Evolution of Bonding Interface during Ultrasonic Welding between Ni and Steels with Various Microstructure

Abstract

In this study, Ni was bonded with steels having various microstructures to investigate the effect of various microstructures in steels on the bonding strength evolution by ultrasonic welding. It is found that at Ni/ferrite interface having similar hardness, the bonding can be produced by flattening of wear particles generated from the abrasion during ultrasonic welding to obtain a higher degree of plastic deformation, which is positive to bonding strength evolution. As for Ni/pearlite and Ni/martensite interfaces having dissimilar hardness, the bonding formation is difficult due to the presence of hard phases that limit the degree of plastic deformation near the interface, and Ni fragments are attached on the steel side. As a result, lower bonding strength evolution is correspondingly obtained due to slower increment of bonded area, whereas longer time is required for bonding formation between attached Ni fragments and the base metal Ni.

1. Introduction

To manufacture multi-material structures in the automotive industry, bonding process is always a crucial aspect for obtaining desired properties, such as weight-reduction¹⁾ and high specific strength.²⁾ To bond high strength steels and other metals such as aluminum alloys³⁾ for the above-mentioned purposes, limitations such as extensive formation of brittle intermetallic compound (IMC)^4,5,6) or significant microstructural change^7,8) at a high temperature by using fusion welding should be concerned due to their negative effect on bonding strength of joints. Also, regarding solid-state bonding techniques, FSW⁹⁾ or diffusion bonding¹⁰⁾ are recognized by their feasibility of bonding dissimilar materials. However, IMC formation at relatively longer welding time and high temperature,¹¹⁾ tool wear issue in bonding high strength steels¹²⁾ in FSW, and long processing time in diffusion bonding¹³⁾ exist. To overcome these problems, a bonding technique at shorter welding time without excessive increase of temperature is desired.

From the above viewpoint, ultrasonic welding (USW) is a potential technique offering advantages such as short welding time (usually in a few seconds), less heat generated at bonding interface, and versatility of bonding dissimilar materials.^14,15) Principle of USW can be depicted as: firstly, a clamping force (500–2000 N) is applied vertically on base metals; a high frequency (15–20 kHz) oscillation (amplitude: 20–60 μm) is then applied on base metals; and heat generated from sliding friction between base metals to make a weld.

To date, investigations of USW applying on bonding structural materials are mainly focusing on feasibility of producing sufficient bonding strength, or interfacial phenomena between base metals during USW, such as macroscopic plastic flow,¹⁶⁾ dynamic recrystallization,¹⁷⁾ and mechanical interlocking caused from plastic deformation¹⁸⁾ in metal combination such as Al¹⁹⁾ or Cu²⁰⁾ alloys. As for bonding harder materials (e.g., steels) by USW, studies regarding the above aspects are still limited. For instance, Levy et al.²¹⁾ demonstrated the bonding of steels, in which the thin mild steel foils are joined on a steel base plate, but weld quality is still insufficient to obtain a strong bonding by using USW only. On the other hand, Nambu et al.²²⁾ reported the development of steel/steel and Ni/steel bonding interfaces by USW, in which wear particles are produced from abrasion between base metals, and subsequently flattened to produce bonding with the disappearance of voids or gaps. To bond high-strength steels and other dissimilar metals by USW, the understanding of bonding interface development is essential for improving bonding strength of joints, especially the effect of steel hardness on dissimilar metal bonding.

Regarding this aspect, it has been reported that more input welding energy is required to produce bonding between harder metals.²³⁾ In bonding steels and other dissimilar metals, it can be interpreted that steels with different hardness resulted from different microstructures have an influence to bonding formation in view of the difficulty of producing plastic deformation. Nevertheless, the understanding from the above aspect is still insufficient. Thus, the primary objective of this study is to clarify the bonding strength evolution mechanism of dissimilar metal bonding interface with the change in steel hardness and microstructure. In this study, nickel was employed to bond with various kind of steels because it will not react with steel to form intermetallic compound.

2. Experimental Procedures

Pure Ni, interstitial-free (IF) steel, 14Ni steel, 0.08C steel and 0.4C steel sheets were employed. Specimens with a dimension of 15 mm × 7.5 mm × 1 mm were prepared for the steels and Ni sheets. In the preparation of different microstructures for metal sheets, Ni and IF steel were heat-treated at 1073 K for 5 hours in an Ar-3%H₂ gas atmosphere to remove hardening effect from rolling process. 14Ni steel and 0.08C steel sheets were austenitized at 1273 K for 2 min, followed by water-quenching to obtain fully martensitic microstructure. 0.4C steel sheets were austenitized at 1273 K for 2 min, and then furnace-cooled to obtain ferrite and pearlite microstructure. Surfaces of the specimens were mirror-polished by 1 μm Al₂O₃ to obtain smooth and flat bonding surfaces. After polishing, hardness of metal sheets with various microstructures were measured by Vickers hardness test (0.2 kg of testing force; HMV-2, Shimadzu). Chemical compositions of metal sheets are listed in Table 1.

Table 1. Chemical compositions, microstructures, and hardness of metal sheets.

Metals	Composition (wt.%)	Microstructure	Hardness (HV0.2)
Ni	99.5% pure Ni	–	105
IF steel	Fe-0.001C-0.008Si-0.16Mn	Ferrite	75
0.4C	Fe-0.4C-1.78Si-0.8Mn	Ferrite + Pearlite	144 (F), 332 (P)
14Ni	Fe-0.002C-0.25Si-0.25Mn-14Ni	Martensite	265
0.08C	Fe-0.08C-1.0Si-1.92Mn	Martensite	336

Ultrasonic welding was conducted by using an ultrasonic metal welder (SW-3500-20/SH-H3K7, Avionics). In the setup of USW, a Ni sheet was placed on a steel sheet. A clamping force was subsequently applied by a horn with contact area 10 mm × 10 mm in the normal direction to the bonding surfaces, and welding was then carried out in the condition of 2000 N of clamping force, 20 kHz of oscillation frequency, 41 μm of oscillation amplitude, and 0.01–5.0 s of welding time. To evaluate bonding strength, lap shear tests were conducted to welded specimens. After the lap shear test, the bonding strengths were obtained by averaging values of five samples tested at each welding condition. Fractured surfaces after the lap shear test and bonding interfaces were observed using field-emission scanning electron microscopy (FE-SEM; JSM-7001FA, JEOL). The compositional analysis using energy dispersive X-ray spectrometry (EDS; EX-64175JMU, JEOL) and the crystallographic analysis using electron backscattered diffraction (EBSD) were also conducted for the fractured surface and bonding interface.

3. Results and Discussion

3.1. Hardness of Steels Varied with Microstructures

The hardness and microstructures of metals obtained after heat-treatments are listed in Table 1. Figure 1 shows the microstructures of steels used in this study. According to images of optical microscopy and inverse pole figure (IPF) map of EBSD, IF steel having fully ferritic microstructure (Fig. 1(a); referred as IF), the 0.4C steel having ferrite (bright phase in Fig. 1(b); referred as F) and pearlite (dark phase in Fig. 1(b); referred as P) phases, and the 14Ni and 0.08C steels having fully martensitic microstructure (referred as M for 14Ni steel, and M’ for 0.08C steel; see Figs. 1(c) and 1(d)) were observed. The combinations are referred to as Ni/IF, Ni/F+P, Ni/M and Ni/M’ in the present study.

Fig. 1.

Microstructures of (a) ferrite in IF steel (b) pearlite and ferrite in 0.4C carbon steel (c) martensite in 14Ni steel and (d) martensite in 0.08C steel. (Online version in color.)

3.2. Strength of Joints

Figure 2 shows the relationship between welding time and bonding strength when the steel microstructure was changed. As shown in the figure, the Ni/IF revealed a rapid increase of bonding strength, which subsequently results in base metal fracture at the IF steel side as the bonding strength evolved to nearly 2 kN after 2.0 s of welding. Compared with the rapid increase in Ni/IF, Ni/F+P exhibited a slower bonding strength evolution, in which the bonding strength was able to be measured after 0.5 s of welding. Subsequently, the bonding strength was reached to approximately 2 kN after 5.0 s of welding, which is longer than that in Ni/IF. As for Ni/M, it exhibited even slower increase with the welding time. After 5.0 s of welding, it was still showed a lower bonding strength than Ni/IF after 2.0 s of welding. Lastly, Ni/M’ exhibited a much slower bonding strength evolution. The bonding strength of Ni/M’ after 3.0 s of welding exhibited a lower bonding strength than that of Ni/M after 3.0 s of welding and Ni/IF after 1.0 s of welding. It is considered that M’ (martensitic microstructure with 0.08 wt.% carbon) having the highest hardness retarded the increase of bonding strength. Compared with Ni/IF after 2.0 s of welding whereas the base metal fracture was occurred at IF steel, all of cases revealed interfacial fracture rather than base metal fracture after the lap shear test due to higher tensile strength possessed in steels and Ni sheets.

Fig. 2.

Relationship between welding time and bonding strength with different cases. (Online version in color.)

3.3. Observations on Ni/F Interface (in Ni/ferrite+pearlite)

Figure 3 shows the SEM images of cross-sectional observation for Ni/F bonding interface in Ni/F+P. Wear particles containing of Ni and Fe were found at the bonding interface, and mixture of steel and Ni in wear particle also indicated the occurrence of plastic flow after 0.25 s of welding (Fig. 3(a)). It is observed that wear particles were existed at bonding interface in a longer welding time of 0.5 s (Fig. 3(b)). After 1.0 s of welding, wear particles containing Ni and Fe were gradually flattened under clamping force to contribute to extensive formation of bonded area (Fig. 3(c)), and bonding was almost achieved with the vanishing of voids or gaps after 2.0 s of welding (Fig. 3(d)). It is considered that bonding formation process of Ni/F interface was similar to Ni/IF interface reported in our previous study.²²⁾

Fig. 3.

SEM images of the cross-section observation of Ni/F interface after (a) 0.25 s (b) 0.5 s (c) 1.0 s and (d) 2.0 s of welding. (Online version in color.)

3.4. Observations on Ni/P Interface (in Ni/ferrite+pearlite)

Figure 4(a) shows the Ni/P bonding interface region in the Ni/F+P after 0.1 s of welding, where a thin layer of Ni was attached on the steel side. Left side of Fig. 4(a) shows the phase map with image quality (IQ) map analyzed by EBSD. Red-colored bcc phase with lamellar structure is the pearlite phase. It is evident that a thin green-colored fcc phase (Ni) was attached on the pearlite side. It is suggested that at the beginning of USW, metals could be fractured from the base metal (Ni). Also, Ni/P interface after 0.25 s of welding exhibited that an attached Ni fragment was possessed at steel side and was further confirmed by EDS compositional mapping analysis, as shown in Fig. 4(b). Figure 4(c) shows the Ni/P bonding interface after 1.0 s of welding. A straight bonded section was observed without notable plastic deformation, or any wavy plastic flow as that observed in Ni/F interface. Figure 4(d) shows the Ni/P bonding interface after 2.0 s of welding. As shown, the fraction of bonded sections was not increased obviously even though a longer welding time was applied. Also, the straight bonded sections indicated the limited plastic deformation during USW.

Fig. 4.

SEM images of Ni/pearlite interface of the cross-section observation after (a) 0.1 s (with phase and IQ maps by EBSD) (b) 0.25 s (with EDS map of Ni) (c) 1.0 s (d) 2.0 s of welding, fractured surfaces of steel (pearlite) side overlaid with EDS map of Ni after (e) 0.1 s and (f) 0.25 s of welding, and (g) SEM image for the comparison between Ni/ferrite and Ni/pearlite interfaces after 1.0 s of welding. (Online version in color.)

In the fractured surface observation, it is revealed that Ni was attached on steel side, confirmed by EDS analysis of Ni composition on the steel side. Thus, it is exhibited that attached Ni fragments with galling were possessed at fractured surfaces after USW, and the fraction of attached Ni was increased with the welding time (Figs. 4(e) and 4(f)). It is known that galling is one form of adhesive wear that usually occurred at hard material sliding against a relatively soft metal with more ductility under heavy load, such as steel sliding on Al,²⁴⁾ to cause material removal from the ductile metal. Thus, it is considered that as hard steel (pearlite phase) sliding on relatively ductile Ni surface, Ni was easily peeled from the base metal to form attached metal fragments.

To compare interfacial microstructure between Ni and steels consisted of ferrite and pearlite, a region containing Ni/F and Ni/P interfaces was observed as shown in Fig. 4(g). As shown, Ni/F interface possessed significant plastic deformation. On the other hand, the Ni/P interface was merely possessed a straight interface indicating limited plastic deformation. Based on the above, it was suggested that the presence of hard pearlite phase could limit the degree of plastic deformation during bonding formation process.

3.5. Observations on Ni/M Interface (14Ni Steel Employed)

Cross-sectional observations at Ni/M interface were conducted to compare with other interfaces. It is shown that a thin Ni fragment was peeled from the Ni base metal and attached on the steel side after 0.1 s of welding, and EDS mapping of Ni revealed that Ni was attached on steel with shear deformation resulting in detachment from the base metal, as shown in Fig. 5(a). After 0.25 s of welding, it is seen that a straight bonding region with direct contact was gradually formed but without any wear particle as observed in Ni/F interface (Fig. 5(b)). After 1.0 s of welding, the fraction of bonded sections was increased, and some attached Ni fragments were observed at bonding interface (Fig. 5(c)). After 3.0 s of welding, bonded section with wavy plastic deformation was produced at the vicinity of the bonding interface (Fig. 5(d)). Also, fine Ni grains were observed at the vicinity of bonding interface, indicating the occurrence of plastic deformation during USW.

Fig. 5.

SEM images of the cross-section observation of Ni/martensite (14Ni steel) interface after (a) 0.1 s with EDS map of Ni (b) 0.25 s (c) 1.0 s and (d) 3.0 s of welding, and fractured surfaces of steel (martensite) side after (e) 0.1 s and (f) 0.5 s of welding. (Online version in color.)

In the fractured surface observations on steel side compared with cross-sectional observation, it is evident that a Ni fragment was attached on steel side after 0.1 s of welding, which was characterized by EDS mapping analysis (Fig. 5(e)). With the increase of welding time, the fraction of attached Ni fragments was correspondingly increased, and the size of contact areas covered by Ni was increased along the oscillation direction, as shown in Fig. 5(f).

3.6. Observations on Ni/M’ Interface (0.08C Steel Employed)

In Ni/M’ interface, low carbon steel consisted of harder martensite microstructure was utilized to examine the bonding interface development with its strength evolution as a comparison with Ni/P interface in which the harder martensite has a similar hardness to the pearlite. In cross-sectional observations, it is evident that Ni was attached on the steel side after 0.1 s of welding (Fig. 6(a)). Bonding was gradually formed accompanied with Ni attached on steel (Figs. 6(b) and 6(c)). As welding time increased to longer than 2.0 s, the fraction of bonded section was gradually increased with a straight bonding interface (Fig. 6(d)). Compared to Ni/M interfaces, a lower degree of plastic deformation was possessed at Ni/M’ interfaces even though welding time was increased to 3.0 s (Fig. 6(e)). Also, in the fractured surface observations, the attachment of Ni on the steel side can be confirmed by EDS mapping of Ni. It is revealed that the fraction of contact areas was increased with the welding time in which the expansion of contact areas was along the oscillation direction (Figs. 6(f) and 6(g)). According to the above, it is suggested that the bonding formation process of harder martensite case is similar to that of pearlite case having similar hardness.

Fig. 6.

SEM images of the cross-section observation of Ni/martensite (0.08C steel) interface after (a) 0.1 s (b) 0.25 s (c) 1.0 s (d) 2.0 s and (e) 3.0 s of welding, and fractured surfaces of steel side with EDS mapping of Ni after (f) 0.1 s and (g) 1.0 s of welding. (Online version in color.)

3.7. Effect of Steel Hardness on Bonding Strength Evolution

According to the above observations, it is evident that the hardness of steels with various microstructures had an influence on the formation of bonding interface and the bonding strength evolution. To understand the contribution of bonded sections observed in the above cross-section observations (Figs. 3, 4, 5, 6) to bonding strength evolution, a correlation between bonding strength and bonded section was considered. The fraction of bonded section was measured based on our previous studies^22,24) by classifying three sections: unbonded section (gaps without contact); partially-bonded section (wear particle or attached metal fragment bonded to either side of base metals with gap or void); and bonded section (fully bonding between base metals without gap or void). At a bonding interface with a length of approximately 500 μm, the fractions of each case were continuously measured by SEM observation to obtain the average values for each welding time. It is noted that the fraction values in Table 2 were measured based on the bonded sections mentioned above, and the partially-bonded and unbonded sections were excluded because they are considered without contribution to bonding strength. A relationship between bonding strength and bonded section was simply estimated using Eq. (1), in which Y_strength represents the average strength of joints (kN), and X represents the fraction of bonded sections measured in cross-sectional observations at Ni/F and Ni/P interfaces respectively to understand the contribution of bonded section to bonding strength of joints. By substituting values obtained from the fraction of bonded sections in Table 2, the relationship is expressed by Eq. (2).   

$${\mathrm{Y}}_{\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}}=\mathrm{a}{\mathrm{X}}_{\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{r}\mathrm{i}\mathrm{t}\mathrm{e}}+\mathrm{b}{\mathrm{X}}_{\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{e}}$$(1)

  

$${\mathrm{Y}}_{\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}}=3.78{\mathrm{X}}_{\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{r}\mathrm{i}\mathrm{t}\mathrm{e}}+1.47{\mathrm{X}}_{\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{e}}$$(2)

Table 2. Fractions of bonded sections and bonding strength with the change of welding time.

	Xferrite	Xpearlite	YStrength (kN)
0.5 s	0.076	0.081	0.38
1.0 s	0.132	0.141	0.68
2.0 s	0.244	0.204	1.13
3.0 s	0.292	0.223	1.41

According to the values obtained from the above relationship, bonding strength evolutions of Ni/F+P and other cases were plotted as a function of the fraction of bonded sections as shown in Fig. 7. It is evident that the slope of Ni/F interface (X_F) in Ni/F+P revealed a similar high bonding strength evolution as Ni/IF interface (i.e., slope of X_IF) with the welding time by higher degree of plastic deformation, induced by flattened wear particles. As for Ni/M interface, a slower bonding strength evolution (i.e., slope of X_M) was revealed compared with Ni/IF or Ni/F interface. In Ni/M’ and Ni/P interfaces, they were possessed much slower bonding strength evolutions (i.e., slopes of X_M’ and X_P) than Ni/M interface. It can be suggested that higher hardness of pearlite (Hv 332) and hard martensite (Hv 336) limited bonding formation with a high degree of plastic deformation. According to the above, it can be considered that softer phase could induce a higher degree of plastic deformation to obtain a rapid bonding strength evolution.

Fig. 7.

Relationship between bonding strength and fraction of the bonded sections. (Online version in color.)

3.8. Effect of Hardness on Wear Modes

To understand the effect of hardness on wear mode during USW, which is related to bonding formation process of USW, the transition of the following two wear modes was investigated. If wear particles are extensively observed at fractured surfaces, the combination was considered as the condition of wear particle formation. On the other hand, the observation of Ni attached on the steel side was confirmed as the condition of attached metal fragments. To obtain additional results, the fractured surfaces of M/M (14Ni steels) and M/M’ (14Ni and 0.08C steels) interfaces were observed as shown in Figs. 8(a) and 8(b), respectively. It is evident that wear particles were observed at bonding interfaces even in steels having higher hardness. Furthermore, Fig. 8(c) also reveals that a wear particle was flattened at bonding interface of 14Ni steels after USW. Figure 8(d) shows the region of IPF map analyzed by EBSD in Fig. 8(c), and it is evident that grains in wear particle were subjected to severe deformation, as reported in our previous study.²²⁾ It is revealed that wear particle in M/M (14Ni martensitic steels) interface possessed a similar size (referring to Fig. 8(d); approximately 10 μm) as that observed in Ni/ferrite interface (Fig. 3(c)) at the same welding time (1.0 s). Both of Ni/ferrite and M/M interfaces revealed wear particle formation, followed by flattening to produce bonding. Based on the above, it is suggested that in the present study, no significant difference in formation process and size of wear particle, regardless of combinations with similar high or low hardness. Thus, the occurrence of wear particle formation was also observed at metal combination having similar high hardness in both of fractured surfaces and interfacial observations.

Fig. 8.

Fractured surfaces of (a) 14Ni steel and (b) 0.08C steel after 1.0 s of welding, (c) 14Ni/14Ni interface with wear particle flattened after 1.0 s of welding, and (d) EBSD IPF map of (c). (Online version in color.)

Regarding the investigations on this aspect, Rabinowicz et al.²⁵⁾ reported that hardness ratio (harder/softer materials; H_hard/H_soft) of mating materials has an influence on the occurrence of wear mode. Figure 9 shows the distribution of wear mode occurred in various kinds of bonding interface with the above regimes, separated by wear particle or attached metal fragment formation. Regarding this difference, it is considered that if the hardness ratio of mating materials is approximately 1.0 (i.e., the similar hardness), both of abrasive and adhesive wear could occur at both of metals to cause material removal from the base metals to produce debris (i.e., wear particles). In other words, under sliding with heavy loading of USW, metals having similar hardness could be simultaneously removed by adhesive wear to generate wear particles, followed by rolling and further peeling materials from the base metal. And this phenomenon was seemingly accompanied with the abrasive wear.²⁶⁾ As the welding time increased, wear particles were flattened to produce bonded areas with higher degree of plastic deformation. Also, it is reported that as the difference in hardness was less than 40% of softer metal hardness, the formation of wear particle was dominant.²⁵⁾ Otherwise, adhesive wear induced at softer metal could be dominant (i.e., attached Ni fragments at Ni/martensite and Ni/pearlite interfaces). Under this condition, longer time was required to produce bonding by disappearing of gaps between attached Ni and the base metal (Ni).

Fig. 9.

Relationship between the difference in hardness and wear modes. (Online version in color.)

Referring to Fig. 9 showing the relationship softer metal hardness and the difference in hardness, the criterion of occurrence of two wear modes in USW can be separated by a dashed line. That is, if a metal combination having similar hardness (the difference is less than 40% of softer metal hardness; region below the dashed line), the formation of wear particle was predominantly occurred. On the other hand, if a metal combination having dissimilar hardness (the difference is more than 40% of softer metal hardness; the region above the dashed line), the formation of attached metal fragments was dominant.

Figure 10 schematically illustrates the development of bonding interface of Ni and steel having various hardness and microstructures. It is considered that the occurrence of two wear modes was affected by the difference in hardness of base metals to further result in difference of bonding strength evolution. If metals having similar hardness such as Ni/ferrite interface, wear particles formation could be dominant in the combination of steel and Ni. Under sliding, fracture generated at both asperities could result in peeling of asperities from the base metal to generate wear particles, followed by flattening of wear particles to achieve bonding with a higher degree of plastic deformation (Figs. 10(a) to 10(d)).

Fig. 10.

Schematic of development of (a)–(d) Ni/ferrite interface; and (e)–(h) Ni/martensite and Ni/pearlite interfaces. (Online version in color.)

In contrast to Ni/ferrite interface, at Ni/pearlite and Ni/martensite interfaces, only Ni was mainly peeled from the base metal to produce attached metal fragments. It is suggested that as the hardness difference became larger, the metal fragments could be peeled from soft metals during the sliding of USW (Figs. 10(e) and 10(f)). Under this condition, bonding achievement required longer time by disappearing of gaps or voids between attached Ni and the base metal (Figs. 10(g) and 10(h)), and the bonding strength evolution was further limited by a lower degree of plastic deformation at the bonding interface.

4. Conclusion

In this study, the effect of steel hardness on bonding interface formation and strength evolution between Ni and steel during USW was investigated, and the findings are summarized as follows:

(1) At Ni/F interface, bonding can be formed by the flattened wear particles generated from the abrasion and wear in the Ni/IF steel to contribute to a higher degree of plastic deformation and a rapid bonding strength evolution.

(2) At Ni/P interface, the bonding formation was difficult compared with the Ni/F probably due to the presence of hard pearlite phase that retards the plastic deformation of the interface. As a result, lower bonding strength evolution was obtained due to decreased bonded area compared with the case of Ni/IF.

(3) In Ni/M interface having softer martensite, bonding was formed uniformly with the welding time and wavy plastic deformation. In comparison, Ni/M’ interface having hard martensite exhibited a further inhibition of plastic deformation at the bonding interface without wavy plastic deformation. It was suggested that the hard martensite phases could also retard the plastic deformation at the bonding interface and resulted in a slow increase of bonding strength.

(4) Regarding the occurrence of two wear modes at the beginning of USW, in a metal combination having similar hardness, both of metals could be simultaneously peeled to generate wear particles, followed by peeling materials from the base metals with abrasive wear. On the other hand, as the difference of hardness in a metal combination increased to more than 40% of softer metal hardness, the formation of attached metal fragments with adhesive wear induced at the softer metal side was dominant.

Acknowledgement

This study was supported by ISIJ Research Promotion Grant.

References

• 1)   T. D.  Claar,  C. J.  Yu,  I.  Hall,  J.  Banhart,  J.  Baumeister and  W.  Seeliger: SAE Trans., 109 (2000), 98.
• 2)   R.  Kuziak,  R.  Kawalla and  S.  Waengler: Arch. Civ. Mech. Eng., 8 (2008), 103.
• 3)   I. N.  Fridlyander,  V. G.  Sister,  O. E.  Grushko,  V. V.  Berstenev,  L. M.  Sheveleva and  L. A.  Ivanova: Met. Sci. Heat Treat., 44 (2002), 365.
• 4)   S. B.  Lin,  J. L.  Song,  C. L.  Yang,  C. L.  Fan and  D. W.  Zhang: Mater. Des., 31 (2010), 2637.
• 5)   H.  Oikawa,  T.  Saitoh,  T.  Yoshimura and  T.  Kiriyama: Tetsu-to-Hagané, 84 (1998), 73 (in Japanese).
• 6)   H.  Oikawa,  T.  Saitoh,  T.  Nagase and  T.  Kiriyama: Tetsu-to-Hagané, 83 (1997), 641 (in Japanese).
• 7)   M. J.  Torkamany,  S.  Tahamtan and  J.  Sabbaghzadeh: Mater. Des., 31 (2010), 458.
• 8)   J. L.  Song,  S. B.  Lin,  C. L.  Yang,  G. C.  Ma and  H.  Liu: Mater. Sci. Eng. A, 509 (2009), 31.
• 9)   R. S.  Mishra and  Z. Y.  Ma: Mater. Sci. Eng. R, 50 (2005), 1.
• 10)   S.  Sam,  S.  Kundu and  S.  Chatterjee: Mater. Des., 40 (2012), 237.
• 11)   Z. W.  Chen,  S.  Yazdanian and  G.  Littlefair: J. Mater. Sci., 48 (2013), 2624.
• 12)   H.  Fujii,  L.  Cui,  N.  Tsuji,  M.  Maeda,  K.  Nakata and  K.  Nogi: Mater. Sci. Eng. A, 429 (2006), 50.
• 13)   N.  Orhan,  T. I.  Khan and  M.  Eroglu: Scr. Mater., 45 (2001), 441.
• 14)   F.  Balle,  G.  Wagner and  D.  Eifler: Adv. Eng. Mater., 11 (2009), 35.
• 15)   S.  Matsuoka: J. Mater. Process. Technol., 75 (1998), 259.
• 16)   H. S.  Shin and  M.  de Leon: J. Mater. Process. Technol., 224 (2015), 222.
• 17)   S.  Shimizu,  H. T.  Fujii,  Y. S.  Sato,  H.  Kokawa,  M. R.  Sriraman and  S. S.  Babu: Acta Mater., 74 (2014), 234.
• 18)   D.  Bakavos and  P. B.  Prangnell: Mater. Sci. Eng. A, 527 (2010), 6320.
• 19)   E.  Mariani and  E.  Ghassemieh: Acta Mater., 58 (2010), 2492.
• 20)   M. R.  Sriraman,  S. S.  Babu and  M.  Short: Scr. Mater., 62 (2010), 560.
• 21)   A.  Levy,  A.  Miriyev,  N.  Sridharan,  T. Y.  Han,  E.  Tuval,  S. S.  Babu,  M. J.  Dapino and  N.  Frage: J. Mater. Process. Technol., 256 (2018), 183.
• 22)   S.  Nambu,  K.  Seto,  J. Y.  Lin and  T.  Koseki: Sci. Technol. Weld. Join., 23 (2018), 687.
• 23)   H. P. C.  Daniels: Ultrasonics, 3 (1965), 190.
• 24)   J. Y.  Lin,  S.  Nambu and  T.  Koseki: Sci. Technol. Weld. Join., 24 (2019), 83.
• 25)   E.  Rabinowicz: Proc. Int. Conf. on Wear of Materials, ASME, New York, (1983), 12.
• 26)   S.  Stupkiewicz and  Z.  Mroz: Wear, 231 (1999), 124.