CC BY
75
УДК 621.7
Влияние режима селективного лазерного плавления на образование пор в алюминиевом сплаве AlSilOMg
P. Ferro1, R. Meneghello1, S.M.J. Razavi2, F. Berto2, G. Savio3
1 Падуанский университет, Виченца, 36100, Италия 2 Норвежский университет естественных и технических наук, Тронхейм, 7491, Норвегия 3 Падуанский университет, Падуя, 35131, Италия
Аддитивные технологии дают уникальную возможность создавать изделия любой геометрии. Аддитивное производство является одним из ключевых направлений в области 3D-печати металлов. Технологии литья имеют больше ограничений в виду неравномерного охлаждения изделий или необходимости использовать литейные формы. Внутри деталей, изготовленных аддитивным способом, могут присутствовать дефекты, такие как оксидные пленки, поры или нерасплавленный порошок, но их количество можно значительно уменьшить или контролировать путем оптимизации параметров технологического процесса. В случае литых изделий такая оптимизация не представляется возможной, поскольку дефекты в разных зонах литого изделия сильно различаются в зависимости от условий затвердевания. В работе проанализировано влияние технологических параметров на образование пор в образцах алюминиевого сплава AlSi10Mg, полученных методом селективного лазерного плавления. Найден оптимальный режим для получения материала с максимальной плотностью и наилучшими механическими свойствами. Предложена модель, позволяющая установить взаимосвязь между количеством пор и пределом прочности сплава на растяжение.
Ключевые слова: аддитивное производство, алюминиевый сплав, селективное лазерное плавление, пористость, механические свойства
DOI 10.24411/1683-805X-2019-15010
Porosity inducing process parameters in selective laser melted AlSilOMg aluminium alloy
P. Ferro1, R. Meneghello1, S.M.J. Razavi2, F. Berto2, and G. Savio3
1 Department of Engineering and Management, University of Padova, Vicenza, 36100, Italy
2 Department of Engineering Design and Materials, Norwegian University of Science and Technology, Trondheim, 7491, Norway 3 Department of Civil, Environmental and Architectural Engineering, University of Padova, Padova, 35131, Italy
Additive manufacturing techniques are known for the unrivalled geometric freedom they offer to designers. It is one of the mainstays of "metal 3D-printing", compared to casting, which, in contrast, implies more restrictions because some shapes do not cool evenly or may need moulds or forms. Despite the possible presence of defects inside additive manufactured components, such as oxide films, pores or unmelted powder, they can be strongly reduced or controlled by process parameters optimization. That seems not true for a casting component, in which defects can vary a lot from zone to zone according to the solidification conditions. Porosity inducing process parameters in selective laser melted AlSilOMg aluminium alloy are carefully analysed with the aim to find optimal conditions that guarantee the maximum material density and the best mechanical properties. Finally, a model is proposed that correlates the amount of pores with the alloy ultimate tensile strength.
Keywords: additive manufacturing, aluminium alloy, selective laser melting, porosity, mechanical property
1. Introduction
Additive manufacturing first emerged in 1987 with
stereolithography from 3D Systems, a process that solidifies thin layers of ultraviolet light-sensitive liquid polymer using a laser. Starting from that period, different additive manufacturing processes developed so that, in 2010, the American Society for Testing and Materials (ASTM) group
ASTM F42-Additive Manufacturing formulated a set of standards that classify the range of additive manufacturing processes into seven categories (Standard Terminology for Additive Manufacturing Technologies, 2012) [1]: VAT photopolymerisation, material jetting, binder jetting, material extrusion, powder bed fusion, sheet lamination, directed energy deposition. Additive manufacturing offers
© Ferro P., Meneghello R., Razavi S.M.J., Berto F., Savio G., 2019
several advantages throughout the design workflow including little lead time, few constraints, little-skill manufacturing. Furthermore, it is a low environmental impact process because it is characterized by less waste and energy saving. As a matter of fact, when compared with traditional manufacturing processes, additive manufacturing can significantly reduce energy usage by using less material and eliminating steps in the production process.
A significant problem with this type of production stands up in failure assessment. The material properties of additive manufacturing components change depending on the fabrication process making it dependent on the geometry of the components. This means that every change in the geometry will change the way that the additive manufacturing machine performs its fabrication routine affecting the properties of the resulting solid [2-4]. Currently, the main challenges for laser and metal powder-based additive manufacturing include the formation of defects (e.g., porosity), low surface finish quality, and spatially nonuniform properties of the material. Such challenges stem largely from the limited knowledge of complex physical processes in additive manufacturing especially the molten pool physics such as melting, molten metal flow, heat conduction, vaporization of alloying elements, and solidification.
Among the aluminium alloys, AlSilOMg is certainly the most used for fabricating components through selective laser melting process. Such alloy is often used in components of motor racing, automotive industry and general engineering. It has historically been a right choice for lightweight and thin-walled casting parts or any components with a complex geometry subjected to high loads. Generally, this aluminium alloy offers an economic alternative to titanium in the case of lightweight components exposed to noncriti-cal fatigue and mechanical loads.
Today, AlSilOMg is the most used aluminium alloy by laser melting techniques mostly because of its reduced susceptibility to cracking during solidification compared to others Al alloys (6xxx series in particular). However, the high thermal conductivity of aluminium and its alloys makes them notoriously difficult to cast and weld. For laser melting, things get even worse. Aluminium powders are highly reflective and have a high thermal conductivity when compared to other materials. Therefore, high laser power is required to overcome the conductive cooling (rapid heat dissipation) into the surrounding material and melt the powder particles.
Another layer of complications comes from the susceptibility of these alloys to porosity formation mechanisms.
In this regard, the sensibility of Al to surface oxidation and moisture pick-up are two major causes. Provided the powder was conditioned properly from the beginning and is still devoid of oxide films, this undesirable effect can be mitigated during the selective laser melting process with strong fluxes of an inert atmosphere. However, since the formation of oxide films cannot be avoided completely, the high laser power recommended to process aluminium alloys is required not solely to compensate heat diffusion, but also to disrupt oxide films.
AlSilOMg parts issued from powder-bed additive layer manufacturing processes exhibit higher or at least, comparable, mechanical properties than casted parts [5]. In particular, the ultimate tensile strength of the as-built aluminium additive layer manufacturing parts is always higher than those obtained by high pressure die casting in either as-casted or in the aged condition. While in the casting of AlSilOMg parts, the hardness and strength increase during the heat treatment with the precipitation of intermetallic compounds (Mg2Si), in additive layer manufacturing parts however, higher hardness and strength are already reached in the as-built state because of the very fine microstructure the rapid cooling allows. Another contribution comes from the fine distribution of the Si phase in the aluminium matrix as well as the presence of intermetallic compounds (i.e. Mg2Si).
Static and fatigue properties of additive manufactured AlSilOMg samples obtained by selective laser melting were investigated by Brandl et al. [6]. Mechanical properties were highly influenced by the building direction. However, the combination of 300°C platform heating and peak-hardening were found to increase the fatigue and static resistance, neutralizing at the same time the building direction effect. Unfortunately, in that work, process parameters optimization was not taken into account.
With the objective to increase the process efficiency, Buchbinder at al. [7] tried to increase the build rate of additive manufacturing AlSilOMg parts by using a 1 kW laser, while reaching a 99.5% material density. They demonstrated the possibility to extend the built rate from approximately 5 to 21 mm3/s. In a recent paper, Han et al. [8] investigated the selective laser melting AlSilOMg cellular lattice strut in terms of molten pool morphology, surface roughness and dimensional accuracy. The results showed that the average width and depth of the molten pool, the lower surface roughness and dimensional deviation decrease with the increase of scanning speed and hatch spacing. In that work the laser power was kept constant (200 W). The
Table
Chemical composition of the alloy AlSilOMg, wt %
Al Si Mg Fe N O Ti Zn Mn Ni Cu Pb Sn
Balance 9-11 0.25-0.45 <0.25 <0.20 <0.20 <0.15 <0.10 <0.10 <0.05 <0.05 <0.02 <0.02
300 ^-1-1-1-1-
0 20 40 60 80 100
Al cSl, wt % Si
Fig. 1. Aluminium-silicon phase diagram
influence of laser power, scan speed, scan spacing and island size on porosity development in AlSi 10Mg alloy builds has been investigated by Read et al. [9]. They found that the laser power, scan speed, and the interaction between the scan speed and scan spacing have a major influence on the porosity development in the builds. They also demonstrated the higher strength and elongation properties of selective laser melted samples compared to those of die-cast samples of similar composition.
Aboulkhair et al. [10] investigated the effect of scanning speed (in the range of250-1000 mm/s) and hatch spacing (from 50 to 250 ^m) on porosity in AlSi10Mg parts processed by selective laser melting. The laser power was kept constant and equal to 100 W. Surveying the windows of parameters, the best combination was found to be a speed of 500 mm/s and hatch spacing 50 ^m when using a layer thickness of 40 ^m and employing the presinter scan strategy yielding a relative density of 99.8%
Trevisan et al. [11] focused their study in porosity inducing hatch distance in fabricated selective laser melting AlSi10Mg parts. The hatch distance was varied from 0.1 to
Fig. 2. Geometry of the specimens (dimensions in mm)
0.15 mm while the laser power, the scanning speed and the layer thickness were kept constant and equal to 350 W, 1650 mm/s and 30 ^m, respectively. The lowest porosity amount was obtained by using a hutch distance of 0.13 mm.
The several papers present in the literature about selective laser melting applied to AlSi 10Mg alloy confirms, from one hand, the high interest of academic and industrial world on this topic and, on the other, the need to still deepen the strong and complex interaction between process parameters, microstructure and mechanical properties that characterises the process. This work is aimed at contributing to covering this gap by analysing the influence of laser power and exposure time on porosity amount in selective laser melted parts. Finally, an equation was proposed that correlates the ultimate tensile strength with porosity.
2. Materials and methods
The analysed material is the AlSi10Mg hypoeutectic alloy, mostly used for aluminium castings. Its chemical composition is collected in Table. The little amount of Mg (0.3-0.5 wt %) allows the reinforcement by natural or artificial aging while its near eutectic composition (Fig. 1) enhances its castability. The particle size of the AlSi10Mg powder is in the range of 15-45 ^m.
The samples were obtained by selective laser melting with the 3d Printer Renishaw 400AM. The building square-based platform has a side length equal to 250 mm. The laser is characterized by 400 W maximum beam power with a diameter of 70 ^m. Argon was used as protective gas against powder oxidation. Both cylindrical (diameter and length equal to 10 mm) and dog bone specimens were produced following the standard ASTM E606 (Fig. 2).
In order to reduce the possible process parameters combinations, the following inputs were kept constant during the tests: layer thickness d = 30 ^m, spot diameter O = = 70 ^m, platform temperature 1700C, point distance s equal to hatch distance h = 80 ^m, duilding direction— sample axis.
their position on the building platform (color online)
Fig. 3. Scanning strategy (a); spot-to-spot fabrication process, where s is the point distance, ® is the laser beam spot and h is the hatch distance (b) (color online)
In Fig. 3 the laser scanning strategy and the definition of point distance and hatch distance are shown. With the aim to minimize the porosity inside the material, the laser beam power and exposure time values were changed starting from those suggested by Renishaw (say, 275 W and 40 ^s, respectively). Three cylindrical samples for each couple of process parameters shown in Fig. 4 were carried out.
In order to optimize the material at disposal, samples for tensile tests were produced with the couples of parameters marked in green in Fig. 4. It is noted that even if it was not possible to change the scan speed vs, this last parameter can be approximated with the following relation:
Vs = J, (1)
where t is exposure time.
Samples for microstructural analyses were embedded in phenolic resin and prepared using standard grinding and polishing procedures. The microstructure was analysed using a scanning electron microscope (Quanta 2580 FEG). Light microscopy pictures were also taken with Leica DMC 2900 microscope. With the help of a dedicated software for image analysis, the microscope is driven to obtain the overall specimen image by means of the combinations of several micrographs carried out by scanning the entire surface of the specimen. The final micrograph is then pro-
cessed for the automatic counting of the area interested by porosity. Figure 5 shows the result of the micrograph processing routine above described for two cylindrical samples.
The tensile tests were carried out by using the MTS Acumen 3 and following the Standard ISO 6892-1:2016. Samples obtained with the power and exposure time equal to 350 W and 80 ^s, respectively, were not tested because of their high deformation resulted after the building process.
3. Results and discussion
3.1. Microstructure
Due to the incomplete homologous wetting and balling effects, pores are easily formed during the selective laser melting process. Figure 6 shows the pores morphology; gas porosity and sometimes balling were the most common defects observed. According to the phase diagram (Fig. 1) the microstructure (Fig. 6, c) is characterized by a-Al dendrites inside a eutectic matrix constituted by a-Al and Si-particles.
Figure 7 shows some micrographs taken from cylindrical samples obtained by varying both the laser power P and the exposure time. It is noted that samples containing lower amounts of pores are those obtained with the lower value of exposure time and laser power. For the sake of simplicity, Fig. 7 collects only the most significative mi-
400
^ 350
300
250
▲ ▲ ▲ ▲ ▲
▲ ▲ ▲ ▲ ▲
Scan speed, mm/s
500 1000 1500 2000 2500 3000
140 100 60
Exposure time, ps
20
Fig. 4. Process parameters tested in the experiment (color online)
r. 1,2 mn \ m
Fig. 5. Processed image for the porosity amount assessment. Samples obtained with power 375 (a), 300 W (b), exposure time 30 |is
Fig. 6. Pores morphology (a), balling (b) and microstructure morphology (c) induced by selective laser melting process. Power 350 W, exposure time 120 |is (color online)
crographs as a function of a limited range of process parameters tested. On the other hand, the mean value of the area percentage occupied by pores as a function of the entire range of tested process parameters is shown in Fig. 8, where the bubble areas are proportional to the porosity percentage detected in the cylindrical samples.
It is observed that minimum values of porosity are reached in the lower right corner of the graph (Fig. 8). Similarly, by using the concept of energy density Ed given by [11]:
Fig. 7. Some micrographs (magnifications x50) showing porosity as a function of laser power P = 350 (a-c), 325 (d-f), 300 W (g-i) and exposure time 80 (a, d, g), 60 (b, e, h), 40 |is (c,f, i) (color online)
350
£
£
£ 300
250 -
140 100 60 20
Exposure time, ]s
Fig. 8. Porosity percentage (%) as a function of process parameters (laser power, exposure time) and energy density (in J/mm3) (color online)
Ed = 57---- Ed = 94 ------------2.00
Ed = 86 - Ed = 172----- 314
10.10/ IE -fj^¡z
/ 3.77 / 217 ! 02
t / /0.43
6-98 / /0.57 / 0.
0.45
it can be easily noted that samples characterized by the highest density, stay in the zone crossed by isoenergy density curves having the lowest values, in the range of tested parameters (Fig. 8). Nevertheless, as observed by Prashanth et al. [12], the applicability of Eq. (2) is still at stake, even though it has been widely used in the literature for optimizing the selective laser melting parameters [11]. As a matter of fact, Eq. (1) may not properly represent the effective energy transferred to melt the powder bed, and thus it needs to be improved involving the material properties. With the aim to reinforce that idea, the graph energy density versus porosity is plotted in Fig. 9. It shows the positive trend of porosity versus energy density but with a high scattering R2 = 0.6. Despite this, results obtained are in agreement with those published by Read et al. [9] who found the minimum pores fraction induced by a critical energy density equal to approximately 60 J/mm3. Finally, it is worth noting to observe that the equivalent mean pore diameter decreases as the energy density decreases (Fig. 10). Furthermore, the lower the energy density, the lower the standard deviation. To the best of the authors knowledge, this selec-
Fig. 9. Porosity and ultimate tensile strength (UTS) versus energy density. y = 2483x"0522, R2 = 0.9739 (7); y = 0.0004xL9°79, R2 = 0.6374 (2)
tive laser melting feature has not been shown in previous work about this topic despite its advantage compared to casting components. Irrespective of the sample dimensions and thickness variations, the solidification defects induced by selective laser melting process depend only by process parameters and do not vary from zone to zone of the component. On the contrary, solidification defects in casted parts strongly depend on both process parameters and casting dimensions. The higher the casting dimensions or thickness variations, the lower the microstructure homogeneity resulting from the corresponding different cooling conditions.
3.2. Tensile tests
As expected, tensile strength is linked to the microstructure morphology, with particular reference to defects. The higher the porosity percentage, the lower the ultimate tensile strength (Fig. 9). A negative trend characterizes the ultimate tansile strength versus the energy density. This is attributed both to stress concentration effect and nominal cross-section area reduction induced by the porosity itself. Basing on this idea the following relation is proposed that correlates the UTS with the porosity (P, %):
UTS = UTS
\n
1 --
100
(3)
Fig. 11. Ultimate tensile strength versus porosity
80
densitv
J/mm
Energy
Fig. 10. Equivalent mean pore diameter as a function of energy density
where UTS0 is the ultimate tensile strength of the sample without porosity and n is a material parameter, the values of which are obtained by the best fitting with experimental data. Figure 11 shows the obtained result with UTS0 and n equal to 338 and 10, respectively.
4. Conclusions
Porosity inducing process parameters in SLM AlSi10Mg samples were studied. Because of the several parameters characterizing the process, layer thickness, spot diameter, platform temperature, point distance, hatch distance and building direction were kept constant. The laser power and the exposure time have been varied from 275 to 375 W and 30 to 120 ^s, respectively. Tensile tests were finally carried out on samples produced with different process parameters. Results were described as a function of energy density and are summarized as follows:
- gas porosity is the main defect detected in the microstructure,
- the mean pore diameter value and its standard deviation decreases as the energy density decreases,
- minimum values of porosity percentage are obtained with the minimum values of energy density, between approximately 60 and 50 J/mm3,
- the higher the porosity percentage, the lower the ultimate tensile strength; an equation was finally proposed to model such behaviour.
References
1. Standard Terminology for Additive Manufacturing Technologies: Designation F2792-12a. / ASTM Committee F42 on Additive Manufac-
turing Technologies. Subcommittee F42.91 on Terminology. - ASTM Int., 2012.
2. Razavi S.M.J., Ferro P., Berto F., Torgersen J. Fatigue strength of blunt V-notched specimens produced by selective laser melting of Ti-6Al-4V // Theor. Appl. Fract. Mech. - 2018. - V. 13. - P. 74-78.
3. Razavi S.M.J., Ferro P., Berto F. Fatigue assessment of Ti-6Al-4V circular notched specimens produced by selective laser melting // Metals. - 20147. - V. 7. - P. 291.
4. Razavi S.M.J., Bordonaro G.G., Ferro P., Torgersen J., Berto F. Fatigue
behavior of porous Ti-6Al-4V made by laser-engineered net shaping // Materials. - 2018. - V. 11. - P. 284.
5. Kempena K., Thijs L., Van Humbeeck J., Kruth J.-P. Mechanical properties of AlSi10Mg produced by selective laser melting // Phys. Proc. -2012. - V. 39. - P. 439-446.
6. Brandl E., Heckenberger U., Holzinger V., Buchbinder D. Additive manufactured AlSi10Mg samples using selective laser melting (SLM): Microstructure, high cycle fatigue, and fracture behaviour // Mater. Design. - 2012. - V. 34. - P. 159-169.
7. Buchbinder D, Schleifenbaum H., Heidrich S., Meiners W., Bultmann J.
High power selective laser melting (HP SLM) of aluminum parts // Phys. Proc. - 2011. - V. 13. - P. 271-278.
8. Han X., Zhu H., Nie X., Wang G., Zeng X. Investigation on selective laser melting AlSi10Mg cellular lattice strut: Molten pool morphology // Surface Roughness Dimens. Accuracy. Mater. - 2018. - V. 11.-No. 3. - P. 392-405.
9. Read N., Wang W., Essa Kh., Attallah M.M. Selective laser melting of AlSi10Mg alloy: Process optimisation and mechanical properties development // Mater. Design. - 2015. - V. 65. - P. 417-424.
10. Aboulkhair N.T., EverittN.M., AshcroftI., Tuck Ch. Reducing porosity in AlSi10Mg parts processed by selective laser melting // Add. Manufact. - 2014. - V. 1-4. - P. 77-86.
11. Trevisan F., Calignano F., Lorusso M., Pakkanen J., Aversa A., Ambrosio E.P., Lombardi M., Fino P., Manfredi D. On the selective laser melting (SLM) of the AlSi10Mg alloy: Process, microstructure, and mechanical properties // Materials. - 2017. - V. 10. - P. 76-99.
12. Prashanth K.G., Scudino S., Maity T., Das J., Eckert J. Is the energy density a reliable parameter for materials synthesis by selective laser melting? // Mater. Res. Lett. - 2017. - V. 5. - No. 6. - P. 386-390.
Received July 07, 2019, revised August 19, 2019, accepted August 26, 2019
CeedeHua 06 aemopax
Paolo Ferro, Prof., University of Padova, Italy, [email protected]
Roberto Meneghello, Prof., University of Padova, Italy, [email protected]
Seyed M. Javad Razavi, PhD Student, Norwegian University of Science and Technology, Norway, [email protected] Filippo Berto, Prof., Norwegian University of Science and Technology, Norway, [email protected] Gianpaolo Savio, Assist. Prof., University of Padova, Italy, [email protected]
