The effects of ZrSi2 on the ablation and insulation performances of low-density carbon-phenolic composites

Carbon-phenolic (C-Ph) composites are typical ablative thermal protection materials. Excellent ablation and insulation performance indicate a decreased thickness of the thermal shield. Thus, ZrSi2 particles were introduced to improve the performance of the low-density C-Ph composite. An oxyacetylene flame torch was used to examine the ablation and insulating characteristics. The curing process of the matrix resin was not affected by ZrSi2. The thermal conductivities of the composites with different ZrSi2 contents ranged from 0.219 ∼ 0.254 W K−1·m−1. A continuous but not very compact cover was formed on the ablating surface of the C-Ph composite with 10% ZrSi2, which limited the escape of the charred matrix. The passageways for the gaseous products of the charring phenolic compounds were also not impeded by this cover. The C-Ph composite with 10% ZrSi2 exhibited the lowest linear loss rate (0.0081 mm s−1) and the best heat insulating performance.


Introduction
Thermal protection systems (TPSs) are designed to protect high-velocity space vehicles from severe aerodynamic heating, which usually occurs when TPSs enter the atmosphere [1][2][3][4].Ablative thermal protection materials, due to their stability and reliability, are widely used for spacecraft [5][6][7][8][9].Phenolic resin (PR) is an irreplaceable material for ablative thermal protection due to its outstanding properties, such as good thermal stability, high char yield and excellent flame resistance [10].
Numerous attempts have been made to further improve the ablation performance of phenolic resins since better ablation performance means greater reliability.Other phases were employed to enhance the ablation property of the phenolic resin, and the carbon fibre-modified phenolic (carbon-phenolic) composite was a typical representative.However, carbon fibres and charred phenolics are sensitive to the oxidizing atmosphere; hence, the loss of mass from carbon-phenolic (C-Ph) composites is very obvious.Many researchers have attempted to enhance its ablation resistance, and the introduction of inorganic particles to C-Ph composites, for example, ZrO 2 , SiO 2 , ZrB 2 , MoSi 2 , ZrC, TaSi 2, B 4 C, TiB 2 , SiC and ZrSi 2 [11][12][13][14][15][16][17][18][19], is effective.
Conventional C-Ph materials do not exhibit excellent thermal insulation at high heat fluxes because of its high thermal conductivity, which resulted from its high density, and the low-density C-Ph composite (lower than 0.5 g cm −3 ) was applied for heat fluxes greater than 400 W cm −2 .Some previous works reported that the oxidation of ZrSi 2 to ZrO 2 and SiO 2 could enhance the ablation resistance of conventional C-Ph materials.However, the influences of ZrSi 2 on the properties of low-density C-Ph composites are not clear, and the main ablation mechanisms are not clearly discussed.
Therefore, this paper aimed to investigate the effects of different amounts of ZrSi 2 on low-density C-Ph composites and analyse the ablation mechanisms.An oxyacetylene flame has oxidizing and extreme ablation properties, so this traditional method was used to examine the ablation and insulation performance of a low-density C-Ph composite modified with ZrSi 2 .
In the present work, the linear ablation rate was the change in the thickness divided by the ablation time, and the back surface temperatures during ablation were monitored to inspect the heat insulation performance.
Finally, the microstructure after ablation was observed to understand the mechanism behind the improvements in ablation performance.

Experiment
The three-dimensional carbon fibre fabric (0.185 g cm −3 ) was impregnated with a mixture for 30 min, and then the fabric was dried at room temperature for 24 h.The mixture used for the impregnation process contained ethanol, phenolic compounds and ZrSi 2 powders.The impregnated fabric was cured at 80 °C for 0.5 h, at 110 °C for 1.5 h and at 150 °C for 3 h in an oven, and the heating rate was 1 °C/min.The low-density carbon-phenolic composites with different ZrSi 2 contents were named sample a, sample b, sample c, sample d, sample e, sample f and sample g, and the ZrSi 2 content and densities of the obtained composites are given in table 1.
The cured composite was chipped into a cylinder with a Φ of 25 mm × 20 mm.The microstructure was determined with scanning electron microscopy (SEM) (Philips FEI Sirion, Holland).The flow rates of oxygen and acetylene were 500 l h −1 and 400 l h −1 in the ablation process, and the oxygen-acetylene gun with Φ 2 mm was perpendicular to the surface of the specimen at a distance of 30 mm.The flame ablation process is presented in figure 1, and the ablation time was 120 s.The back surface temperature was measured with an experimental Ktype thermocouple, and all temperature data were recorded by a computer.In addition, the thermal conductivity was measured by a steady-state plate device (DP-PL600, China).

Influence of ZrSi 2 on the curing process of phenolic resin
To explore the influence of ZrSi 2 on the curing process of phenolic resin, the Fourier transform infrared (FT-IR) spectra of cured phenolic materials [20] with different ZrSi 2 contents were analysed with a Nicolet iS50 FT-IR spectrometer device, and the results are shown in figure 2.
The FT-IR spectrum of the cured pure phenolic resin is presented in figure 2(a), and there was an absorption peak at 3394 cm −1 attributed to the presence of a hydroxyl group.The vibration absorption peak at 2973 cm −1 was attributed to methylene.Moreover, the peaks at 1592 cm −1 and 1477 cm −1 were the absorption peaks induced by the double bond vibration in the benzene ring, and the peak at 1376 cm −1 corresponded to the absorption peak of the boron-oxygen bond.Furthermore, there was a peak at 1222 cm −1 ascribed to the phenolic hydroxyl group.The peaks at 1098 cm −1 and 753 cm −1 corresponded to the vibration absorption peak of the ether bond and the vibrational absorption peak of the phenol substituted by boron in the ortho-position of the aromatic ring, respectively.As shown in figure 2, the FT-IR spectra of cured phenolic compounds with  different ZrSi 2 contents were all identical to those of the cured phenolic compounds.These results indicated that the curing reactions and processing of phenolic resin were not affected by ZrSi 2 .

Thermal properties
The thermal conductivity was a main factor influencing the thermal insulation performance (table 2).As displayed in table 2, the values were all between 0.219 ∼ 0.254 W K −1 •m −1 with increasing ZrSi 2 content, implying that the thermal conductivity of the C-Ph composites was hardly influenced by the amount of ZrSi 2 .

Ablation properties
The linear loss rate revealed the ablation resistance of the ZrSi 2-modified C-Ph composites, and figure 3 shows these properties.As shown in figure 3, the linear loss rate of the C-Ph composites was 0.0182 mm s −1 , and that of the ZrSi 2-modified C-Ph composites was below this value, which demonstrated that the modified C-Ph composites had better ablation resistance than the unmodified C-Ph composite.A minimum value of 0.0081 mm s −1 was obtained when 10% ZrSi 2 was introduced.When the ZrSi 2 content was less than 10%, the linear loss rate of the modified composites decreased with increasing ZrSi 2 content.When 12% ZrSi 2 was introduced into the C-Ph composite, the linear loss rate was slightly greater than that of the C-Ph composite with 10% ZrSi 2 .

The back surface temperature
The back surface temperatures of the ZrSi 2 modified C-Ph composite during the 120 s ablation process are shown in figure 4. The temperatures of the modified composites were all lower than those of the C-Ph composites when the ablation process lasted more than 70 s.The temperature of the unmodified C-Ph composite quickly increased to 514 °C when the ablation time was 120 s.When the amount of ZrSi 2 was less than 8%, the temperatures of the modified C-Ph composites approached similar levels.The 10% ZrSi 2-modified C-Ph composite showed better thermal insulation, and the temperature increased slowly especially when the flame ablation time exceeded 70 s.In addition, the temperature of this composite was lower than 335 °C during the 120 s ablation process, which was the minimum value.The C-Ph composites modified with 12% ZrSi 2 also revealed good heat shielding during the initial ablation process.However, when the ablation lasted more than 70 s, its back surface temperature was higher than that of the 10% ZrSi 2-modified C-Ph composites.

The ablated surface topography
The ablated surfaces of the composites with different ZrSi 2 contents were analysed, as shown in figure 5.The ablated surface of the C-Ph composite was black, and there were some holes on the surface.When the ZrSi 2 powders were introduced into the C-Ph composite, some white substances were observed on the ablated surface, and the sizes of the existing holes were smaller than those of the sample without ZrSi 2 powders.A 'coat' formed on the ablated surface when the ZrSi 2 content increased to 8% (figure 5(e)).This 'coat' was uniform and continuous when the content was 10% (figure 5(f)).However, a continuous but uneven 'coat' was formed with partial exfoliation at the edge when the ZrSi 2 amount was 12% (figure 5(g)).

Microstructure of the ablated surface
The ablated surface microstructure of the composites was investigated by SEM, and the microstructures of the modified C-Ph composites with less than 6% ZrSi 2 are displayed in figure 6.As shown in figure 6(a), there was little charring matrix between the carbon fibres, and the tips of some fibres had needle-like shapes.Moreover, there were many cavities on the fibres.The C-Ph composite with 2% ZrSi 2 exhibited a similar appearance (figure 6(b)).The ends of some fibres were sharper, and a small amount of matrix was found among the fibres.A few spherical particles were attached to the fibres after 4% ZrSi 2 was added (figure 6(c)), and the tips of more fibres evolved into needle shapes.The C-Ph composite with 6% ZrSi 2 (figure 6(d)) exhibited a similar structure, as shown in figure 6(c).
To confirm the chemical composition of the 'coat' mentioned above, chemical analysis was performed.The energy dispersive spectroscopy (EDS) results of the ablated surface of the composite with 6% ZrSi 2 are shown in

Discussion
An SEM image of the manufactured C-Ph composite is shown in figure 9.The matrix filled the spaces between the carbon fibres and was tightly bonded to the fibre.When the C-Ph composite was tested by an oxyacetylene flame, the phenolic resin converted to char, CH 4 , CO, H 2 , etc [21,22].The charred resin exhibited a honeycomb structure, as shown in figure 10.
The charred matrix and carbon fibre fabric were the different forms of carbon atoms; consequently, they were all strongly oxidized during the oxyacetylene ablation process.The honeycomb matrix had more contact area with the oxidizing atmosphere, which resulted in a high oxidation rate.Accordingly, this high oxidation rate heavily reduced the strength of the matrix.Thus, more matrix was stripped off by the shear force of the oxyacetylene flame flow.Because of the absence of a matrix, only fibres were observed on the ablated surface of the C-Ph composite.The cross-section of the ablated surface is shown in figure 11.A small amount of matrix far from the ablative surface was found, and at a certain depth (indicated in red), there were almost only carbon fibres.Hence, in the C-Ph composite, ablation occurred not only on the surface but also in the interior, and the ablation rate of the matrix was greater than that of the fibre matrix.
The white ZrO 2 resulting from the grey ZrSi 2 was the white substance present on all the ablated surfaces (as shown in figures 5(b)-(g)).The melting point of SiO 2 is approximately 1600 °C, and its boiling point is 2230 °C.The maximum temperatures of all the ablated surfaces monitored by the E1RH-F2-L-0-0 device ranged from 1700 °C to 2000 °C.As a consequence, the SiO 2 was liquid during the ablation process, and the solid ZrO 2 particles (with a boiling point of 2680 °C) were adhered together and bonded to the surface by this viscous liquid.SiO 2 liquid also existed in the holes of the C-Ph composite because of its fluidity.Therefore, the oxyacetylene flame needed to pass through more heat harriers, such as solid ZrO 2 and liquid SiO 2, to reach the interior.Thus, the solid ZrO 2 and liquid SiO 2 slowed the interior ablation.
For the modified C-Ph composite with less than 6% ZrSi 2 (in figures 5(b)-(d)), some of the surface area was exposed to the oxidizing flame due to the lack of a protective cover formed by ZrO 2 and SiO 2, resulting in ZrSi 2 .The oxidizing oxyacetylene flame entered the interior of the composite, and the ablation rate of the matrix was  greater than that of the fibre fabric.Hence, the introduction of less than 6% ZrSi 2 into the C-Ph composite slightly decreased the ablation loss.
As shown in figures 5(e)-(g), when the amount was greater than 8%, the solid ZrO 2 cooperated with the liquid SiO 2 to form a continuous coating.The escape of the stripped-off matrix was limited by this coating, which was similar to a 'cage'.Consequently, the time it took for the matrix to withstand the ablated flame increased.In addition, the fibres also experienced a longer ablation time because a greater amount of matrix could hold them in place.The coating was also similar to an oxygen barrier through which the oxyacetylene flame needed to diffuse to oxidize the interior substances.Some fibres (marked in figures 8(a), (b)) were thinner, and more substances (matrix and fibres) were present in figures 8(a)-(c); both provided some support for the 'cage' statement.When 10% ZrSi 2 particles were introduced, a coating with the optimized thickness (figure 5(f)) was obtained, and the SEM image is shown in figure 12.This coat was continuous but not very compact.The cross section of the ablated surface of the composite with 10% ZrSi 2 is shown in figure 13.A greater amount of matrix next to the cover coat was observed, and the ablation rate of the matrix was almost identical to that of the fibre fabric.
When the ZrSi 2 content was 12%, a too-thick cover was formed, which impeded the passageways for the oxidizing flame and for the gaseous products of the charring phenolic compound.These passageways were very important for the release of gaseous products.The obstruction of interior gaseous product release resulted in too high of an inner pressure.The gaseous products may break through the coating when the inner pressure exceeds the maximum pressure that the coating can bear.The open-hole shell shown in figure 8(c) might be evidence of the release of gaseous substances.In addition, this breakthrough resulted in an uneven cover (as shown in figure 5(g)).The cross-section of the ablated surface is shown in figure 14.The size of the matrix in the position where it protruded out was less than that of its neighbours, which resulted from the breakage of the above protective cover.
Thermal insulation is another important property of thermal protection materials.Thus, the influence of ZrSi 2 on this performance is discussed below.
The ZrSi 2 particles used in the C-Ph composite were approximately 2.5 μm in size and were surrounded by a resin matrix.The continuous matrix was the main heat shield substance through which the thermal energy  needed to pass during the thermal conductivity measurement.Hence, the dispersed ZrSi 2 with a thermal conductivity of approximately 23 ∼ 27 W mK −1 did not noticeably increase the thermal conductivity of the composite (as revealed in table 2).The thermal conductivity constant was obtained by the steady-state method, and an oxyacetylene flame was used to simulate the real unsteady application environment of the C-Ph composite.The back surface temperature reflects the actual heat-insulating performance during the ablation process.Thus, the thermal conductivity, linear loss rate and ablation resistance behaviour all significantly influenced the back surface temperature.
A better ablation resistance indicates a lower loss rate of the heat shield material.From the results in figure 3, it can be inferred that more heat shield substances were retained on the ablating surface of the ZrSi 2 -modified C-Ph composite; thus, the thermal energy needed to pass through a greater shield thickness to reach the back surface.In addition, SiO 2 might obviously reduce thermal transmission [23] since the thermal conductivity of SiO 2 is less than 1.4 W mK −1 .
The back temperature results confirmed that ZrSi 2 could improve the actual thermal insulating property of the C-Ph composite, which was most critical for thermal protection systems (TPSs), and 10% was also the optimum amount for this property.

Conclusion
The carbon fibre fabric was impregnated with a mixture to manufacture a ZrSi 2-modified C-Ph composite with low density.These results indicated that the matrix curing process was not affected by ZrSi 2 and that the thermal conductivity of the composites was also hardly influenced.
When the amount of ZrSi 2 was insufficient, the partial areas of the surface were exposed to the oxidizing flame due to the lack of a protective cover formed by ZrO 2 and SiO 2 , resulting in ZrSi 2 .Accordingly, the oxidizing flame could reach the interior of the composite, and the ablation rate of the matrix was greater than that of the fibre fabric.When the ZrSi 2 content was 8%, a continuous 'coat' was formed, which reached the optimized thickness when the content was increased to 10%.The 'cage'-like coating limited the escape of the charred matrix and did not obstruct the release of the gaseous products from the pyrolysis of the phenolic resin.The C-Ph composite with 10% ZrSi 2 displayed the lowest linear loss rate (0.0081 mm s −1 ), and the ablation rate of the resin matrix was almost identical to that of the fibre fabric.When the ZrSi 2 content was 12%, a too-thick coating was formed, which impeded the passageways for the oxidizing flame and for the gaseous products of the charring phenolic compound.
The back surface temperature results demonstrated that the heat-insulating performance of the C-Ph composite was enhanced by ZrSi 2 .
The C-Ph composite with 10% ZrSi 2 exhibited optimized ablation resistance properties and heat insulation performance.This work provided an effective way to improve the performance of low-density C-Ph composites.

Figure 3 .
Figure 3.The linear loss rates of modified C-Ph composites with different ZrSi 2 contents.

figure 7 .
figure 7. Zr, O, and Si were detected on the surface of the charred matrix (figure 7(a)), and they also existed on the fibres (figure 7(b)).It can be confirmed that ZrO 2 and SiO 2 were the oxidation products of ZrSi 2 .In addition, different quantities of ZrO 2 and SiO 2 were observed on the ablated surface of all the ZrSi 2 -modified C-Ph composites (figures 5(b)-(g)).The microstructures of the composites with more than 8% ZrSi 2 after being ablated are presented in figure 8.As shown in figure 8(a), the fibres and the matrix in the C-Ph composite with 8% ZrSi 2 were covered with many globular particles.The diameters of some fibres were obviously reduced.More globular particles adjoined each other, forming a continuous 'coat' layer in the C-Ph composite with 10% ZrSi 2 (in figure 8(b)).Figure 8(c) shows the microscopic structure of the 12% ZrSi 2 -modified C-Ph composite.A large shell with an open hole was observed in the 'coat' layer.

Figure 8 (
figure 7. Zr, O, and Si were detected on the surface of the charred matrix (figure 7(a)), and they also existed on the fibres (figure 7(b)).It can be confirmed that ZrO 2 and SiO 2 were the oxidation products of ZrSi 2 .In addition, different quantities of ZrO 2 and SiO 2 were observed on the ablated surface of all the ZrSi 2 -modified C-Ph composites (figures 5(b)-(g)).The microstructures of the composites with more than 8% ZrSi 2 after being ablated are presented in figure 8.As shown in figure 8(a), the fibres and the matrix in the C-Ph composite with 8% ZrSi 2 were covered with many globular particles.The diameters of some fibres were obviously reduced.More globular particles adjoined each other, forming a continuous 'coat' layer in the C-Ph composite with 10% ZrSi 2 (in figure 8(b)).Figure 8(c) shows the microscopic structure of the 12% ZrSi 2 -modified C-Ph composite.A large shell with an open hole was observed in the 'coat' layer.

Figure 9 .
Figure 9. SEM of the manufactured C-Ph composite.

Figure 10 .
Figure 10.SEM image of the charred matrix.

Figure 11 .
Figure 11.Cross-section of the ablated surface.

Table 1 .
ZrSi 2 content and densities of the obtained composites.

Table 2 .
The thermal conductivity values of C-Ph composites with different ZrSi 2 contents (%).