(RSSN) and RSSN with SiC particles have been infiltrated with CaF, and CaSiO 3 melts at a temperature of. 1450 and 1600~ respectively. The wear resistance ...
Refractories and Industrial Ceramics
VoL 41, Nos. 3
4. 2000
UDC 666.762.93
W E A R R E S I S T A N C E OF C O M P O S I T E M A T E R I A L S B A S E D ON S I L I C O N N I T R I D E AND C A R B I D E W I T H C E R A M I C MELT I N F I L T R A T I O N S V. G. Gilev t Translated from Ogneupory i Tckhnicheskaya Keramika, No. 3, pp. 7 - 10, March, 2000.
Fabrication of composite materials is considered. Porous pretbrms from reaction-sintered silicon nitride (RSSN) and RSSN with SiC particles have been infiltrated with CaF, and CaSiO 3melts at a temperature of 1450 and 1600~ respectively. The wear resistance of the composite (with infiltrations) materials turned out to be mt, ch higher than that of original (without infiltrations) materials of the RSSN and RSSN 50% SiC type. A com'posite material with the composition RSSN 50% SiC CaSiO~ is shown to possess the highest hardness and abrasive resistance.
Materials based on silicon nitridc can be used successfully in highly loaded structures, have a high resistance to thermal shocks because of their low temperature coefficient of linear expansion, possess a high hardness and wear resistance, and are used for the production o f cutting tools, parts of friction units, and internal combustion engines [ I - 4]. Friction and wear of silicbn nitride materials has been studied in pairs with steels, alloys, and ceramic materials at a temperature _ 10 me/g (2). Materials based on RSSN were reinforced with a SiC powder synthesized from a mixture of Si and 30% graphite. We used silicon Krl milled in a 16ch ball mill and graphite of grade S-I. Belbre the synthesis the mixture was milled in an AGO-3 planetary mill fbr 5 rain in a dry state. In order to synthesize SiC the mixture was heat treated for 1 h in nitrogen at 1450~ The synthesized powder had a specific surface o f 3.4 me/g. In accordance with x-ray phase analysis the predominant phase in the composition of the powder is SiC: no admixtures of silicon and graphite were detected. The x-ray pattern exhibits a weak line of an admixture o f milled iron. After pressing with a binder of 5% aqueous solution o f polyvinyl alcohol in a steel press-mold at a pressure o f 2 0 0 - 4 0 0 MPa we obtained pressings with a porosity o f 35 - 40% (Table I ). Block specimens tbr the friction test had a cross section of 10 x 16 mm and a thickness of 5 - 10 nun. The growth in the mass o f the specimens after sintering in nitrogen for 5 h at 1350~ and tbr 2 h at 1450~ was at best
I Research Center of Powder Materials, Perm, Russia. 80
1083-.1877/00/0304-01)80525.00 ~;' 21100Kluwer Academic/Plenum Publishers
Wear Resistance of Composite Materials B a s e d on Silicon N i t r i d e and Carbide
TABLE 1, Properties of RSSN Specimens" Silicon
Pone, %
2 2 1 1 1
35.6 40.8 40.2 35.3 38.1
powdcr
-
~
%
52.0 55.6 50.1 48.8 47.2
Siret , ~176
22.0 16.6 24.9 26.8 29.2
81
TABLE 2. PropertiesofRSSN - CaF~ and RSSN Composites" PRSSN ,
CaSiO 3
o/o
g/cm3
PRSgX9 "
2.28 2.18 2.09 2.24 2.12
24.0 27.7 28.6 23.0 26.5
Silicon powder
Pc, g/cm3
Pth, g/cm3
Pc, %
hnpregnation with CaF2.[br I0 rain at 1450~
2 2 1
2.950 2.923 2.765
3.00 3.06 2.93
1.7 4.5 5.6
hnpregnation with CaSiO3fin" 10 rain at 1600~
/ 1
Porig is the porosity of the pressings, AP is the increment of the mass as a result ofsintering in nitrogen for 5 h at 1350~ and 2 h at 1450~ Siret is the calculated amotmt of retained silicon, PRSSN is the density of sintered specimens, PRSSN is the porosity of sintered specimens.
5 2 - 56% (the theoretical growth in the mass is 66.7%); in high-porosity materials fabricated from the same powders the mass increased by 60 - 62% [ I 1]. The calculated values o f the amount of retained silicon (in accordance with the growth in the mass) was > 17%. The porosity o f the RSSN specimens was 25 - 30% and the density was 2.2 - 2.4 g/era 3, which are quite typical values and provide an ultimate compressive strength of 400 - 500 MPa [1 ,.12]. An attempt to impregnate RSSN specimens with molten calcium metasilicate CaSiO 3 (Tmelt = [ 5 4 0 ~ has shown that their sizes change (swelling), which diminishes the density o f the material (Table 2). The results o f an x-ray phase analysis show that the impregnation causes a decrease in the intensity o f the lines of silicon nitride and an increase in the intensity o f the lines of silicon carbide present in the original RSSN in the form of an admixture, which indicates that the nitride skeleton interacts with the oxide melt and does not interact with silicon carbide. In order to improve the resistance o f the porous skeleton to the ceramic melt, preserve the sizes o f the parts, and increase the wear resistance, the initial mixture based on silicon nitride milled in a ball mill was then enriched with 50 wt.% silicon carbide. After pressing the mixture of Si ~50% SiC and nitriding (5 h at 1350~ and 2 h at 1450~ the mass and the density of the sintered specimens increased substantially (Table 3). The additive of the SiC powder prorooted the nitriding process. The obtained specimens o f R S S N - S i C were impregnated with molten CaSiO 3 at 1600~ A mixture o f CaO and Si(), powders in the requisite proportion was used to press block preforms as calculated belbrehand. They were placed onto sintered pretbrms of R S S N - SiC. One extra specimen was pressed from SiC powder and impregnated like the specimens o f R S S N - SiC. In the course o f the impregnation a part o f the infiltrate ran down the porous prefonns of R S S N - S i C and theretbre the density and growth in the mass o f the infiltrated specimens were lower than the calcu-
2.610 2.737
2.91 2.91
10 6
Pc is the density of the composite. Pth is the theoretical density, and Pc is the porosity of the composite.
TABLE 3. Results of Sintcring of Specimens ofSi + 50% SiC Mixtures and Their Impregnation with CaSiO 3 Melt" Number of specimen P' (see Ta- MPa
Pong" %
blc 6) 5 6
300 400
36.1 31.8
~t'l, Propertiesof composite fraction AP, PRSSN, of the % % theore- 9c, Pc, HRA ticaI g/cm3 % value 62.7 60.0
29.2 25.3
0.77 0.56
2.92 2.80
6 10
88 87
P is the pressing pressure, AP I is the increase in the mass upon impregnation.
lated ones (see Table 3). The specimen of pure silicon carbide cracked during the infiltration. The results o f friction tests of RSSN specimens sliding over steel are presented in Table 4. A part o f the specimens was impregnated by oil. Oil impregnation decreased somewhat the coefficient of friction of RSSN over steel. The friction coefficientj~ r o f the RSSN specimens was much higher than the one presented in [3, p. 212]. The elevated value of.If.~ can be caused by incomplete nitriding and the presence o f retained silicon. The elevation oi~fl-~in the presence o f silicon is known for materials fabricated from siliconized graphite [10, p. 192]. In accordance with [13] the porosity o f the material affects considerably the value of/i.~, and the friction coefficient increases with the porosity. The results o f the tests of R S S N - C a F ~ and RSSN CaSiO 3 specimens rubbed against steel are presented in Table 5. The introduction of a hard CaF z lubrication virtually does not decrease./i, r but increases considerably the wear resistance, which is a consequence of the densification. The wear resistance o f RSSN - CaSiO 3 composites rubbed against steel is not high because of their porosity, but is better than in the unimpregnated RSSN. The pair of RSSN - CaSiO 3 composite and steel is characterized by high friction coefficients just like tile pair o f conventional RSSN and steel.
82
V.G. Gilev
TABLE 4. Results of Friction and Wear Tests of RSSN against Steel Without Lu0rication Density of RSSN, g/cm3
Load N. State of specimen N
2.37
Sintered
2.37
Impregnated with oil
2.20
2.20
Sintered
Impregnated with oil
Sintered
Wear inten- Coefficient sity I h. o f friction)i? lam/kr n
23.75 59.75 23.75
16.1 1055 16.9
1.29 I. 19 I. 17
59.75 159.7 23.75 59.75 159.7 23.75 59.75 159.7 10"
22.5 59.1 13.3 173 287 42.5 53.0 356 --
1.16 0.56 1.29 1.26 0.63 1.05 0.93 0.70 0.75 +0.1"
Data of [3].
TABLE 5. Tests ofRSSN - CaF, and RSSN Sliding over Steel without Lubrication Composition of the composite
p,
p ,,
g/em 3
N, N
100 200 300 23.7 59.7 100 150 200
RSSN
CaF 2
2.92
RSSN
CaSiO 3
2.60
CaSiO 3 Composites It,,
MPa
1.5 3.0 4.5 0.35 0.89 1.49 1.68 2.98
/am/km
li'r
4.2 I0 45.8 1.6 3.0 5.3 5.1 2.7
1.53 0.93 0.78 1.17 1.07 0.83 0.81 0.96 0.58 0.72
Specific load. Coarse particles chipped off from the front edge.
Under a load N = 150 - 200 N the value ofJ~-r in the pair o f R S S N - C a S i O 3 and steel fluctuates periodically. The scattering with respect to the mean value varies from 17% at N = 150 N to 21% at N = 200 N. The s e l f fluctuations are connected with the periodic formation and disruption o f fihn structures in the process o f friction. The decrease in.[i.~ with growth in the load typical tbr all the studied silicon nitride materials (see Tables 4 and 5) is a sign o f the presence o f plastic contact in the friction zone. Specimens o f the type o f RSSN - SiC infiltrated with CaSiO 3 have shown a very high resistance to abrasive wear (Table 6): the wear Am per 1 min is less than 0.1 mg. The HtL4 o f specimen No. 6 is 87 (see Table 3), which corresponds to an H V o f 1400. The HRA o f specimen No. 5 is also high (8~), but in this case the indentation results in spalling
TABLE 6. Results of Tests for Abrasive Wear upon Sliding over a Monolith o f a 25A25PSM26K5B Abrasive Material Number ofspecimen
1 2 3 4 5 6
Material
RSSN RSSN 50% SiC RSSN -- CaF 2 RSSN- CaSiO 3 RSSN 50% SiC CaSiO 3 RSSN -- 50% SiC CaSiO 3
P,
Am,
Ih,
0.24 0.34 0.57 0.66
9 5.1 0.32 1.2
6.85 3.00 0.45 2.13
6
0.49
0.05
0.06
10.3
0.41
0.05
0.05
P, g/cm 3
Ptot, %
Ml'a
2.19 2.40 2.92 2.61
31.7 25.0 4.5 10
2.92 2.80
rag/rain p.m/km
spots absent in specimen No. 6. The advantage o f the latter seems to consist in the lower porosity o f the sintered specimen, i.e., 25.3% instead o f 29.2% in specimen No. 5. The introduction o f SiC is necessary, because it provides a high-temperature chemical resistance o f the skeleton with respect to the melt o f the infiltrate at the impregnation temperature. In the absence o f SiC in the RSSN thc impregnation causes a change in the sizes (swelling) o f the preforms and a decrease in the density and wear resistance. In addition, SiC can be treated as an additive that increases the abrasive resistance. The introduction o f Si3N 4 into the composition provides a higher strength o f the pretbnns as compared to materials from siliconized graphite [ I 0] and makes it possible to fabricate parts with a more complex configuration. The main advantage provided by the introduction o f Si3N 4 into the c o m position o f the material consists in the fact that Si3N 4 interacts with the refractory oxides that infiltrate the skeleton at 1500 - 1600~ As a result o f the interaction, it transtbrms into an amorphous state or an oxide glass phase and provides hardening o f the oxide phase and a strong coherence o f the components o f the composite material. The resistance to abrasive wear enhances. In addition, when the pores o f the RSSN and R S S N - SiC materials are filled with refractory oxides, the resistance to high-temperature oxidation and aggressive media increases too. A typical value o f f ; . tbr a ceramic composite rubbed against an abrasive is 0.7, which is much lower than in friction against steel and is connected with the absence o f plastic con,i~'t.
It can be seen from the data o f Table 6 that infiltrated composites and especially those with a stronger binder o f calcium metasilicate have considerable advantages. The high hardness and wear resistance o f the obtained material can provide operation o f pats and units under severe conditions represented by abrasion, high temperature, heat cycles, and aggressive media. The material can be used for gaskets, journal bearings, and stop val.ves. The high coefficientf.~ o f composite materials serving in a pair with steel can provide efficient braking systems and
Wear Resistance of Composite Materials Based on Silicon Nitride and Carbide
friction transmissions where the worn element is a less expensive steel and the ceramic friction element has a longer life. CONCLUSIONS
4. Y. Ogawa, M. Mashida, N. Miyamura, K. Tashiro, and M. Sugano, "Ceramic rocker arm insert for intemal combustion engines,"in:MetalPowderRep.,Februao'(1987),pp. 113 119. 5. Z. P. Wang and C. Ruiz, "Damage process of ceramics in contact with metals," Matel: Sci. Eng., A127, 105 114 (1990). 6. Z. P. Wang and C. Ruiz, "Characterization of contact damage of
We have shown that it is possible to increase substantially (by 2 - 3 orders o f magnitude) the abrasive resistance o f reaction-sintered silicon nitride by impregnating it with ceramic melts. The friction coefficient o f a RSSN-steel pair is 1 - !.5 and decreases with growth in the load, which indicates the presence o f plastic contact. Impregnation o f RSSN with oil reduces somewhat the coefficient./i~ o f friction against steel without lubrication and the wear of a material with a density o f 2.37 g / c m 3. The introduction o f a hard CaF~ lubrication does not have a substantial effect on the coefficient o f friction o f R S S N against steel.
10.
REFERENCES
I I.
I. R. A. Andrievskii and I. I. Spivak. Silicon Nitride andMatelqals Based on It [in Russian], Metallurgiya, Moscow (I 984). 2. T. Ya. Kosolapova, T. V. Andreeva, T. S. Bartnitskaya, et al., Nomnetallic Reji'uctoty Compounds [in Russian], Metallurgiya. Moscow (1976). 3. G. G. Gnesin, I. I. Osipova, G. D. Rontal' et al., (ed. by G. G. Gcnsin), Ceramic lbol Materials [in Russian], Tekhnika, Kicv (1991).
83
7.
8. 9.
12.
13.
Si3N4 in contact with Waspaloy," Bt: Ceram. Trans. J., 89. 12 16(1990). 14. Tomizawa and T. E. Fisher, "Friction and wear of silicon nitridc at 150~ to 800~ '' Asle Transact., 29(4), 481 488 (1986). A. S. Tarabanov and V. I. Kostikov, Siliconized Graphite [in Russian], Metallurgiya, Moscow (1977). D. Seyferth, G. N. Wiseman, and Chr. Prud'homme, "Silicon-nitrogen polymers and ceramics derived from reactions of dichlorosilasane H2SiC12," in: Emergent Process Meth. High-Tech. Ceram., Proc. Conj.. Raleigh N.C. 8 IONo~: 1982, New- }brk-- London (1984), pp. 263 - 269. W. M. Hillig, Ceramic matrix composites by means of melt infiltration.'* Ceram. Eng. Sci. Proc., 8(7 8), 834 - 838 (1987). V. N. Antsiferov and V. G. Gilev, "'Special features of fabrication of porous silicon nitride materials from thin-wall elements," Poroshk. Metall., No. 8, 44 48 (1991). V. G. Gilev, Development o f Materials Based on ReactionSintered Silicon Nitride. Author ~s"Abstract o f Candidate's Thesis [in Russian], Perm (1990). A. D. Panasyuk, L. E. Struk, A. I. Yuga, et al., )'Tribological properties of ceramic materials based on nonmetallic nitridcs. I. Friction characteristics of composite materials based on silicon nitride," Poroshk. Metall., No. 8, 76 81 (1990).
