Available via license: CC BY 3.0
Content may be subject to copyright.
Journal of Physics: Conference Series
PAPER • OPEN ACCESS
A new 3D printing method based on non-vacuum electron beam
technology
To cite this article: Shuhe Chang et al 2018 J. Phys.: Conf. Ser. 1074 012017
View the article online for updates and enhancements.
This content was downloaded from IP address 216.74.111.254 on 09/10/2018 at 13:46
1
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
A new 3D printing method based on non-vacuum electron
beam technology
Shuhe Chang1, Stefan Gach2, Aleksej Senger2, Haoyu Zhang1 and Dong Du1, 3
1 Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China;
2 Welding and Joining Institute, RWTH Aachen University, Aachen Germany
3 E-mail: dudong@tsinghua.edu.cn
Abstract. Electron beam freeform fabrication (EBF3) is one the of rapid manufacturing
methods, which produces metal parts using an electron beam and wire feed unit in a layer
additive fashion without need of mold or jigs. The main disadvantage of EBF3 is that the
process is carried out in a chamber, which not only takes a lot of time to evacuate but also
limits the size of the part being printed. Non-vacuum electron beam (NVEB) welding is widely
used in industry field as a reason for its high production volumes. So a NVEB 3D printing
device based on a non-vacuum electron beam welding machine from SST is designed. A serial
of experiments is carried out to get a deep understand of 3D printing processing procedure
using non-vacuum electron beam system and find a suitable process window. Result shows that
droplet transfer mode is one of the most important parameters which determines the quality of
the deposition.
1. Introduction
Now days, there are several ways for metal additive manufacturing (AM), such as EBF3 (Electron
Beam Freeform Fabrication), EBSM (Electron Beam Selective Melting), SLM (Selective Laser
Melting), WAAM (Wire and Arc Additive Manufacturing), etc. [1] [2] [3] [4]. There are also some
other nicknames for AM, like 3-D printing, solid freeform fabrication or direct digital manufacturing
[5] [6] [7]. After more than 20 years development, additive manufacturing can create sophisticated
parts directly, without need of molds or jigs [4]. Comparing to other rapid manufacturing methods,
EBF3 has a higher deposition rate and better quality [8-9]. It has advantages on energy efficiency, size
adaptability, material cost and particular on suitability in 0-g environment. This technology is
especially suitable for manufacturing reactive alloys, such as aluminum or titanium.
The mechanism of EBF3 is illustrated in figure 1. In the process, the wire is fed into the baseplate
and melted by the electron beam. The electron beam gun and the wire feed or the baseplate can be
manipulated by the same way as CAM, which has a predetermined trajectory stored in a computer.
Solidification occurs after the electron beam passes. Repeating this progress layer by layer, a near-net
shape part will be manufactured at last. It is much simpler than the traditional subtractive machining
methods [10].
But the main disadvantage of EBF3 is that the process is carried out in a chamber, which not only
takes a lot of time to evacuate but also limits the size of the part being printed [11]. Non-vacuum
electron beam welding has been firstly used in automotive industry as a reason for its high production
volumes. The electron beam welding at atmospheric pressure is now used in many fields, such as
welding laboratories, equipment construction [11].
2
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
Researchers in the Institute for Material Science at Leibniz University Hannover construct a 3D
printing device using NVEB welder PTR NV-EBW 25-175 TU [13]. But only some basic deposited
parts are shown. To get a deep understand of 3D printing processing procedure using non-vacuum
electron beam system and find a suitable process window for it, a serial of experiments is carried out
at ISF in RWTH-Aachen.
2. Experimental layouts
The experiments were carried out on a non-vacuum electron beam-welding machine from SST:
EBONOVA G300 DSS. It was used to do high-speed electron beam welding at atmospheric pressure
[14]. This machine has following specifications:
Maximum beam power – 30kw
Maximum beam current – 200mA
Maximum accelerating voltage – 170kv
Working distance(nozzle to work piece) – 5-25mm
Building volume – 1060mm x 700mm x 1340mm
2.1. Adaptation of the NVEBW system for additive manufacturing
To transform the welding machine to a 3D printing device, as shown in figure 2, a wire feed unit was
installed to feed the material into the molten pool. As shown in figure 3, a collimator is placed behind
the motor to make the wire straight enough to be fed into the right position of the molten pool. In the
first few pre-experiments without collimator, the wire cannot be melt completely, illustrated in figure
4.
Figure 2. System composition of 3d printer
based on NVEBW machine.
Figure 3. Wire feed unit.
Figure 1. EBF3 manufacturing system.
3
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
Figure 4. Deposition without collimator.
Figure 5. Penetration in baseplate.
Pay attention that 4 brackets was placed between the operation platform and baseplate for
protection. As shown in figure 5, sometimes the high power electron beam can penetrate the baseplate
as a reason of too much heat input.
More stable clamps should be adopted in the future experiments. The waviness of the part built by
EBF3 is big [15-16]. The printing procedure will be disrupted when there is a large deformation of the
baseplate. The baseplate need to be fixed very tightly to stand up to the residual stress.
3. Experiment procedure
Process repeatability is one of the most important factors that limit the large-scale application of AM
[16]. The main aim of the research is to get a deep understand of 3D printing processing procedure
using non-vacuum electron beam system and find a suitable process window for it.
3.1. Pre-experiments
Several pre-experiments were carried out to get basic knowledge of the processing procedure. One of
the important advantage of EBF3 is the highest deposition rates, which makes it suitable for making
large-scale metal structures [18]. To get a high deposition rate, the accelerating voltage was set to
150kv. The energy loss of electron beam is around 50% due to the collision with atmosphere and inner
wall of the nozzle.
According to the cooling capacity, the beam current was set to 30mA. In the pre-experiments, the
deposition figure was just the simplest line. The deposition length is 150mm.
First, the deposition was carried on a high translation and feeding speed. The translation speed was
500mm/min. The feeding speed was set to 1m/min on the control box. (Due to the ration in the feeding
wheel, the real feeding speed was 1.34m/min. To make the comparison easier, the feeding speed
mentioned below all meant the display speed on the control box.) As shown in figure 6, the wire was
not melt completely at some places due to the lack of heat input.
Figure 6. Deposition on a high translation and
feeding speed.
Figure 7. Deposition on a low translation and
feeding speed.
4
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
On the second pre-experiments, the translation speed was decreased to 400mm/min. The feeding
speed was set to 0.8m/min. The deposition result was shown in figure 7. The wire was melt completely
onto the baseplate. Further experiments was carried out to get a better understand of the influence of
feeding speed as shown below.
3.2. Influence of wire feeding speed
Manufacturing cost can be greatly reduced with a high deposition rate. But the wire is not melt
completely when the feeding speed is too high. To optimize the deposition rate of this NVEB machine,
several experiments were carried out using different feeding speed.
Table 1. Deposition with different feeding speed.
No.
1
2
3
4
5
6
Translation speed(m/min)
0.40
0.40
0.40
0.40
0.40
0.40
Wire feeding speed(m/min)
0.70
0.75
0.80
0.85
0.90
0.95
As shown in table 1, the wire feeding speed varied from 0.70m/min to 0.95m/min. The beam
current, accelerating voltage, deposition length was the same with the pre-experiments.
3.3. Influence of deposition gap between layers
In the deposition procedure, the wire is fed into the moving molten pool, which forms a line. A layer is
formed line by line. The final part is formed layer by layer. The deposition gap between layers is
important to the quality of the parts. To get a better understand of the influence of deposition gap on
part quality, three 10-layer walls with single line were deposited. The deposition gaps between layers
in each group are 0.5mm, 0.6mm, 0.7mm respectively. The translation speed was 400mm/min. The
feeding speed was set to 0.9m/min. Other parameters is the same with the pre-experiments.
3.4. Influence of droplet transfer mode
In welding process, droplet transfer mode is very important to the final quality of the seam [19]. The
wire is fed into different places of the molten pool with the variation of the wire feed unit position. An
experiment was designed to find the suitable transfer modes for 3D printing using NVEB machine.
Droplet transfer mode is mainly decided by the distance between the baseplate and the intersection
of the center line of the electron beam and the extension line of the wire. Because a, b and α cannot be
changed easily. The distance is changed by adjusting the value of c from 12.5mm to 18mm.
As we can see from figure 8, a, b, c were measured before each experiment as follows, 8mm,
11mm, 62°. From formula (1), h is calculated，which varies from -1.3mm to 4.2mm.
h=c-a-b*cot(α) (1)
A camera was fixed on the nozzle by aluminum elements to take pictures of the transfer droplets.
This camera can shot a video of 1280*720 with a frame rate of 50fps.
Figure 8. Position relation between the wire feed unit and the nozzle.
5
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
4. Results & discussion
4.1. Influence of wire feeding speed
The result is shown in figure 9. In deposition of the sixth line, the wire feed unit didn’t work properly.
So another line was processed with the same parameters.
Figure 9. Deposition with different feeding speed.
As we can see from line 1~3, the deposition is not consistent due to lack of material. And big
droplet forms when too much wire is fed into the molten pool. Within the translation speed of
400mm/min, a wire feeding speed during 0.85~0.9m/min seems suitable.
A sample was made to take Macro-photo in the center cross section of the baseplate. As we can see
from figure 10, penetration depth decreases with a higher feeding speed. The baseplate takes less
energy when the wire consumes more energy. SEM or 3D imaging techniques will be used to see the
microstructual defects, such as process-induced porosity, which lead to poor performance without
mastering [20] [21].
Figure 10. Macro image of samples deposited with different feeding speed.
4.2. Influence of deposition gap between layers.
As shown in figure 11, three 10-layer walls were deposited. There is also no obvious difference when
deposited single layer within different deposition gap between layers.
A sample was made to take Macro-photo in the center cross section of the baseplate. The
microstructures of different layers are different, as shown in figure 12. There is grain refinement in the
bottom of the deposition due to heat treatment when depositing the top.
To get more understand on the moving strategy, further experiment focused on influence of
different gap between both lines and layers should be done in the future.
Figure 11. Deposition within different gap
between layers.
Figure 12. Macro images of samples
deposited with different gap between layers.
6
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
4.3. Influence of droplet transfer mode
As we can see from figure 13, deposition quality varied obviously within different deposition distance.
In line 1~5, big droplets formed when there is a long distance between the baseplate and the point
where the wire is melt. Droplets fell into the molten pool mostly by gravity. In line 6~9, the wire was
continuously melted in the center of molten pool within a shorter distance between the baseplate and
the point where the wire is melted. Within the distance decreased more, the wire is melted at the edge
of the molten pool, as shown in line 10~12. The wire bend when touching the baseplate, which makes
the process unstable. The penetration depth decreases just a little due to a longer distance between the
baseplate and the nozzle.
Transfer modes in other experiments is shown in table 2. A distance during 0.2~1.7mm is much
more suitable for deposition using an NVEB machine, which make the droplet transfer time decrease
to 0 as we can see from figure 14.
Table 2. Transfer modes within different distance.
h = 4.2mm
h = 3.7mm
h = 3.2mm
h = 2.7mm
h = 2.2mm
h = 1.7mm
h = 1.2mm
h = 0.7mm
h = 0.2mm
h = -0.3mm
h = -0.8mm
h = -1.3mm
7
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
5. Summary and conclusion
In this project, a NVEB 3D printing device based on a non-vacuum electron beam welding machine
from SST is designed. A serial of experiments are carried out to get a deep understand of 3D printing
processing procedure using non-vacuum electron beam system and find a suitable process window.
The translation speed need to be set under 400mm/min to form a constant molten pool due to the
maximum heat input mentioned above. Some typical figures has been deposited successfully. A
number of controlled experiments were designed.
Here are some conclusion,
The suitable wire feeding speed is around 0.85~0.9m/min. The deposition is not consistent due
to lack of material with a low feeding speed. Big droplet forms when too much wire is fed into
the molten pool.
A distance (h: the distance between the baseplate and the point where the wire is melt) during
0.2~1.7mm is much more suitable for deposition using an NVEB machine. Big droplets
formed when there is a long distance between the baseplate and the point where the wire is
melt. The wire is melted at the edge of the molten pool within a short distance, which make
the processing unstable.
There is also no obvious difference when deposited single layer within different deposition
gap between lines or layers.
To get more understand on the moving strategy, further experiment focused on influence of
different gap between both lines and layers should be done in the future. A powerful baseplate cooling
device need to be designed to get a higher deposition rate instead of waiting several minutes between
each lines and layers. The wire feed unit need to be synchronized with the electron beam generator.
Then more complicated figures can be test to find the potential of this NVEB 3D printing device.
Acknowledgement
The National Key Research and Development Program of China (2017YFB1103100).
Project xx--079 supported by Ministry of Industry and Information Technology of China.
Tsinghua Fudaoyuan Research Fund.
References
[1] Huang S H, Liu P, Mokasdar A and Hou L 2013 The International Journal of Advanced
Manufacturing Technology 67(5-8) 1191-1203
[2] Gao W et al. 2015 Computer-Aided Design 69 65-89
[3] DebRoy et al. 2017 Progress in Materials Science
[4] Herderick E 2011 Materials Science and Technology 1413-1425
Figure 13. Deposition with different
droplet transfer modes.
Figure 14. Droplet transfer time rises with
the increase of h.
8
1234567890 ‘’“”
MEIE 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1074 (2018) 012017 doi :10.1088/1742-6596/1074/1/012017
[5] Berman B 2012 Business horizons 55(2) 155-162
[6] Chiu W K and Yu K M 2008 Computer-Aided Design 40(12) 1080-1093
[7] Guessasma et al. 2015 International Journal for Simulation and Multidisciplinary Design
Optimization 6 A9
[8] https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20160006912.pdf
[9] https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20080013538.pdf
[10] Zalameda et al. 2013 In Thermosense: Thermal Infrared Applications XXXV 8705 87050M
[11] https://www.researchgate.net/profile/Wessam_Elnaggar/publication/271076327_WESSAM_M
Sc_Thesis_AIEBCQ_200607/links/54bd4bd90cf218da9391ae4b.pdf
[12] Hassel et al. 2015 In Proceedings of VIII International Conference on Beam Technologies and
Laser Application 82-90
[13] G Klimov, N Murray and A Beniyash 2017 International electron beam welding conference
[14] https://www.ptreb.com/sites/default/files/papers/PTR-EBONOVA-Non-Vac-EB-Welder.pdf
[15] Taminger K M and Hafley R A 2003 In Proceedings of the 3rd annual automotive composites
conference 9-10
[16] https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20080021301.pdf
[17] Wallace et al. 2004 National aeronautics and space administration hampton va langley research
center
[18] http://www.sciaky.com/additive-manufacturing/wire-am-vs-powder-am
[19] Liu S, Siewert T A 1989 Welding Journal 68(2) 52-58
[20] Nouri H, Guessasma S, Belhabib S 2016 Journal of Materials Processing Technology 234 113-
124
[21] Guessasma S, Nouri H and Roger F 2017 Polymers 9(8) 372
