Type: Process Essays
Sample donated: Lucy Sharp
Last updated: December 29, 2019
INVESTIGATION INTO THE EFFECT OFBUILD PARAMETERS AND BUILD ANGLE ON THE DOWN SKIN SURFACES FOR STAINLESS STEELLASER MELTED PARTSAurelia Pana – 1723161Faculty of Science and Engineering, University of Wolverhampton,Telford, TF2 9NT, UK 1.
INTRODUCTION INTO LASER MELTINGIn the Additive Manufacturing (AM)industry several technologieshave been developed, powder bed fusion being the leading technology forobtaining metal parts.Someof powder bed fusion processes are alternately known as selective lasersintering, selective laser melting, direct metal laser sintering, direct metallaser melting, and electron beam melting. All powder bed fusion processes havea similar basic operating principle.
The main differences between processes are in the way layers are deposited tocreate parts and in the materials that are used. (W.E.
King et al., 2015) The designstage of the process starts bycreating 3D CAD model of the desired object. Using a pre-defined slicingprogram, the 3D CAD file is converted into a series of thin parallel layers thatfully describe the geometry of the desired object creating a 2D image of each layer (Wikipedia,2015).The data obtained is transferred to a computer controlled laser device.Depending on the equipment and the method used, thin layers of powder arespread with a recoating system onto a platform and a laser or electron beam isused to fuse the powder at locations specified by the model of desiredgeometry. When one layer is completed, a new layer of powder is applied and theprocess is repeated until a 3D part is produced (S. Bose et al.
,2017). Theapplications of rapid prototyping are vast. Over the last years, AMtechnologies are continuously expanding in industrial sectors likearchitectural, medical, dental, automotive, furniture and jewellery, and manyothers. (Metal AM, 2014)2. THEORY AND BACKGROUNDIn AM processes,laser sintering is a technique that uses the directed energy generated by alaser to melt particles of metal powder, layer by layer forming a solidstructure (Wikipedia, 2014).
For the experiment discussed in this paper, the laser sintering method appliedis Selective Laser Melting (SLM). SLM is a process based on powder bed fusion where metal powder iscompletely melted in order to obtain dense parts (J. Jhabvala, 2010). SLM method involves a number ofsteps that move from the virtual CAD description to the physical resultantpart.
The first step is to create a 3D CAD and then convert the CAD file onto anSTL file format because nearly every SLM machine accepts this type of file format.After completing these steps, using computer software the STL file will be mathematicallysliced it in 2D cross sections, representing the layers that will form thesolid structure of the part. The data obtained is converted into a readablefile for the SLM equipment and transferred to the machine that will be used tocreate the physical object. The machine must be properly set up prior to thebuild process. Such settings would relate to the build parameters.
The primaryprocess parameters for SLM that are related to the quality of the part in termsof surface finish are the scanning speed, laser power, layer thickness, hatchdistance and beam offset (Y.Tian,2017),shown in Figure 1.Figure 1- Process parameters The build parameters can be standarddepending on the machine used, or either modified by entering them manually bya trained operator. (I. Gibsonet. al.
,2015) Laser power is the energy brought by the laser beam to thepowder bed and influences the melting temperature. (P.Hanzl et. Al., 2015) Scanningspeed (speed of beam over the powder bed) has an important role in lasersintering because it can affect the mechanical proprieties of the final part.When increasing the scanning speed, a better quality of the surface finish canbe obtained (L.
Taimisto, 2009). Layer thickness usually influences the building time andrepresents the depth of each successive addition of metal powder to thebuilding platform. A lower thickness of layers can decrease surface roughness (S.
Dadbakhsh, L. Hao, 2014). Therefore, all build parametersneed to be carefully chosen to avoid creating a defective part. 3.
EXPERIMENTATIONIn the following experiment two setsof parts was manufactured and an investigation into the effects of the buildangle and downskin condition on laser melting Stainless Steel powderedmaterial, has been carried out. The SLM equipment used in this work is an EOSM270 machine which has a Yb-fibre laser with a variable focus diameter 100?m – 500 ?m and a maximum power output of 200 W (3RSystems,2014). Followingthe generic steps for the SLM process, first of all, the 3D CAD model for bothsets of test samples was created using Solidworks2017 software. In Figure 2 is illustrated the 3D CAD model and its dimensions. Figure 2- 3D CAD model of test samples In Table 1 the angle dimensions (?) forall 11 A parts and B parts are presented.Table 1 – Built Angle parts “A” and “B” Built Angle (?) for parts “A” and “B” 1 2 3 4 5 6 7 8 9 10 11 18 21 24 27 30 33 36 37 39 42 45 After designing the 3D CAD modelsfor both sets of test samples, the CAD files ware converted onto STL files.Using Magics Software both sets ofparts ware fixed onto a virtual platform, without the aid of support structures,although for overhanging structures inclined from 0 to 45 degrees supportstructures are needed (Y.
Kajima et.al., 2017). Forthe next step, the STL files for all test samples and the support files, ware slicedand converted using EOS RP TOOLSsoftware in the SLI format, which in the EOS language represents the part layerby layer. The thickness of each layer was set at 0.02 mm. After completing thisstep, the files were transferred to the machine, where settings for the buildparameters and exposure strategies ware made as it follows in Table 2: Table 2 – Built parameters test samples Build parameters for set “A” Build parameters for set “B” Pre-contour Post-contour Pre-contour 1 Pre-contour 2 Post-contour1 Post-contour 2 Power = 40W Power = 40W Power = 40W Power = 40W Power = 40W Power = 40W Scan speed = 700 mm/s Scan speed = 700 mm/s Scan speed = 800 mm/s Scan speed = 1200 mm/s Scan speed = 1600 mm/s Scan speed = 1800 mm/s Offset = 0.
020 µm Offset = 0.00 µm Offset = 0.03 µm Offset = 0.02 µm Offset = 0.01 µm Offset = 0.00 µm In the exposure strategy, pre-contours and post-contours canbe given by the standard settings of the machine. Pre-contours are used todefine sections to be melted and post-contours are used to define the finalsize of the component. Contours are an important aspect because the quality of surfacefinish is influenced by them.
(F. Caligano et. Al. 2012) In Figure 3 below, thepre-contour is illustrated as “contour without beam offset”, and post-contouris presented as “contour with beam offset” Figure 3- Contour strategy In this experiment, the contouring values for group “B” oftest samples ware modified in order to investigate if the surface finish of thedownskin surfaces will be improved when using two pre-contours and twopost-contours with different scanning speed. Once the build contours and other parametersware set on the EOS M270 machine, the build platform was fixed and orientatedinside the build chamber. The build chamber was pre-filled with argon gas toprotect the parts from the effects of oxidation.
After the completion of theSLM process, the test samples were cut off from the build platform usingWireEDM process. In Figure 4 both sets of test samples obtained by SLM are presented.Figure 4 – test samples obtained by SLM Using Guyson Euroblast 4SF dry blastingmachine, all parts were blasted with aluminium oxide and after this process,for each test sample, the downward facing surface was analysed using theconfocal laser microscope Olympus LEXT TS 150 (Figure 5). Figure 5- Analysingsurface finish 3.RESULTS AND ANALYSISSurface roughness it is characterised by the deviations in the directionof the normal vector of a real surface fromits ideal form (G. Strano et. al., 2013).
In Table 3 the average values obtained by analysing thesurface roughness using the confocal microscope are presented, where Ra is the arithmetical mean deviationof the roughness profile. Ra can be calculated by the formula shown in Figure 6,where Z(x)is the deviation of surface height at x from the mean heightover the profile. Figure 6 – Ra representation For this experiment, all values forthe average surface roughness have been measured by the confocal lasermicroscope. Knowing that the parts were designed with varyingangles of the overhang surface to determine the quality of the downward-facingsurface when changing the build parameters, and knowing that for StainlessSteel the angle of 45? is used as the minimum build angle for the processwithout support structures, the surface roughness values obtained for set A andset B will be compared and discussed. Table 3 – Ra average valuesfor downskin surfaces. Part Number Overhanging Angle (?) Ra for Set A Ra for Set B 1 18 30.
9338 29.8409 2 21 21.8308 22.4608 3 24 24.5165 27.3188 4 27 36.7436 25.
0643 5 30 24.5165 29.126 6 33 22.1805 30.0908 7 36 14.2019 18.4023 8 37 30.1682 10.
0956 9 39 9.2447 13.4115 10 42 6.4263 13.
7339 11 45 14.9983 11.5549 For part number A1 and B1, Ra values are approximately thesame and the aspect of the overhang surface built at 18? is poorly. See imagesfrom Figure 7 below:Figure 7– Downskin surface Part A1 and B1 Part A1 Part B1 In the chartbelow, Figure 8 a profile of the surface roughness for part A1 and B1 isdrowned.Figure 8- Profile of surface roughness forpart A1 and B1 Looking at Figure 4, a slightimprovement can be observed regarding the quality of surface roughness whenincreasing the overhang angle. However, even when modifying the buildparameters for contours, for the overhang angles that are lower than 30 degreesthe quality of surface finish is still unsatisfying.
An image of parts A5 andB5 with the overhanging angle of 30 degrees is presented in Figure 9.Figure 9 – Surface finish part A5 and B5 In the chart from Figure 10, aprofile of the surface roughness for part A5 and B5 is showing that there is athere is a small difference between the profiles created. For set B, even ifthe average Ra values are higher (see Table 3), the quality of the surfaceseems to improve.Figure 10-Profile of surface finish for partA5 and B5 Looking at Table 3, the Ra averagevalue for part A8 is considerably increased compared to Ra average value for B8with the overhanging build angle at 37 degrees. In figure 11 the profile of thesurface for both parts is highlighting the fact that for the prescribed dimensionsof sample A8 and B8, set B has a better quality of the downskin surface wheretwo pre-contours and two post contours ware applied.
Figure 11-Profile of surface finish for partA8 and B8 In Table 4, the quality of thesurface finish can be compared by simply visualising the images showed. Thedifferences between set A and set B are clearly visible.Table4 – Images of set A8 to A 11 and B 8to B 11 Comparing the Ra average values forsample A11 and B11 (Table 3), the difference between them is insignificant. InFigure 12 the profile of surface finish shows that the quality of downskinsurface for both sets is almost similar with small differences of Ra values.Figure 12- Profile defining surface finish for part A11 and B11 Therefore, itis revealed that surface roughness by SLM can be varied by modifying theprocessing parameters.4.DISCUSSION Analysisof Ra average values for both set of test samples in relation with build angle(?) of the downward facing surfaces, can be seen in Figure 13, which shows aclear dependence of surface angle on Ra.
As ? increases, the value of Radecreases. Figure13- Ra in relation with ? The variable parameters for contouring as the offset of the beam,and the scanning speed function of the laser ware modified for set B, and thestandard settings ware used for set A. Looking at Figure 13, for set B when theSLM process was optimised by changing the contouring parameters, the averagevalues for Ra are variating between 10µm-30µm, and for set A Ra values arebetween 7µm-37µm. Even though the difference between both sets of parts it isnot enormous in terms of surface finish, modifying the exposure type by addingone more pre-contour and a post-contour, seems favourable when increasing thescanning speed and laser beam offset. A study made by F.Caligano (2014) revealsthe importance of using support structures for overhanging surfaces in order toavoid “staircase effect” of angled wallsand surfaces.
It was discovered that for surfaces with an angle smaller than 30?,the staircase effect tended to increase. The minimum orientation of theoverhang surface that provided an acceptable quality of surface finish, was theorientation at 45?.The experiment was carried out on two types of material AlSi10Mgand Ti6Al4V alloys. Referring to Figure 13, the minimum value of Ra is obtainedat the 36? angle for set B, and for set A at 39?. For both sets of testsamples, surface roughness seems to remain stabilized. A good quality of thedownward-facing surfaces is difficult to obtain by SLM process without the aidof supports. Introducing support structures for the overhanging surfaces willmake the process less cost-effective, however the surface that has the lowest requirement for surfacefinish can be considered as the bottom surface when orientating and fixing thepart in the design stage of the process, since the surface attached to the supportstructure will have an increased roughness after removal of the supportstructure (K. Zeng 2015).
5.CONCLUSIONSAn investigation into the effect ofbuild angle and downskin condition obtained in SLM using Stainless Steel hasbeen presented to further the understanding of the relationship between surfaceroughness and process parameters. It was found that for Stainless Steel,considering ideal the Ra average value obtained at 45?, and comparing it withall the other values, at the standard parameters settings, the minimum angle valuesfor overhanging surfaces that can be considered acceptable for SLM processwithout the use of supports are ranging between 39?-45?. When the processparameters ware modified, the minimum angle values for overhanging surfaces areranging between 36?-45?.To conclude, for overhanging surfaces with an anglebelow 36? support structures are needed. The orientation of thepart must be considered in the design stage in order to reduce the number ofsupports and avoid damaging the quality of the surface after their removal.
6. REFERENCESS1. F.Calignano, D. Manfredi, E. P. Ambrosio, L. Iuliano, P.
Fino (2012) Influenceof process parameters on surface roughness of aluminium parts produced by DMLS,Accessedon 26th November 2017,Available at < http://porto.polito.it/2505584/2/2505584.pdf>2.F.Caligano (2014), Materials& Design, Design optimization of supports for overhangingstructures in aluminum and titanium alloys by selective laser melting, 64, pp203-2133. G.
Strano, L. Hao, R.M. Everson, K.E. Evans(2013), Journal of Materials Processing Technology, Surface roughness analysis, modelling and prediction in selective lasermelting, 213, pp. 589-5974. I.
Gibson, D. Rosen, B. Stucker (2015),Additive ManufacturingTechnologies,3D Printing, RapidPrototyping, and Direct Digital Manufacturing, Second Edition, ISBN 978-1-4939-2113-3,pp3-55. J. Jhabvala, E.
Boillat, T. Antignac, R. Glardon (2010)Virtual and Physical Prototyping, On theeffect off scanning strategies in selective laser melting process,5,pp. 99-1096. K.
Zeng (2015) Optimization of supportstructures for selective laser melting, Accessed on 28thNovember 2017, Available at < https://ir.library.louisville.edu/cgi/viewcontent.
cgi?article=3250&context=etd >7. L.Taimisto(2009), Bachelor’s thesis, Processparameters in laser sintering process, Accessed on 24th November2017, Available at < https://www.doria.fi/bitstream/handle/10024/70831/nbnfi-fe201109082361.pdf?sequence=3 >8.
Metal AM (2014), Applications for metalAdditive Manufacturing technology, Accessed on 15th November 2017, Available at < http://www.metal-am.com/introduction-to-metal-additive-manufacturing-and-3d-printing/applications-for-additive-manufacturing-technology/ >9. P. Hanzl,M. Zetek,T. Bakša,T. Kroupa (2015),Procedia Engineering, The Influence of Processing Parameters onthe Mechanical Properties of SLM Parts,100, pp.
1405-141310. S. Bose,D. Ke,H. Sahasrabudhe,A.
Bandyopadhyay (2017), Progress in Materials Science, Additive Manufacturing of Biomaterials, 92, pp 45-11111. S. Dadbakhsh, L.
Hao (2014),TheScientific World Journal ,Effect of LayerThickness in Selective Laser Melting on Microstructure of Al/5?wt. % Fe2O3 Powder Consolidated Parts, 2014.12. W. E. King, A.
T.Anderson, R. M. Ferencz , N. E. Hodge, C. Kamath, S.
A. Khairallah and A. M. Rubenchik (2015),Applied PhysicsReviews, Laserpowder bed fusion additive manufacturing of metals; physics, computational, andmaterials challenges,2,pp1-313. Wikipedia (2015) Selectivelaser melting, Accessed on 12 November 2017, Available at < https://en.
wikipedia.org/wiki/Selective_laser_melting >14. Wikipedia(2013) Sintering, Accessed on 21stNovember 2017, Available at < https://en.wikipedia.org/wiki/Sintering >15. Y.
Tian, D.Tomus,P. Rometsch, X. Wu (2017), Additive Manufacturing, Influences of processing parameters onsurface roughness of Hastelloy X produced by selective laser melting,13,pp103-11216.Y. Kajima,A. Takaichi,T.
Nakamoto,T. Kimura,N. Kittikundecha,Y. Tsutsumi,N. Nomura,A. Kawasaki,H.
Takahashi,T. Hanawa (2017), Journal of the MechanicalBehaviour of Biomedical Materials, Effect of adding support structuresfor overhanging part on fatigue strength in selective laser melting,78,pp. 1-917. 3R Systems 2014, EOSINTM270 – metal,Accessed on 24th November 2017, Available at < http://www.3rsystems.cz/en/3d-print/3d-print-dental/eosint-m270-metal >