Materials echnology ELSEVIER Joumal of Materials Processing Technology 63(1997)399-404 Injection Moulding of SiCw/ALO3 Composites R.K.Y lent of Physics and Materials Science, City University of Hong Kong, 83 Tat Chee Avenue, Hong Kong Abstract This investigation is a feasibility study of the injection moulding of alumina matrix composites reinforced with silicon carbide whiskers. The flow properties of the ceramic/polymer blend with various polymeric binder systems were studied. The dispersion of the silicon carbide whiskers in the ceramic/polymer blend from two mixing routes were compared. The ceramic weight composition and whisker length degradation of the ceramic injection mouldings with various whisker content were explored. As silicon carbide whisker may cause wearing on the processing equipment, an investigation on the samples iron content was also carried out Keywords: alumina, silicon carbide whisker and ceramic injection moulding 1. Introduction In this pilot study, alumina powder with and without SiC reinforcement are injection moulded into rectangular bar shape Although ceramic materials have the advantages of high using a polystyrene based binder system. Materials and pro- specific strength and good thermomechanical performance they cessing characterizations, which include mixing route selection are well known for low toughness and poor ductility. These prior to compounding of the polymer and the ceramic powder deficiencies can be improved by the incorporation of fiber ceramic weight composition measurement after injectior reinforcements into ceramic materials [1-5]. For example, Wei moulding, machinery abrasion and whisker length degradation and Becher [6] found that the fracture toughness of a 20 vol have been carried out before and after injection moulding silicon carbide whisker (SiCw)reinforced alumina was nearly twice that of the unreinforced alumina Ceramic injection moulding has the potential as a leading 2. Experimental details lape-forming process for ceramic materials, as it can provide igh production rate of the product with high dimensional 2.1. Materials accuracy. Before ceramic injection moulding, a polymeric binder is first blended with the ceramic powder. Subsequently, Alumina powder(Grade A152, Alcoa Chemicals Ltd. UK and SiCw(Grade SCw-1, Tateho Chemical Industries Ce slow heating, before sintering. As ceramic injection moulding Ltd, Japan) were selected as matrix and reinforcement provides many manufacturing advantages, this technology may materials respectively in this investigation. The morphology of be extended to ceramic matrix composite materials [7-101 the A152 alumina powder and the as-received silicon carbide whisker(SiCw)are shown in Figs. 武验神 ig. 1. Scanning electron micrograph of the alum 0924-0136/97/S15.00@ 1997 Elsevier Science S.A All rights reserved PIs09240136(9602654-4
ELSEVIER Journal ofMaterials Processing Technology 63 (1997) 399-404 Injection Moulding of SiCw/ Al20 3 Composites Journal of Materials Processing Technology T. L. Wong, R. K. Y. Li and C. M. L. Wu Department ofPhysics and Materials Science, City University ofHong Kong, 83 Tat Chee Avenue, Hong Kong Abstract This investigation is a feasibility study of the injection moulding of alumina matrix composites reinforced with silicon carbide whiskers. The flow properties of the ceramic/polymer blend with various polymeric binder systems were studied. The dispersion of the silicon carbide whiskers in the ceramic/polymer blend from two mixing routes were compared. The ceramic weight composition and whisker length degradation of the ceramic injection mouldings with various whisker content were explored. As silicon carbide whisker may cause wearing on the processing equipment, an investigation on the samples' iron content was also carried out. Keywords: alumina, silicon carbide whisker and ceramic injection moulding. 1. Introduction Although ceramic materials have the advantages of high specific strength and good thermomechanical performance, they are well known for low toughness and poor ductility. These deficiencies can be improved by the incorporation of fiberreinforcements into ceramic materials [1-5]. For example, Wei and Becher [6] found that the fracture toughness of a 20 vol % silicon carbide whisker (SiCw) reinforced alumina was nearly twice that of the unreinforced alumina. Ceramic injection moulding has the potential as a leading shape-forming process for ceramic materials, as it can provide high production rate of the product with high dimensional accuracy. Before ceramic injection moulding, a polymeric binder is first blended with the ceramic powder. Subsequently, the polymeric binder is removed from the moulding, usually by slow heating, before sintering. As ceramic injection moulding provides many manufacturing advantages, this technology may be extended to ceramic matrix composite materials [7-10]. Fig. 1. Scanning electron micrograph of the alumina powder. 0924-0136/97/$15.00 © 1997 Elsevier Science SA All rights reserved PII 80924-0136(96)02654-4 In this pilot study, alumina powder with and without SiCw reinforcement are injection moulded into rectangular bar shape using a polystyrene based binder system. Materials and processing characterizations, which include mixing route selection prior to compounding of the polymer and the ceramic powder, ceramic weight composition measurement after injection moulding, machinery abrasion and whisker length degradation, have been carried out before and after injection moulding. 2. Experimental Details 2.1. Materials Alumina powder (Grade A152, Alcoa Chemicals Ltd., UK) and SiCw (Grade SCW-I, Tateho Chemical Industries Co., Ltd., Japan) were selected as matrix and reinforcement materials respectively in this investigation. The morphology of the AI52 alumina powder and the as-received silicon carbide whisker (SiCw) are shown in Figs. 1 and 2 respectively. Fig. 2. Scanning electron micrograph of the as-received SiCw
T L Wong et al. /Journal of Materials Processing Technology 63(1997)399-404 Table 1 Details of the polymeric binders constituents Constituent Density(g/cm') Molecular weight (My) Grade 525 polystyrene Kaofu Chemical Ltd Riedel-de Haen Laboratory Chemicals Grade N15 polypropylene wax Eastman Chemical Co., Tennessee, USA 12000 Grade N34 polyethylene wax Eastman Chemical Co, Tennessee, USA Grade E43 chemically modified Eastman Chemical Co., Tennessee, USA m W. D. Callister, Jr, "Materials Science and Engineering An Introduction, wiley 2nd ed Not avaible from the manufacturer. Lubricants for study. The polymeric binder system consists of ()polystyrene(PS) solution was prepared by dissolving the stearic acid into as the major binder, (ii)stearic acid, and (ii)a lubricant. appropriate quantity of carbon tetrachloride. After the pre- Three types of lubricant were considered, The one which gave coated alumina/SiCw mixture was dried thoroughly at 100C,it the best flow behaviour of the ceramic/polymer blend was hosen for subsequent compounding and injection moulding container. Finally, both of the mixtures were compounded by The selection method of the lubricant for the binder system will the internal mixer for 20 minutes at 30 r.p. m. and 210C. In be described in Section 2. 2. The three types of lubricant this stage of study an internal mixer was used for the investigated were (i)a low-molecular-weight polypropylene wax impounding as it had the advantage of simple operation and 15),(i)a low-molecular-weight polyethylene wax(N34)and only a small quantity of material was needed. However, during (iii) a chemically modified low-molecular-weight polypropylene the injection moulding feeds preparation twin-screw (E43). Details of the polymeric binder system are given in compounder was used for the compounding of the Table 1 ceramic/polymer blends Alumina/whisker/polymer (A/W/P)blends with the tio of Al,O, to SiCw of 95: 5, 85: 15, and prepared from these two mixing routes. The A/W/P Three types of polymeric binder system, each with a were compression moulded into rectangular bars. The reason different lubricant incorporated, were prepared. Alumina for using compression moulding rather than injection moulding powder was then added to the three binder systems. The weight was because of the small quantities of blends prepared in this ercentage of the alumina powder: polystyrene: stearic aci part of work. Fracture surfaces of the bars were then studied bricant is 85: 10: 1.7: 3. 3. Alumina/polymer(A/P)blends by the scanning electron microscope (SEM) to examine the were prepared by compounding the pre-mixed alumina/polymer extent of the whisker dispersion in the A/W/P blends. The mixtures with an internal mixer which was attached to a mixing route which provided better whisker dispersion ability Brabender Plastic-corder 2000. The compoundings were was chosen for the mixing procedures to prepare the injection carried out at 210C. The A/P mixtures were prepared by moulding feeds tumbling the alumina powder with the polymeric binder system in a container. The aim of compounding was to achieve a 2. 4. Compounding and injection moulding thorough A/p blend so that more reliable experimental results can be obtained. A melt flow index (MFD) test was performed A/W/P mixtures with 0, 5. 15, and 30 vol on the a/P blends by a Ceast Melt Flow Indexer(Model 6542) (prepared by the choosen mixing route as described in Section at 210C with a load of 12.5 kg. The weight of the extrudates 2.3) were compounded by using a Brabender counter-rotating were measured at 5-minute intervals. The MfI values quoted twin-screw compounder. The corresponding weight percentage were the averages from four measurements. The lubricant of alumina powder, SiCw, polystyrene, stearic acid and lubricant which gave the greatest MFI to the A/P blend was selected for which were used for the A/W/P blends and their designated symbols are given in Table 2. The compounding conditions are given in Table 3. The extrudates were granulated for 2.3. Selection of mixing route It is well known that better mechanical properties of a Henposite can be obtained from a thorough fiber dispersion. Table 2 ic materials and the Weight composition of the ceramic/polymer blends polymer binders prior to compounding were studied in order to Ceramic/polymer A152 30SC compare the dispersion characteristics of the whiskers For mixing route 1, the desired weight composition of 57 74.50 63.32 alumina powder, SiCw, stearic acid and the lubricant selected SiC whisker 3,43 (as described in Section 2.2)were well mixed by shaking in a SiC whisker(vol%吗 container before compounding. For mixing route 2, the Polystyrene 10 lumina powder and SiCw were first pre-coated with stearic Lubricant* Sic, into a solution of stearic acid by continuous stirring. The *The lubricant selected in Section z 2 ative to the alumina powder acid. It was done by dispersing the alumina powder and the a The calculated approximate vol rel
400 7:L. Wong et al. / Journal of Materials Processing Technology 63 (1997) 399-404 Constituent Table 1 Details of the polymeric binders constituents Grade 525 polystyrene Stearic acid Grade NI5 polypropylene wax# Grade N34 polyethylene wax# Grade E43 chemically modified polypropylene# Source Kaofu Chemical Ltd., Riedel-deHaen Laboratory Chemicals Eastman Chemical Co., Tennessee, USA Eastman Chemical Co., Tennessee, USA Eastman Chemical Co., Tennessee, USA Density ( g/cm3 ) 1.04@ ---* 0.902 0.910 0.934 Molecular weight ( Mw ) ---* ---* 12000 6200 9100 @ From W. D. Callister, Jr., "Materials Science and Engineering An Introduction,» Wiley 2nd ed. * Not avaible from the manufacturer. # Lubricants for study. The polymeric binder system consists of (i) polystyrene (PS) as the major binder, (ii) stearic acid, and (iii) a lubricant. Three types of lubricant were considered. The one which gave the best flow behaviour of the ceramic/polymer blend was chosen for subsequent compounding and injection moulding. The selection method of the lubricant for the binder system will be described in Section 2.2. The three types of lubricant investigated were (i) a low-molecular-weight polypropylene wax (NI5), (ii) a low-molecular-weight polyethylene wax (N34) and (iii) a chemically modified low-molecular-weight polypropylene (E43). Details of the polymeric binder system are given in Table 1. 2.2. Selection of lubricant Three types of polymeric binder system, each with a different lubricant incorporated, were prepared. Alumina powder was then added to the three binder systems. The weight percentage of the alumina powder : polystyrene : stearic acid : lubricant is 85 : 10 : 1.7 : 3.3. Alumina/polymer (A/P) blends were prepared by compounding the pre-mixed alumina/polymer mixtures with an internal mixer which was attached to a Brabender Plastic-corder 2000. The compoundings were carried out at 210°C. The A/P mixtures were prepared by tumbling the alumina powder with the polymeric binder system in a container. The aim of compounding was to achieve a thorough A/P blend so that more reliable experimental results can be obtained. A melt flow index (MF!) test was performed on the A/P blends by a Ceast Melt Flow Indexer (Model 6542) at 210 °C with a load of 12.5 kg. The weight of the extrudates were measured at 5-minute intervals. The MFI values quoted were the averages from four measurements. The lubricant which gave the greatest MFI to the A/P blend was selected for subsequent work. 2.3. Selection ofmixing route It is well known that better mechanical properties of a composite can be obtained from a thorough fiber dispersion. Hence, two mixing routes for the ceramic materials and the polymer binders prior to compounding were studied in order to compare the dispersion characteristics of the whiskers. For mixing route 1, the desired weight composition of alumina powder, SiCw , stearic acid and the lubricant selected (as described in Section 2.2) were well mixed by shaking in a container before compounding. For mixing route 2, the alumina powder and SiCw were first pre-coated with stearic acid. It was done by dispersing the alumina powder and the SiCw into a solution of stearic acid by continuous stirring. The solution was prepared by dissolving the stearic acid into appropriate quantity of carbon tetrachloride. After the precoated alumina/SiCw mixture was dried thoroughly at 100°C, it was mixed with the polymeric binder by tumbling in a container. Finally, both of the mixtures were compounded by the internal mixer for 20 minutes at 30 r.p.m. and 210 °C. In this stage of study an internal mixer was used for the compounding as it had the advantage of simple operation and only a small quantity of material was needed. However, during the injection moulding feeds preparation a twin-screw compounder was used for the compounding of the ceramic/polymer blends. Alumina/whisker/polymer (A/W/P) blends with the volume ratio of Alz0 3 to SiCw of 95 : 5, 85 : 15, and 70 : 30 were prepared from these two mixing routes. The A/W/P blends were compression moulded into rectangular bars. The reason for using compression moulding rather than injection moulding was because of the small quantities of blends prepared in this part of work. Fracture surfaces of the bars were then studied by the scanning electron microscope (SEM) to examine the extent of the whisker dispersion in the A/W/P blends. The mixing route which provided better whisker dispersion ability was chosen for the mixing procedures to prepare the injection moulding feeds. 2.4. Compounding and injection moulding A/W/P mixtures with 0, 5, 15, and 30 vol% of SiCw (prepared by the choosen mixing route as described in Section 2.3) were compounded by using a Brabender counter-rotating twin-screw compounder. The corresponding weight percentage of alumina powder, SiCw , polystyrene, stearic acid and lubricant which were used for the A/W/P blends and their designated symbols are given in Table 2. The compounding conditions are given in Table 3. The extrudates were granulated for subsequent injection moulding. Table 2 Weight composition of the ceramic/polymer blends Ceramic/polymer AI52 5SC 15SC 30SC blend (wt%) Alumina 85 81.57 74.50 63.32 SiC whisker 0 3.43 10.50 21.68 SiC whisker (vol %@) 0 5 15 30 Polystyrene 10 10 10 10 Stearic acid I.7 I.7 I.7 1.7 Lubricant* 3.3 3.3 3.3 3.3 @ The calculated approximate vol% relative to the alumina powder. * The lubricant selected in Section 2.2
T.L. Wong et aL. /Journal of Materials Processing Technology 63(1997)399-404 Table 3 Such a study was carried out by using the energy Twin-screw compounding conditions spectrum analysis system attached to the Joel JSM 820 SEM Die diameter(mm) 2.7. Whisker length measurements Screw length/dia, ratio Screw speed (r.p. m feed 200210220210 The whisker length distributions of the three A/W/P blends A 6 mm die was used for the alumina/ whisker/polymer blends which after compounding and after injection moulding were measured cannot be extruded from the 3mm die The procedures for carrying out the measurements were described as follows:()small pieces of samples were cut from Rectangular bars with size70×10×6mm3 injection the corresponding sprues, followed by ashing at 600C to moulded from the four types of A/W/P blends prepared using a remove the polymeric binders; (i)the ashed sample was Hitech WellTec Industrial Equipment Ltd, HK). The injection minutes with a Thermolyne magnetic stirrer; (in) a drop of moulding conditions are given in Table 4 and the injection lution was taken and spread onto a glass slide(in order to moulded rectangular bar with sprue is shown in Fig 3 avoid errors caused by sedimentation the drop of solution was taken while the solution was still being stirred) length of 800 whiskers per sample were measured by using ar optical image analysis system. Cumulative frequency of the 3. 1. Lubricant and mixing route During the MFI measurement, the a/P blends with N15 and cm E43 lubricant could not be extruded from the heated barrel at Fig. 3. The injection moulded rectangular bar the specified MFI testing conditions stated in Section 2.2 However, the A/P blend with the N34 lubricant had an average MFI value of 0. 17 g/10 mins. Clearly, the lubricant N34 2.5. Ceramic weight composition measurements provided better flow properties to the A/P blend and consequently was selected as the lubricant for subsequent The ceramic weight composition after injection moulding compounding and injection moulding. From Table 1, it is noted Sprues were cut from the as-moulded that N34 has the lowest molecular weight. Thus, it possesses a rectangular bars and some of these were used as samples for lower viscosity than the other two types of lubricant [111 this measurement. It was assumed that the ceramic wt at the Usually, for a given ceramic content, a lower binder viscosity sprue and that at the body of the rectangular bar were the same ill give a lower ceramic/polymer blend viscosity. Therefore Five sprues from each type of rectangular bars were ashed at there is no doubt that N34 provides better flow properties 600C in a Nabertherm Model L3/17S programable furnace to among the a/P blends remove the polymeric binders. The weight of the specimens Fig. 4 shows the SEM micrographs of the fracture surface of and after ashing were measured by a Ohaus ToP-pan the compression moulded bars prepared by mixing route I with 5, 15 and 30 vol SiC w respectively. Examination on the micrographs revealed that the whiskers were 2. 6. Iron content analysis homogeneously inside the moulded bars. The SiCwconcen- tration are low in some regions (as shown by the photographs SiCw is a very abrasive material, and can cause we on the left-hand-side of Figs. 4a-c), but are exceptionally high the barrels and of the compounder and the in some other regions(see photographs on the right-hand-side of moulding machine. A study of the composition of the Figs. 4a-c). However, for the compression moulded bars content in the compounded blends and the injection prepared by mixing route 2 a uniform bars can indicate the extent of wear in the barrels and Injection moulding conditions A/W/P bler Screw Length/dia, ratio Barrel temperature(feed-nozzle, C) 200-210200 200-210-200 200-210-200 ould Temperature (C) Cycle time(s
TL. Wong et al. I Journal ofMaterials Processing Technology 63 (1997) 399-404 401 Table 3 Twin-screw compounding conditions Screw diameter (mm) 41.8 Die diameter (mm) 3* Screw length/dia. ratio 7 Screw speed (r.p.m.) 30 Barrel temperature (feed to die, 0c) 200-210-220-210 * A 6 mm die was used for the alumina/whisker/polymer blends which cannot be extruded from the 3mm die. Rectangular bars with size 70x IOx6 mm3 were injection moulded from the four types of A/W/P blends prepared using a reciprocating-screw machine (Cosmo TTI-220/80 Hitech, WellTec Industrial Equipment Ltd., HK). The injection moulding conditions are given in Table 4 and the injection moulded rectangular bar with sprue is shown in Fig. 3. c:."" Fig. 3. The injection moulded rectangular bar. 2.5. Ceramic weight composition measurements The ceramic weight composition after injection moulding was measured. Sprues were cut from the as-moulded rectangular bars and some of these were used as samples for this measurement. It was assumed that the ceramic wt% at the sprue and that at the body of the rectangular bar were the same. Five sprues from each type of rectangular bars were ashed at 600 °C in a Nabertherm Model L3117S programable furnace to remove the polymeric binders. The weight of the specimens before and after ashing were measured by a Ohaus Top-pan electronic balance. 2.6. Iron content analysis SiCw is a very abrasive material, and can cause wearing on the barrels and screws of the compounder and the injection moulding machine. A study of the composition of the iron (Fe) content in the compounded blends and the injection moulded bars can indicate the extent of wear in the barrels and screws. Such a study was carried out by using the energy dispersive spectrum analysis system attached to the Joel JSM 820 SEM. 2. 7. Whisker length measurements The whisker length distributions of the three A/WIP blends after compounding and after injection moulding were measured. The procedures for carrying out the measurements were described as follows: (i) small pieces of samples were cut from the corresponding sprues, followed by ashing at 600 °C to remove the polymeric binders; (ii) the ashed sample was dispersed in a beaker of alcohol by continuous stirring for 20 minutes with a Thermolyne magnetic stirrer; (iii) a drop of this solution was taken and spread onto a glass slide (in order to avoid errors caused by sedimentation the drop of solution was taken while the solution was still being stirred); and (iv) the length of 800 whiskers per sample were measured by using an optical image analysis system. Cumulative frequency of the whisker length were then plotted. 3. Results and Discussion 3.1. Lubricant and mixing route During the MFI measurement, the AlP blends with N15 and E43 lubricant could not be extruded from the heated barrel at the specified MFI testing conditions stated in Section 2.2. However, the AlP blend with the N34 lubricant had an average MFI value of 0.17 g/1O mins. Clearly, the lubricant N34 provided better flow properties to the A/P blend and consequently was selected as the lubricant for subsequent compounding and injection moulding. From Table 1, it is noted that N34 has the lowest molecular weight. Thus, it possesses a lower viscosity than the other two types of lubricant [11]. Usually, for a given ceramic content, a lower binder viscosity will give a lower ceramic/polymer blend viscosity. Therefore, there is no doubt that N34 provides better flow properties among the A/P blends. Fig. 4 shows the SEM micrographs of the fracture surface of the compression moulded bars prepared by mixing route 1 with 5, 15 and 30 vol% SiCw respectively. Examination on the micrographs revealed that the whiskers were not distributed homogeneously inside the moulded bars. The SiCw concentration are low in some regions (as shown by the photographs on the left-hand-side of Figs. 4a-c), but are exceptionally high in some other regions (see photographs on the right-hand-side of Figs. 4a-c). However, for the compression moulded bars prepared by mixing route 2, a uniform dispersion of SiCw Table 4 Injection moulding conditions A/W/Pblend Screw diameter (mm) Screw Length/dia. ratio Barrel temperature (feed - nozzle, °C) Mould Temperature ("C) Injection pressure (MPa) Hold pressure (MPa) Hold time (s) Cycle Time (s) A152 33 18 200-210-200 20 90 90 5 20 5SC 33 18 200-210-200 20 100 100 5 20 15SC 33 18 200-210-200 20 130 130 5 20 30SC 33 18 200-210-200 20 130 130 5 20
T.. Wong et al. /Journal of Materials Processing Technology 63(1997)399-404 k Fig. 4. Scanning electron micrographs of fracture surfaces of the compression moulded bars prepared by mixing route 1, (a)5 vol% SiCw,(b)15 vol% SiCw, and (c)30 vol% SiC can be seen in all samples. Moreover, further investigation on Table 5. The average weight percentage of ceramic constituents the fracture surfaces which had the same whisker volume in the injection moulded bars with 0, 5, 15, and 30 vol SiCw fraction showed that there was no noticeable differences in are 83.83, 83.91, 84.94 and 85.30 wt% respectively. The whisker content. It is clear that mixing route 2 provided largest deviation from the expected value 85 wt% of ceramic better whisker dispersion than that of mixing route 1 constituents is 1. 17 wt % So, the weight composition of the The wetting and dispersion of the powder into the binder can ceramic constituents can be maintained during the whole be improved by reducing the contact angle and hence lowering fabrication process which includes twin-screw compounding and the surface energy of the powder/binder interface. Stearic acid injection moulding is a typical surfactant or wetting agent for such purpose. In mixing route 2, stearic acid was used to treat the ceramic 3. 3. Wear of processing equipment wder/whisker mixture. As a result, mixing route 2 provide a better whiskers dispersion structure than that in mixing route The variation of Fe content with Sic content in the I and hence it was chosen for the mixing route prior to the ceramics/polymer blends are shown in Fig. 5. The Fe content compounding of the ceramic/polymer mixtures present in the ceramics/polymer blends after compounding and after injection moulding are shown in the same graph for amic weight lost mples after compounding(curve A in Fig 5)were found to have a higher Fe content in the A/p blend (i.e The weight percentages of the ceramic constituents(alumina no SiCw present)than expected. This is because in extruding and SiCw)in each sample from the ashing results are shown in the a/P blend, a 3 mm diameter die was used. However, the
402 T.£. Wong et al. I Journal of Materials Processing Technology 63 (1997) 399-404 (a) (b) (c) Fig. 4. Scanning electron micrographs of fracture surfaces of the compression moulded bars prepared by mixing route I, (a) 5 vol% SiCw ' (b) 15 vol% SiCw ' and (c) 30 vol% SiCw • can be seen in all samples. Moreover, further investigation on the fracture surfaces which had the same whisker volume fraction showed that there was no noticeable differences in whisker content. It is clear that mixing route 2 provided a better whisker dispersion than that of mixing route I. The wetting and dispersion of the powder into the binder can be improved by reducing the contact angle and hence lowering the surface energy of the powder/binder interface. Stearic acid is a typical surfactant or wetting agent for such purpose. In mixing route 2, stearic acid was used to treat the ceramic powder/whisker mixture. As a result, mixing route 2 provided a better whiskers dispersion structure than that in mixing route I and hence it was chosen for the mixing route prior to the compounding of the ceramic/polymer mixtures. 3.2. Ceramic weight lost The weight percentages of the ceramic constituents (alumina and SiCw) in each sample from the ashing results are shown in Table 5. The average weight percentage of ceramic constituents in the injection moulded bars with 0, 5, 15, and 30 vol% SiCw are 83.83, 83.91, 84.94 and 85.30 wt% respectively. The largest deviation from the expected value 85 wt% of ceramic constituents is 1.17 wt%. So, the weight composition of the ceramic constituents can be maintained during the whole fabrication process which includes twin-screw compounding and injection moulding. 3.3. Wear ojprocessing equipment The variation of Fe content with SiCw content in the ceramics/polymer blends are shown in Fig. 5. The Fe content present in the ceramics/polymer blends after compounding and after injection moulding are shown in the same graph for comparison. The samples after compounding (curve A in Fig. 5) were found to have a higher Fe content in the A/P blend (i.e. no SiCw present) than expected. This is because in extruding the A/P blend, a 3 rom diameter die was used. However, the
TL Wong et aL. /Journal of Materials Processing Technology 63(1997)399-404 A/w/p blends could not be extruded with the 3 mm diameter die, so a 6 mm diameter die had to be used. The use of a die with a small diameter for the a/p blend increases its time inside the heated barrel of the compounder. Thus, more Fe will be taken up from the barrel Table 5 Wt% of ceramic constituents of injection moulded samples 838483.6584.8585.34 85 84, 84.94 (0.08)* vol% SiCw, (B)1 *Standard deviation mounding. D)is the length distribution of the as-received whiskers Usually, the loss of Fe from the barrel will increase with content. However, it seems that this effect do not apply to the compounded A/w/P blends(see Fig. 5 curve A), since the Fe content did not increase significantly with whisker content. It might be due to the fact that the screw and barrel of the compounder had been significantly hardened by the manufacturer. For the A/w/p blends after injection moulding the tendency of increasing Fe content with whisker content can be observed since the screw and the barrel were prepared fo plastic injection moulding and so were not hardened to significant extent. 3. 4. Whisker length degrad The whisker length distribution of the a/w/p blends with 5 15 and 30 vol of SiCw after compounding are shown in Fig. 6. The longest whisker length measured from the as-received SiC Fig. 7. Variation of whisker length distribution of ceramic is 70 mm, while in the twin-screw compounded A/w/P blend polymer blend with(A)5 vol% SiCw,(B)15 vol% SiCw and(C) the longest measured whisker length is 40 mm. The mean 30 vol SiCw after injection moulding. D)is the length whisker length of the as-received SiCw was 13.8 mm. distribution of the as-received whiskers However, the mean whisker length of the compounded A/W/P blends ranged from 4 to 5 mm. Obviously, the overall whisker length of the A/w/p blends after compounding were seriousl decreased. Also, it seems that the degree of degradation of the whisker length of the compounded A/W/P blends did not Fig. 7 shows the whisker length distribution of the A/W/P increase with the whiskers content, as all the compounded blends after injection moulding with mean whisker between 3.5 A/w/P blends had nearly the same mean whisker length and 4.5 There is no significant damage on the me whisker length due to an increase in whisker content. Fig. 8 nows the optical micrographs for the as-received SiCw, A/W/P blends with 15 vol% SiCw after compounding and after injection moulding. It can be seen from the micrographs that the whisker lengths of the as-received SiCw are definitely longer than those of the whiskers after compounding or after injection moulding Moreover, the whiskers after compounding and after injection moulding have similar lengths. This observation showed that the decrease in whisker length was mainly due According to Frangen et al [12], as the whiskers are injection moulded, they interact with the alumina powder, the polymer whisker content/volz melt or the whiskers themselves. Such interaction may lead to bending force acting perpendicular to the whisker length. The whisker will be broken at some critical bending force whick Fig. 5. Variation of Fe content in the ceramic/polymer blends depends on the length and diameter of the whisker. During with whisker content(A)after compounding and (B)after compounding(or injection moulding), the whiskers continue to break until the bending force acting on the whiskers fall below
T.L. Wong et al.!Journal of Materials Processing Technology 63 (1997) 399-404 403 A/WIP blends could not be extruded with the 3 mm diameter die, so a 6 mrn diameter die had to be used. The use of a die with a small diameter for the AlP blend increases its time inside the heated barrel of the compounder. Thus, more Fe will be taken up from the barrel. Table 5 Wt% of ceramic constituents of injection moulded samples Ceramic blend A152 5SC 15SC 30SC Sample 1 83.95 83.50 84.97 85.57 Sample 2 83.84 83.65 84.85 85.34 Sample 3 83.80 84.38 84.97 85.36 Sample 4 83.83 83.57 85.12 85.16 Sample 5 83.72 84.44 84.78 85.08 Mean 83.83 83.91 84.94 85.30 (0.08)* (0.46)* (0.13)* (0.19)* *Standard deviation Usually, the loss of Fe from the barrel will increase with increasing SiCw content. However, it seems that this effect does not apply to the compounded A/W/P blends (see Fig. 5 curve A), since the Fe content did not increase significantly with whisker content. It might be due to the fact that the screw and barrel of the compounder had been significantly hardened by the manufacturer. For the A/W/P blends after injection moulding, the tendency of increasing Fe content with whisker content can be observed since the screw and the barrel were prepared for plastic injection moulding and so were not hardened to a significant extent. 3.4. Whisker length degradation The whisker length distribution of the A/W/P blends with 5, 15 and 30 vol % of SiCw after compounding are shown in Fig. 6. The longest whisker length measured from the as-received SiCw is 70 mm, while in the twin-screw compounded A/WIP blends the longest measured whisker length is 40 mm. The mean whisker length of the as-received SiCw was 13.8 mm. However, the mean whisker length of the compounded A/WIP blends ranged from 4 to 5 mm. Obviously, the overall whisker length of the A/WIP blends after compounding were seriously decreased. Also, it seems that the degree of degradation of the whisker length of the compounded A/WIP blends did not increase with the whiskers content, as all the compounded A/W/P blends had nearly the same mean whisker length. 3 '" I'~=~ o L....:0-~-:5-~.,..10=--~-:"15=--~--=2::-0~--:2'="5~--:3'::-O whisker content / vol7- Fig. 5. Variation of Fe content in the ceramic/polymer blends with whisker content (A) after compounding and (B) after injection moulding. 100 '" 90 """- v 80 0 c 70 v " 60 0- 50 v 40 0- 2 30 E" 20 0" 10 , 0 20 30 40 50 60 70 80 whisker length / /-lm Fig. 6. Variation of whisker length distribution of aluminal whisker/polymer blend with (A) 5 vol% SiCw , (B) 15 vol% SiCw and (C) 30 vol % SiCw after compounding. (D) is the length distribution of the as-received whiskers. '""""- v 80 0 c 70 0 v " 60 0- J:' 50 v 40 0- 0 30 E" 20 U" 10 10 20 30 40 50 60 70 80 whisker length / /-lm Fig. 7. Variation of whisker length distribution of ceramicl polymer blend with (A) 5 vol % SiCw, (B) 15 vol % SiCw and (C) 30 vol % SiCw after injection moulding. (D) is the length distribution of the as-received whiskers. Fig. 7 shows the whisker length distribution of the A/W/P blends after injection moulding with mean whisker between 3.5 and 4.5 !-lm. There is no significant damage on the mean whisker length due to an increase in whisker content. Fig. 8 shows the optical micrographs for the as-received SiCw , A/WIP blends with 15 vol % SiCw after compounding and after injection moulding. It can be seen from the micrographs that the whisker lengths of the as-received SiCw are definitely longer than those of the whiskers after compounding or after injection moulding. Moreover, the whiskers after compounding and after injection moulding have similar lengths. This observation showed that the decrease in whisker length was mainly due to the compounding process. According to Frangen et al [12], as the whiskers are injection moulded, they interact with the alumina powder, the polymer melt or the whiskers themselves. Such interaction may lead to a bending force acting perpendicular to the whisker length. The whisker will be broken at some critical bending force which depends on the length and diameter of the whisker. During compounding (or injection moulding), the whiskers continue to break until the bending force acting on the whiskers fall below
TL. Wong et aL/ Journal of Materials Processing Technology 63(1997)399-404 injection moulded A/w/P blends, further significant degradation of the whisker length was not noticed. It may be due to the fact that the injection moulding process could not provide sufficient bending moment to break the shortened whiskers 4. Conclusions This investigation has shown that a silicon carbide whisker reinforced alumina matrix composites with SiCw content up to 30 vol can be injection moulded. Also, a polymeric binder (a) system with polystyrene as the major binder and stearic acid the wetting agent can give good flow properties of the ceramic/polymer blend when a low-molecular polyethylene lubricant is used. Moreover, the dispersio whiskers in the alumina/ polymer blends is better if both powder and Sic whiskers are pre-coated with the stearic acid prior to compounding The iron loss from the screw and barrel incre increasing whisker content. The length of the as-received Sic twIn-screw compounding, but there was no further decrease in the whisker length after the compounded blends were being injection moulded. Furthermore, the degree of degradation of the whiskers did not increase with whisker content (b) This work was supported by the City University of Hong Kong Strategic Research Grant through grant number 700185 The authors would like to thank Alcoa Chemicals Ltd. UK for their kind donation of alumina powder for this investigation The lubricants used for the study were kindly donated by Eastman Chemical Co., Tennessee. USA References [1] R. W. Davidge, Composites, 18(1987)92 Fig. 8. Whiskers morphology of the(a)as-received SiCw,(b [2 I. W. Donald and P. W. Memillan, J. Mat. Sci., 11 alumina/ whisker blend with 15 vol SiC. after compounding and (c)alumina/whisker/polymer blend with 15 (1976)949 [3] K. P. Gadkaree, J. Mat Sci., 26(1991)4845 vol% SiCw after injection moulding [4 Satoshi lio, Matanabe Watanabe, Masaru Matsubara and Yasushi Matsuo, J. Am. Ceram Soc., 72 (1989)1880. [51 G. Y. Lin, T. C. Lei, Y. Zhou and s.x. Wang, J. Mat the critical bending force of breakage or it has beer Sci.,28(1993)2745 compounded(or injection moulded) from the compounder(or [6]G. C. Wei and P. F. Becher. Am. ceram. Soc. BulL.64 injection moulding machine). For the compounding of A/w/P (1985)298 mixture, a relatively high percentage of solid particle exists and [7] T. Kandori,S. Kobayashi, S. Wada and O. Kamigaito, J. the melt viscosity is very high. The significant increase in the melt viscosity enhances the bending force acting on the whiskers [8]SJ. Stedman, J.R. G. Evans and J Woodthorpe,J.Mar and hence increase the whisker damage result the Sci.,25(1990)1025 compounded A/w/P blends suffer severe whisker damage [9] S. J. Stedman, J.R. G. Evans and J. Woodthorpe, J. Mat Moreover, the interaction between the whisker and the alumina Sci.,25(1990)1833 powder seems to dominate the whisker breakage process, so that [0 Inlin Tsao and S. C. Danforth, J. Am. Ceram. Soc., 76 he effect of increased whisker degradation with increasi whisker content cannot be observed from the injection moulded [II]N. G. McCrum, Principles of Polymer Engineering A/w/P blends(see Figs. 6 and 7). As the whiskers are severely Oxford University Press, (1988)280 degraded after compounding, they possess a much shorter [12] B. Franzen, C, Klason, J. Kubot and T. Kitano length which can withstand higher bending force. For the Composites, 20( 1989)65
404 TL. Wong et al.IJournal ofMaterials Processing Technology 63 (1997) 399-404 (a) (b) (c) 20 !-lm 10 !-lm injection moulded A/W/P blends, further significant degradation of the whisker length was not noticed. It may be due to the fact that the injection moulding process could not provide sufficient bending moment to break the shortened whiskers. 4. Conclusions This investigation has shown that a silicon carbide whisker reinforced alumina matrix composites with SiCw content up to 30 vol % can be injection moulded. Also, a polymeric binder system with polystyrene as the major binder and stearic acid as the wetting agent can give good flow properties of the ceramic/polymer blend when a low-molecular-weight polyethylene lubricant is used. Moreover, the dispersion of the whiskers in the alumina/polymer blends is better if both alumina powder and SiC whiskers are pre-coated with the stearic acid prior to compounding. The iron loss from the screw and barrel increased with increasing whisker content. The length of the as-received SiC whiskers were seriously decreased after twin-screw compounding, but there was no further decrease in the whisker length after the compounded blends were being injection moulded. Furthermore, the degree of degradation of the whiskers did not increase with whisker content. Acknowledgements This work was supported by the City University of Hong Kong Strategic Research Grant through grant number 700185. The authors would like to thank Alcoa Chemicals Ltd., UK for their kind donation of alumina powder for this investigation. The lubricants used for the study were kindly donated by Eastman Chemical Co., Tennessee, USA. References Fig. 8. Whiskers morphology of the (a) as-received SiCw , (b) alumina/whisker/polymer blend with 15 vol % SiCw after compounding and (c) alumina/whisker/polymer blend with 15 vol% SiCw after injection moulding. the critical bending force of breakage or it has been compounded (or injection moulded) from the compounder (or injection moulding machine). For the compounding of A/W/P mixture, a relatively high percentage of solid particle exists and the melt viscosity is very high. The significant increase in the melt viscosity enhances the bending force acting on the whiskers and hence increase the whisker damage. As a result, the compounded A/W/P blends suffer severe whisker damage. Moreover, the interaction between the whisker and the alumina powder seems to dominate the whisker breakage process, so that the effect of increased whisker degradation with increasing whisker content cannot be observed from the injection moulded A/W/P blends (see Figs. 6 and 7). As the whiskers are severely degraded after compounding, they possess a much shorter length which can withstand higher bending force. For the [1] R. W. Davidge, Composites, 18 (1987) 92. [2] 1. W. Donald and P. W. Memillan, J. Mat. Sci., 11 (1976) 949. [3] K. P. Gadkaree, J. Mat. Sci., 26 (1991) 4845. [4] Satoshi lio, Matanabe Watanabe, Masaru Matsubara and Yasushi Matsuo, J. Am. Ceram. Soc., 72 (1989) 1880. [5] G. Y. Lin, T. C. Lei, Y. Zhou and S. X. Wang, J. Mat. Sci., 28 (1993) 2745. [6] G. C. Wei and P. F. Becher, Am. Ceram. Soc. Bull., 64 (1985) 298. [7] T. Kandori, S. Kobayashi, S. Wada and O. Kamigaito, J. Mat. Sci. Letter., 6 (1987) 1356. [8] S. J. Stedman, J. R. G. Evans and J. Woodthorpe, J. Mat. Sci., 25 (1990) 1025. [9] S.1. Stedman, J. R. G. Evans and J. Woodthorpe, J. Mat. Sci., 25 (1990) 1833. [10] Inlin Tsao and S. C. Danforth, J. Am. Ceram. Soc., 76 (1993) 2977. [11] N. G. McCrum, Principles of Polymer Engineering, Oxford University Press, (1988) 280. [12] B. Franzen, C. Klason, J. Kubot and T. Kitano, Composites, 20 (1989) 65