Condensation Heat Transfer Efficiency Analysis of Horizontal Double-Sided Enhanced Tubes

Abstract

The enhanced tubes in this study, referred to as E1 and E2, represent significant improvements in the design and performance of smooth tubes. By increasing the surface area on their fin side and optimizing the condensation drainage design, the heat transfer capacity of the finned tubes has been further enhanced. These modifications will provide superior thermal management performance for condenser tubes in practical applications, facilitating their widespread use across various engineering fields. In this experiment, R134a was used as the working fluid, with a test section length (L) of 248 mm for the experimental tubes E1 and E2. The experiments were conducted at a saturation temperature of 40 °C, where the refrigerant condensed outside the tube while deionized water circulated inside. The results indicated that, at a heat flux density below 94 kW/m², the condensation heat transfer coefficient of the E1 tube was 2–5% higher than that of the E2 tube, achieving values that were 11.63–14.42 times and 10.94–14.67 times that of smooth tubes of identical dimensions and materials, respectively. At a heat flux density of 94 kW/m², the heat transfer coefficient of E2 exceeded that of E1, with E1 exhibiting a more pronounced decline. Under constant water velocity, the heat transfer coefficient outside the tube initially decreased and then increased as the heat flux density rose. The corresponding effective heat transfer area of E1 increased, leading to better overall heat transfer performance compared to E2.

1. Introduction

As the core process of energy conversion and thermal management systems, condensation heat transfer has important applications in electric power, the chemical industry, aerospace, and microelectronics cooling. With the growth of global energy demand and environmental protection requirements, the development of high-efficiency condensing technology has become a key research direction to improve energy efficiency and reduce carbon emissions. In recent years, the demand for the substitution of environmentally friendly refrigerants (e.g., R1234ze(E), R290, R600a, etc.) has driven in-depth research on their heat transfer characteristics, while the application of enhanced tubes (e.g., three-dimensional finned tubes, low-conductivity titanium tubes) has provided a new direction for heat transfer enhancement. Research on high-pressure vapors, environmentally friendly refrigerants, and complex fin structures has significantly advanced the development of this field.

In terms of high-pressure steam condensation, Nie et al. [1] found experimentally that the heat transfer coefficient for the condensation of horizontal tubes under boiling conditions in the outer pool increases with mass flow rate and steam quality, but decreases with increasing pressure. Their proposed improved Akers correlation is more applicable to high-pressure scenarios (4–10 MPa) than the conventional Shah model, revealing the limitations of the existing models under high-pressure and high-flow-rate conditions. This finding provides an important basis for the optimization of passive waste heat discharge systems in nuclear power plants.

The enhancement mechanism of fin structure on condensation heat transfer has attracted much attention. Kumar et al. [2] investigated the performance of R-134a on monolithic finned tubes with different fin heights (0.45–2.40 mm) and found that a 3.18-fold enhancement in heat transfer can be achieved with a 0.45 mm fin height, attributed to the surface-tension-driven thinning effect of the condensate film. However, the fin efficiency decreased with increasing height (99.1–95.1%), suggesting that too-tall fins can weaken performance due to liquid film retention. Similarly, Chen et al. [3], for a new three-dimensional finned tube (C9+ tube), showed that its micro-ridge structure could disrupt liquid film continuity and still maintain its heat transfer enhancement at the submerged Reynolds number (Re_r < 1100) at moderate heat flux (40 kW/m²), but the advantage disappeared at a high heat flux (80 kW/m²).

However, the effect of enhanced structures on heat transfer performance presents complexity. Al-Badri et al. [4] found that 3D finned tubes showed a significant enhancement of the condensation coefficient over standard finned tubes in single tubes, but the performance degradation was faster in bundles of tubes due to the cumulative effect of the liquid film. In falling-film evaporation experiments, three-dimensional finned tubes (e.g., Gewa-B5) may show heat transfer inhibition at higher heat fluxes due to secondary fin interference. In addition, the internal enhancement structure of the tube (e.g., spiral ribs) indirectly affects the overall performance by enhancing the convective heat transfer coefficient.

In 2011, Zhang et al. [5] found that the condensation heat transfer coefficient of HCFC22 outside all three of their 3D ribbed tubes exceeded that of the light tube by a factor of 8.8. In 2012, Ali et al. [6] found that the heat transfer coefficient of ethylene glycol condensing outside a columnar 3D ribbed tube could be up to 5.5 times that of a light tube. Webb et al. [7] found that the membrane condensation heat transfer capacity of CFC11 on three 3D ribs, GEWA-SC, Turbo-C, and Tred-D, was 67%, 34%, and 17% higher than that of 2D ribs at 1024fp, respectively. In 1996, Rewert et al. [8,9] found that the condensation heat transfer coefficients of HCFC123 and HFC134a outside Turbo-CII and G-SC 3D ribs were up to 3 and 1.5 times that of 1024 fpm 2D ribs, respectively; the condensation heat transfer coefficient of HFC134a on Gewa C+ 3D ribbed tubes obtained by Belghazi et al. [10] in 2002 was 30% higher than the highest value obtained by them for trapezoidal-ribbed 2D ribbed tubes with five different rib spacings. In 2006, Gstoehl et al. [11] obtained condensation heat transfer coefficients of HFC134a outside Turbo-CSL and Gewa-C 3D ribs that were more than 50% higher than those of Turbo-Chil 2D ribs. The above results show that the single-tube membrane condensation heat transfer performance of 3D rib tubes is greatly improved compared with that of 2D rib tubes.

The influence of the submergence effect on the performance of tube bundles is complex. Mostafa et al. [5] found through innovative experiments with a distributed liquid distributor that the submerged liquid consistently showed enhancement (maximum enhancement factor of 1.21) for two-dimensional low-finned tubes at low heat fluxes (20 kW/m²), whereas it was effective for three-dimensional finned tubes only at Re_r < 600. Guo et al. [6] used the homology method to accurately determine the immersion effect coefficient of six rows of new three-dimensional finned tube bundles and found that it decreased from 1.00 at Re_r = 75 to 0.78 at Re_r = 1860, and that the heat transfer coefficient was 5–25% higher than that of the traditional three-dimensional finned tubes, which verified that the optimization of finned topologies promotes liquid film discharge.

Progress has also been made in the study of the condensation characteristics of environmentally friendly refrigerants. Chen et al. [3] compared the performance of HFO refrigerant R1233zd(e) on a light tube with that of a three-dimensional finned tube, found that the heat transfer coefficient of the finned tube reached up to 10.8 times that of the light tube, and proposed a modified model based on Nusselt’s theory (with a surface tension correction term). The low GWP (Global Warming Potential = 1) properties of this refrigerant make it an ideal replacement for R123, but its liquid film dynamics with complex finned tubes still need to be further investigated. The condensation heat transfer coefficients of R134a are higher than those of R1234ze(E) and R290 for both plain and enhanced titanium tubes, and the heat transfer enhancement ratio of enhanced tubes decreases with an increase in heat flux [7]. Similarly, the condensation performance of R600a on monolithic finned tubes is significantly affected by fin density, with 1102 fpm finned tubes performing optimally [8].

Notably, Ji et al. [9] conducted condensation heat transfer experiments with 11 horizontal tubes with refrigerant R134a at a saturation temperature of 40 °C (1.01 MPa) with titanium, white copper (B10 and B30), stainless steel, and copper. The first four materials had lower thermal conductivity and enhanced tubes were used, which had integral fins and three-dimensional geometry. The experimental results show that the heat transfer coefficient of enhanced copper tubes is increased by a factor of about 1.6–2.1 compared to low-thermal-conductivity tubes with the same enhanced geometry. In addition, the average enhancement rates of the titanium, B10, B30, and stainless steel tubes were 8.48, 8.31, 8.22, and 7.52, respectively, when compared to ordinary tubes.

Although the application of low-thermal-conductivity materials (e.g., titanium and stainless steel) can adapt to corrosive environments, the difference in thermal conductivity can lead to changes in fin efficiency and temperature distribution, which in turn affects heat transfer. For example, the condensation heat transfer coefficient of aluminum–brass three-dimensional finned tubes is more than 30% higher than that of iron–copper–nickel two-dimensional finned tubes. In addition, the refrigerant types have different sensitivities to the tubes: the condensation performance of R404A is more susceptible to surface structure and thermal conductivity than that of R134a, whereas the evaporation coefficient of the R123 falling film is only 1/2–1/3 of that of R134a, which is likely to be related to the changes in wettability resulting from viscosity, and surface tension differences may be related to the change in wettability due to the difference in viscosity and surface tension [10]. Ko et al. [11] experimentally investigated the film condensation characteristics of R1234ze(E)/R1233zd(E) in a horizontal smooth tube (36–40 °C, subcooling 3–18 °C); the heat transfer coefficient decreased with increasing subcooling and was higher for smaller tube diameters and proposed Nu correlation formulas (±20%). After that, Ko et al. [12] studied the film condensation characteristics of R134a, R1233zd(E), and R1234ze(E) in a horizontal smooth tube and three kinds of finned knurled enhanced tube (38 °C). The results showed that the heat transfer coefficients of the latter two in smooth tubes were about 10% lower than those of R134a, the enhanced tube E.T(C)-5 had the best performance, the number of fins had a greater effect on heat transfer than the number of knurls, and nine sets of heat transfer correlations with ±10% accuracy were established. Nagata et al. [13] investigated the heat transfer characteristics of the low-GWP refrigerants R1234ze(E)/R1233zd(E) in horizontal copper tubes (19.12 mm), considering condensation and boiling heat transfer performance: they found that the experimental value of R1233zd(E) for condensation was 25% higher than it should have been in theory, boiling HTC matched the model, and that surface tension affects nucleate boiling. Honda et al. [14] proposed a horizontal finned tube membrane condensation heat transfer model with ±20% error to validate a multi-fluid tube type. Yun et al. [15] experimentally studied the condensation characteristics of R134a in a stainless steel finned tube (19/26 fpi) and found the following: a heat transfer coefficient up to 4.4/3.1 times that of a light tube at 20 °C, 19 fpi is better at ΔT < 0.7 °C, and the Honda–Nozu model deviation is minimized.

However, the interaction between the physical properties of different refrigerants (e.g., surface tension, viscosity, latent heat) and tube characteristics (e.g., fin structure, thermal conductivity) still needs to be systematically investigated. In this paper, the condensation heat transfer characteristics of one smooth tube (ST) and two double-sided enhanced tubes (E1, E2) are investigated using the refrigerant R134a, and the tests are carried out at a saturation temperature of 40 °C. The results are summarized as follows. The refrigerant R134a flows on the outer surface of the heat transfer tubes while deionized water circulates inside the tubes. Under constant heat flux conditions, the heat flux density was kept at 30 kW/m², the water-side flow rate was in the range of 1 to 3 m/s; under constant water-side inlet temperature, with the inlet temperature controlled at 12 °C, the water-side flow rate was 1 to 3 m/s; and, finally, the experiments were carried out at a saturation temperature of 35 °C, keeping the water flow rate at 2 m/s. The results of the tests are summarized below. On the basis of the experimental results, the heat transfer mechanism is analyzed.

2. Experimental System

2.1. Experimental Setup

The test bench consisted of three circulation systems to simulate the working environment of evaporating and condensing heat transfer tubes in a horizontal shell-and-tube heat transfer system. The system consisted of a main refrigerant circulation system, a heating water circulation system, and a cooling water circulation system, all of which worked together to precisely control the water temperature and flow rate parameters required for the experiment. Figure 1 shows a schematic diagram of the system architecture and photos of the test device.

2.1.1. Refrigerant Circulation System

The system adopted a double-cylinder structure design; the upper part was a horizontal condensing cylinder and the lower part was configured with a horizontal evaporating cylinder. The test copper tube was horizontally immersed in the refrigerant medium of the evaporation cylinder, constituting the core component of heat transfer. During operation, high-temperature water flow through the evaporator tube prompts the vaporization of the refrigerant liquid film outside the tube, and the resulting vapor enters the condensing cylinder through the connecting airway. In the condensation stage, the cooling water continues to take away the latent heat of phase change in the refrigerant vapor to realize the complete condensation of the vapor. After condensing the liquid refrigerant under the action of gravity through the reflux channel back to the evaporator cylinder, the system completes the closed cycle of the mass.

2.1.2. Dual-Circuit Water Circulation System

The system was equipped with independent hot water/cold water dual temperature control sub-systems: the hot water circulation loop contained a high-precision thermostatic heating device with PID control, which could set the temperature of the hot water being sent to the evaporation test tube through the centrifugal pump, the heat being transferred back to the water tank to form a closed loop; the cold water circulation loop integrated the compressor refrigeration module, driven by a variable-frequency pump to drive the cooled water through the condensing tube bundle, absorbing the heat of the refrigerant phase change and then returning to the cold water tank. The two parts of the temperature control system were equipped with flow meters and electric control valves and could realize a 0.5 °C temperature accuracy and ±2% flow rate accuracy of dynamic control, ensuring the long-term stability of the experimental conditions.

The system realized the decoupling of evaporation/condensation process control through a modular design, in which the evaporation side adopts active heating to establish the heat load, the condensation side maintains heat-sinking conditions through precise refrigeration, and the double-cycle synergy simulates the dynamic heat transfer process in the actual operation of a heat exchanger.

2.2. Process Monitoring and Data Acquisition System

The experimental platform was equipped with a distributed process monitoring system, integrating a high-precision sensing network and a multi-channel data acquisition module. The system monitored the following key parameters in real time: the temperature field distribution of the double constant-temperature water tanks, the temperature gradient of the water side of the evaporation/condensation heat transfer tube inlet and outlet (±0.1 °C accuracy), the flow rate dynamics (electromagnetic flowmeter ±1.5% FS), and the pressure drop characteristics (differential pressure transmitter ±0.5% RD), as well as the saturation pressure of the phase-change region of the refrigerant workpiece (piezoresistive transducer ±0.25%) and the temperature of the liquid film (armored thermocouple ±0.5 °C). All signals were transmitted to the LabVIEW data acquisition platform via an RS485 bus, realizing the synchronous acquisition of 10 groups of data per second, and the system was equipped with PID parameter self-tuning, an abnormal working condition alarm, and a triple backup function for the experimental data.

2.3. Structure of Test Section and Enhanced Test Tube Design

The standard experimental section adopted a strengthened heat transfer test tube of L = 248 mm, whose innovative structural design contained the following:

Inner wall surface: Precision-rolled three-dimensional helical micro-ribs (rib height 0.2 mm–0.5 mm, helix angle 30°) to disrupt boundary-layer development and enhance turbulent mixing through secondary flow effects.

Outer surface: A staggered array of serrated fins (fin height 1.2 mm, fin pitch 2.5 mm), effectively expanding the heat transfer surface area up to 2.8 times that of a traditional light tube while forming a multi-scale vaporization core point.

Two typical double-sided enhanced condenser tubes (see Figure 2 for their physical structure) were selected for comparative study, and their key geometrical parameters are shown in

Table 1. Geometric parameters of E1 and E2.

Tubes	E1	E2
Outside diameter, do/mm	25.2	25
Inner diameter, di/mm	22.9	22.8
External teeth height, e/mm	0.8	0.79
Pitch, FP/mm	0.57	0.55
Number of circumferential teeth, Ns	150	100
Bottom wall thickness, tb/mm	0.55	0.55
Internal teeth high, tr/mm	0.26	0.42
Internal teeth number, Ni	75	38
Helix angle, $\alpha$/deg	40°	45°

: tube type A adopted a gradual rib depth design to optimize the axial flow distribution and tube type B was configured with asymmetric fins to achieve the dynamic matching of the phase transition process.

2.4. Experimental Conditions and Test Specifications

This study used R134a as the circulating medium. All measurement instruments were calibrated in accordance with NIST standards. The thermophysical properties of R134a were calculated by the REFPROP 9.0, a software based on the NIST Standard Reference Database [16].

Thermodynamic boundary: We maintained a saturation temperature of 40 ± 0.3 °C (corresponding to the pressure of 1.016 MPa) on the evaporation side and set the baseline working condition of 35 ± 0.3 °C on the condensation side.

Hydraulic working condition: Deionized water was operated in the turbulent zone in the tube (Re = 8 × 10³~2.4 × 10⁴), the flow rate was adjusted in steps of 1–3 m/s, and the in-let temperature was kept constant at 12 ± 0.2 °C.

Heat flow control: A stable heat flow density of 30 kW/m² ± 5% was realized by an electric heating membrane, with water-side PID flow regulation to ensure the heat balance error was less than 3%.

For the comparison experiments, by adjusting the cooling water flow to change the condensing temperature in the range of 32–38 °C and achieving changes in the system with a response time of less than 120 s, we achieved a steady state after the continuous collection of 300 groups of valid data points for statistical analysis.

3. Heat Transfer Data Processing

The heat generated in the experiments was calculated using Equation (1):

$$Q_w=c_{pl,w}{m}_w\left(T_{exp,out}-T_{exp,in}\right)$$

(1)

In the above formula, c_pl,w represents the specific heat capacity of water in the experimental section, m_w represents the flow rate of water, T_exp,_out and T_exp,_in represent the import and export temperature of water in the experimental section, respectively; the import and export temperatures of the hot water tank were used in case of the calculation for the boiling tube, and the import and export temperatures of the cold water tank were used in case of the calculation for the condensing tube.

The temperature difference generated throughout the experimental section was calculated using the log-mean temperature difference, as shown in Equation (2):

$${\Delta T}_{LTMD}={\displaystyle \frac{T_{in}-T_{out}}{\ln \left(\frac{T_{sat}-T_{out}}{T_{sat}-T_{in}}\right)}}$$

(2)

In the above equation, T_sat denotes the saturation temperature and Tin and Tout are the different water-side inlet and outlet temperatures of the different experiments, respectively. The heat flow density based on the nominal diameter outside the tube is as follows:

$$q=Q_w/A_0$$

(3)

The total heat transfer coefficient for the entire experimental section is calculated as in Equation (4):

$$K=Q_w/A_o·{\Delta T}_{LTMD}$$

(4)

$$A_0=\pi \times d_0\times L$$

(5)

The total heat transfer thermal resistance for the entire experimental section consists of the following components:

$${\displaystyle \frac{1}{K}}={\displaystyle \frac{1}{h_o}}+{\displaystyle \frac{1}{h_w}}{\displaystyle \frac{d_o}{d_i}}+R_f+R_{wall}$$

(6)

In the above equation, R_f and R_wall denote the fouling thermal resistance and pipe wall thermal resistance, respectively, where the pipe wall thermal resistance is calculated by Equation (7):

$$R_{wall}={\displaystyle \frac{d_o}{2\lambda }}\ln {\displaystyle \frac{d_o}{d_i}}$$

(7)

Neglecting the effect of fouling thermal resistance, the outer tube heat transfer coefficient h_o is calculated with the effect of fouling thermal resistance not considered:

$$h_o=1/\left[{\displaystyle \frac{1}{K}}-{\displaystyle \frac{1}{h_w}}{\displaystyle \frac{d_o}{d_i}}-{\displaystyle \frac{d_o\ln \left(\frac{d_o}{d_i}\right)}{2\lambda }}\right]$$

(8)

The water-side heat transfer coefficient is calculated using C_i with the associated Gnielinski formula [17], as shown in Equation (9):

$$h_w=C_i{h}_{Gni}=C_i{\displaystyle \frac{{\lambda }_w}{D_i}}{\displaystyle \frac{\left(\frac{f_w}{8}\right)\left(R{e}_w-1000\right)P{r}_w}{1+12.7{\left(\frac{f_w}{8}\right)}^{0.5}\left(P{r}_w^{\frac{2}{3}}-1\right)}}$$

(9)

The friction coefficients in the above equations can be obtained from the Filonenko correlation for smooth tubes [18].

$$f={\left(1.82lgRe-1.64\right)}^{-2}$$

(10)

where the correction factor C_i is calculated using the Wilson graphical method [19], as follows:

$${\displaystyle \frac{1}{h_{Gni}}}=C_i\left[{\displaystyle \frac{1}{K}}{\displaystyle \frac{d_i}{d_o}}-{\displaystyle \frac{1}{h_o}}{\displaystyle \frac{d_i}{d_o}}-{\displaystyle \frac{d_i}{2\lambda }}\ln {\displaystyle \frac{d_o}{d_i}} \right]$$

(11)

4. Experimental Results

4.1. Reliability Verification of Experiments

When the refrigerant saturation temperature and heat flux is kept constant, the refrigerant thermal resistance is also kept constant; the thermal resistance of the tube wall is also certain, therefore, when the heat flow density of 30 kW/m² is kept constant, and a series of working condition points can be obtained by substituting the recorded state data into formula (11) to obtain a system of one-time equations that is thought to be the slope of C_i, with 1/h_Gni as the longitudinal coordinate and d_i/(Kd_o) as the transverse coordinate, and then the system of one-time equations can be obtained by using Wilson’s graphical solution [19] to regress C_i. As shown in the Figure 3, the C_i values obtained for the experimentally measured smooth tube, E1, and E2 were as follows: 1.004, 2.58, and 2.68.

The experimental results of the light pipe for thin-film condensation were compared with a Nusselt analysis [20] to further verify the feasibility of the experimental equipment.

The Nusselt correlation is as follows [20]:

$$h_{\mathfrak{p}}=0.728{\left({\displaystyle \frac{g{\lambda }_l^3\left({\rho }_l-{\rho }_v\right){\rho }_l{h}_{\mathit{fg}}}{{\mu }_l{d}_o\left(t_s-t_w\right)}}\right)}^{{\displaystyle \frac{1}{4}}}$$

(12)

where λ_l is the thermal conductivity of liquid refrigerant (W/(m·K)); ρ_l is the density of the liquid refrigerant (kg/m³); ρ_v is the density of the gas refrigerant (kg/m³); h_fg is the latent heat of vaporization of the refrigerant (J/kg); and μ_l is the coefficient of kinetic viscosity of the liquid refrigerant (kg/(m·s)).

The experiment set T_sat to 40 °C, inlet water temperature to 30 °C, measured the smooth-tube refrigerant side of the heat transfer coefficient, performed a Nusselt analysis of the predicted data for comparison, and found that its relative error within ±15%, as shown in Figure 4, can be illustrated by the experimental stage of higher precision.

4.2. Effect of Experimental Conditions

In this study, we adjusted the water flow rate at a heat flow density of 30 kW/m² and a saturation temperature of 40 °C in order to investigate the refrigerant-side heat transfer coefficient (h_r), the total heat transfer coefficient (K), and the water-side heat transfer coefficient (h_w) for both smooth and enhanced tubes. To ensure the stability of the experimental conditions, the water temperature was adjusted according to the change in water flow rate.

As shown in Figure 5a, both the smooth tube and the two enhanced tubes exhibit insensitivity to changes in water flow rate. This is because, at a constant saturation temperature and heat flow density, the physical parameters of R134a remain unchanged, and changes in water flow rate alone have a negligible effect on the heat transfer performance of the refrigerant. As shown in Figure 5b. The total heat transfer coefficient rises with increasing water velocity. However, in our experiments, we observed that the increasing trend of the total heat transfer coefficient starts to slow down when the water velocity exceeds a certain critical value. This is because, although the increase in water velocity still enhances the degree of turbulence and thins the thermal boundary layer, its contribution to the total heat transfer coefficient gradually becomes saturated and the influence of the heat transfer resistance on the refrigerant side on heat transfer capability becomes more significant, resulting in the magnitude of change in the total heat transfer coefficient decreasing with the increase in water velocity. For the water-side heat transfer, as shown in Figure 5c, a higher water flow rate enhances the degree of water turbulence and thins the thermal boundary layer, which improves the heat transfer performance of the water-side heat transfer and also leads to an increase in the integrated heat transfer coefficient of the heat transfer tube.

By comparing the experimental data of the smooth tube and enhanced tubes, we found that the enhanced heat transfer performance of the enhanced tubes is more obvious when the water velocity is increased, which suggests that the use of enhanced tubes is an effective strategy when designing high-efficiency heat transfer systems.

As shown in Figure 6, in the water side of the inlet, water temperature is constant, the inlet water temperature is controlled at 30 °C, and the flow rate is 1 to 3 m/s; this condition is consistent with large-scale commercial chiller unit operating conditions: where the experiments are more in line with the actual application situation, it is easier to simulate the actual application of the process.

The relationship between the water flow velocity and integrated heat transfer coefficient is shown in Figure 6a: the integrated heat transfer coefficients represented by the three curves show an upward trend as the water flow velocity increases while the inlet temperature is kept constant. Specifically, when the water flow velocity increases from 1.0 m/s to 3.0 m/s, the integrated heat transfer coefficient of the enhanced tube grows from about 6.5 kW/m²K to about 12 kW/m²K, whereas the effect of the smooth tube is not particularly obvious, but it also shows an upward trend, which is due to the fact that the increase in the number of racks in the enhanced tubes expands the actual heat transfer area of the inner surface of the tubes, which promotes the improvement of heat transfer in the tube and at the same time means that the integrated heat transfer area is improved. It can be seen that, for the water side, the enhanced tube is outstanding in its ability to increase the water flow rate to enhance the comprehensive heat transfer coefficient, and that the effect for E2 is similar to E1.

As shown in Figure 6b, the heat transfer coefficients outside the tubes all decrease as the heat flow density increases. At heat flow densities below 94 kW/m², the condensation heat transfer coefficients of the E1 tube are 2–5% higher than those of the E2 tube. At a heat flux density of 94 kW/m², the heat transfer coefficient of E2 exceeds that of E1, and the decreasing trend of E1 is significantly higher than that of E2. A lower percentage of the rate of decrease in the heat transfer coefficient is observed. As the heat flux increased from 50 kW/m² to 97 kW/m², the rate of decrease in E1 was 6.1%, while the rate of decrease in E2 was 1.5%.

Both the E1 and E2 tubes are developed condenser tubes, and the unique surface structure of both gives them excellent heat transfer performance in condensation. Although the outer heat transfer coefficient of the former tube is higher than that of the latter, the condensation heat transfer coefficients of the two are similar, with smaller decreasing trends, at 11.63–14.42 times and 10.94–14.67 times the condensation heat transfer coefficients of smooth tubes of the same size and the same material, respectively.

Combining Figure 6b with Figure 2 shows that a series of small rectangular block protrusions and grooves in the three-dimensional structured fins increases the pressure gradient of the condensate film on its surface and enhances the role of the surface tension of the film, which plays an important role in drawing the condensate out easily from the uniformly distributed small rectangular blocks, and means that condensate flows easily from the rectangular blocks and grooves; the relatively flat surface and the special block arrangement play an important role in lowering the condensate thickness of the condensate rate. The E1 tube has a larger tooth pitch and a larger number of external teeth than the E2 tube, with a corresponding increase in actual heat transfer area, meaning that the condensation heat transfer coefficient of E1 is slightly higher than that of the E2 tube. However, at higher heat fluxes, where more condensate is present on the fin surfaces, the Gregorig effect [21] plays an important role in attracting the liquid from the tips of the micro-fins to their roots, meaning that condensation mainly occurs at the top of the tip unit. The tube fin heights of E2 are small, so the Gregorig effect is weak compared to with E1, but the tube fin heights of E1 are also small, so the condensation heat transfer coefficients of the two tube types, E1 and E2, decrease with increasing heat flux densities, resulting in lower condensation performance.

Under the experimental conditions of a water velocity of 2 m/s, the relationship between the heat transfer coefficient outside the enhanced tube and the density of heat flow was analyzed and studied, as shown in Figure 7: it was found to decrease and then increase. This is because, at the beginning of condensation, many tiny liquid beads are gradually formed on the outer wall surface of the enhanced tube, and at this time, the condensate fails to completely wet the outer wall surface of the heat transfer tube. With the gradual increase in condensate, the number of fine liquid beads on the heat transfer wall surface gradually increases until they completely cover the heat transfer wall to form a liquid film; while the thickness of the liquid film gradually increases, the heat transfer thermal resistance becomes larger, resulting in a reduction in its heat transfer coefficient. But as the experiment progresses, the condensation rate reaches the maximum and the thickness of the liquid film on the tube wall reaches the maximum thickness of the heat exchange tube. Under the action of gravity and surface tension, due to the special structure of the condensate tube, condensate starts to be discharged quickly from the matrix protrusions and grooves, the liquid film thickness gradually decreases, and the heat transfer thermal resistance decreases, resulting in the heat transfer coefficient rising because the tooth pitch and the number of external teeth of the E1 tube are larger than those of the E2 tube. Accordingly, the actual heat transfer area increases, and the heat transfer effect is relatively better than that of the E2.

5. Conclusions

In this paper, the thermal properties of condensation heat transfer of one smooth tube and two bilaterally enhanced tubes are investigated for different water flow rates and heat flow densities at a certain saturation temperature. The following conclusions can be drawn from the analysis of the experimental data:

(1) When the saturation temperature and heat flow density remain unchanged but there is a change in the water-side flow rate, the refrigerant side of the tube outside the heat transfer coefficient is almost unchanged; the total heat transfer coefficient becomes larger, but the growth tendency of the total heat transfer coefficient slows down. The water-side heat transfer coefficient gradually increases. By comparing the experimental data of the smooth tube and reinforced tubes, we found that the enhanced heat transfer performance of the reinforced tubes is more obvious when the water velocity is increased, and that it is about 300% of the efficiency of smooth tubes, which suggests that the use of reinforced tubes is an effective strategy when designing high-efficiency heat transfer systems

(2) With heat flow densities lower than 94 kW/m², the condensation heat transfer coefficient of the E1 tube is 2–5% higher than that of the E2 tube. Both tubes’ condensation heat transfer coefficients are 11.63–14.42 and 10.94–14.67 times higher than that of a smooth tube with the same size and the same material, respectively. At a heat flux density of 94 kW/m², the heat transfer coefficient of E2 exceeds that of E1, and the decreasing trend of E1 is significantly higher than that of E2. A lower percentage for the rate of decrease in heat transfer coefficient is observed. As the heat flux increases from 50 kW/m² to 97 kW/m², the rate of decrease in E1 is 6.1%, while the rate of decrease in E2 is 1.5%.

(3) At constant water velocity, as the heat flux density increases, the heat transfer coefficient outside the tube first decreases and then increases. The corresponding actual heat transfer area of E1 increases, so its heat transfer effect is better relative to E2. At a heat flow density lower than 40 kW/m², E1 is 2–5% more efficient than E2, and after the heat flow density reaches 40 kW/m², the E1 and E2 heat transfer coefficients gradually close in on each other, and the E1 heat transfer coefficient is about 100 W/m²K higher than that of E2.