00:00Thank you very much for the introduction and in this presentation I want to talk about
00:13the laser polishing of 3D printed plastic components.
00:18And to start off with a small introduction, I want to briefly mention currently popular
00:26surface finishing methods for 3D parts. You probably know most of them, if not all, like
00:32the vibratory grinding. Manual grinding polishing is also quite common for the 3D printing of
00:39plastic parts. Chemical etching, sandblasting painting are also common. Now they are commonly
00:46used but some of them or all of them have typical deficits. I want to talk about a few
00:54of them, like some of them might introduce foreign particles into the surface that might
01:01have to be cleaned afterwards. Others have rather low reduction in roughness, especially
01:08sandblasting for example. Others have higher costs, especially in manual processes or they
01:17might change slightly the geometry with edge rounding or have low selectivity, like the
01:24vibratory grinding. Now, those are typical deficits, not all processes have all of them obviously,
01:32but that's the reason why we want to develop a new post-process that might be able to solve
01:42some or all of the deficits and that's the laser polishing. Because it's probably not known
01:49to everyone, I want to explain the process principle because this is quite different from classical
01:56approaches. So we have our laser source coming from the top here in this sketch onto the material
02:08surface, then it's absorbed in a thin layer on at the surface, heats up this thin layer up until it melts.
02:17And in this melted state, the surface roughness can sort of flow out through the surface tension.
02:25Now the sketch is rather applied for metals because here you have a sharp transition between the liquid
02:33phase and the solid phase. For glass and plastics this transition is not that sharp obviously,
02:40but the process principle is still the same. On the right hand side you can see a few examples where this
02:48technology is already used for glass materials, primary for optic manufacturing, also metals and plastics
02:56are possible to polish with this technology. Today I only want to talk about plastics, of course,
03:06and I want to briefly explain how the process strategy is for those parts. Now for plastic components,
03:17we typically use a CO2 laser because with its 10.6 micrometer wavelength, it's absorbed very near the surface,
03:26surface. And then we guide this laser beam into a galvanometer scanner. There are two mirrors that can
03:35move the laser spot across a surface. The movement of this scanning strategy looks like this. So we have our laser spot
03:48that's moved in a b-directional way across the surface and actually the whole polished area
04:00is processed at once. So we scan across this polished area with a very high scanning velocity, high track
04:07distance and also a defocus laser beam to heat up the whole polished area as homogeneously as
04:18possible and at once. Now if we would use simply a constant laser power, we would heat up the surface
04:28more and more. So we need some kind of control mechanism and what we do is measuring the surface
04:34temperature with a pyrometer and then we have a closed loop control that controls the laser power
04:42power to keep the surface temperature of the polished area at a constant level. And with this we reduce the
04:52process to its two main process parameters, sorry, which are the process temperature and then the polishing time,
05:03basically. That's the time how long you keep the melt pool, yeah, hold this melt pool at the surface.
05:10Now to verify how homogeneous the temperature actually is, we can look at the actual measurement of the
05:19pyrometer. You can see here on the graph in the red line here is the set point temperature, so this
05:28should be the polishing temperature. In black you can see the measured temperature of the surface and you can
05:35see after a short heat up phase the temperature control kicks in and controls the laser power here in blue
05:45to keep the actual temperature at the surface at a constant level. And this can be hold now for an
05:50arbitrary amount of time as long as it's needed to reduce the surface roughness. On the right-hand side
05:59I have a short video of a thermographic camera where you can see in the first few seconds this heat up phase
06:06and after this heat up phase the temperature distribution, in this case in a 10 by 10 millimeter field size,
06:13is rather homogeneous. And actually it is similar to a temperature distribution that you wouldn't
06:23would achieve if you would use a top head intensity distribution of your laser source and that's why we
06:33call this quasi top head strategy. Now this simplification of the process leads to a very simple experimental design
06:48because you only have two main process parameters temperature and time. So you can set up an experimental
06:54design to get the influence of temperature and time on the resulting surface roughness.
07:00And on the right-hand side you can see the example for an SLS printed PA12 sample. The initial roughness was about 16 micrometer.
07:13I show it on the next slide in more detail.
07:17But you can see from this graph here on the x-axis the temperature and the y-axis shows the interaction time,
07:25polishing time and you can simply see a few things. If the temperature is too small obviously the melting
07:33temperature is not reached and nothing happens. If the temperature is too high then well a thermal degradation
07:41can occur. In this case for this material a colorization occurs so that's not the desired parameter range.
07:50But then you have your desired parameter range in this case at 210 degrees and for 200 seconds interaction times
07:59you get the lowest roughness here. Obviously from this graph it's easy to interpret that for longer
08:06interaction times the roughness is even further reduced which is actually the case but not included in this graph here.
08:14And now this was all a little bit theoretical. Now I want to show an actual picture of such a surface.
08:25You can see here on the left-hand side a 20 by 20 millimeter test field on an SLS printed PA12 sample.
08:36And in the center here you can see a white light interferometer measurement of the initial state. So after the printing,
08:44before laser polishing, there we have a roughness of about 60 micrometers, typical for 3D printing.
08:53And after the laser polishing, well on this scale the surface basically flat. The left roughness in this
09:01is in this case about 0.3 micrometers so far below 1 micrometer which is most of the time some kind of goal point.
09:12Now that I talked about the process principle and showed an example where we want to polish or we want to use the laser polishing
09:24for typical 3D printing materials. And then the first question arising is which materials can actually be laser polished.
09:34And it's quite easy to answer. The general requirement is that the material has to be thermoplastic because it needs a melting state.
09:45But that's basically all of the requirements. Some might need a drying step before the polishing to avoid bubble formation or something.
09:57But then with this shown quasi top head scanning strategy, you can easily find a suitable parameter set
10:06that at least shows the processability of a new material. And here at the bottom, you can see a few materials that we already tested.
10:18Most of them are 3D printed with either SLS or FDM technology. And on the right hand side,
10:25you can see a few examples of surfaces. Here TPU, polypropylene and peak. Now the roughness values here, I have to mention this,
10:38are measured in a 1 by 1 millimeter area in the center of the test view. Now the next question after we have a material that is processable
10:50is of course, which geometries can actually be laser polished. And I want to split this into four categories or four types of geometries.
11:03Starting with a simple small two-dimensional area. With this quasi top head strategy, we can simply apply this onto the surface.
11:14Even if it has a different shape than the square that I showed before. Because of the scanning strategy, we can actually make arbitrary shapes of the resulting melting pool basically.
11:32So that's the easy case. It's just stationary polishing. Now if we want to go to larger areas.
11:41So this starts at around 30 by 30 millimeters typically. Then we come to the limit of the stationary process. And we have to move this quasi top head field with, for example, a mechanical axis across the area.
11:59This might look like this. So we have here in the dark orange color, this quasi top head field.
12:06And move this across the larger area with a certain feed velocity to polish a larger area.
12:19Now, because of the scanning of the quasi top head field, we can also change the shape of the quasi top head field to
12:29consider slight changes in geometry or changes in the processed area.
12:39Now, the next step would be what we call a 2.5 dimensional structures.
12:47This is basically a 2D structure with an 8 profile.
12:52This actually doesn't need much adjustment of the laser process, as you can see in the following picture here on the left hand side.
13:03So curvatures of 30 or 30 degrees can actually be processed without a significant adjustment of the process.
13:13But then, of course, we are talking about actual 3D printing.
13:21If we have a real 3D part, we need some kind of rotation or part handling system like a 5D, a 5D axis system or something to rotate the
13:35part. So every side of the part can be accessed by the laser.
13:41Now, this has one limitation. It is important to mention though, hidden areas where the laser spot can't reach the surface of the material cannot be processed.
13:57Okay, here on the right hand side, you see a few examples of peak material.
14:03Okay, now I want to briefly summarize what I talked about.
14:12So I presented a new approach for the 3D printing surface or the surface finish of 3D printed parts, which is laser polishing.
14:21And this brings some really good advantages.
14:27The first one is that the process is completely contactless and it is non-abrasive.
14:33So it's a remelting of the surface and you don't actually remove surface material.
14:40This has also the advantage that it might be able to close pores from the rather high porosity of the surface.
14:51And because we use the laser, we have a 100% digital and completely automatable process, which really well fits into the idea of the digital manufacturing
15:04or industry 4.0 and you know all these passwords.
15:11The limitation is that we have to have thermoplastic materials.
15:17But here the advantage is SLS parts or FDM parts, for example, they are always thermoplastic materials.
15:26The roughness reduction is typically a factor of at least 20, sometimes even a factor of 100.
15:36So a roughness below one micrometer is commonly used, in some cases even 0.1 micrometer.
15:45But I have to say this is a research or this process is in a research state and there's actually not an industrial application yet.
16:00But this brings me to the outlook in the near future.
16:04The next step would be applying this process to an actual business case or an actual used case for a real 3D part.
16:12Also, what's surely always interesting is looking at the combination of different processes
16:23and maybe integrate the laser polishing into an already existing process chain to further reduce
16:30roughness, reduce polishing times or finishing times and so on.
16:36Okay, well that's it for me and thank you very much for your attention.
16:42Thank you so much for having me.
16:44And first...
16:45I'm ready to
Comments