Trends in Metal Additive Manufacturing: The Importance of Powder
Hello everyone, thanks for joining me. We are here today to talk about the importance of powder to metal additive manufacturing. I want to thank 3dpbm for being the sponsor for this event. 3dpbm is a platform that provides media, insight, and analysis in the additive manufacturing market, which is a very useful service recognizing that this industry is fast-paced and fast-growing, so again thanks to them.
I’m very excited that we have a very diverse group of experts here today, to talk about the importance of powder to metal additive manufacturing. We have a group of people that span the spectrum from powder production, all the way to powder use. On the powder production side, I’m happy to have Dr. Martin Conlon from my company, Equispheres. Equispheres has a unique technology for producing powder which is specifically designed for additive manufacturing.
Then moving up the supply chain, I’m very pleased to have Filip Francqui, the General Manager of Granutools. Granutools is a manufacturer of high precision, high technology instrumentation for measuring powder characteristics and powder behaviour.
Moving up further in the supply chain, I’m pleased to have Dr. Mathieu Brochu who is a professor at McGill University. Dr. Brochu conducts a considerable amount of research into metal additive manufacturing and powder bed fusion, in particular. He does research on powder but also in the area of printing parts and mechanical performance. Dr. Brochu, in addition to being great academically, is also deeply integrated with the aerospace community here in Canada.
Finally, and I’m excited to have with us today Simon Pun, Lead Metallurgist at Divergent 3D. Divergent 3D is an exciting company in the automotive space, they’ve come up with technology leveraging metal additive manufacturing, that allows for the production of vehicles at a much lower cost than the traditional factory. Divergent 3D is “dogfooding” at the moment, as they are using their technology to produce a car, the Zinger Automobile. If you haven’t seen this car, I’d encourage you to look it up, it looks like the Batmobile! It’s a very, very cool looking automobile.
And of course myself, Doug Brouse, I’m the Vice President of Strategic Partnerships and Alliance Equispheres, I’m very pleased to be hosting this event!
The first question I’d like to ask today, I’m going to present to the group as a whole, which is, metal additive manufacturing holds a lot of promise, for lightweight production, bespoke design, improved supply chain, but it still hasn’t completely taken hold in the industry. It’s still used mostly for prototyping and R&D and hasn’t disrupted the 13 trillion dollar a year manufacturing industry, and I was wondering if the group could give me their thoughts of what they think the obstacles are that are preventing metal additive manufacturing from going mainstream?
I guess from my perspective the main barrier to adoption is just the cost associated with the technology, which is mainly linked to the speed of production. There are also somewhat limited materials available, and generally speaking, the reliability of the printed parts tends to be rather variable which impacts the mechanical design. Those are the three that would come to mind for me.
I want to piggyback off of Martin’s comments there on the materials selection. I think this panel and many of the users, like Divergent, know that there are only a very select few materials that can be printed. Specifically, if we talk about laser bed powder fusion and aluminum, right, there’s only a handful with the famous AlSi10Mg which was developed from a weld wire composition. From a user’s perspective that really limits what we can do and provide in terms of applications. One challenge we see is the trend or need to develop new materials that are specifically designed for 3D printing, in an application-driven manner as well.
There is a need to develop new material specifically designed
for 3D printing and specialized 3D printing applications.
On my side, what I will add is reliability. This is one of the things I believe in strongly. How can we increase the reliability, reproducibility of the part that we are actually fabricating? This is to me a big issue, to reassure all the final users of those AM parts, that the part will sustain both the life and the performance that we are looking into. This is a big thing that we have to solve, how we can increase any aspect that can improve the reliability of parts is critical.
If I may add upon what Dr. Brochu is saying, if I consider the main industries at the moment which are aerospace and implants – medical implants for example – they are both strongly regulated industries. And in order to increase the mass production in those two industries, I think we have to be able to speak a common language, and we have to be able to progress in terms of standardization. There are a lot of efforts that are currently focused on those areas, but it takes time. It takes time to develop a common language, a common understanding so that it can be spoken to the whole supply chain. It’s something we are putting a lot of effort into. Granutools and many other companies are currently meeting a couple of times a day so that we can progress, move faster, develop a common language, develop the same standards so that we can integrate those into aerospace and medical implements industries, and then move forwards into volume production.
Thank you, those are interesting perspectives. I’d like to ask a question directly to Simon, because Simon is our user in the group. There is an adage in business that you can have it good, fast, or cheap, but you can only pick 2. I was wondering if you could tell us what the priorities are at Divergent 3D with regard to what’s most important to your organization, and also how that relates to feedstock. What exactly are you looking for in feedstock, what kind of issues are you having with feedstock. What are your thoughts?
Being a tech startup company focused heavily on automotive, I’d say productivity is a hurdle we are trying to get over when competing with conventional manufacturing techniques like forging and casting. Productivity has always been the key milestone that we try to reach to compete with these other techniques. Right now with powder stock, I believe the industry standard is 2 cuts, 15 microns to 45 microns PSD, D10 to D90, and then there is the 23-25 micron to 70 microns but as we see, the maximum layer is 73 microns.
If we wanted to increase productivity, we’d have to shift and have the specs, standards in order to shift that up to a courser cut, and at the same time, we need the powder feedstock to pack well so that we get the repeatability that Dr. Brochu had mentioned. So that’s a big challenge, and that’s where we are trying to work with our powder vendors to develop our machines.
We have a strategic partnership with SLM Solutions and that we are basically taking our feedback like this to the machine manufacturer and saying hey, how can we print courser powder and increase productivity.
To increase productivity you need to increase printed layer thickness
and to do that you need to print with powder feedstock that packs and performs well in a thicker
application to ensure consistent printed part quality.
I’d like to explore this a little bit more. In the automotive sector, cost is very important, but we’re seeing this also in the aerospace industry, and a lot of printer manufacturers have been adding more lasers to their systems to get the cost of production down. We gave recently heard they were up to twelve. Certainly, as you describe, Simon, thicker powder layers area good way to increase production speed. But I’d like to ask the group, what more could the powder do? What other features and characteristics of the powder are possible that could help with the speed of production?
Dr. Conlon (11:11):
I think what Simon mentioned is the primary one, which is increasing build speed by increasing layer thickness. But in order to do that you have to have a nice spread density in the layers, so you don’t end up with porosity in the final part and impact mechanical properties, so that’s a big lever to pull there.
Along with what Martin just mentioned, I believe the spread density is going to go as well with the spreading speed, how can you do this optimization? You also have to take into consideration, that within the spread density you need to minimize the fluctuation of the laser interaction with the powder bed. Then in terms of reproducibility for your mechanical properties, you have all the factors that come into play with the surface chemistry; the moisture pickup, all of those aspects that are often neglected, in terms of handling characteristics that need to be taken into account. If you can control all of this, you minimize all of those detrimental effects of moisture pickup, you’re going to maximize a chance of obtaining high mechanical performance parts.
If you can minimize fluctuations of the laser’s interaction with the powder bed by
reducing inconsistencies in the powder’s surface chemistry and reducing moisture pickup,
the result is higher mechanical performance in your printed parts.
So, Filip, Dr. Brochu had mentioned some characteristics of the powder which he considered to be important such as spread density and moisture pickup, what can you tell us about the class of tools that you’ve developed or seen, that are helpful in identifying these properties, and what can you tell us about them in terms of their ability to be useful in a production environment?
The way we ended up working in additive manufacturing was a little bit by accident actually. We published a paper focused on spreadability – an academic paper – and we discovered that what actually drives the ability or the inability to spread, is the coupling between the cohesive index, and the cohesion inside the powder and the printer; the re-coater of the printer. So that was the first requirement was the ability to spread equally. It is very important that powder can actually be spread across a layer where it is flat and homogenous. That was the first requirement.
The second requirement and this is still under discussion, is the property related to the packing. We just mentioned it is important for the feedstock to be able to pack well, we talked about repeatability and we talked about how fast it can pack. My current understanding is for most applications, the repeatability is most stringent. The problem is when you have layer after layer, and you have differences in the packing of those layers. That will induce non-homogeneity inside the build, and inside your parts. So, it’s more about repeatability than it is about the speed of the packing.
The third one which we are working a lot on is the electrostatic properties. The electrostatic properties are very often overlooked, but they are very important for two reasons. First, because electrostatics can play a role in the cohesiveness of the powder and therefore in the spreadability, but more importantly if the powder gets charged, then the charge can build up and the powders can actually stick to the walls, stick everywhere, create clogging and so on. From a holistic approach, I would say spreadability is very important, the packing, not necessarily speed of the packing, but more the repeatability across the build and finally, electrostatics.
Three important powder characteristics for better performance are overall spreadability,
consistent, repeatable, homogenous packing ability, and minimal electrostatic pickup.
Thank you. So, Dr. Conlon, both Filip and Dr. Brochu have both identified key features that they believe quite firmly are important in metal additive manufacturing. As a company that’s producing powder for the industry and producing powder that is specific for metal AM, do you agree with their assessment and what is your organization doing to provide powder that has these kinds of features and characteristics?
Yes, I do agree with them, I think that they’ve hit on many of the important properties in the actual application and those properties or that behaviour is derived from the attributes of the powder. That’s why we strive to make powder that’s highly spherical, with narrow particle size distribution, an absence of fine particles that are prone to that electrostatic charging that Filip mentioned, and a smooth surface so that they adsorb minimal moisture and that if you bake them they desorb really well. That’s our focus.
Now I want to talk a bit more generally about the industry. Simon, your operation at Divergent 3D, you’ve got close to a dozen printers, you have a relationship with SLM as you mentioned, what would you say to the industry as a whole, not to just SLM, but to the printer industry as a whole; what they could do to improve quality, improve production speed, which we’ve been hearing about, improve powder handling, are there any comments you have about those issues?
I’d say you hit on the entire portfolio of areas where they need to be improved. But something that’s closest to me as a materials engineering leader, Divergent is – again I have to bring it back to the productivity problem – and I know we have a lot of experts with powder here, as we increase the print speed by manipulating process parameters, we see a lot of defects that could be mitigated if the powder quality is better. By that I mean they are spherical, they don’t have satellites, could be well-packed, so all those things contribute to the lessening of the charging of the powder, so they become less sticky and can actually spread well, and we can get more homogenous parts and microstructure. I would say there’s definitely room for improvement in the powder quality. The other thing I think you also mentioned, post-processing. Right now, and I think I witnessed this in aerospace, you have a team of post-processing technicians where they are manually removing support structures and other features that won’t go on the actual part. So, if post-processing could be automated or semi-automated, I think that is the trend of the industry. We talk about 4.0 and digital trends, and I think this vision of automating post-processing fits well into the digital factory.
The automation of post-processing (for example, the removal of support structures)
is a likely trend we’ll see next in the metal additive manufacturing industry.
I’ll put this one to the group, if you had to give advice to powder producers about what they should focus on to improve the quality to address some of these issues that have been raised, what advice would you give?
I mentioned the limitations of using these “off the shelf” commercial alloys that are 3D printable, my advice for any powder maker would be to be open to trying new chemistry. Investing in groups like Dr. Brochu’s, to make a couple of new batches of new powder where you can put them in the printer, and see if you get better properties than what is commercially available. I think that is key to unlocking many applications, at least for the automotive industry.
Dr. Brochu (20:55):
I don’t want to call it ‘advice,’ but a ‘suggestion’. I really like the concept of trying to understand what are the powder characteristics? What are the limitations? What are the advantages of all of the powders that people are actually supplying? To fully grasp what is the variability, and to foresee what might the issues be that people might face when it is time to print. So sadly, I have a feeling that too many people are believing that all of the powders behave the same, which is not really true. All of those minor differences can sometimes be healed by the process, but sometimes cannot, so I firmly believe that this aspect needs to be thoroughly understood. How do we ensure that there can be good compatibility between the powder and the processes?
Too many people believe all powders behave the same,
which is not true.
Great, Great. We’ve been talking an awful lot about powder bed fusion almost exclusively. Certainly, there’s also a lot of interest, and we talked about production the speed of production, there’s also a lot of interest in binder jet technology because it has the promise to produce parts 100x faster and at a cost that is comparable to casting. What role does everyone feel that binder jet printing will have within (additive) manufacturing? Will it dominate? Will it have a role to play along with powder bed fusion? What’s the thinking there?
Dr. Brochu (22:58):
So, from the academic perspective, binder jet is a sintering-based process, so it’s a totally different ball game in terms of the cohesion between each particle. The surface chemistry and possible contaminations within the binder during the process will play a huge role in the process. There’s still a lot to the aspect of binder jet printing that needs to be studied. Are we going full solid-state sintering? Slight sub-solidus sintering? There are many aspects that need to be studied on that front to fully grasp the potential of this technology, so yes, for mass-production of potentially lower-cost application, binder jet has good potential, but still as I mentioned, we are still having the same hurdles, what will be the reliability, repeatability of the process, and all of this now comes down to the sinterability and powder properties.
Yeah absolutely, and again, coming from the automotive industry perspective, automotive unlike aerospace is very cost-sensitive. Even with raw material, as opposed to aerospace, the buy-to-fly ratio is more important. So, we look closely at feedstock, and this is something I learned from working with the procurement team and several vendors is how can we maximize the usage of the yield? And when we talk about traditional gas atomization techniques, binder jet we know utilizes a cut, a PSD similar to MIM. So, what we are trying to do is see if we can utilize the entire yield by allocating the MIM grade yield of the atomized lot for binder jet, we already set that middle for SLM or MLS, and for a courser cut maybe for DED or EB. So from the automotive perspective, we think that there is a future for binder jet, because we can utilize the yield with this technology.
When it comes to powder properties, binder jetting’s most important property is the packing of the power. If the build is too tight, then the binder is not able to really go through the whole layer of powder. Now if it’s not tight enough, the density won’t be there, and the parts won’t be there in terms of meeting the requirements for density. So packing is even more important. It has to be very well controlled. On the other hand, I’ve been working a lot with binder jetting customers, and there is a huge future for this technology because basically all the MIM customers, all the metal injection molding, all the casting customers, can reuse parts of their plants. They can divert their process of the parts and use binder jetting, then put all the parts back in the oven and sinter them, so there is huge productivity gain when that can be done.
Last question I think for the evening. I know for Filip it’s getting late in Belgium. In general for the group, what are your predictions for metal additive manufacturing over the next twelve months? Are you expecting to see greater adoption, are you hoping to see some of these new alloys that Simon is hopeful to be able to utilize? Where do you see the industry going?
I think I read somewhere, it was probably an Ernst and Young report, probably from late 2019 that said 18% of end-users have already adopted metal additive manufacturing in their production line. That’s across the entire manufacturing platform, but specific for automotive, in the past few years we are seeing a lot of buy-in, from the CTO and the upper C-suite folks having tremendous interest in investing in metal AM for the production line. I believe it’s going to continue to grow, especially if we can unlock other applications with new chemistry. I would imagine the growth to be five-fold over the next twelve months.
There you are, bold prediction.
I’d like to go and bet on a number, but I do believe as Simon just mentioned, we are going to piggyback on the hard work that the industry has done in the last couple of years. Now we have quad lasers that are starting to bring a positive economic balance to the production of parts. There are still some issues that need to be solved, but I believe that this is going to grow more this year.
Another thing that Simon touched on that is really important is the buy-in, we have more and more buy-in because again, piggybacking on the work that was done in the previous years, people are starting to believe more and more in the properties and reliability of parts. There’s going to be growth around getting buy-in from the (manufacturers) as well as the authorities, who will actually push more. The more restriction or more severe conditions applied to the goods we are working with, (which AM can meet) will allow AM to fully capture this opportunity and deliver those complex parts, serve those more stringent applications.
Metal AM is getting more and more buy-in from the C-Suite and that
openness to adoption will be fueled even more by increased manufacturing regulations,
which AM can help meet.
In 2019, at the end of the year, we were seeing people that had been buying a lot of powder bed fusion instruments, and they were basically getting money to buy two or three of the systems to try them out. Then we saw a strong push from their managers saying “Let’s buy more powders, let’s get this process under control, and let’s make parts.” There was a huge push for that at the end of last year. Then we saw at the beginning of this year was binder jetting picking up, but still, there was a strong push from the managers to get parts produced and get these parts used in their products, and the products sold. I think this is still the case today. Now, if people find problems with the process, they have to fix those problems, they have to buy more powders, and they have to produce more parts. And it is what everybody is expecting, it’s really the proof in the pudding. This is happening now, and I think it’s going to happen for the next twelve months.
Those are good insights. I noticed that nobody has taken a concern about COVID-19, but COVID-19 has certainly impacted the aerospace industry. Dr. Brochu, you’re deeply integrated with that community. What’s the prognosis there in your opinion?
Dr. Brochu (32:13):
With the aerospace industry, the manufacturing will restart when COVID disappears. If we can say, in full quotes, “it’s going to disappear.”
There is a side tangent to this, Doug, that COVID was a real reality check to show how AM could spin on a dime and start producing parts that the conventional manufacturing industry could not. And if you go back to the months of February, March, April, the first big period of COVID, the thing that you were hearing was that “AM helped develop this,” “AM helped to develop that.” And there was a lot of exposure that happened with this, so I think that yes it impacted very badly many industries, but the AM field keeps progressing, and it made really really really, total progress during that period. When the economy revs back up and comes back to normal, we are going to take advantage of all those developments that were done during the COVID period.
I’d like to add to that, I’d heard that the automotive industry was not impacted as much as aerospace, but for us being a late-stage start-up tech company, we have actually picked up a lot of interest from external customers, new customers, and investors. One point of feedback that we got was, in Europe one of our customers, their lines had to be stopped for (COVID) regulation reasons and whatnot. In the digital manufacturing platform that Divergent has, we can put minimal people on the production line and still get the same parts out and meet the timeline. If anything, the COVID case actually helped us demonstrate to our customers and investors that the new trend is using the digital manufacturing platform.
That’s fascinating, those are very good anecdotes about the flexibility of additive manufacturing which has been promoted as one of the key benefits, and clearly, COVID has been disruptive and it’s good to hear that it’s helped people adapt.
I want to thank everybody for participating here tonight. I found this very useful and I appreciate you joining me and again, my gratitude. Thank you all for the wonderful discussion.
Thank you our guests
Dr. Mathieu Brochu – McGill University
Dr. Martin Conlon – Equispheres
And our sponsor – 3dpbm