer was giving us was an Oklahoma silica. And we did a full characterization of that, and we discovered it’s just regular silica sand, but it’s screened to a different level. So, we went to different sand suppliers and said, ‘Hey, can you make something that looks like this?’ And they did. We worked with U.S. Silica, which became Covia, and also Badger Mining. And we asked them to provide us with silica that matched the properties of what we characterized for the machine. That gave us kind of a one-to-one silica comparison, and we went from $800 a ton to about $80. 

We also qualified some specialty aggregates––things like zircon and chromite, and the engineered ceramics; things like CARBO or CeraBeads come to mind. We were the first to print with those as well. When you print with specialty aggregates, that increases the number of alloys that your process is compatible with. So, we discovered you can produce steel castings with 3D printed molds and cores, because you’re not limited to just silica. It was another discovery that the machine manufacturer originally told us was not possible––but through our long history of testing sand, we found a way to do so.

Modern Casting: You’ve also done a fair amount of work with surface finish and coatings related to 3D printing––what led you into that area of research? 

Bryant: We were providing cores as a service, and we frequently would receive feedback that the castings produced with our 3D molds and cores were rough, which is not desirable. Alongside that, the federal government was very interested in 3D printing for low-run castings for what they call critical sustainment parts––for things like jets, where they might only need two or three. But the problem was, the surface finish was not adhering to their specifications. 

So, from these complaints, we decided we would look into what actually influences your surface finish. And what are some strategies that we can employ to improve the surface finish of 3D printed molds and cores. In my work, for the most part, I show that you can achieve very smooth finishes with traditional casting means, but you cannot achieve that with 3D sand printing. That was the first paper that I wrote: “Critical Characteristics Affecting Surface Finish.”

We tried different things, but we could not improve the surface finish to the point where it was it matched traditional manufacturing. That sort of led to this whole research area where coating companies are saying they think they can develop something that will cover up the layer lines and produce a finish that is at least comparable to traditional means. That work is still ongoing––we haven’t quite reached it yet. There have been some recent innovations, like one out of like Fraunhofer Institute, that I think has a lot of potential in improving surface finish. I wish I could say that we discovered the holy grail of settings or materials that make your castings beautiful, but we just haven’t gotten there yet. 

Modern Casting: Did you learn why 3D printing roughs up the surface of a casting more than a traditionally-made sand mold?

Bryant: We think it relates to the density of the mold that is produced. When you make a traditional, green sand mold, you’re using incredible hydraulic pressure to squeeze your sand as tight as it possibly can be against a pattern surface that’s made out of metal or wood, which is incredibly smooth. What happens then, is you get a very, very uniform and smooth surface that’s representative of your pattern surface. Now, when you’re creating your mold directly, you don’t have that super smooth surface to adhere against or to match. When we do 3D printing, we’re essentially laying down a layer of sand, and there’s not usually that much compaction or pushing those layers together to make a tight surface. We usually see about a 15% density drop in 3D printed sand versus sand that’s blown or mixed or squeezed. I think that has a lot to do with it.

Modern Casting: In addition to the surface finish conundrum, what are some other major challenges with additive manufacturing, and what kind of role will you play?

Bryant: One challenge right now is that I really feel like we need a domestic producer of this equipment––the equipment is made in Germany and we do not have a U.S.- based supplier, which can make technical support very slow. There are also a lot of patents around the equipment, which makes it very difficult to produce one. 

I actually think there’s room to improve the equipment, as well. A lot of the machines use a small printhead, that rasters across the bed surface instead of going in one pass. At UNI, we have a prototype machine that does that, and we can finish a layer in four seconds instead of 38. So, there’s a lot of room for us to improve the speed at which we print molds and cores. But bottom line, I’d like to see less overseas dependence. I think price point is another big limitation right now––this equipment is really expensive.

I’m also interested in the emerging topic of digital twin technology and how it can assist in 3D sand printing. I did a presentation on this at an additive conference last year in Michigan. We can essentially have a digital replica of a manufacturing process and simulate what would happen if we changed things in real time. So, you would make a change on your digital model, and then the physical version of that would match what’s happening on the digital side. I think there’s a lot of potential in that for us to optimize how we create these molds and cores. I’ve started the framework of what that could be, but what we lack is the support from the manufacturers of the equipment to allow us to continue down that path. You need to be able to get into the PLC [programmable logic controller] of the equipment, which controls the brain of the system. 

Manufacturers don’t want you to get in on our prototype unit––but we have managed to get in. However, if we had that support, we could really start to optimize the way that we print. I think maybe eight years from now, we’ll look back on that presentation I did and think, okay, we’re finally starting to get somewhere on this. 

But I think it’s a major area for further development, because we can get AI involved. And we can start making some really interesting decisions in real time as we print. I definitely want to continue on the development of digital twin infrastructure. I think that’s very important. I think that could help us make informed decisions about how to improve surface finish. With AI, it’s going to see things that humans won’t, and it might be able to inform us on a different distribution of sand that we never thought of that would densify better.

Modern Casting: You’re pretty convinced that 3DP adoption is going to really take off among U.S. foundries. What’s going to drive that?

Bryant: I think foundries are going to start to understand that these aren’t that difficult to run––it’s not that different than running a CNC, for example. You’ll also have graduates coming out of undergraduate programs who know how to use these things and know how to design for it. I think it’s going to make foundries more agile and also much more efficient. 

For instance, say you win a job for 1,000 parts where tooling is going to have to be cut. In order to make the best hard tool possible that won’t have to go back for modifications, I would suggest going to your 3D printer and make eight different gating designs––simulate all of them, and find the best one that makes good quality parts that adhere to the specifications of the customer. Then send that design out to get cut for hard tooling. While that’s happening, you’re printing molds of that same geometry, and you’re delivering those castings while your tooling is being made. And then, when your tooling is made, you stop printing and start making them the old-fashioned way: fast.

And since every foundry has multiple jobs going on simultaneously, once you get ahead of it, and you’re making this run of 1,000, you’re prototyping the next one. By the time you’re done with those 1,000, you’ve got your next tool ready. I just feel like it facilitates a very efficient casting production sequence.