An experimental investigation of subcooled flow boiling in a large width-to-height-ratio, one-sided heating rectangular mini-gap channel was conducted with deionized water as the working fluid. The super-hydrophobicity micro-porous structured copper surface was utilized in the experiments. High speed flow visualization was conducted to illustrate the effects of heat flux and mass rate on the heat transfer coefficient and flow pattern on the surfaces. The mass fluxes were in the range of 200–500 kg/m 2 s, the wall heat fluxes were spanned from 40–400 kW/m 2 . With increments of imposed heat flux, the slopes of boiling curves for superhydrophobic micro-porous copper surfaces increased rapidly, indicating the Onset of Nucleate Boiling. Heat transfer characteristics were discussed with variation of heat fluxes and mass fluxes, the trends of which were analyzed with the aid of high speed flow visualization.

Fuel cell vehicles (FCVs) are facing more severe heat dissipation challenges since the fuel cell stack is required to operate at a lower temperature and thus smaller heat exchanger temperature difference. Thorough analysis of the parametric effects is required to maximize the thermal performance. In this paper, a numerical analysis of the tube-strip heat exchanger is conducted for the targeted application in a high performance passenger FCV. The representative unit cell is used to model the detailed fluid flow and heat transfer at both hot and cold sides in the theoretical framework of volume averaging. Based on the numerical computation over the representative unit cells, the flow, temperature and pressure fields are obtained, which are then utilized to obtain the cell-level heat transfer coefficient between the hot and cold fluids. The obtained heat transfer coefficients are used for the estimation of the heat exchanger thermal performance based on the effectiveness-NTU method. Different air and liquid water flow rates are first examined. Various design parameters such as fin height, fin spacing, fin thickness and fin material are examined through the heat transfer analysis at the unit cell level. Attention is also paid on the improvement of the air-side performance by changing fin shapes to increase the heat transfer coefficient of the heat exchanger. The result shows that the total exchanged heat of the aluminum louvered fin heat exchanger, with fin thickness of 0.06mm, fin height of 5mm and fin spacing of 1mm, can reach 59.24 kW at the liquid flowrate of 120L/min and air velocity of 5m/s.

Higher energy densities and the potential for nearly instantaneous recharging make microscale fuel cells very attractive as power sources for portable technology in comparison with standard battery technology. Heat management is very important to the microscale fuel cells because of the generation of waste heat. Waste heat generated in polymer electrolyte membrane fuel cells includes oxygen reduction reaction in the cathode catalyst, hydrogen oxidation reaction in the anode catalyst, and Ohmic heating in the membrane. A novel microscale fuel cell design is presented here that utilizes a half-membrane electrode assembly. An ANSYS Fluent model is presented to investigate the effects of operating conditions on the heat management of this microscale fuel cell. Five inlet fuel temperatures are 22°C, 40°C, 50°C, 60°C, and 70°C. Two fuel flow rate are 0.3 mL/min and 2 mL/min. The fuel cell is simulated under natural convection and forced convection. The simulations predict thermal profiles throughout this microscale fuel cell design. The exit temperature of fuel stream, oxygen stream and nitrogen stream are obtained to determine the rate of heat removal. Simulation results show that the fuel stream dominates heat removal at room temperature. As inlet fuel temperature increases, the majority of heat removal occurs via convection with the ambient air by the exposed current collector surfaces. The top and bottom current collector removes almost the same amount of heat. The model also shows that the heat transfer through the oxygen channel and nitrogen channel is minimal over the range of inlet fuel temperatures. Increasing fuel flow rate and ambient air flow both increase the heat removal by the exposed current collector surfaces. Ultimately, these simulations can be used to determine design points for best performance and durability in a single-channel microscale fuel cell.

Heat transport mediated by near-field interaction in particulate system (e.g. chain of particles) is one of the research focuses in thermal transport in micro-nanoscale. Near field radiative heat transfer (NFRHT) characteristics of metallic nanoparticle chains (separation distance between neighboring particles is h ) are analyzed by means of both coupled electric-magnetic dipole approximation and quadrupole approximation. Thermal conductance ( G ) between the central particle and other particle with different separation gaps (Δ x ) is calculated at both 300K and 1200K. Corrected polarizability is used to take quadrupole effect into consideration when calculating the NFRHT in extreme near field where dipole approximation ceases to be valid. Temperature distributions along several different chains of particles due to NFRHT are also predicted. Results show that, according to the asymptotic behavior of distribution of G along metallic chains similar as that observed in SiC chains, heat super-diffusion is demonstrated at both 300K and 1200K in metallic nanoparticle chains. At 300K, the contribution of quadrupole results in that thermal conductance responses to h in different way in metallic and dielectric particle chains. Temperature distribution and heat flow are the two key parameters used to characterize the heat transport in chains of particles. Ag particles in SiC chain act as barriers during the radiative heat transport process. Heat super-diffusion, as well as some other characteristics of NFRHT, observed in metallic nanoparticle chains may help for insight of heat transport in particulate system and new design of device in micro-nanoscale.

Inspired by a few phenomena in nature such as the lotus leaf, red rose petal, gecko’s feet and Nepenthes Alata plant, much attention has been paid to use simple and feasible means to achieve remarkable wetting behaviour for many applications in various areas including self-cleaning for building exteriors and windshields, oil/water separation, anti-icing, liquid collecting, anti-fogging and anti-corrosion. Based on the established theoretical models, wetting behaviour of a liquid droplet obtained by molecular dynamics simulation method is generally in good agreement with the experimental results. In macro and micro scale, the previous theories can explain and predict the wetting behaviors well. However, these theories are invalid for nanoscale. It is essential to reveal the underlying physical mechanism of the wetting behaviors of the droplet on solid surface with nanoroughness. Extensive studies on nanosale wettability focus on the effect of nano structures on wettability state. Desired wetting behavior of rough material surface achieved by nanosize reentrant geometry like “T” or mushroom shape and other variant geometry with solid overhangs has been widely used in self-cleaning surfaces, heat exchange and many applications. For example, “T” shape groove with different depths and widths under nanoscale has been considered to confer superhydrophobicity to hydrophilic surfaces gradually. In this paper, wettability transition of a liquid droplet on geometrically heterogeneous solid substrate with nanoscale structures of inverted triangular grooves is investigated by using molecular dynamics simulation method under the parameter space spanned by structure geometry and solid-liquid molecular interaction potential strength. Three wettability states, namely Cassie nonwetting state, Cassie-to-Wenzel transition state and Wenzel wetting state, are identified with various geometries and potential strength. For Cassie nonwetting state, increasing height of the triangles has less effect on wettability transition with weak solid-liquid molecular interaction. Besides, the Cassie nonwetting state is less sensitive to different interval between the triangles as solid-liquid molecular interaction is weak. For Cassie-to-Wenzel transition state, increasing height of the triangles and decreasing interval between the triangles decrease wettability. For Wenzel wetting state, increasing interval between the triangles with low height increases wettability. With strong solid-liquid molecular interaction, different interval between the triangles results in wetting state transition from Wenzel to transition state. What’s more, liquid droplet changes its state from Wenzel wetting state to Cassie-to-Wenzel transition state with increasing height of the triangles or decreasing interval between the triangles. Three wettability transition regions are identified in the parameter space.

A detailed mathematical model is modified that describes heat and mass transfer through Combined Compact Evaporative Cooler CCEC using mono-scale sintered particles coating. The model accounts for the effects of particles size, Reynold’s number, and channel spacing on system performance for certain operating conditions. In this study, a Kelvin-Clapeyron equation is utilized to obtain numerical data for an evaporating meniscus in the wet channels and compared to the plain surface. The results indicate that porous coatings enhanced evaporative heat and mass transfer in the wet channels significantly, compared to the plain wet channels walls, due to wicking water that covers entire heated surfaces.

Boiling heat transfer is one of the most effective methods to meet the challenge of heat dissipation of high heat flux devices. Wetting hybrid surface has been demonstrated to have better performance than hydrophilic or hydrophobic surface. But this kind of wetting hybrid modification is always carried out on plain or flat surface. In this paper, we coated PTFE powders on superhydrophilic micro copper dendrite fin surface to build wetting hybrid surface. The pool boiling experimental results showed that after coating, the wall superheat dramatically decreased to 8 K, which is 9K lower than that on original surface at 250 W·cm −2 . Besides, it also has much better performance than Silicon pin fin based wetting hybrid surface.

Numerical simulation was performed to establishing a two-dimensional pulsating heat pipe model, to investigate the flow and heat transfer characteristics in the pulsating heat pipe by using the Mixture and Euler models, which were unsteady models of vapor-liquid two-phase, based on the control-volume numerical procedure utilizing the semi-implicit method. Through comparing and analyzing the volume fraction and velocity magnitude of gas phase to decide which model was more suitable for numerical simulation of the pulsating heat pipe in heat and mass transfer research. It was showed there had gas phase forming in stable circulation flow in the heating section, the adiabatic section using the Mixture and Euler models respectively, and they were all in a fluctuating state at 10s, besides, the pulsating heat pipe had been starting up at 1s and stabilizing at 5s, it was all found that small bubbles in the heat pipe coalescing into large bubbles and gradually forming into liquid plugs and gas columns from the contours of volume fraction of the gas phase; through comparing the contours of gas phase velocity, it could be seen that there had further stably oscillating flow and relatively stabler gas-liquid two-phase running speed in the pulsating heat pipe used the Mixture model, the result was consistent with the conclusion of the paper[11] extremely, from this it could conclude that the Mixture model could be better simulate the vaporization-condensation process in the pulsating heat pipe, which could provide an effective theoretical support for further understanding and studying the phase change heat and mass transfer mechanism of the pulsating heat pipe.

Velocity slip and temperature jump at the solid-liquid interface are important phenomena in microchannel heat transfer. A comprehensive mathematical model considering both velocity slip condition and temperature jump at the solid-liquid interface is developed to understand the mechanisms of heat and mass transfer during thin-film evaporation in this paper. The model structure is established based on the lubrication theory, Clausius-Clapeyron equation and Young-Laplace equation. To better formulate the film evaporation process, three dimensionless parameters representing the effects of slip length coefficient, temperature jump and wall superheat degree respectively, are introduced in the present model. The analytical solution provides insight of film thickness and heat transfer characteristics for the evaporating thin film. It shows that as the slip length and temperature jump coefficient decrease, the length of evaporating thin film region is shortened and the location of maximum heat flux moves closer to the initial evaporating point. The effect of slip condition on heat flux is small, but the increase of temperature jump can reduce the peak heat flux significantly. Furthermore, the analysis on the three thermal resistances which are caused by temperature jump, conduction through liquid film and evaporation on liquid-vapor interface result in a better understanding for effective heat transfer during thin-film evaporation.

The heat transfer performance of fluid flowing in a microchannel was experimentally studied, to meet the requirement of extremely high heat flux removal of microelectronic devices. There were 10 parallel microchannels with rectangular cross-section in the stainless steel plate, which was covered by a glass plate to observe the fluid flowing behavior, and another heating plate made of aluminum alloy was positioned behind the microchannel. Single phase heat transfer and fluid flow downstream the microchannel experiments were conducted with both deionized water and ethanol. Besides experiments, numerical models were also set up to make a comparison with experimental results. It is found that the pressure drop increases rapidly with enlarging Reynolds number (200), especially for ethanol. With comparison, the flow resistance of pure water is smaller than ethanol. Results also show that the friction factor decreases with Reynolds number smaller than the critical value, while increases the velocity, the friction factor would like to keep little changed. We also find that the water friction factors obtained by CFD simulations in parallel microchannels are much larger than experiment results. With heat flux added to the fluid, the heat transfer performance can be enhanced with larger Re number and the temperature rise could be weaken. Compared against ethanol, water performed much better for heat removal. However, with intensive heat flux, both water and ethanol couldn’t meet the requirement and the temperature at outlet would increase remarkably, extremely for ethanol. These findings would be helpful for thermal management design and optimization.

This paper presents the effects of heat dissipation performance of pin fins with different heat sink structures. The heat dissipation performance of two types of pin fin arrays heat sink are compared through measuring their heat resistance and the average Nusselt number in different cooling water flow. The temperature of cpu chip is monitored to determine the temperature is in the normal range of working temperature. The cooling water flow is in the range of 0.02L/s to 0.15L/s. It’s found that the increase of pin fins in the corner region effectively reduce the temperature of heat sink and cpu chip. The new type of pin fin arrays increase convection heat transfer coefficient and reduce heat resistance of heat sink.

Two-phase boiling in advanced microchannel heat sinks offers an efficient and attractive solution for heat dissipation of high-heat-flux devices. In this study, a type of reentrant copper microchannels was developed for heat sink cooling systems. It consisted of 14 parallel Ω-shaped reentrant copper microchannels with a hydraulic diameter of 781μm. Two-phase pressure drop characteristics were comprehensively accessed via flow boiling tests. Both deionized water and ethanol tests were conducted at inlet subcooling of 10°C and 40°C, mass fluxes of 125–300kg/m 2 ·s, and a wide range of heat fluxes and vapor qualities. The effects of heat flux, mass flux, inlet subcoolings and coolants on the two-phase pressure drop were systematically explored. The results show that the two-phase pressure drop of reentrant copper microchannels generally increased with increasing heat fluxes and vapor qualities. The role of mass flux and inlet temperatures was dependent on the test coolant. The water tests presented smaller pressure drop than the ethanol ones. These results provide critical experimental information for the development of microchannel heat sink cooling systems, and are of considerable practical relevance.

Absorption air conditioning system could be driven by low grade energy, such as solar energy and industrial exhaust heat, for the purposes of energy conservation and emission reduction. Its development is limited by its huge volume and high initial investment. The nanofluids, which possess the superior thermophysical properties, exhibit a great potential in enhancing heat and mass transfer performance. In this paper, nanofluids of H 2 O/LiBr with Fe 3 O 4 nanoparticles were introduced into absorption air conditioning system. The effects of some parameters, such as the flow rate of H 2 O/LiBr nanofluids, nanoparticle size and mass fraction, on the falling film absorption were investigated. The H 2 O/LiBr nanofluids with Fe 3 O 4 nanoparticle mass fractions of 0.01 wt%, 0.05 wt% and 0.1 wt%, and nanoparticle size of 20 nm, 50 nm and 100 nm were tested by experiment. The results imply that the water vapour absorption rate could be improved by adding nanoparticles to H 2 O/LiBr solution. The smaller the nanoparticle size, the greater enhancement of the heat and mass transfer performance. The absorption enhancement ratio increases sharply at first by increasing the nanoparticle mass fraction within a range of relatively low mass fraction, and then exhibits a slow growing even reducing trends with increasing the mass fraction further. For Fe 3 O 4 nanoparticle mass fraction of 0.05wt% and nanoparticle size of 20nm, the maximum mass transfer enhancement ratio is achieved about 2.28 at the flow rate of 100 L·h −1 .

Chemical vapor deposited (CVD) graphene together with a superior gate dielectric such as Al 2 O 3 , is a promising combination for next-generation high-speed field effect transistors (FET). These high-speed devices are operated under high frequencies and will generate significant heat, requiring effective thermal management to ensure device stability and longevity. It is thus of importance to characterize the interfacial thermal resistance (ITR) between graphene/Al 2 O 3 gate dielectric and graphene/metal contacts. In this work, ITRs across the single-layer graphene/Al 2 O 3 and the graphene/metal (Al, Ti, Au) interfaces were characterized from 100 K to 330 K using the differential 3ω method. Unlike previous works which mostly used exfoliated single or few-layer graphene, we used CVD large-scale graphene, which is most promising for FET fabrication due to cost and quality control, in the experiments. To ascertain the measured results and reduce uncertainty, different sandwich configurations including metal/graphene/metal, Al 2 O 3 /graphene/Al 2 O 3 and metal/graphene/Al 2 O 3 were used for the measurements. The effects of post annealing on different interfaces were also investigated. Measurements of numerous samples showed an average ITR at 300K of 9×10 −8 m 2 K/W for graphene/Al 2 O 3 , 6×10 −8 m 2 K/W for graphene/Al, 5×10 −8 m 2 K/W for graphene/Ti, and 7×10 −8 m 2 K/W for graphene/Au interfaces. For the metal interfaces with graphene, the results are within the same order of magnitude as previous measurement results with graphite. However, ITR for graphene/Al 2 O 3 is one order of magnitude higher than those reported for graphene/SiO 2 interfaces. The measured ITRs for both metal and dielectric interfaces with graphene are almost temperature-independent from 100 K to 330 K, indicating that phonons are the major heat carrier. Annealing was found to have different effects on different interfaces. For graphene/Ti interfaces, ITR results measured before and after annealing consistently show a reduction of around 20%. However, such improvements on interfacial conductance were not observed for graphene/Al, graphene/Au and graphene/Al 2 O 3 interfaces. The reduction of ITR of graphene/Ti interface is perceived to stem from the formation of Ti-C covalent bonds. However, neither the commonly used maximum transmission model nor the diffuse mismatch model explicitly considers bonding effects at the interface, which is why they poorly predict and explain all the aspects of the measurements. An improvement to the classic anisotropic DMM model was proposed by taking into account different bonding types and bonding area between graphene and Al 2 O 3 /metal layer, resulting in a better fitting with the experimental data.

In this article, the heat transferring property of the copper-water nanofluids in self-exciting mode oscillating flow heat pipe under different laser heating power is experimented, as well as is compared with that of the distilled water medium in self-exciting mode oscillating flow heat pipe under same heating condition. The objective of this article is to provide the heat transfer characteristics of Cu-H 2 O nanofluids in self-exciting mode oscillating-flow heat pipe under different laser heating input, and to compare with the heat transfer characteristics of the same heat pipe with distilled water as working fluids. The SEMOS HP used in this experiment is made of brass tube with 2mm interior diameter, which is consisted of 8 straight tubes with 4 turns’ evaporation section and 12 turns’ condensation section. The heat resource for the evaporation zone is eight channel quantum pitfall diode array semi-conductor laser heater with 940nm radiation wave length, while the radiation power of each channel is changeable within 0–50W and the facular size is 1×30mm 2 . The condensation section is set in a cooling water tank in which water is from another higher tank. The actual transferring rate could be calculated by the flow rate of the cooling water and the change of the temperature. The change of the temperature of the heat pipe wall is measured by those thermo-couple fixed in different section in the heat pipe and data is collected by a data acquisition. In the heat pipe the fluid filling rate is 43%, the pressure is 2.5×10 −3 Pa, and the heat pipe inclination angle is 55° while the size of the brass particle in the nanofluids is less than 60nm and volume proportion is 0.5%. In this paper, the particularity of heat transfer rate of the SEMOS heat pipe with Cu-H 2 O fluid has been experimentally confirmed by changing the proportion of working fluid and Cu nonsocial particles in the heat pipe. By comparing the experimental result of these two different medium in the SEMOS HP, it is shown that the heat transferring rate with brass-water nanofluids as medium is much better than that with distilled water as medium under same volume proportion.

We present the rarefaction effects on diffusive mass transport in micro- and nanoscales using the results of direct simulation Monte Carlo DSMC method. Unlike the previous investigations, the momentum and heat contributions are eliminated from the computations via uniform velocity, pressure, and temperature field considerations. The effects of global Knudsen number on the diffusion phenomenon are studied for the same Peclet number and a unique mixer shape. The results indicate that there is considerable weakening in diffusion mechanism for high Knudsen number cases. As a result, the non-dimensional diffusive mass fluxes would decrease and the non-dimensional mixing length would increase as the Knudsen number increases. The effective diffusion coefficient is calculated throughout the mixer using the diffusive mass fluxes and the species mass fraction gradients. It is observed that the effective diffusion coefficient can vary considerably as a result of local rarefaction variations. It reaches to the lowest value at the point of confluence, where the maximum mass fraction gradient magnitude would occur for the species. Moving away from this point, the local rarefaction effects would weaken and the effective diffusion coefficient would reinforce subsequently. All the presented results indicate that there would be a convergent to a limiting behavior, which corresponds to the continuum mass diffusion case. Despite this, the local rarefaction level decreases continuously. Unfortunately, because of a considerable increase in the statistical fluctuations at very low rarefaction levels, the simulations do not provide reliable results in the limit of continuum regime.

Thermophotovoltaic (TPV) energy conversion enables millimeter scale power generation required for portable microelectronics, robotics, etc. In a TPV system, a heat source heats a selective emitter to incandescence, the radiation from which is incident on a low bandgap TPV cell. The selective emitter tailors the photonic density of states to produce spectrally confined selective emission of light matching the bandgap of the photovoltaic cell, enabling high heat-to-electricity conversion efficiency. The selective emitter requires: thermal stability at high-temperatures for long operational lifetimes, simple and relatively low-cost fabrication, as well as spectrally selective emission over a large uniform area. Generally, the selective emission can either originate from the natural material properties, such as in ytterbia or erbia emitters, or can be engineered through microstructuring. Our approach, the 2D photonic crystal fabricated in refractory metals, offers high spectral selectivity and high-temperature stability while being fabricated by standard semiconductor processes. In this work, we present a brief comparison of TPV system efficiencies using these different emitter technologies. We then focus on the design, fabrication, and characterization of our current 2D photonic crystal, which is a square lattice of cylindrical holes fabricated in a refractory metal substrate. The spectral performance and thermal stability of the fabricated photonic crystal thermal emitters are demonstrated and the efficiency gain of our model TPV system is characterized.

Thermal management is essential to compact devices particularly for high heat flux removal applications. As a popular thermal technology, refrigeration cooling is able to provide relatively high heat flux removal capability and uniform device surface temperature. In a refrigeration cycle, the performance of evaporator is extremely important to the overall cooling efficiency. In a well-designed evaporator, effective flow boiling heat transfer can be achieved whereas the critical heat flux (CHF) or dryout condition must be avoided. Otherwise the device surface temperature would rise significantly and cause device burnout due to the poor heat transfer performance of film boiling. In order to evaluate the influence of varying imposed heat fluxes, saturated flow boiling in the evaporator is systematically studied. The complete refrigerant flow boiling hysteresis between the imposed heat flux and the exit wall superheat is characterized. Upon the occurrence of CHF at the evaporator wall exit, the wall heat flux redistributes due to the axial wall heat conduction, which drives the dryout point to propagate upstream in the evaporator. As a result, a significant amount of thermal energy is stored in the evaporator wall. While the heat flux starts decreasing, the dryout point moves downstream and closer to the exit. The stored heat in the wall dissipates slowly and leads to the delay in rewetting or quenching, which is the key to understand and predict the flow boiling hysteresis. In order to reveal the transient heat releasing mechanism, an augmented separated-flow model is developed to predict the moving rewetting point and minimum heat flux at the evaporator exit, and the model predictions are further validated by experimental data from a refrigeration cooling testbed.

Microscale heat transfer in macro geometry has been proven to yield comparable heat transfer performance to that of typical microchannels. This paper takes it a step further by looking at three passive heat transfer enhancement techniques to improve the heat transfer performance of the newly proposed system for single-phase liquid flow. The novelty of the study lies in that the enhancement features are designed based on inspiration from nature. Fish scale, Durian (a thorny tropical fruit), and Inverted Fish Scale enhancement profiles are considered. In this study, an annular microchannel is formed by securing a cylindrical insert of mean diameter 19.4 mm within a cylindrical pipe of internal diameter 20 mm. The enhancement features are introduced on the surface profile of the insert, while the heat is supplied to the flow via the cylindrical pipe of fixed surface area. Therefore, heat transfer is improved by increasing convective heat transfer coefficient, for a constant heat transfer area. The enhancement features serve to increase heat transfer coefficient by disturbing the flow and thermal boundary layers. Experiments were carried out to investigate the effect of the three enhancement profiles on the heat transfer and flow characteristics of the microscale flow. The extent of enhancement is computed with the Plain profile as benchmark. The constant parameters include microchannel length of 30 mm, mean hydraulic diameter of 600 μm, and heat input of 1000 W. Reynolds number range is 1,300 to 4,600, with water as working fluid. Results show that the Inverted Fish Scale profile doubles the Nusselt number as compared to the Plain profile. However, when friction factor increment is considered, Durian profile yields the best overall thermal performance, nearly 1.4 times better than the Plain profile. In the whole study, the maximum convective heat transfer coefficient achieved is 45.0 kW/m 2 ·K, using Inverted Fish Scale profile at Reynolds number of 4,300. The pressure drop values of the system are all less than 3 bars, which may easily be achieved by a commercially available pump.

The heat dissipation of current busbur in power plant is one of the important issues in power transmission, usually through the cylinder slotted to strengthen heat dissipation. Natural convection in a cylinder with an internal slotted annulus is the computational model abstracted from it. Natural convection in a cylinder with an concentric slotted annulus is concerned. Attention is focused on the effects of different slotted sizes on natural convection. Numerical results showed that, the equivalent thermal conductivity increases with the increase of Rayleigh number. At high Ra, the system heat transfer exhibit rich nonlinear characteristics. When the slotted direction or the slotted degree changed, it would have an important impact on the flow and heat transfer in the system, and also influence the related nonlinear characteristics.