Evaluation of an original use of spark plasma sintering to laminate carbon fibres reinforced aluminium

Abstract

An alternative route for producing aluminium matrix reinforced with continuous carbon fibres is proposed in this paper. On the one hand, liquid aluminium does not wet carbon; on the other hand, however, the two form a reactive system leading to carbide formation. A novel way to obtain continuous carbon fibre-reinforced aluminium was investigated, using spark plasma sintering with aluminium foils as raw material. Sintering parameters were adjusted to achieve the effective welding of aluminium foils and penetration of the metal between the filaments. A quality assessment of the fibre/aluminium coupling is presented. Interfaces were then investigated by scanning electron microscopy, transmission electron microscopy and energy-dispersive ray spectroscopy. An effective cohesion of fibres with the matrix was shown. The manageable fibre positioning could result in unidirectional architecture and reinforcement rate should be handled through foil thickness and yarn properties. Using tensile tests, cohesion between aluminium and carbon fibres can be quantified.

Keywords

Metal matrix composites spark plasma sintering fibre/aluminium coupling

Introduction

Aluminium matrix composite materials (AMCs) represent tremendous potential for many industrial applications, particularly for heat sink. An aluminium matrix reinforced by highly thermal conductive continuous carbon fibres represents a high potential for thermal applications. Over the last decades, many processing routes have been explored to develop AMCs, both in liquid and solid form.^1,2

However, coupling liquid aluminium and carbon fibre can be difficult, due to their incompatibility. For example, a really low wetting of carbon by liquid aluminium^3,4 and its chemical reactivity can lead to the formation of aluminium carbide Al₄C₃ which is harmful to the properties of the material.^5–7 Many studies have tried to find solutions to these problems. These include the use of ceramic or metal fibre coatings.^8,9 Metallic coatings mainly of nickel or copper are used to improve the wettability of fibres by aluminium.^10,11 These coatings can be obtained through electrolytic methods, with current or by electroless techniques.^12,13 Other techniques consist on the chemical activation of wetting by deposits of fluorinated salts K₂ZrF₆¹⁴ or K₂TiF₆^15,16 or even processes involving the use of infiltration by ultrasound or by means of a pressurized chamber.^4,17,18

The use of the solid processing route is another alternative for developing carbon fibre-reinforced AMCs. These processes consist of maintaining a discontinuous material under pressure at a temperature below its melting temperature. This solution eliminates wetting problems. In addition, the diffusion mechanisms are largely slowed down, as the carbon can be considered insoluble in solid aluminium.¹⁹

In order to employ continuous reinforced metal matrix composites (MMCs) by means of powder metallurgy, it is essential to be able to control the placing of the reinforcement in the matrix. By using slurry^20,21 or a continuous titanium powder pre-deposition technique on silicon carbide (SiC) filaments that have been positioned before being sintered,^22,23 MMCs can be produced. However, as the diameter of the carbon filaments is more than 10 times smaller than that of the SiC filaments, this process proves to be unsuitable for obtaining sufficient fibre volume fraction.

The use of sheets instead of powders may in some cases be a suitable alternative as this will considerably restrict fibre swimming^24,25 during implementation. This technique is used to produce titanium matrix composites reinforced with SiC filaments.²⁶

In the case of aluminium, however, the presence of a thin layer of alumina makes penetration difficult and it is essential that it is disrupted during the process to ensure a good penetration of the aluminium.²⁷ To encourage this disruption and promote good consolidation, some authors have suggested initially using a powder with an irregular morphology,²⁸ or sintering in a nitrogen atmosphere,²⁹ or adding magnesium to the aluminium powder.³⁰ However, as sheets have a flat surface, they are not sufficiently able to ensure that this disintegration takes place during conventional hot compression.

Spark plasma sintering (SPS) is a technique that has proved effective for quickly and easily sintering a large number of materials, even those which are difficult to sinter, and obtaining new microstructures at room temperature.³¹ It differs from traditional hot compression as its heating mode is by pulse current directly through the mold and the sample if it is a conductor. With this technique, very rapid heating rates can be achieved (up to several 100 K/min), notably making it possible to obtain vitreous structures.³² In addition, there are many mechanisms at the interfaces associated with the strong SPS currents that can activate sintering more effectively than the traditional hot compression processes.^33–37 Al-C_f composite materials with particulate reinforcement^38,39 or with discontinuous fibres have already been obtained by powder metallurgy by mixing chopped fibres with the powder before hot compression by SPS.⁴⁰ Continuous fibre-reinforced aluminium has also been achieved by SPS by sinking fibres into an ethanol suspension solution before sintering.⁴¹ Thus, SPS can be a solution for sintering aluminium while doing away with the layer of oxide.

Until now, its use has mostly been restricted to the sintering of powders, and the use of metallic foils has only been investigated within this process in a few cases.⁴² Hence, a new way to manufacture aluminium matrix composite reinforced by continuous carbon fibres using the fibre–foil technique processed by SPS was investigated. The application only aims at providing a good local bonding between fibre and aluminium to ensure a thermal conductive interface. The objective is not to obtain good mechanical properties for the composite material.

After carrying out studies to adapt the sintering parameters (temperature, pressure and time) in order to obtain well penetrated carbon fibres, qualitative and chemical analyses of the interface are presented.

Experimental setup and starting material

Pitch-based carbon fibres (YS-80 A, Nippon Graphite Fibre) with high-modulus and high thermal conductivity (800 GPa; 320 W/(m.K) – manufacturer’s data) and made up of 3 k filaments were cut into 10 mm lengths then placed between 25 µm-thick sheets of aluminium 1050 A (EN 573-1, Goodfellow).

There are different techniques for removing the seizing from carbon fibres;^4,5,17,43 this has to be done to avoid any fibre–matrix interface contamination by polymer species. After carrying out a thermogravimetric analysis on our fibres, a 30-min finish removal at 400℃ was put in place. There is also a preparation of the surface of the aluminium foils before the SPS: degreasing with deionized water/pickling 10 s in a NaOH solution – 4 g/L at 50℃ with movement/several prolonged rinses with deionized water/drying in an oven.

The different stacks of sheets and fibres were placed in a graphite mold (Graphitech) comprising a floating matrix of 15 mm in diameter and two pistons (Figure 1) then placed between the two electrodes of the SPS equipment (Dr. Sinter LAB (SPS-515), SPS Syntex Inc.). Sintering was carried out in a vacuum. Uniaxial pressure was applied gradually until reaching 50 MPa after two minutes. The heating sequence was 12:2, a heating rate of 50℃/min up to a temperature level of 500℃, then 25℃/min for the last 50° to reduce overshoot. The 12:2 pulse used in the study is the pulse condition recommended by the manufacturer SPS Syntex Inc: the first number corresponds to the time ON and the second to the time OFF. The current pulses are ON: OFF sequences of 3 ms. In addition, a minimal force of 3.3 kN (i.e. pressure of about 17 MPa for our setup) was applied before beginning the experiments to ensure adequate electrical contact. Temperature was measured and controlled by a type K thermocouple placed 3 mm from the outer surface of the matrix. At the end of the sintering process, the electric current was turned off and the sample cooled in a vacuum in the mold. The sample was kept separate from the different parts of the mold by sheets of graphite (papyex) to prevent them from sticking.

Figure 1.

Experimental setup (SPS). SPS: spark plasma sintering.

After cutting, polishing and lastly ionic abrasion, the resulting samples were characterized by optical microscope (Keyence), scanning electronic microscope (SEM, Hitachi S-3000 N), electron dispersion spectroscopy (EDS, Bruker XFlash 6l10 or Oxford Instrument Xmax) and transmission electron microscopy (TEM, JEOL JEM-ARM 200 F).

Results and comments

Adapting experimental parameters

Tests without reinforcement

Before processing composite samples, it was necessary to ensure that SPS was effectively able to bond aluminium foils and determine the sintering temperature for further experiments. Sintering trials were carried out on a stack of simple sheets (without reinforcement) comprising two 500 µm sheets between which were placed three 25 µm sheets (Figure 2). Different sintering temperatures were tested (500, 550 and 575℃) with isothermal hold for 10 min at a pressure of 50 MPa.

Figure 2.

Cross-section diagram of sheet stack for adjustment of sintering temperature.

The 575℃ test caused the aluminium to melt, although the temperature measured by the thermocouple was lower than the melting point of aluminium. Many numerical and experimental studies have indeed reported strong temperature gradients in the SPS assembly.^44–47 When dealing with a conductive material, the current passes directly into the sample, which therefore heats up by the Joule effect. Thus, the temperature in and around the sample is higher than at the point used as the temperature control point. The temperature measured by the thermocouple placed 1 mm from the sample did indeed show that there can be a difference in temperature of 80℃ between this point and the control thermocouple for a control temperature of 550℃, which confirms the results found in the literature.

Samples obtained for temperatures of 500 and 550℃ appeared to be well consolidated. The 500℃ sample was prepared for microscopic examination and showed delamination along its entire length (Figure 3). The samples obtained at 550℃ were well consolidated, although the interface between the different sheets remained visible under the SEM (Figure 4). We also noted the presence in the aluminium of intermetallic compounds appearing white, which under EDS analysis proved to contain iron, the common main additional element present in 1050 aluminium.

Figure 3.

Cross section of a sample of aluminium alone produced at 500℃ (optical microscope).

Figure 4.

Demonstrating bonds between sheets (SEM). SEM: scanning electronic microscope.

Experiments with reinforcement

After these encouraging results, we moved on to composite samples, retaining the sintering temperature at 550℃. Carbon fibres were introduced manually between each of the 25 µm-thick sheets, in the same direction (Figure 5). The filaments were laid out widthways in order to reduce the number of filaments that made up the thickness of the fibre and facilitate infiltration by the aluminium. This step was carried out by hand, resulting in a non-uniform distribution of filaments within the consecutive layers. Two configurations of pressure application and temperature were studied, with pressure applied either at the beginning before heating or when hot just before the sintering stage (Figure 6).

Figure 5.

Diagram of composite stack to study the influence of time.

Figure 6.

Different pressure and temperature profiles during sintering: (a) pressure applied on the specimen at room temperature; (b) pressure applied on the specimen from 500℃.

Pressure application

Observations of polished cross sections show that the fibres retained their original position, thus the structure of the material can be compared to that of a multilayer material (Figures 7 and 8). When pressure is applied at room temperature, the filaments are mostly compacted and partially crushed, forming a continuous mass of carbon that the aluminium is unable to infiltrate (Figure 7). Around the edges, however, the filaments in contact with the aluminium are well embedded in the metal. When the thickness of the fibre is a single filament, these filaments are completely infiltrated by the aluminium.

Figure 7.

Cross section of a sample produced at 550℃ with pressure applied on the specimen at room temperature (SEM). SEM: scanning electronic microscope.

Figure 8.

Cross section of a sample produced at 550℃ with pressure applied on the specimen from 500℃ and holding time of 10 mn (SEM). SEM: scanning electronic microscope.

The samples produced by applying hot compression included fibres that were perfectly infiltrated by the aluminium (Figure 8). When pressure is applied as the temperature reaches 500℃, the aluminium penetrates the fibre, which shows no signs of crushing. At this temperature, the viscosity of the aluminium state is assumed to be sufficiently low to enable it to deform between the filaments without crushing the fibre: each filament is then completely surrounded by aluminium, even when the filaments are very close together, with spacing of less than 1 µm. Nevertheless, some damaged filaments were observed.

Isothermal holding time

Samples were produced by being held at 550℃ for 5, 10 and 20 min with hot pressure applied. A holding time of five minutes proved insufficient to obtain a satisfactory result (Figure 9): the sample revealed considerable defects linked with infiltration, cohesion and filament scouring. After a 10 min holding time (Figure 8), the infiltration we obtained was very satisfactory and the samples showed no fibre scouring and few areas of debonding. The results are similar after a holding time of 20 min (Figure 10). For the continuation of the study, the holding time is equal to 10 min.

Figure 9.

Debonding between filaments and aluminum in a cross-section of a sample produced at 550℃ with pressure applied on the specimen from 500℃ and holding time of 5 mn (SEM). SEM: scanning electronic microscope.

Figure 10.

Cross section of a sample produced at 550℃ with pressure applied on the specimen from 500℃ and holding time of 20 mn (SEM). SEM: scanning electronic microscope.

Summary

Early studies to develop a carbon fibre-reinforced aluminium matrix composite by the SPS process from aluminium sheets led to the choice of applying uniaxial pressure of 50 MPa at 500℃ immediately before 10 min sintering at 550℃. These experimental parameters gave well-infiltrated fibres which maintained their position and had limited damage. These results were reassuring with the view to obtain full composite samples.

Implementing a multilayer composite

Samples were developed comprising a stack alternating only thin sheets and carbon fibres, giving a sample containing 20 layers of carbon fibres (Figure 11). Our observations suggest that the fibres are very well infiltrated, with position and orientation conserved despite alignment defects due to being put in place manually (Figure 12). In addition, filament density is not uniform across the width of the sample for the same fibre. The volume fraction of carbon fibres is estimated from image analyses to be around 40% and may be higher, thus changing the raw materials’ properties resulting in thinner foils and/or yarns with more filaments.⁴⁸

Figure 11.

Cross-section diagram of stack of a multilayer composite sample.

Figure 12.

Cross sections inside a 20-layer composite sample for two magnifications (SEM). SEM: scanning electronic microscope.

Quality assessment of fibre/matrix coupling

The study of the interfaces gives an idea of the quality of the MMC obtained. It is important to examine them closely, carrying out physico-chemical analyses, mechanical tests and observations of the fibres (orientation and degradation). For instance, aluminium carbides may include notch effects that decrease the ultimate composite strength. There are two types of interface to study: the interface between the aluminium sheets, when it is discernible, and the interface between the carbon fibres and the metal matrix.

Mechanical coupling

In order to check if the bonding between aluminium and carbon fibres shows cohesion, tensile tests were carried out on halter-shaped composite samples with 5 × 0.5 cm² cross section. With a fibre volume fraction of 40%, the modulus predicted by the rule of mixture is 355 GPa.

The stress-strain curve first shows an elastic behaviour up to a strain close to 0.02%. Then the curve bends slightly before rupture, which occurs at a strain rate equal to three times the strain at the yielding point (Figure 13). This change in slope appears to have occurred following an effect presumed to be damaging because unloading curves are linear and return to a null strain. Reloads then follow the same path with a slope designated E_damaged. This damage effect is then not due to plasticity but is related to “damageable elasticity”.

Figure 13.

Stress-strain curve of composite sample until rupture (blue), discharge curve (red).

The average measured modulus for the processed composites is 269 GPa, which is lower than expected by the rule of mixture. This result can be attributed to fibre disorientation. Indeed, fibres are placed manually and their disorientation could be measured at between 1 and 14°. Hence, we could evaluate the expected modulus E^θ using the formula related to a disoriented unidirectional (UD) ply.⁴⁹ This leads to an expected modulus of 302 GPa. Fibre disorientation actually impacts the composite modulus by more than 10%.

Another phenomenon that has an impact on composite modulus is fibre damage. Carbon fibres may indeed suffer multiple ruptures. Although fibre damage was low, SEM observations showed that it did occur. This suggests that fibre fragmentation already occurs during the manufacturing process. In addition, in a discontinuous fibre-reinforced composite, there is a critical length l_c required for the load to be fully transferred to the fibres (Figure 14). Hence, each time a fibre breaks, its length supporting the load is lowered (Figure 15). This fragmentation phenomenon may then occur gradually during the tensile tests, lowering the composite modulus without resulting in any non-elastic behaviour for the material. A coefficient representing this damaging effect was set up, assuming that the load is linearly transferred along the critical length l_c.

η = \frac{l}{l + {nl}_{c}} = \frac{1}{1 + n \frac{l_{c}}{l}}

(1)

with l the composite length and n the number of ruptures. When η = 1/2, the damage is maximum as nl_c = l, meaning there is no anymore section of fibres in which the nominal load can be reached). Assuming an equivalent fibre modulus E_f^θ, as the stiffness of the composite is equal to the E^θ estimated earlier, η can be computed as follows

η = \frac{E_{c} - E_{m} (1 - V_{f})}{E_{f}^{θ} V_{f}}

(2)

Figure 14.

Representation of the transfer of charge between a filament and the matrix in a composite material.

Figure 15.

Representation of the distribution of the stress around a filament break.

The evolution of this coefficient could then be drawn together with the strain rate (Figure 16). On the tensile test previously shown in Figure 13, η is initially 0.87 until the composite modulus lessens around 0.022%. The rupture occurs for η = 0.66, before maximum damage (η = 0.5). This result could be predicted as the composite rupture took place for 0.13%, whereas the maximum fibre elongation is 0.5%. Aluminium carbides may also be the cause of that rupture.

Figure 16.

Young’s modulus and efficiency coefficient evolution during tensile test of a composite sample.

The observation of fractographs has shown very interesting behaviour of the sintered materials. Samples obtained with aluminium foils show only ductile rupture but delamination occurred between each pair of initial foils (Figure 17). When we add carbon fibres, aluminium still appears ductile but the bonding between sheets is more effective as much less delamination occurs (Figure 18(a)). Aluminium is ductile deformed between filaments even though some left their cavities following the rupture (Figure 18(b)).

Figure 17.

Fracture surface of aluminium sample (sample produced at 550℃ with pressure of 50 MPa applied on the specimen from 500℃).

Figure 18.

Fracture surface of composite sample (sample produced at 550℃ with pressure of 50 MPa applied on the specimen from 500℃).

Hence, the addition of carbon fibres to aluminium through the foil–fibre–foil technique resulted in high specific stiffness but also improvements in the bonding between aluminium sheets. This mechanical analysis provides information about the quality of the coupling between aluminium and fibres. Physico-chemical analysis of bonding may explain such behaviour and give additional information on the process and the presence of carbides.

Analysis of interfaces

Aluminium sheet bonding

Interfaces between sheets remain discernible when there is contact between two sheets with no carbon fibre (Figure 19). They are clearly discernible under the electron microscope and are free of apparent cohesion defects. EDS analysis reveals peaks in the presence of oxygen within the samples. These oxygen peaks coincide with the boundaries of the layers of aluminium sheets, but also with the ferrous compounds present in the sample. An analysis shows that this alumina layer may be discontinuous along the junctions between sheets (Figure 20).

Figure 19.

(a) Cross section of a sample without fibres showing interfaces between sheets (SEM – BSE); (b) Associated EDS line analysis of Al and O along the arrow (non-quantitative data – normalized with Al signal – 10 kV). EDS: electron dispersion spectroscopy; SEM: scanning electronic microscope; BSE: Back Scattering Electron.

Figure 20.

EDS mapping of oxygen (blue) and carbon (green) in a cross section exhibiting oxide break along sheet junction (10 kV). EDS: electron dispersion spectroscopy.

Analysis of fibre–matrix interfaces

One of the reasons for choosing the solid form was specifically to avoid the formation of aluminium carbides. Observations using the SEM at lower acceleration tension (5 or 10 kV) to show cross sections of composite samples reveal compounds entirely surrounding the filaments (Figure 21). Although the matrix shows great penetration between filaments, even in spaces tighter than a few 100 nm, a few spots seem to be still free of aluminium and a lack of bonding between fibre and matrix remains discernible.

Figure 21.

Compounds surrounding filaments and partial debonding between filaments and matrix (SE). SE: Secondary Electron.

An EDS mapping of Al, C and O was performed; Fe was not detected within the area (Figure 22). This shows that carbon does not overfill the filament section and aluminium is not detected within the filaments. Meanwhile, the compounds surrounding the carbon clearly contain oxygen and aluminium but no carbon, leading to the conclusion that it is made of alumina. The origin of such alumina compounds is assumed to be the result of the rupture of the original surface alumina layer of foils used in manufacture. Because of the loss of resolution due to the nature of the analysis, carbon fibres seem to be bonded. Still, the detection of oxygen within these gaps and higher magnification captions (Figure 21) show that fibres may indeed be down to a few nanometres and close together or even in contact with each other but that they are not bonded. The presence of aluminium carbides was not shown within this scale of analysis.

Figure 22.

(a) SEM – BSE picture and associated EDS mapping of (b) carbon, (c) aluminium and (d) oxygen on a cross section of composite sample (5 kV); red square represents an area where thin blade was processed for TEM analyses. EDS: electron dispersion spectroscopy; SEM: scanning electronic microscope; TEM: transmission electron microscopy.

TEM analyses were carried out on a thin blade to study the foil–fibre interfaces in greater detail. The blade was processed using focused ion beam between two filaments closer than 1 µm. Such tight areas – where aluminium infiltrated – are the most crucial for interfaces since they are the most suitable for aluminium carbide (Al₄C₃) to create, as will be discussed later. These analyses have shown that aluminium carbide was indeed created within the process. Lath-like crystals under 0.5 µm in length and a few tenths of a nanometre wide were sometimes found stuck along the carbon fibres (Figure 23). EDS analysis of these crystals confirmed that they are made of aluminium carbides and their morphology matches the literature descriptions. However, the number and size of the observed carbides are lower than the amounts created through liquid routes,⁵⁰ confirming the short duration of local melting.

Figure 23.

(a) STEM observation of aluminium penetrated between two carbon filaments showing the presence of lath-like crystals along filaments; (b) magnification on a carbide crystal. STEM: Scanning Transmission Electron Microscopy.

In order to complete the study of the quality assessment of fibre/aluminium coupling, we have characterized the diffusivity of the composite material.

Thermal coupling

To measure the thermal diffusivity of the composite material in the direction of the fibres, we used a technique developed specifically for thin plates.⁵¹ It consists in sending continuous power onto a sample surface using a bonded heated resistance. The difference in temperature change at two measurement points enabled us to estimate thermal diffusivity. The first results obtained show that the thermal diffusivity of the composite material is 1.7 times higher than that of aluminium alone, thus confirming the previous results.

Discussion

Processing aluminium foils and carbon fibres by SPS resulted in samples with perfect penetration of the matrix into reinforcements. To achieve such a result, the dwell temperature was set to 550℃ for 10 min under vacuum. A mechanical pressure of 50 MPa was applied between 500 and 550℃ during the last two minutes of the heating step after it was found that filaments suffered heavy damage and the matrix did not penetrate the fibres.

In cases where the aluminium is molten, such as those at 575℃, the metal leaks out along the pistons as the assembly is not watertight. In the tests carried out at 550℃, resulting in full penetration of the metal, this phenomenon was not observed. It is therefore acknowledged that full melting of the aluminium is excluded.

In addition, it was shown that applying pressure when the metal is hot rather than cold results in good penetration of aluminium into tight gaps between carbon filaments. Hence, aluminium creep may be the first mechanism involved. When hot, the metal is in a viscoplastic state and tends to flow between the filaments with the action of the applied pressure. This phenomenon requires a certain amount of time to reach full penetration, which would explain the insufficient penetration of the matrix observed when the hold time was under 10 min (Figure 9). Still, the presence of aluminium carbides suggests that the liquid state was reached, in accordance with the literature.^19,50 Localized melting over a short period must therefore be considered.

Other studies have estimated that in the case of powders sintered by SPS, the temperature can therefore be very much higher at the contact points between two aluminium particles than in the bulk of the material because of the convergence of the current in necking representing a tight section.^52,53 Within our study of foil–fibre–foil technique, similar hot spots may be formed at the contact between foils and fibres. These can then result in localized melting of the aluminium which then infiltrates between the carbon filaments into tight areas under the action of pressure. Carbides may form during this step. However, the further the infiltration progresses, the wider the interfaces between the sheets and carbon become. Hot spots resorb as the current becomes more uniform.

Consequently, it seems that a gradient of several tens of degrees/µm may occur during the process. This is part of the numerous debates about SPS that are still ongoing and for which many studies are willing to determine the real temperature reached by the sample, whether it is an electrical conductor or not.^{31,44,54–57}

Conclusion

We conclude that continuous carbon fibre-reinforced aluminium could be manufactured by SPS using the foil–fibre–foil technique. The sintering temperature was selected to ensure good infiltration of the fibres. Applying pressure (50 MPa) when hot and with a holding time above 10 min, we observed good infiltration into the fibres by the aluminium with little filament damage. Apart from a few occasional defects, little lack of bonding between fibre and matrix could be seen in the samples. Local melting of aluminium has been identified as playing a role in achieving full infiltration as aluminium carbides were found. However, the amounts observed are moderate compared to liquid processing routes and hopefully could be controlled in further studies along with a better understanding of SPS technique. In addition, the foil–fibre–foil technique led to a high specific stiffness. The overall results show that AMCs could be a solution to improving the properties of aluminium, such as thermal conductivity.

Footnotes

Thanks to Alain Largeteau and U-Chan Chung Seu of ICMCB (University of Bordeaux – France) for providing the SPS facilities and to the Raimond Castaing Characterization Center for TEM analyses (University of Toulouse – France).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge the financial support from BPI France.

References

Rosso

. Ceramic and metal matrix composites: Routes and properties. J Mater Process Technol 2006; 175: 364–375.

Shirvanimoghaddam

Hamim

Akbari

et al.

Carbon fiber reinforced metal matrix composites: fabrication processes and properties. Compos Part 2017; 92: 70–96.

Rocher

Quenisset

Naslain

. A new casting process for carbon (or SiC-based) fibre-aluminium matrix low-cost composite materials. J Mater Sci Lett 1985; 14: 1527–1529.

Hufenbach

Gude

Czulak

et al.

Development of textile-reinforced carbon fibre aluminium composites manufactured with gas pressure infiltration methods. J Achiev Mater Manuf Eng 2009; 35: 177–183.

Ureña

Rams

Escalera

et al.

Characterization of interfacial mechanical properties in carbon fiber/aluminium matrix composites by the nanoindentation technique. Compos Sci Technol 2005; 65: 2025–2038.

Daoud

. Microstructure and tensile properties of 2014 Al alloy reinforced with continuous carbon fibers manufactured by gas pressure infiltration. Mater Sci Eng A 2005; 391: 114–120.

Lee

Choi

Sugio

et al.

Effect of aluminum carbide on thermal conductivity of the unidirectional CF/Al composites fabricated by low pressure infiltration process. Compos Sci Technol 2014; 97: 1–5.

Payan S. Comportement à la corrosion galvanique de matériaux composites à matrice d’alliage d’aluminium renforcée par des fibres de carbone haut-module. PhD Thesis, Université Sciences et Technologies-Bordeaux I, France, 2001.

Margueritat-Regenet C. Elaboration et caractérisation de fils composites C/Al: infiltration spontanée et continue par activation chimique du mouillage. PhD Thesis, École Nationale Supérieure des Mines de Paris, France, 2002.

10.

Silvain

Heintz

Lahaye

. Interface analysis in Al and Al alloys/Ni/carbon composites. J Mater Sci 2000; 35: 961–965.

11.

Seong

Lopez

Rohatgi

. Microsegregation during solidification of graphitic fiber-reinforced aluminum alloys under external heat sinks. Metall Mater Trans A 2007; 38: 138–149.

12.

Rams

Urena

Escalera

et al.

Electroless nickel coated short carbon fibres in aluminium matrix composites. Compos Part A Appl Sci Manufact 2007; 38: 566–575.

13.

Abraham

Pai

Satyanarayana

et al.

Copper coating on carbon fibres and their composites with aluminum matrix. J Mater Sci 1992; 27: 3479–3486.

14.

Rocher

Quenisset

Naslain

. Wetting improvement of carbon or silicon carbide by aluminium alloys based on a K2ZrF6 surface treatment: application to composite material casting. J Mater Sci 1989; 24: 2697–2703.

15.

Baumli

Sytchev

Kaptay

. Perfect wettability of carbon by liquid aluminium achieved by a multifunctional flux. J Mater Sci 2010; 45: 5177–5190.

16.

Juhasz

Baumli

Kaptay

. Fabrication of carbon fibre reinforced, aluminium matrix composite by potassium iodide (KI) – potassium hexafluoro-titanate (K 2TiF 6) flux. Mater Werkst 2012; 43: 310–314.

17.

Blucher

Narusawa

Katsumata

et al.

Continuous manufacturing of fiber-reinforced metal matrix composite wires — technology and product characteristics. Compos Part A 2001; 32: 1759–1766.

18.

Deming

Xinfang

Jin

. Continuous yarn fibre-reinforced aluminium composites prepared by the ultrasonic liquid infiltration method. J Mater Sci Lett 1993; 12: 252–253.

19.

Massalski

Okamoto

. Binary alloy phase diagrams. ASM International (ed), 2nd ed. Materials Park, Ohio: ASM International, 1990.

20.

Even

. Conception, élaboration et caractérisation d’un matériau composite à matrice de titane renforcée par des fibres continues de carbone, France: Université Sciences et Technologies – Bordeaux I, 2000.

21.

Even

Arvieu

Quenisset

. Powder route processing of carbon fibres reinforced titanium matrix composites. Compos Sci Technol 2008; 68: 1273–1281.

22.

Sanguinetti Ferreira

Yadava

Arvieu

et al.

Matrix consolidation mechanism in 1D-Ti/SiC/C composites produced by continuous binder-powder coating. Mater Res 2013; 16: 310–314.

23.

Sanguinetti Ferreira

Arvieu

Guillaume

et al.

Titanium matrix composites processed by continuous binder-powder coating: an alternative fabrication route. Compos Part A 2006; 37: 1831–1836.

24.

Nicolaou

Semiatin

Piehler

. Design of metal-matrix composite consolidation practices based on the foil/fiber/foil approach. Scr Metall Mater 1995; 32: 57–62.

25.

Goetz

Kerr

Semiatin

. Modeling of the consolidation of continuous-fiber metal matrix composites via foil-fiber-foil techniques. J Mater Eng Perform 1993; 2: 333–340.

26.

Watt GC, Kyle-Henney S, Crocombe AD, et al. Tensile stress-strain behaviour of multiaxial metal matrix composites. In: 20th International Conference on Composite Materials, 19–24 July 2015, Copenhagen, Denmark.

27.

Anderson

Foley

. Determining the role of surfaces and interfaces in the powder metallurgy processing of aluminum alloy powders. Surf Interface Anal 2001; 31: 599–608.

28.

Liu

Sercombe

Schaffer

. The effect of particle shape on the sintering of aluminum. Metall Mater Trans A 2007; 38: 1351–1357.

29.

Schaffer

Hall

. The influence of the atmosphere on the sintering of aluminum. Metall Mater Trans A 2002; 33: 3279–3284.

30.

Lumley

Sercombe

Schaffer

. Surface oxide and the role of magnesium during the sintering of aluminum. Metall Mater Trans A 1999; 30: 457–463.

31.

Voisin

Monchoux

Hantcherli

et al.

Microstructures and mechanical properties of a multi-phase β-solidifying TiAl alloy densified by spark plasma sintering. Acta Mater 2014; 73: 107–115.

32.

Yan

Imai

et al.

The critical role of heating rate in enabling the removal of surface oxide films during spark plasma sintering of Al-based bulk metallic glass powder. J Non-Cryst Solids 2013; 375: 95–98.

33.

Bonifacio

Holland

Van Benthem

. Evidence of surface cleaning during electric field assisted sintering. Scr Mater 2013; 69: 769–772.

34.

Munir

Anselmi-Tamburini

Ohyanagi

. The effect of electric field and pressure on the synthesis and consolidation of materials: a review of the spark plasma sintering method. J Mater Sci 2006; 41: 763–477.

35.

Sweet

Brochu

Hexemer

Jr et al.

Microstructure and mechanical properties of air atomized aluminum powder consolidated via spark plasma sintering. Mater Sci Eng A 2014; 608: 273–282.

36.

Anselmi-Tamburini

Garay

Munir

. Fundamental investigations on the spark plasma sintering/synthesis process: III. Current effect on reactivity. Mater Sci Eng A 2005; 407: 24–30.

37.

Xie

Ohashi

Chiba

et al.

Frequency effect on pulse electric current sintering process of pure aluminum powder. Mater Sci Eng A 2003; 359: 384–390.

38.

Ghasali

Pakseresht

Alizadeh

et al.

Vanadium carbide reinforced aluminum matrix composite prepared by conventional, microwave and spark plasma sintering. J Alloys Compd 2016; 688: 527–533.

39.

Ghasali

Pakseresht

Rahbari

et al.

Mechanical properties and microstructure characterization of spark plasma and conventional sintering of Al–SiC–TiC composites. J Alloys Compd 2016; 666: 366–371.

40.

Lalet

Kurita

Miyazaki

et al.

Microstructure of a carbon fiber-reinforced aluminum matrix composite fabricated by spark plasma sintering in various pulse conditions. J Mater Sci 2014; 49: 3268–3275.

41.

Ueno

Yoshioka

Sato

et al.

Fabrication of high thermal conductive aluminum/ graphitic fiber composites by pulsed electric current sintering. J Jpn Soc Powder Powder Metall 2007; 54: 595–600.

42.

Pakseresht

Javadi

Bahrami

et al.

Spark plasma sintering of a multilayer thermal barrier coating on Inconel 738 superalloy: microstructural development and hot corrosion behavior. Ceram Int 2016; 42: 2770–2779.

43.

Bushby

Scott

. Evaluation of aluminium-copper alloy reinforced with pitch-based carbon fibres. Compos Sci Technol 1997; 57: 119–128.

44.

Anselmi-Tamburini

Gennari

Garay

et al.

Fundamental investigations on the spark plasma sintering/synthesis process: II. Modeling of current and temperature distributions. Mater Sci Eng A 2005; 394: 139–148.

45.

Voisin

Durand

Karnatak

et al.

Temperature control during Spark plasma sintering and application to up-scaling and complex shaping. J Mater Process Technol 2013; 213: 269–278.

46.

Muñoz

Anselmi-Tamburini

. Temperature and stress fields evolution during spark plasma sintering processes. J Mater Sci 2010; 45: 6528–6539.

47.

Yu-Cheng

Zheng-Yi

Wei-Ming

et al.

Temperature field distribution in Spark Plasma Sintering of BN. J Wuhan Univ Technol Mater Sci Ed 2002; 17: 19–21.

48.

Guo

Derby

. Fibre uniformity and cavitation during the consolidation of metal-matrix composite via fibre-mat and matrix-foil diffusion bonding. Acta Metall Mater 1993; 41: 3257–3266.

49.

Gay D. Composite materials: design and applications, 3rd ed, Taylor and Francis Group, US: CRC Press.

50.

Steffens

Reznik

Kruzhanov

et al.

Carbide formation in aluminium-carbon fibre-reinforced composites. J Mater Sci 1997; 32: 5413–5417.

51.

Hadisaroyo

Batsale

Degiovanni

. Un appareillage simple pour la mesure de la diffusivité thermique de plaques minces J. Phys III 1992; 2: 111–128.

52.

Nagae

Yokota

Nose

et al.

Effects of pulse current on an aluminum powder oxide layer during pulse current pressure sintering. Mater Trans 2002; 43: 1390–1397.

53.

Xie

Ohashi

Song

et al.

Behavior of oxide film at the interface between particles in sintered Al powders by pulse electric-current sintering. Metall Mater Trans A 2003; 34: 699–703.

54.

Räthel

Herrmann

Beckert

. Temperature distribution for electrically conductive and non-conductive materials during field assisted sintering (FAST). J Eur Ceram Soc 2009; 29: 1419–1425.

55.

Wei

Zhang

Shen

et al.

Simulation of temperature and stress distributions in functionally graded materials synthesized by a spark plasma sintering process. Comput Mater Sci 2012; 60: 168–175.

56.

Wang

Cheng

Zhao

. FEM analysis of the temperature and stress distribution in spark plasma sintering: modelling and experimental validation. Comput Mater Sci 2010; 49: 351–362.

57.

Vanmeensel

Laptev

Hennicke

et al.

Modelling of the temperature distribution during field assisted sintering. Acta Mater 2005; 53: 4379–4388.