[vc_row][vc_column][vc_column_text]On June 29, 2021, we held an online webinar entitled Improving Performance and Quality in Aluminum Additive Manufacturing.
Event Overview
Increasing productivity in metal additive manufacturing will accelerate the widescale adoption of the technology in manufacturing. As the technology matures, it is becoming increasingly evident that the feedstock used in Laser-Powder Bed Fusion (L-PBF) is a primary factor driving the limits of the process.
In this webinar, learn how to improve Aluminum L-PBF productivity by leveraging the unique attributes of Equispheres proprietary metal powder.
Topics covered will include:
- The impact of powder properties on L-PBF process stability
- Key considerations for achieving productivity increases without sacrificing quality
- Achievable print speeds with different classes of L-PBF machines
Video Replay
Transcript
Doug Browse, VP Strategic Partners and Alliances, Equispheres: Good morning, thank you for attending our webinar. We have people here today from all over the world, so we’re very excited about that, and thank you for taking time out of your busy day to join us here. We know your time is valuable, so we plan to do our best to provide you with some useful and helpful information. My name is Doug Brouse, and I’m the Vice President of Strategic Partners and Alliances here at Equispheres.
I think, as most people know, Equispheres has developed a new technology for producing metal powder. The powder that comes out of this technology has some unique features and characteristics which allow our customers to get improved mechanical properties, as well as to rapidly and quite economically print with our powder. We have about 40 people here in Ottawa producing this powder. We have three reactors and we focus mostly on aluminum alloys.
We realized about a year ago that for customers to get the best value out of our powder, it was best if we delivered the powder with application engineering services, so about a year ago we hired Evan Butler-Jones to lead the creation of our Applications Engineering Division within the company. Evan is an engineer – he has an aerospace degree – and he has a broad base of experience bringing new technologies to industry, everything from welding to software. He’s been a product manager, he’s been a business development manager, he’s been a CEO, and most importantly, he spent four years at an advanced manufacturing lab at Canadore College.
So we’re very excited to have recruited Evan, and I’m going to hand it over to you, Evan. Thank you.
Evan Butler-Jones, Head of Applications Engineering Division: All right, thanks Doug! I see a few people are still joining up, so I’m gonna switch over to the topics agenda slide and give it just a minute, and then we’ll get into the details.
All right, so thanks again, Doug, and thanks everybody for joining the webinar today. Just a note, you should be able to see on the right-hand side of your screen a window that gives you the ability to ask questions and answer polls when they come up. If you can’t see that screen, there’s a little bar at the bottom right-hand side of your window that you can click, and it should pop up a menu on the right-hand side. I’ll also show a slide in a little bit that has that menu and what you need to see to turn that on.
So, getting started! What I want to start with today, and kind of summarize, is a preview of what we’re going to be talking about. I want to start with some context about what’s driving our work here, and really the goal for us is to make additive manufacturing a viable alternative to traditional manufacturing for a wider range of applications, a wider range of parts, and in order to reduce those costs, there are certainly some barriers that we need to get through. A little bit of a spoiler to the presentation, the big barrier is production throughput — so the ability to make more parts faster; the ability to reduce those reduce costs by producing more with the investment that we’ve made. And again, another little spoiler, but one of the biggest challenges and the biggest barriers to doing that, particularly with aluminum, is the stability of the process, when we move up to higher speeds and add thicker build layers. So we’re going to talk a bit about how we can achieve some of that stability with aluminum at these higher speeds, and then go through some of the results of our work and the work that we’re doing with our partners on mechanical results and surface roughness with high-speed printing and then we’ll get into some questions.
So before we get right into the meat, you have one poll question that should pop up, just to get a lay of the land, you know, read the room a little bit. What purpose is your organization using aluminum for in AM, if you are, and “not using aluminum” is an option in the poll, too. So if you can, go over to your poll window – you can see it on the right, if you don’t see the sidebar – and just select your answer. Once you submit your vote, you should be able to see the responses of the other folks on the call.
Right, so as those answers come in, you’ll be able to continue watching the poll results and see the responses. We’re seeing a number of people not using aluminum yet, but a number using it in research and development or for manufacturing production, which is excellent to see. I think we’re going to have some good questions coming into the end of this.
All right, so reducing costs in metal AM with aluminum, again to give that context, on the left-hand side of the screen we’ve got two charts on the screen. On the left-hand side of the screen is a chart that talks about what the cost premium is for additive manufacturing and where we need to get to for additive to expand in terms of the viable business cases for production parts. So if we look at the left, the axis here is the price per cubic centimetre, and the blue line is indicating the price per cubic centimetre for additive manufacturing with aluminum. That’s the cubic centimetre of printed aluminum, and that’s contrasting to the potential marketplace for additive manufacturing to make sense, for there to be a viable business case. Additive brings a lot of advantages along with it that many of us know about — whether that’s part consolidation or light weighting or other optimization strategies — but there is a very significant cost premium and if the cost premium is too high, the business case just isn’t there.
So as that cost premium reduces, as the cost per cubic centimetre reduces, more and more valid business cases become available with different applications in different parts. The years, in this case, are notional, but really the chart is saying that as these costs come down more and more, applications open up to the point where additive manufacturing with aluminum really is a mainstream option for manufacturing. So if we look on the right-hand side of the screen now, and say okay, if we need to reduce that cost per cubic centimetre, what is really driving the cost? This charge is coming from data from a number of manufacturers, including BMW, so it’s really representative of the automotive industry more than aerospace. You can reference it for aerospace too, though there is a difference in the cost profile.
But really, for any industry where additive is, what’s driving the cost right now is in the process and in the overhead of that process. What that means is the cost of the printer, the cost of operating the printer, the cost of handling all that infrastructure and managing all that infrastructure to make parts. And so, what’s really driving that is very expensive equipment, very expensive infrastructure, that can only produce a small number of parts. And so, the more parts that we can make with that investment in infrastructure, the smaller the cost per part will be. The part cost will be coming down. So our goal then is to increase that production throughput to make more parts with the infrastructure that’s been put in place. So just to give a bit of a case study here, in terms of the time and what it really means when we say it takes a “long time” to make a part, if you look at this case study here, we’ve got this topology optimized bracket putting 12 of these on a build plate.
So this is on the top right here of the screen, so if we look at this build job – and just for reference, these brackets are, in the current orientation, they’re about 20 centimetres high, so a typical print using a typical 400-watt Additive Manufacturing machine with typical parameters of 30 micron layer thickness, which is a standard offered by many of the manufacturers that offer a 400-watt machine — with these standard parameters, to make this print requires over 6600 layers. It takes about 30 seconds to scan each layer, and if you include the recoding or, if it’s not the recording type, if you include the contours and recoding time, just estimating about six seconds the total build time is 68.1 hours. Now, if we look at an improved print where we have really optimized the parameters — in this case, we’re definitely using a higher power laser but moving up to 120 microns and faster scan speed and a wider distance between each scan, that’s the hatch distance — that now brings us to just over 1600 layers. That’s a shorter scan time and a total build time of about 12 and a half hours.
So that brings us from just under three days to produce this part to just over half a day. It’s approximately a 5.4 times improvement. Now, just for clarity, and just to give a little bit more context here, this is single laser build time. So this is a machine with a single laser. There are several machines out there, production machines, that have multiple lasers — most notably the SLM NXG 12, which has 12 lasers all at one kilowatt. So, if you were to combine the benefits of these improved parameters along with some of these high production machines and bigger build volumes, this print time and this total build time is going to be improved very significantly, even more than what we’re seeing here. We might just say, okay, well, that’s great, so if we just get a higher power printer, spend more money and get a bigger laser, we’re going to be able to print a whole lot faster. But the reality is that as we move faster, and move to thicker layers, what we start to see is a real impact on process stability. So we can’t get consistent part results when we print when we move to a higher speed, or when we move to higher thickness. And really, to improve those speeds, other than adding multiple lasers, increasing layer thickness and increasing scan speed is how we’re going to get to that higher productivity with the SLM build technology. So when I talk about stability or consistency, again, a little bit of context — what we’re talking about is how much can we rely on that laser to produce a very stable weld as it’s melting each layer.
So what we do to look at what this process looks like at the microscopic scale is, we’ll take apart a part or a sample — one up here if you can see my cursor, I will slice it and we’ll take a look at that cross-section. We’re going to look, and when we look at that cross-section, what we can see is each track of the laser, what I’ve outlined in blue here. You can see, there’s lots of overlaps. They only tend to see half of one, but each one of these is the cross-section of a track of the laser as it’s building. The laser will go across — it will melt one kind of what is called the melt pool, and then it’s going to come back. You pick your pattern, but it’s going to be making these tracks layer by layer and if you’re wondering why if you haven’t looked at this type of image before, if you’re wondering why we see some odd shapes down here, it is because as the layer builds, the laser is going through different orientations. So it’s scanning in one direction and then it’s going to turn by say 60 degrees and it’s going to scan in a different direction. So as we go down through the layers, we’re seeing different cross-sections of this.
But if we just focus on the top here, where we’re seeing a real cross-section of the layers, what we want to see is very consistent weld beads with a consistent mount pull shape, a consistent depth, and a consistent overlap, and if we can get this consistent shape and depth and size, then we’re going to get a consistent overlap based on the settings that we define. Here, you can see there’s a little bit of variation, but for the most part, this is a very consistent, very good melt pool, and we’re not seeing any areas where we’re not getting that overlap between layers where there might be powder in-between. We’re not getting any really deep melt pools where we might get gas trapped inside, and this is what we want a melt pool to look like. But as we go up in speed, as we go up in layer thickness, what we typically run into is some chaos in this process, that gives us very big variations in each one of those melt pools, which leads to porosity in the part, which leads to poor part quality.
So just to look at some examples of that poor part quality, on the left and on the right-hand side of the screen here, we have two prints done with the same machine on the same day with the same parameters — one that produced a pretty good part. So these percents here are indicating the percent or relative density of the part, and this is at just under 99.5 which is not a bad density – it’s not great, but it’s pretty good. Then we have one here, again with the same parameters, it’s 97.9 so just under 98 dense, so something happened in this print to cause that inconsistency. We see the same result in mechanical properties where we have one part, that one sample that broke where we expected it to, we could see a little bit of the sign in necking. Probably a pretty good part, but it’s another sample that broke well outside where we expected it to, and probably due to some line of defects like this where it’s indicating that the part is just not a high-quality part. It’s not something that we want to put into production. So what are the causes of these instabilities?
There’s a number of reasons that we can get inconsistency in the print. The first is in the machine itself. It could be due to calibration or damaged or dirty optics or other components; we could have issues with gas flow, or there could be other process settings that may be off or may have been changed for some reason or another. It could be in the environment, whether that’s humidity or temperature swings that the machine or your facility can’t handle. It could be in the powder with moisture or other contaminants getting in, but you know as much as we that machine manufacturers are really working on improving the image for these machines, making them more robust, as much as the process operators folks like myself are working to make sure we get very good consistency in the way that we run our process. We have one other factor, which is the powder, in the inconsistency in the powder, in the feedstock itself, and this is something where Equispheres has identified a problem. This is what we’re working on here.
So if we look at that powder, and very few people actually put their powder under a microscope, but if you do put the powder under the microscope what you would typically see with aluminum powder is something like what we have on the left. We have a pretty wide size distribution, we have lots of particles that are not spherical, very rough. There are lots of very tiny particles, there are particles that are kind of agglomerated or stuck together, and what this results in is a powder that has a really high surface area. With aluminum, because aluminum is reactive with oxygen and with moisture, that higher surface area means there’s a bigger tendency to develop oxides and pick up moisture. It also really impacts the flow and some of the other behaviours that are really important in the print.
If we look at Equispheres of course, we’ve got the reason for the company name. We’ve got very consistent spheres, very consistent particles that are very tight in size distribution, very smooth, and there are not nearly as many opportunities for moisture pickup, for oxidization, and in fact, the smoothness and the sphericity of these particles also lead to better flowability and some other factors that can help produce a much better product in the bed. It’s in the powder bed itself. So just going through a little bit more detail on this, not to harp on it too long, but when we talk about the uniform size powder and fine free, what we mean is that we’ve got a much tighter size distribution. We don’t have as wide a distribution of powder that leads to that wide size distribution; lack of fines leads to, as I said, a lower surface area for the powder itself compared to the mass of the powder. The result, when we talk about moisture contamination is that — I guess I’ll give a little bit of context on this, on these charts. When we expose the powder to moisture, the reference powder tends to pick up powder and you can see on the red line here, it tends to pick up that moisture much more. So essentially, what we’re doing and what was done in this paper is the researchers exposed the powder to humidity, measured the weight gain, measured how much moisture was picking up, and then they dried that powder out and measured how much it let go of or desorbed.
So you can see with the reference powder, there’s a significant increase compared to the atmospheres of moisture pickup, and in the end, as it was, even as it was dried out the reference powder was left with significantly more moisture. What that moisture and also the inconsistency in the particles does, is it leads to poor flowability. On the left – hopefully, the video shows up okay on everyone’s screens – but you can see a very significant difference in flowability. So atmospheric powder on the right, and the traditional powder on the left you can see clumping – you can see some dust. On this dust, on the jar, and just significantly less flowability, and the importance there is what that results in — in the powder bed itself we get much less consistent spread of powder with reference powder and it is much less dense. With Equispheres, we have a much more consistent powder bed that is much denser, as it’s actually spread. So if you want to learn more about this, the work that was done here, you can see the reference here. This is really you know one of the primary factors that lead to this improved stable melting process. So the result of using this powder, along with dealing with those items on the machine and dealing with the items in your process – if we get all three of those working together, we get a very very stable melting process even at very high speeds and thicker powder layers.
So this image, this is the same image as I showed a few slides ago, but just for reference, and to give some context, this is actually a cross-section of a print that we did at 120-micron layer thickness and a melt rate of about 130 cubic centimetres an hour. This image is very close to that theoretical case study that I presented, in terms of the improved powder parameters. In fact, as we’ll go through we’ll see some parameters that show that we’re going even faster now. So great, we’ve talked a little bit about stability, we’ve talked about how the powder along with the other factors in the machine and your process can affect stability, so what are the results? How fast can we go? What can we actually achieve? So this is a picture of our printer in our internal lab here at Equispheres — just to give some context, it’s an Iconity Mini 1-kilowatt laser — it’s very obviously not a plug-and-play machine like you see from many of the industrial, big machine manufacturers. This is very good for research, in that it allows us to tune a lot of different parameters and a lot of different factors. We also have partners testing on systems including other economy 3D machines – EOS, Renishaw, SLM solutions and Trump — so we’re not just limiting our work to a single machine, but we’re looking across kind of the span of the OEM offerings, to see whether this does hold, do these results hold, across machines, across laser powers, et cetera.
I’ll pause here, and just remind folks to add questions into the side panel, we will take some time at the end of the presentation to go through as many questions as we can.
So this chart summarizes the results of the work today, and what it’s showing is a relationship between the density or relative density. We’ve talked a bit about relative density before — relative density of the parts as a function of build rate — and that is then related to the power that we’re using. So when I talk about build rate, I just want to step back and take a minute to describe build rate, because it can be confusing. There’s a lot of things that impact build rate in the SLM process, and because there are so many things that affect build rate including your recoder time and the orientation of your part or the number of lasers on the machine or the powder bed size, all those things can affect your build rate. So what we do to kind of normalize across that is talk about kind of a theoretical build rate or melt rate which is simplified down to if one laser is working, and one laser is melting powder, how fast can that one laser – getting rid of all the other factors – melt the powder? So we calculate that by multiplying the speed that that laser is travelling, times the hatch distance or the distance between each track that the laser follows, multiplied by the layer thickness — how deep that layer is, that the laser is melting. So that gives us that volume of cubic centimetres per hour. So not to shortcut or ignore things like recording time, but to normalize it across different printers, different print jobs, et cetera.
So what we can see here is, we’ve cut off the results at 99.8 density, and if you recall the slide that I showed earlier that was about 99.5 percent dense, but we really want to target a very high density for very high-quality parts, so we kind of set our cutoff at 99.8, and then looking at what build speeds, what build rates we’re able to achieve. So with a typical 400-watt machine, that 30-micron print example that I used in the slide up at the beginning, that build rate is again a typical approach at 30 microns and that’s about a 20 cubic centimetre an hour build rate as we move up to a 60-micron build parameter that some manufacturers offer that tends to get us between 30 to 40 cubic centimetres an hour. As we continue to push those parameters, because we’ve got that stable build, we can push that a little farther. We’re seeing that we’re able to get to around 50 cubic centimetres an hour with a 400-watt machine very very consistently and reliably now. If we move up in power to a 700-watt machine, we can see that that gives us the opportunity to move much faster now. We’re moving from your 20 to 40 cubic centimetres an hour, up to well into the 80 to 100 and even above 100 while still achieving that 99.8 and above percentage. And that density is really important because as I say, you can go faster, you can make the laser go faster, but you can see that decrease due to that lack of stability. You can see that decrease in density. So we want to get that speed while still achieving high density and good mechanical properties as we move up into the even higher-powered lasers. We’ve got 800 watts, 900, 950 watts, and we can see that we’re now getting into a range where we can achieve well above that. Here we’re looking at a little bit above 150 cubic centimetres an hour and these build rates are actually the example that I used in the early case study.
So just for those who might be wondering why we have such a tight distribution here of build rates, a lot of this is just the work that’s been done so far, and a lot of the focus has been on the laser focus or the size of the laser beam. That size of the laser beam that we change doesn’t affect the build rate but it does affect the part quality, so each one of these dots represents a different parameter set. It’s not multiple results from the same parameters, yet. So getting into a little bit more of the details here, showing some of the actual mechanical results as well as the densities from a 400-watt laser. Again, these results, just for reference as we get into mechanical results, are with ALSI 10 mg. ALSI 10 mg is probably the most common aluminum alloy used in SLM today. It’s a reference that we use – it’s a product that we offer, but it’s a reference that we use so that we can have good comparisons and have comparisons to what our customers are building with today. So 37 cubic centimetres an hour and 60-micron layer, these images are showing density. We’ve got a 99.96 percent density, and we’ve got excellent mechanical properties at yield strength at 260 MPA or 37 KSI, and very tight standard deviation. Again, that stability in the process, that stability in the mill process means that we have a much tighter distribution of our mechanical results, then we move up in speed and up in layer thickness again with a 400-watt machine. We’re now up to just under 50 cubic centimetres an hour and again above 99.9 density, very good density and very good mechanical properties, and again, very tight standard distribution.
Now moving up to the higher power, this is at 83 cubic centimetres an hour and now we’re at 100-micron layer thickness, 99.91 dense; at 103 cubic centimetres an hour, 99.84 dense. These are still well above 99.8 and again, very good mechanical properties, very tight distribution in the tensile range. These are all the z-axis, so again for those who don’t know, I should have pointed that out earlier, the z-axis is the vertical direction. So these are built in the vertical direction, kind of up from the build plane.
So then, moving up further again to using almost a full one-kilowatt laser power, we’ve got above – at 128 cubic centimetres an hour – above 99.9 percent dense, above 100 or 152 cubic centimetres an hour. We’re at 99.86 percent dense, and again very good mechanical properties a very tight distribution.
So taking all those slides together and just kind of summarizing the yield strength, the elongation versus build rate, what you can see is, on the left-hand chart, is yield strength versus that build rate again. That’s the single laser build rate again, the build rate with the multi-laser system is — the actual build rate with the multi-laser system would be would be quite a bit faster. We’ve got our yield strength trend, this dashed line on each of the charts represents the ASTM as a printed target, so it’s not that everybody uses that as a target but it is a good reference that other folks can reference. So we can see that even up at greater than 150 cubic centimeters an hour build rate, we’re still hitting those targets, both at yield strength and elongation, again with a very tight distribution on the yield strength.
I’m going to talk a little bit about the contours and surface roughness that we get. Because we’re talking now at a 120-micron layer thickness, so one of the obvious questions that comes up is, what does that do to our surface roughness?
But before I do that, I’d like to have another poll question. Just what laser powers are available on the machines that you have at your organization, if you have machines at your organization currently? “No printers” is also an option. So again, if you go over to the right hand side of the screen you should see a poll. We can go back and look at some of the previous responses as well. So we’ve got great a number of people making production parts out of aluminum and some in research and development, that’s excellent.
Now we’ve got a good range, so there’s a good range of powers available as well, so great. Hopefully everybody can see the results of those polls on the right-hand side. Once you’ve submitted your answer you should be able to see what everybody’s answered.
So as I mentioned, I want to talk about surface roughness, because my engineer’s brain says, great, we’ve hit some density but are we able to get reasonable surface properties at these thicknesses, in these speeds? The answer so far is yes. So on the left-hand side again, we see another one of these cross-sections where we’ve taken the sample and sectioned it to take a look at what our melt pool looks like, and this is now looking at the side skin. So the orientation of this image is the orientation that it was built on the plate, so that the vertical surface is a cross-section of the actual vertical surface that we’re seeing, that was printed on the plate. Again, we can see very good, consistent weld beads here as we go through the layers. We can also see on the side skin here very consistent melting, a very smooth surface. We had both a contour and a border parameter, so we can see that we’re filling in that gap between the main bulk very well; we’re not seeing any porosity, we’re not seeing any gaps, and then this border parameter comes in and it again fills that in. We’re not seeing many particles conglomerating to the surface, so again, very good, this is what we want to see. This is a picture of a print taken at 120-micron layer thickness so just for reference as we go up in build rate and go up in layer thickness, we’re not seeing a significant increase in surface roughness. Now I believe we could do better at these lower speeds. In fact, I’ve seen results much much better. But still, we’re sitting at about 12 and a half micron surface roughness measured on the RA scale, so we’re pretty happy with that so far.
All right, so just to go back and summarize the flow of the story here… So along with controlling for the machine, controlling for the environment, controlling for your process, your process having that proper feedstock and controlling that feedstock and using feedstock that’s designed for additive with the proper properties gives us better processing characteristics, like flowability, less moisture, better spread density, and more consistent spread density. The result is that we got much more consistent melting and solidification behaviour and the result of that is that we’re able to move to much higher speeds. We’re able to take advantage of the high powers and the big layer thicknesses because we’ve got a stable melt pool and we can continue to build prints that are very consistent, very strong, even at much higher speeds. So back to the case study where we look at the reference print versus this improved print, the parameters are in fact the parameters that I showed at the upper end of what we’ve tested, what we’ve run so far. And we’re seeing, so just you know contextually when we print an equivalent part or say tensile samples at our original 60-micron build — that’s not even 30, but 60-micron build versus you know 120 in the higher speeds. We’re moving our prints down from, say it’s a 12 and a half hour print, down to three hours or less. So this is a 5.4 times improvement, depending on your starting point, and it is actually achievable and something that we’re seeing. What that means for the cost in SLM is significant savings by allowing for the printers to produce more parts with the same investment and making sure that we’re actually able to distribute those costs across those parts. So we’re getting into that curve where we can actually see additive being much more competitive, and the business case for those parts expanding, and the number of applications expanding. Once again, just a reminder that’s your single laser build time; when we talk about bigger beds, multi-high powered lasers, we’ve got even more potential that gets added into this. So I talked about this a bit, I’m not gonna read through the text again, but I think what we’ll do now is move on to questions.
Doug: Thanks Evan. We’ve got a lot of good questions here, so in no particular order here, I’ve got to do them in themes as best I can. So the first question we have is, Equispheres is focusing only on ALSI 10 mg or other alloys as well, and are you only focusing on bed fusion, or are you looking at other processes such as a directed energy device? Somebody else also asked about binder jets. That’s a multi-part question for you.
Evan: Okay, great. To answer the first question of are we only working with ALSI 10 mg, the answer is no. ALSI 10 mg is just the most common alloy that’s currently used, so it’s something that we use as our reference. We do produce ALSI 10 mg, we also produce other alloys. We do focus on aluminum, but we work across a spectrum of alloys of both kinds of traditional casting alloys and others, working with partners to develop and produce specialty aluminum alloys as well. If you have any specific questions on which alloys, or have an interest in a specific alloy, please reach out after and we’ll be happy to talk about it. In terms of the question about other technologies, so while this presentation is focused on SLM, we are looking at and we have worked on other technologies, in fact, cold spray being one of them, directed energy binder jet absolutely. So one of the challenges the binder jet has is that it’s very difficult, if not impossible, to sim cincher typical aluminum powders. One of the unique features of Equispheres’ powder that has been discovered and that we’re continually developing is the ability to center, and so we are working with a number of the binder jet manufacturers to bring to market offerings in aluminum for a binder jet. Hopefully, that answered both those questions.
Doug: Excellent. So another themed question here is, one question asking why we stopped at 950 in terms of the power on the laser — why not higher? And another question said that many of the advantages that we seem to be seeing are as a result of the laser power, and how much of the increased build rate is really from the powder, and can other aluminum powders achieve these speeds?
Evan: Right, okay. So the first question on why not higher than one kilowatt. It’s a very good question, and I would love a more powerful laser, but really what it gets down to is that we want our work to be applicable to our customers, and for the market, and right now for SLM machines. And when I say SLM, I’m not just talking about SLM solutions, but the technology of powder bed fusion laser, or also known as selective laser melting. The largest laser typically offered is one kilowatt, and so doing research into higher and higher powers is something we want to get to, but it’s not currently in our focus and so we will be expanding into that in the future for sure. Then the other question on, is achieving those rates possible with any powder, is it just the laser power or is there something to do with the powder as well? The short answer is it absolutely has to do with the powder as well as the machine. Both of those work together, but what we’ve seen and what many of our partners have seen is that if you get beyond about 80 microns layer thickness, in some cases 90, but about 80 microns layer thickness you really start to see the process become unstable, and you end up with porosities that you just can’t eliminate. You get into a state where the variation in your melt pool is causing different types of porosities that normally you can correct for by tuning one way or the other, but you start seeing both types of porosity at the same time so you’ve kind of run out of room to build. So absolutely, improvements in the machines and the laser power are pushing the boundaries now, and more companies offer 60-micron parameters versus 30-micron parameters, and some are offering 90-micron parameters or 100 for prototyping. But for actual production parts what we’ve seen so far, and I’d love to see research to the contrary, but what we’ve seen so far is that you really need to have that combination of the machine and the powder to achieve these speeds and get that stability.
Doug: As a quick follow-up, which is a question here, just to clarify — are the laser powers mentioned the actual powers used, or what the capacity of the laser is?
Evan: Right, I would say it’s the powers used in many of the tests. In some cases, it’s the capacity of the laser, but a 400-watt laser usually only runs at about 370, 380, sometimes 390 watts. So if we’re working with a laser that is a 400-watt laser, we’d be working at 370, 380, but we also have as I pointed out in one of the slides, we have a one-kilowatt laser at our lab. Then we will tune it down to 370, 380 to represent what the capabilities of a smaller laser would be. So what is being shown here is the power capability of the laser essentially.
Doug: That makes sense. Okay, so we have a few more themes here that we’d like to get to. One of them being we have a number of questions about powder reusability. Can you comment on how reusable the powder is?
Evan: Yeah, great question. So in fact we’re in the midst of a collaborative study with a couple of academic institutions in Canada. We are doing a direct reusability study. The results we’re expecting a little bit later this year, but anecdotally and from the work that we’ve done, we are seeing an increase in reusability, because again there are a few factors in there, but that lack of moisture absorption or that reduced tendency to absorb moisture has a significant effect. The consistent particle size as well — we don’t see that particle size-shifting with reuse to the same degree that we see otherwise. But that is all anecdotal. I guess for reference, all of the studies shown here were printed with reuse powder, not with fresh powder. But we are going to have more comprehensive data coming over the next little while so I can give you a better quantified value soon.
Doug: We have a couple of questions here about pricing, particularly relative to gas atomization. Can you comment on that?
Evan: We offer a powder at typical market rates. We absolutely recognize that the powder price is part of the equation for reducing the cost of the technology and so right now we offer our powder at market rates. I would add to that that we’re just installing now a new high capacity, high volume reactor so our anticipation is that our cost production is going to go down significantly throughout 2022, and our plan is to drive powder pricing down so that manufacturing becomes more cost-competitive. So we are at market rates right now but driving down to lower prices moving forward.
Doug: There are questions about the layer thickness increase and on the overhang angle that can be achieved.
Evan: Right. That’s a very good question and I’ve only shown kind of the the 90 degree side skin here. Some work that we’ve been doing with our partners showing that in fact again, even at higher layer thicknesses, we’re getting very good density, we have not yet fully quantified all the parameters and all the layer thicknesses with the various down skinned angles. That’s work that’s ongoing now and unfortunately, some of that work is not publicly available yet. But it is a very good question, and I’d be happy to follow up later with the asker of that question when we can talk in more detail about some of that work.
Doug: Okay. So I’m just on more themes — a few people asking about the technology to produce the powder, so I think we should probably speak to that a little bit.
Evan: Yeah, so I did not talk about our atomization technology, that’s correct. Folks noticed that! So our atomization technology is a proprietary technology. It’s not for those who are following atomization technologies, it is not gas atomization, it is not plasma atomization. It is proprietary atomization technology developed by Equispheres, and that is not something that we share or discuss, only under proper NDAs and with proper agreements. That’s something that we could potentially get into more detail on.
Doug: Okay, I think we’re running out of time here so time for two more questions. The question here about the narrow distribution — what technology of recoder is most adequate roller, polymer recorder or hard blade?
Evan: Right, yeah, that’s another good question. So we have seen success with a number of different recorder methods. I’m not certain of all the recoder technologies on all the machines but generally, we used brush blade polymer. We have seen success with all of them. We haven’t done a direct one-to-one comparison on the same machine — for example, with a hard blade versus a carbon brush versus a polymer recoder or a roller. That would be very interesting to see. But so far we’ve seen very consistent spreading with with all the recoding methods that we’ve tested, I’m going to take note of your question because it’s an interesting thing to follow up on.
Doug: Yeah, okay. I think this is our last question, and we get this a lot. Your website claims that you have three powder lines – performance, production, precision. What’s the difference, and what was used for this research?
Evan: Sure. So the powder that was used for this research was our performance powder, which is kind of our middle-of-the-road powder in terms of size and kind of the most robust powder for allowing you to get fine features as well as high productivity. Our precision powder is, among other things, a finer cut of powder, so smaller particles again, very tight distribution, but smaller particles and that allows us to get very fine features and allows us to go do some work down at the thinner layer thicknesses. So if we’re not looking for speed but we’re looking for very fine features, whether that’s a thermal management application or other applications where we want very fine fines and features, that precision powder is targeted at those types of applications. Although, we still can get speed increases, but it is targeted for those fine features and then our production powder is kind of designed purely for high-speed production. So probably, in that case, you’re going to be sacrificing some of the very fine feature resolution, but allowing you to really take advantage of the high speeds.
Doug: Excellent. Well, I think we should probably wrap it up there, Evan. I want to thank everyone for participating today. I hope you found this useful and that you’ve got some good information. Feel free to reach out to us if you have any more questions about better powder and its performance, and how it can help you. We will be providing a link to this webinar on our website, so you can watch it at your leisure again if you’re so inclined, but thank you, everybody. Unless you have any more parting words?
Evan: Not for me, thanks, everybody!
Doug: Thanks for all your questions, I hope you enjoyed it![/vc_column_text][/vc_column][/vc_row]