Use of an Abrasive Water Cavitating Jet and Peening Process to Improve the Fatigue Strength of Titanium Alloy 6Al-4V Manufactured by the Electron Beam Powder Bed Melting (EBPB) Additive Manufacturing Method
- 607 Downloads
Metal components made by additive manufacturing have large inherent surface roughness, and, as such, their strength and fatigue life can be reduced significantly versus wrought products. In order to improve these properties, a novel mechanical surface treatment that introduces compressive residual stress while simultaneously reducing the surface roughness is proposed. The proposed treatment uses cavitation peening combined with an abrasive slurry. The impact of the kinetic energy-charged abrasive particles, induced by collapsing water cavitation vapor bubbles, produces compressive residual stress, while the abrasive reduces the surface roughness. Plane-bending fatigue tests were carried out to determine the effectiveness of this treatment on the fatigue life and strength of titanium alloy Ti6Al4V manufactured by electron beam melting. It was demonstrated that the fatigue strength of an as-built specimen was improved from 169 MPa to 280 MPa by the proposed treatment.
Additive manufacturing (AM) such as electron beam powder bed melting (EBPB) and laser beam melting (LBM) are attractive processes for the aerospace and biomedical industries, as complex metal components can be produced directly using computer-aided design systems, and “near-net-shape” manufacturing with shorter lead times and reduced material waste can be achieved.1,2 However, the fatigue life and strength of components manufactured by AM can be very small and also variable due to the surface roughness,1,3,4 One method that has been reported to improve the fatigue strength of titanium alloy Ti6Al4V manufactured by EBPB is shot peening, which introduces compressive residual stress into the surface.1 However, to reduce the surface roughness of AM metal components, further surface treatment is needed.
Chan et al. have reported that the fatigue life of titanium alloy Ti6Al4V manufactured by EBPB and LBM, including rolled and cast specimens, is proportional to the maximum surface roughness.4 Rafi et al. and Gong et al. studied the effect of defects on the mechanical properties of Ti6Al4V fabricated by EBPB and LBM 5,6; however, their specimens were machined or ground. Furthermore, Seifi et al. carried out a review of the flaws found in metal specimens made by AM, such as voids, layer defects and inclusions, and also the porosity.7 The effects of surface roughness, however, were not discussed. As is well known, AM metals have large inherent surface roughness due to powder. Fousova et al. discussed the influence of the surface roughness and internal defects on the fatigue properties of Ti6Al4V manufactured by EBPB and selective laser melting (SLM), and concluded that the surface roughness is the most critical property.8 Thus, investigations into the effect of surface roughness on the fatigue performance of AM Ti6Al4V have been investigated by many research groups,9–14 and several processes to be performed during and after AM have been proposed.15
Edwards and Ramulu, et al. showed that the fatigue strength of Ti6Al4V manufactured by EBPB and SLM in various stacking directions could be improved by machining and shot peening.1,9 Improvements in the fatigue strength of Ti6Al4V manufactured by AM by milling were confirmed by Sato et al. and Bagehorn et al.11,16 Machining and milling can reduce roughness of the surface; however, it is very difficult to apply these processes to the inner walls of components or undercut structures. Other methods used to improve the quality of the surface are chemical polishing17 and laser polishing18,19; however, neither of these methods introduce compressive residual stress into the surface. One of key factors determining the fatigue properties of AM Ti6Al4V is residual stress; 9 thus, as well as polishing, the introduction of compressive residual stress is very important. A typical method used to introduce compressive residual stress is shot peening. Unfortunately, this produces sparks and dust, and it has been pointed out that these could lead to dust explosions.20 Also, the contact made between the shot and the metal being treated can cause material to be transferred to the surface, and there is a risk that this might become the source of corrosion. In order to avoid this risk and to enhance the peening intensity, shotless peening such as laser peening21,22 and cavitation peening,23,24 have been proposed. Compressive residual stress can be introduced into the surface of AM Ti6Al4V by cavitation peening and laser peening,16,25 and these processes have been shown to improve the fatigue strength of Ti6Al4V manufactured by EBM.26 The great advantage of cavitation peening is that it can be used to treat undercut parts and the internal walls of holes,27,28 as the cavitation bubbles can get into these regions before the bubbles collapse. In a previous study, we demonstrated that the fatigue strength of Ti6Al4V manufactured by EBPB was improved by 84% by cavitation peening; however, the surface roughness of the treated specimen was similar to that of an as-built specimen.26 Thus, a process that combines the introduction of compressive residual stress by cavitation peening with a process that reduces the surface roughness should improve the fatigue properties.
In this paper, we propose a novel surface finishing method that uses an abrasive cavitating jet, which reduces the surface roughness by abrasion while simultaneously introducing compressive residual stress by cavitation peening. Zhang et al. have reported that EBPB can produce components with higher densities compared with SLM,29 and Chan et al. examined the relationship between the fatigue life and the surface roughness of AM Ti6Al4V manufactured both by EBPB and LBM. Once improvements in the fatigue properties of EBPB metals have been demonstrated, these can be applied to other AM metals. In the present experiment, specimens made by EBPB were treated with an abrasive cavitating jet, and the fatigue life and strength were evaluated using plane-bending fatigue tests. The residual stress and the hardness of the surface of the specimens were also measured as well as the surface roughness.
Experimental Facilities and Procedures
The geometry of the specimens used for the plane-bending fatigue tests was the same as in the previous report.26 The thickness of the specimens was 2 ± 0.2 mm and all were manufactured by EBPB. The powder used in the EBPB process was Ti6Al4V with an average diameter of about 75 μm. The diameter of the spot size in the EBPB process was 0.2 mm and the stacking pitch was 90 μm. The stacking direction was in the width direction of the specimen. The width of the specimen at the center was 20 mm with a radius of curvature of 45 mm. The specimens were heat-treated at 1208 K under vacuum for 105 min, then cooled in argon gas. Then, aging was carried out at 978 K under vacuum for 2 h before the specimens were cooled in argon gas. After that, the edges of all the specimens were polished by hand using rubber whetstones of #80 and #180, to reduce crack initiation from the edges, as described in a previous report.26
In this experiment, c1 and c2 were obtained from 3 experimental data points for the non-treated specimens by the method of least squares. The c3 was obtained from c1 and the experimental data points, i.e., σaT and NfT, for each specimen treated for the different processing times using Eq. 5.
From Eq. 7, we obtain Nf 330 for each processing time. In order to investigate the fatigue strength, fatigue tests were carried out for the optimum processing time that maximizes the fatigue life. The tests were terminated when 107 cycles were exceeded.
The fatigue strength of the specimens is affected primarily by both the surface roughness and the compressive surface/sub-surface residual stress. The arithmetic mean roughness Ra and the maximum height of the roughness Rz were measured by a profilometer with stylus cutoff lengths, λc, of 0.25 mm, 0.8 mm and 2.5 mm. As the surfaces of the specimens were very rough, the surface hardness was measured using a Rockwell superficial hardness tester. For the superficial hardness test, both a 120° diamond spheroconical indenter and a 1/16-inch-diameter (1.588 mm) steel sphere were used. The initial load was 3 kgf (29 N), and the applied load was 15 kgf (147 N). The hardness was measured seven times in each case, and the mean and standard deviation were obtained from all the values excluding the highest and lowest values. The thickness of material removed by the abrasive cavitating jet was measured by a caliper with a measurement accuracy of 0.01 mm.
The residual stress σR of the specimen surface was evaluated by a 2D method31 using x-ray diffraction with a two-dimensional detector. The x-ray diffraction patterns were obtained using Cu Kα x-rays from a tube operated at 40 kV and 40 mA through a 0.8-mm-diameter collimator with an incident monochromator. The lattice planes (h k l) used for these measurements were the Ti (2 1 3) and (3 0 2) planes, and the diffraction angles without strain were 139.5° and 148.4°, respectively. Using the conditions established in a previous study, 24 diffraction rings were measured at various angles.32 The exposure time per frame for locating the diffraction ring at each single position was 5 min.
Surface roughness and ratio between roughness at tp= 0 s and 15 s
Arithmetical mean roughness, Ra
Maximum height of the roughness, Rz
Cutoff lengths, λc
Processing time tp
6.9 ± 0.8
13.4 ± 1.9
18.2 ± 1.9
47 ± 6
81 ± 17
109 ± 26
3.4 ± 1.0
9.2 ± 2.8
15.1 ± 5.5
30 ± 11
65 ± 20
95 ± 28
1.9 ± 0.4
6.8 ± 1.4
11.5 ± 2.5
15 ± 4
43 ± 11
66 ± 11
0.8 ± 0.1
3.8 ± 0.9
7.6 ± 1.6
6 ± 1
21 ± 4
42 ± 10
1.0 ± 0.3
4.0 ± 0.5
9.3 ± 2.3
8 ± 3
25 ± 2
54 ± 13
Ratio at tp= 0 s/tp= 15 s
Here, b is constant with 0 < b < 1. In Fig. 7, the 5 data points plotted in Figs. 3 and 5 were used to obtain a and b by the method of least squares. Namely, a and b were obtained from the relationship between Nf330exp and Nf 330 est by the method of least squares. In the present calculation, Rz was used the value of λc = 2.5 mm, as the correlation coefficient was better than that of the others. The correlation coefficient for the 5 data points is 0.958, and the probability of non-correlation is less than 1.0%. Note that, when non-correlation is less than 1%, it can be concluded that the relationship is highly significant. Thus, it can be concluded that the relationship between Nf 330 exp and Nf 330 est is highly significant. That is, Nf 330 is closely related to Rz′ with λc = 2.5 mm, the surface hardness HR15T′ and the compressive residual stress σCR, each of which were used for the estimation. Note that the values of a and b obtained were 0.081 and 0.660, respectively. The results suggest that the contribution of the compressive residual stress to the improvement made in the fatigue life is about 8% of the total contribution, and that the effects of Rz and HR15T are larger than that due to σCR at the present condition.
The abrasive cavitating jet and peening process improved both the fatigue life and strength. There was an optimum processing time for the proposed treatment, at which the fatigue strength of the treated specimen was improved by 66%. The improvements in the fatigue properties were obtained as a result of the smoothing of the roughness, work-hardening and the introduction of surface/sub-surface compressive residual stress. Under the conditions used here, the effects of smoothing and work-hardening were greater than the effect of introducing compressive residual stress.
The abrasive cavitating jet and peening process smoothed the inherent surface roughness of the Ti6Al4V. There was an optimum processing time for this. At the optimum processing time, the surface roughness was reduced by factors of approximately 8 with λc = 0.25 mm, 4 with λc = 0.8 mm and 2.5 with λc = 2.5 mm, where λc is the cutoff length of the stylus in the profilometer.
The abrasive cavitating jet introduced compressive residual stress of 220 MPa into the surface of the metal specimens. Under the conditions used here, the contribution of the compressive residual stress to the improvement made in the fatigue life is about 8% of the total contribution.
The surfaces of the Ti6Al4V specimens were work hardened by the abrasive cavitating jet.
This work was partly supported by JSPS KAKENHI Grant Number 17H03138. The abrasive cavitating jet apparatus was financially supported by Boeing Research and Technology (BR&T).
- 16.M. Sato, O. Takakuwa, M. Nakai, M. Niinomi, F. Takeo, and H. Soyama, Mater. Sci. Appl. 7, 181 (2016).Google Scholar
- 17.A. Dolimont, E. Riviere-Lorphevre, F. Ducobu, and S. Backaert, in Proceedings of the 21st international esaform conference on material forming, 140007–1, (2018).Google Scholar
- 24.H. Soyama, Int. J. Peen Sci. Technol. 1, 3 (2017).Google Scholar
- 28.H. Soyama, Mater. Sci. Appl. 5, 430 (2014).Google Scholar
- 29.L.C. Zhang, Y.J. Liu, S.J. Li, and Y.L. Hao, Adv. Eng. Mater. 20, 16 (2018).Google Scholar
- 33.H. Soyama, R. Oba, and H. Kato, in Proceedings of the institute of mechanical engineering, 3rd international conference of the cavitation, vol. 103 (1992).Google Scholar
- 34.R.E. Little, ASTM STP 511, 29 (1972).Google Scholar
- 35.National Research Institute for Metals, Japan, Fatigue Data Sheet, No. 85, 1 (2000).Google Scholar
- 36.T. Kokubun and H. Soyama, Trans. JSME 83, 1 (2017).Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.