A three-dimensional physical model was developed for both no-load and potato storage experiments (Figure 4 illustrates the potato storage model). The dimensions of the potato pile, platform, and internal facilities match those of the actual experimental setup. The humidification inlet is located near the fan supply port, enabling moisture to mix with the air from the cooler. For steady-state analysis, the humidification port is simplified to coincide with the cooler port. To address distinct flow characteristics and heat/mass transfer mechanisms in the potato porous medium, the warehouse space is divided into two zones: free airflow and potato stacking. This division facilitates detailed analysis of flow field distributions in each region.

The fluid flow, heat transfer, and mass transfer within the experimental platform are governed by three fundamental equations: the continuity equation, the energy equation, and the momentum equation. When accounting for mass transfer processes like air humidification, the system must also satisfy the law of conservation of components. The high-velocity airflow generated by the chiller creates turbulent conditions within the storage platform, with heat transfer dominated by convection. In contrast, airflow within the potato piles is significantly reduced due to obstruction, resulting in heat transfer primarily driven by conduction. The flow and heat transfer characteristics in each subdomain are described by their respective governing equations (Verboven et al., 2006).

In the free airflow region, the distribution of airflow and the temperature field can be approximately described as follows:

where v→ is velocity vector (m/s), ρ is the density (kg/m³), t is the time (s), S_φ is the generalized source term, Γ_φ is the generalized diffusion coefficient, φ is the general variables, the value of which is different, the equation is a different equation.

When φ = 1, the equation is the continuity equation (mass conservation equation):

where u, v, and w are velocity scalars in the three directions of the x, y, and z axes.

When φ = v = [uvw], this equation is the conservation equation of momentum:

where p is the pressure on a microelement on a fluid (Pa), τxx,τyx,τzx is the component of viscous stress τ acting on the surface of fluid microparticles in x, y, z directions due to molecular viscosity (Pa).

When φ = T, this equation is the energy conservation equation. Among them, T is the local temperature (°C), cp is the specific heat capacity (J/(kg °C)), λ is the thermal conductivity (W/(m°C)).

To investigate the adjustment of environmental humidity during potato storage, the component transport equation should also be introduced:

where Yi is the mass fraction of component i, ρ_a is the air density (kg/m³), R_i is the net generation rate of component i, G_i is the additional generation rate caused by discrete phases and custom source terms, J_i is diffusion flux density, vi is the component flow velocity (m/s). In turbulent motion, the diffusion flow density, J_i of component, i can be calculated using the following:

where D_i,m is the mass diffusion coefficient of component, i in the mixed gas, Sc_t is the turbulent Schmidt number, which takes the value of 0.7, μ_t is the turbulent viscosity (N s/m²), D_i,T is the diffusion coefficient of heat.

The ellipsoidal shape of potatoes creates gaps when stacked, enabling air to flow through. Thus, the potato pile can be modeled as a porous medium. Within this medium, porosity exists, and fluid flow interacts with the solid non-porous regions, generating resistance. This resistance causes a pressure drop and velocity reduction due to inertia and viscous losses. In CFD simulations, the effect of the potato pile on airflow is incorporated by introducing a momentum source term S into the momentum conservation equation. The resistance of the porous medium to gas flow can be described using Darcy’s law, expressed as follows:

where α is the permeability, C is the inertial drag coefficient, which takes the value of 0.7, μ is fluid viscosity (N s/m²).

The thermal and physical properties of the materials used in the storage experimental platform, including the envelope, fan shell, potatoes, air, and internal heat source, are summarized in Table 1. The air supply port of the chiller is modeled as a velocity inlet, with turbulence characteristics defined by a turbulence intensity of 5% and a hydraulic diameter of 0.4 m. Temperature, humidity, and velocity parameters are determined based on subsequent simulations. The chiller returns air outlet and the potato storage model are designated as free outflow boundaries, requiring no additional specifications. All wall surfaces of the storage platform are defined as non-slip adiabatic boundaries. The internal heat source and the initial temperature of the potatoes are set to 10°C, based on the average temperature during the potato storage month, as illustrated in Figure 4.

In selecting a turbulence model, the shear stress transport k-ω model (SST k-ω) was chosen for this study because the SST k-ω model combines the advantages of the standard k-ω model, the standard k-ε model, and the RNG k-ε model. This combination results in an accurate and reliable turbulence prediction model for both the boundary layer and the free shear layer regions (Gong et al., 2021; Defraeye and Radu, 2018). Considering the heat and mass transfer processes in this numerical simulation, it is necessary to enable the energy equation and the species transport model and set the gravitational acceleration to 9.81 m/s². The semi-implicit method for pressure-linked equations (SIMPLE) is used to couple the pressure to velocity. Additionally, to better describe turbulence and flow, the convective terms are implemented using the second-order upwind scheme. A convergence criterion of 10⁻⁶ was imposed for momentum, continuity, and turbulence and 10⁻²⁰ for the scalar and energy equations. This numerical simulation was conducted on a 64-bit computer equipped with dual 64-core Intel^® Xeon^® 6430 processors (2.10 GHz) and 256 GB of memory.

To enhance the model’s accuracy, local refinement is applied at the inlet and outlet of the air cooler. Based on the physical structure of the experimental platform, a polyhedral mesh is utilized. Considering the balance between computation time and model accuracy, a sensitivity analysis of the mesh is necessary. In this study, the temperature values at two measurement points after 120 s of simulated cooling are selected for comparative analysis to verify the independence of the grid. And the number of divided meshes used for testing is presented in Figure 5.

The temperature values of two designated points following a simulated cooling period of 120 s were compared to assessing the mesh’s independence. Measuring point 1 (coordinates: 1,000 mm, 1,000 mm, 4,000 mm) is situated adjacent to the chiller’s air supply, while measuring point 2 (coordinates: 0 mm, 0 mm, 2,500 mm) represents the three-dimensional center of the storage experimental platform, as shown as Figure 5a. Across seven grid resolutions, the simulation outcomes remained consistent. Line graphs depicting the temperatures at the two measuring points for various grid resolutions are presented in Figure 5b. As the number of mesh increases from 13 × 10⁴ to 214 × 10⁴, the temperature fluctuations at the two measurement points significantly decrease and stabilize once the grid count reaches 166 × 10⁴. Therefore, the influence of grid count on the calculation results can be disregarded when it reaches 166 × 10⁴, indicating that the temperature value at this mesh count falls within an acceptable error range, thereby meeting the criterion for mesh independence. Consequently, the mesh quantity for the potato storage model should exceed 166 × 10⁴. Following the mesh independence verification, the potato storage model was locally refined at the chiller inlet and moisture outlet, resulting in a final mesh count of 172 × 10⁴.

Two parameters, the root mean square error (RMSE) and the mean relative deviation (ARD), were introduced to assess the agreement between numerical simulation results and experimental data (Gong et al., 2021). RMSE and ARD are defined as follows:

where S_i is the current simulation result, E_i is the existing experimental data, and n is the total number of samples data points.

Uniformity of temperature and humidity distribution refers to the equilibrium of air temperature and humidity across the experimental platform (Defraeye and Radu, 2018; Han et al., 2023). A more uniform distribution of temperature and humidity results in a more consistent storage effect, indicating that the performance of the temperature and humidity control equipment is being fully utilized. To evaluate the uniformity of airflow organization, we introduce the temperature coefficient of variation and the relative humidity coefficient of variation. The temperature coefficient of variation is typically calculated using thermodynamic temperature. A smaller coefficient of variation indicates a more uniform distribution of airflow (Pu et al., 2014). The coefficients of variation for temperature and relative humidity are calculated as follows:

where CV_T is the temperature variation coefficient, m is the number of measuring points, T_i is the temperature of the measurement point i, T_ave is the average temperature of all measurement points, CV_H is the coefficient of variation of relative humidity, H_i is the relative humidity of measuring point i, H_ave is the average relative humidity of all measurement points.

Air humidity ratio is a critical factor in determining the degree of changes in air humidity, as it can mitigate the effects of the coupling relationship between temperature and humidity in environmental assessments. In experimental settings, the inter-play between air humidity and temperature suggests that when the humidity content remains constant, variations in temperature will alter the air’s capacity to dissolve water vapor. Specifically, an increase in temperature results in a decrease in the relative humidity of the air, while a decrease in temperature leads to an increase in relative humidity (Song et al., 2024). Since the air humidity data collected in experiments is typically expressed as relative humidity, conversion is necessary to derive accurate air humidity ratio measurements. The equation for calculating air humidity ratio is as follows:

where d is the air humidity ratio (g/kg), ε is the ratio of the dry air specific gas constant to the water vapor specific gas constant, which is 0.622; Hr is the relative humidity of air (%), Ps is the saturation pressure of water vapor (Pa), and Pl is the air pressure (Pa), Pl= 9 × 10⁴ Pa according to local conditions.

The variables of temperature, humidity, and velocity appearing in the numerical results are normalized by defining dimensionless variables utilizing the scaling factors, which were defined by Equations 13, 14:

where T∗ is denoted the dimensionless temperature, Tp is the arithmetic average value of the volume-averaged temperatures, T∞ is the air temperature, TW is the initial temperature, V∗ is the dimensionless velocity, Vp is the arithmetic average value of the volume-averaged velocity, V∞ is the air velocity, VW is the initial velocity.

Figures 6a,b display the temperature and humidity curves for each measuring point in Section 1 and Section 2 during a 60-min temperature control period on the experimental platform. The initial temperature and humidity at each measurement point ranged from 20°C to 22°C and from 25 to 32%, respectively. After 18 min of operation, the air within the experimental platform enters a cyclic state characterized by “slow passive heating” in response to elevated external temperatures and “rapid active cooling” triggered by the refrigeration system when temperatures surpass a set threshold, as depicted in Figure 6. This transition occurs when the sensor linked to the refrigeration system reaches the desired temperature, prompting the system to stop operating. As a result, temperatures at various measurement points exhibit fluctuations and decline, ultimately stabilizing within the target temperature range of 3 to 5°C after 60 min. The relative humidity of each measurement point is affected by the decrease in temperature fluctuations, and it also shows a phased rise and fall, and the overall slow rise is shown. At 60 min, the humidity of each measurement point is between 44 and 50%.

As illustrated in Figure 6a, each measurement point exhibits a rapid decrease within 18 min after the onset of cooling. Notably, the temperatures at measurement points 1-1 and 1-2 decrease gradually, remaining above 7°C after 18 min of cooling. This phenomenon can be attributed to the higher air temperature flowing towards the ground, slower flow rate, and weaker convective heat exchange effect, resulting in a slower temperature decline at these points. Conversely, other measurement points situated at higher spatial positions experience lower temperatures and higher flow velocities during the influx of cold air, leading to a pronounced convective heat transfer effect and consequently faster cooling rates, with temperatures dropping below 5°C by the 18-min mark. Simultaneously, the relative humidity at measuring point 1-1 was relatively high, reaching 32%, due to the influence of human respiration during the setup of the data acquisition equipment. Once cooling commenced, the humidity at each measurement point increased significantly as the temperature rapidly decreased, exceeding 40% within 18 min due to the coupling effect of temperature and humidity.

Figure 6b illustrates the changes in temperature and humidity at each measuring point in Section 2 during a 60-min temperature control of the storage experimental platform. The data indicates that the temperature at each measuring point decreases significantly 18 min after cooling commences. However, due to the obstruction caused by the data acquisition equipment, the rate of temperature decreases at measurement points 2-1 and 2-4 is slightly slower than at the other points, although the overall temperature drop rates across all measurement points remain relatively consistent. Within 18 min, all points, except for 2-1 and 2-4, record temperatures below 5°C. Following this period, the temperature fluctuations at each measuring point are consistent with those observed in Section 1, exhibiting similar variation amplitudes. This synchronous behavior likely results from the distance of these points from the data acquisition equipment, which minimizes its influence. After four cycles of temperature fluctuation, the temperatures at each measuring point stabilize around 3°C to 4°C at the 60-min mark, with minimal differences between the points, indicating effective cooling.

Correspondingly, the initial humidity at each measuring point is approximately 25%. During the first 18 min of cooling, the humidity at all measurement points increases rapidly. However, the humidity at points 2-1 and 2-4 rises more slowly due to the coupling effect between temperature and humidity. The temperature at these points decreases more rapidly than at the other locations, resulting in a slower increase in humidity. By the 18-min mark, the humidity at all measuring points exceeds 40%. Following this period, the humidity at each point rises gradually due to minor temperature fluctuations, and overall, the humidity fluctuations remain minimal. After 60 min, the humidity at all measurement points slowly increases to between 45 and 50%.

As depicted in Figures 6c,d, the temperature and humidity change curves for each measurement point in both Section 1 and Section 2 are presented during the temperature and humidity coordinate control of the experimental platform. The initial temperature at each measuring point ranges from 16 to 19°C, and the initial humidity at each measuring point is approximately 30%. After 15 min of operation, the refrigeration system reduced the temperature of all measurement points to below 5°C, with humidity levels exceeding 75%. Each measurement point exhibited a temperature fluctuation pattern similar to that observed during temperature control alone, albeit with a notably weaker overall temperature decrease compared to a simple cooling process. This diminished temperature drop can be attributed to the elevated air humidity within the storage experimental platform, leading to an increase in the specific heat capacity of the air and subsequently reducing the amplitude of temperature fluctuations. Initially, each measurement point reached temperatures ranging from 3 to 5°C upon the initial operation of the refrigeration system. The introduction of humidified air at the inlet significantly influenced the heat exchange process within the platform, enhancing its efficiency.

Upon initiation of the cooling process, Figure 6c illustrates a rapid decrease in temperature across all measurement points. Notably, measurement points 1-4 and 1-3, positioned near the top of the platform in proximity to the air cooler, exhibit accelerated cooling due to enhanced heat exchange with the colder air. Conversely, the cooling rate is comparatively slower at other measurement points. Subsequently, activation of the humidifier leads to a swift increase in humidity levels at all measurement points. Particularly, measurement points 1-4, situated closer to the humidification port, experience a more rapid humidity increase, reaching a peak of 81%. Between 15 and 26 min, a temporary cessation of the refrigeration system causes a rise in temperature at all measurement points, consequently decelerating the rate of humidity increase. Measurement points 1-2 and 1-3, located farther from the humidification port, exhibit a slower rate of humidity increase compared to other measurement points. At 26 min, humidity levels exceeded 85% at all measurement points, with points 1-4 and 1-5 reaching 90%. Subsequently, the humidifier ceased operation, and the cooling system resumed, leading to condensation of water vapor and a rapid decline in humidity levels across all points. Following another pause of the cooling system, the temperature within the storage experimental platform gradually rose, resulting in a gradual decrease in humidity levels at all points due to the interplay of temperature and humidity. At 39 min, the humidifier initiated a new cycle of humidification, causing a subsequent increase in humidity levels at all points. By 56 min, the impact of cold air led to another decline in humidity levels at all points. Overall, following the initial humidification phase by the humidifier, humidity levels at each point generally remained between 85 and 90%.

After the initiation of the refrigeration process, Figure 6d illustrates a rapid decrease in temperature at each measuring point, with a cooling rate of approximately 1.0°C/min. Within 17 min, all measuring points fell below 6°C, entering a fluctuation state consistent with the temperature control process. Concurrently, as the humidification process commenced, there was a swift increase in humidity at each measuring point. Points 2-5 experienced a faster rise in humidity due to increased air flow, while points 2-1 exhibited a slower increase due to equipment obstruction. By the 27th minute, the humidity levels at all experiment points exceeded 80%, with points 2-5 reaching 95%. Subsequently, the humidifier ceased operation, and the refrigeration system resumed, causing condensation of water vapor due to the cold air, leading to a rapid drop in humidity at each measuring point. At the 32nd minute, the humidifier initiated a new round of humidification, resulting in a subsequent rise in humidity at each measuring point. By the 47th minute, the humidity levels decreased again at each point due to the influence of the cold air. At the 53rd minute, the humidifier recommenced operation, leading to a rise in humidity at each point. Following the initial humidification phase, the humidity levels at each point fluctuated around 80%, predominantly maintained between 85 and 90% most of the time, indicating an effective humidification process.

Given the proximity of Section 1 to the sensor of the monitoring system, its air-flow distribution significantly influences the analysis. Therefore, only the temperature and humidity variation coefficients of the temperature control process in Section 1 are examined.

Figure 7a shows the temperature variation coefficient curve of Section 1 at different initial temperatures (16°C and 21°C) during the temperature control process. When the initial temperature is 21°C, the coefficient of temperature variation remains consistently below 0.35%. However, it exhibits a phased fluctuation pattern: during the initial 0–6 min of cooling, there is a rapid increase in the coefficient of temperature variation from 0.16% to over 0.30% due to varying cooling rates at different measuring points. Subsequently, from 6 to 15 min, as the refrigeration equipment continues to operate, there is a slight fluctuation in the temperature variation coefficient, although the overall value remains high. Between 15 and 18 min, as the refrigeration process progresses, the temperatures at each point gradually converge, leading to a decline in the temperature variation coefficient. Following this, from 18 to 25 min, as the refrigeration equipment ceases operation and natural convection facilitates heat transfer, the temperatures at each point further converge, resulting in a continued decrease in the temperature variation coefficient.

Lastly, from 25 to 60 min, the coefficient temperature variation fluctuates periodically due to the intermittent operation of the refrigeration equipment and natural convection heat transfer, with the value generally staying below 0.20%. During the temperature control process starting at 16°C, the temperature coefficient of variation remains below 0.35%. However, it peaks at 8 min, starts fluctuating slightly, and gradually decreases. By the 39th minute, it drops to around 0.16%, after which the rate of decline slows significantly. Subsequently, the temperature coefficient of variation stabilizes within a narrow range. Notably, after 60 min of temperature control at various initial temperatures, the temperature coefficient of variation remains consistently below 0.2% in Section 1. This indicates excellent air temperature uniformity within the storage experimental platform.

The coefficient of variation trend of temperature in Figure 7b mirrors that of Figure 7a, exhibiting an initial increase, followed by a decrease, and subsequent fluctuations over time. The maximum temperature coefficient of variation in both instances was initially below 1%, stabilizing at less than 0.4% post-initial refrigeration. In comparison, the coefficient of variation for relative humidity surpasses that of temperature, exceeding 3% with a peak surpassing 7%, as shown as Figure 7c. Notably, when the initial temperature is 21°C, significant fluctuations in relative humidity coefficient of variation occur at 2–26 min and 26–56 min due to elevated external temperatures and the concurrent operation of the air cooler and humidifier, aligning with air temperature and humidity patterns. Conversely, at an initial temperature of 16°C, the relative humidity coefficient of variation remains relatively stable, hovering between 3 and 5% after 8 min, reflecting minimal changes in relative humidity. Overall, the relative humidity coefficient of variation aligns with findings in literature (Wang et al., 2023; Guo et al., 2019), underscoring the effective humidification uniformity of the experimental setup.

The humidification process of the humidifier impacts on the moisture content of the air. Therefore, the analysis focuses solely on the moisture content of the air in Sections 1 and 2 during the temperature control process. Figure 8 illustrates the variation in air humidity ratio across different initial temperatures and cross sections during the temperature control process. Overall, the trend of humidity ratio in the air at each point aligns with the temperature trend.

Upon cooling initiation, hot air enters the return air outlet of the air cooler and contacts the cooling pipe, leading to pre-condensation of water vapor and subsequent water adherence to the air cooler coil. This process results in a rapid decrease in air humidity ratio at each measurement point, as depicted in Figure 8a. Subsequently, after 18 min, as air temperature naturally rises, water solubility in the air increases, causing some water to evaporate and increase moisture content. Upon the resumption of the fan operation for cooling, some water vapor condensed, further decreasing air moisture content. Numerically, between 18 and 60 min, the fluctuation amplitude of air humidity ratio at all measurement points remains below 0.2 g/kg. The natural recovery period of air humidity ratio aligns with the temperature recovery cycle, approximately 10 min, indicating a natural recovery rate of humidity ratio at each point below 0.02 g/(kg min).

As depicted in Figure 8b, the decrease in air moisture content halts 10 min after cooling due to the low initial temperature, transitioning into a cyclic fluctuation phase. However, the alteration in air humidity ratio differs from that illustrated in Figure 8a. Following the onset of the fluctuation phase, a gradual increase in air humidity ratio is observed at all points, with particularly pronounced fluctuations at measurement point 1-1. The impact of the low initial temperature results in heightened influence from the heat source of the data acquisition equipment, leading to the evaporation and ascent of water under heat. Generally, the air’s moisture content across all points remains within the range of 2.25 to 2.75 g/kg, with the moisture content exhibiting subtle fluctuations, except notably at measurement points 1-4.

The air humidity ratio curve changes in Section 2 at initial temperatures of 21°C and 16°C are depicted in Figures 8c,d, respectively. Following the initial cooling phase, the air humidity ratio exhibits slight fluctuations within a limited range, with fluctuation amplitudes consistently below 0.2 g/kg.

Additionally, there is a subtle increasing trend in the rising phase of humidity ratio over time. Specifically, at 21°C, this period extends to 8 min, whereas at 16°C, it extends to 9 min. The growth rates of humidity ratio are below 0.025 g/(kg min) and 0.022 g/(kg min) for initial temperatures of 21°C and 16°C, respectively. These observations indicate minimal variation in air humidity ratio within the experimental platform, with fluctuations primarily attributed to temperature increases and chiller operation.

To assess the numerical model’s efficacy, a steady-state simulation was conducted on the no-load experimental platform model, with an inlet wind speed of 6 m/s and an inlet temperature of 3°C. Due to the proximity of cross Section 1 to the data acquisition device, which significantly influences it through heating and flow obstruction, data from this section were cooled for 60 min after data from cross Section 2, which is less affected, were selected for comparison. Table 2 presents a comparison between actual measured temperatures and simulated values after 60 min of refrigeration at an initial temperature of 21°C. The RMSE and ARD defined by Equations 8, 9, for points 2-1 to 2-5 were calculated. Results indicated that the maximum RMSE was 0.8°C, and the maximum ARD was 19.5%. The above values appear at the point 2-1. Considering the influence of the data acquisition equipment, the model’s simulation results are deemed more accurate.

To validate the precision of the potato storage numerical model, an equivalent-scale numerical model was constructed based on the experimental platform layout detailed in Section 2.4. Steady-state simulation calculations were conducted to regulate the environmental temperature and humidity within the model, using identical boundary conditions as outlined in Section 2.4. During the experiment, the equipment operates until the temperature and humidity data recorded by the sensor stabilizes and the fluctuation amplitude decreases significantly. Calculating the RMSE and ARD by comparing the experimental results with those from numerical simulations. Table 3 presents the measured and simulated temperature and humidity values at each experiment point, along with the corresponding RMSE and ARD. The data reveals a minor disparity between the simulated and measured outdoor air temperatures. Conversely, the air temperature within the potato pile consistently registers lower than the measured values. This discrepancy can be attributed to the hindrance of airflow by the woven bag during storage, resulting in suboptimal temperature regulation. Notably, the maximum RMSE of 1.1°C occurs at measurement point 6.

The modeled external air humidity of the potato pile consistently exceeds the measured values, with a maximum RMSE of 5.4% (and a maximum ARD of 6.39%) observed at measurement point 6. Similarly, the simulated air humidity within the potato pile surpasses the measured values, with higher RMSE and ARD values at each measurement point compared to the external air. This discrepancy can be attributed to the elevated moisture holding capacity of the air, impeded by the woven bags hindering water vapor transmission.

In summary, the error between the simulated and measured temperature and humidity values at each point was minimal, indicating that the numerical simulation method used for setting the internal heat source produces more accurate results. This model will be employed to study the flow field and the temperature and humidity fields of the potato storage environment under different operating conditions, and to analyze the optimal conditions for regulating the temperature and humidity in the potato storage environment.

Temperature is the primary factor influencing the storage effectiveness of potatoes. By setting the air supply temperature at 2°C, 3°C, 4°C, and 5°C, the impact of varying temperatures on the storage conditions can be compared and analyzed.

Figure 9 illustrates the distribution characteristics of potato storage environmental factors at the center of the potato pile, specifically at Z = 2.5 m and X = 0 m, under varying air supply temperatures. Overall, the temperature distribution within the experimental platform can be divided into two regions based on different air supply temperatures. Region 1 represents the relatively cooler air domain, which encompasses most of the potato pile area. In this region, the temperature distribution is relatively uniform, and the temperature regulation is effective. Region 2, on the other hand, is situated above the warmer potato piles, where the temperature decreases rapidly with an increasing distance from the center, exhibiting a distinct temperature gradient that extends into the air domain. This phenomenon occurs due to the obstruction created by the potato piles and the presence of internal heat sources, which hinder air circulation above the piles, resulting in lower airflow speeds and consequently poor cooling effects.

At an air supply temperature of 2°C, the temperature in Region 1 ranges from 2.15°C to 2.48°C, which is relatively low and poses a risk of frostbite to the potatoes, making it unsuitable for storage. In Region 2, the peripheral temperature varies between 2.67°C and 3.83°C, while the central temperature ranges from 4.12°C to 5.50°C. When the air supply temperature increases to 3°C, the temperature in Region 1 increases to between 3.13°C and 3.56°C. In Region 2, the peripheral temperature ranges from 3.71°C to 4.99°C, and the central temperature varies from 5.21°C to 5.92°C. Notably, the area within the 3°C to 5°C range has expanded significantly. At an air supply temperature of 4°C, the temperature in Region 1 reaches between 4.12°C and 4.37°C, while in Region 2, the peripheral temperature ranges from 4.61°C to 5.51°C, and the central temperature rises to between 5.67°C and 6.37°C, indicating an increase in high-temperature areas. At an air supply temperature of 5°C, the temperature in Region 1 ranges from 5.10°C to 5.43°C, with the peripheral temperature in Region 2 between 5.43°C and 6.27°C, and the central temperature reaching 6.40°C to 6.95°C. The expansion of high-temperature regions correlates with an increase in potato respiration intensity, which is detrimental to storage. Overall, the most suitable potato storage environment occurs at an air supply temperature of 3°C, aligning with the experimental results. Therefore, for practical adjustments in storage conditions, an air supply temperature of 3°C is recommended. This study will continue to explore conditions based on this 3°C air supply temperature.

The air supply velocity of the chiller significantly influences the environmental flow field. Different air supply velocities primarily affect the cooling rate; in a steady-state scenario, they also impact the distribution characteristics of both the flow and temperature fields. By adjusting the input frequency between 0 and 50 Hz, the chiller’s conveying speed can be modified relatively smoothly. While maintaining constant air supply humidity and temperature, the chiller’s conveying speed can be set to 3 m/s, 4 m/s, 5 m/s, and 6 m/s, respectively, to investigate the effect of air supply velocity on the flow fields within the potato storage environment.

Figure 10 illustrates the velocity distribution at section X = 0 m and Z = 2.5 m under various air supply velocities. The figure reveals a zigzag airflow pattern, with noticeable differences observed primarily above the potato pile. As the air supply velocity increases, the air velocity within the potato stack gradually rises, with a more pronounced increase observed below the potato pile. At air supply velocity of 3 m/s and 4 m/s, some air flows from the cold air blower and enters the potato pile from the ground. However, at the air supply velocity of 5 m/s and 6 m/s, this portion of the air exits away from the blower. The variation in flow rates above the potato pile can be attributed to changes in air supply velocity from the cold air blower, resulting in different pressure levels above the pile. When the air supply velocity is low, the airflow above the potato pile is slower, leading to similar pressure levels both inside and outside the pile, which allows air to pass through by inertia. Conversely, at higher air supply velocity, the airflow above the potato pile increases significantly, while the internal airflow does not increase proportionately. This discrepancy creates a larger pressure difference between the potato pile and its surroundings, resulting in air inside the pile flowing outwards, away from the blower. At an air supply velocity of 3 m/s, a small backflow occurs in the middle of the south side. When the air supply velocity increases to 4–6 m/s, a small backflow is observed in the lower part of the north side. However, at an air supply velocity of 5 m/s, no significant backflow is evident. From the perspective of energy utilization, the optimal air supply velocity is determined to be 5 m/s.

Figure 11 illustrates the temperature distribution at the Z = 2.5 m and X = 0 m sections of the experimental platform under varying wind velocity. It is evident that, at the Z = 2.5 m section, the temperature distribution characteristics across different wind velocities are generally similar, with most air domains and potato regions exhibiting low temperatures, while the area above the potato pile shows higher temperatures. However, the distribution pattern of the high-temperature region above the potato pile varies at air supply velocity of 3–4 m/s, the high-temperature region tilts towards the fan side, whereas at velocity of 5–6 m/s, it tilts away from the fan direction, corresponding to the velocity distribution at those positions. Additionally, as the air supply velocity increases, the area of the high-temperature zone in the center of the potato pile gradually decreases. In the X = 0 m section, the high-temperature areas above the potato pile lean towards the south, and the area extending into the air domain diminishes with increased air supply velocity and enhanced heat transfer rates. At air supply speeds of 3 m/s, 4 m/s, 5 m/s, and 6 m/s, the average temperatures of the potato pile were 3.23°C, 3.21°C, 3.19°C, and 3.20°C, respectively. The average temperature of the potato pile initially decreases and then increases with increasing air supply velocity, reaching its minimum at 5 m/s. This indicates that the spatial extent of high temperatures within the potato pile is minimized at this air supply velocity. Considering the velocity profiles of the sections, there is no significant distinction between the actual cooling rates at 5 m/s and 6 m/s. Consequently, an air supply velocity of 5 m/s is deemed appropriate for regulating the temperature of the potato storage environment.

Appropriate humidity distribution plays a crucial role in maintaining potato moisture levels. The humidity of the air supplied by the chiller significantly impacts the storage environment humidity for potatoes. Altering the air supply humidity directly affects the humidity distribution in the storage environment. Thus, this study will investigate the impact of varying air supply humidity levels (86, 88, 90, and 92%) while keeping air supply speed and temperature constant on the humidity field within the potato storage environment.

Figure 12 illustrates the relative humidity distribution at section Z = 2.5 m and X = 0 m under different fan air supply humidity levels, with an air supply velocity of 5 m/s and a temperature of 3°C. It can be observed that the relative humidity distribution remains consistent across different cross-sections and varying fan air supply humidity levels; however, the humidity is slightly lower above the potato pile because of potato respiration heat. Under varying operational conditions, the humidity distribution in the experimental platform can be broadly categorized into two areas: Section 1, which encompasses the air volume with high relative humidity predominantly in the potato stacking zones, and Section 2, located above the potato pile, where relative humidity is low. In this second section, humidity decreases rapidly with increasing distance from the center of the potato pile, exhibiting a distinct humidity gradient. Under conditions of 86% relative humidity, the humidity in Section 1 ranges from 82.71 to 86.35%, while in Section 2, the peripheral humidity ranges from 75.44 to 79.08%, and the central humidity ranges from 60.90 to 75.44%. Overall, the environmental humidity is low, leading to significant moisture loss in the potatoes, which is detrimental to their storage. At 88% relative humidity, Section 1’s humidity increases to between 84.52 and 88.36%. In Section 2, the peripheral humidity falls between 76.85 and 80.69%, while the central section ranges from 61.51 to 76.85%. When the air humidity reaches 90%, the humidity in Section 1 is optimal, ranging from 85.16 to 90.36%, making it suitable for potato storage. In Section 2, the peripheral humidity is between 75.66 and 80.27%, while the central humidity ranges from 63.35 to 71.31%. However, at 92% relative humidity, the humidity in Section 1 rises to between 88.62 and 92.36%, which increases the risk of potato rot. In Section 2, the peripheral humidity is between 63.22 and 81.85%, while the central humidity ranges from 62.47 to 77.42%. Considering the effects of humidification, it is crucial to note that during high humidity conditions, some condensation may occur. Therefore, maintaining an air supply humidity of 90% is most appropriate for regulating the environmental humidity.