Thakur Sudesh Kumar Raunija, Rajeev K. Gautam, Sharad Chandra Sharma, Anil Verma

(1. Carbon and Ceramics Laboratory, Materials and Mechanical Entity, Vikram Sarabhai Space Centre,Indian Space Research Organisation, Thiruvananthapuram 695022, Kerala, India;2. Sustainable Environergy Research Lab (SERL), Department of Chemical Engineering,Indian Institute of Technology Delhi, Hauz-Khas, New Delhi 110016, India;3. Materials and Metallurgy Group, Materials and Mechanical Entity, Vikram Sarabhai Space Centre,Indian Space Research Organisation, Thiruvananthapuram 695022, Kerala, India)

Abstract: Conventional ways of fabricating C/C composites are unable to achieve wetting in the core of the carbon fiber tow by molten pitch because a single tow has thousands of fibers. Pitch stabilization after infiltration into the carbon fiber preform results in local oozing of the pitch and a non-uniform microstructure due to non-uniform pitch stabilization in the preform. The objective of this research was to devise a method by which complete wetting of fibers in the tow with mesophase pitch is attained by mixing stabilized pitch and chopped and scattered carbon fibers, followed by hot-pressing and carbonization. The influence of process parameters on the carbon yield of the stabilized mesophase pitch is evaluated. The carbon yield divided by processing time and the carbon yield multiplied by the ratio of the apparent density to real density of the C/C composites are used to optimize process parameters. Results show that complete wetting of fibers with mesophase pitch is attained by this method. The carbon yield of the stabilized mesophase pitch decreases with heating rate and the mass ratio of stabilized pitch to carbon fibers, and increases with hot-pressing pressure. At high pressures and high mass ratios, local oozing of the stabilized pitch occurs. The optimum C/C composite was prepared under a hot-pressing pressure of 15 MPa, a heating rate of 0.2 ℃/min and a mass ratio of stabilized pitch to carbon fibers of 1∶1.

Key words: Petroleum pitch; Mesophase pitch; Short carbon fibers; Hot-pressing; Yield

1 Introduction

In space and aerospace applications, almost all types of advanced composite materials are used, but carbon fiber-reinforced carbon matrix composite has its specific application. Carbon fiber-reinforced carbon matrix composite, more specifically, called as carbon/carbon (C/C) composite was first synthesized in 1960 s[1]. It was first used for nose cones in the rocket reentry. Thereafter, the C/C composite was diversified to different components such as wing leading edges, rocket nozzles, exhaust cones, disc brakes in supersonic aircrafts, dies for hot molding metal powders, tools for the superplastic forging of titanium, connecting rods and pistons, turbine rotors, heat exchanger tubes, car brake discs, biomedical implants, molecular sieves, fasteners, fixtures, and fuel cells[2-6]. Among these applications, the requirement varies from high thermal and electrical conductivities to high mechanical strength.

Chemical vapor infiltration (CVI) derived carbon matrix is used for the fabrication of C/C composite in applications where very high mechanical properties are needed. However, pitch derived carbon matrix is used over CVI derived carbon matrix to fabricate C/C composite for the applications where high conductivity is needed[7]. The use of pitch as matrix precursor makes the process rapid due to its high yield. However, high yield from pitch can be obtained on account of slow heating and high pressure processing[2]. High yield of pitch matrix precursors can be assured by oxidative stabilization of pitches prior to pyrolysis and carbonization[8-14]. Besides that, the oxidative stabilization avoids the issues of[14]exudation/oozing of the pitches during pyrolysis and carbonization[9]. The yield of the pitch increases with the increase in oxygen mass gain. However, the increased mass gain of pitch beyond a certain limit makes them lose their fluidity permanently and further oxygen mass gain starts carbon burn off[10,13,15]. The influence of all these parameters on deriving carbon matrix from the pitches has been studied extensively through conventional route like hot isostatic pressure impregnation carbonization (HIPIC)[16]; however, this route is highly time consuming and costly. Hence, researchers have alternatively derived carbon matrix, in the preform of carbon fibers and carbon fabrics, from the pitches by unidirectional pressing through hot-pressing route[17]. Nevertheless, these routes are unable to obtain the filament level wetting by molten pitch in the core of the carbon fiber tow as a single tow has thousands of filaments. Hence, the methods, which can provide wetting and matrix distribution on filament level opens up the avenues to improve the properties of the C/C composite.

Owing to the above challenges, a rapid and cost-effective processing of short carbon fiber reinforced C/C composite through ISRO (Indian Space Research Organization) Medium Density Slurry Moulded C/C Composite (IMDSC) process using stabilized mesophase pitch (SP) as primary matrix precursor was demonstrated by our group[5]. In this method, hot-pressing operation was the key to fabricate C/C composite. Various research groups have worked on the parameters of impregnation for a small sample of carbon fiber preforms with pitches and then oxidative and thermal stabilization followed by carbonization on lab scale[8,10,12,18]. Further, we studied the influence of various hot-pressing parameters on the morphology, density and mechanical properties. However, the influence of real time process parameters on the yield of mesophase pitch in the fabrication of bigger C/C composite artifacts through the hot-pressing route was not reported[4]. It is important to note that unlike conventional routes to fabricate C/C composites using pitch matrix precursors, the infiltration with pitches is done after the stabilization in IMDSC. Therefore, the aim of this research work is to evaluate the influence of heating rate, pressure, and the ratio of matrix to reinforcement on yield, yield rate, and yield impact of stabilized mesophase pitch in short carbon fiber reinforced randomly oriented C/C composite fabricated through the novel IMDSC process. Moreover, the influence of matrix precursor and oxygen interaction with reinforcement on filament level as compared to the conventional process is reported. It is to be noted that the yield rate and yield impact were not reported in the literature.

2 Materials and methods

2.1 Raw materials

In this study, in-house synthesized and stabilized mesophase pitch (SP) was used as matrix precursor. Commercially available carbon fibers were chosen for reinforcing the composite. Polyacrylonitrile (PAN) based T-800 grade and pitch-based P-75 grade carbon fibers were used as hybrid reinforcement. Hybrid form was used for the better combination of mechanical, thermal and electrical properties in the composite[5]. The fibers were chopped into discrete lengths. The ratio between the aspect ratios of PAN and pitch-based carbon fibers was kept 2.5 to better utilize both the mechanical properties of PAN-based carbon fibers and densification characteristics of pitch-based carbon fibers. On the other hand, the weight ratio of pitch to PAN-based carbon fibers was taken as 2.33, which avails optimizing the mechanical properties and densification behavior[5].

2.2 Methods

2.2.1 Pitch preparation and stabilization

The mesophase pitch was prepared from commercially available raw petroleum pitch (RP) through thermotropic route. The mesophase was named as “isroaniso” matrix precursor from industrial application point of view. In the preparation sequence of isroaniso matrix precursor, first of all, the raw pitch (RP) lumps as received were crushed into powder with an average particle size (volume weighted mean) of ～15 μm in order to get uniform melting during heat treatment. Thereafter, the crushed RP was heat treated in an inert atmosphere at 350 ℃ for 7 h to obtain inert treated pitch (ITP). The entire operation of heat treatment was carried out in a cylindrical vessel made up of steel. The agitator was kept in the vessel along with sealed lid. The provision for thermocouple, inlet and outlet gas was made in the vessel through sealed lid. The ITP was extracted from the reactor in the form of solid lumps. These lumps were powdered and further refined through vacuum distillation at 400 ℃ for 2 h. The operation of vacuum distillation was carried out in a glass distillation reactor of size around Ø 200 mm. The vacuum distilled pitch was extracted from the vacuum distillator in the form of lumps, which were further crushed into powder with an average particle size of ～15 μm. The fine control of particle sizes was needed to ensure uniform oxygen diffusion during stabilization and uniform distribution with carbon fibers during slurry preparation and moulding. The vacuum treated pitch (VTP) powder, also known as mesophase pitch, thus obtained was stabilized in air to obtain stabilized pitch (SP). The oxygen mass gain of SP was maintained around 0.8 wt%.

2.2.2 Preform preparation

In order to prepare preforms, the chopped carbon fibers were first exfoliated and then SP powder was added into distilled water with agitation to get slurry[5]. The slurry was then agitated for 1 h at 120 rpm. The slurry preparation and moulding was done in an indigenously made slurry moulding equipment. Uniform slurry was then vacuum moulded over a sieve screen in a rectangular die of cross-section 70 mm×50 mm. All the preforms were prepared with the similar conditions. The preforms thus obtained were hot-pressed in an indigenously fabricated hot-press. The arrangement of compact in the die system and the load direction are shown in Fig.1.The hot-pressing of the compacts was carried out by varying the hot-pressing parameters (pressure and heating rate).

Fig. 1 Die system and loading arrangement.

2.2.3 Carbonization

Carbonization of the hot-pressed compacts was carried out in a tubular furnace at a heating rate of 1 ℃/min and kept at 1 050 ℃ for 1 h. The inert atmosphere inside the furnace was maintained by purging argon gas at a rate of 5 L/min. All the parameters of carbonization were kept constant for all of the samples.

2.3 Characterization

2.3.1 Physicochemical characterization

Softening point of pitch was determined by a softening point analyzer (Mettler-Toledo, FP900). CHN analysis of pitch and C/C composite samples was carried out using a CHNS analyzer (Perkin Elmer 2400 Series II). Thermogravimetric analysis (TGA) of pitch sample was carried out in a thermal analyzer (TA Instruments) by heating the samples to 1 100 ℃ at a rate of 10 ℃/min in a nitrogen flow of 150 mL/min at atmospheric pressure. The main purpose of doing TGA of the pitch samples was to gain the preliminary knowledge of the pyrolysis behavior of pitches and yield thereafter. After the pyrolysis behavior of various pitches at fixed higher heating rate (10 ℃/min) was known, the heating rates were lowered down and varied in the actual system of C/C composite fabrication. The bulk density of all the compacts after hot-pressing and carbonization was calculated from their mass and volume.

2.3.2 Yield of SP

For the calculation of exact yield of SP in actual C/C composite system, the weight loss of carbon fibers was first calculated separately through TGA at a maximum temperature of carbonization at which compacts were fabricated.

It is to be noted that the weight loss of carbon fibers is considered for a better accuracy though the weight loss of carbon fibers is low, in general. Moreover, the powder SP experiences weight loss during vacuum moulding due to slippage of powde SP with water in slurry making. The weight loss of powder SP during vacuum moulding (wmpv) was calculated and found to be around 0.5 wt% of SP taken initially for the slurry preparation and in all the calculations (wmpv) was considered as 0.5 wt%. Thus, taking account of all these losses, the yield of SP is defined as following.

(1)

(2)

where,

yhp= yield after hot-pressing (%),

yc= yield after carbonization (%),

wh= weight of the compact after hot pressing,

wcfh= weight loss of carbon fibers at hot-pressing temperature (650 ℃),

wfh= fiber weight in the compact after hot-pressing,

wmp= weight of SP taken initially,

wmpv= weight loss of SP during vacuum moulding,

wc= weight of the compact after carbonization,

wcfc= weight loss of carbon fibers at carbonization temperature (1 050 ℃),

wfc= fiber weight in the compact after carbonization.

It may be noted that Li et al[18]used the similar method to evaluate the yield of SP but they have not considered the weight loss of the carbon fibers. Some more methodologies for the calculation of yield of pitches have been used by various researchers on the line close to shown above[13,14]. However, the yield rate and yield impact are being reported first time.

2.3.3 Yield rate and yield impact

The yield rate and yield impact of the SP are computed as follows.

yr=yc/pt

(3)

yi=(yc×ρ0)/ρa

(4)

where,

yr= yield rate (%/h)

yi= yield impact (%)

yc= carbonized yield (%),

pt= total process time (h),

ρ0= obtained density,

ρa= maximum possible density of the compact, which is calculated, based upon the practically possible density of the carbon matrix obtained from matrix precursor (SP) at carbonization temperature and quoted density of carbon fibers in supplied MSDS.

The percentage errors for the yield, yield rate, and yield impact were found to be less than 5%.

3 Results and discussion

Fig. 2 shows the TGA plots of T-800, P-75, RP, ITP, VTP, and SP. It can be seen that the carbon fibers (T-800 and P-75) have very low weight loss up to 1 100 ℃, which is due to the fact that the carbon fibers experience the temperature of around 1 100 ℃ during their manufacturing process. The weight loss of carbon fibers happens due to the evaporation of hydrocarbons and H2upon heating at 1 100 ℃, which is basically a reheating[19,20]. The weight loss of T-800 and P-75 carbon fibers at carbonization temperature (1 050 ℃) is found to be around 1.53% and 4.2%, respectively.

Fig. 2 Thermogravimetric analysis of pitch and carbon fiber samples. RP: raw petroleum pitch; ITP: inert treated pitch;VTP: vacuum treated pitch; SP: stabilized mesophase pitch; T-800: PAN based carbon fiber; P-75: pitch based carbon fiber.

Table 1 summarizes the softening point, carbon, hydrogen and nitrogen contents of the RP, ITP, VTP, and SP. From Table 1, it can be seen that the softening point of the pitches increases with the treatment sequence from RP to ITP, VTP and SP. Similarly, with the sequence of pitch treatment upto VTP the carbon content increases and hydrogen content decreases. After oxidative stabilization the carbon content of the SP reduces slightly due to an addition of oxygen, which is obvious. Further, the TGA as depicted in Fig. 2 shows that the yield of the pitch samples increases with the treatment sequence from RP to ITP, VTP and SP. This is the main motive behind the conversion of RP into VTP and the oxidative treatment of VTP thereafter. The carbon yield increase in TGA with treatment sequence till the conversion to VTP might have happened due to removal of aliphatic carbons and their conversion into aromatic carbons. As the resultant pitch after these treatments is more aromatic, hence the increase in carbon yield is obvious. Further, an increase in yield after stabilization would have happened due to the cross-linking reactions with oxygen, which increases carbon yield in the pyrolysis and carbonization.

Table 1 Properties of pitch.

The detailed discussion about the interaction of oxygen with matrix precursor and stabilized matrix precursor with reinforcement thereafter, and the influence of real time processing parameters on the yield, yield rate, and yield impact of SP in C/C composite system are presented in subsequent sections.

3.1 Matrix precursor interaction

High yield of pitch-based matrix precursors can be gained by high processing pressure, slow heating rate[2], and oxidative stabilization[8-14]. However, for the conventional processes to prepare C/C composites, like HIPIC, it is very difficult to obtain high carbon yield from matrix precursor due to the lack of micro level interaction of oxygen with matrix precursor. In order to understand these micro level issues, Fig. 3 shows a typical micro level matrix precursor interaction with the reinforcement for conventional process of C/C composite fabrication. In the conventional process, carbon fiber tow is used to weave out the desired preform and the pitch powder is melted and infiltrated into the preform. It is very important to note that the infiltration of pitch matrix precursor into the woven preform is done prior to stabilizing the pitch. Therefore, the infiltration of molten pitch matrix precursor in preform pores is highly dependent on the viscosity and the softening point of the precursor. Post stabilization of pitch is employed in the conventional method. As a result, core of the preform is incompletely stabilized as depicted in the Fig.3. This non-uniform stabilization generally results in localized oozing of the matrix precursor and low net carbon yield along with non-uniform microstructure. Further, the fiber preform used in conventional routes comprises tow level weaving. In such kind of carbon fiber woven preforms, the uniform filament level wetting by pitch matrix precursor is hard to attain. However, the use of these composites in real application desires the filament level wetting. These are technically difficult with conventional routes (e.g., HIPIC, et al), besides time consuming, costly process and limitation to fabricate big composites. The above issues are tackled by the novel C/C fabrication process devised by us to provide a high degree of oxygen interaction with matrix precursor, and interaction of stabilized matrix precursor with reinforcement thereafter.

Fig. 3 A typical conceptual arrangement of the interaction of matrix precursor in conventional technique.

Fig. 4 Conceptual arrangement of the interaction of matrix precursor during C/C composite fabrication.

Fig. 4 shows the process in which the matrix precursor i.e. isroaniso matrix precursor or mesophase pitch powder of 15 μm was uniformly stabilized by the oxygen diffusion from air. It is noteworthy that the stabilization of mesophase pitch was achieved on micro level by stabilizing it in powder form whereas the same stabilization cannot be achieved in conventional processes. The SP thus obtained was used to prepare the slurry. The chopped and exfoliated carbon fibers were mixed thoroughly with SP with the help of distilled water. It can be seen that in the well mixed powder of SP and carbon fibers, the SP powder interacts with the exfoliated carbon fibers on filament level. This level of matrix precursor to reinforcement interaction was achieved through the exfoliation assisted dispersion of reinforcement in the stabilized matrix precursor. It is noteworthy that in this process the SP with very high viscosity and softening point can be used without infiltration into the deep pores of preform. Further, this process provides the filament level wetting of reinforcement by SP. Hence, the yield even at low pressure by the usage of high softening point matrix precursor would be significant and the microstructure of the carbon matrix derived from this SP would be uniform. Fig. 5 shows representative SEM micrograph of the compact processed at 0.2 ℃/min, 15 MPa and the ratio of matrix to reinforcement of 50∶50. It can be seen that the interaction of the carbon matrix derived from matrix precursor, i.e. SP on filament level is excellent and the carbon fibers are well scattered within SP.

Fig. 5 SEM micrograph of the compact processed at 0.2 ℃/min, 15 MPa and the ratio of matrix to reinforcement of 50∶50.

3.2 Influence of hot-pressing pressure

It is a well-known fact in the processing of C/C composites with pitch matrix precursors that the carbon yield of pitch increases with a decrease in heating rate and an increase in pressure[7,16]. However, the combination of these two important parameters varies from process to process. For example, a high heating rate coupled with a high pressure may work in HIPIC route may result in a significant yield, whereas the same conditions are either not possible or may not result in the same carbon yield in hot-pressing route. Therefore, an optimized combination of hot-pressing pressure and heating rate with respect to carbon yield of mesophase pitch for hot-pressing of short carbon fiber reinforced mesophase pitch matrix precursor derived carbon matrix composite was studied. A matrix precursor to reinforcement ratio of 50∶50 and a heating rate of 0.5 ℃/min were used and the influence of hot-pressing pressure on the yield (yhpandyc), yield rate (yr), and yield impact (yi) of SP was studied.

Fig. 6 shows the yield behavior of SP with hot-pressing pressure. It is evident from Fig.6 that the carbon yield increases as the hot-pressing pressure increases from 5 to 20 MPa. Though the yield increases with the pressure but at and above 20 MPa hot-pressing pressure the compact was failed due to the severe exudation of SP during hot-pressing. The higher hot-pressing pressure may be used by further enhancing the oxidative stabilization degree of SP. Further, it can be seen that the hot-pressing yield is more than the carbon yield due to retention of some of the heteo atoms by SP during hot-pressing, which are liberated during carbonization at elevated temperatures. The trend of increase in the yield of SP is well in concurrence with that reported[7,16]in the literature. It is noteworthy that the yield studies in the literature are on coupon level samples whereas in the present study the influence of real time hot-pressing pressure on the yield of comparatively larger compact is reported. The pressure at which a maximum yield is obtained and exudation is free in compact is most important. Hence, the pressure of 15 MPa was used for further studying the exudation free compact.

The influence of hot-pressing pressure on the newly introduced denominations (yield rate and yield impact) is shown in Fig. 7. The yield rate is derived to assess the yield of SP with respect to processing time. The higher the yield rate, the shorter will be the processing time. The other denomination, yield impact, determines how efficiently the process transforms the yield of the SP into the end density of the C/C compact. If two systems with the same yield have different yield impacts then the compact with a higher yield impact is more efficient subject to the exudation free processing. It is evident from Fig.7 that the yield rate increases with hot-pressing pressure but the yield impact attains a maximum at 15 MPa. On a further increase in the pressure, the yield impact suddenly drops down due to failure of the compact as explained in the earlier section. Therefore, the pressure of 15 MPa is found most suitable considering the yield, yield impact, and exudation free compact, which are crucial in deciding the processing parameters with an efficient transformation of yield into end density.

Fig. 7 Yield rate and yield impact behavior of mesophase pitch as a function of hot-pressing pressure.

Further the carbon contents of all the compacts were found to be almost the same ((95.8±0.2)% after hot pressing and (98.8±0.2)% after carbonization) for a range of pressure 5-20 MPa (results not shown). It implies that the carbon content in the compacts at the end of hot-pressing and carbonization is independent of the pressure employed. However, it is interesting to note that the carbon content is higher after carbonization than that after hot-pressing process. Since two types of carbon components are present in C/C composites, the retention of hydrogen by secondary carbon component is very common[20]. As a result of this retention of hydrogen by secondary carbons in the compacts, the carbon content is less after hot-pressing. This hydrogen is liberated only at carbonization. Even after carbonization, a minute quantity of hydrogen is retained by the secondary carbons. As a result, the net carbon content is always less than 100% even after the compact is carbonized at high temperature. It indicates that the carbon content of the compact is more temperature dependent than pressure.

3.3 Influence of heating rate

Literature shows that the heating rate critically influences the carbon yield of pitch-based matrix precursor[21]. Therefore, it is important to investigate the heating rate effect on the yield, yield rate and yield impact for SP in the randomly oriented C/C composite system. The influence of heating rate on the yield (yhpandyc), yield rate and yield impact of the SP is shown in Fig.8 and 9 at 15 MPa pressure and the mass ratio of matrix to reinforcement of 1.0. It can clearly be seen from Fig.8 that the yield of the SP in the composite decays almost exponentially with the increase in heating rate. Fig. 9 shows that the yield rate increases with the increase in the heating rate but the yield impact decreases very sharply. Hence, with the help of yield, yield rate and yield impact, one can decide a suitable heating rate for the development of the randomly oriented C/C composite. It is also found (results not shown) that the carbon content of the compact is not influenced by the heating rate.

Fig. 8 Yield behavior of mesophase pitch as a function of hot-pressing heating rate.

Fig. 9 Yield rate and yield impact behavior of mesophase pitch as a function of hot-pressing heating rate.

3.4 Influence of the ratio of matrix to reinforcement

The influence of the ratio of matrix to reinforcement in a wide range on the yield (yhpandyc), yield rate and yield impact of the SP is shown in Fig.10 at 15 MPa pressure and a heating rate of 0.2 ℃/min. Fig. 10 shows that the yield and yield rate of the SP decrease with increasing the ratio of matrix to reinforcement. This might be due to more absorption of the pressure in the case of a higher ratio of matrix to reinforcement compared to a lower ratio of matrix to reinforcement. It is due to the fact that pressure drops more quickly for the compacts having high ratios of matrix to reinforcement. It can be further clarified that carbon fibers are almost incompressible at the hot-pressing and carbonization temperatures and only SP with self-sintering and plasticity characteristics undergoes compressibility. The higher the amount of compressible substance in the compacts, the more the absorption of the pressure. As a result, the net pressure experienced by the SP during hot-pressing compacts with higher ratios of matrix to reinforcement would be less as compared to that with lower ratios of matrix to reinforcement and once the pressure is reduced the yield decreases[16]. Hence, a lower yield is obtained for a compact with a higher ratio of matrix to reinforcement.

Fig. 10 Yield, yield rate and yield impact behavior of mesophase pitch as a function of matrix to reinforcement ratio.

4 Conclusions

The filament level wetting of carbon fibers by carbon matrix was successfully achieved by exfoliating the carbon fibers prior to slurry preparation. The influence of real time hot-pressing parameters was successfully investigated on the yield of indigenously synthesized SP matrix precursor. The higher yield is found at 20 MPa, however, the compact fails due to exudation of SP. Further, the yield is found to increase linearly with decreasing the hot-pressing heating rate, the ratio of matrix to reinforcement, and with increasing pressure. The yield of SP is found to be high at a ratio of matrix to reinforcement of 30∶70, however, the compact is almost failed due to insufficient binder. The new denominations i.e. yield rate and yield impact were introduced to assess the effectiveness of the yield of the matrix precursor and the transformability of the yield into end density of the composite system. The new denominations are found prominent in assessing and optimizing process parameters. The higher yield, yield rate and yield impact are obtained even at a moderate hot-pressing pressure. The study gives a useful insight into the behavior of yield related parameters of SP with real time hot-pressing parameters. The optimum yield (yhpandyc) of SP is found with a heating rate of 0.2 ℃/min, hot-pressing pressure of 15 MPa, and the ratio of matrix to reinforcement of 50∶50.

Acknowledgments

The authors are thankful to Mr. V.K. Vineeth, Mr. Omendra Mishra, Ms. N. Supriya, Ms. Soumyamol, Mr. K. S. Abhilash, Ms. Bismi, Mr. Sushant K. Manwatkar, Mr. Ranjith, Mr. Vijendra Kumar, Mr. Biswa Ranjan Mohanty, Mr. Narayan Murthy, Ms. Mariamma Mathew and Dr. Koshy M. George for their kind support rendered in various capacities.