Effects of gas-liquid flow and dehumidification performance of a liquid desiccant dehumidifier A num

Abstract

Keywords: Dehumidifier Performance; Heat and Mass Transfer; Liquid Desiccant; Moist Air; Vertical and Inclined Plates

Introduction

Nowadays, liquid desiccant air conditioning (LDAC) systems and other dehumidification systems use a liquid desiccant to remove moisture and heat from the air to improve the air quality, which can result in significant energy savings. The liquid desiccant dehumidifier (LDD) is a vital part of LDAC systems that act as a heat and mass exchanger that removes moisture from the air as it passes over plates coated with liquid desiccant. The liquid desiccant regenerator (LDR) then removes the absorbed moisture from the desiccant, allowing it to be reused. Common liquid desiccants used in LDAC systems include lithium bromide (LiBr), calcium chloride (CaCl), and lithium chloride (LiCl). On the contrary, solid desiccant cooling systems use materials like silica gel or zeolites to absorb moisture from the air at lower temperatures, requiring separate heating for regeneration. They offer lower corrosion risk and potential energy efficiency benefits but can be costly initially. Therefore, liquid desiccant cooling systems use solutions such as lithium chloride for faster moisture absorption, requiring higher regeneration temperatures and careful material selection to avoid corrosion. They provide flexibility in heat sources but may have higher operational energy costs. In general, liquid desiccants play an important role in improving indoor air quality, reducing energy consumption, and improving the efficiency of air conditioning and dehumidification systems [1]. However, lithium chloride (LiCl) is a highly hygroscopic substance that readily absorbs water vapor from the air. The mechanism by which a LiCl solution absorbs water vapor through the process of water vapor diffusion, where the water vapor molecules in the air move from areas of higher concentration to areas of lower concentration. The computational fluid dynamics (CFD) analysis of a liquid desiccant dehumidifier was performed by Lu and Lu [2] and enhancing the dehumidification performance of LiCl solution was investigated by Wen et al. [3]. When a LiCl solution is exposed to air, the water vapor molecules in the air diffuse into the solution, attracted by the high concentration of water molecules in the LiCl solution. As the water vapor molecules diffuse into the LiCl solution, they bond with the lithium cations (Li+) and chloride anions (Cl-) in the solution to form a stable hydrated complex. This process continues until the concentration of water vapor in the air and the solution reaches equilibrium, at which point the solution is said to be saturated with water vapor [4]. The extent to which a LiCl solution absorbs water vapor depends on several factors, including the concentration of the solution, the surface area of the exposed solution, and the relative humidity of the surrounding environment. The more concentrated the solution, the more water vapor it can absorb. Similarly, the greater the surface area of the exposed solution, the more water vapor it can absorb. In a high-humidity environment, the LiCl solution will

absorb more water vapor than in a low-humidity environment [5]. In the meanwhile, Zhang et al. [6] studied the shrinkage of a liquid film running downhill down the heated solid surface, which results from the presence of a surface tension gradient and is significantly impacted by the Marangoni effect. Another study by Zhang et al. [7] optimized the operating parameters of a LiCl dehumidification system, including the concentration of the LiCl solution and the regeneration temperature, to maximize its performance. They found that a concentration of 60 % and a regeneration temperature of 70 °C resulted in the best performance. Das and Jain [8] investigated solar energy utilization as a heat source for the regeneration of a LiCl dehumidification system. They found that the use of solar energy reduced the energy consumption of the system and increased its overall efficiency. While Zhang et al. [9] conducted the exergy calculation and analysis of a dehumidification system using LiBr liquid desiccant. A study by Chen et al. [10] analyzed temperature and humidity with incorporated dehumidification technology. Air dehumidification employs low-temperature, low-concentration liquid desiccant; condensation heat may be used to resupply the desiccant solution. Stevens et al. [11] outlined a computationally effective model for mass and heat exchangers with packed beds and liquid desiccant that encompasses the creation and extraction of the efficacy model, testing against experimental data, and a finite-difference model. A study by Liu et al. [12] proposed a new model that incorporated a heat pipe for heat recovery and a multistage configuration for improved dehumidification performance. The liquid film in the LDD is very thin, making it challenging to measure the properties of the film and accurately predict dehumidification operations [13]. Earlier studies have used the NTU model [14] and the finite difference model [15] to obtain results. Nevertheless, a CFD analysis [12] was performed on smooth plates to express the dehumidification process of LDDs, and it successfully predicted the heat and mass transfer characteristics. Liu et al. [16] presented the results of a study on the mass transfer capabilities of a cross-flow dehumidifier employing a liquid desiccant. The dehumidifier was equipped with Celdek structured packings, and the liquid desiccant was an aqueous solution of LiBr. Tao et al. [17] reported an investigation of the liquid desiccant dehumidification process based on CFD technology and Wen et al. [3] also reported the enhancement of the ability of the LiCl solution dehumidification process with surfactant Polyvinylpyrrolidone (PVP) K-30. Jayanti and Hewitt [18] employed CFD techniques to calculate the flow field and heat transfer through a wavy liquid film in order to study the heat transfer and hydrodynamics in a thinfilm flow. Meanwhile, Del Carlo et al. [19] established a straightforward model to assess the efficiency of mass transfer and dynamics in distillation columns. Banerjee et al. [20] assessed the heat and mass transfer that takes place on the surface of liquid ethanol using a computer

model. Haroun et al. [21] explored the influence of liquid flow rate and structured packing geometry on both liquid hold-up and mass transfer. In addition, the authors of Turgut and Coban [22], Wen et al. [23], and Xu et al. [24] conducted their numerical and experimental research on the liquid desiccant dehumidifier to examine the performance of a liquid falling film dehumidifier. Meanwhile, Wen and Lu [25] studied film shrinkage form and vapor condensation in internally cooled LDD, and Luo et al. [26] reported a fin-tube type internally cooled LDD both for experimental and CFD simulation. Again Qi and Lu [27] optimized a survey of an internally cooled/heated LDAC system. Qi et al. [28, 29] investigated the liquid contact angle and its influence on the LDD system and the wetted area and film thickness for the falling film LDR system, respectively. Bouzenada et al. [30] and Dong et al. [31] reported the dehumidification performance of the LDAC system and enhanced the dehumidification performance by TiO2 super-hydrophilic coating on the plate of LDD, whereas Lu et al. [32] investigated the dynamic characteristics of the counter-current flow for LDD system. Das et al. [33] assessed the effectiveness of surface-modified LiCl and CaCl2-based falling film dehumidification in a cooling system, while another study by Das et al. [34] explored the performance of liquid desiccant absorber for the corrugated triangular structure where the volume of fraction (VOF) approach is utilized to monitor the interface of gas-liquid and penetration theorem is employed to estimate the mass transfer coefficients. Lyu et al. [35] demonstrated the effects of air parameters such as air velocity, flow pattern, and pressure on LiCl-H2O flow behavior in a liquid desiccant system. Das et al. [36] reported a numerical model of an LDD system with the trapezoidal baffled surface that involves a conjugate heatmass interaction in a 2-D transient, multiphase, multicomponent system. Du et al. [37] developed an internally cooled dehumidifier by incorporating the corrugated fins into the system to seek more contact area and enhance the system’s efficiency as well as compatibility. The temperature adjustments are optimized as per recommendation and the reliability is ensured by the excellent precision of the correlation, which is within ±25 percent. In another study, Du et al. [38] analyzed the thermal effect of a flat tube with many channels and corrugated fins that are internally cooled with its strengthening mechanism. Gao and Lu [39] reported a parametric analysis of a membrane distillation LDR with heat/moisture recovery and potable water production. Khan et al. [40] studied heat and mass transfer in adiabatic regeneration of LiCl desiccant on vertical and modified polypropylene surfaces under full and partial wetting conditions. Peng et al. [41] used CFD with the RNG k-ε model and a user-defined function to optimize a falling film dehumidifier by varying the number and width ratio of rectangular cylinders. Moreover, Rokhman et al. [42] conducted a CFD analysis to model and enhance dehumidification performance in

a three-dimensional evaporative cooling-type film dehumidifier. Zhao et al. [43] reported also a CFD analysis to survey the performance of falling film LDD dehumidifier over sinusoidal and corrugated plates. Based on the findings, a mass transfer correlation was established and the effects of plate shape, moist air psychrometric characteristics, and solution parameters were examined. Zhou et al. [44] developed a mathematical model to describe the falling film flow on the surface of the flat-finned tube heat exchanger in a closed-type heat source tower, where the interface characteristics of gas-liquid falling film flow consider the interface forces. Meanwhile Mondal et al. [45, 46] revealed the results of an experimental study on a small-scale vapor absorption cooling system and bend tube water to the air heat exchanger, while Mondal and Islam [47] investigated the intermittent ammonia absorption refrigeration system. To aid in selecting the most suitable refrigerant for future air conditioning and refrigeration systems, Mondal et al. [48–52] measured the heat transfer coefficients and transport properties of various low GWP refrigerants. Moreover, Das et al. [53] conducted a thermodynamic study of a dual cascading vapor compression cycle and an organic Rankine cycle driven by solar energy, where two cases of power generation and cooling effects are studied. Several studies have shown this by combining vapor compression (VC) technology with liquid desiccant systems. Kumar et al. [54] reported a review of the progressive study for hybrid cooling in liquid desiccant-vapor compression systems, using a dehumidification method can provide a reasonable degree of cooling while saving 40–80% on energy. Zhao et al. [55] presented an overview of mass, heat, and hydrodynamic transfer in falling film evaporation, absorption, cooling, and dehumidification in order to investigate the effect mechanism of a gas stream and obtain precise profiles of flow, temperature, and concentration. Fahad et al. [56] investigated a survey of materials, systems, and applications for the progressions in liquid desiccant technologies. Koronaki et al. [57] conducted a study of an adiabatic dehumidifier with various liquid desiccants. Park et al. [58] demonstrated an energy benefit of a hybrid LDD and evaporative cooling-assisted system for building air conditioning, whereas Mohammad et al. [59] detailed an examination of an integrated vapor compression system-based LDD system for use in buildings. Moreover, Rafique et al. [60] presented a review work on the regenerative method of solid/liquid desiccant system and its application in the drying system, while Jani et al. [61] reported a review of liquid desiccant dehumidifiers where summarized the performance parameters have been made to analyze the system performance. Meanwhile, Misha et al. [62] reported a review on liquid desiccant dehumidification-assisted cooling systems. From the debate mentioned earlier, some studies were performed with smooth and rough vertical plates for LDD. However, there is an issue of considering the plate at some

angle with the vertical axis. To date, little to no research has been done taking the angle of inclination of the plate at any angle. There is no need for the plate or channel of moist air to be always vertical. Therefore, this study aimed to analyze the desiccant dehumidifier performance in three configurations: vertical smooth, vertical rough, and inclined rough plates, respectively, where LiCl is used as a liquid desiccant to absorb moisture from the air. The angle of inclination was considered and vapor absorption simulation was done in inclined plate LDD. In the test section, the desiccant will flow downwards for gravitational force and airflow will be set opposite to the desiccant flow from vertically downward to upward. Then, the amount of moisture collected from the air was investigated by the simulation using ANSYS workbench 2020 R1. The investigation was also carried out by placing the plates inclined and then the mass transfer rate was observed and compared to the vertical plate observations.

Methodology

Wen et al. [3] specifically developed an experimental bench to find the impacts of various parameters on the dehumidification features, and Lu and Lu [2] then provided numerical justification for this work. Figure 1 exhibits the schematic diagram of the test bench of this experiment. It had three loops, each of which was identified by a different color of arrow. They were made up of three loops: one for liquid desiccant, one for processing air, and one for cooling water. All loops were stuffed with neoprene foam on the surfaces for thermal isolation. The entire setup was built in a lab with the ability to maintain a relatively constant level of moisture and temperature. To mitigate corrosion from LiCl

solution and to improve dehumidification efficiency, stainless steel 316L, known for its excellent corrosion resistance, was chosen for the plate dehumidifier [3]. In this study, the material of different plate configurations is considered as stainless steel. The vertical and inclined smooth and rough plates of falling film LDD are optimized in the main test section which is presented in Figure 2. In the main test section, the LiCl solution of different concentrations is used as a working substance for the various plate geometries: vertical smooth, vertical rough, and inclined rough. Physical Model The three different forms of plate configurations for falling film liquid desiccant dehumidifiers are simplified and depicted in Figure 2. One is vertical smooth, one is a rough vertical plate and another is an inclined rough plate, respectively. The moist air flows in a counter-current direction at the bottom, whereas the desiccant film falls from top to bottom. The liquid desiccant dehumidifier (LDD) has a length of 150 mm and a width of 10 mm, with a width of 2 mm at the top of the liquid desiccant inlet, where the liquid desiccant is supplied through the length of the dehumidifier. In the vertical smooth plate shown in Figure 2(a), the liquid desiccant solution is flown along a 150 mm long plate, whereas the vertical rough plate shown in Figure 2(b) is spaced 10 mm apart and 1 mm in height. To enhance heat transfer, the vertical rough plate casing features ribs that are 1 mm high and 10 mm apart; the rib pitch-to-height ratio is regarded as 10. Both the first rib and the outlet of liquid desiccant are 30 mm apart, as is the distance between the liquid desiccant input and the last rib. Figure 2(c) represents the geometry of the inclined rough plate. The inclination of the rough plate is 300 from the vertical axis or 600 from the

Figure 1. Illustration of experimental falling film dehumidifier of vertical plates of LDD. [From Wen et al. [3], with permission from Elsevier, License Number: 5880760582683].

Figure 2. Geometry of simplified falling film dehumidifiers (a) vertical smooth, (b) vertical rough, and (c) inclined rough plates of LDD.

horizontal axis, rather all other specifications are the same as the vertical rough plate. Moreover, Figure 3 depicts the flow chart of the proposed model wherein the entire steps have been incorporated to gain insights into the work. The concentration of LiCl solution is considered as 30%, 33%, 36%, 40%, and 44%, respectively, for the justification of the influence of various concentrations of LiCl solution. While the inlet air velocity is considered as 0.5, 1.0, 1.5, and 2.0 m/s, respectively. Formation of the Problem In the current study, simulations have been conducted using ANSYS workbench 2020 R1. The choice of a pressure-based solver was made because the flow of both moist air and desiccant is incompressible with the constant density. The energy equation has also allowed us to understand the heat transfer mechanisms involved in the system. It is found that the liquid-dropping film is somewhat turbulent, while the airflow is turbulent overall. To model the gas-liquid flow, various turbulence models have been used, including the conventional k-ε model, RNG k-ε model, and Reynolds stress model, based on the work done by Banerjee and Isaac [63]. This simulation methodology has proved effective in analyzing the behavior of the moist air and desiccant flows and could be useful in designing and optimizing similar systems in the future. The modeling of the gas-liquid interface has been solved accurately using the volume of fraction (VOF) model. This VOF model is utilized in this study to track the free surface

of the two-phase flow. The piecewise linear interface calculation (PLIC) technique, in particular, makes the VOF model flexible and efficient [1, 2]. The mass and momentum conservation equations for the gas-liquid two-phase system rely on the volume fractions of the gas and liquid. Nonetheless, the density (ρ) and viscosity (µ) characteristics for every computational cell are displayed as follows: (1) (2) Where, αg and αl are the gas and liquid volume fractions, respectively. Again, ρg and ρl present the gas and liquid phase’s density, respectively. The fluid computational cell is defined as qth for αq = 1 while qth for αq=0 and 0 < αq < 1 for empty fluid indicating that the cell has contact between the qth fluid and other fluids. So, the total volume fraction of liquid-gas adds up to 1 which is as follows: (3) Therefore, the equations [1, 2] are to be solved for the volume fraction in order to track the gas and liquid interface and written as: (4)

Both the air flow and the liquid-dropping film exhibit some degree of turbulent behavior. Banerjee and Isaac [63] projected gas-liquid flow using several turbulence models, such as the RNG k-ε model and the conventional k-ε model. Comparing the RNG k-ε model with the experimental measurement yielded superior results than other turbulence models in terms of accuracy. Thus, the RNG k-ε turbulence model was used in this investigation. The following are the details of the governing equations used in the study to describe how moist air passes through a liquid desiccant, including the VOF model, energy, momentum, species, and turbulence equations [1, 2, 63]. Thus, the continuity equation is written as follows:

The momentum equation is written as follows: (6) The energy equation is written as follows: (7) Where, ρ and u are the fluid density and velocity, respectively. While keff is the effective thermal conductivity, F is the momentum source term and SE is the energy source term, respectively. To calculate the turbulent kinetic energy and its rate of dissipation, the transport equation is repre-

(8) Where, Gk represents the generation of turbulence kinetic energy due to mean velocity gradients, while Gb represents the generation of turbulence kinetic energy due to buoyancy. The turbulence dissipation rate transport equation can be expressed as follows: (9) Where ε is the turbulence dissipation rate by considering Cε1 = 1.42, Cε2 = 1.68 , and αk = 1.393. The following is an expression for the transport equation: (10) Where, Yk,q denotes the k component’s mass fraction of qth phase, and Slg,k presents the source term of mass transfer. Again, the continuous surface tension model is to be used to explain the surface tension at the gas-liquid free surface, and it looks like this: (11) Where, σi,j represents the coefficient of surface tension. Moreover, the dynamic dehumidification process was examined by applying the mass transfer theory of penetration. It was not taken into account the mass transfer barrier at the gas-liquid contact [1, 2, 7]. The coefficient of overall mass transfer (Mg) may be written as follows: (12) Where, hm,g and hm,l are the gas and liquid phases’ respective local mass transfer coefficients, while ψ is expressed as a function of temperature and concentration of liquid desiccant in the following way: (13) According to the literature by Zhang et al. [7], the values of αi (where i = 0, 1, 2, and 3) are estimated and the coefficients of local mass transfer are calculated as follows:

Where, Dg and Dl are the diffusion coefficients of the gas and liquid phases, respectively, at a certain contact time (tc). Again, the mass transfer source is estimated at a contact area of A by considering the bulk air humidity ratio (Wg,b) and equilibrium air humidity ratio (Wg,e) to the desiccant, respectively. Thus, the source of mass transfer (Slg,k) is written as follows: (15) (16) Then, the mass transfer’s latent heat is the energy source term (SE), which is shown as follows: (17)

Boundary Conditions and Setup In this work, the numerical problem is solved using an ANSYS fluent pressure-based solver. The mass transfer mechanism is used at a constant rate of 50 mol/s. The LiCl solution mixture used consists of 30-44 % LiCl and 56-70 % water liquid, while the moist air consists of 1.6-2.4 % water vapor and 97.6-98.40 % air. The simulation utilizes a velocity inlet boundary at the air exit whereas the solution is an inlet and a pressure outlet boundary for the entry of the moist air and the exit of the solution. The plate is set as a wall boundary, and the right side of the moist air flow uses a symmetry boundary. The VOF model tracks the interface of the two-phase flow, incorporating surface tension effects within the interfacial forces considered in the simulation. To complete the simulation the following fluid flow assumptions are considered: (i) the flowing fluid is incompressible, (ii) the fluid is two-phase flow, (ii) the unsteady gas-liquid fluid flow conditions, (iv) the no-slip boundary condition, and (v) negligible body force. However, Table 1 lists the input parameters for this study, while Table 2 displays the properties of fluids. Once more, Table 3 lists the numerical setup along with the solver setting parameters.

Table 1. Input parameters for the simulation in this study Plates

Table 2. Properties of air, LiCl, and liquid & water vapor [64] Parameter

Table 3. Numerical setup with solver setting parameters and solution methods Parameter

Simple

Figure 4. Illustration of the mesh geometry and structural grids of (a) vertical smooth, (b) vertical rough, and (c) inclined rough plates of LDD.

Table 4. Mesh information at various plate configurations of LDD Information

Grids Independence and Validation The mesh independence test was carried out using a series of simulations for vertical smooth, vertical rough, and inclined rough plates. The commercial software ANSYS meshing was used to mesh the inclined and vertical plates. Better convergence and wall function were the goals behind the development of face meshing. The meshing with the quadrilateral element of the blunt body for the whole domain is shown in Figure 4. The mesh geometry and structural grids of the vertical smooth plates and the vertical rough plates are illustrated in Figures 4(a) and (b), respectively, while Figure 4(c) represents the mesh geometry and structural grids of the inclined rough plate where the inclination of the rough plate is 300 from the vertical axis or 600 from the horizontal axis. Moreover, Figure 5 demonstrates the study carried out to investigate the impact of grid resolution on the simulation outcomes which is commonly known as grid independence. When the gird number exceeds 150270, there is no effect on the computational result of the absolute moisture content of air. Even though, the moisture content of the air passing through the LDD plates did not noticeably change with additional grid refining. As a result, the grid numbers 182500, 178520, and 160270 were adopted in this work for vertical smooth, vertical rough, and inclined rough plates,

Figure 6. Variation of mass fraction of outlet air corresponding to the width of the plate.

respectively, while Table 4 shows the mesh information about the various plate configurations. Figure 6 depicts the variation of mass fraction outlet air with respect to the width of vertical smooth and rough plates. From this figure, the mass fraction of outlet air is gradually increased along the plate width which indicates a similar trend to the findings by Lu and Lu [2]. The data suggest that the deviation in the graphs for smooth plates is around 2.5%, while for vertical rough plates, it is approximately 3.9%. By the way, the deviation of this work with Lu and Lu [2] is relatively small and it varies depending on the plate being analyzed. This implies that the deviation in the results is less pronounced in the case of smooth plates when compared to rough plates.

Results And Discussion

In this study, the impact of air velocity on the mass transfer enrichment and temperature increments of LDD was numerically studied by ANSYS. The 30% LiCl solution is used as an absorbent and hence, the result section presents the findings, while the discussion section interprets and explains their meaning in the context of the research. Moreover, the LiCl solution’s mass concentration is taken into account as 30%, 33%, 36%, 40%, and 44%, respectively,

for the justification of the influence of various concentrations of LiCl solution. While the inlet air velocity is considered as 0.5, 1.0, 1.5, and 2.0 m s-1, respectively. This analysis was validated with the study of Lu and Lu [2], where the deviations are found 2.5% and 3.9% for vertical smooth and vertical rough plates, respectively, corresponding to the measurement of the mass fraction of outlet air. Overall, this simulation was effectively carried out for all the variabilities such as moisture content, outlet air temperature, and concentration of LiCl solution within the error limit of 6.75%. Influence of Air Velocity The research has been done to examine how air velocity affects LDD’s mass transfer enhancement for the three geometries as the main test section (vertical smooth, vertical rough, and inclined rough plates). For this consequence, it is necessary to adjust the air velocity in practical dehumidification applications. However, Figure 7 shows the contours of the moist air mass fraction under various air inlet velocities for vertical smooth, vertical rough, and inclined rough plates, respectively. The red area denotes

the wet air, while the blue area is the liquid desiccant. It is noticed that the reduction in the air velocity leads to a significant decrease in the outlet mass fraction of water vapor from moist air. When the velocity of the incoming moist air is higher, the outlet air mass fraction increases, which suggests that the liquid LiCl solution cannot absorb enough moisture due to insufficient time. As the velocity of inlet moist air increases in Figure 7(a), the outlet air mass fraction increases up to 2.1 % which indicates a decrease in water vapor absorption as the LiCl liquid solution did not get sufficient time to absorb moistness. This goes the same for the vertical rough plate up to 2 %, shown in Figure 7(b). But in the case of an inclined rough plate, the amount of mass fraction decreases which means an increase in water vapor absorption because of inclination and ribs the friction increases which is observed in the contour in Figure 7(c). It is seen that the red area is getting yellow from the red zone as air gradually goes upward. At the ribs, it is blue because there is no mass of water vapor at the boundary but it is getting greener at the contact surface with moist air.

Figure 7. Contours of outlet air mass fraction under various air inlet velocities for (a) vertical smooth, (b) vertical rough, (c) inclined rough plates of LDD (for 30% LiCl solution).

For 30% LiCl solution Inclined rough plate Vertical rough plate Vertical smooth plate

Figure 8. Outlet air mass fraction under various air inlet velocities.

Figure 8 describes the variation of absolute moisture content of air under various air inlet velocities with 30% LiCl solution that indicates the comparison of moisture absorption vertical smooth, vertical rough, and inclined rough plate. To put it simply, the outcomes show that as the velocity of incoming moist air rises, the ability of the liquid LiCl solution to absorb moisture decreases. This effect is seen in both vertical smooth and vertical rough plates. The absolute moisture contents of outlet air are recorded as the maximum of 2.1% and 2.0% for vertical smooth and rough plates, respectively. In the case of an inclined rough plate, the moist air mass fraction decreases, indicating an increase in water vapor absorption with an increase in the inlet air velocity. This is because the inclination and ribs increase friction, which helps to absorb more moisture. In this case, the absolute moisture content of outlet air is found a maximum of 1.8%. Figure 9 shows the contours of outlet air temperature under various air inlet velocities for vertical smooth, vertical

Figure 9. Contours of air temperature under various air inlet velocities for (a) vertical smooth, (b) vertical rough, and (c) inclined rough plates of LDD (for 30% LiCl solution).

Inclined rough plate Vertical rough plate Vertical smooth plate

Figure 10. Variation of outlet air temperature under various air inlet velocities.

Figure 11. Variation of outlet air temperature along the length of the plates.

rough, and inclined rough plates, respectively. As the velocity of inlet moist air increases in Figure 9(a), the outlet air mass fraction increases indicating a decrease in water vapor absorption and hence, the temperature of the outlet air does not decrease much from inlet. As the velocity is increasing the red area is increasing because of the short time of absorption. This goes almost the same for vertical smooth and rough plates but slightly good findings are observed in the vertical rough plate shown in Figure 9(b). In an inclined rough plate shown in Figure 9(c), the amount of temperature decreases more than in vertical smooth and rough plates which means an increase in water vapor absorption due to the inclination and ribs, which create friction. When moist air flows that test section at 0.5 m/s in the upward direction into the vertical smooth, vertical rough, and inclined rough plate configurations, the minimum outlet air temperatures are recorded as 301.3 K, 301.2 K, and 301.1 K, while maximum temperatures are found as 302.14 K, 302.0 K, and 301.85 K at a moist air velocity of 2.0 m/s, respectively. Figure 10 illustrates the variation of air temperature under various air inlet velocities of vertical and inclined plates of LDD, respectively. The temperature of outlet air rapidly increases with raising the inlet air velocity until 1.1 m/s and then gradually increases along the vertical and inclined plates. The rapid increase of outlet air temperature indicates less water vapor absorption and hence, the temperature of outlet air for the vertical smooth plate is greater than the rough plate. Moreover, in the inclined rough plate, the temperature of outlet air is quite smaller than the vertical smooth and vertical rough plate. So, inclination is quite efficient for dehumidification purposes.

velocity increases. The change in air temperature is affected by two factors. Initially, the desiccant cools the air through sensible heat exchange because its temperature is lower than the air. Subsequently, as water condenses from the air, latent heat is released, raising the air temperature. The interplay between these two processes results in the observed fluctuations in the outlet air temperature. Figure 11 shows the temperature distribution of outlet air along the length of the plates of LDD. The length of plates (150 mm) from bottom to top, the temperature of the moist air is decreasing because the water vapor is absorbed by LiCl solution. Along the plate length, the outlet air temperature decreasing is quite less for the vertical smooth plate. But it is quite higher in vertical rough plates and highest in inclined rough plates. The outlet air temperatures were measured through the plate width by considering 30% LiCl solution and 1.5 m/s moist air inlet velocity. The maximum air temperature was measured at 302.8 K for the plate length of 0.02 m, while the minimum temperatures were recorded as 299.8 K, 299.5 K, and 299.15 K at the plate length of 0.15 m for vertical smooth, vertical rough, and inclined rough plate configurations, respectively. Figure 12 interprets the temperature distribution of LiCl solution along the length (150 mm) of the plates of LDD. The temperature is going upward as the liquid LiCl desiccant is falling down from the top of the plates. Because of the absorption of moisture from the air, the temperature rises as comparatively warm air releases heat and liquid desiccant absorbs it. The trend is quite similar for all cases, but the temperature rise is less in the case of the inclined rough plate. Because of inclination, the absorption is more and so the temperature rise is less. The minimum temperature of LiCl solution was measured at 298 K, while the maximum temperatures were recorded as 302.5 K, 302.05 K, and 301.14 K at the plate length of 0.15 m for vertical smooth, vertical rough, and inclined rough plate configurations, respectively.

Influence of the Varying Lengths of the LDD Plates The temperature of air initially increases in the inlet section and then slightly decreases along the plate length when passing the outlet section by variation as the inlet air

Figure 12. Variation of the temperature of LiCl solution along the length of the plates.

Impact of Water Concentration of Inlet Air The outlet air mass fraction of the different plates is significantly influenced by the entrance air vapor percentage. Figure 13 shows the contours of the outlet air mass fraction under various inlet air vapor fractions for the vertical smooth & rough and inclined rough plates of LDD. The mass fraction of outlet air increases with increasing the inlet air vapor fraction along the plate length. The amount of water vapor absorption decreases with the increase of vapor percentage in inlet moist air. Figure 13(b) shows relatively good results for the vertical rough plate to the water absorption indicating less moisture content of outlet air than the vertical smooth plate from the incoming moist air to the LiCl solution. However, the moisture content of outlet air for the vertical rough plate increased instead of decreasing trend than the vertical smooth plate after the inlet air vapor quality of 0.018. Moreover, in an inclined rough plate shown in Figure 13(c), the friction increases due to the inclination, which is responsible for increasing the contact

Figure 13. Contours of outlet air mass fraction under various inlet air vapor fractions for (a) vertical smooth, (b) vertical rough, and (c) inclined rough plates of LDD (for 30% LiCl solution).

Conclusion

This study sheds light on the importance of plate configurations in liquid desiccant dehumidifiers and their impact on the performance of dehumidification and gas-liquid flow. The findings highlight the significant role of inclined rough plates in enhancing dehumidification effectiveness by promoting increased water vapor absorption. The CFD simulation using the RNG k-ε turbulence model effectively predicted the dehumidification performance of a falling film LDD with various plate configurations under unsteady gas-liquid flow conditions. The key findings of this work are as follows: • Increased inlet air velocity results in higher outlet air mass fractions, indicating reduced water vapor absorption due to insufficient time for the moist air and LiCl solution to come into contact. • Among the plate configurations, the inclined rough plate significantly enhances dehumidification efficiency by generating liquid film waves and increasing the contact time, leading to a greater reduction in the outlet mass fraction of water vapor. • In the inclined rough plate, it is noticed that the dehumidification enhancement goes up to 10.8% compared to the use of a smooth plate, particularly when the inlet air velocity is 1.5m/s. • Higher inlet desiccant concentrations improve water vapor absorption, with an optimal range identified between 30% to 40% LiCl concentration. The study outcomes specify that the efficiency of enrichment remains consistent even with solution concentrations exceeding 36%, while the benefits may level off beyond a certain point. • The inclined rough plate configuration consistently showed superior performance compared to vertical smooth and rough plates, making it the most effective design for improving dehumidification in LDD systems. This research highlights the importance of factors like inlet desiccant concentration and air velocity in optimizing falling film liquid desiccant systems for dehumidification. Moving forward, further exploration and experimentation in this area can lead to advancements in liquid desiccant technology, ultimately contributing to energy savings and improved indoor air quality in various applications. By identifying these key factors and their effects, this research provides valuable insights into the design and operation of efficient and sustainable dehumidification systems.

Coefficient of diffusion, m2/s Energy, J/kg Source term momentum, N/m3 Kinetic energy of turbulence produced by mean velocity gradients, kg/(m. s3) Buoyancy-induced kinetic energy creation of turbulence, kg/(m. s3) Gravitational acceleration, m/s2 Enthalpy, J/kg Coefficient of local mass transfer, kg/(m2.s) Latent evaporation heat, J/kg Mass flux, kg/(m3 .s) Thermal conductivity, W/ (m. K) Distance of liquid flow, m Local total mass transfer coefficient, kg/(m2.s) Pressure, Pa Atmospheric pressure, Pa The bulk desiccant solution's partial vapor pressure, Pa Source of mass transfer at phase interface, kg/m3 Source term energy, W/m3 Additional term, kg/(m. s3) Time, s Contact time, s Temperature, K Bulk air humidity ratio, kg/kg Air humidity ratio in equilibrium with desiccant solution, kg/kg

Greek symbols α Volume fraction µ Dynamic viscosity, kg/ (m. s) ψ Concentration ratio ρ Density, kg/m3 ε Turbulence energy dissipation rate σij Coefficient of surface tension, kg/m3 Subscripts a Air eff Effective g Gas phase l Liquid phase s Solution i,j Indication of the coordinate axis

Authorship Contributions

Md. Tamzid Shaharier (M. T. Shaharier): Conceptualization, methodology, investigation, data acquisition, analysis, writing original draft, and editing manuscript. Dipayan Mondal (D. Mondal): Conceptualization, methodology, investigation, data acquisition, analyzing and interpreting the data, writing, review and editing, discussion, supervising.

Md. Abdul Hasib (M. A. Hasib): Analyzing and interpreting the data, reviewing and editing, and writing. Md. Ashraful Islam (M. A. Islam): Reviewing editing, and writing of reviewer response.

Data Availability Statement

Data will be available upon reasonable request from the corresponding author.

Conflict Of Interest

The authors declared no conflicts of interest associated with this publication, and there had been no significant financial support for this work that could have influenced its outcome.

Ethics

There are no ethical issues with the publication of this manuscript.

Share and Cite