Silica aerogels were fabricated using a polyethoxydisiloxanes (E-40) based recipe which is a solution with ingredient of n(ethanol): n(E-40): n(H₂O): n(HCl)=3:1:0.2:0.1. After mixing, the mixed solution was stirred for 30 min at room temperature and placed in the water bath at 60℃ for 2 h. Silica hydrogels was obtained after completion of hydrolysis and polycondensation reactions. Thereafter, Al₂O₃ fiber mat was impregnated by the hydrogels. Here, the Al₂O₃ fiber mat was prepared viaSol-Gel process. The chemical composition of Al₂O₃ fiber mat is 95%Al₂O₃+ 5%SiO₂. The fiber diameter is about 3 μm to 7 μm, and the length is about 100 mm. To strengthen the gel network, the hydrogel filled mat was immersed into 50vol% ethanol solution at 50℃ to exchange part of water in the hydrogels^{[ 4]}.

After aging, the composite was immersed in hexamethyldisiloxane (HMDSO) for several days to complete solvent exchange, and then proper trimethylchlorosilane (TMSCl) was added for surface modification. Finally, the mat filled by the modified hydrogel was dried at 200℃ with an infrared lamp for 2 h and the silica hydrogel in mat changed into silica aerogels^{[ 5]}.

The structure of traditional MLIMs is shown in Fig. 1. MLIMs composes of repetitive unit, and every unit is consisted of a heat shield with low emissivity and a spacer with low thermal conductivity, the heat shield is made from Polyimide(PI) coated with Al thin film, and spacer is made from Al₂O₃ fiber mat. The traditional MLIMs are marked as Sample I. In this paper, the new MLIMs are marked as Sample II, in which the spacer is made from Al₂O₃ fiber mat and silica aerogels is implanted on the surface of Al₂O₃ fiber. In order to make a comparison between the two samples, we made them into a similar thickness.

Fig. 1

	Figure Option ViewDownloadNew Window
	Fig. 1 The structure of sample I and II

The micro morphology of silica aerogels is analyzed by SEM, and the thermal conductivity is tested by calorimetric emissometer designed according to the steady-statecalori- metric method^{[ 6]}.

The effective thermal conductivity ( λ_e) of MLIMs was measured based on the steady state calorimeter method^{[ 7]}. For a MLIM, its effective thermal conductivity, λ_e is the most critical performance factor. In this work, λ_e is determined according to the cylinder steady-state conduction model, λ_e could be expressed as formula (1)^{[ 8]}:

(1)

Where, r₁, r₂ is the radius of the sample inside and outside (m), L is the length of the state-temperature area (m), T₁, T₂ is the inner and outer temperature of the sample (K) , Q is the power of heat consumption (W).

In this work, the testing equipment has a heat pipe with coatings and heater associated with thermostat. The sample was mounted in the vacuum chamber. To reduce errors, we fetch about 40 mm length located the middle of the sample as measuring area, which will match the requirement of cylinder steady-state conduction model.

Figure 2(a) shows the microstructure of silica aerogel-Al₂O₃ fiber composite. It can be seen that the silica aerogels is a porous structure material. It is proposed that the silica aerogels has 80%-99% open porosity and a three-dimensional solid network like silica framework which changes the heat transfer mode in MLIM. The heat conduction path in the porous material is extended. The mean free path of air molecules is about 70-100 nm (300 K), which is close to the particle size of silica aerogels^{[ 9]}, hence, air molecules could be bound by pores in aerogels. As a result, the thermal conductivity of this new material has a significant reduction.

Fig. 2

	Figure Option ViewDownloadNew Window
	Fig. 2 SEM images of silica aerogels granules (a) and silica aerogels-fiber composites (b)

From Fig. 2(b), it could be seen that a special structure was formed with some silica aerogels clung to Al₂O₃fibers (point A) and parts of silica aerogels filled in the space of Al₂O₃ fibers ( point B ). This particular steric configuration could reduce the heat transfer between fibers. It also could make the material has a certain strength, and other special properties.

The effective thermal conductivity of Sample I and II at different temperatures, was tested, and the air pressure is 10^-3 Pa. Figure 3(a) shows. the λ_e of sample II is always lower than that of sample I. From 300 K to 550 K, the thermal conductivity of sample II changes from 1.27×10^-3 W/(m·K) to 2.24×10^-3 W/(m·K). Such as, when the temperature is 450 K, λ_e of sample I and sample II are 2.25×10^-3and 1.76×10^-3 W/(m·K), respectively. It has 21.8% reduction amplitude.

Fig. 3

	Figure Option ViewDownloadNew Window
	Fig. 3 Thermal conductivity with temperature changes of Sample I and II

Because 100 Pa is a typical air pressure in Near space (20-100 km from the earth surface)^{[ 10]}, thermal conductivities of two samples were tested at this pressure. As shown in Fig. 3(b), Sample II has obviously better performance than Sample I. From 300 K to 550 K, the thermal conductivity of sample II changes from 6.53×10^-3 W/(m·K) to 8.57×10^-3 W/(m·K). At 450 K, λ_eof Sample I is 1.02×10^-2W/(m·K), and it is 7.68×10^-3W/(m·K) for sample II which is reduced by 24.7%. Sample II still shows good insulation performance at low vacuum.

Sample II has better thermal barrier performance than sample I in all the temperature and air pressure because of the silica aerogels. In the experiment, we also found that compared with sample I, the weight of the Sample II is lighter. Sample II has 67.2% reduction amplitude, which is very important for the space science and technologies. It can be clearly seen from above results that silica aerogels pacer has great potential in multilayer insulation field. Under the condition of quantity production, the cost of new MLIMs system can be reduced to about 40 € /m².

At high temperature(>1000 K), PI heat shield is easily melted down, and the structure of silica aerogels is also destroyed. For high temperature test, we prepared a new Sample II with 30 units as shown in Fig. 4 and Table 1. In the sample, the MLIM was divided into two parts, which were high-temp section and low-temp section.

Fig. 4

	Figure Option ViewDownloadNew Window
	Fig. 4 The structure of new MLIMs in high-temp experiment

Table 1

Table 1

Table 1 Two parts of MLIMs used for different temperatures Section Temperature range/K Low-emissivity shield Spacer material Low-temp 300-600 Aluminized polyimide film (PI) ( T=30 µm) Silica aerogel-Al2O3 fiber mat High-temp 600-1300 Ni foil ( T=15 µm) Ceramic fiber blanket

Table 1 Two parts of MLIMs used for different temperatures

In order to predict the thermal performance of the MLIM at high temperature, we developed a method to simulate the temperature response by using the software ANSYS. ANSYS is a kind engineering simulation software based on the Finite Element Method^{[ 11]}.

To simplify the problem, the MLIMs were considered as a uniform thermal conductivity at both two sections. The heating boundary kept constant temperature of 1200 K for 240 s, and the cold border was adiabatic. The background temperature was 293 K. Table 2 shows thickness, density, and specific heat capacity of two samples.

Table 2

Table 2

Table 2 Spacers parameters of two samples Spacers Thickness/mm Density/(kg·m-3) Specific heat capacity/(J·kg-1·K-1) Al2O3 fiber mat 0.5 210 1.04 silica aerogel-Al2O3 fiber composite 0.8 180 0.53

Table 2 Spacers parameters of two samples

As shown in Fig. 5, the cold boundary of traditional sample rose to 300 K while cold boundary of new sample only rose to 294 K. The temperature difference became larger with prolonging the time.

Fig. 5

	Figure Option ViewDownloadNew Window
	Fig. 5 Simulated result of two samples in high-temp test

From the simulation, it can be seen that the new MLIMs shows better thermal barrier properties than traditional MLIMs at high temperature. At 240 s, the cold boundary of new MLIMs keeps at room temperature, which is very important to precision machines normal operation and astronaut’ s life safety.

In order to verify the accuracy of the simulation results, we make an actual test of the new MLIMs. The high-temp section of the sample is contacted a heat source whose temperature is 1200 K, the low-temp section face the black environment and the air pressure is high vacuum (10^-4Pa) in high-temp experiment. To monitor the tem- perature variation of two section of sample, three thermal couples were bonded on the surface of unit in heating surface, low-temp section and the cold boundary.

Figure 6 shows the time dependence of temperature in sample II. Line (1) is the temperature of heating surface, Line (2) is the temperature of high-low temp boundary, and Line (3) is the temperature of cold boundary. It can be found that although the sample is contacted with high temperature heat source, the temperature of the low temp-section keep at room temperature 295 K, which demonstrates the excellent thermal barrier properties of the Sample II. And this experimental data is in accordance with the simulation result above.

Fig. 6

	Figure Option ViewDownloadNew Window
	Fig. 6 The high-temp experiment results of new MLIMs