Metallurgical and Materials Transactions A

, Volume 40, Issue 3, pp 625–636

Boride Formation Induced by pcBN Tool Wear in Friction-Stir-Welded Stainless Steels

Authors

    • Department of Materials Processing, Graduate School of EngineeringTohoku University
    • Hitachi Research LaboratoryHitachi Ltd.
  • Yutaka S. Sato
    • Department of Materials Processing, Graduate School of EngineeringTohoku University
  • Hiroyuki Kokawa
    • Department of Materials Processing, Graduate School of EngineeringTohoku University
  • Kazutaka Okamoto
    • Hitachi Research LaboratoryHitachi Ltd.
  • Satoshi Hirano
    • Hitachi Research LaboratoryHitachi Ltd.
  • Masahisa Inagaki
    • Hitachi Research LaboratoryHitachi Ltd.
Article

DOI: 10.1007/s11661-008-9709-9

Cite this article as:
Park, S.H.C., Sato, Y.S., Kokawa, H. et al. Metall and Mat Trans A (2009) 40: 625. doi:10.1007/s11661-008-9709-9
  • 330 Views

Abstract

The wear of polycrystalline cubic boron nitride (pcBN) tool and its effect on second phase formation were investigated in stainless steel friction-stir (FS) welds. The nitrogen content and the flow stress were analyzed in these welds to examine pcBN tool wear. The nitrogen content in stir zone (SZ) was found to be higher in the austenitic stainless steel FS welds than in the ferritic and duplex stainless steel welds. The flow stress of austenitic stainless steels was almost 1.5 times larger than that of ferritic and duplex stainless steels. These results suggest that the higher flow stress causes the severe tool wear in austenitic stainless steels, which results in greater nitrogen pickup in austenitic stainless steel FS welds. From the microstructural observation, a possibility was suggested that Cr-rich borides with a crystallographic structure of Cr2B and Cr5B3 formed through the reaction between the increased boron and nitrogen and the matrix during FS welding (FSW).

1 Introduction

Friction stir welding (FSW) has been widely studied and commercially used in low-softening temperature material structures since it was invented nearly 15 years ago.[116] The feasibility studies of FSW on high-softening temperature material (HSTM) were also made relatively early. Some initial feasibility studies of FSW on 12 pct Cr alloy and low carbon steel were demonstrated by The Welding Institution (TWI).[17] Thereafter, FSW feasibility was examined in several types of HSTMs such as ferritic steels,[18,19] stainless steels,[2024] and heat-resistant steels.[20] Transverse tensile specimens failed in regions corresponding to the base material (BM), and their transverse tensile properties were governed by the BM properties in most of FS-welded steels; i.e., yield and ultimate tensile strengths of the weld were comparable to those of the BM.

However, the research on FSW of HSTMs such as steels representing most of the welded structures is still limited, compared to that of LSTM. One of the major causes of the limited studies on HSTM is the lack of suitable welding tools. The tool must resist physical and chemical wear, possess sufficient mechanical strength at elevated temperatures, and effectively dissipate the heat carried to the tool during the welding process. For initial tool materials for HSTM, W-series alloys such as W alloys and WC-Co alloy have been used.[17,21] The FSW of mild steel using Mo-based alloy tool has also been demonstrated.[18] However, tool wear was inevitable. The previous study reported the changes in tool dimensions arising from both rubbing wear and deformation of the tool.[18] The greatest changes in tool dimensions occurred during the initial plunging stage of the tool. Recently, a new tool made of polycrystalline cubic boron nitride (pcBN) that appears capable of meeting these requirements, and especially the wear resistance, has been developed.[20] Development of a tool material with excellent properties offers a pathway to more active studies on FSW of steels.

Several studies have demonstrated the potential of pcBN tool for ferrous materials.[2226] A refined microstructure formed in the stir zone (SZ), and adequate mechanical properties were achieved in the welds, suggesting that pcBN is one of the promising tool materials for HSTM. Unfortunately, wear and damage of pcBN tools still occur. It is important to examine the wear behavior of the pcBN tool during FSW, because the tool wear affects the mechanical and corrosion properties of the welds.

In the present study, the fundamentals of pcBN tool wear during FSW were investigated in stainless steels. The FSW was applied to ferritic (type 430), duplex (type 329J4L), and austenitic (types 304, 316L, and 310) stainless steels using pcBN tool. The microstructures were observed by electron microscopy. Their microstructural evolution and the boride formation induced by pcBN tool wear in austenitic stainless steel were examined in this study.

2 Experimental procedure

2.1 Materials and Process Details

The material type and chemical composition used in the present study are shown in Table I. Five types of as-received stainless steels with 6-mm thickness, including 430 ferritic, 329J4L duplex, 304, 316L, and 310 austenitic types, were FS welded. The stainless steel plate was tightly clamped by clamping fixtures, and then FSW was performed on the plate. Bead-on-plate FS welds were produced at a travel speed of 1.33 mm/s and a rotational speed of 550 rpm using an MS100 pcBN tool with a pin length of 4.29 mm. All of the welding trials were carried out on a manual vertical milling machine with 11 kW spindle drive motor power. The tool was tilted at 3.5 deg from the plane normal, and an inert Ar gas shroud was used for shielding, which prevented surface oxidation. The welding temperature was measured in each stainless FS weld using K-type thermocouples.
Table I

Materials and the Chemical Composition Used in the Present Study

Type

C

Si

Mn

P

S

Ni

Cr

N

Mo

Cu

W

430

0.070

0.42

0.88

0.029

0.002

0.09

16.34

0.039

329J4L

0.016

0.41

0.78

0.028

0.001

7.27

24.81

0.166

3.13

0.43

0.41

304

0.040

0.59

1.08

0.032

0.003

8.56

18.10

0.066

316L

0.019

0.43

0.87

0.026

0.001

12.19

17.70

0.028

2.19

310

0.040

0.57

0.85

0.023

0.001

19.19

25.35

0.025

2.2 Evaluation of Tool Wear

When pcBN tool wear occurs during FSW, both boron and nitrogen content should increase in the weld because the pcBN consists of boron and nitrogen. Following FSW, the nitrogen contents of the SZ and BM were measured in an inert gas atmosphere by a fusion gas chromatographic analyzer produced by LECO1 (model TC-436DR oxygen and nitrogen analyzer). The samples of about 0.5 g for nitrogen content analysis were carefully cut from only the SZ using an electrical-discharge machine (EDM). The samples were then cleaned by filing away the surface and rinsing them with acetone for 300 seconds using an ultrasonic cleaner. At least three samples were analyzed for both the SZ and BM. The average value was used for the nitrogen contents of each region.

It is likely that tool wear is related to the resistance force from the material to be welded to the tool during FSW. The resistance force would be presented by the flow stress of welded material at welding temperatures. The steady-state flow stress during hot compression test was measured using a hot working simulator “Thermecmastor Z (Fuji Electronic Industrial Co. Ltd., Saitama, Japan).” The measurement of flow stress is schematically shown in Figure 1. Thermecmastor Z consists of a chamber, compression bases, and a high-frequency induction heating system. Cylindrical-shaped specimens were prepared, with a 6-mm diameter and a 9-mm height. Thermocouples were spot welded on the surface of specimen, and then the specimen was placed on the lower base. The chamber was evacuated using a rotary vacuum pump. The specimen was heated to the temperature from 1373 to 1573 K by a high-frequency induction system. The actually measured peak temperatures during FSW were applied for hot compression test. The flow stress at the measured average temperature were used in this study, because there is a high possibility that the tool wear occurs in the vicinity of the interface between the tool surface and the welded material, where the temperature is the highest. When the temperature of each specimen reached the target value, the hot compression test was conducted by displacement of the upper base toward the compression direction. The applied strain rate ranged from 15 to 80 s−1. The steady-state flow stress was defined as the average value in the range of strain between 0.25 and 0.40.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig1_HTML.gif
Fig. 1

Schematic illustration of the measurement of flow stress

2.3 Microstructural Observation

Cross sections perpendicular to the welding direction were observed using optical microscopy (OM). Specimens for OM were etched electrolytically in a solution of 10 pct oxalic acid + 90 pct water with a power supply set to 30 V for about 10 seconds. Elemental analysis using an electron probe microanalyzer (EPMA) was carried out for particles found on the cross section. Thin disks with a diameter of 3 mm were cut from the various locations of the weld using an EDM, and then the electron-transparent thin sections were electrolytically made by twin-jet polishing in a 10 pct perchloric acid + 90 pct ethanol solution. These were observed at 200 kV with a JEOL2-2000EXII transmission electron microscope (TEM) and a Hitachi HD-2000 (Tokyo, Japan) scanning transmission electron microscope (STEM) equipped with an energy-dispersive X-ray (EDX) spectroscopy analysis system, using a 0.5-nm electron probe with a spatial resolution of 1.0 nm. The particles in the advancing side of the SZ were identified with both a selected area electron diffraction (SAED) pattern and STEM-EDX.

3 Results and discussion

3.1 Wear of pcBN Tool

The cross sections perpendicular to the welding direction of FS-welded 430, 329J4L, 304, 316L, and 310 stainless steels are shown in Figure 2. A counterclockwise rotation direction of the welding tool was applied to these welds. The retreating and advancing sides correspond to the left- and right-hand sides on the cross sections shown in Figure 2. The SZ, which is seen around the weld center, showed different contrast between the advancing and the retreating sides. The SZ on the advancing side (SZ-AS) had a band structure with deeply etched lines and pits, which is typically found in FS-welded austenitic stainless steels,[25,26] so that the strong contrast of this region on the cross section would be attributable to the corroded lines and pits.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig2_HTML.gif
Fig. 2

Cross section of FS-welded stainless steels perpendicular to the welding direction

Analysis of the nitrogen content was conducted to investigate the wear behavior of pcBN tool in five types of stainless steels including ferritic (type 430), duplex (type 329J4L), and austenitic (types 304, 316L, and 310) stainless steels. The typical regions analyzed are drawn in Figure 2(e) as white squares. The nitrogen contents of the advancing side of the SZ were compared with those of the BM and the retreating side of the SZ. The advancing and retreating sides are notated as SZ-AS and SZ-RS throughout the article, respectively.

Nitrogen contents of the BM, SZ-AS, and SZ-RS in five types of stainless steel welds are shown in Figure 3. Types 430 ferritic and 329J4L duplex stainless steels showed very little increase in nitrogen content in both the SZ-AS and SZ-RS, although the SZ-AS contained slightly higher nitrogen content than the BM and SZ-RS. On the other hand, a large increase of nitrogen contents was detected in the SZ-AS in austenitic stainless steel FS welds. In contrast, the SZ-RS of the austenitic stainless steel FS welds showed roughly the same amount of nitrogen content as the BM. The increase in nitrogen content in the SZ-AS of FS-welded austenitic stainless steels is summarized in Figure 4. The nitrogen content in the advancing side increased in the following order: 304, 316L, and 310 austenitic stainless steels. Type 310 steel exhibited the largest increase in nitrogen content among these austenite stainless steels, roughly 5 times greater than the BM. These results suggest that the pickup of both boron and nitrogen is greater in austenitic stainless steels during FSW.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig3_HTML.gif
Fig. 3

Nitrogen content in the SZ of FS weld and BM

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig4_HTML.gif
Fig. 4

Nitrogen content in the advancing side of the SZ in FS-welded austenitic stainless steel

3.2 Factor Governing Wear Behavior of pcBN Tool

The nitrogen content was analyzed in stainless steel FS welds, which showed that the pickup of boron and nitrogen is more significant in the austenitic stainless steels. Since FSW is a mechanical process accompanying recrystallization at high strain and high temperature, the flow stress at steady state would be one of the important factors that represent the tendency of tool wear. During recrystallization, the steady-state flow stress is not strongly dependent on strain so that the flow stress predominantly depends on the strain rate and temperature, which is generally known to be proportional to the Zener–Hollomon parameter (Z =  \( \ifmmode\expandafter\dot\else\expandafter\.\fi{\varepsilon } \) exp (Q/RT)).[27,28] In the present study, the peak welding temperatures were measured to be about 1373 K (1100 °C) in 430 ferritic FS weld and greater than or equal to 1473 K (1200 °C) in the other welds. Some previous studies roughly estimated the strain rate of about 10 to 15 s−1 during FSW of Al and Mg alloys.[29,30] Therefore, temperature between 1373 K (1100 °C) and 1523 K (1250 °C) and strain rates greater than or equal to 15 s−1 were chosen in the hot compression test. Effects of temperature and strain rate on stress-strain curves in five types of stainless steels are shown in Figure 5. Effects of temperature on steady-state flow stress at the strain rate of \( \ifmmode\expandafter\dot\else\expandafter\.\fi{\varepsilon } \) = 15 s−1 are presented in Figure 6. The steady-state flow stress decreases with increasing temperature and slightly with decreasing strain rate. The steady-state flow stress followed the following order: 430 ferritic < 329J4L duplex < austenitic stainless steels. This order is incident with the tendency of the boron and nitrogen pickup in stainless steels obtained by the analysis of nitrogen content. These results reveal that the boron and nitrogen pickup during FSW occurs more significantly in steels having higher steady-state flow stress. Tribochemical reaction would also affect tool wear during FSW, because FSW contains tribological phenomena. Formation of the borides related to tribochemical wear will be presented in Section C.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig5_HTML.gif
Fig. 5

Stress-strain curve in stainless steels using the hot working simulator

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig6_HTML.gif
Fig. 6

Effect of the temperature on steady-state flow stress

3.3 Boride Formation Induced by pcBN Tool Wear

In this section, the evolution of B-rich second phases in austenitic stainless steel FS welds will be discussed. The OM and TEM images of the BM, SZ-RS, and SZ-AS in type 304, 316L, and 310 weld microstructures are shown in Figure 7. The observed regions are roughly coincident with the regions where the nitrogen content was analyzed. The BM consisted of the annealed grain structure with low dislocation density. The SZ had relatively fine and equiaxed recrystallized grains. Much research has reported that this recrystallized grain structure generally originates from the high-temperature exposure and severe plastic deformation during FSW.[25,30,31] The SZ-AS had alternating bands similar to the “onion ring” structure that is often seen in Al alloy FS welds. The TEM images revealed that all the welds contained fine particles both along grain boundaries and in grain interiors in the SZ-AS. In the 304 FS weld, sigma phase was detected among these particles.[25,26]
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig7_HTML.gif
Fig. 7

Optical micrographs and typical TEM images of the weld

The nitrogen content analysis clarified the presence of nitrogen arising from the wear of the pcBN tool. In order to confirm the distribution of the solute elements, boron and nitrogen in the weld, relatively large-scale elemental analysis on the cross sections perpendicular to the welding direction was conducted for some particles in the SZ-AS by EPMA. The EPMA maps and line profiles of Cr, Ni, B, and N in austenitic stainless steel FS welds are shown in Figure 8. The corresponding backscattered electron images are also presented in Figure 8. Some particles contained high Cr and low Ni contents compared to the austenite matrix. A boron peak was detected in these particles, although the austenitic stainless steels used in the present study did not contain any boron, as shown in Table I. It is certain that the boron comes from pcBN tool, which indicates that some wear of the pcBN tool occurs during FSW. On the other hand, a nitrogen peak was not detected in some particles in the EPMA map, except for relatively large particles with at least 1-μm diameter, as shown in Figure 8 in 316L weld. Because the large particles contained lower Cr and Ni contents than the austenite matrix, they may have been from the pcBN debris.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig8_HTML.gif
Fig. 8

EPMA compositional map on the advancing side of the SZ in FS-welded austenitic stainless steel

Detailed TEM observations revealed that the SZ-AS had other precipitates different from the sigma phase, with various morphologies along the grain boundaries and in the grain interiors. Typical TEM image, SAED pattern, and EDX spectra of the particles obtained in the 304, 316L, and 310 FS weld by TEM and STEM are shown in Figures 9 through 11. The size of the precipitates ranged between about 100 and 1000 nm. These particles exhibited a high Cr content of about 50 to 60 wt pct, while the Ni content was very low. The boron peak is clearly found in the EDX spectra of particle, but the nitrogen peak is not detected. These SAED patterns suggest that the particles are Cr5B3 with tetragonal and Cr2B with orthorhombic crystallographic structures. The zone axis, index pattern, and measured lattice parameter of each particle are also shown in Figures 9 through 11. These are roughly coincident with the powder diffraction data, although the lattice parameter is 3 to 10 pct larger than the powder data.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig9_HTML.gif
Fig. 9

STEM-EDX analysis of the advancing side of the SZ in FS-welded 304 austenitic stainless steel. (a) Bright-field image, (b) SAED pattern, and (c) through (f) EDX spectra of the particle and the austenite matrix shown in (a)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig10_HTML.gif
Fig. 10

TEM-EDX analysis of the advancing side of the SZ in FS-welded 316L austenitic stainless steel. (a) Bright-field image, (b) SAED pattern, and (c) and (d) EDX spectra of the particle and the austenite matrix shown in (a)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig11_HTML.gif
Fig. 11

STEM-EDX analysis of the advancing side of the SZ in FS-welded 310 austenitic stainless steel. (a) Bright-field image, (b) SAED pattern, and (c) through (f) EDX spectra of the particle and the austenite matrix shown in (a)

It should be noted that nitrogen was not detected in most of the Cr-rich borides, although the SZ-AS exhibited about 2 to 5 times higher nitrogen than the austenite matrix of the BM, as shown in Figure 4. This suggests that the nitrogen comes from the pcBN during FSW but exists in the matrix after FSW. Because the solubility of nitrogen is higher in austenite than in ferrite, the nitrogen from the pcBN would dissolve into the austenite matrix during the boride formation. Therefore, tribochemical reaction arising from differences in the solubility of nitrogen as well as the flow stress would also be one of reasons why the nitrogen pickup during FSW is shown more significantly in the austenitic steel welds.

The TEM observation and large-scale elemental analysis showed a possibility that chemical compounds containing nitrogen as well as boron coexist with Cr borides in the advancing side of the austenitic stainless steel weld, as shown in Figure 8. The size of this chemical compound is 1 μm or more, which is larger than that in the Cr carbides observed in the present study. This evidence suggests that relatively large wear particles remain completely unreacted due to insufficient dissolution of BN into the matrix. The aforementioned results strongly suggest that boron and nitrogen from the tool react with the austenite matrix during FSW. The reaction would be expressed as follows:
$$ {\text{BN}} + 2{\text{Cr}} \to {\text{Cr}}_{{\text{2}}}{\text{B}} + {\text{N}} $$
$$ {\text{3BN}} + {\text{5Cr}} \to {\text{Cr}}_{{\text{5}}} {\text{B}}_{{\text{3}}} + {\text{3N}} $$

Nonoxide ceramic borides generally have strongly negative free energies of formation, giving them excellent stability under many conditions.[32] Because the available thermodynamic data on Cr borides are limited, it is difficult to calculate the standard free energy of the boride formation. It has been reported that the standard free energy of Cr5B3 boride formation is much lower than that of BN at 1500 K (1227 °C).[33] This temperature is roughly equal to the welding temperature during FSW of austenite stainless steels in the present study. A previous study[34] on Cr boride formation has shown that Cr2B and Cr5B3 borides can coexist at temperatures between 1273 K (1000 °C) and 1573 K (1300 °C) and that Cr2B is thermodynamically more stable than Cr5B3.[34] These data suggest that Cr-rich borides with crystallographic structure of Cr2B and Cr5B3 possibly form by the reaction between BN and austenite matrix during FSW. As mentioned previously, high nitrogen solubility in austenite may help these reactions during FSW.

On the other hand, there is a possibility that Fe-rich boride containing Cr, such as (Fe, Cr)2B, forms in the early stage of boride formation, because the affinity of Fe to B is greater than that of Cr.[35] However, (Fe, Cr)2B is less stable than (Cr, Fe)2B at high temperature due to the lower melting temperature of (Fe, Cr)2B. This would cause the substitution of Cr atoms for Fe atoms in the (Fe, Cr)2B phase to lower the free energy of the system. As the Cr content in the (Fe, Cr)2B exceeds the solubility limit, (Fe, Cr)2B begins to transform into the (Cr, Fe)2B phase by atomic shifts of B on {110} crystallographic planes.[35] It has been reported that the (Fe, Cr)2B → (Cr, Fe)2B phase transformation takes place through the generation of stacking faults in (Cr, Fe)2B along the [100] crystallographic direction.[35] The present study also found many borides with a number of stacking faults. Figure 12 shows an example of Cr2B having staking faults. Many stacking faults with one-dimensional disordered structures presented as “SF” are observed in the highly magnified bright-field image (Figure 12(b)). These faults lie on {100} crystallographic planes exactly perpendicular to the direction of streaking in Figure 12(c). These crystallographic features are in good agreement with a previous study.[35] This evidence suggests that the boride is formed through the BN → (Fe, Cr)2B → (Cr, Fe)2B phase transformation during FSW.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-008-9709-9/MediaObjects/11661_2008_9709_Fig12_HTML.gif
Fig. 12

(a) Bright-field image, (b) magnified TEM image, and (c) SAED pattern of the particle in the advancing side of the SZ in FS-welded 304 austenitic stainless steel

The analysis of nitrogen content and the Cr-rich borides shows that the increase in boron and nitrogen content was mainly found in the SZ-AS in austenitic stainless steels. This microstructural heterogeneity in the SZ could be attributed to the material movement from the retreating side to the advancing side at the trailing edge of the tool during FSW. The material undergoes large shear deformation along the pin surface toward the rotating direction during stirring.[31,36,37] The Cr borides and BN compounds from the tool are simultaneously captured into the material. The shear deformation probably moves most regions to the advancing side at the trailing edge of the tool, and then the regions accumulate around the SZ-AS. This would be one of the possible reasons why the higher boron and nitrogen contents are detected in the SZ-AS.

The present study clarified that the Cr-rich borides are formed in the SZ-AS in austenitic stainless steel FS welds when the pcBN tool severely wears during FSW. Because the Cr-rich borides consume the Cr, the Cr-depleted zone, whose width is generally a hundred and several tens of nanometers, would be created in the vicinity of the Cr-rich borides, as well as the sigma phase.[26] The Cr-depleted zone was not observed in the EPMA map, which is attributed to the fact that the spatial resolution of EPMA is usually larger than 1 μm. Coexistence of the Cr-rich boride and sigma phase would cause the strong contrast consisting of the corroded grain boundaries and pits in the SZ-AS on the cross sections (Figure 2). This result implies that suppression of the pcBN tool wear is a requirement to produce the high-quality welds without the preferentially corroded SZ-AS in austenitic stainless steels. However, the pcBN tool having the much higher wear resistance (an MS80 pcBN tool) has been recently developed, and the degree of the tool wear occurring during FSW of austenitic stainless steels is remarkably reduced. The effect of the pcBN tool grade on the tool wear will be reported in a separate article.

4 Conclusions

The FSW was applied to five types of ferritic, duplex, and austenitic stainless steels, and interstitial pickup and B-rich phase evolution were investigated in their welds. The largely increased nitrogen content was detected in the advancing side of the SZ in austenitic stainless steel FS welds. It was shown that the level of boron and nitrogen increases in the SZ of austenitic stainless steel FS welds. A possibility was suggested that tool wear during FSW can be attributed to high flow stresses in austenitic stainless steels. The increase boron and nitrogen in the SZ resulted in the formation of Cr-rich borides, of between 100- to 1000-nm diameter, with a crystallographic structure of Cr2B and Cr5B3 through the reaction between the boron and nitrogen and the matrix during FSW.

Footnotes
1

LECO is a trademark of LECO Corporation, St. Joseph, MI.

 
2

JEOL is a trademark of Japan Electron Optics Ltd., Tokyo.

 

Acknowledgments

The authors sincerely thank Professors K. Ikeda, K. Maruyama, T.W. Nelson, C.D. Sorensen, and Z.J. Wang, for many helpful suggestions and advice, and Mr. A. Honda, Mr. M. Doi, and H. Matsumoto for there technical assistance. Financial support from the Japan Ministry of Education, Culture, Sports, Science and Technology for the Promotion of Science with a Grant-in-Aid for Young Researchers and for the Global COE Program in Materials Integration International Center of Education and Research at Tohoku University is gratefully acknowledged.

Copyright information

© The Minerals, Metals & Materials Society and ASM International 2009