Feasibility study on hybrid manufacturing – combining laser‐based powder‐bed fusion and chill casting on the example of EN AC‐42000 alloy

Laser‐based powder‐bed fusion can be combined with casting into a hybrid manufacturing process. This is done to simultaneously utilize advantages of both processes. To examine related challenges stemming from this combination when using aluminium alloys, first experiments were conducted with EN AC‐42000 alloy, an alloy common in both standalone processes. EN AC‐42000 alloy samples with different surface treatments were placed into a sand mould and cast on to with an aluminium melt. The contact region between the partially molten insert and the cast material was examined with light microscopy. The main challenges were shown to be comprised of bonding issues due to present oxides and a high porosity. The high porosity was traced back to a porosity increase in the laser‐based powder‐bed fusion insert.


S C H L Ü S S E L W Ö R T E R
Aluminium, Fügen, Gießen, Laserstrahl, Pulverbett

| INTRODUCTION
Additive Manufacturing is a highly focused area in research and development and is constantly gaining relevance in industry.The ability to directly manufacture metal parts without the need to undergo post-processing is reserved to a few techniques, of which single-step powder bed based processes are most prominent [1,2].These use a laser or electron beam to melt scan tracks in a powder layer, each layer built on top of the last one, finally resulting in the required geometry.
Casting is amongst the oldest techniques for shaping products using metals.It is still one of the most important processes up to this day speaks for itself.The process itself consists of pouring a liquid material into a mould.Once the melt is solidified and the part is cooled down sufficiently, the part is removed from the mould [1].The combination of these two fabrication processes, additive manufacturing and casting, referred to as hybrid manufacturing, has multiple advantages, Figure 1.It also aims at overcoming the disadvantages of additive manufacturing, namely the slow build rate, high energy cost and limited build space [2,3].
For achieving a mechanical connection between laser-based powder-bed fusion manufactured part and the cast, three fit options are available: force fit, form fit and metallic continuity [4].Force fit requires mechanical pressing or clamping mechanisms incorporated into the design.Form fit requires locking mechanisms like teeth or notches, which require space and add complexity to a component.To achieve metallic continuity, there are several mechanisms available, like diffusion or mixture in the liquid phase [4].In this research, the aim was to achieve metallic continuity to avoid the drawbacks of the remaining options for obtaining a mechanical connection.As the material of choice was an aluminium alloy, the greatest expected obstacle for metallic continuity was the characteristic oxide layer present on the surface of aluminium alloys.This oxide layer acts like a barrier both for diffusion and for continued growth of preexisting grains on the solid liquid phase boundary [5,6].
Finally, the choice of material was a precipitationhardenable hypoeutectic aluminium silicon alloy, here EN AC-42000 alloy.Reasons for this choice have been firstly, aluminium alloys are versatile metallic materials most commonly found in lightweight constructions, which is a good fit to the features available through laser-based powder-bed fusion.Secondly, the properties of EN AC-42000 alloy are well-suited for casting as well as laser-based powder-bed fusion, which, thirdly, is one of the reasons the material is commonly used and well-proven for both manufacturing techniques [2].
The goal of this study was to find the most pressing challenges and obstacles when attempting hybrid manufacturing of EN AC-42000 alloy parts through combining laser-based powder-bed fusion and casting.In this study, the fusion zone between a laser-based powder-bed fusion insert and the cast was studied.Multiple surface conditions of the insert were considered and comprised of untreated, sand blasted, lathed and caustic soda etched samples.The quality of the connection was evaluated based on residual oxides and level of porosity in the vicinity of the fusion zone.

| Laser-based powder-bed fusion fabrication
Rod shaped samples were manufactured with a SLM280HL machine (SLM Solutions GmbH, Lübeck, Germany).Their diameter was 12 mm with a length of 60 mm, with an upright orientation and build direction in axial direction of the sample.The manufacturing parameters are typical for the chosen alloy, Table 1.

| Surface treatments
Cylindrical laser-based powder-bed fusion manufactured EN AC-42000 alloy samples receive four different surface treatments.One sample was tested per configuration.
1. Sand blasting with 200 μm-250 μm aluminium oxide, pressure 6 bar for 7 minutes.Resting time at atmospheric conditions longer than one week.2. Sand blasting with 200 μm-250 μm aluminium oxide, pressure 6 bar for 7 minutes.3. Lathing with 300 μm of removed material, 600 m/min 4. Caustic soda etching at room temperature.The procedure was conducted by degreasing the sample with water based detergent, 5 minutes submersion in 10 % sodium hydroxide solution, rinsing with desalinated water and removal of the soda etching coat by submerging the sample in 35 % nitric acid for 1 minute.Afterwards, the acid was removed by rinsing the sample in desalinated water, rinsing in ethanol alcohol and finally drying with a hot airstream for 20 seconds.
Following this, samples were stored for approximately 5 minutes in ambient environment prior to being cast onto.This accounts for the delay from completion of the surface treatment until the cast procedure.As such, a renewed formation of an oxide layer was possible.

| Casting setup
A sand mould was prepared, Figure 2. The setup was stabilized using forming sand (bentonite bound quartz sand).The mould was formed by AW6060 aluminium tubes embedded in the forming sand and a laser-based powder-bed fusion sample placed within each tube.The gating system was implemented through channels in the forming sand.The AW6060 tubes had an inside diameter of 40 mm and a height of 80 mm.The mould has been filled with aluminium melt of a similar composition, Table 2.The temperature of the melt was measured at 750 °C directly before casting.After the cast had sufficiently cooled down, the sand mould was removed.The samples were separated and cut into pieces suitable for light microscopy analysis.

| Microsection
The microscopy analysis was prepared by first cutting the cast samples with a band saw along their longitudinal axis, followed by several grinding and polishing steps using manually operated grinding machines.First, the saw grooves were removed using 120 grit sand paper.This was followed by grinding steps with 360 grit and 1000 grit.Finally, polishing was done with 6 μm, 3 μm and 1 μm polishing paste.
Microscopy itself was performed using a Leica DM LM light optical microscope.

| RESULTS
In this chapter, experimental results are presented.The detailed microscopy images are located at specific points of the cross-sectioned sample, Figure 3.A comparison of resulting microscopy images is provided, Figure 4.These images are finally assessed regarding metallic fusion and porosity, Table 3.

| DISCUSSION
In all samples, high porosity was observed in the laserbased powder-bed fusion fabricated insert.A porosity increase due to elevated temperatures exhibited in the casting process is implied, as usual porosity levels of laser-based powder-bed fusion manufactured parts are well below what was observed [7].Usual porosity levels for the as-fabricated EN AC-42000 alloy samples with these fabrication parameters were documented earlier with about 0.3 % [8].Porosity levels in the laser-based powder-bed fusion fabricated inserts after the casting process in this study were above 10 %, Figure 4.This has been unexpected, as porosity increase due to solution annealing is much less pronounced [8].The most probable cause of this difference in porosity between this study and solution annealing is the high temperature the samples' heat affected zones were exhibited to.Porosity at the level found in this study is not feasible for most applications, with few exceptions [9].The most probable reasons for porosity increase after heat treatment of laser-based powder-bed fusion fabricated parts are hydrogen and/or argon contamination.Argon contamination may stem from its use as inert gas during the laser-based powder-bed fusion process, where it is present in the building chamber at all times.Argon may be trapped inside the material during fabrication because of the rapid cool down (10 4 to 10 6 K/s) of the material during solidification not allowing enough time for residual Argon bubbles to rise to the top of the melt and exit [7].
Hydrogen pore formation in laser-based powder-bed fusion part fabrication with aluminium alloys is mostly caused by the presence of moisture in the powder during fabrication.Hydrogen has a high solubility in the liquid aluminium phase, causing it to accumulate there.When the solubility limit is reached locally, pore growth is initiated.The supply of hydrogen into the pore is limited by diffusion through the pores' surface area and stops as soon as it reaches the solidification front.For a high solidification rate, solubility of hydrogen in aluminium is increased.This in turn means that with the right laserbased powder-bed fusion build parameters, it is possible to trap more hydrogen in the material in a metastable state, reducing porosity in the as-built state of the part.During heat treatment, this metastable state can relax, causing pore growth [10].An extended form of this phenomenon, where porosity increase was more extreme due to higher temperatures, is expected to have taken place in this study.
Lack of fusion of various degrees was present in all samples.There is a direct correlation between metallic fusion and thoroughness of oxide layer removal, as is evident by the best material fusion result in the lathed sample.Another possible influence is surface roughness, which was lowest in the lathed sample and comparatively high in all other samples.Sand blasting seems unsuitable for sufficient oxide layer removal or requires different parameters.It should be noted, that during sand blasting, the abrasive particles are carried by an airstream, which in itself has an oxidizing property.The etching treatment applied in this study was unsuitable for achieving material fusion, although a slight improvement was observed compared with the first sample.This may be due to re-oxidation during rinsing and drying of the sample and in the time between treatment and casting.To overcome this re-oxidation, zincate treatments are a state of the art method for stabilization of aluminium surfaces by preventing re-oxidation.Because of the low melting point of the zincate layer, it is removed during the casting process in hybrid manufacturing.The relatively small amounts of elementary zinc introduced into the part during manufacturing don't have a negative effect on material properties [11].
The cast material in this study did not show any unusual properties related to the novelty of hybrid casting.Some shrinkage cavities could be observed, which are related to the casting setup.As the quality of the cast material was not the focus of this study, no excessive effort was invested into the casting setup, which in effect was lacking inert gas, argon post-treatment for hydrogen removal or a casting filter for oxide removal.

| CONCLUSIONS
The goal of this study was to obtain first insights into the challenges involved with combining laser-based powderbed fusion additive manufacturing with casting.The results showed that a significant porosity increase occured in the case of EN AC-42000 alloy, an hypoeutectic aluminium silicon alloy.This porosity increase is a major issue for the process feasibility and possible solutions for the prevention of this significant gain in porosity should be studied further.
Apart from porosity, some degree of material fusion between the cast and the laser-based powder-bed fusion fabricated insert was evident.For process feasibility, those results imply the need for thorough oxide layer removal and preventing renewed oxide layer growth after removal.Methods for removing the oxide layer and stabilizing this deoxidized state suitable for laser-based powder-bed fusion fabricated components should be examined in further research.

F I G U R E 1
Hybrid casting aims to combine the advantages of both laser-based powder-bed fusion and die casting.B I L D 1 Hybrides Gießen versucht, die Vorteile von laserstrahlbasiertem Pulverbettschmelzen und Guss zu vereinen.

T A B L E 1
Parameters for EN AC-42000 alloy laser-based powder-bed fusion fabrication.T A B E L L E 1 Parameter des EN AC-42000 alloy laserstrahlbasierten Pulverbettschmelzens.

FT A B L E 2
I G U R E 2 a) Shows the scheme of the experimental method.The cylindrical aluminium sample was being fully cast in.The cross section of the sprue was 8 mm×12 mm.b) shows footage from conducting the casting process with the samples inserted into the four tubes within the sand mould.B I L D 2 a) Zeigt das Schema der experimentellen Methode.Die zylindrischen Aluminiumproben werden vollständig eingegossen.Der Querschnitt des Angusskanals ist 8 mm×12 mm.b) Zeigt die Ausführung des Gießprozess' mit den Proben innerhalb der vier Rohre in der Sandform.Chemical composition of the materials used in mass percentage.T A B E L L E 2 Chemische Zusammensetzung der verwendeten Werkstoffe in Massenprozent.

F I G U R E 3
Overview of a cut, ground and polished sample where evaluation area 1 and 2 are marked.B I L D 3 Übersicht einer gesägten, geschliffenen und polierten Probe in der die Auswertungsgebiete 1 und 2 markiert sind.F I G U R E 4 Microscopy images of the samples: Areas 1 and 2 depict the visible contact surface and porosity.Area 2 at higher magnification attempts to capture evidence of metallic fusion or lack thereof.B I L D 4 Mikroskopaufnahmen der Proben: Die Flächen 1 und 2 bei 2,5-facher Vergrößerung stellen die sichtbare Kontaktfläche und die Porosität dar.Fläche 2 bei 50-facher Vergrößerung versucht zu zeigen, ob Stoffschluss erfolgt ist.