Thermomechanical Investigation of Overhang ...

5 downloads 134 Views 1MB Size Report
Jun 9, 2014 - Manufacturing: Powder Characterization, Process. Simulation and Metrology," Proceedings of ASME Early. Career Technical Conference, Vol.
Proceedings of the ASME 2014 International Manufacturing Science and Engineering Conference MSEC2014 June 9-13, 2014, Detroit, Michigan, USA

MSEC2014-4063

THERMOMECHANICAL INVESTIGATION OF OVERHANG FABRICATIONS IN ELECTRON BEAM ADDITIVE MANUFACTURING Bo Cheng, Ping Lu and Kevin Chou Mechanical Engineering Department The University of Alabama Tuscaloosa, AL 35487, USA

KEYWORDS Electron beam additive manufacturing (EBAM), Finite element analysis, Overhang geometry, Thermomechanical modeling; Simulations

ABSTRACT

geometric model. This group of technologies offers many design and manufacturing advantages such as short lead time, design freedom in geometry, and tooling-free productions. Powder-based electron beam additive manufacturing (EBAM) is a relatively new AM technology [1]; it utilizes a high-energy electron beam, as a moving heat source, to melt and fuse metal powder and produce a solid part in a layer-by-layer fashion. In building each layer, the processes involve powder spreading, pre-heating, contour melting and hatch melting. In powder spreading, a metal rake is utilized to uniformly distribute one layer of powder, supplied from the hoppers. Then, pre-heating is applied using a single beam at a high speed (e.g., 14 m/s), with multi-pass scans, to reach a high temperature across the entire powder-bed surface. The purpose of pre-heating is to slightly sinter the powder so to prevent them from expelling when bombarded by high-energy electrons. Thus, pre-heated powder will possess a certain level of porosity. Following preheating, contour-melting and hatch-melting stages take place, during which an electron beam, either multiple split or single, moves across the powder-layer surface tracing the cross-section contour of the model and then raster-scan throughout the inside of the contour at a lower scanning speed (e.g., around 0.5 m/s). Once all layers are completely built, the system is cooled down, remained in vacuum, until it is close to the room temperature before the build chamber can be opened. Then, the entire powder bed is removed from the machine for cleaning: sand blasting to clean off loose sintered-powder from the build part. In post-processing, support structures, if any, will be removed,

Electron beam additive manufacturing (EBAM) is one of powder-bed-fusion additive manufacturing processes that are capable of making full density metallic components. EBAM has a great potential in various high-value, small-batch productions in biomedical and aerospace industries. In EBAM, because a build part is immersed in the powder bed, ideally the process would not require support structures for overhang geometry. However, in practice, support structures are indeed needed for an overhang; without it, the overhang area will have defects such as warping, which is due to the complex thermomechanical process in EBAM. In this study, a thermomechanical finite element model has been developed to simulate temperature and stress fields when building a simple overhang in order to examine the root cause of overhang warping. It is found that the poor thermal conductivity of Ti6Al-4V powder results in higher temperatures, also slower heat dissipation, in an overhang area, in EBAM builds. The retained higher temperatures in the area above the powder substrate result in higher residual stresses in an overhang area, and lower powder porosity may reduce the residual stresses associated with building an overhang. INTRODUCTION Additive manufacturing (AM) based on “layer-adding” fabrications is a group of technologies, by which a physical solid part is produced directly from digital data of the part

1 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

mostly by a mechanical means. Recently, Oak Ridge National Laboratory published a video animation of the EBAM process [2] that offers good illustrations of the process details. EBAM is one of a few AM technologies capable of making full-density metallic parts, which drastically broaden AM applications in a wide variety of industries [1,3]. Because of the high energy density, EBAM has the potential to work with many material classes, e.g., aluminum alloys [4], tool steel [5], cobalt-based superalloys [6], Cu [7], and Inconel alloys [8], etc. One titanium alloy, Ti-6Al-4V, was the first material extensively researched [9,10], also widely used, in EBAM technologies for aircraft parts and medical implants. Moreover, Intermetallic groups such as titanium aluminide have also been studied for EBAM applications, but the process development was considered very time consuming [11]. Another intriguing EBAM capability is to fabricate complex geometries and structures (e.g., meshed, porous, cellular). Schwerdtfeger et al. applied EBAM to fabricate a unique structure that exhibits the “auxetic behavior,” a negative Poisson’s ratio, good for high impact and shear resistances [12]. Despite of an intense interest in attainable part accuracy by EBAM [19], few studies have been emphasized on the geometric aspects in EBAM. Cooke and Soon [20] studied the geometric accuracy of metallic test parts manufactured by EBAM and other powder-based metal AM processes. The authors reported that overall, the observed errors of EBAM parts are significantly larger, at least an order of magnitude, than those of typical machined parts. The errors seem to be repeatable, providing opportunities for compensation strategies. In research conducted by Koike et al. [21], an enlarged view of the gauge section of the EBAM Ti-6Al-4V specimen shows a rough exterior appearance comparable to as-cast counterparts. The surface of typical EBAM specimens generally shows a feature of rippled layers. EBAM is a complicated process and process physics modeling of EBAM is only sparsely found in literature. Zah and Lutzman developed a simplified heat transfer model of EBAM, based on a finite element (FE) method, to study the melt-pool geometry [22]. Shen and Chou recently initiated a study to investigate the thermal phenomenon of the EBAM process. An FE heat transfer simulation incorporating Gaussian heat flux distribution, fusion latent heat, temperature-dependent thermal properties was developed to investigate the thermal response of metal powder in EBAM [23]. Moreover, process measurements such as temperatures have not been frequently reported for EBAM either. Zah and Lutzman applied thermocouples, attached to the build plate, to evaluate temperature response during EBAM to evaluate the model developed by this research group [22]. Dinwiddie et al. [24] and Rodriguez et al. [25] applied different thermal imaging systems to attempt for in-situ process monitoring and for feedback controls, respectively. On the other hand, Price et al. reported of using a near-infrared thermal imager for process temperature measurements in EBAM [26]. The result indicates that using an infrared thermography, with a correct spectral range, for temperature measurements in EBAM is feasible. One of the advantages in AM technologies is freedom in part geometry designs, virtually no limit (except resolution). However, for some AM processes, e.g., the material extrusion

type AM, also known as fused deposition modeling (FDM), part designs with overhang (or undercut) geometry are considered not favorable. In such an AM process, a part overhang will require the so-called “support structure” to carry the weight of the overhang portion. However, the support structure has to be removed during post-processing, which may be time consuming and add additional production cost. Hence, overhang geometry may be considered as undesired design geometries. Some approaches have been used to tackle support structure removals, e.g., designing break-away structures or using water solvable materials in FDM. On the other hand, powder-bed-fusion AM processes including EBAM are considered not requiring a support for an overhang because the powder bed itself is able to bear the weight of a built overhang. However, in common practices, still, the general rule is to arrange support structures underneath an overhang, or defects such as warping may occur. Defects such as warping associated with overhang geometry problems in EBAM have not been frequently studied in literature. Recently, Vora et al. attempted to benchmark the problems with overhangs and reported cases that overhang warping occurs [27]. For laser-based powder-bed AM processes such as selective laser sintering or melting (SLS or SLM), there are a few studies related to support structures in literature. The focus was, however, on the design and fabrication of support structures. Jihabvala et al. reported using pulse laser, instead of a continuous mode, to fabricate support structures and claimed the fabricated supports are much easier to be removed [28]. Yan et al. studied “cellular lattice” structures for support and investigated the effect of cellular geometry parameters. The evaluation criterion was the ease of removal and access. The authors also intended to use cellular-type support to minimize the material volume needed for support [29]. In general, the above mentioned studies have not comprehensively addressed the aspect of the need of supports for an overhang and for suitable support types. EBAM has a potential to be a cost-effective alternative to conventional discrete component manufacture for high-value, small-batch, and custom-designed metallic parts. However, one of the challenges in part designs is inevitable overhang geometries. The overhang area will require support structures, or there will be defects around the overhang area, shown in Figure 1, including size inaccuracy and, most noticeably, distortions.

Figure 1. An example of overhang model and associated warping defect.

2 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

Furthermore, post-processing for support removal can be of a high cost. Figure 2 below shows an example of support patterns from a complex build part (condyle of a temporomandibular joint prosthesis pair). The part has such geometry inevitably requiring support structures in certain areas, which was obtained by default from commercial software. Figure 2b displays the area of the support showing contact (solid) with the build part. Though a mechanical tool may be used to remove the support structures, small piece by piece (Figure 2c), it is labor intense and time consuming, with a risk of damaging the part surface.

thermal conductivity compared to a solid, and thus, heat dissipation around the overhang area may be less efficient compared to an area with a solid substrate. The assumption of overhang defect causes, therefore, is that in the powder area beneath an overhang, the ineffective heat dissipation and repeated thermal gradient cycles result in greater thermal stresses, leading to warping defects. Hence, the research needs would be a fundamental thermomechanical study of EBAM for temperature and stress evaluations. The knowledge obtained may be applied to facilitate investigations of different overhang configurations, thermal and mechanical responses, and to enable effective designs for heat load supports and simultaneously for ease of support structure removals. The objective of this research is to understand the thermal and mechanical characteristics in fabricating overhang layers during the EBAM process. A three-dimensional (3D) finite element (FE) model was developed to simulate the thermal and mechanical phenomena in EBAM in a raster-scanning pattern across an overhang area.

(a)

Support

MODELING APPROACH AND SIMULATION METHODOLOGY A 3D FE thermomechanical model was developed by using ABAQUS to simulate the process temperature history, residual stress distributions and deformation in a multi-layer raster scanning and deposition on an overhang model. Figure 3(a) shows the geometry of the overhang model, which has a substrate dimension of 8 × 8 × 10 mm (length, width, and height). A smaller model was used in order to keep a reasonable computational time. The thickness of the substrate is small comparing to the actual size of common parts; therefore, only the top section of the actual substrate was modeled. During simulations, powder layers were added to the substrate sequentially. The model volume was composed of a powder substrate on one side and a solid substrate on the other side (half and half) to represent an overhang fabrication (on top of the powder substrate), and the powder layers added were Ti6Al-4V particles with a given levels of porosity, with each layer of 70 µm thickness. The FE simulations consisted of a preheat phase, a melting phase, and a cooling phase. The preheating phase (also called initial phase) was simplified and considered as the initial thermal conditions. During the multi-layer deposition process, the layers to be deposited were not activated until the melting phase on that particular layer started. An 8-node coupled thermal-displacement type of C3D8T was used to mesh the substrate and the powder layer, and the element size in the scanning area was 0.2 × 0.2 × 0.05 mm. The total element number for the double layer deposition model was 6084. Figure 3(b) displays the model after meshing.

(b)

Support

(c)

Figure 2. (a) and (b) EBAM build condyle (front and back sides) showing support structures, and (c) same part showing partial support removed. There are commercial software packages, e.g., [30], which can generate support structures with a variety of options (patterns/size); however, there is no clear guideline of support design, relying on trial and error and experiences. Moreover, it needs to be pointed out that those support designs were developed more for the weight-carrying purpose, to avoid overhang area deformation due to gravity. It does not address the powder bed conditions for the process like EBAM. Moreover, additional questions are: what is the root cause of overhang defects, and what type of “support” configurations will function to minimize warping defects, and yet, without burdening post-processing? To effectively design necessary supports, it is essential to understand the source of defects associated with overhang geometries in EBAM. It is argued that for EBAM, heat load, instead of gravity load, is more the source of the problems. Because the powder bed (with sintered powder) has a far poor

3 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

substrates horizontally with a raster length of 3.2 mm. The raster scanning area is 3.2 × 3.2 mm2 and the width of the raster is the same as the beam diameter, assumed as 0.4 mm. There were total 9 scanning paths along the x direction (longitudinal) on each layer. The stress and temperature characteristics along these scanning paths were analyzed from the simulation results. Since each scan (or layer) was considered as an intermediate scan (or layer) in a continuous multi-layer part building, the previous layer was considered to be transformed into solid bulk materials, meaning solidified from the powder in the previous melting scan.

(a)

(a)

(b)

(b)

Figure 3. (a) Part geometry, and (b) element meshing of the EBAM overhang model. The material properties used, including both thermal and mechanical, temperature dependent, were from previous studies, [23,31] and [32], respectively. In particular, the thermal conductivity of Ti-6Al-4V solid and powder were measured at different temperatures, up to 750 °C. It is worth to point out that the thermal conductivity of the Ti-6AL-4V powder is much smaller than the solid counterpart, e.g., 0.63 W/m·K vs. 6.2 W/m·K at 25 °C. During the melting and cooling phase, the bottom surface of the model was fully constrained as the boundary condition, Figure 4(a). The contact between the substrate material and the powder layers was assumed to be rigid. Convection between the powder layer and environment was not included due to the vacuum environment. Hence, only the radiation was considered in the heat transfer between the powder/part and surroundings. The substrate and the powder layer were assigned with a uniform temperature distribution of Tpreheat as the initial thermal condition. The temperature of the substrate bottom was confined with a constant temperature of Tpreheat as the thermal boundary condition. For the multi-layer raster scan, two layers of powder were simulated; the raster scan is shown in Figure 4(b). The left side of the dashed line is above top of a powder substrate (the overhang area), while the right side is on top of a solid substrate. During the simulation, the electron beam traveled along the top surface of the powder and solid

Figure 4. (a) Boudary conditions, and (b) raster scan pattern applied in multi-layer analysis (the left side of the dashed line is above of a powder substrate, while the right side is on top of a solid substrate). After the melting phase, the simulation of the cooling phase was followed. In the multi-layer raster scan, there were two cooling steps; the first cooling phase was the part temperature self-balancing corresponding to a few seconds interval before a new powder-layer spreading, and the second one was the final cooling step to the room temperature. The time for the first cooling step was assumed as 3 sec which is approximately the duration for the next powder layer spreading during the EBAM process. The thermal boundary conditions were the same as those in the melting step. The last cooling step was realized using general static analysis method in which the temperature of the EBAM part dropped to room temperature of 25 °C. Results and Discussion FE analysis was first performed for the overhang model with a powder porosity of 50%. Figure 5 shows the temperature

4 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

distribution at the top surface in the model at the end of a twolayer deposition. It can be seen that the maximum temperature occurs at the end of the scanning with the magnitude around 2900 °C. In addition, the longer lasting of high temperatures in the scanned area of the left part (powder substrate) is significantly higher than the right side of the part (solid substrate), which is due to the low thermal conductivity of the powder substrate.

(b)

Figure 6. (a) Temperature history, and (b) magnified area around the peak temperature of the 2 nodes above powder and solid substrates. Figure 7 shows the temperature profile around the electron beam center location (approximately at 0 mm in the distance axis) when the beam traveled above the powder and solid substrates during the scanning path 2 and 7. Both Figure 7(a) and 7(b) show that a higher maximum temperature occurred when the beam traveled above the powder substrate; this is again attributed to the significant lower thermal conductivity of the powder substrate.

Figure 5. Temperature contour (top view) at the end of twolayer deposition. Figure 6 displays temperature history of 2 nodes (A and B in Figure 4) above the powder and solid substrates. Nodes A and B are on the scanning path 4, with the same distance to the center of the EBAM part model (also the interface between the powder and solid substrates). Node A is above the powder substrate, while B above the solid substrate. The result shows that the peak temperature of Node A is about 200 °C higher than Node B, also the cooling rate of Node A is noticeably slower than that of Node B.

(a)

(a)

(b)

Figure 7. Temperature profiles around electron beam center (0 mm) above solid and powder substrates for path 2 and path 7.

5 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

Figure 8 illustrates the melt pool size shape when the electron beam traveled above the powder and solid substrates at the initial (P1) and final (P9) scanning paths. The jagged meltpool shape was due to element meshing, not fine enough. Table 1 further lists the dimensions of the melt pools. It can be noted when the beam traveled from the powder substrate to the solid substrate area, the melt pool sizes became smaller due to a higher thermal conductivity of the solid substrate.

Figure 10(a) displays the von Mises stress evolution of the nodes (Node A and B, same nodes as in Figure 6) above the solid and powder substrates. The results indicate that the maximum von Mises stress at Node A is around 190 MPa higher than that of Node B after the final cooling process. It has been indicated earlier that the maximum temperature of Node A (corresponding to the overhang area) is higher than Node B during the scanning process. The residual stress after the final cooling is resulted from the temperature at the end of the second layer deposition, which is consistent with the observation from the temperature results of the two nodes (Figure 6).

(a)

Figure 8. Melt pool size and shape in the powder and solid substrates at initial and final scanning paths.

(b)

Table 1. Melt pool size dimension (unit: µm). Melt pool Path 1 Path 9 Powder Solid Powder Solid Length 950 900 1180 1100 Width 380 375 400 380 Depth 146 144 160 150 Figure 9 illustrates the von Mises stress distribution in the overhang EBAM model after the final cooling stage. It can be found that the maximum von Mises stress occurs at the end of electron beam scanning location with the magnitude of about 1250 MPa, which exceeds the yield strength of Ti-6Al-4V alloy, and thus, plastic deformation will occur in this region after the EBAM process.

Figure 10. (a). von Mises stress evolution, and (b) magnified time zone around the peak stress of 2 nodes above solid and powder substrates. Figure 11 shows the von Mises stress profile at different scanning paths on one-layer deposition after final cooling process. It can be noted that the average residual stress on the scanning path increases with the scanning path number. The trend of residual stress on each path is relatively uniform along the longitudinal direction in the individual left or right sides, except the 7th scanning path, which is close to the edge of the scanning area. Moreover, the stress on the left part (powder substrate) is greater than that on the right part (solid substrate).

Figure 9. von Mises stress contour after final cooling (MPa).

6 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

the difference in temperatures in the areas between above the powder and above the substrate.

Figure 11. von Mises stress at different paths from one-layer deposition after final cooling.

Figure 13. Temperature profiles on the scanning paths for two different powder porosity levels.

Figure 12 displays the von Mises stresses at the top surface after the cooling process comparing the single and double-layer depositions. The results indicate that the average residual stresses are slightly different between single vs. double-layer depositions. For example, the average residual stress on the 2nd scanning path for the single layer deposition model is about 480 MPa after final cooling, and the average residual stress decreases to about 470 MPa for the double layer deposition. It is also noted that the residual stress on the left half (corresponding to the overhang part) is consistently greater than that in the right half (above solid substrate) for all scanning paths.

Figure 14 shows the von Mises stress contours for the EBAM model with porosity of 35% and 50%. It can be noted that with the increase of powder porosity, the residual stress in the left half (above powder substrate) of the model becomes larger, 1252 MPa vs. 1074 MPa, comparing to the lower porosity model, which is due to the smaller thermal conductivity of the powder with a higher porosity level.

(a)

Solid Powder

(b)

Figure 12. von Mises stress after final cooling deposited with single and double layers.

Figure 14. von Mises stress contours after final cooling for different powder porosity levels: (a) 50%, and (b) 35%

Simulations of cases with different porosity levels, 50% vs. 35%, were conducted to examine the porosity effects. Figure 13 shows the temperature profiles on the paths for the EBAM model with different porosity levels. It can be noted that the average temperature on the scanning path increases slightly with the scanning paths. It is also clearly evident that for the powder with a larger porosity, the temperatures above the overhang area are much higher, over 300 °C in average, so is

Figure 15 displays the von Mises stress on the three paths (2nd, 5th, and 7th) of the top layer after the final cooling process for the EBAM model with different powder porosity levels. It can be noted that the average von Mises stress in the overhang area increases, on all scanning path, when the powder porosity is larger; for example, the average von Mises stress on the 2nd scanning path (powder substrate side) increases from about 350 MPa to about 500 MPa when the powder porosity increases from 35% to 50%.

7 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

[3] Biamino, S., A. Penna, U. Ackelid, S. Sabbadini, O. Tassa, P. Fino, M. Pavese, P. Gennaro, and C. Badini, (2010), “Electron beam melting of Ti-48Al-2Cr-2Nb alloy: Microstructure and mechanical properties investigation,” Intermetallics, 19(6), pp. 776-781. [4] Yu, P., M. Qian, D. Tomus, C.A. Brice, G.B. Schaffer, and B.C. Muddle , (2009), “Electron beam processing of aluminium alloys,” Materials Science Forum, 618-619, pp. 621-626. [5] Cormier, D., O. Harrysson, and H. West, (2004), “Characterization of H13 steel produced via electron beam melting,” Rapid Prototyping Journal, 10(1), pp. 35-40. [6] Gaytan, S.M., L.E. Murr, E. Martinez, J.L. Martinez, B.I. Machado, D.A. Ramirez, F. Medina, S. Collins,. and R.B. Wicker, (2010), “Comparison of microstructures and mechanical properties for solid and mesh cobalt-base alloy prototypes fabricated by electron beam melting,” Metallurgical and Materials Transactions A, 41(12), pp. 3216-3227. [7] Ramirez, D.A., L.E. Murr,E. Martinez, D.H. Hernandez, J.L. Martinez, B.I. Machado,F. Medina,P. Frigola, and R.B. Wicker, (2011), “Novel precipitate-microstructural architecture developed in the fabrication of solid copper components by additive manufacturing using electron beam melting,” Acta Materialia, 59, pp. 4088-4099. [8] Murr, L.E., E. Martinez, S.M. Gaytan, D.A. Ramirez, B.I. Machado, P.W. Shindo, J.L. Martinez, F. Medina, J. Wooten, D. Ciscel, U. Ackelid, and R.B. Wicker, (2011), “Microstructural architecture, microstructures, and mechanical properties for a Nickel-base superalloy fabricated by electron beam melting,” Metallurgical and Materials Transactions A. [9] Parthasarathy, J., B. Starly, S. Raman, and A. Christensen, (2010), “Mechanical evaluation of porous titanium (Ti6Al4V) structures with electron beam melting (EBM),” Journal of the Mechanical Behavior of Biomedical Materials, 3(3), pp. 249-259. [10] Murr, L.E., E.V. Esquivel, S.A. Quinones, S.M. Gaytan, M.I. Lopez, E.Y. Martinez, F. Medina, D.H. Hernandez, E. Martinez, J.L. Martinez, S.W. Stafford, D.K. Brown, T. Hoppe, W. Meyers, U. Lindhe, and R.B. Wicker, (2009), “Microstructures and mechanical properties of electron beam-rapid manufactured Ti-6Al-4V biomedical prototypes compared to wrought Ti-6Al-4V,” Materials Characterization, 60(2), pp. 96-105. [11] Sabbadini, S., O. Tassa, P. Gennaro, and U. Ackelid, (2010), “Additive manufacturing of gamma titanium aluminide parts by electron beam melting,” TMS 2010. 139th Annual Meeting & Exhibition. Materials Processing and Properties, Seattle, WA, pp. 267-274. [12] Schwerdtfeger, J., P. Heinl, R.F. Singer, and C. Körner, (2010), “Auxetic cellular structures through selective electron-beam melting,” Physica Status Solidi (b), 247, pp. 269-272. [13] Gulzar, A., J.I. Akhter, M. Ahmad, G. Ali, M. Mahmood, and M. Ajmal, (2009), “Microstructure evolution during surface alloying of ductile iron and austempered ductile iron by electron beam melting,” Applied Surface Science, 255, pp. 8527-8532.

Figure 15. von Mises stress on different scanning paths of the top layer after final cooling for the models with different porosity levels. CONCLUSIONS In this study, a 3D FE thermomechanical model was developed for temperature and stress simulations in EBAM, particularly, applied to investigate temperatures and stresses when fabricating an overhang feature, which by nature is above a powder substrate. Ti-6Al-4V powder materials with two different porosity levels, resulted from different preheating conditions, were studied. The major findings can be summarized as follows. 1. The poor thermal conductivity of Ti-6Al-4V powder results in higher temperatures, also slower heat dissipations, in overhang areas during an EBAM build, 2. The retained higher temperatures in the area above the powder substrate, when building an overhang feature in EBAM, result in higher residual stresses after final cooling, and 3. A smaller porosity (e.g., 35% vs. 50%) may noticeably reduce the residual stresses, also process temperatures, associated with building an overhang. Future work will investigate stress and defect severity of different overhang patterns and study various configurations for designs of effective functional support structures for any specific types of overhang. ACKNOWLEDGEMENTS The materials presented in this paper are supported by NSF (Award No. 1335481) and NASA (Award No. NNX11AM11A). This research is in collaboration with Marshall Space Flight Center (MSFC), Advanced Manufacturing Team. Helpful discussions with K. Cooper, J. Lydon and M. Babai at MSFC are acknowledged. REFERENCES [1] Gong, X., T. Anderson and K. Chou, (2012), “Review on Powder-based Electron Beam Additive Manufacturing Technology,” Proceedings of ASME 2012 International Symposium on Flexible Automation (ISFA), St. Louis, MO, June 18-21, 2012, ISFA2012-7256. [2] http://www.youtube.com/watch?v=BxxIVLnAbLw, Oak Ridge National Laboratory, accessed in June 2012.

8 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME

[14] Murr, L.E., S.M. Gaytan, F. Medina, E. Martinez, D.H. Hernandez, L. Martinez, M.I. Lopez, R.B. Wicker, and S. Collins, (2009), “Effect of build parameters and build geometries on residual microstructures and mechanical properties of Ti-6Al-4V components built by electron beam melting (EBM),” Solid Freeform Fabrication Proceedings, Austin, TX, August 3-5, 2009, pp. 374-397. [15] Bontha, S., N.W. Klingbeil, P.A. Kobryn, and H.L. Fraser, (2009), “Effects of process variables and size-scale on solidification microstructure in beam-based fabrication of bulky 3D structures,” Materials Science and Engineering A, 513-514 (15), pp. 311-318. [16] Svensson, M., and U. Ackelid, (2010), “Titanium alloys manufactured with electron beam melting mechanical and chemical properties,” Medical Device Materials V Proceedings of the Materials and Processes for Medical Devices Conference, Minneapolis, MN, pp. 187-194. [17] Gaytan, S.M., L.E. Murr, D.H. Hernandez, E. Martinez, S.A. Quinones, F. Medina, and R.B. Wicker, (2009), “Structure-property-process optimization in the rapid-layer manufacturing of Ti-6A1-4V components by electron beam Melting,” TMS 2009 - 138th Annual Meeting and Exhibition, San Francisco, CA, February 15-19, pp. 363369. [18] Khalid Rafi, H., N.V. Karthik, T.L. Starr, and B.E. Stucker, (2012), “Mechanical Property Evaluation of Ti6Al-4V Parts Made Using Electron Beam Melting,” 23rd Annual International Solid Freeform Fabrication Symposium, Austin, TX, Aug. 6-8, 2012, pp. 526-535. [19] Liu, W., L. Li, and K. Kochhar, (1998), “A method for assessing geometrical errors in layered manufacturing. Part 2: Mathematical modeling and numerical evaluation,” The International Journal of Advanced Manufacturing Technology, 14, pp. 644-650. [20] Cooke, A.L., and J.A. Soons, (2010), “Variability in the Geometric Accuracy of Additively Manufactured Test Parts,” Proceedings of 21st Annual International Solid Freeform Fabrication Symposium, Austin, TX, August 810, 2010, pp. 1-12. [21] Koike, M., K. Martinez, L. Guo, G. Chahine, R. Kovacevic, and T. Okabe, (2011), “Evaluation of titanium alloy fabricated using electron beam melting system for dental applications,” Journal of Materials Processing Technology, 211, pp. 1400-1408. [22] Zah, M. F., and S. Lutzmann, (2010), “Modelling and simulation of electron beam melting,” Production Engineering, 4(1), pp. 15-23. [23] Shen, N., and K. Chou, (2012), “Thermal Modeling of Electron Beam Additive Manufacturing Process – Powder Sintering Effect,” Proceedings of ASME International Manufacturing Science and Engineering Conference, Notre Dame, IN, June 4-8, 2012, MSEC2012-7253. [24] Dinwiddie, R., Dehoff, R., Lloyd, P., Lowe, L, Ulrich, J., (2013) “Thermographic In-Situ Process Monitoring of the Electron Beam Melting Technology used in Additive Manufacturing,” Proceedings of SPIE, v8705, pp. 87050K1 – 87050K-9. [25] Rodriguez, E., Medina, F., Espalin, D., Terrazas, C., Muse, D., Henry, C., MacDonald, E., Wicker, R., (2012)

“Integration of a Thermal Imaging Feedback Control System in Electron Beam Melting,” Proceedings of 23rd Solid Freeform Fabrication Symposium, pp. 945-961. [26] Price, S., K. Cooper, and K. Chou, (2012), “Evaluations of Temperature Measurements by Near-infrared Thermography in Powder-based Electron-beam Additive Manufacturing,” 23rd Annual International Solid Freeform Fabrication Symposium, Austin, TX, Aug. 6-8, 2012, pp. 761-773. [27] Vora, P., D. Stapleton, M. Kamran, F. Derguti, I. Todd, and N. Hopkinson, (2012) “Benchmarking Metal Powder Bed Additive Manufacturing Processes (SLM and EBM) to Build Flat Overhanging Geometries Without Supports,” 23rd Annual International Solid Freeform Fabrication Symposium, Austin, TX, Aug. 6-8, 2012. [28] Jhabvala, J., E. Boillat, C. André, and R. Glardon, (2012), “An innovative method to build support structures with a pulsed laser in the selective laser melting process,” The International Journal of Advanced Manufacturing Technology, 59(1-4), pp. 137-142. [29] Yan, C., L. Hao, A. Hussein, and D. Raymont, (2012), “Evaluations of cellular lattice structures manufactured using selective laser melting,” International Journal of Machine Tools and Manufacture, 62(0), pp. 32-38. [30] http://software.materialise.com/magics-0, Software for Additive Manufacturing, Materialise, accessed in June 2012. [31] Gong, X., B. Cheng, S, Price and K. Chou, (2013), "Powder-Bed Electron-Beam-Melting Additive Manufacturing: Powder Characterization, Process Simulation and Metrology," Proceedings of ASME Early Career Technical Conference, Vol. 12, pp. 59-66, 2013. [32] Shen, N. and K. Chou, (2012), “Simulations of Thermomechanical Characteristics in Electron Beam Additive Manufacturing,” Proceedings of ASME 2012 International Mechanical Engineering Congress and Exhibition (IMECE), Houston, TX, Nov. 9-15, 2012, IMECE201288276.

9 Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/07/2015 Terms of Use: http://asme.org/terms

Copyright © 2014 by ASME