EFFICIENT SELECTIVE LASER MELTING OF ALSI10MG0.5
Sabrina Koss1, André Edelmann2, Ralf Hellmann3
1 M. Sc., research associate
2 Dr.-Ing., Head of center of additive manufacturing
3 Prof. Dr., Head of applied laser and photonics group
1 2 3 Applied laser and photonics group, University of Applied Sciences Aschaffenburg, 63743 Aschaffenburg, Germany
ABSTRACT
Although selective laser melting is one type of rapid-prototyping methods, the manufacturing of a part can be quite time-consuming due to the layer-wise process.
By using alternative laser sources with higher powers a strong increase of the build- up rate of the process is possible when laser and scanning parameters such as hatch distance, point distance and exposure time are adjusted. In this study it was possible to use high scanning rates while obtaining a relative density of more than 99% with comparable or even higher hardness values as for conventionally fabricated parts.
MOTIVATION
Additive manufacturing is a promising technology that has been developed and improved over the last decade. The layer-wise build-up design enables the production of complex geometries, e. g. light-weight structures which cannot be processed by using conventional methods. Different materials like plastics, ceramics and metals can be processed additively by different methods.
Selective Laser Melting (SLM) is a process using a metallic powder as base material and is applicable in medical, automotive and aerospace industries, as well as in the field of tool construction [1, 5]. Due to the layer-wise fabrication mode with layer thicknesses of 20 to 100.µm, SLM is a time-consuming process. Time is an important factor to control, especially in the industrial branch as it directly correlates with profit. In order to minimize the production time in the SLM process it is e. g.
possible to use a multi-laser system where two or more lasers scan simultaneously [2]. Another approach is the usage of a high power laser which enables higher possible scanning velocities at constant energy input [3].
The following research focuses on the optimization of the process time by using laser powers of up to 1 kW. The influence of layer thickness, scanning velocity and scanline spacing where investigated before [3]. The next step will be a study with even higher scanning velocities in order to accelerate the production further. The aluminum alloy AlSi10Mg0.5 is chosen as it is a potential material for light-weight structures. It is processed with different sets of parameters followed by an evaluation of the resulting porosity and hardness. After analysis of the results it will be discussed whether the resulting material properties justify the higher process speed.
MultiScience - XXXI. microCAD International Multidisciplinary Scientific Conference University of Miskolc, Hungary, 20-21 April 2017
ISBN 978-963-358-132-2
DOI: 10.26649/musci.2017.020
SELECTIVE LASER MELTING
Selective laser melting is a powder-based additive manufacturing method. The process is depicted in Fig. 1 and can be characterized as an iterative process of two basic steps: In the first one the building platform is covered with a thin powder layer which is applied by a wiper. In the second step a fiber laser scans the particular pattern of the current layer in the powder bed and melts the material which solidifies within fractions of seconds. Subsequently, the platform is lowered in z-direction by the defined layer thickness. The chamber is flooded with argon gas to prevent oxidation.
Fig. 1
Process of Selective Laser Melting
Excess powder is not lost but can be recycled after adequate sieving which is especially beneficial for high-priced base materials. There is a great variety of materials which can be processed by selective laser melting: Particularly weldable materials such as nickel or aluminum alloys, titanium, gold, steels, or Co-Cr alloys are used, but it is also possible to handle hard-metal alloys as WC-Co [4].
Many process parameters such as gas flow, part orientation or temperature influence the resulting building part properties, for instance density, hardness and mechanical characteristics. Further parameters which directly influence the building rate and thus the total process duration are illustrated in Fig. 2.
Fig. 2
Time-influencing parameters in SLM power
layer thickness speed
point distance hatch
distance
laser scanner
building part
excess powder wiper
powder tank
platform powder
The process time is among others dependent on the spacing between the single melt traces in a plane which is called hatch distance hd. The point distance pd as well as the exposure time texp of one single laser spot define the scanning speed vscan
according to (1) which also contributes to the total building time.
vscan = pd
texp (1)
Furthermore, with a higher layer thickness Dl an increase of build-up speed would be possible, however, at the cost of precision. Another parameter which directly enables higher building rates is the laser power PL. With a higher power an equal energy input is realizable in a smaller time span which allows higher scanning speeds while the bonding in between the single laser spots is still strong. Subsequently, a loss in precision is avoided compared to saving time by increasing the layer thickness.
The build-up rate of an SLM-process can be determined according to (2), where 𝑣̇ is the building rate, Dl the layer thickness, vscan the scanning speed and hd the hatch distance [3].
v̇ = Dl ∙ vscan ∙ hd (2) EXPERIMENTAL PROCEDURE
For the executed experiments a Realizer SLM-300i machine with a maximum laser power of PL = 1 kW is used. The building chamber has dimensions of 300 x 300 x 300 mm³. The chamber is flooded with argon gas and the platform is heated to a temperature of 100 °C to obtain a better binding between the built parts and the platform.
The used powder material is AlSi10Mg0.5 (TLS Technik) with a grain size of 10 to 63 µm, a theoretical density of 2.67 g/cm³ and a hardness of about 80 HB with reference to the cast material. The optimization of the parameters point distance, hatch distance and scanning speed which were introduced in the above mentioned paragraph is performed by a step-by-step adjustment while varying one parameter at a time. An adequate point distance is necessary for receiving homogeneous melt tracks. As the melting point diameter is dependent on the used laser power this value is adjusted in the first step for laser powers of 0.5, 0.7 and 1 kW. In the subsequent step the maximum hatch distance at which it is still possible to obtain a continuous bonding of the single melt traces in the plane is determined. As an increase in production time is the goal of these investigations this is only performed with a laser power of 1 kW. The same holds for the determination of the shortest exposure time of the laser beam on a spot which directly correlates with a higher scanning speed.
The samples are printed in cubic forms of 10 x 10 x 10 mm³ each for easy removal, post-processing, measurement and evaluation (see Fig. 3).
Fig. 3
Cubes on the building platform
After printing five replicates of each variation in order to assure reproducibility, the density of all samples is determined by buoyant force method according to Archimedes. The obtained data is confirmed by metallographic preparation of the samples containing grinding and polishing. In the microsection of the cubes not only the porosity of the sample can be determined but also the amount of non-molten powder in the sample indicating insufficient energy input. In the end the Brinell hardness HB2.5/31.25/10 of the samples is measured and compared to the theoretical value of the alloy.
RESULTS AND DISCUSSION
Adjustment of the optimal point distance for different powers
As the melting spot diameter depends on the energy input per time, it is also dependent on the applied laser power. In the following, the optimum point distances are determined by producing single melt traces in the shape of frames with an edge length of 10 mm using laser powers of 0.5, 0.7 and 1 kW and point distances of 100, 150, 200, 250, 300, 350 and 400 µm. The special frame shape in comparison to sheets is chosen to confirm perfect melting behavior and to exclude tear-off in case of curved shapes.
Selected results are depicted in Fig. 4. Good results are characterized by continuous melting tracks with homogeneous width, results are stated as poor if the track tears off during one line. For PL = 0.5 kW (Fig. 4 first row) a point distance up to 150 µm would be possible, although pd = 100 µm produces notably better results.
With an increase to PL = 0.7 kW (Fig. 4 second row) the optimal point distance of 100 µm stays consistent with the lower power. Only at the maximum power of PL = 1 kW (Fig. 4 third row) a higher point distance of 150 µm is possible without loss of quality.
The increased point distance from pd = 100 µm to 150 µm is equivalent to a time gain of 50% which already increases the build-up rate from 4.7 mm3s−1to 7mm3s−1.
Fig. 4
Overview of different point distances at various powers Determination of optimum hatch distance for 1 kW
A great time-saving potential can furthermore be found in adjusting the hatch distance towards higher values. The distance has to be selected in a way which still guarantees a complete fusion of the tracks in the plane. In order to investigate the 2- and 3-dimensional bonding, the frame shape is replaced by a filled cubic sample form.
Furthermore, variations of the hatch distance are only performed for 1 kW power due to higher possible point distances compared to 0.5 and 0.7 kW. An extract of macroscopic results can be found in Fig. 5. For hatch distances hd < 0.2 mm the energy input is so high that the material starts to deform under thermal stress. Some samples showed a severe distortion in such a way that they had to be removed from the building process to avoid damage of the wiper. The surfaces using larger values of hd = 0.2-0.3 mm (Fig. 5 left) show a good joining behavior of the single melt tracks and can be considered as potential values for the hatch distance. For hd > 0.3 mm (Fig. 5 middle) the bonding is getting worse up to a complete loss of linkage. For values of hd > 0.5 mm (Fig. 5 right) even the formation of a lattice structure can be observed which indicates a way too high hatch distance value.
By setting the value to hd = 0.3 mm the distance between two melt tracks can be doubled compared to standard settings. Therefore, the build-up rate increases from 7 mm3s−1 to 14 mm3s−1 which indicates that the scanning time during the process can be reduced by 50%.
0.5 kW
0.7 kW
1.0 kW
100 µm 150 µm 200 µm 250 µm
Fig. 5
Comparison of different hatch distances at 100 µs exposure time Investigation of different exposure times
On the basis of the previous results the last time-influencing variable, i. e. the exposure time can be adjusted. For a constant laser power of 1 kW and hatch distances of 0.2, 0.25 and 0.3 mm the exposure time is varied between 50, 100, 150 and 200 µs.
The conversion between exposure time and scanning speed can be found in Tab. 1.
Tab.1
Conversion exposure time and scanning speed Exposure time/µs Scanning speed/(mm/s)
50 3000
100 1500
150 1000
200 750
This is followed by an investigation of the surface quality and deformation behavior as well as a microscopical inspection of the microsection regarding porosity.
For an exposure time of texp = 50 µs – which results in the highest scanning speed – no continuous melting tracks could be observed which results in a very rough surface and, hence, a low density. When exposure times are set to values texp > 150 µs the thermal input is too high due to the small scanning speed, similarly to the small hatch distances. The result is again a strong deformation of the parts as well as a dark tarnishing of the surface which indicates an excess energy input. The surface quality of exposure times of texp = 100-150 µs is comparably good with sufficient bonding and few spatter particles. To decide which parameter set is the one to choose microsections are prepared and investigated, see Fig. 6. As expected, at high scanning speeds (Fig. 6 left) the powder is not fully melted and large pores with unmolten powder are visible all over the samples for different hatch distances. For low scanning speeds (Fig. 6 right) large gas pores with diameters of more than 100 µm occur. Gas porosity can be found in every sample while the amount is minimum for an exposure time of 100 µs (Fig. 6 middle). As the pores appear using completely different sets of parameters it is conceivable that not only scanning parameters contribute to the porosity but also the powder quality.
0.3 mm 0.4 mm 0.5 mm
Fig. 6
Metallographic structure using different exposure times at 0.3 mm hatch distance Aboulkhair et al. [6] state that there is a correlation between gas porosity, humidity and subsequent oxidation of the powder. As humidity can increase during storage and recycling of the powder this could be a possible reason for the overall gas porosity.
Density and hardness examination of different parameter sets
To finally decide which values are appropriate to use at very high build-up rates it is not sufficient to evaluate the surface roughness and microsections. Therefore, the density is determined according to Archimedes for samples with a hatch spacing of 0.25 and 0.30 mm and exposure times of 50, 100, 150 and 200 µs at 150 µm point distance and a power of 1 kW. The trend shown in Fig. 7 confirms the metallographic results: due to the high share of unmolten powder and gas porosity in samples of high and low scanning speeds these samples have low relative densities of < 99%. The only sample with a density of 99.4% relative is the one with a hatch distance of 0.30 mm and a scanning speed of 1500mm3s−1 .
Fig. 7
Density and hardness for hatch distances 0.25 mm (grey) and 0.30 mm (black) The measurement of the Brinell hardness results in the highest hardness of 94±1 HB for the optimal values. For parameters with lower density the hardness decreases strongly while the standard deviation increases to values >10 HB due to porosity.
50 µs 100 µs 150 µs
scanning speed / mm
s
relative density/%
exposure time/µs
hardness/HB
CONCLUSION AND OUTLOOK
By adjusting the relevant parameters a great improvement of process time is possible. The build-up rate of initially 4.7 mm3s−1 at a power of 0.4 kW could be increased to 22.5 mm3s−1 at 1 kW, corresponding to an increase by 480%. The relative density amounts to 99.4% and the Brinell hardness of 94 HB is even higher than the reference value for conventionally fabricated parts of 80 HB.
Tasks for the future will be the further improvement of the side roughness of the built parts by setting special values for the boundary melt tracks. The remaining surface roughness could for instance be minimized by multiple re-melting of one layer which can additionally have positive effects on the density [7]. Apart from the roughness improvements powder investigations will follow to find whether the remaining gas porosity is a product of degenerating powder quality or a result of still insufficiently adapted scanning parameters which need further research.
REFERENCES
[1] GIBSON, I. ET AL.: Additive Manufacturing Technologies: 3D Printing, Rapid Prototyping, Direct Digital Manufacturing. – 2nd edition, Springer, 2015.
[2] WIESNER, A.; SCHWARZE D.: Multi-Laser Selective Laser Melting. – Proc. of 8th Conference on Photonic Technologies LANE 2014.
[3] BUCHBINDER, D. ET AL.: High Power Selective Laser Melting (HP SLM) of Aluminum Parts. – Physics Procedia 12, 2011. p. 271 – 278.
[4] UHLMANN, E. ET AL: Investigation on Additive Manufacturing of Tungsten carbide-cobalt by Selective Laser Melting. – Proc. of 15th Machining Innovations Conference for Aerospace Industry, 2015. p. 8-15.
[5] GIBSON, I. ET AL.: Additive Manufacturing Technologies – 3D Printing, Rapid Prototyping, and Direct Digital Manufacturing. – 2nd edition, Springer, 2015.
[6] ABOULKHAIR, N. ET AL.: On the formation of AlSi10Mg single tracks and layers in Selective Laser Melting. – Journal of Materials Processing Technology 230, 2015. p. 88-98.
[7] YASA, E.; KRUTH, J.: Application of laser re-melting on selective laser melting parts. – Advances in Production Engineering & Management 6, 2011.
p. 259-270.
