- Email: [email protected]
h-BN composites
Accepted Manuscript Effects of h-BN on mechanical properties of reaction bonded β-SiAlON/h-BN composites Yanjun Li, Bangzhi Ge, Zhihong Wu, Guoqing Xiao, Zhongqi Shi, Zhihao Jin PII:
S0925-8388(17)30388-2
DOI:
10.1016/j.jallcom.2017.01.318
Reference:
JALCOM 40685
To appear in:
Journal of Alloys and Compounds
Received Date: 7 September 2016 Revised Date:
26 January 2017
Accepted Date: 29 January 2017
Please cite this article as: Y. Li, B. Ge, Z. Wu, G. Xiao, Z. Shi, Z. Jin, Effects of h-BN on mechanical properties of reaction bonded β-SiAlON/h-BN composites, Journal of Alloys and Compounds (2017), doi: 10.1016/j.jallcom.2017.01.318. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Effects of h-BN on mechanical properties of reaction bonded β-SiAlON/h-BN composites Yanjun Li1,*, Bangzhi Ge2, Zhihong Wu1, Guoqing Xiao1, Zhongqi Shi2,*, Zhihao Jin2 College of Materials and Mineral Resources, Xi’an University of Architecture and Technology, Xi’an
RI PT
1
710055, China
State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an 710049,
SC
2
China
M AN U
*Corresponding authors: Yanjun Li, E-mail: [email protected], Tel: +86 29 82205798, Fax: +86 29 82205798; Zhongqi Shi, E-mail: [email protected], Tel: +86 29 82667942, Fax: +86 29 82663453. Abstract: β-SiAlON/h-BN composites with improved machinability and thermal shock resistance were fabricated via a reaction bonding procedure using raw materials of Si, Al2O3, AlN and h-BN. Effects of h-BN on
TE D
properties of the composites, including density, strength, hardness, fracture toughness, machinability and thermal shock resistance, were investigated. XRD analysis was performed to analyze the phase composition of the
EP
composites, and SEM was employed to analyze their microstructure. The results showed that the density of the composites went through a maximum value with h-BN content increased from 0 to 50 wt.%. Vickers hardness of
AC C
the composites with less h-BN addition was decreased without obvious strength loss. Fracture toughness of the composites was decreased with h-BN content increased from 0 to 50 wt.%. Drilling hole and thread tapping test showed that the machinability of the composites with h-BN addition was commendable. Thermal shock resistance of the composites with h-BN content higher than 20 wt.% was excellent, and the retained strength of the composites was decreased slightly with the temperature difference and thermal shock times increased. Key words: SiAlON; Machinable; Thermal shock resistance
ACCEPTED MANUSCRIPT 1 Introduction
β-SiAlON solid solutions inherit the crystal structure of β-Si3N4 with equivalent substitution of Al-O for Si-N, and have chemical formula of Si6-ZAlZOZN8-Z (Z=0~4.2), where Z is the number of (Si-N) bonds being replaced by
RI PT
(Al-O) bonds [1-3]. β-SiAlON is a promising material in nonferrous metal industry due to its poor wettability to nonferrous metals melts, high strength at room temperature and high temperature, high toughness, low thermal expansion, and well oxidation resistance [4]. The components with complicated shapes, such as thread or
SC
high-precision, are very difficult to be fabricated via a shaping and sintering procedure. Thus, machining is
M AN U
necessary to fabricate the components. However, the machinability of the β-SiAlON is restricted by its high hardness. Although the thermal shock resistance of β-SiAlON is better than that of other ceramics, the β-SiAlON components would still be damaged due to the thermal stress from the temperature variation. Therefore, both the machinability and the thermal shock resistance of the β-SiAlON need to be improved.
TE D
h-BN and graphite are both soft phase with lamer crystalline structure. Lots of researches were carried out to improve the machinability and the thermal shock resistance of ceramics by addition of h-BN or graphite [5-13]. Due to its chemical inertness to ceramics and oxygen, h-BN is preferable than graphite to incorporate with
EP
ceramics to fabricate composites with excellent machinability and thermal shock resistance [14].
AC C
Drilling test was mostly carried out to assess the machinability of the machinable ceramics. The drilling velocity, wear of the drill, morphology of the dilled hole, wall’s morphology of the drilled hole were calculated or observed to evaluate the machinability [5, 8, 15, 16]. Few further work was published to investigate machinability of the machinable ceramics.
Quenching with water as medium was usually carried out to measure the thermal shock resistance of ceramics. Retained strength of the ceramics after thermal shock was measured, and degradation or retained ratio of the strength was used to evaluate the thermal shock resistance [12, 17-19]. In most of the researches, samples
ACCEPTED MANUSCRIPT undergo the temperature variation just one time. However, the components would suffer repeated thermal shock from the temperature variation for the application at high temperature environment. The temperature variation may from the frequent stops and runs of the devices, or from the process variations during the serving period.
RI PT
Thus, the retained strength of the samples undergoing sudden temperature variation one time cannot evaluate the thermal shock resistance of the materials sufficiently, and repeated thermal shock tests would be imperative to evaluate the thermal shock resistance of ceramics.
SC
In this study, β-SiAlON/h-BN composites with different h-BN content were fabricated via a reaction bonding
M AN U
procedure with Si, Al2O3, AlN, h-BN as raw materials and Sm2O3 as sintering additives. The effects of h-BN on properties of the composites, including density, strength, hardness, fracture toughness, were investigated. Drilling holes and thread tapping were performed to evaluate the machinability of the composites, once and repeated thermal shock tests with water quenching were performed to evaluate the thermal shock resistance.
TE D
2 Experimental
The raw materials were as following: Si (>99 %, particle size: 15 µm, Shandong Huahao Silicon Co., Ltd.,
EP
China), α-Al2O3 (>99.5 %, particle size: 1 µm, Zibo Dongchangye Alumina Co., LTD., China), AlN (>99 %, particle size: 0.5 µm, Hefei MoK Advanced Material Technology Co., LTD., China), h-BN (>99 %, particle size: 2
AC C
µm, YingKou Liaobin Fine Chemicals Co., Ltd., China), Sm2O3 (>99.5 %, particle size: 5 µm, Shanghai Yuelong New Materials Co., Ltd., China). The composition of β-SiAlON with Z = 1 could be obtained by equation (1):
3( 6 − Z ) Si+2 ( 6 − Z ) N2 + ZAl2O3 + ZAlN = 3Si6−Z AlZ OZ N8−Z
(1)
here, assuming that all the Si reacted with nitrogen to form β-SiAlON without any mass loss. The amount of Sm2O3 additive were 5 wt.%, which was chosen according to a previous study on reaction bonded β-SiAlON product [20]. h-BN contents in the aim products were 0, 5 wt.%, 10 wt.%, 15 wt.%, 20 wt.%, 30 wt.%, 40 wt.% and 50 wt.%, respectively.
ACCEPTED MANUSCRIPT The starting powders with different compositions were mixed with ethyl alcohol and alumina balls, then placed in plastic bottles and ball milled for 20 hours to form mixed slurry. The slurry was dried at 80 ºC. Then, PVA solution was mixed with the powder in a mortar. The powder mixture was granulated by pressing at 30 MPa,
× 30 mm, height: 5~6 mm) with cold isotactic pressing (CIP) at 200 MPa.
RI PT
crushing and sieving through 0.5 mm mesh. The resulting granulated powder was shaped to piece (square: 30 mm
The geometric densities (ρg) of the dried green bodies after CIP were calculated by measuring their weights
SC
and the dimensions. The theoretical density of the green body (ρth) without pores was calculated via a direct mixing method by the equation (2):
∑ϕ ϕ ∑ρ
M AN U ρth =
i
(2)
i
i
where, φi and ρi are the mass ratio and theoretical densities of the raw materials in the green body. The content of the temporary bond of PVA (less than 1 wt.%) was omitted for calculation. The theoretical density values of each
TE D
raw material for calculation are shown in Table 1. And the relative density of green body (ρr) was the ratio of ρg to ρth, calculated via the equation (3):
EP
ρr =
ρg ρ th
(3)
AC C
The reaction bonding of the green body was performed in a MoSi2 resistance furnace in nitrogen. The samples were heated to 1150 °C with heating time of 1.5 hours, then annealed for 1 hour. In the next, the samples were heated to 1300 °C with heating rate of 5 °C·min-1 and annealed for 1.5 hours. Finally, the samples were heated to 1450 °C with heating rate of 5 °C·min-1, the holding time at 1450 °C is 2 hours. The nitrogen pressure during the whole heating process is 0.1 MPa. The weights of the samples before and after nitridation were measured, and the nitridation ratio (NR) of the silicon was calculated using the weight gain of the samples. The liner change (∆L) of the samples was calculated according to the dimensions before and after nitridation.
ACCEPTED MANUSCRIPT Archimedes’ principle was used to measure the apparent porosity (Pa) and bulk density (ρb) of the as-sintered samples with distilled water as an immersion medium. For strength test, the as-sintered samples were machined to bars with the dimension of 3 mm × 4 mm × 30 mm, the surfaces of the bars were polished and the edges were
RI PT
beveled. The bending strength (σf) was measured by a three-point bending method with a span of 16 mm and a loading speed of 0.5 mm·min-1. Vicker’s hardness (HV10) was tested by a Vicker’s hardness tester (HVC-10A1, Shanghai Jimin Measuring Equipment Co., Ltd, China) with a load of 10 kg for 15 s. Fracture toughness (KIC) was
SC
measured by a single edge notched beam (SENB) method. The samples for SENB were machined to bars with
M AN U
dimension of 2 mm × 4 mm × 30 mm and a notch was cut with a wheel cutter at the center of the bars along the direction of the 4 mm. The depth of the notch was approximate 2 mm. The thickness of the wheel saw is 0.25 mm and the edge of the saw was modified to make sure the notch radius was approximate 100 µm. Machinability test of the composites was carried out by drilling the samples using the common carbide drill
TE D
and thread tapping. The micrograph for the surface of screw threads was also observed. Thermal shock test was performed by water quenching methods. Both once and repeated thermal shock tests were carried out to investigate the thermal shock resistance of the fabricated composites. For once thermal shock test, the bars of the
EP
samples were put into a furnace, heated up to the preset temperature and maintained for 20 min to ensure the
AC C
temperature of the bars was uniform, then dropped in to room temperature water. The temperature differences (∆T) for the once thermal shock test were 500, 600, 700, 800 and 900 °C, respectively. A repeated thermal shock test with ∆T of 900 °C was performed, the repeat times were 1, 3, 5, 7, and 9, respectively. The retained strength of the bars (σr) after the thermal shock was measured. For all mechanical tests, at least five specimens were measured, mean values and standard deviations were calculated subsequently. An X-ray diffraction instrument (XRD, X’Pert Pro) was used to analyze the phase composition. XRD used Cu Kα radiation (λ=0.15405 nm) as a radiant source operating at 40 kV voltage and 40 mA current. Scanning
ACCEPTED MANUSCRIPT range is 10~80 ° and the scanning speed is 4 °·min-1. A field emission scanning electron microscopy (FE-SEM, FEI Quanta600) was employed to observe the microstructure of the materials.
3 Results and disscussions
process, Pa and ρb of the as-sintered samples are all shown in Table 2.
RI PT
ρg and ρr of the green samples, NR of the silicon and ∆L of the samples before and after reaction bonding
SC
It can be seen in Table 2 that the ρg is increased with addition of the h-BN, and reached the maximum value when the h-BN content is 15 wt.%. The ρr is increased accordingly with the addition of h-BN, and reached the
M AN U
maximum value when the h-BN content was 30 wt.%. The particle size of the Si, Al2O3, AlN, Sm2O3 are approximate 15 µm, 1 µm, 5 µm, and 5 µm, respectively, meanwhile the particle size of the h-BN is approximate 2 µm. During the CIP shaping process, the h-BN could go into the space between the Si particles, the denser particle packing could be obtained. So the ρr is increased with addition of the h-BN. In addition, h-BN has excellent
TE D
lubricity. During the pressing process, h-BN particles, existing between the hard particles (Si, Al2O3, AlN, Sm2O3), could act as lubricant. Thus, the densifying process could be enhanced by the promoted mobility of particles
EP
during the pressing process. So the density of the green body is increased with addition of the h-BN, and the ρr of the green body reached the maximum when the h-BN content was 30 wt.%. However, due to the hindering effect
AC C
of the platelike h-BN particles on the packing of the raw materials, the ρr is decreased with the h-BN content increased from 30 wt.% to 50 wt.%. Therefore, the ρr of the green body goes through a maximum value with the h-BN content increased from 0 to 50 wt.%. It can be noticed from Table 2 that the nitridation ratios for all the samples are higher than 92%. This indicates that a sufficient nitridation of the raw silicon can be achieved at the present experimental conditions and h-BN addition has no obvious contribution on the nitridation ratio. From Table 2, it can be noted that the ∆L for samples without h-BN is approximate -4.2 %. The negative value of the reaction bonded β-SiAlON samples
ACCEPTED MANUSCRIPT means the shrinkage. The liner shrinkage could be attributed to the appearance of the liquid phase during the reaction bonding process [20, 21]. With the h-BN content increases from 0 to 50 wt.%, the ∆L increased from -4.2 % to 1.8 % gradually. During the CIP shaping process of the green body, the soft h-BN was compressed with
RI PT
the elastic energy storage. When the samples were heated to hundreds degree Celsius, the temporary bond of PVA disappeared, the residual stress within the h-BN would be released, and a volume expansion would occur. Moreover, the addition of h-BN could hinder the shrinkage of the samples in the following sintering process
SC
because of the plate-like h-BN particles. Both the volume expansion and the h-BN’s hindering effect on the
M AN U
shrinkage depend on the amount of the h-BN in the samples. Thus, the ∆L of the samples is increased with the amount of h-BN increases.
Table 2 also shows that when the h-BN content increases from 0 to 50 wt.%, the Pa of the samples goes through a minimum while the ρb goes through a maximum, accordingly. The Pa of the samples reached the
TE D
minimum of about 17.0 % when the h-BN content is between 15 wt.% ~ 20 wt.%, and the ρb reaches a maximum of 2.41 g·cm-3 when the h-BN content is 10 wt.%. This indicates that the h-BN content corresponding to the minimum Pa is different with that of the maximum ρb, which can be attributed to the density difference between
EP
h-BN and β-SiAlON. For the composite with h-BN content of 10 wt.%, although its Pa is higher than those of the
AC C
composites with h-BN content of 15 wt.% and 20 wt.%, its ρb is higher due to the higher density of the β-SiAlON (3.10~3.18 g·cm-3 is applicable according to our previous work on dense β-SiAlON with Sm2O3 as sintering additive [20]) than that of the h-BN (2.27 g·cm-3). In addition, the corresponding h-BN content for the green body with highest ρr is also different with that of the as-sintered sample with the lowest Pa. The volume of solid substance in the samples is increased after the sintering due to the reaction of Si with N2, and the volume increase is proceeding in a manner of "inner volume expansion". The inner volume expansion of the samples with lower h-BN content is larger due to more Si in the
ACCEPTED MANUSCRIPT raw materials. Moreover, it has been approved that the samples with lower h-BN content show larger liner shrinkage than those with higher h-BN content. Therefore, the corresponding h-BN content for the composites with the lowest Pa is lower than that of the green body with the highest ρr.
RI PT
σf and HV10 of the as-sintered samples with different h-BN content are shown in Figure 1. It can be seen that σf of the samples with h-BN content no more than 15 wt.% is approximate 160 MPa, without significant decrease with addition of h-BN. However, σf of the as-sintered samples was decreased significantly when the h-BN content
SC
is higher than 20 wt.%, and finally decreases to about 50 MPa for the composite with 50 wt.% h-BN. The trend for σf of the composites is corresponding to their Pa. When the h-BN content is no more than 15 wt.%, Pa of the
M AN U
samples is decreased, so σf of which is not decreased instantly; however, with the h-BN content higher than 20 wt.%, both increased Pa and h-BN content deteriorate the strength of the composites increasingly, so σf shows a significant degradation.
TE D
Different with the strength performance, Vickers hardness of the composites is decreased significantly with the addition of h-BN, and decreased continuously with its content increases (Figure 1). The HV10 of the as-sintered sample without h-BN (the reaction bonded β-SiAlON) is approximate 3.5 GPa, and the HV10 of the
EP
composites with 50 wt.% h-BN is only about 0.3 GPa. Hardness is the main factor affecting the machinability of
AC C
ceramic materials. Generally, Vickers hardness shows highest correlation with the subjective machinability, and lower Vickers hardness results in better subjective machinability of the ceramics [22]. It could be conclude from the strength and Vickers hardness of the as-sintered samples that the machinability of the composites with h-BN content no more than 15 wt.% could be improved obviously without significant strength degradation. KIC of the as-sintered samples measured by a SENB method are shown in Figure 2. It should be noted out that the samples were fabricated via a reaction bonding procedure at 1450 °C, and contain through pores. The microstructure of the samples is coarse and the notch-tip radius of 100 µm would not cause much overestimation
ACCEPTED MANUSCRIPT of the KIC values. The measured KIC value would be valid in this work. It can be seen in Figure 2 that KIC of the as-sintered samples is decreased with the increase of h-BN content from 0 to 50 wt.%, and the trend of KIC is similar with that of σf (Figure 1). KIC of the as-sintered samples without h-BN content (the reaction bonded
RI PT
β-SiAlON) is approximate 2.8 MPa·m1/2, and the value of the as-sintered samples with 50 wt.% h-BN is decreased to approximate 1.0 MPa·m1/2.
XRD patterns of the as-sintered samples with h-BN content of 0, 10 wt.%, 20 wt.%, 30 wt.%, 40 wt.%, 50
SC
wt.%, are shown in Figure 3. No diffraction peak of Si could be observed in the as-sintered samples, which means
M AN U
that all the Si reacted with N2 without residual Si in the samples. Only diffraction peaks of β-SiAlON could be observed in the as-sintered samples without h-BN. The highest diffraction peak of h-BN is the diffraction peak of the (002) lattices plane, which is very close to that of the β-SiAlON (200) lattices plane. Thus, the diffraction peak of the β-SiAlON (200) lattices plane for the sample with h-BN content of 10 wt.% shows a slight widening to the
TE D
left. With the increase of h-BN content, the intensity of the β-SiAlON’s diffraction peaks become weaker, and the intensity of the h-BN diffraction peaks become stronger. For the as-sintered samples with h-BN content of 50
lattices plane.
EP
wt.%, the height of the diffraction peak of h-BN (002) lattices plane is higher than that of the β-SiAlON (200)
AC C
The micrograph for the fracture surface of the as-sintered samples with h-BN content of 0, 10 wt.%, 20 wt.%, 30 wt.%, 40 wt.%, 50 wt.%, are shown in Figure 4. For the samples without h-BN (reaction bonded β-SiAlON), pores can be observed clearly (Figure 4(a)). However, for the samples with small amount of h-BN, most of the pores are filled with the h-BN particles (Figure 4(b) and (c)). With the increase of h-BN content, more and more h-BN can be observed on the fracture surfaces of the samples. Fracture surfaces of the samples with h-BN content of 40 wt.% and 50 wt.% are completely covered with the h-BN, and β-SiAlON matrix could rarely be observed (Figure 4 (e) and (f)). Table 2 shows that Pa of the as-sintered samples is firstly decreased and then increased with
ACCEPTED MANUSCRIPT the h-BN content increase from 0 to 50 wt.%. h-BN acts as pore filler in the composites with less h-BN introduced, however, it would hinder the densification process during the sintering process when more h-BN was added into the β-SiAlON. Thus, the porosity of the as-sintered samples is decreased when less h-BN (≤ 15 wt.%) was
RI PT
introduced into the composites, and increased when more h-BN (> 20 wt.%) was added. The distribution sketch of the h-BN in the composites is shown in Figure 5. For the sample without h-BN addition, that is, the reaction bonded β-SiAlON, pores are the flaws weakening the strength of the materials, as
SC
shown in Figure 5(a). Thus, the reaction bonded β-SiAlON samples show inferior strength than the sintered ones.
M AN U
For the samples with small amount of h-BN, due to filling of the h-BN in the pores, the porosity of the composites is decreased, the flaw size is not increased significantly (as shown in Figure 5(b)). Thus, the strength of the composites is not decreased significantly with addition of a small amount of h-BN. For the samples with large amount of h-BN, h-BN particles would connect with each other, and the flaw size in the composites would be
TE D
increased (as shown in Figure 5(c)). The increase of flaw (or crack) size ascribing to the addition of h-BN can be verified by the calculation based on the Irwin equation [23]. According to Irwin’s work, the relationship among
EP
KIC, σf and the crack size (2c) in materials is as equation (4):
K IC = Y σ f c
(4)
AC C
where, Y is the dimensionless parameter, for Inglis loading geometry, Y=√π [23]. Hence, c for Inglis loading geometry can be calculated according to equation (4), and is shown in Figure 6. It can be seen that the crack size (2c) is increased with increase of h-BN content. The drilled sample of the composite with 20 wt.% h-BN is shown in Figure 7(a). It can be seen that the holes are cleanly drilled with no evidence of cracking or chipping. Due to the improved machinability of the composites, component with complicated shapes could be obtained by a machining process. A square nut was made out by machining of the composite with 30 wt.% h-BN, and the profile of the square nut is shown in Figure 7(b). The
ACCEPTED MANUSCRIPT section photo of the threaded hole is shown in Figure 7(c), and the micrograph of the screw threads is shown in Figure 7(d). It can be seen from Figure 7(c) and (d) that the screw threads are complete without obvious defects. Therefore, it can be concluded that the machinability of the composites is improved significantly and the
RI PT
complicated components could be obtained easily via a machining procedure. The machinability of the composites is commendable.
σr of the samples with different h-BN content after once and repeated thermal shock tests are shown in Figure
SC
8. Due to lots of data points in the figure, the error bars were omitted to make sure the curves can be read easily. It can be seen in Figure 8(a) that σr of the samples with h-BN content no more than 20 wt.% are decreased obviously
M AN U
when ∆T is higher than 500 °C. However, the strength degradation of the samples after thermal shock are decreased with the increase of h-BN content from 0 to 20 wt.%. σr of the samples with h-BN content of 30 wt.% and 40 wt.% is decreased not suddenly but gradually. The composite with 50 wt.% h-BN shows the slightest
TE D
strength degradation among the ∆T values in this work.
It can been seen from Figure 8(b) that σr of the samples without h-BN is decreased with the increase of thermal shock times from 1 to 9, and the σr after 9 times thermal shock is approximate 1/10 of that of the
EP
as-sintered samples. σr of the samples with 10 wt.%, 20 wt.% and 30 wt.% h-BN suffering one time thermal shock
AC C
test is decreased significantly, and then decreased slowly with the increase of thermal shock times from 1 to 9. It should be noted that the strength degradation of the samples with h-BN content in the range of 10 wt.% ~ 30 wt.% is lower than that of the samples without h-BN during the thermal shock tests. σr of the samples with h-BN content more than 30 wt.% is decreased with the thermal shock times increases from 1 to 5, and tends basically to steady with the thermal shock times increases from 5 to 9. Generally, ceramics are brittle and may show catastrophic damage when they are cooled or heated quickly. The thermal shock resistance of ceramics can be stated as the ability to avoid catastrophic damage, or to maintain
ACCEPTED MANUSCRIPT integrity, after sudden temperature variation. In this work, the strength of the composites were decreased with addition of the h-BN. For samples undergo once thermal shock, the retained strength of the composite without h-BN was higher than that of the composite with h-BN addition. It seems that the thermal shock resistance of the
RI PT
samples with h-BN addition is not improved. However, it should be noted that when the samples undergo repeated thermal shock, the retained strength of the sample without h-BN were further decreased with the thermal shock times increased from 1 to 9. It could be predicted that the higher strength of the samples without h-BN addition
SC
would be decreased to zero, which means fracture of the samples. The samples with h-BN addition would remain
M AN U
their strength for much more thermal shock times. That is the ability of the composite to avoid catastrophic damage or to maintain its integrity is improved with h-BN addition. So, the composites with h-BN addition, even their initial strength is lower, would be more reliable in the thermal shock environment, especially in the repeated thermal shock environment. Therefore, the thermal shock resistance of the as-sintered samples is improved
TE D
efficiently by addition of h-BN, and the fabricated composites with h-BN content more than 20 wt.% have excellent thermal shock resistance.
Soft phase of h-BN in the composites could act as flaws (or cracks) in the β-SiAlON matrix. When more
EP
h-BN is introduced into the composites, the size and the density of the cracks would be increased, and the cracks
AC C
propagation during the thermal shock would be quasi-static, which makes the retained strength of the composites after once or repeated thermal shock be decreased indistinctively [24, 25]. In addition, h-BN is different with the common cracks. h-BN can exhibit a series of energy absorbing mechanism under thermal stress [26-30]. Therefore, the thermal shock resistance of the composites is improved with addition of the h-BN, and the composites would present higher reliability than monomial reaction bonded β-SiAlON in the application.
4 Conclusion
β-SiAlON/h-BN composites with improved machinability and thermal shock resistance were fabricated via a
ACCEPTED MANUSCRIPT reaction bonding procedure. The relative density of the green bodies and the bulk density of the composites go through a maximum value with the h-BN content increased from 0 to 50 wt.%. Due to the increased bulk density of the composites with small amount of h-BN addition, the Vickers hardness of the composites is decreased
RI PT
without obviously strength loss, so the machinability of the composites could be improved with good strength. Fracture toughness of the composites is decreased with h-BN content increased from 0 to 50 wt.%. Drilling holes and thread tapping tests show that the composites with h-BN have commendable machinability. Due to the
SC
quasi-static crack propagation and energy absorbing mechanism of h-BN, the thermal shock resistance of the
M AN U
composites with h-BN content higher than 20 wt.% is excellent, and the retained strength of the composites is decreased slightly with the increase of temperature difference and thermal shock times.
Acknowledgement
This research was financially supported by the Scientific Research Program Funded by Xi’an University of
TE D
Architecture and Technology (No. RC1529; QN1523), the Young Doctoral Research Foundation in College of Materials and Mineral Resources, the Scientific Research Program Funded by Shaanxi Provincial Education
EP
Department (No. 14JK1393), the Program for Young Excellent Talents in Shaanxi Province (2013KJXX-50), National Natural Science Foundation of China (No. 51272203).
AC C
References
[1] K. Jack, Sialons and related nitrogen ceramics, J. Mater. Sci., 11 (1976) 1135-1158. [2] T. Ekstrom, M. Nygren, SiAlON ceramics, J. Am. Ceram. Soc., 75 (1992) 259-276. [3] F.L. Riley, Silicon nitride and related materials, J. Am. Ceram. Soc., 83 (2000) 245-265. [4] C. Raju, S. Vermab, M. Sahua, P. Jaina, S. Choudary, Silicon nitride/SiAlON ceramics-A review, Indian J. Eng. Mater. S., 8 (2001) 36-45.
ACCEPTED MANUSCRIPT [5] Y.L. Li, G.J. Qiao, Z.H. Jin, Machinable Al2O3/BN composite ceramics with strong mechanical properties, Mater. Res. Bull., 37 (2002) 1401-1409. [6] T. Kusunose, Y.H. Kim, T. Sekino, T. Matsumoto, N. Tanaka, T. Nakayama, K. Niihara, Fabrication of Al2O3/BN
RI PT
nanocomposites by chemical processing and their mechanical properties, J. Mater. Res., 20 (2005) 183-190. [7] T. Kusunose, N. Sakayanagi, T. Sekino, Y. Ando, Fabrication and characterization of aluminum nitride/boron nitride nanocomposites by carbothermal reduction and nitridation of aluminum borate powders, J Nanosci.
SC
Nanotechno., 8 (2008) 5846-5853.
M AN U
[8] H.Y. Jin, W. Wang, J.Q. Gao, G.J. Qiao, Z.H. Jin, Study of machinable AlN/BN ceramic composites, Mater. Lett., 60 (2006) 190-193.
[9] W.L. Yang, Z.Q. Shi, Y.F. Li, C.J. Zhao, G.J. Qiao, Z.H. Jin, Thermal shock and oxidation resistance of SiC composite ceramic sheath for immersion heater, J. Chin. Ceram. Soc., 42 (2014) 698-702.
TE D
[10] H. Jin, S.H. Meng, Y.W. Zhu, Y.J. Zhou, Effect of environment atmosphere on thermal shock resistance of the ZrB2-SiC-graphite composite, Mater. Des., 50 (2013) 509-514. [11] S.B. Zhou, Z. Wang, W. Zhang, Effect of graphite flake orientation on microstructure and mechanical properties of
EP
ZrB2-SiC-graphite composite, J. Alloys Compd., 485 (2009) 181-185.
AC C
[12] Z. Wang, C.Q. Hong, X.H. Zhang, X. Sun, J.C. Han, Microstructure and thermal shock behavior of ZrB2-SiC-graphite composite, Mater. Chem. Phys., 113 (2009) 338-341. [13] X.H. Zhang, Z. Wang, X. Sun, W.B. Han, C.Q. Hong, Effect of graphite flake on the mechanical properties of hot pressed ZrB2-SiC ceramics, Mater. Lett., 62 (2008) 4360-4362. [14] T. Jiang, Research progress and development of machinable ceramics composites, Mater. Rev., 26 (2012) 49-53, 65. [15] X.D. Wang, G.J. Qiao, Z.H. Jin, Fabrication of machinable silicon carbide-boron nitride ceramic nanocomposites, J.
ACCEPTED MANUSCRIPT Am. Ceram. Soc., 87 (2004) 565-570. [16] H.Y. Jin, H. Xu, G.J. Qiao, J.Q. Gao, Z.H. Jin, Study of machinable silicon carbide-boron nitride ceramic composites, Mater. Sci. Eng., A, 483-484 (2008) 214-217.
metals as heating medium, Ceram. Int., 41 (2015) 6117-6121.
RI PT
[17] Y.J. Li, H.L. Yu, H.Y. Jin, Z.Q. Shi, G.J. Qiao, Z.H. Jin, Fast heating thermal shock test for β-SiAlON with molten
[18] M. Chen, H.J. Wang, H.Y. Jin, X.D. Pan, Z.H. Jin, Transient thermal shock behavior simulation of porous silicon
SC
nitride ceramics, Ceram. Int., 42 (2016) 3130-3137.
M AN U
[19] T. Jiang, C.C. Tian, Investigation of microstructure and thermal shock resistance of the B4C/BN composites fabricated by hot-pressing process, Key Eng. Mater., 512-515 (2012) 748-752. [20] Y.J. Li, D.H. Liu, C.F. Zeng, Z.Q. Shi, Z.H. Jin, Effects of Sm2O3 content on the microstructure and mechanical properties of post-sintered reaction-bonded β-SiAlON, J. Mater. Eng. Perform., 25 (2016) 1143-1149.
TE D
[21] Y.J. Li, D.H. Liu, H. Jin, D.H. Ding, G.Q. Xiao, Z.Q. Shi, Z.H. Jin, Effect of Z values on the microstructure and mechanical properties of post-sintered reaction bonded β-SiAlON, High Temp. Mater. Proc., (DOI: 10.1515/htmp-2015-0256).
EP
[22] J.B. Quinn, I.K. Lloyd, R.N. Katz, G.D. Quinn, Ceramic "Machinability"-What does it mean, in: W.M. Kriven,
AC C
H.-a. Lin (Eds.) 27th International Cocoa Beach Conference on Advanced Ceramics and Composites, 2003, pp. 511-516.
[23] D.J. Green, An introduction to the mechanical properties of ceramics, Cambridge University Press, Cambridge, 1998.
[24] D.P.H. Hasselman, Unified theory of thermal shock fracture initiation and crack propagation in brittle ceramics, J. Am. Ceram. Soc., 52 (1969) 600-604. [25] Y.J. Li, Fabrication and properties research for β-SiAlON matrix composites with high strength and thermal shock
ACCEPTED MANUSCRIPT resistance for molten metals, Xi'an, Xi'an Jiaotong University, Ph. D. (2015). [26] B. Zhong, G.L. Zhao, X.X. Huang, L. Xia, X.H. Tang, S.C. Zhang, G.W. Wen, Microstructure and mechanical properties of ZTA/BN machinable ceramics fabricated by reactive hot pressing, J. Eur. Ceram. Soc., 35 (2015)
RI PT
641-649. [27] D.L. Cai, Z.H. Yang, X.M. Duan, B. Liang, Q. Li, D.C. Jia, Y. Zhou, A novel BN-MAS system composite ceramics with greatly improved mechanical properties prepared by low temperature hot-pressing, Mater. Sci. Eng., A, 633
SC
(2015) 194-199.
M AN U
[28] X.M. Duan, D.C. Jia, Z.L. Wu, Z. Tian, Z.H. Yang, S.J. Wang, Y. Zhou, Effect of sintering pressure on the texture of hot-press sintered hexagonal boron nitride composite ceramics, Scripta Mater., 68 (2013) 104-107. [29] Z.Q. Shi, J.P. Wang, G.J. Qiao, Z.H. Jin, Effects of weak boundary phases (WBP) on the microstructure and mechanical properties of pressureless sintered Al2O3/h-BN machinable composites, Mater. Sci. Eng., A, 492 (2008)
TE D
29-34.
[30] Y.T. Zheng, H.B. Li, T. Zhou, Microstructure and mechanical properties of h-BN-SiC ceramic composites prepared
AC C
List of table(s)
EP
by in situ combustion synthesis, Mater. Sci. Eng., A, 540 (2012) 102-106.
Table 1 Theoretical densities of raw materials for calculation
Raw materials
-3
Density/g·cm
Si
Al2O3
AlN
Sm2O3
h-BN
2.33
3.97
3.26
8.35
2.27
Table 2 ρg, ρr of the green samples, NR and ∆L of the samples before and after reaction bonding process, Pa and ρb of the as-sintered samples BN content/wt.%
ρg/g·cm-3
ρr/%
NR/%
∆L/%
Pa/%
ρb/g·cm-3
0
1.71
63.1
93.0
-4.2
22.0
2.39
5
1.80
67.3
93.6
-2.4
20.1
2.40
10
1.85
70.0
93.3
-0.6
17.6
2.41
ACCEPTED MANUSCRIPT 1.88
72.4
93.1
-0.1
17.0
2.34
20
1.87
72.7
93.5
0
16.9
2.32
30
1.86
73.8
93.0
1.1
18.4
2.19
40
1.79
72.8
93.6
1.5
21.1
2.06
50
1.74
71.9
92.9
1.8
24.8
1.92
List of figure(s)
RI PT
15
Figure 1 σf and HV10 of the as-sintered samples with different h-BN content
SC
Figure 2 KIC of the as-sintered samples with different h-BN content
Figure 3 XRD patterns of the as-sintered samples with different h-BN content
M AN U
Figure 4 Micrograph for the fracture surface of as-sintered samples with different h-BN content: (a) without h-BN, (b) 10 wt.%, (c) 20 wt.%, (d) 30 wt.%, (e) 40 wt.%, (f) 50 wt.%
Figure 5 Sketch of the h-BN distribution in the composites: (a) without h-BN, (b) less h-BN, (c) more h-BN Figure 6 Half-length (c) of the cracks for Inglis loading geometry
TE D
Figure 7 Photograph of the samples after machinability tests: (a) drilled sample with h-BN of 20 wt.%, (b) square nut made from the composite with h-BN of 30 wt.%, (c) photograph of the screw threads made from the
EP
composite with h-BN of 50 wt.%, and (d) micrograph of the screw threads
AC C
Figure 8 σr of the samples with different h-BN content after once and repeated thermal shock tests
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Density of the composites could be increased with addition of small amount of h-BN Thread tapping test showed that the composites have commendable
RI PT
machinability Repeated thermal shock showed excellent thermal shock resistance of the
AC C
EP
TE D
M AN U
SC
composites
S0925-8388(17)30388-2
DOI:
10.1016/j.jallcom.2017.01.318
Reference:
JALCOM 40685
To appear in:
Journal of Alloys and Compounds
Received Date: 7 September 2016 Revised Date:
26 January 2017
Accepted Date: 29 January 2017
Please cite this article as: Y. Li, B. Ge, Z. Wu, G. Xiao, Z. Shi, Z. Jin, Effects of h-BN on mechanical properties of reaction bonded β-SiAlON/h-BN composites, Journal of Alloys and Compounds (2017), doi: 10.1016/j.jallcom.2017.01.318. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Effects of h-BN on mechanical properties of reaction bonded β-SiAlON/h-BN composites Yanjun Li1,*, Bangzhi Ge2, Zhihong Wu1, Guoqing Xiao1, Zhongqi Shi2,*, Zhihao Jin2 College of Materials and Mineral Resources, Xi’an University of Architecture and Technology, Xi’an
RI PT
1
710055, China
State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an 710049,
SC
2
China
M AN U
*Corresponding authors: Yanjun Li, E-mail: [email protected], Tel: +86 29 82205798, Fax: +86 29 82205798; Zhongqi Shi, E-mail: [email protected], Tel: +86 29 82667942, Fax: +86 29 82663453. Abstract: β-SiAlON/h-BN composites with improved machinability and thermal shock resistance were fabricated via a reaction bonding procedure using raw materials of Si, Al2O3, AlN and h-BN. Effects of h-BN on
TE D
properties of the composites, including density, strength, hardness, fracture toughness, machinability and thermal shock resistance, were investigated. XRD analysis was performed to analyze the phase composition of the
EP
composites, and SEM was employed to analyze their microstructure. The results showed that the density of the composites went through a maximum value with h-BN content increased from 0 to 50 wt.%. Vickers hardness of
AC C
the composites with less h-BN addition was decreased without obvious strength loss. Fracture toughness of the composites was decreased with h-BN content increased from 0 to 50 wt.%. Drilling hole and thread tapping test showed that the machinability of the composites with h-BN addition was commendable. Thermal shock resistance of the composites with h-BN content higher than 20 wt.% was excellent, and the retained strength of the composites was decreased slightly with the temperature difference and thermal shock times increased. Key words: SiAlON; Machinable; Thermal shock resistance
ACCEPTED MANUSCRIPT 1 Introduction
β-SiAlON solid solutions inherit the crystal structure of β-Si3N4 with equivalent substitution of Al-O for Si-N, and have chemical formula of Si6-ZAlZOZN8-Z (Z=0~4.2), where Z is the number of (Si-N) bonds being replaced by
RI PT
(Al-O) bonds [1-3]. β-SiAlON is a promising material in nonferrous metal industry due to its poor wettability to nonferrous metals melts, high strength at room temperature and high temperature, high toughness, low thermal expansion, and well oxidation resistance [4]. The components with complicated shapes, such as thread or
SC
high-precision, are very difficult to be fabricated via a shaping and sintering procedure. Thus, machining is
M AN U
necessary to fabricate the components. However, the machinability of the β-SiAlON is restricted by its high hardness. Although the thermal shock resistance of β-SiAlON is better than that of other ceramics, the β-SiAlON components would still be damaged due to the thermal stress from the temperature variation. Therefore, both the machinability and the thermal shock resistance of the β-SiAlON need to be improved.
TE D
h-BN and graphite are both soft phase with lamer crystalline structure. Lots of researches were carried out to improve the machinability and the thermal shock resistance of ceramics by addition of h-BN or graphite [5-13]. Due to its chemical inertness to ceramics and oxygen, h-BN is preferable than graphite to incorporate with
EP
ceramics to fabricate composites with excellent machinability and thermal shock resistance [14].
AC C
Drilling test was mostly carried out to assess the machinability of the machinable ceramics. The drilling velocity, wear of the drill, morphology of the dilled hole, wall’s morphology of the drilled hole were calculated or observed to evaluate the machinability [5, 8, 15, 16]. Few further work was published to investigate machinability of the machinable ceramics.
Quenching with water as medium was usually carried out to measure the thermal shock resistance of ceramics. Retained strength of the ceramics after thermal shock was measured, and degradation or retained ratio of the strength was used to evaluate the thermal shock resistance [12, 17-19]. In most of the researches, samples
ACCEPTED MANUSCRIPT undergo the temperature variation just one time. However, the components would suffer repeated thermal shock from the temperature variation for the application at high temperature environment. The temperature variation may from the frequent stops and runs of the devices, or from the process variations during the serving period.
RI PT
Thus, the retained strength of the samples undergoing sudden temperature variation one time cannot evaluate the thermal shock resistance of the materials sufficiently, and repeated thermal shock tests would be imperative to evaluate the thermal shock resistance of ceramics.
SC
In this study, β-SiAlON/h-BN composites with different h-BN content were fabricated via a reaction bonding
M AN U
procedure with Si, Al2O3, AlN, h-BN as raw materials and Sm2O3 as sintering additives. The effects of h-BN on properties of the composites, including density, strength, hardness, fracture toughness, were investigated. Drilling holes and thread tapping were performed to evaluate the machinability of the composites, once and repeated thermal shock tests with water quenching were performed to evaluate the thermal shock resistance.
TE D
2 Experimental
The raw materials were as following: Si (>99 %, particle size: 15 µm, Shandong Huahao Silicon Co., Ltd.,
EP
China), α-Al2O3 (>99.5 %, particle size: 1 µm, Zibo Dongchangye Alumina Co., LTD., China), AlN (>99 %, particle size: 0.5 µm, Hefei MoK Advanced Material Technology Co., LTD., China), h-BN (>99 %, particle size: 2
AC C
µm, YingKou Liaobin Fine Chemicals Co., Ltd., China), Sm2O3 (>99.5 %, particle size: 5 µm, Shanghai Yuelong New Materials Co., Ltd., China). The composition of β-SiAlON with Z = 1 could be obtained by equation (1):
3( 6 − Z ) Si+2 ( 6 − Z ) N2 + ZAl2O3 + ZAlN = 3Si6−Z AlZ OZ N8−Z
(1)
here, assuming that all the Si reacted with nitrogen to form β-SiAlON without any mass loss. The amount of Sm2O3 additive were 5 wt.%, which was chosen according to a previous study on reaction bonded β-SiAlON product [20]. h-BN contents in the aim products were 0, 5 wt.%, 10 wt.%, 15 wt.%, 20 wt.%, 30 wt.%, 40 wt.% and 50 wt.%, respectively.
ACCEPTED MANUSCRIPT The starting powders with different compositions were mixed with ethyl alcohol and alumina balls, then placed in plastic bottles and ball milled for 20 hours to form mixed slurry. The slurry was dried at 80 ºC. Then, PVA solution was mixed with the powder in a mortar. The powder mixture was granulated by pressing at 30 MPa,
× 30 mm, height: 5~6 mm) with cold isotactic pressing (CIP) at 200 MPa.
RI PT
crushing and sieving through 0.5 mm mesh. The resulting granulated powder was shaped to piece (square: 30 mm
The geometric densities (ρg) of the dried green bodies after CIP were calculated by measuring their weights
SC
and the dimensions. The theoretical density of the green body (ρth) without pores was calculated via a direct mixing method by the equation (2):
∑ϕ ϕ ∑ρ
M AN U ρth =
i
(2)
i
i
where, φi and ρi are the mass ratio and theoretical densities of the raw materials in the green body. The content of the temporary bond of PVA (less than 1 wt.%) was omitted for calculation. The theoretical density values of each
TE D
raw material for calculation are shown in Table 1. And the relative density of green body (ρr) was the ratio of ρg to ρth, calculated via the equation (3):
EP
ρr =
ρg ρ th
(3)
AC C
The reaction bonding of the green body was performed in a MoSi2 resistance furnace in nitrogen. The samples were heated to 1150 °C with heating time of 1.5 hours, then annealed for 1 hour. In the next, the samples were heated to 1300 °C with heating rate of 5 °C·min-1 and annealed for 1.5 hours. Finally, the samples were heated to 1450 °C with heating rate of 5 °C·min-1, the holding time at 1450 °C is 2 hours. The nitrogen pressure during the whole heating process is 0.1 MPa. The weights of the samples before and after nitridation were measured, and the nitridation ratio (NR) of the silicon was calculated using the weight gain of the samples. The liner change (∆L) of the samples was calculated according to the dimensions before and after nitridation.
ACCEPTED MANUSCRIPT Archimedes’ principle was used to measure the apparent porosity (Pa) and bulk density (ρb) of the as-sintered samples with distilled water as an immersion medium. For strength test, the as-sintered samples were machined to bars with the dimension of 3 mm × 4 mm × 30 mm, the surfaces of the bars were polished and the edges were
RI PT
beveled. The bending strength (σf) was measured by a three-point bending method with a span of 16 mm and a loading speed of 0.5 mm·min-1. Vicker’s hardness (HV10) was tested by a Vicker’s hardness tester (HVC-10A1, Shanghai Jimin Measuring Equipment Co., Ltd, China) with a load of 10 kg for 15 s. Fracture toughness (KIC) was
SC
measured by a single edge notched beam (SENB) method. The samples for SENB were machined to bars with
M AN U
dimension of 2 mm × 4 mm × 30 mm and a notch was cut with a wheel cutter at the center of the bars along the direction of the 4 mm. The depth of the notch was approximate 2 mm. The thickness of the wheel saw is 0.25 mm and the edge of the saw was modified to make sure the notch radius was approximate 100 µm. Machinability test of the composites was carried out by drilling the samples using the common carbide drill
TE D
and thread tapping. The micrograph for the surface of screw threads was also observed. Thermal shock test was performed by water quenching methods. Both once and repeated thermal shock tests were carried out to investigate the thermal shock resistance of the fabricated composites. For once thermal shock test, the bars of the
EP
samples were put into a furnace, heated up to the preset temperature and maintained for 20 min to ensure the
AC C
temperature of the bars was uniform, then dropped in to room temperature water. The temperature differences (∆T) for the once thermal shock test were 500, 600, 700, 800 and 900 °C, respectively. A repeated thermal shock test with ∆T of 900 °C was performed, the repeat times were 1, 3, 5, 7, and 9, respectively. The retained strength of the bars (σr) after the thermal shock was measured. For all mechanical tests, at least five specimens were measured, mean values and standard deviations were calculated subsequently. An X-ray diffraction instrument (XRD, X’Pert Pro) was used to analyze the phase composition. XRD used Cu Kα radiation (λ=0.15405 nm) as a radiant source operating at 40 kV voltage and 40 mA current. Scanning
ACCEPTED MANUSCRIPT range is 10~80 ° and the scanning speed is 4 °·min-1. A field emission scanning electron microscopy (FE-SEM, FEI Quanta600) was employed to observe the microstructure of the materials.
3 Results and disscussions
process, Pa and ρb of the as-sintered samples are all shown in Table 2.
RI PT
ρg and ρr of the green samples, NR of the silicon and ∆L of the samples before and after reaction bonding
SC
It can be seen in Table 2 that the ρg is increased with addition of the h-BN, and reached the maximum value when the h-BN content is 15 wt.%. The ρr is increased accordingly with the addition of h-BN, and reached the
M AN U
maximum value when the h-BN content was 30 wt.%. The particle size of the Si, Al2O3, AlN, Sm2O3 are approximate 15 µm, 1 µm, 5 µm, and 5 µm, respectively, meanwhile the particle size of the h-BN is approximate 2 µm. During the CIP shaping process, the h-BN could go into the space between the Si particles, the denser particle packing could be obtained. So the ρr is increased with addition of the h-BN. In addition, h-BN has excellent
TE D
lubricity. During the pressing process, h-BN particles, existing between the hard particles (Si, Al2O3, AlN, Sm2O3), could act as lubricant. Thus, the densifying process could be enhanced by the promoted mobility of particles
EP
during the pressing process. So the density of the green body is increased with addition of the h-BN, and the ρr of the green body reached the maximum when the h-BN content was 30 wt.%. However, due to the hindering effect
AC C
of the platelike h-BN particles on the packing of the raw materials, the ρr is decreased with the h-BN content increased from 30 wt.% to 50 wt.%. Therefore, the ρr of the green body goes through a maximum value with the h-BN content increased from 0 to 50 wt.%. It can be noticed from Table 2 that the nitridation ratios for all the samples are higher than 92%. This indicates that a sufficient nitridation of the raw silicon can be achieved at the present experimental conditions and h-BN addition has no obvious contribution on the nitridation ratio. From Table 2, it can be noted that the ∆L for samples without h-BN is approximate -4.2 %. The negative value of the reaction bonded β-SiAlON samples
ACCEPTED MANUSCRIPT means the shrinkage. The liner shrinkage could be attributed to the appearance of the liquid phase during the reaction bonding process [20, 21]. With the h-BN content increases from 0 to 50 wt.%, the ∆L increased from -4.2 % to 1.8 % gradually. During the CIP shaping process of the green body, the soft h-BN was compressed with
RI PT
the elastic energy storage. When the samples were heated to hundreds degree Celsius, the temporary bond of PVA disappeared, the residual stress within the h-BN would be released, and a volume expansion would occur. Moreover, the addition of h-BN could hinder the shrinkage of the samples in the following sintering process
SC
because of the plate-like h-BN particles. Both the volume expansion and the h-BN’s hindering effect on the
M AN U
shrinkage depend on the amount of the h-BN in the samples. Thus, the ∆L of the samples is increased with the amount of h-BN increases.
Table 2 also shows that when the h-BN content increases from 0 to 50 wt.%, the Pa of the samples goes through a minimum while the ρb goes through a maximum, accordingly. The Pa of the samples reached the
TE D
minimum of about 17.0 % when the h-BN content is between 15 wt.% ~ 20 wt.%, and the ρb reaches a maximum of 2.41 g·cm-3 when the h-BN content is 10 wt.%. This indicates that the h-BN content corresponding to the minimum Pa is different with that of the maximum ρb, which can be attributed to the density difference between
EP
h-BN and β-SiAlON. For the composite with h-BN content of 10 wt.%, although its Pa is higher than those of the
AC C
composites with h-BN content of 15 wt.% and 20 wt.%, its ρb is higher due to the higher density of the β-SiAlON (3.10~3.18 g·cm-3 is applicable according to our previous work on dense β-SiAlON with Sm2O3 as sintering additive [20]) than that of the h-BN (2.27 g·cm-3). In addition, the corresponding h-BN content for the green body with highest ρr is also different with that of the as-sintered sample with the lowest Pa. The volume of solid substance in the samples is increased after the sintering due to the reaction of Si with N2, and the volume increase is proceeding in a manner of "inner volume expansion". The inner volume expansion of the samples with lower h-BN content is larger due to more Si in the
ACCEPTED MANUSCRIPT raw materials. Moreover, it has been approved that the samples with lower h-BN content show larger liner shrinkage than those with higher h-BN content. Therefore, the corresponding h-BN content for the composites with the lowest Pa is lower than that of the green body with the highest ρr.
RI PT
σf and HV10 of the as-sintered samples with different h-BN content are shown in Figure 1. It can be seen that σf of the samples with h-BN content no more than 15 wt.% is approximate 160 MPa, without significant decrease with addition of h-BN. However, σf of the as-sintered samples was decreased significantly when the h-BN content
SC
is higher than 20 wt.%, and finally decreases to about 50 MPa for the composite with 50 wt.% h-BN. The trend for σf of the composites is corresponding to their Pa. When the h-BN content is no more than 15 wt.%, Pa of the
M AN U
samples is decreased, so σf of which is not decreased instantly; however, with the h-BN content higher than 20 wt.%, both increased Pa and h-BN content deteriorate the strength of the composites increasingly, so σf shows a significant degradation.
TE D
Different with the strength performance, Vickers hardness of the composites is decreased significantly with the addition of h-BN, and decreased continuously with its content increases (Figure 1). The HV10 of the as-sintered sample without h-BN (the reaction bonded β-SiAlON) is approximate 3.5 GPa, and the HV10 of the
EP
composites with 50 wt.% h-BN is only about 0.3 GPa. Hardness is the main factor affecting the machinability of
AC C
ceramic materials. Generally, Vickers hardness shows highest correlation with the subjective machinability, and lower Vickers hardness results in better subjective machinability of the ceramics [22]. It could be conclude from the strength and Vickers hardness of the as-sintered samples that the machinability of the composites with h-BN content no more than 15 wt.% could be improved obviously without significant strength degradation. KIC of the as-sintered samples measured by a SENB method are shown in Figure 2. It should be noted out that the samples were fabricated via a reaction bonding procedure at 1450 °C, and contain through pores. The microstructure of the samples is coarse and the notch-tip radius of 100 µm would not cause much overestimation
ACCEPTED MANUSCRIPT of the KIC values. The measured KIC value would be valid in this work. It can be seen in Figure 2 that KIC of the as-sintered samples is decreased with the increase of h-BN content from 0 to 50 wt.%, and the trend of KIC is similar with that of σf (Figure 1). KIC of the as-sintered samples without h-BN content (the reaction bonded
RI PT
β-SiAlON) is approximate 2.8 MPa·m1/2, and the value of the as-sintered samples with 50 wt.% h-BN is decreased to approximate 1.0 MPa·m1/2.
XRD patterns of the as-sintered samples with h-BN content of 0, 10 wt.%, 20 wt.%, 30 wt.%, 40 wt.%, 50
SC
wt.%, are shown in Figure 3. No diffraction peak of Si could be observed in the as-sintered samples, which means
M AN U
that all the Si reacted with N2 without residual Si in the samples. Only diffraction peaks of β-SiAlON could be observed in the as-sintered samples without h-BN. The highest diffraction peak of h-BN is the diffraction peak of the (002) lattices plane, which is very close to that of the β-SiAlON (200) lattices plane. Thus, the diffraction peak of the β-SiAlON (200) lattices plane for the sample with h-BN content of 10 wt.% shows a slight widening to the
TE D
left. With the increase of h-BN content, the intensity of the β-SiAlON’s diffraction peaks become weaker, and the intensity of the h-BN diffraction peaks become stronger. For the as-sintered samples with h-BN content of 50
lattices plane.
EP
wt.%, the height of the diffraction peak of h-BN (002) lattices plane is higher than that of the β-SiAlON (200)
AC C
The micrograph for the fracture surface of the as-sintered samples with h-BN content of 0, 10 wt.%, 20 wt.%, 30 wt.%, 40 wt.%, 50 wt.%, are shown in Figure 4. For the samples without h-BN (reaction bonded β-SiAlON), pores can be observed clearly (Figure 4(a)). However, for the samples with small amount of h-BN, most of the pores are filled with the h-BN particles (Figure 4(b) and (c)). With the increase of h-BN content, more and more h-BN can be observed on the fracture surfaces of the samples. Fracture surfaces of the samples with h-BN content of 40 wt.% and 50 wt.% are completely covered with the h-BN, and β-SiAlON matrix could rarely be observed (Figure 4 (e) and (f)). Table 2 shows that Pa of the as-sintered samples is firstly decreased and then increased with
ACCEPTED MANUSCRIPT the h-BN content increase from 0 to 50 wt.%. h-BN acts as pore filler in the composites with less h-BN introduced, however, it would hinder the densification process during the sintering process when more h-BN was added into the β-SiAlON. Thus, the porosity of the as-sintered samples is decreased when less h-BN (≤ 15 wt.%) was
RI PT
introduced into the composites, and increased when more h-BN (> 20 wt.%) was added. The distribution sketch of the h-BN in the composites is shown in Figure 5. For the sample without h-BN addition, that is, the reaction bonded β-SiAlON, pores are the flaws weakening the strength of the materials, as
SC
shown in Figure 5(a). Thus, the reaction bonded β-SiAlON samples show inferior strength than the sintered ones.
M AN U
For the samples with small amount of h-BN, due to filling of the h-BN in the pores, the porosity of the composites is decreased, the flaw size is not increased significantly (as shown in Figure 5(b)). Thus, the strength of the composites is not decreased significantly with addition of a small amount of h-BN. For the samples with large amount of h-BN, h-BN particles would connect with each other, and the flaw size in the composites would be
TE D
increased (as shown in Figure 5(c)). The increase of flaw (or crack) size ascribing to the addition of h-BN can be verified by the calculation based on the Irwin equation [23]. According to Irwin’s work, the relationship among
EP
KIC, σf and the crack size (2c) in materials is as equation (4):
K IC = Y σ f c
(4)
AC C
where, Y is the dimensionless parameter, for Inglis loading geometry, Y=√π [23]. Hence, c for Inglis loading geometry can be calculated according to equation (4), and is shown in Figure 6. It can be seen that the crack size (2c) is increased with increase of h-BN content. The drilled sample of the composite with 20 wt.% h-BN is shown in Figure 7(a). It can be seen that the holes are cleanly drilled with no evidence of cracking or chipping. Due to the improved machinability of the composites, component with complicated shapes could be obtained by a machining process. A square nut was made out by machining of the composite with 30 wt.% h-BN, and the profile of the square nut is shown in Figure 7(b). The
ACCEPTED MANUSCRIPT section photo of the threaded hole is shown in Figure 7(c), and the micrograph of the screw threads is shown in Figure 7(d). It can be seen from Figure 7(c) and (d) that the screw threads are complete without obvious defects. Therefore, it can be concluded that the machinability of the composites is improved significantly and the
RI PT
complicated components could be obtained easily via a machining procedure. The machinability of the composites is commendable.
σr of the samples with different h-BN content after once and repeated thermal shock tests are shown in Figure
SC
8. Due to lots of data points in the figure, the error bars were omitted to make sure the curves can be read easily. It can be seen in Figure 8(a) that σr of the samples with h-BN content no more than 20 wt.% are decreased obviously
M AN U
when ∆T is higher than 500 °C. However, the strength degradation of the samples after thermal shock are decreased with the increase of h-BN content from 0 to 20 wt.%. σr of the samples with h-BN content of 30 wt.% and 40 wt.% is decreased not suddenly but gradually. The composite with 50 wt.% h-BN shows the slightest
TE D
strength degradation among the ∆T values in this work.
It can been seen from Figure 8(b) that σr of the samples without h-BN is decreased with the increase of thermal shock times from 1 to 9, and the σr after 9 times thermal shock is approximate 1/10 of that of the
EP
as-sintered samples. σr of the samples with 10 wt.%, 20 wt.% and 30 wt.% h-BN suffering one time thermal shock
AC C
test is decreased significantly, and then decreased slowly with the increase of thermal shock times from 1 to 9. It should be noted that the strength degradation of the samples with h-BN content in the range of 10 wt.% ~ 30 wt.% is lower than that of the samples without h-BN during the thermal shock tests. σr of the samples with h-BN content more than 30 wt.% is decreased with the thermal shock times increases from 1 to 5, and tends basically to steady with the thermal shock times increases from 5 to 9. Generally, ceramics are brittle and may show catastrophic damage when they are cooled or heated quickly. The thermal shock resistance of ceramics can be stated as the ability to avoid catastrophic damage, or to maintain
ACCEPTED MANUSCRIPT integrity, after sudden temperature variation. In this work, the strength of the composites were decreased with addition of the h-BN. For samples undergo once thermal shock, the retained strength of the composite without h-BN was higher than that of the composite with h-BN addition. It seems that the thermal shock resistance of the
RI PT
samples with h-BN addition is not improved. However, it should be noted that when the samples undergo repeated thermal shock, the retained strength of the sample without h-BN were further decreased with the thermal shock times increased from 1 to 9. It could be predicted that the higher strength of the samples without h-BN addition
SC
would be decreased to zero, which means fracture of the samples. The samples with h-BN addition would remain
M AN U
their strength for much more thermal shock times. That is the ability of the composite to avoid catastrophic damage or to maintain its integrity is improved with h-BN addition. So, the composites with h-BN addition, even their initial strength is lower, would be more reliable in the thermal shock environment, especially in the repeated thermal shock environment. Therefore, the thermal shock resistance of the as-sintered samples is improved
TE D
efficiently by addition of h-BN, and the fabricated composites with h-BN content more than 20 wt.% have excellent thermal shock resistance.
Soft phase of h-BN in the composites could act as flaws (or cracks) in the β-SiAlON matrix. When more
EP
h-BN is introduced into the composites, the size and the density of the cracks would be increased, and the cracks
AC C
propagation during the thermal shock would be quasi-static, which makes the retained strength of the composites after once or repeated thermal shock be decreased indistinctively [24, 25]. In addition, h-BN is different with the common cracks. h-BN can exhibit a series of energy absorbing mechanism under thermal stress [26-30]. Therefore, the thermal shock resistance of the composites is improved with addition of the h-BN, and the composites would present higher reliability than monomial reaction bonded β-SiAlON in the application.
4 Conclusion
β-SiAlON/h-BN composites with improved machinability and thermal shock resistance were fabricated via a
ACCEPTED MANUSCRIPT reaction bonding procedure. The relative density of the green bodies and the bulk density of the composites go through a maximum value with the h-BN content increased from 0 to 50 wt.%. Due to the increased bulk density of the composites with small amount of h-BN addition, the Vickers hardness of the composites is decreased
RI PT
without obviously strength loss, so the machinability of the composites could be improved with good strength. Fracture toughness of the composites is decreased with h-BN content increased from 0 to 50 wt.%. Drilling holes and thread tapping tests show that the composites with h-BN have commendable machinability. Due to the
SC
quasi-static crack propagation and energy absorbing mechanism of h-BN, the thermal shock resistance of the
M AN U
composites with h-BN content higher than 20 wt.% is excellent, and the retained strength of the composites is decreased slightly with the increase of temperature difference and thermal shock times.
Acknowledgement
This research was financially supported by the Scientific Research Program Funded by Xi’an University of
TE D
Architecture and Technology (No. RC1529; QN1523), the Young Doctoral Research Foundation in College of Materials and Mineral Resources, the Scientific Research Program Funded by Shaanxi Provincial Education
EP
Department (No. 14JK1393), the Program for Young Excellent Talents in Shaanxi Province (2013KJXX-50), National Natural Science Foundation of China (No. 51272203).
AC C
References
[1] K. Jack, Sialons and related nitrogen ceramics, J. Mater. Sci., 11 (1976) 1135-1158. [2] T. Ekstrom, M. Nygren, SiAlON ceramics, J. Am. Ceram. Soc., 75 (1992) 259-276. [3] F.L. Riley, Silicon nitride and related materials, J. Am. Ceram. Soc., 83 (2000) 245-265. [4] C. Raju, S. Vermab, M. Sahua, P. Jaina, S. Choudary, Silicon nitride/SiAlON ceramics-A review, Indian J. Eng. Mater. S., 8 (2001) 36-45.
ACCEPTED MANUSCRIPT [5] Y.L. Li, G.J. Qiao, Z.H. Jin, Machinable Al2O3/BN composite ceramics with strong mechanical properties, Mater. Res. Bull., 37 (2002) 1401-1409. [6] T. Kusunose, Y.H. Kim, T. Sekino, T. Matsumoto, N. Tanaka, T. Nakayama, K. Niihara, Fabrication of Al2O3/BN
RI PT
nanocomposites by chemical processing and their mechanical properties, J. Mater. Res., 20 (2005) 183-190. [7] T. Kusunose, N. Sakayanagi, T. Sekino, Y. Ando, Fabrication and characterization of aluminum nitride/boron nitride nanocomposites by carbothermal reduction and nitridation of aluminum borate powders, J Nanosci.
SC
Nanotechno., 8 (2008) 5846-5853.
M AN U
[8] H.Y. Jin, W. Wang, J.Q. Gao, G.J. Qiao, Z.H. Jin, Study of machinable AlN/BN ceramic composites, Mater. Lett., 60 (2006) 190-193.
[9] W.L. Yang, Z.Q. Shi, Y.F. Li, C.J. Zhao, G.J. Qiao, Z.H. Jin, Thermal shock and oxidation resistance of SiC composite ceramic sheath for immersion heater, J. Chin. Ceram. Soc., 42 (2014) 698-702.
TE D
[10] H. Jin, S.H. Meng, Y.W. Zhu, Y.J. Zhou, Effect of environment atmosphere on thermal shock resistance of the ZrB2-SiC-graphite composite, Mater. Des., 50 (2013) 509-514. [11] S.B. Zhou, Z. Wang, W. Zhang, Effect of graphite flake orientation on microstructure and mechanical properties of
EP
ZrB2-SiC-graphite composite, J. Alloys Compd., 485 (2009) 181-185.
AC C
[12] Z. Wang, C.Q. Hong, X.H. Zhang, X. Sun, J.C. Han, Microstructure and thermal shock behavior of ZrB2-SiC-graphite composite, Mater. Chem. Phys., 113 (2009) 338-341. [13] X.H. Zhang, Z. Wang, X. Sun, W.B. Han, C.Q. Hong, Effect of graphite flake on the mechanical properties of hot pressed ZrB2-SiC ceramics, Mater. Lett., 62 (2008) 4360-4362. [14] T. Jiang, Research progress and development of machinable ceramics composites, Mater. Rev., 26 (2012) 49-53, 65. [15] X.D. Wang, G.J. Qiao, Z.H. Jin, Fabrication of machinable silicon carbide-boron nitride ceramic nanocomposites, J.
ACCEPTED MANUSCRIPT Am. Ceram. Soc., 87 (2004) 565-570. [16] H.Y. Jin, H. Xu, G.J. Qiao, J.Q. Gao, Z.H. Jin, Study of machinable silicon carbide-boron nitride ceramic composites, Mater. Sci. Eng., A, 483-484 (2008) 214-217.
metals as heating medium, Ceram. Int., 41 (2015) 6117-6121.
RI PT
[17] Y.J. Li, H.L. Yu, H.Y. Jin, Z.Q. Shi, G.J. Qiao, Z.H. Jin, Fast heating thermal shock test for β-SiAlON with molten
[18] M. Chen, H.J. Wang, H.Y. Jin, X.D. Pan, Z.H. Jin, Transient thermal shock behavior simulation of porous silicon
SC
nitride ceramics, Ceram. Int., 42 (2016) 3130-3137.
M AN U
[19] T. Jiang, C.C. Tian, Investigation of microstructure and thermal shock resistance of the B4C/BN composites fabricated by hot-pressing process, Key Eng. Mater., 512-515 (2012) 748-752. [20] Y.J. Li, D.H. Liu, C.F. Zeng, Z.Q. Shi, Z.H. Jin, Effects of Sm2O3 content on the microstructure and mechanical properties of post-sintered reaction-bonded β-SiAlON, J. Mater. Eng. Perform., 25 (2016) 1143-1149.
TE D
[21] Y.J. Li, D.H. Liu, H. Jin, D.H. Ding, G.Q. Xiao, Z.Q. Shi, Z.H. Jin, Effect of Z values on the microstructure and mechanical properties of post-sintered reaction bonded β-SiAlON, High Temp. Mater. Proc., (DOI: 10.1515/htmp-2015-0256).
EP
[22] J.B. Quinn, I.K. Lloyd, R.N. Katz, G.D. Quinn, Ceramic "Machinability"-What does it mean, in: W.M. Kriven,
AC C
H.-a. Lin (Eds.) 27th International Cocoa Beach Conference on Advanced Ceramics and Composites, 2003, pp. 511-516.
[23] D.J. Green, An introduction to the mechanical properties of ceramics, Cambridge University Press, Cambridge, 1998.
[24] D.P.H. Hasselman, Unified theory of thermal shock fracture initiation and crack propagation in brittle ceramics, J. Am. Ceram. Soc., 52 (1969) 600-604. [25] Y.J. Li, Fabrication and properties research for β-SiAlON matrix composites with high strength and thermal shock
ACCEPTED MANUSCRIPT resistance for molten metals, Xi'an, Xi'an Jiaotong University, Ph. D. (2015). [26] B. Zhong, G.L. Zhao, X.X. Huang, L. Xia, X.H. Tang, S.C. Zhang, G.W. Wen, Microstructure and mechanical properties of ZTA/BN machinable ceramics fabricated by reactive hot pressing, J. Eur. Ceram. Soc., 35 (2015)
RI PT
641-649. [27] D.L. Cai, Z.H. Yang, X.M. Duan, B. Liang, Q. Li, D.C. Jia, Y. Zhou, A novel BN-MAS system composite ceramics with greatly improved mechanical properties prepared by low temperature hot-pressing, Mater. Sci. Eng., A, 633
SC
(2015) 194-199.
M AN U
[28] X.M. Duan, D.C. Jia, Z.L. Wu, Z. Tian, Z.H. Yang, S.J. Wang, Y. Zhou, Effect of sintering pressure on the texture of hot-press sintered hexagonal boron nitride composite ceramics, Scripta Mater., 68 (2013) 104-107. [29] Z.Q. Shi, J.P. Wang, G.J. Qiao, Z.H. Jin, Effects of weak boundary phases (WBP) on the microstructure and mechanical properties of pressureless sintered Al2O3/h-BN machinable composites, Mater. Sci. Eng., A, 492 (2008)
TE D
29-34.
[30] Y.T. Zheng, H.B. Li, T. Zhou, Microstructure and mechanical properties of h-BN-SiC ceramic composites prepared
AC C
List of table(s)
EP
by in situ combustion synthesis, Mater. Sci. Eng., A, 540 (2012) 102-106.
Table 1 Theoretical densities of raw materials for calculation
Raw materials
-3
Density/g·cm
Si
Al2O3
AlN
Sm2O3
h-BN
2.33
3.97
3.26
8.35
2.27
Table 2 ρg, ρr of the green samples, NR and ∆L of the samples before and after reaction bonding process, Pa and ρb of the as-sintered samples BN content/wt.%
ρg/g·cm-3
ρr/%
NR/%
∆L/%
Pa/%
ρb/g·cm-3
0
1.71
63.1
93.0
-4.2
22.0
2.39
5
1.80
67.3
93.6
-2.4
20.1
2.40
10
1.85
70.0
93.3
-0.6
17.6
2.41
ACCEPTED MANUSCRIPT 1.88
72.4
93.1
-0.1
17.0
2.34
20
1.87
72.7
93.5
0
16.9
2.32
30
1.86
73.8
93.0
1.1
18.4
2.19
40
1.79
72.8
93.6
1.5
21.1
2.06
50
1.74
71.9
92.9
1.8
24.8
1.92
List of figure(s)
RI PT
15
Figure 1 σf and HV10 of the as-sintered samples with different h-BN content
SC
Figure 2 KIC of the as-sintered samples with different h-BN content
Figure 3 XRD patterns of the as-sintered samples with different h-BN content
M AN U
Figure 4 Micrograph for the fracture surface of as-sintered samples with different h-BN content: (a) without h-BN, (b) 10 wt.%, (c) 20 wt.%, (d) 30 wt.%, (e) 40 wt.%, (f) 50 wt.%
Figure 5 Sketch of the h-BN distribution in the composites: (a) without h-BN, (b) less h-BN, (c) more h-BN Figure 6 Half-length (c) of the cracks for Inglis loading geometry
TE D
Figure 7 Photograph of the samples after machinability tests: (a) drilled sample with h-BN of 20 wt.%, (b) square nut made from the composite with h-BN of 30 wt.%, (c) photograph of the screw threads made from the
EP
composite with h-BN of 50 wt.%, and (d) micrograph of the screw threads
AC C
Figure 8 σr of the samples with different h-BN content after once and repeated thermal shock tests
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Density of the composites could be increased with addition of small amount of h-BN Thread tapping test showed that the composites have commendable
RI PT
machinability Repeated thermal shock showed excellent thermal shock resistance of the
AC C
EP
TE D
M AN U
SC
composites