J. Cent. South Univ. Technol. (2011) 18: 953-959
DOI: 10.1007/s11771-011-0786-3
Bonding mechanism of X10CrNi18-8 with Ni/Al2O3 composite ceramic by pressureless infiltration
YANG Shao-feng(杨少锋), CHEN Wei-ping(陈维平), HAN Meng-yan(韩孟岩),
YANG Chao(杨超), ZHU De-zhi(朱德智)
National Engineering Research Center of Near-net-shape Forming for Metallic Materials,
South China University of Technology, Guangzhou 510640, China
? Central South University Press and Springer-Verlag Berlin Heidelberg 2011
Abstract: An alloy steel/alumina composite was successfully fabricated by pressureless infiltration of X10CrNi18-8 steel melt on 30% (mass fraction) Ni-containing alumina based composite ceramic (Ni/Al2O3) at 1 600 °C. The infiltration quality and interfacial bonding behavior were investigated by SEM, EDS, XRD and tensile tests. The results show that there is an obvious interfacial reaction layer between the alloying steel and the Ni/Al2O3 composite ceramic. The interfacial reactive products are (FexAly)3O4 intermetallic phase and (AlxCry)2O3 solid solution. The interfacial bonding strength is as high as about 67.5 MPa. The bonding mechanism of X10CrNi18-8 steel with the composite ceramic is that Ni inside the ceramic bodies dissolves into the alloy melt and transforms into liquid channels, consequently inducing the steel melt infiltrating and filling in the pores and the liquid channels. Moreover, the metallurgical bonding and interfacial reactive bonding also play a key role on the stability of the bonding interface.
Key words: pressureless infiltration; steel/alumina composites; interface bonding; infiltration mechanism
1 Introduction
Metal matrix composites (MMCs) combine high strength and hardness, wear and oxidation resistance of ceramics with ductility as well as thermal and electrical conductivity of metals. Therefore, they are very suitable for certain applications, e.g. aerospace, military, automotive industry, food and pharmaceutical industry. Among various MMCs, high-melting alloy steel matrix composites reinforced by oxide ceramics (e.g. Al2O3 and ZrO2) not only have various advantages of MMCs but also possess ideal performance at high temperature, which have been produced by melts infiltration [1–9] and tape casting [10]. In comparison with the tape casting, the melts infiltration is a more economic and attractive method. Due to special requirement for common mould caused by high melting point of alloying steel (above 1 000 °C), it is difficult to fabricate these MMCs by squeeze casting or gas pressure infiltration. So, pressureless infiltration may be a better choice, which relies on the wettability between ceramic and metal melts [3]. But the contact angle between the steel melt and oxide ceramics still remains above 90° at 1 600 °C [4]. This poor wettability was improved by Ti-activated infiltration [1-4] or coating metal layers (e.g. plating Ni) on a surface of alumina ceramic [5]. However, Ref.[1] reported that Ti particles inside the ceramic will block the melt infiltration channels when Ti content increases, and the processing of coating or plating Ni is very complex and expensive. Therefore, it becomes an interesting problem whether a more economic and simple method for the fabrication of alloy steel/alumina composites can be developed or not.
In the past, in order to achieve the infiltrating behavior of steel melts on alumina ceramic and the interface bonding, Ni particles with different contents (5%, 10%, 15% and 20%, mass fraction) were mixed into the ceramic bodies and Ni-containing alumina ceramic based composites were obtained by powder metallurgy method. The pressureless infiltration experiments indicated that when the Ni content was 20%, the firm bonding between the steel and ceramic samples could be formed, the strength reached about 60 MPa and the steel melt infiltration depth was above 400 μm. However, the mechanism for interface bonding and infiltration had not been revealed particularly. The focus of the present work is to analyze the bonding mechanism by using X10CrNi18-8 steel melt and 30% Ni-containing alumina ceramic based composite.
2 Experimental
The compositions of alumina ceramic and alloying steel are summarized in Table 1. A mixture of alumina powders (70%, mass fraction) and Ni powders (30%, mass fraction) were milled for 50 h with a QM-3SP2 ball mill. After that, the mixture powders were dried at 100 °C and held for 4 h, leached by 150 meshes boult, and pressed into column green bodies (diameter: 22 mm, height: 10 mm). The Ni-containing alumina based composite ceramic (briefly called AN30) with a diameter of about 20 mm was prepared by sintering the pressed green bodies at 1 450 °C with a dwell time of 2 h in vacuum. The X10CrNi18-8 steel samples were cut into cubes with dimensions of 8 mm×8 mm×8 mm (used for wetting experiments and microstructure observation) and 14 mm×14 mm×25 mm (used for bonding strength tests), respectively. After ground and cleaned with acetone in ultrasonic bath, the steel samples were placed on the top of AN30 plates. Then, the steel-AN30 samples were placed in corundum crucibles (aperture: 20-21 mm, height: 50 mm) for bonding experiment by pressureless infiltration. The infiltration processing was conducted in a graphite vacuum furnace, and the processing parameters were set as heating to 1 600 °C at a heating rate of 5 °C/min and holding for 4 h.
Table 1 Compositions of alumina ceramic and alloying steel
The microstructures of the obtained samples were observed by scanning electron microscopy (SEM, FEI Quanta 200) with an energy dispersive system (EDS, IE350MT). The elemental content of metal phase inside the fabricated alloy steel/alumina composite was analyzed by EDS. The microhardness of metal phase on both sides of the bonding interface was measured with a micro-Vickers measurement system (HVS-1000). The interface bonding strength was performed under tensile stress with a universal testing machine (SANS CMT 5105). Both surfaces of the tensile sample (Fig.1) were ground and adhered to special fixtures by an adhesive. The phase analysis for the bonding interface was studied by X-ray diffractometry (XRD, X’pert Pro MRD).
Fig.1 Crucible and X10CrNi18-8 steel/AN30 composite sample used for tensile test prepared by pressureless infiltration processing at 1 600 °C and holding for 4 h
3 Results and discussion
3.1 Wetting and infiltration analysis
In order to determine the wettability of X10CrNi18-8 steel and alumina ceramic with and without Ni addition, simple sessile drop experiments were made to estimate the contact angle by measuring the specimen after high-temperature infiltration. The contact angle of X10CrNi18-8 steel and alumina ceramic without Ni addition is about (115.5±0.5)°. Meanwhile, interface bonding does not occur. In comparison, the contact angle of X10CrNi18-8 steel and alumina based composite ceramic with 30% Ni addition decreases to (94±0.5)°, as shown in Fig.2, and their interface reaches an excellent bonding. Obviously, the addition of Ni particles in alumina ceramic improves not only its wettability by X10CrNi18-8 steel but also the interface bonding behavior.
Fig.2 Wettability of X10CrNi18-8 alloy steel and AN30 substrate
Figure 3 shows the BSE image of the sintered AN30 composite ceramic and the SEM morphologies of the cross-section of the alloy steel/alumina composite after infiltration. Figure 3(a) shows that the Ni particles are dispersed homogeneously inside the ceramic, and the particles size of Ni is different because of the existence of agglomerates. It is found from Fig.3(b) that excellent bonding between the alloy steel and AN30 emerges, and there are no cracks and pores at the bonding interface. High resolution observation reveals the formation of a transition interface layer with a thickness of about 6- 8 μm, as shown in Fig.3(c). The excellent bonding is further demonstrated by the EDS analysis of alloy particles inside the melt-infiltrated AN30 bodies. After infiltration, EDS analysis (Fig.3(d)) proves that new alloy particles with Ni, Cr and Fe are formed while pure Ni particles inside the starting AN30 specimen disappear (Fig.3(a)). This indicates that liquid steel melt has infiltrated into and reacted with the pure Ni particles. In addition, a few pores still exist inside the infiltrated ceramic bodies.
Fig.3 BSE image of AN30 specimen obtained by sintering at 1 450 °C and holding for 2 h (a); SEM morphology of cross-section of alloy steel/alumina composite, presenting microstructure of composite (b); formed interfacial transition layer (c); EDS pattern of alloy particles with different distances from bonding interface inside AN30 after infiltration (d): (d1) 30 μm; (d2) 300 μm; (d3) 400 μm
Figure 4 shows the variations of Fe, Cr and Ni contents of the metal phase with distance from the bonding interface inside the obtained alloying steel/alumina composite. It can be seen that Fe and Cr contents decrease and Ni content increases with the distance along the infiltration direction inside the infiltrated AN30 composite. When the distance from the bonding interface is about 400 μm, the Fe and Cr contents decrease to 7.93% and 9.32% (mass fraction), respectively; while the Ni content increases to 77.74% (mass fraction). The variations of Fe and Ni contents with distance from the bonding interface may be attributed to various infiltration resistances, e.g. the melt freezing resistance, the anti-pressure generated by remaining gas in pores of the sintered ceramic bodies, the viscous resistance of steel melt, and the friction between melt and ceramic wall. The inverse variation trend takes place in the alloy steel with the distance from the bonding interface, which is due to the fact that Ni element in AN30 can diffuse into the steel melt. The variations of the Fe, Cr and Ni contents in both sides of the bonding interface indicate that an elemental diffusion zone exists adjacent to the bonding interface.
Fig.4 Variations of Fe, Cr and Ni contents of metal phase with distance from bonding interface
3.2 Microhardness test
In order to prove the existence of content variations of Fe, Cr and Ni with distance from bonding interface, microhardness test was carried out accordingly. Figure 5 shows the hardness values of metal phase with distance from the bonding interface. In the alloying steel’s side, the hardness value on the positions of P1 and P2 are HV433 and HV434, the distance of which is over 100 μm from the bonding interface. In zones adjacent to the bonding interface, the steel’s hardness decreases to HV398 on P3. In the infiltrated AN30 side, compared with the microhardness of starting Ni particles (HV210), the hardness of the alloy particles as seen in Fig.3(b) increases gradually with decreasing the distance from the bonding interface. When the distance from the bonding interface is about 50 μm, the hardness on P4 is about HV370. The variation of the hardness for the alloy particles reveals the formation of new alloying phases, which is in good agreement with the EDS analysis results in Fig.3(c). Therefore, the variation of the hardness in both sides of the bonding interface demonstrates again that, the alloy elements can diffuse between the alloying steel and the AN30 composite specimen, which forms an element diffusion zone with a thickness of above 600 μm adjacent to the bonding interface.
Fig.5 Microhardness values of metal phase with distance from bonding interface
3.3 Bonding strength and phase characterization of X10CrNi18-8/AN30 interface
Figure 6 shows the axial tensile loading- displacement curve for the fabricated composite sample in Fig.1. Apparently, the maximum loading value reaches 22.27 kN, and the diameter of the composite sample is about 20.5 mm. The bonding strength is as high as about 67.5 MPa determined by the formula σ=F/πr2 (σ is the strength, MPa; F is the loading, kN). BAO et al [10] reported that the bonding strength of alloying steel and alumina with Ni-coating is only 4.045 MPa (shearing strength). Reference [11] reported that alumina ceramic coatings were fabricated on 304 stainless steel by cathodic plasma electrolytic deposition, and the maximum bonding strength is 22 MPa. So, it can be concluded that, the method of infiltration of alloying steel melt on the Ni-containing alumina based ceramic composite has obviously improved the bonding behavior.
Fig.6 Room-temperature tensile loading–displacement curve for X10CrNi18-8 steel/AN30 composite sample in Fig.1
Figure 7(a) shows a transition layer formed on the surface of the infiltrated AN30 composite. The transition layer is smooth and compact. No dimple structure of ductile fracture occurs, which indicates that a brittle fracture occurs between the alloying steel and the transition layer. There are few of pores formed by the pull-out of alloying particles, and this is the cause of the existence of little strain under tensile stress. Also, cracking emerges inside the infiltrated AN30 composite beside on the transition layer, as seen in Fig.7(b). Parts of fractured bodies are still strongly adhered to the surface of the steel specimen. These reveal that the bonding force is diverse in different zones at the bonding interface.
Fig.7 Fracture surfaces of steel melt-infiltrated AN30 composite: (a) Transition layer showing formed pores due to pull-out of alloy particles; (b) Different fracture zones on surface of infiltrated AN30 composite (M: Formed Ni-Fe-Cr alloy particles; N: Fracture zone within infiltrated AN30 composite)
Figure 8 shows the XRD pattern of the fracture surface of the infiltrated AN30 composite in Fig.7. By the index of the XRD patterns, it is found that except for the austenitic phase in the X10CrNi18-8 steel, two new compound phases of (FexAly)3O4 and (AlxCry)2O3 are identified, and their elemental compositions are (Fe0.793Al0.207)(Fe1.793Al0.207)O4 and (Al0.9Cr0.1)2O3, espectively. Formation of new compound phases has been reported, e.g. the FeAl2O4 spinel [12] by the reaction between Fe and Al2O3 sintered at 1 700 °C in vacuum for 1.5 h and the (Fe,Al)3O4 phase [13] investigated by phase equilibria in the Al2O3-Fe2O3 system in air and at 0.1 MPa O2 pressure by equilibration and quenching. Meanwhile, it has been proved that there is dissolution of the oxide in liquid steel melt at 1 600 °C [1], which can affect the wettability at the steel/ceramic interface. In this study, the oxidation of Fe and Cr due to the remains of oxygen in the environment can not completely be excluded. LIN et al [14] has reported when the Cr/Al2O3 composite is sintered, Cr can be oxidized and Cr2O3 can form; Al2O3 and Cr2O3 can form substitutional solid solution at high temperature. This can lead to the shift of the Al2O3 peaks to lower angles of the substitutional solid solution in the XRD patterns. It can be concluded that the transition layer formed at the X10CrNi18-8/ alumina interface is composed of (FexAly)3O4 and (AlxCry)2O3, which also plays a role in decreasing the contact angle. In the present cases, the formation of the (FexAly)3O4 and (AlxCry)2O3 compound phases can be understood as follows:
2.586(Fe)+3.379(O)+0.207Al2O3→
(Fe0.793Al0.207) (Fe1.793Al0.207)O4 (1)
2(Cr)+3(O)→Cr2O3 (2)
0.1Cr2O3+0.9Al2O3→(Al0.9Cr0.1)2O3 (3)
Fig.8 XRD pattern for fracture surface of infiltrated AN30 composite in Fig.7
3.4 Discussion
3.4.1 Infiltration mechanism
The metal melt infiltration behavior lies on its wettability with ceramic in vacuum, and the relation of pressureless infiltration depth and contact angle is derived by WANG et al [15]:
H=R/2·{[(P1+(2·σ·cosθ)/R)·τ]/η}1/2 (4)
where H is the infiltration depth, R is the capillary radius, P1 is the vacuum suction force, σ is the liquid metal surface tension, θ is the contact angle, τ is the dwell time, and η is the liquid metal viscous value. It can be concluded from Eq.(4), when ceramic can not be wetted by metal liquid, θ is above 90° and 2σcosθ/R is below 0. If |P1| is less than |2σcosθ/R|, Eq.(4) means that metal liquid infiltration can not occur. In general, when the metal liquid can not wet ceramics, an external force generated by squeeze casting or gas pressure infiltration is needed to achieve infiltration. In these cases, their mechanism is that the external force makes |P1| greater than |2σcosθ/R|.
In the present study, although the contact angle θ between X10CrNi18-8 melt and AN30 is above 90° and no external force is applied, infiltration occurs. It can be seen from Fig.9(a) that, the infiltrated alloying steel can form interconnected agglomerates inside the ceramic bodies. Such infiltration behavior can be explained below. Ni particles are dispersed homogeneously inside the AN30 specimen; the pore size in the AN30 is about 1- 10 μm (Fig.3(a)). Both gravity force and capillary force play a major role in melt infiltration according to LEMSTER et al’s report [3]. At 1 600 °C and with a dwell time of 4 h, steel melt initially meets Ni melt distributed on the surface of the AN30 plate, then wetting and dissolving occur between steel and Ni. In this case, the Ni particles inside the AN30 have formed interconnected liquid channels similar to the capillaries, which can be well wetted by the steel melt. Under the gravity force and capillary force driving, steel melt can infiltrate and fill up the pores through the liquid Ni channels (Fig.9(b)).
Fig.9 Steel melts-infiltrated AN30 composite, presenting interconnected Ni melts channels formed at 1 600 °C marked by white lines (a), and proposed model of melts infiltration (b)
3.4.2 Bonding mechanism
As mentioned above, the interface between the X10CrNi18-8 steel and AN30 specimen reaches excellent bonding and the bonding strength value is so high. The stability of bonding interface mainly relies on four styles of forces as shown in Fig.9(b). 1) At high temperatures, the steel melt and Ni melt inside alumina ceramic are contacted and the Fe, Cr and Ni elements can diffuse in both melts, which forms a good metallurgical bonding force (Fm). 2) Steel melt can infiltrate and fill up pores of ceramic through Ni liquid channels, further combine steel with ceramic, which can be called infiltrating bonding force (Fi). 3) The formation of a transition layer composed of the (FexAly)3O4 and (AlxCry)2O3 phases between the X10CrNi18-8 steel and alumina ceramic, which can be called reaction bonding (Fr). 4) The surface tension (Fs) makes shrinking of steel melt surface during cooling. Only if the surface shrinking effect is weaker than that of comprehensive interfacial bonding effect caused by Fm, Fi and Fr, the bonding interface will be stable. According to the experimental results, the X10CrNi18-8/AN30 system obviously meets this requirement. However, when X10CrNi18-8 steel contacts with pure alumina substrate, there is neither metallurgical bonding nor infiltrating bonding, where the surface shrinking effect is stronger than the interfacial bonding effect. Therefore, the steel is easily pulled off from the pure alumina substrate.
These basic experiments have convincingly demonstrated that it is possible to improve the bonding behavior between the alloying steel and ceramic by adding Ni particles to the ceramic bodies. There is no doubt that this simple method can be used to fabricate the alloying steel matrix composites with the layered ceramic or ceramic coatings using the infiltration of metal melt. Besides, three-dimensional interconnected structure of alloying steel/ceramic composites with ideal bonding interfaces can be obtained, if the porous Ni-containing ceramic composite frame is prepared, e.g. impregnating a polymeric sponge with aqueous ceramic slurry.
4 Conclusions
At 1 600 °C with a dwell time of 4 h in vacuum, X10CrNi18-8 steel can bond firmly with the 30% Ni-containing alumina based ceramic composite, and the bonding strength reaches about 67.5 MPa. Steel melt can infiltrate into the ceramic composite through the liquid channels formed by the Ni particles, further dissolve and react with Ni, which forms new Ni-Fe-Cr alloying particles. The pull-out of alloy particles embedded in ceramic matrix after infiltration increases the interface bonding strength. Adjacent to the bonding interface, there exists a diffusion zone of alloying elements. An interface reaction emerges between the X10CrNi18-8 steel and alumina, and the reactive products are (FexAly)3O4 intermetallic and (AlxCry)2O3 solid solution. Besides, the metallurgical bonding can be formed by steel and Ni inside the ceramic bodies.
References
[1] BAHRAINI M, MINGHETTI T, ZOELLIG M, SCHUBER T J, BERROTH K, SCHELLE C, GRAULE T, KUEBLER J. Activated pressureless in?ltration of metal-matrix composites with graded activator content [J]. Composites: Part A, 2009, 40: 1566-1572.
[2] LEMSTER K, DELPORTE M, GRAULE T, KUEBLER J. Activation of alumina foams for fabricating MMCs by pressureless in?ltration [J]. Ceram Inter, 2007, 33: 1179-1185.
[3] LEMSTER K, GRAULE T, KUEBLER J. Processing and microstructure of metal matrix composites prepared by pressureless Ti-activated in?ltration using Fe-base and Ni-base alloys [J]. Mater Sci Eng A, 2005, 393: 229-238.
[4] VASIC S, GROB?TY B, KUEBLER J, GRAULE T. Activation mechanism and infiltration kinetic for pressureless melt infiltration of Ti activated Al2O3 preforms by high melting alloy [J]. Adv Eng Mater, 2009, 11: 658-665.
[5] WANG En-ze, XU Yan-ping, BAO Chong-gao, XING Jian-dong. Fabrication of Al2O3/heat-resistance steel composite and its wear-resistance at high temperature and abrasive [J]. Acta Materiae Compositae Sinica, 2004, 21: 56-60. (in Chinese)
[6] WITTIG D, ANEZIRIS C G, GRAULE T, KUEBLER J. Mechanical properties of three-dimensional interconnected alumina/steel metal matrix composites [J]. J Mater Sci, 2009, 44: 572-579.
[7] MANFREDI D, PAVESE M, BIAMINO S, FINO P, BADINI C. NiAl(Si)/Al2O3 co-continuous composites by double reactive metal penetration into silica reforms [J]. Composites Sci Tech, 2008, 16: 580-583.
[8] BAHRAINI M, MINGHETTI T, ZOELLIG M, SCHUBERT J, BERROTH K, SCHELLE C, GRAULE T, KUEBLER J. Activated pressureless infiltration of metal-matrix composites with graded activator content [J]. Composites: Part A, 2009, 40: 1566-1572.
[9] ZHANG S Y, LI Y Y, QU S G, CHEN W P. Friction and wear behaviour of brake pads dry sliding against semi-interpenetrating network ceramics/Al-alloy composites [J]. Tribol Lett, 2010, 38: 135-146.
[10] BAO Chong-gao, WANG En-ze, GAO Yi-min, XING Jian-dong. Effect of interface strength of Al2O3/steel on wear resistance of composites at high temperature [J]. Journal of Xi’an Jiaotong University, 2000, 34: 41-45.
[11] WANG Y L, JIANG Z H, LIU X R, YAO Z D. In?uence of treating frequency on microstructure and properties of Al2O3 coating on 304 stainless steel by cathodic plasma electrolytic deposition [J]. Appl Surf Sci, 2009, 255: 8836-8840.
[12] KONOPKA K, OZI?B?O A. Microstructure and the fracture toughness of Al2O3-Fe composites [J]. Mater Characterial, 2001, 46: 125-129.
[13] ROBERT H, PETER C H, EVGUENI J. Experimental study of phase equilibria in the Al-Fe-Zn-O system in air [J]. Metallurgical and Materials Transactions B, 2004, 35: 633-642.
[14] LIN H T, WANG S C, HUANG J L, CHANG S Y. Processing of hot pressed Al2O3-Cr2O3/Cr-carbide nanocomposite prepared by MOCVD in ?uidized bed [J]. J Eur Ceram Soc, 2007, 27: 4759-4765.
[15] WANG En-ze, ZHENG Yan-qing, XING Jian-dong, BAO Chong-gao, XU Xue-wu, WANG Hai-yan. Fabricating steel-based composites containing ceramic particles with vacuum infiltration process [J]. Acta Materiae Compositae Sinica, 1998, 15: 12-17. (in Chinese)
(Edited by PENG Chao-qun)
Foundation item: Project(2009ZM0296) supported by the Fundamental Research Funds for the Central Universities in China
Received date: 2010-04-09; Accepted date: 2010-01-30
Corresponding author: CHEN Wei-ping, Professor, PhD; Tel: +86-20-87113832; E-mail: mewpchen@scut.edu.cn
