Design and Validation of Thermal Insulation for Deep-sea Fluid Sampler of Jiaolong Human Occupied Vehicle

In this paper, a thermal insulation structure with silica aerogel felt as filler material was designed for the requirements of deep-sea fluid thermal insulation sampling technology for Jiaolong human occupied vehicle. Simulation analysis of thermal insulation performance was carried out and an experimental prototype was developed for the thermal insulation structure. Experimental study on thermal insulation performance was conducted with the variation characteristics of the operation environment for Jiaolong human occupied vehicle being taken into account. Results show that the silica aerogel felt with a thickness of 30 mm filled in the radial space between the inner and outer cylinders can achieve the expected thermal insulation effect during the diving-sampling-transferring process, with maximum temperature rise of 8.5 °C, and can meet the requirements of deep-sea fluid thermal insulation sampling technology.


Introduction
Investigation on submarine cold-seep ecological communities has become a hot spot in geoscience and life science research since microbial communities were discovered in deep-sea extreme environment. As deep-sea cold-seep fluids contain rich information, collecting and analyzing their information on physical and chemical properties, and microbial composition, etc. is of significance for promoting the exploitation of deep-sea fluid microbial resources and deep-sea mineral resources [1][2][3][4][5][6]. To obtain fluid samples from deep-sea extreme environment, deep-sea samplers of many types have been developed in China and other countries. Deep diving equipment carries such a sampler and dives to a specified depth so that the sampler conducts sampling. However, at present, most samplers for deep-sea cold seeps have only realized pressure holding function, and the influences of temperature variation on the physical and chemical properties and microbial activities of sample during sampling and transferring have not been taken into account. Because of the activity and later cultivation value of the sampled organisms, the temperature under thermal insulation is generally controlled in the range of 2-8 ℃, with maximum of no more than 10 ℃. The passive thermal insulation mode of filling thermal insulation material is preferred [7][8][9]. In this paper, a high-fidelity sampler with sampling, withdrawing and transferring functions integrated together and with maximum operation water depth of 5000 m, which could realize thermal insulation, pressure holding, and sample dynamic characteristics monitoring, was developed for the requirements of deep-sea cold-seep sampling thermal insulation technology. It could be carried by Jiaolong human occupied vehicle and operated after the vehicle dived. The thermal insulation performance of the The sampler designed in this paper was planned to be carried by Jiaolong human occupied vehicle. During installing-diving-sampling-withdrawing-transferring, the external operation environmental conditions changed constantly, and the temperature and operation time varied with the operation environmental conditions. Therefore, the heat exchange process between the sampler and the external environment needed to be studied according to the variations in external temperature and operation time. Figure 2 shows the curves of seawater temperature and salinity versus water depth in a sea area. (1) The sampler was installed and fixed on the deck 30 minutes before diving; (2) It took 30 minutes to arrange the deep-diving submersible on the sea surface; (3) The deep-diving submersible started to dive half an hour after the sea surface state was confirmed; (4) It took ~1.5 hours for the deep-diving submersible to dive to water depth of 4 000-5 000 m. And then, it operated under the sea for 6 hours. Afterwards, it took ~1.5 hours for the deep-diving Total time 780 Difference operation states were delimited according to the operation processes of the deep-diving submersible and the external environmental temperature variation. The temperature variation was not even under different operation states, so the maximum temperature value under each state was taken as the temperature in corresponding time under this state.

Design of Thermal Insulation for Fluid Sampler
Due to long period and high thermal insulation performance requirement for deep-sea sampling, the novel deep-sea cold-seep thermal insulation and pressure-holding sampler designed in this paper was mainly composed of four parts: pressure-resistant structure, thermal insulation filler material, nitrogenfilling pressure-holding device, and data acquisition system. The schematic diagram of assembly of the sampler is shown in Figure 3. The pressure-resistant structure mainly included inner and outer pressure-bearing cylinder bodies and sample-entering needle valve body. The radial thickness of the inner and outer cylinder bodies was 30 mm, and the axial thicknesses between inner and outer cylinder bodies were 122 mm, and 100 mm, respectively, on both sides [10][11][12][13][14]. The thermal insulation filler material was primarily silica aerogel felt, which had such advantages as small density (146 kg m -3 ), low thermal conductivity (0.022 W m -1 K -1 ), and small specific heat capacity (502.42 J kg -1 K -1 )) [15]. It was filled between inner and outer cylinders, to reduce the heat transfer between the housing and the inner. In addition, to further improve the thermal insulation effect, the surfaces of the inner and outer cylinders were coated with thermal insulation layers.

Simulation Study on Thermal Insulation Performance of Fluid Sampler
To analyze the effectiveness of design of the thermal insulation device, simulation study on the thermal insulation performance of the sampler was carried out according to the external temperature variation under actual operation conditions.

Three Dimensional Model of Thermal Insulation Material Filling
A three dimensional (3D) model of thermal insulation material filled structure was established using 3D design software, as shown in Figure 4. The employed thermal insulation material was silica aerogel felt, and its 3D dimensions are shown in Figure 4.

Simulation Analysis and Results
The established 3D model of thermal insulation material was imported into the transient thermal analysis module of the Ansys Workbench software for simulation. The temperature variation sub-steps were set at different stages according to the actual external environmental temperature variation, as shown in Figure 5. Stage 1 was that the simulated sampler was at initial temperature of 2 ℃ in the low temperature chamber. Stage 2 was that the simulated sampler was transferred from the low temperature chamber to room temperature state. Stage 3 was that the simulated sampler was exposed to the air at 50 ℃ on the deck. Stages 4-10 were that the simulated sampler dived with Jiaolong, operated, and then floated up to the sea surface. Stage 11 was that the simulated sampler was   Figure 6 was obtained through simulation. The upper curve is the set external environmental temperature variation curve, and the lower one is the internal sample temperature variation curve under the influence of the external environment. After the heating was over, the internal sample temperature reached its maximum. The internal sample temperature distribution contour diagram (Figure 7) after the heating was over was obtained accordingly. The maximum temperature of sample was 7.4971 ℃, meeting the design requirement that the maximum temperature rise shall not be more than 10 ℃.

Experimental Study on Thermal Insulation for Sampler
Based on the simulation study results, we further fabricated an experimental prototype of thermal insulation structure of sampler, of which the structural parameters were the same as the model in Figure 3. The pressure-resistant shell of the experimental prototype was made of TC4 titanium alloy, and silica aerogel felt (density 146 kg m -3 , thermal conductivity 0.022 W m -1 K -1 ) was used as filler material for thermal insulation, as shown in Figure 8. The inner core of the prototype was equipped with a seawater container with a volume of 150 mL, in which a Pt100 thermal resistor temperature sensor was provided, to acquire the internal seawater temperature in a real time manner. Based on the operation flow of Jiaolong and the seawater temperature variation, the external temperature variation during the experiment was designed rationally, to carry out experimental study on thermal insulation.

Main Apparatus
The data from temperature sensor was acquired by an M-7018 series analog conversion module, and a B-TH-1000C-S high-low temperature tester was used for simulation of high and low temperature environments.

Experimental Study on Thermal Insulation under Constant Temperature Condition
Thermal insulation experiments were carried out at 40℃ 4 h, 45 ℃ 4 h and 50 ℃ 3 h, respectively, with 150 mL of seawater as sample. Before the heating experiment, the sampler was cooled down to the specified temperature, which was kept constant, to ensure that the temperature was the same in the whole sampler. During the heating experiment, the temperature variation data was extracted once per 5 min. Temperature variation curves were plotted based on the extracted data, as shown in Figures 9-11.    Figure 11. Seawater temperature variation curve at 50 ℃, 3 h (rise of 4.92 ℃), and variation curve of temperature difference between adjacent time points. It can be discerned from the three groups of temperature variation curves under different experimental conditions that, within 1 h in the early stage of constant temperature heating, the curve tended to flat, and the seawater temperature rise was very small, only ~0.5 ℃. With the increase in heating time, the temperature variation curve went up continually and the curve slope increased constantly, indicating increased temperature rise. It can be seen from the three groups of variation curves of temperature difference between adjacent time points under different experimental conditions that, within 30 min in the early stage of heating, the variation in temperature difference between adjacent time points was very small, basically ~0.05 ℃. After ~3 h of heating, the variation in temperature difference between adjacent time points tended to stable, ranging from 0.15 to 0.2 ℃. From then on, external heat was constantly and rapidly transferred to the sample in the inner cylinder, enhancing the rise in sample temperature. Therefore, the operation time under relatively high external temperature condition should not be too long, and the time of exposure of the sampler to external environment at 50 ℃ should be decreased as far as possible. The times for temperature to rise to different degrees were calculated further for three groups of data, as shown in Table 2. It can be seen from Table 2 that, in the first ~30 min, the temperature rise was 0.1 ℃; therefore, it could be thought that the temperature did not change in this stage. When the heating reached ~50 min, the temperature reached 5% of total temperature rise in the experiment. When the heating reached ~70 min, the temperature reached 10% of total temperature rise in the experiment. When the heating reached ~155 min, the temperature reached 50% of total temperature rise in the experiment. The time for 50% of temperature growth accounted for 60%-70% of total time, so the other 50% of temperature rise consumed a shorter time during later temperature growth. Analysis showed that the main causes were as follows: (1) it took some time for the heated chamber to be heated to a specified temperature; (2) it took some time for the heat to be transferred to the liquid through the thermal insulation layer; (3) it took some time for the temperature to achieve even distribution in the liquid; (4) with the increase in heating time, the curve of variation in temperature difference between adjacent time points went up constantly till tending to stable. The experiment of sample temperature rise in a certain time under constant external temperature showed that, the temperature rise was not evident in the early stage of heating. With the increase in heating time, the temperature difference between adjacent time points increased continually, and the sample temperature rose constantly. When the temperature difference between adjacent time points tended to stable, the sample temperature rose constantly at a nearly fixed temperature difference, till the heating was over, achieving the maximum temperature rise.

Experimental Study on Thermal Insulation Performance under Actual Operation Conditions
To fully understand the influence of external temperature variation on the temperature rise of sample in the inner cylinder under actual operation conditions, experimental study on thermal insulation was carried out with a high-low temperature tester used to simulate the external temperature variation under actual operation conditions. The temperature rise or drop rate of the high-low temperature tester could be adjusted by setting the temperature slope; however, in the actual operation, the variation in temperature rise or drop rate in unit time was inaccurate. Through repeated no load experiments, the high-low temperature tester was set as shown in Table 3 for simulation of temperature variation under actual external environment. Table 3. Settings of high-low temperature tester for simulation of temperature variation under actual external environment.

Operation environment Temperature (℃) Duration (min)
Temperature of high-low temperature tester rose from 2 ℃ to deck temperature 2→50 10 Sampler was installed on deck and deep-diving submersible was arranged on sea surface 50 30+30 Temperature of high-low temperature tester became the same as that of sea surface 50→25 30 Deep-diving submersible dived from sea surface to underwater 1,000 m 25→5 30 Deep-diving submersible lowered from underwater 1,000 m to 5,000 m, operated at 5,000 m for 6 h, and rose from 5,000 m to underwater 1,000 m 5 532 Temperature of high-low temperature tester became the same as temperature at underwater 500 m 5→10 5 Deep-diving submersible rose from underwater 1,000 m to 500 m 10 10 Temperature of high-low temperature tester became the same as that of sea surface 10→25 6 Deep-diving submersible rose from underwater 500 m to sea surface 25 10 Temperature of high-low temperature tester became the same as that of deck 25→50 8 Sampler was transferred from sea surface to deck, removed on deck, and then transferred to laboratory 50 30+30+30 Total time consumed 792 The high-low temperature tester heating temperature was set according to the actual external environmental temperature variation, and the total time consumed was 13.2 h. Figure 12 shows the temperature variation curve of high-low temperature tester under ideal state. Figure 12 shows the actual temperature variation curve of high-low temperature tester. Comparing Figures 12 with 13, temperature fluctuation occurred in the heated chamber during temperature change, but the temperature variation basically agreed with the set variation state.     It can be discerned from the three groups of experimental data that, in the initial stage of heating, the sample temperature rose slowly, fundamentally maintained at ~2 ℃, and the temperature difference between adjacent temperature points was basically 0. As the sampler was exposed to external operation environment of 50 ℃ for a long time, the sample temperature rose evidently, and the 10 temperature difference between adjacent temperature points started to increase. When the sampler was arranged on the sea surface and dived with the deep-diving submersible, the sample temperature and the temperature difference between adjacent temperature points remained rising. As the deep-diving submersible dived, the surrounding seawater temperature dropped constantly, the temperature difference between adjacent temperature points started to decrease, and the sample temperature rise started to slow down, with the "second highest peak 1" reached. As the deep-diving submersible dived to water depth of more than 3,500 m and conducted underground operation, the sample temperature dropped constantly and the temperature difference between adjacent temperature points became negative value and tended to stable. During the rising of the deep-diving submersible after completing the operation, the sample temperature remained dropping, with the "second lowest valley 2" reached. At this moment, the temperature difference between adjacent temperature points started to increase. In the process that the deep-diving submersible rose to the sea surface and was withdrawn to the deck and the sampler was removed from the deep-diving submersible and transferred to a cultivation kettle, the sample temperature remained rising constantly. After the sampler was transferred to the cultivation kettle, the temperature difference between adjacent temperature points started to decrease. After the sampler was transferred to the cultivation kettle at constant temperature of 2 ℃, the temperature remained rising for a time period, with "the highest peak 3" reached. In the three groups of experimental data, the values of the "second highest peak 1" of sample temperature were 8.19 ℃, 8 ℃, and 7.65 ℃, respectively, those of the "second lowest valley 2" were 5.84 ℃, 5.59 ℃, and 5.52 ℃, respectively, and those of the "highest peak 3" were 8.4 ℃, 8.2 ℃, and 8.5 ℃, respectively. All of these values did not exceed the design indicator required maximum temperature rise of 10 ℃. It can be discerned from the above experiments that, in the experiment of thermal insulation (temperature variation) under the simulated actual operation conditions, the total time consumed was 13.2 h, and the maximum temperature variation was 8.5 ℃. The temperature variations basically met the requirements that the mean temperature variation shall be within 8 ℃ and the maximum temperature rise shall not be more than 10 ℃.

Comparative Analysis between Simulation and Experiment of Thermal Insulation Performance
Comparative analysis was further conducted between the experimental results and the simulation results in this section. Figure 17 shows the temperature variation curves of the sample under simulation and experimental conditions. It can be seen from the sample temperature variation curve obtained from experiment (Figure 17a) that, with the change in external temperature, the sample temperature rose slowly first. With the rise in external heating temperature, the slope of the sample temperature variation curve increased constantly, and the sample temperature rise was sped up. When the external heating temperature dropped gradually and was maintained at 5 ℃, the slope of the sample temperature variation curve became negative value, and the sample temperature started to continually drop. With the external heating