MATERIALS HIENGE& ENGIEERING ELSEVIER Materials Science and Engineering A 444 (2007)120-129 www.elsevier.com/locate/msea Erosion wear behaviours of Sic/(w,Ti)c laminated ceramic nozzles in dry sand blasting processes Deng jianxin. Liu Lili. dir Department of Mechanical Engineering, Shandong University, Jinan 250061, Shandong Province, PR China Received 13 April 2006: accepted 17 August 2006 In sand blasting processes, the nozzle entry region suffers form severe abrasive impact, which may cause large tensile stress and lead to an increased erosion wear at the nozzle entry area. In this paper, SiC/W, Ti)c laminated ceramic nozzles were produced by hot pressing. The purpose to reduce the tensile stress at the entry region of the nozzle. Due to the different thermal expansion coefficients and shrinkage of the Sic and W,Ti)C solid-solution, the entry region of the Sic/(w,Ti)C laminated ceramic nozzles in the fabricating process exhibit a compressive residual stress. The value of this residual stress was calculated by means of the finite element method. The erosion wear behaviour of the laminated ceramic nozzle and of a stress-free nozzle, with the same composition, was assessed by dry sand blasting Results showed that the laminated ceramic nozzles have superior erosion wear resistance to that of the homologous stress-free nozzles. The mechanism responsible was explained as the formation of compressive residual stresses in nozzle entry region in fabricating process of the laminated ceramic nozzles, which may partially counteract the tensile stresses resulting from extermal loadings, and leads to an improvement of erosion wear resistance. It is indicated that laminated structures in ceramic nozzles is an effective way to improve the erosion wear resistance of the stress-free nozzles C 2006 Elsevier B. V. All rights reserved. Keywords: Nozzles: Ceramic materials: Laminated materials: SiC 1. Introduction The nozzle is the most critical part in the sand blasting treatment equipment. There are many factors that influence the Sand blasting treatment is an abrasive machining process and nozzle wear such as: the mass flow rate and impact angle [5-7] is widely used for surface strengthening[l], surface modification the erodent abrasive properties [8-10, the nozzle material and 21, surface clearing and rust removal [3, 4], etc. It is suitable for its geometry [11-16], and the temperatures [17, 18]. Ce the treatment of hard and brittle materials, ductile metals, alloys, being highly wear resistance, have great potential as and nonmetallic materials, and can provide perfect surface treat- blasting nozzle materials ment to all kinds workpieces from hull, steel structure, container Several studies [11-15] have shown that the entry area of to watchcase, button and inject needle In the sand blasting pro- a ceramic nozzle exhibited a brittle fracture induced removal cess, a very high velocity jet of fine abrasive particles and carrier process, while the centre area showed plowing type of material gas coming out from a nozzle impinges on the target surface and removal mode. As the erosive particles hit the nozzle at high erodes it. The fine particles are accelerated by the gas stream, angles(nearly 90%) at the nozzle entry section in sand blasting commonly compressed air at a few times atmospheric pressure. (see Fig. 1), the nozzle entry region suffers form severe abrasive The particles are directed towards the surfaces to be treated. impact, which may cause large tensile stresses. The stress alon As the particles impact the surface, they cause a small fracture, the axial direction of the nozzle decreases from entry to centre, and the gas stream carries both the abrasive particles and the and increases from centre to exit. The highest tensile stresses fractured particles away are located at the entry region of the nozzle. While the wear of the nozzle centre area changes from impact to slidin the tensile stresses caused by the abrasive impact in this area are much smaller than those at the entry section. Thus, the erosion Corresponding author. Tel: +86 531 88392047 wear of the nozzle entry region is always serious in contrast with E-mail address sdu.edu. cn(D. Jianxin) that of the centre area [11-15] )6 Elsevier B v. All rights reserved
Materials Science and Engineering A 444 (2007) 120–129 Erosion wear behaviours of SiC/(W,Ti)C laminated ceramic nozzles in dry sand blasting processes Deng Jianxin ∗, Liu Lili, Ding Mingwei Department of Mechanical Engineering, Shandong University, Jinan 250061, Shandong Province, PR China Received 13 April 2006; accepted 17 August 2006 Abstract In sand blasting processes, the nozzle entry region suffers form severe abrasive impact, which may cause large tensile stress and lead to an increased erosion wear at the nozzle entry area. In this paper, SiC/(W,Ti)C laminated ceramic nozzles were produced by hot pressing. The purpose is to reduce the tensile stress at the entry region of the nozzle. Due to the different thermal expansion coefficients and shrinkage of the SiC and (W,Ti)C solid-solution, the entry region of the SiC/(W,Ti)C laminated ceramic nozzles in the fabricating process exhibit a compressive residual stress. The value of this residual stress was calculated by means of the finite element method. The erosion wear behaviour of the laminated ceramic nozzle and of a stress-free nozzle, with the same composition, was assessed by dry sand blasting. Results showed that the laminated ceramic nozzles have superior erosion wear resistance to that of the homologous stress-free nozzles. The mechanism responsible was explained as the formation of compressive residual stresses in nozzle entry region in fabricating process of the laminated ceramic nozzles, which may partially counteract the tensile stresses resulting from external loadings, and leads to an improvement of erosion wear resistance. It is indicated that laminated structures in ceramic nozzles is an effective way to improve the erosion wear resistance of the stress-free nozzles. © 2006 Elsevier B.V. All rights reserved. Keywords: Nozzles; Ceramic materials; Laminated materials; SiC 1. Introduction Sand blasting treatment is an abrasive machining process and is widely used for surface strengthening [1], surface modification [2], surface clearing and rust removal [3,4], etc. It is suitable for the treatment of hard and brittle materials, ductile metals, alloys, and nonmetallic materials, and can provide perfect surface treatment to all kinds workpieces from hull, steel structure, container to watchcase, button and inject needle. In the sand blasting process, a very high velocity jet of fine abrasive particles and carrier gas coming out from a nozzle impinges on the target surface and erodes it. The fine particles are accelerated by the gas stream, commonly compressed air at a few times atmospheric pressure. The particles are directed towards the surfaces to be treated. As the particles impact the surface, they cause a small fracture, and the gas stream carries both the abrasive particles and the fractured particles away. ∗ Corresponding author. Tel.: +86 531 88392047. E-mail address: jxdeng@sdu.edu.cn (D. Jianxin). The nozzle is the most critical part in the sand blasting treatment equipment. There are many factors that influence the nozzle wear such as: the mass flow rate and impact angle [5–7], the erodent abrasive properties [8–10], the nozzle material and its geometry [11–16], and the temperatures [17,18]. Ceramics, being highly wear resistance, have great potential as the sand blasting nozzle materials. Several studies [11–15] have shown that the entry area of a ceramic nozzle exhibited a brittle fracture induced removal process, while the centre area showed plowing type of material removal mode. As the erosive particles hit the nozzle at high angles (nearly 90◦) at the nozzle entry section in sand blasting (see Fig. 1), the nozzle entry region suffers form severe abrasive impact, which may cause large tensile stresses. The stress along the axial direction of the nozzle decreases from entry to centre, and increases from centre to exit. The highest tensile stresses are located at the entry region of the nozzle. While the wear of the nozzle centre area changes from impact to sliding erosion, the tensile stresses caused by the abrasive impact in this area are much smaller than those at the entry section. Thus, the erosion wear of the nozzle entry region is always serious in contrast with that of the centre area [11–15]. 0921-5093/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2006.08.090
D Jianxin et al. Materials Science and Engineering A 444 (2007)120-129 Nozzle Shallow impact or abrasive zone Fig. 1. Schematic diagram of the interaction between the erodent particle and the nozzle in sand blasting processe A novel technique, by which a compressive residual stress was investigated in comparison with the homologous stress-free n be generated into the surface of a material, is the produc- nozzles. The purpose is to evaluate whether laminated ceramic tion of laminated structures designed to combine the advanta- nozzles have superior erosion wear resistance to that of the geous characteristics of the different materials involved, thereby homologous stress-free nozzles. improving the overall mechanical behaviour of the materials. It has been shown that laminated hybrid structures constituted by 2. Materials and experimental procedures alternate layers of different materials can be properly designed in order to induce a surface compressive residual stress leading to 2.1. Preparation of sic/(w, Ti)C laminated ceramic nozzle an improved surface mechanical properties and wear resistance materials [19-22]. Residual stresses arise from a mismatch between the coefficients of thermal expansion(CTE), sintering rates and elas- The starting materials were(W,Ti C solid-solution powders ic constants of the constituent phases and neighbouring layers. with average grain size of approximately 0.8 um, purity 99.9%0 Compressive residual stresses are induced in layers with lower and Sic powders with average grain size of 1 um, purity 99.8%. CTE, while tensile stresses arise in those with higher CtE. The Six different volume fractions of (W,Ti)C(55, 57, 59, 61, 63, residual stress field also depends on the geometry of the layered 65 vol %)were selected in designing the SiC/(W,Ti)C laminated structure and on the thickness ratio among layers [23-26] nozzle material with a six-layer structure The effectiveness of laminated hybrid structures in improving The compositional distribution of the laminated ceramic noz the sliding wear resistance of alumina has been already reported zle materials is shown in Fig. 2. It is indicated that the composi by Toschi [22]. In the present study, Sic/(W,Ti)C laminated tional distribution of the laminated nozzle materials changes in ceramic nozzles were produced by hot pressing. The residual nozzle axial direction(see Fig. 2(a)and( b)). As the heat conduc thermal stress of the laminated nozzle in the fabricating process tivity of SiC is higher than that of (w,Ti)C solid-solution, while was calculated by means of the finite element method(FEM). its thermal expansion coefficient is lower than that of (,Ti)C, The erosion wear behaviour of the laminated ceramic nozzle the layer with the highest volume fraction of SiC was put in the Nozzle exit Nozzle exit Nozzle exit C/65Vol%(W,Tn)c iC/57Vol%(W,TO)C SiC/55Vol%(W, TO)C Nozzle entry Nozzle entry Nozzle entry Fig. 2. Compositional distribution of (a) the ceramic nozzle laminated only in entry area. (b) the ceramic nozzle laminated both in entry and exit area, and (c)the stress-free nozzles (a) GN-2 laminated nozzle, (b) GN-3 laminated nozzle, (c)CN-2 stress-free nozzle
D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 121 Fig. 1. Schematic diagram of the interaction between the erodent particle and the nozzle in sand blasting processes. A novel technique, by which a compressive residual stress can be generated into the surface of a material, is the production of laminated structures designed to combine the advantageous characteristics of the different materials involved, thereby improving the overall mechanical behaviour of the materials. It has been shown that laminated hybrid structures constituted by alternate layers of different materials can be properly designed in order to induce a surface compressive residual stress leading to an improved surface mechanical properties and wear resistance [19–22]. Residual stresses arise from a mismatch between the coefficients of thermal expansion (CTE), sintering rates and elastic constants of the constituent phases and neighbouring layers. Compressive residual stresses are induced in layers with lower CTE, while tensile stresses arise in those with higher CTE. The residual stress field also depends on the geometry of the layered structure and on the thickness ratio among layers [23–26]. The effectiveness of laminated hybrid structures in improving the sliding wear resistance of alumina has been already reported by Toschi [22]. In the present study, SiC/(W,Ti)C laminated ceramic nozzles were produced by hot pressing. The residual thermal stress of the laminated nozzle in the fabricating process was calculated by means of the finite element method (FEM). The erosion wear behaviour of the laminated ceramic nozzle was investigated in comparison with the homologous stress-free nozzles. The purpose is to evaluate whether laminated ceramic nozzles have superior erosion wear resistance to that of the homologous stress-free nozzles. 2. Materials and experimental procedures 2.1. Preparation of SiC/(W,Ti)C laminated ceramic nozzle materials The starting materials were (W,Ti)C solid-solution powders with average grain size of approximately 0.8 m, purity 99.9%, and SiC powders with average grain size of 1 m, purity 99.8%. Six different volume fractions of (W,Ti)C (55, 57, 59, 61, 63, 65 vol.%) were selected in designing the SiC/(W,Ti)C laminated nozzle material with a six-layer structure. The compositional distribution of the laminated ceramic nozzle materials is shown in Fig. 2. It is indicated that the compositional distribution of the laminated nozzle materials changes in nozzle axial direction (see Fig. 2(a) and (b)). As the heat conductivity of SiC is higher than that of (W,Ti)C solid-solution, while its thermal expansion coefficient is lower than that of (W,Ti)C, the layer with the highest volume fraction of SiC was put in the Fig. 2. Compositional distribution of (a) the ceramic nozzle laminated only in entry area, (b) the ceramic nozzle laminated both in entry and exit area, and (c) the stress-free nozzles. (a) GN-2 laminated nozzle, (b) GN-3 laminated nozzle, (c) CN-2 stress-free nozzle.
D Jianxin et al./ Materials Science and Engineering A 444(2007)120-129 Fig. 3. Schematic diagram of the sand blasting machine tool(1, air compressor, 2, control valve; 3, filter; 4, desiccator; 5. press adjusting valve; 6, dust catcher, 7, blasting gun: 8, abrasive hopper: 9, ceramic nozzle). 851115K沁总;:mm nozzle entry with the compositional distribution changing from the entry layer to the exit layer with the lowest volume frac Fig. 5. SEM micrograph of the SiC abrasives used for dry sand blasting ion of SiC(see Fig. 2(a)). While in Fig. 2(b), the compositional istribution of the laminated ceramic nozzle is symmetrical, the trolled by the valves and regulators. The abrasive air jet is formed layer with the highest volume fraction of Sic was put both in in the blasting gun using a suction-type process as schematically the entry layer and in the exit layer. The homologous stress-free illustrated in Fig. 4. The gas flow rate is controlled by the com- nozzle with no compositional change is shown in Fig. 2(c).The pressed air, and the abrasive particle velocity through the nozzle ceramic nozzle laminated only in entry area is named GN-2, the is adjusted to 60m/s ceramic nozzle laminated both in entry and exit area is named The erodent abrasives used in this study were of silicon GN-3. while the stress-free nozzle is named CN-2. carbide(Sic) powders with 50-150 um grain size. The SEM Six Sic/(w,Ti)c composite powders of different mixture micrograph of the SiC powders used for the dry sand blasting is ratios were prepared by wet ball milling in alcohol with shown in Fig. 5. As these abrasives are more durable and create cemented carbide balls for 80h. Following drying, the mixtures less dust than sand, and typically are reclaimed and reused. composite powders with different mixture ratios were laminate Nozzles with internal diameter mm and length 30 mm made into the mould in turn. The sample was then hot-pressed in flow- from SiC/(W, Ti)C laminated structure(GN-2 and GN-3)and ing nitrogen for 40 min at 1900C temperature with 30MPa stress-free structure(CN-2)were manufactured by hot-pressing as can be seen in Fig. 6. The mass loss of the worn nozzles was measured with an accurate electronic balance(minimum 2.2. Sand blasting tests 0. 1 mg). All the test conditions are listed in Table 1. The erosion rates(W) of the nozzles are defined as the nozzle mass loss(mD) .. Fig. 3 shows the schematic diagram of the abrasive air-jet divided by the nozzle density (d) times the mass of the erodent chine tool(GS-6 type), which consists of an air compressor, abrasive particles(m2) a blasting gun, a control valve, a particle supply tube, a filter, a desiccator, an adjusting press valve. a dust catcher, an abra- W= ml (1) sive hopper, and a nozzle. The dust catcher was used to prevent fugitive dust emissions. The air and grit flow adjusting was con- where the Whas the units of volume loss per unit mass(mm/g) Abrasive flow orifice Fig. 4. Schematic diagram of blasting gun structure(1, gun support; 2, air flow nozzle; 3, adjusting gasket; 4, ceramic nozzle; 5, plastic jacket for the nozzle)
122 D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 Fig. 3. Schematic diagram of the sand blasting machine tool (1, air compressor; 2, control valve; 3, filter; 4, desiccator; 5, press adjusting valve; 6, dust catcher; 7, blasting gun; 8, abrasive hopper; 9, ceramic nozzle). nozzle entry with the compositional distribution changing from the entry layer to the exit layer with the lowest volume fraction of SiC (see Fig. 2(a)). While in Fig. 2(b), the compositional distribution of the laminated ceramic nozzle is symmetrical, the layer with the highest volume fraction of SiC was put both in the entry layer and in the exit layer. The homologous stress-free nozzle with no compositional change is shown in Fig. 2(c). The ceramic nozzle laminated only in entry area is named GN-2, the ceramic nozzle laminated both in entry and exit area is named GN-3, while the stress-free nozzle is named CN-2. Six SiC/(W,Ti)C composite powders of different mixture ratios were prepared by wet ball milling in alcohol with cemented carbide balls for 80 h. Following drying, the mixtures composite powders with different mixture ratios were laminated into the mould in turn. The sample was then hot-pressed in flowing nitrogen for 40 min at 1900 ◦C temperature with 30 MPa pressure. 2.2. Sand blasting tests Fig. 3 shows the schematic diagram of the abrasive air-jet machine tool (GS-6 type), which consists of an air compressor, a blasting gun, a control valve, a particle supply tube, a filter, a desiccator, an adjusting press valve, a dust catcher, an abrasive hopper, and a nozzle. The dust catcher was used to prevent fugitive dust emissions. The air and grit flow adjusting was conFig. 5. SEM micrograph of the SiC abrasives used for dry sand blasting. trolled by the valves and regulators. The abrasive air jet is formed in the blasting gun using a suction-type process as schematically illustrated in Fig. 4. The gas flow rate is controlled by the compressed air, and the abrasive particle velocity through the nozzle is adjusted to 60 m/s. The erodent abrasives used in this study were of silicon carbide (SiC) powders with 50–150 m grain size. The SEM micrograph of the SiC powders used for the dry sand blasting is shown in Fig. 5. As these abrasives are more durable and create less dust than sand, and typically are reclaimed and reused. Nozzles with internal diameter 8 mm and length 30 mm made from SiC/(W,Ti)C laminated structure (GN-2 and GN-3) and stress-free structure (CN-2) were manufactured by hot-pressing as can be seen in Fig. 6. The mass loss of the worn nozzles was measured with an accurate electronic balance (minimum 0.1 mg). All the test conditions are listed in Table 1. The erosion rates (W) of the nozzles are defined as the nozzle mass loss (m1) divided by the nozzle density (d) times the mass of the erodent abrasive particles (m2): W = m1 (d × m2) (1) where the W has the units of volume loss per unit mass (mm3/g). Fig. 4. Schematic diagram of blasting gun structure (1, gun support; 2, air flow nozzle; 3, adjusting gasket; 4, ceramic nozzle; 5, plastic jacket for the nozzle).
D Jianxin et al. Materials Science and Engineering A 444 (2007)120-129 Hardness of different layers of the SiC/(,TiC laminated nozzle(GN-2) Layer (W,TiC content (vol %) Vickers hardness, Hy(GPa) 26.52 3 4 5 24.67 6 where P is the indentation load (N), 2a is the catercorner length (um) due to indentation. Hardness of each layer of SiC/(W,Ti)C laminated nozzle(GN-2)material is presented in Table 2 ig. 7 illustrates -ray diffraction analysis Sic/W,Ti)C laminated ceramic nozzle(GN-2) material after Fig. 6. Photo of the SiC/w,Ti)C laminated ceramic nozzles. sintering at 1900 C for 40 min. It can be seen that both (w,Tic and Sic existed in the sintered specimens. SEM micrograph of each polished layer of Sic/(W,Ti)C laminated ceramic noz- The finite element method(FEM) was used as a means of zle(Gn-2)material are shown in Fig 8. The black areas were numerically evaluating the residual thermal stress and its dis- identified by EDX analysis as SiC, and the white phases with tribution of the laminated ceramic nozzle in the fabricating clear contrast were(W,Ti)C. It can be seen that the Sic particles processes are quite uniformly distributed throughout the microstructure. For observation of the micro-damage and determination of porosity is virtually absent. erosion mechanisms, the worn nozzles were sectioned axially The eroded bore surfaces of the nozzles were examined by scan- ning electron microscopy. 3.2. Residual thermal stress analysis of sic/(W,Ti)C laminated nozzle material 3. Results and discussion The residual thermal stress of the laminated ceramic noz- zle in the fabricating process was calculated by means of the 3.1. Microstructural characterization and properties of finite element method by assuming that the compact is cooled Sic(W, Ti)C laminated nozzle materials from sintering temperature 1900C to room temperature 20oC Thermo-mechanical properties of (W,Ti)C and Sic are as fol- Hardness measurements were performed by placing Vick- lows: rs indentations on every layer of the cross-sectional surface of SiC/W,Ti)C laminated nozzle (GN-2)material. The indentation (W, Ti)C: E=480 GPa, v=0.25, a=85x10 K load was 200N and a minimum of three indentations were tested for each layer. The Vickers hardness( GPa)of each layer is given k=214W/mK) P Hy=1.8544 (2a)2 4000 Table 1 3000 Dry sand blasting test conditions Sand blasting equipment GS-6 type sand blasting machine tool Sic/(W,Ti)C ceramic nozzle laminated only in Nozzle material entry area( GN-2 SiC/W, Ti)C ceramic nozzle laminated both in entry and exit area(GN-3) SiC/(W,Ti)C stress-free nozzle(CN-2) Dimension of nozzle omm(internal diameter)x 30 mm(length) 50-150um SiC powders 0.4 MPa Cumulative mass weigh Accurate electronic balance(minimum 0.1 mg) Fig. 7. X-ray diffraction analysis of the SiC/(W,Ti)c laminated ceramic nozzle material(GN-2)after sintering at 1900C for 40 min
D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 123 Fig. 6. Photo of the SiC/(W,Ti)C laminated ceramic nozzles. The finite element method (FEM) was used as a means of numerically evaluating the residual thermal stress and its distribution of the laminated ceramic nozzle in the fabricating processes. For observation of the micro-damage and determination of erosion mechanisms, the worn nozzles were sectioned axially. The eroded bore surfaces of the nozzles were examined by scanning electron microscopy. 3. Results and discussion 3.1. Microstructural characterization and properties of SiC/(W,Ti)C laminated nozzle materials Hardness measurements were performed by placing Vickers indentations on every layer of the cross-sectional surface of SiC/(W,Ti)C laminated nozzle (GN-2) material. The indentation load was 200 N and a minimum of three indentations were tested for each layer. The Vickers hardness (GPa) of each layer is given by: Hv = 1.8544 P (2a) 2 (2) Table 1 Dry sand blasting test conditions Sand blasting equipment GS-6 type sand blasting machine tool Nozzle material SiC/(W,Ti)C ceramic nozzle laminated only in entry area (GN-2) SiC/(W,Ti)C ceramic nozzle laminated both in entry and exit area (GN-3) SiC/(W,Ti)C stress-free nozzle (CN-2) Dimension of nozzle Ø 8 mm (internal diameter) × 30 mm (length) Erodent abrasives 50–150m SiC powders Compressed air pressure 0.4 MPa Cumulative mass weigh Accurate electronic balance (minimum 0.1 mg) Table 2 Hardness of different layers of the SiC/(W,Ti)C laminated nozzle (GN-2) materials Layer (W,Ti)C content (vol.%) Vickers hardness, Hv (GPa) 1 55 26.89 2 57 26.52 3 59 25.93 4 61 25.70 5 63 24.67 6 65 24.15 where P is the indentation load (N), 2a is the catercorner length (m) due to indentation. Hardness of each layer of SiC/(W,Ti)C laminated nozzle (GN-2) material is presented in Table 2. Fig. 7 illustrates the X-ray diffraction analysis of the SiC/(W,Ti)C laminated ceramic nozzle (GN-2) material after sintering at 1900 ◦C for 40 min. It can be seen that both (W,Ti)C and SiC existed in the sintered specimens. SEM micrographs of each polished layer of SiC/(W,Ti)C laminated ceramic nozzle (GN-2) material are shown in Fig. 8. The black areas were identified by EDX analysis as SiC, and the white phases with clear contrast were (W,Ti)C. It can be seen that the SiC particles are quite uniformly distributed throughout the microstructure, porosity is virtually absent. 3.2. Residual thermal stress analysis of SiC/(W,Ti)C laminated nozzle material The residual thermal stress of the laminated ceramic nozzle in the fabricating process was calculated by means of the finite element method by assuming that the compact is cooled from sintering temperature 1900 ◦C to room temperature 20 ◦C. Thermo-mechanical properties of (W,Ti)C and SiC are as follows: (W, Ti)C : E = 480 GPa, ν = 0.25, α = 8.5 × 10−6 K−1, k = 21.4 W/(m K). Fig. 7. X-ray diffraction analysis of the SiC/(W,Ti)C laminated ceramic nozzle material (GN-2) after sintering at 1900 ◦C for 40 min.
D Jianxin et al./ Materials Science and Engineering A 444(2007)120-129 05e1115gy氵i:u a) ()L6561169:636n ?虑 Fig 8. SEM micrographs of each polished layer of SiC/(W,Ti)C laminated ceramic nozzle material (GN-2):(a)the first layer(entry zone). (b)the second layer, (c) the third layer, (d)the fourth layer, (e)the fifth second layer, and(f)the sixth layer(exit zone) SiC:E=450GPa,u=0.16.,a=46×10-6K-1 and the maximum value is.003 MPa,-130949 MPa, and k=33.5W/(mK) -265.368 MPa, respectively. Therefore, laminated structures in ceramic nozzles can form an excess compressive residual stresses in the nozzle entry (or exit)region during fabricating Owing to the symmetry, an axisymmetric calculation was process preferred Presume that it was steady state boundary conditions The FEM gridding model of the laminated nozzle is shown in Fig 9. The results of the distribution of axial (oz), radial(or), and circumferential(oe)residual thermal stresses in the GN-2 laminated nozzle in cooling process from sintering temperature to room temperature are showed in Fig. 10. As can be seen, an excess residual thermal stress is formed in the nozzle entry region for the GN-2 laminated nozzle. It is indicated that axial(oz), radial(o), and circumferential(oe) residual thermal stresses at the nozzle entry zone are compressive, and the maximum value is -71.018 MPa,-121578 MPa, and -276 204 MPa, respec- tively Fig. 11 shows the distribution of axial (oz), radial (or). and circumferential(oe)residual thermal stresses in the GN-3 laminated nozzle in cooling process from sintering tempera ture to room temperature. It is obvious that an excess resid ual thermal stress is formed both in nozzle entry and exit region for the GN-3 laminated nozzle, and the axial(oz), radial (or), and circumferential(oe) residual thermal stresses both at the nozzle entry zone and at the exit zone are compressive Fig. 9. FEM gridding model of the laminated nozzle
124 D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 Fig. 8. SEM micrographs of each polished layer of SiC/(W,Ti)C laminated ceramic nozzle material (GN-2): (a) the first layer (entry zone), (b) the second layer, (c) the third layer, (d) the fourth layer, (e) the fifth second layer, and (f) the sixth layer (exit zone). SiC : E = 450 GPa, ν = 0.16, α = 4.6 × 10−6 K−1, k = 33.5 W/(m K). Owing to the symmetry, an axisymmetric calculation was preferred. Presume that it was steady state boundary conditions. The FEM gridding model of the laminated nozzle is shown in Fig. 9. The results of the distribution of axial (σz), radial (σr), and circumferential (σ) residual thermal stresses in the GN-2 laminated nozzle in cooling process from sintering temperature to room temperature are showed in Fig. 10. As can be seen, an excess residual thermal stress is formed in the nozzle entry region for the GN-2 laminated nozzle. It is indicated that axial (σz), radial (σr), and circumferential (σ) residual thermal stresses at the nozzle entry zone are compressive, and the maximum value is −71.018 MPa, −121.578 MPa, and −276.204 MPa, respectively. Fig. 11 shows the distribution of axial (σz), radial (σr), and circumferential (σ) residual thermal stresses in the GN-3 laminated nozzle in cooling process from sintering temperature to room temperature. It is obvious that an excess residual thermal stress is formed both in nozzle entry and exit region for the GN-3 laminated nozzle, and the axial (σz), radial (σr), and circumferential (σ) residual thermal stresses both at the nozzle entry zone and at the exit zone are compressive, and the maximum value is −94.003 MPa, −130.949 MPa, and −265.368 MPa, respectively. Therefore, laminated structures in ceramic nozzles can form an excess compressive residual stresses in the nozzle entry (or exit) region during fabricating process. Fig. 9. FEM gridding model of the laminated nozzle.
D Jianxin et al. Materials Science and Engineering A 444 (2007)120-129 m-m5a-m1723-m (a 3939423.854 50.349 213s9(b) 48093 (c) 94347 51.139 Fig 10. Distribution of (a) axial(oz).(b)radial(o, ) and (c)circumferential(aa) residual thermal stresses of GN-2 laminated nozzle 3.3. Erosion behaviour of the Sic/w,li)C laminated nozzle that of the worn GN-2 and GN-3 laminated nozzles, especially at the nozzle entry region. The exit diameter of the worn GN-3 The erosion behaviour of Sic/w,Ti)c laminated ceramic laminated nozzle is greatly reduced compared with that of the nozzle(GN-2 and GN-3) in dry sand blasting processes was GN-2 laminated nozzle. investigated in comparison with the stress-free ceramic nozzle The results of the nozzle entry bore diameter variation with (CN-2). Fig. 12 shows the cumulative mass loss of GN-2, GN- the erosion time for GN-2, GN-3, and CN-2 nozzles are shown 3, and CN-2 nozzles in dry sand blasting processes. It can be in Fig. 15. It is indicated that the entry bore diameter enlarges seen that the cumulative mass loss continuously increased with greatly with the operation time for CN-2 stress-free nozzle the operation time. Compared with GN-2 and GN-3 laminated While the entry bore diameter increases slowly with the opera nozzle, the cn-2 stress-free nozzle showed higher cumulative tion time for gn-2 and gN-3 laminated nozzles. Fig. 16 shows mass loss under the same test conditions the comparison of the erosion rates for GN-2, GN-3, and CN-2 The entry bore profiles of worn GN-2, GN-3, and CN-2 noz- nozzles in sand blasting processes. It is obvious that the erosion zles after dry sand blasting for 540 min are shown in Fig. 13. rate of the stress-free nozzles is higher than that of the laminated It is showed that the entry bore of CN-2 stress-free nozzle was nozzles. Therefore, It is apparently that the GN-2 and GN-3 lam everely worn. While the entry bore of GN-2 and GN-3 lami- inated nozzles exhibited higher erosion wear resistance over the nated nozzle had worn slightly compared with that of the former CN-2 stress-free nozzle under the same test conditions Fig. 17 shows the SEM micrographs of the entry bore surface The worn ceramic nozzles were cut after operation in longitu- of the worn ceramic nozzles. From these SEM micrographs, dinal directions for failure analysis. Fig. 14 shows the photos of different morphologies and fracture modes of the nozzles can the inner bore profile of the whole ceramic nozzle after 540 min be seen clearly. The CN-2 stress-free nozzle at the entry area operation. It is showed that inner bore diameter of the worn CN- failed in a highly brittle manner, and exhibited a brittle fracture 2 nozzle along the nozzle longitudinal directions is larger than induced removal process. There are a lot of obvious pits located
D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 125 Fig. 10. Distribution of (a) axial (σz), (b) radial (σr), and (c) circumferential (σ) residual thermal stresses of GN-2 laminated nozzle. 3.3. Erosion behaviour of the SiC/(W,Ti)C laminated nozzle The erosion behaviour of SiC/(W,Ti)C laminated ceramic nozzle (GN-2 and GN-3) in dry sand blasting processes was investigated in comparison with the stress-free ceramic nozzle (CN-2). Fig. 12 shows the cumulative mass loss of GN-2, GN- 3, and CN-2 nozzles in dry sand blasting processes. It can be seen that the cumulative mass loss continuously increased with the operation time. Compared with GN-2 and GN-3 laminated nozzle, the CN-2 stress-free nozzle showed higher cumulative mass loss under the same test conditions. The entry bore profiles of worn GN-2, GN-3, and CN-2 nozzles after dry sand blasting for 540 min are shown in Fig. 13. It is showed that the entry bore of CN-2 stress-free nozzle was severely worn. While the entry bore of GN-2 and GN-3 laminated nozzle had worn slightly compared with that of the former one. The worn ceramic nozzles were cut after operation in longitudinal directions for failure analysis. Fig. 14 shows the photos of the inner bore profile of the whole ceramic nozzle after 540 min operation. It is showed that inner bore diameter of the worn CN- 2 nozzle along the nozzle longitudinal directions is larger than that of the worn GN-2 and GN-3 laminated nozzles, especially at the nozzle entry region. The exit diameter of the worn GN-3 laminated nozzle is greatly reduced compared with that of the GN-2 laminated nozzle. The results of the nozzle entry bore diameter variation with the erosion time for GN-2, GN-3, and CN-2 nozzles are shown in Fig. 15. It is indicated that the entry bore diameter enlarges greatly with the operation time for CN-2 stress-free nozzle. While the entry bore diameter increases slowly with the operation time for GN-2 and GN-3 laminated nozzles. Fig. 16 shows the comparison of the erosion rates for GN-2, GN-3, and CN-2 nozzles in sand blasting processes. It is obvious that the erosion rate of the stress-free nozzles is higher than that of the laminated nozzles. Therefore, It is apparently that the GN-2 and GN-3 laminated nozzles exhibited higher erosion wear resistance over the CN-2 stress-free nozzle under the same test conditions. Fig. 17 shows the SEM micrographs of the entry bore surface of the worn ceramic nozzles. From these SEM micrographs, different morphologies and fracture modes of the nozzles can be seen clearly. The CN-2 stress-free nozzle at the entry area failed in a highly brittle manner, and exhibited a brittle fracture induced removal process. There are a lot of obvious pits located
D Jianxin et al. Materials Science and Engineering A 444(2007)120-129 p (b)/-.949 4.717 59.146 Fig. 11. Distribution of (a) axial (oz), (b)radial(or), and(c)circumferential (ae)residual thermal stresses of GN-3 laminated nozzle on the nozzle bore surface indicating that brittle fracture took temperature to room temperature, which may partially counter- place. Characteristic SEM pictures taken on the eroded entry act the tensile stresses in the nozzle entry section resulting from bore surface of the gn-2 and gn-3 laminated ceramic nozzle external loadings effect may lead to the increase in resis- are shown in Fig. 18. It is shown that the appearance of the tance to fracture, and thus increase the erosion wear resistance eroded areas of the gn-2 and gn-3 laminated nozzle showed a of the laminated nozzle relative smooth surface by contrast with that of the CN-2 stress- free nozzle Ceramic nozzle failure by erosion wear is generally caused by fracture owing large the tensile stress at the nozzle entry zone -o CN-2 stress-free nozzle [11-15]. Because the nozzle entrance region suffers form severe 日GN2 laminated nozzle SN-3 laminated nozzle abrasive impact, and generates large tensile stress, which may ause the subsurface lateral cracks and facilitates removal of the material chips. Thus, the erosion wear of the nozzle depends on the stress distribution in the entry region. Once the maximum tensile stress exceeds the ultimate strength of the nozzle material. will occur The higher erosion wear resistance of the GN-2 and GN-3 laminated nozzle compared with the CN-2 stress-free nozzle can be analysed in terms of the formation of compressive residual stresses on the entry region. As calculated above, compres Erosion time(min) ive residual stresses were formed in the entry region of the 12. Cumulative mass loss of GN-2. GN-3 laminated nozzle and CN-2 Sic/(W,Ti)C laminated nozzle in cooling process from sintering -free nozzle in dry sand blasting processes
126 D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 Fig. 11. Distribution of (a) axial (σz), (b) radial (σr), and (c) circumferential (σ) residual thermal stresses of GN-3 laminated nozzle. on the nozzle bore surface indicating that brittle fracture took place. Characteristic SEM pictures taken on the eroded entry bore surface of the GN-2 and GN-3 laminated ceramic nozzle are shown in Fig. 18. It is shown that the appearance of the eroded areas of the GN-2 and GN-3 laminated nozzle showed a relative smooth surface by contrast with that of the CN-2 stressfree nozzle. Ceramic nozzle failure by erosion wear is generally caused by fracture owing large the tensile stress at the nozzle entry zone [11–15]. Because the nozzle entrance region suffers form severe abrasive impact, and generates large tensile stress, which may cause the subsurface lateral cracks and facilitates removal of the material chips. Thus, the erosion wear of the nozzle depends on the stress distribution in the entry region. Once the maximum tensile stress exceeds the ultimate strength of the nozzle material, fracture will occur. The higher erosion wear resistance of the GN-2 and GN-3 laminated nozzle compared with the CN-2 stress-free nozzle can be analysed in terms of the formation of compressive residual stresses on the entry region. As calculated above, compressive residual stresses were formed in the entry region of the SiC/(W,Ti)C laminated nozzle in cooling process from sintering temperature to room temperature, which may partially counteract the tensile stresses in the nozzle entry section resulting from external loadings. This effect may lead to the increase in resistance to fracture, and thus increase the erosion wear resistance of the laminated nozzle. Fig. 12. Cumulative mass loss of GN-2, GN-3 laminated nozzle, and CN-2 stress-free nozzle in dry sand blasting processes.
D Jianxin et al. Materials Science and Engineering A 444(2007)120-129 rnm (a Fig 13. Entry bore profile of the nozzles: (a)entry bore profile of the nozzle before operation. (b)entry bore profile of GN-2 laminated nozzle after 540 min operatic (c)entry bore profile of GN-3 laminated nozzle after 540 min operation, and(d) entry bore profile of CN-2 stress-free nozzle after 540 min operation. Entry 。[ Fig. 14. Photos of the won inner bore profile of the whole ceramic nozzle in longitudinal direction after 540 min operation:(a)GN-2 laminated nozzle, (b)GN-3 laminated nozzle, and(c)CN-2 stress-free nozzle. 一cN2 stressfree noz -GN-2 laminate 6=c OcN-2 stress-free nozzle GN-2 laminated nozzle GN-3 laminated nozzle 0120240360480 Erosion time(min) Fig. 15. Nozzle entry bore diameter variation with the erosion time for GN-2, Fig. 16. Comparison of the erosion rates of different ceramic nozzles in sand GN-3 laminated nozzle and CN-2 stress-free nozzle in sand blasting processes. blasting processes
D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 127 Fig. 13. Entry bore profile of the nozzles: (a) entry bore profile of the nozzle before operation, (b) entry bore profile of GN-2 laminated nozzle after 540 min operation, (c) entry bore profile of GN-3 laminated nozzle after 540 min operation, and (d) entry bore profile of CN-2 stress-free nozzle after 540 min operation. Fig. 14. Photos of the worn inner bore profile of the whole ceramic nozzle in longitudinal direction after 540 min operation: (a) GN-2 laminated nozzle, (b) GN-3 laminated nozzle, and (c) CN-2 stress-free nozzle. Fig. 15. Nozzle entry bore diameter variation with the erosion time for GN-2, GN-3 laminated nozzle and CN-2 stress-free nozzle in sand blasting processes. Fig. 16. Comparison of the erosion rates of different ceramic nozzles in sand blasting processes
D Jianxin et al. Materials Science and Engineering A 444(2007)120-129 5e1115KvX1.5②K三9.0un Fig. 17. SEM micrographs of the entry bore surface of the worn CN-2 stress-free ceramic nozzle. (a) 05081115Kv 5ee1115K Fig. 18. SEM micrographs of the entry bore surface of the wom laminated ceramic nozzle: (a and b) GN-2 laminated nozzle, (c and d) GN-3 laminated nozzle. 4. Conclusions in nozzle entry region in fabricating process of the laminated ceramic nozzles, which may partially counteract the tensile Sic/W, Ti)C laminated ceramic nozzles were produced by stresses resulting from external loadings. Laminated struc hot pressing. The purpose is to reduce the tensile stress at the ture in ceramic nozzles is an effective way to improve the entrance area of the nozzle during dry sand blasting processes erosion wear resistance of the stress-free ceramic nozzles Particular attention was paid to the erosion wear behaviours of 2. The ceramic nozzle laminated both in entry and exit area this kind laminated ceramic nozzle. Results showed that (GN-3)exhibited higher erosion wear resistance over the one laminated only in entry area(GN-2) 1. The laminated ceramic nozzles(GN-2 and GN-3)have supe- Acknowledgements rior erosion wear resistance to that of the homologous stress ee ceramic nozzle(CN-2). The mechanism responsible was This work was supported by the"National Natural explained as the formation of compressive residual stresses F oundation of China (50475133)", "Natural Science Founda-
128 D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 Fig. 17. SEM micrographs of the entry bore surface of the worn CN-2 stress-free ceramic nozzle. Fig. 18. SEM micrographs of the entry bore surface of the worn laminated ceramic nozzle: (a and b) GN-2 laminated nozzle, (c and d) GN-3 laminated nozzle. 4. Conclusions SiC/(W,Ti)C laminated ceramic nozzles were produced by hot pressing. The purpose is to reduce the tensile stress at the entrance area of the nozzle during dry sand blasting processes. Particular attention was paid to the erosion wear behaviours of this kind laminated ceramic nozzle. Results showed that: 1. The laminated ceramic nozzles (GN-2 and GN-3) have superior erosion wear resistance to that of the homologous stressfree ceramic nozzle (CN-2). The mechanism responsible was explained as the formation of compressive residual stresses in nozzle entry region in fabricating process of the laminated ceramic nozzles, which may partially counteract the tensile stresses resulting from external loadings. Laminated structure in ceramic nozzles is an effective way to improve the erosion wear resistance of the stress-free ceramic nozzles. 2. The ceramic nozzle laminated both in entry and exit area (GN-3) exhibited higher erosion wear resistance over the one laminated only in entry area (GN-2). Acknowledgements This work was supported by the “National Natural Science Foundation of China (50475133)”, “Natural Science Founda-
D Jianxin et al. Materials Science and Engineering A 444(2007)120-129 tion of Shandong Province(Y2004F08)", "the Program for New [13] J. Deng, X. Zhang, P. Niu, L Liu, J. Wang, Tribol. Int. 39(3)(2006) Century Excellent Talents in University (NCET-04-0622)",an 74-280. The Taishan Scholar Program of Shandong Province [14 J. Deng, Adv Appl. Ceram. 104(2005)59-64. [15] J. Deng Z. Zheng Z Ding, J. Wang, Br. Ceram. Trans. 102(2003)61-65 References [16] R.J.K. Wood, D w. Wheeler, D.C. Lejeau, Wear 233-235(1999)134-150 [17] J. Deng, Z Ding, D. Yuan, Mater. Sci Eng, A 417(1-2)(2006)1-7 [18] Z. Ding, J. Deng, J. Li, G. Sun, X. Ai, Ceram. Int. 34 (4)(2004)591-596 [1] G. Li, Surface Engineering, Mechanical Industry Publishing House, Bei- [19]R Lakshminarayanan, D.K. Shetty, R.A. Cutler, J Am Ceram Soc. 79(1) jing, 1998 (1996)7987 2]J.Deng,T.Lee, Surf.Eng.16(5)(2000411-414. [20] P.Z. Cai, D.J. Green, G.L. Messing, J. Eur. Ceram. Soc. 5(1998) [3]A. Raykowski, M. Hader, Wear 249(2001)127-132. 2025-2034. [4] B Djurovic, E. Jean, Wear 224(1999)22-37 [21] Tarlazzi, E. Roncar, P Pinasco, S. Guicciardi, C. Melandri, G. de Portu, 5] YI. Oka,H. Ohnogi,Wear203-204(1997)573-579 wear244(200029-40. [6] 1. Finnie, G.R. Stevick, J.R. Ridgely, Wear 152(1992)91-9 [22] F. Toschi, C. Melandri, P Pinasco, E. Roncar, S Guicciardi, G. De Portu, [7R.G. Wellman, C. Allen, Wear 186-187(1995)117-123. J.Am. Ceran.Soc.86(9)(2003)1547-1553 []S Srinivasan, R.O. Scattergood, Wear 128(1988)139-215 [23] C. Hillman, Z Suo, F.F. Lange, J. Am. Ceram Soc. 79(1996)2127 [9] P.H. Shipway, I.M. Hutchings, Wear 149(1991)85-98. [24] D B. Marshall, J.J. Ratto, FF. Lange, J. Am. Ceram Soc. 74(1991)2979 [10S. Bahadur, R. Badruddin, Proceedings of the International Conference [25] V. Sergo, D.M. Lipkin, G De Portu, D.R. Clarke, J. Am. Ceram. Soc. 80 Wear of Materials, ASME. New York, 1989, pp. 143-153 7)(1997)1633-1638 [Il J. Deng, Mater. Sci Eng, A 408(1-2)(2005)227-233 26]G. de Portu, L Micele, Y Sekiguchi, G. Pezzotti, Acta Mater. 53(2005 [12] J. Deng, Y. Feng, Z. Ding, J. Eur. Ceram Soc. 23(2003)323-329 1511-1520
D. Jianxin et al. / Materials Science and Engineering A 444 (2007) 120–129 129 tion of Shandong Province (Y2004F08)”, “the Program for New Century Excellent Talents in University (NCET-04-0622)”, and “The Taishan Scholar Program of Shandong Province”. References [1] G. Li, Surface Engineering, Mechanical Industry Publishing House, Beijing, 1998. [2] J. Deng, T. Lee, Surf. Eng. 16 (5) (2000) 411–414. [3] A. Raykowski, M. Hader, Wear 249 (2001) 127–132. [4] B. Djurovic, E. Jean, Wear 224 (1999) 22–37. [5] Y.I. Oka, H. Ohnogi, Wear 203–204 (1997) 573–579. [6] I. Finnie, G.R. Stevick, J.R. Ridgely, Wear 152 (1992) 91–97. [7] R.G. Wellman, C. Allen, Wear 186–187 (1995) 117–123. [8] S. Srinivasan, R.O. Scatterrgood, Wear 128 (1988) 139–215. [9] P.H. Shipway, I.M. Hutchings, Wear 149 (1991) 85–98. [10] S. Bahadur, R. Badruddin, Proceedings of the International Conference on Wear of Materials, ASME, New York, 1989, pp. 143–153. [11] J. Deng, Mater. Sci. Eng., A 408 (1–2) (2005) 227–233. [12] J. Deng, Y. Feng, Z. Ding, J. Eur. Ceram. Soc. 23 (2003) 323–329. [13] J. Deng, X. Zhang, P. Niu, L. Liu, J. Wang, Tribol. Int. 39 (3) (2006) 274–280. [14] J. Deng, Adv. Appl. Ceram. 104 (2005) 59–64. [15] J. Deng, Z. Zheng, Z. Ding, J. Wang, Br. Ceram. Trans. 102 (2003) 61–65. [16] R.J.K. Wood, D.W. Wheeler, D.C. Lejeau, Wear 233–235 (1999) 134–150. [17] J. Deng, Z. Ding, D. Yuan, Mater. Sci. Eng., A 417 (1–2) (2006) 1–7. [18] Z. Ding, J. Deng, J. Li, G. Sun, X. Ai, Ceram. Int. 34 (4) (2004) 591–596. [19] R. Lakshminarayanan, D.K. Shetty, R.A. Cutler, J. Am. Ceram. Soc. 79 (1) (1996) 79–87. [20] P.Z. Cai, D.J. Green, G.L. Messing, J. Eur. Ceram. Soc. 5 (1998) 2025–2034. [21] Tarlazzi, E. Roncari, P. Pinasco, S. Guicciardi, C. Melandri, G. de Portu, Wear 244 (2000) 29–40. [22] F. Toschi, C. Melandri, P. Pinasco, E. Roncari, S. Guicciardi, G. De Portu, J. Am. Ceram. Soc. 86 (9) (2003) 1547–1553. [23] C. Hillman, Z. Suo, F.F. Lange, J. Am. Ceram. Soc. 79 (1996) 2127. [24] D.B. Marshall, J.J. Ratto, F.F. Lange, J. Am. Ceram. Soc. 74 (1991) 2979. [25] V. Sergo, D.M. Lipkin, G. De Portu, D.R. Clarke, J. Am. Ceram. Soc. 80 (7) (1997) 1633–1638. [26] G. de Portu, L. Micele, Y. Sekiguchi, G. Pezzotti, Acta Mater. 53 (2005) 1511–1520