Food Chilling Methods and CFD Analysis of a Refrigeration Cabinet as a Case Study

This chapter presents the most significant facts about food chilling. For food chilling, the chilling medium in mechanically cooled chillers may be air, water or another secondary cooling agent (slurry ice), or metal surfaces (heat exchangers). Ice chilling is also presented. CFD simulation is applied to a vertical display cabinet with four shelves. In order to evaluate the temperature gradient, the following stages are taken into account: preprocessing—geometry set-up and design of the discretization scheme; processing— introduction of the boundary conditions and calculation; and post-processing—visualization of the velocity and temperature fields.


Introduction
Refrigeration slows down the chemical and biological processes in foods, such as the accompanying deterioration and the loss of quality, extending the shelf life of the products, with minimum changes to the sensory characteristics and nutritional properties.
Temperatures in the range of 0-5 C slow down the development and growth of microorganisms, but some pathogenic agents can grow to large numbers at these temperatures or are still sufficiently virulent to cause poisoning. Because the activity of most of the pathogenic agents is only slowed down and not stopped, long-term storage of refrigerated products can finally cause food poisoning [1]. In these cases, when a longer preservation period is needed, freezing must be used to minimize any physical, biochemical, and microbiological changes affecting quality during storage. The storage life of fresh perishable foods, such as meats, fish, fruits, and vegetables, can be extended by several days by cooling, and by several weeks or months by freezing [2,3]. During freezing, most of the water content of the meat, about 80%, solidifies into pure ice crystals, accompanied by a separation of dissolved solids [2,4].
In order to chill fresh foods, it is necessary to remove the sensible heat and also the heat generated by the respiratory activity of vegetables and fruits; in animal tissues, aerobic respiration rapidly declines when the supply of oxygenated blood is stopped at slaughter [1,4].

Cooling rate and duration
Cooling rate may be defined as [4]: where dt is the temperature variation during the period dτ.
Several assumptions are taken into account in order to obtain the temperature-time function [4]: • homogenous product; • at any moment, the temperature is the same in the entire mass of the product; • the temperature of the cooling medium is constant; • there is no mass transfer between the product and the cooling medium.
The assumption that the sensible heat removed from the product equals the convective heat transfer from the product to the cooling medium leads to [4]: Eq. (2) finally leads to the temperature-time equation: : (3) The chilling duration is [4]:

Chilling equipment
Food chilling is performed with mechanical refrigeration systems or with ice [4]. The temperature of the product should be lowered as quickly as possible through the critical warm zone (5-10 C), where maximum growth of microorganisms occurs [1,3,4].
Batch or continuous operation is possible when using mechanical refrigeration systems, while batch operation is used in the ice chilling systems.
The chilling medium in mechanically cooled chillers may be air, water, or metal surfaces.

Air chilling
Air chillers use forced convection to circulate cold air at high speed (4 m/s); thus, the thickness of the boundary film is reduced, and the heat transfer rate is increased.
Air chilling of foods is performed in chilling tunnels or chilling rooms; usually, the chilling tunnels operate continuously, while the batch chilling is used for chilling rooms [4].
In a refrigeration tunnel, air circulates longitudinally, transversally, or vertically (Figures 1 and 2). Two-phase chilling can also be used in order to speed up the process and limit the evaporative weight losses [4]. Figure 3 presents the schematic of the two-phase chilling tunnel. In the first  chilling phase, the air temperature is around À10 C ,a n di t ss p e e di sa p p r o x i m a t e l y1m / s ;t h e products, hanged on the conveyer, travel through section (I) of the tunnel in about 4-5h .T h e surface of the products is chilled rapidly in this first section. In the second section of the tunnel (II), air temperature is about 0 C, and its speed is 0.3 m/s; the duration of the chilling process is 10-15 h, until the product reaches a relatively uniform temperature in its entire mass [4]. This system allows the diminishing of the evaporative weight loss compared with the one-phase chilling systems because the surface of the product is quickly cooled in the first phase, and a lower temperature difference between the product and the cooling medium is achieved in the second phase.
Chilling rooms have a lower capacity than chilling tunnels; because of the lower air speed (0.3 m/s), the duration of the chilling process increases. Figure 4 presents some examples of chilling rooms: the air discharge ducts are placed in the upper side of the room, while the air intake ports are placed in the lower part.

Chilling with liquid secondary agents
In this case, a chilled secondary agent is used in order to refrigerate the product; the product is sprayed with cold agent or is immersed into the chilling agent [1,4,5]. Due to the higher values of the convective coefficients, chilling with cold liquid agents requires less time than air chilling. The procedure is used for chilling poultry, fish, and some vegetable products [4,5].

Refrigeration
Depending on the final temperature of the product and on the type of product, chilling may be achieved with water, salt water (brine), or slurry ice; ice or vapor compression refrigeration systems are used in order to cool the secondary agent [4]. Figure 5 presents the schematics of a device for the immersion chilling of poultry [1]; the carcasses are placed on the conveyor (2) and then immersed into the cold agent. The secondary agent is chilled in the heat exchanger (9) by a vapor compression refrigeration system.
Slurry ice is a phase-changing secondary agent, containing small ice crystals (typically 0.1-1mm in diameter), suspended within a solution of water and a freezing point depressant. Some   commonly used compounds are salt (sodium chloride), ethylene glycol, propylene glycol, various alcohols (isobutyl, ethanol), and sugar (sucrose, glucose) [6]. This type of ice has many advantages in comparison with the traditional ice (flake ice, shell ice, crushed ice, etc.): it can be used in direct contact with the object to be chilled; due to the large contacting area, it has very good cooling performances; slurry ice can be pumped to the point of use (Figure 6), eliminating costly and maintenance intensive rakes, augers, and ice conveying systems [7]; operating at temperatures below the freezing point of water, ice slurry facilitates several efficiency improvements such as lowering the required temperature difference in heat exchangers due to the beneficial thermo-physical properties of ice slurry [6].
Slurry ice is produced in scrapped surface heat exchangers (SSHE); Figure 7 presents the operating principle of the SSHE for slurry ice. The ice slurry generator consists of a cylindrical metal shell (1). The exterior surface is cooled by the evaporating refrigerant passing through the cooling jacket (2), while water freezes in contact with the cold inner surface of the shell. Spring-loaded rotating blades (4) scrap off the ice crystals formed on the inner cylindrical surface of the metallic shell.

Ice chilling
This method is used for chilling fish and vegetables [1,4].

Refrigeration
Ice is often produced in the form of lumps or blocks, of various weights, from 10 to 200 kg. Alternatively, ice may be made in the form of smaller pieces (granular ice). Different types of granular ice exist (flake ice, tube ice, plate ice) [4].
In order to obtain ice blocks, the water to be frozen is filled into large metal molds, which are placed in a tank containing refrigerated brine, for up to 24 h. When the ice blocks are completely frozen, they are removed from the freezing tank and dipped into hot water. This melts the surface of the block so that it can be tipped out of its mold. The molds can then be refilled with water and returned to the freezing tanks. The large ice blocks may then be broken down into smaller pieces in an ice-crushing machine ( Figure 8).   One of the most common types of granular ice is flake ice, which is obtained by freezing water onto the surface of a rotating, refrigerated drum ( Figure 9). The water freezes into a 2-3-mmthick layer of ice, which is then scraped off the drum as flakes of ice.
Storing fish in ice is largely used aboard fishing ships; apart from chilling the fish, ice removes heat from the surrounding structure of the box or storage compartments, absorbs the heat input through the structure from the warm air and sea outside, and removes the heat produced by the spoilage process in the fish themselves [8]. It is therefore essential that plenty of ice is properly distributed throughout the catch to ensure efficient cooling. Ideally, each fish should be in contact only with ice; in practice, there are alternating layers of ice and fish ( Figure 10). The ice-to-fish ratio is comprised between 1/3 and 1/1; there should be at least 5 cm thick layers of ice (1,3) between fish and the compartment walls [8]. In order to avoid the lower layers being damaged under the weight of the upper layers, fish must be placed on shelves, in order to keep the depth of the storage compartment (2) lower than 0.5 m [8].
Ice chilling is also used for vegetables and fruits [4]; the products are placed in wooden or cardboard crates, filled with ice. Figure 11 presents the schematics of an ice-filling machine; the ice flakes flow into the machine hopper (2) through the chute (1) and are poured into the crates through the hopper (4).

Chilling in tanks and heat exchangers
This chilling method is used for liquid food products (milk, cream, juices, beer, wine, etc.). Refrigerants or secondary cooling agents are used on the cold side of the chilling device, which is operated in batch or continuous mode [4].
In the batch operation mode, the product fed into the tank is cooled using an external cooling jacket or an internal cooling coil. Figure 12 presents the schematics of a tank equipped with an external cooling jacket (2). The product is fed into the tank through the pipe (7); when the desired final temperature is reached, the product is purged from the tank through the discharge pipe (5).

Refrigeration
Heat exchangers are used in order to continuously chill the liquid products [3,4]. Figure 13 presents a double pipe heat exchanger; the cooling medium circulates through the shell (3), while the product circulates through the inner pipe (4). The exchanger contains several sections that are coupled through flanges; the number of sections depends on the final temperature and flow rate of the fluids [9].

Retail refrigeration cabinets
Retail refrigeration cabinets use cold air circulated through natural or forced convection. The cost of chill storage is high and, in order to reduce costs, large stores may have a centralized plant to circulate the refrigerant to all cabinets; the heat generated by the condensers of the refrigeration system may be used for in-store heating. Computer control of multiple cabinets detects excessive rises in temperature and warns of any requirement for emergency repairs or planned maintenance. Other energy-saving devices include night blinds or glass doors on the front of cabinets to trap cold air.   According to the cabinet geometry, the retail refrigeration cabinets are [10]: • horizontal, single-deck units; • vertical, multi-deck units.

Refrigeration 54
Horizontal refrigeration cabinets (Figure 14) are open-top and are designed for self-service; the wall-site units (Figure 14a) allow shopping from one side, while the island type units (Figure 14b) are accessible from all sides [2,10].  Axial fans (6) are used to provide air flow over the evaporator coils (1), and grills placed at the upper side of the cabinet deflect the air current from one side of the cabinet to the other, over the stacked products. Due to cold air stratification, air infiltrations from the environment are relatively low; the heating load is due to the radiant heat transfer and conductive heat transfer through the insulated walls of the cabinet case [10].
Vertical multi-deck units have an open front or are provided with a glass front door [2,10]. The open front refrigeration cabinets ( Figure 15) use air curtains (a, b, c) in order to prevent the infiltrations of warm air into the cabinet. The axial fans (1) and (3) circulate the cold air over the evaporator coils (2,4) and produce the stream of air necessary for obtaining the inner air curtains; the outer air curtain (c) is generated by the axial fan (7). Grills are used in order to direct the air flow and obtain the air curtains.
The three air curtains form a very complex system, and the proper design imposes an in-depth understanding of the cabinet's fluid dynamics [2].
Some stores are equipped with combination cabinets (Figure 16), comprising an open front vertical cabinet and an open-top, horizontal cabinet [10].
The closed vertical refrigeration cabinets ( Figure 17) use a glass door (4) in order to contain the products into the refrigerated space and prevent the infiltration of warm air [10].

CFD modeling and simulation in a vertical refrigeration cabinet
Vertical refrigerating cabinets are widely used in supermarkets because the products have good storage conditions and are adequately presented to the consumers.
As mentioned above, the open type refrigerating cabinets use one or more air curtains in order to separate the cold interior from the warm surroundings.
The study of air curtains is necessary because these can be easily disrupted by air circulation in front of the cabinet or by the consumers taking foods from the shelves; the disturbing effect increases with the height of the cabinet. The aerodynamic non-homogeneity of the air curtain  increases when the number of "holes" increases, leading to the increase of the temperature inside the display case and of its power consumption; as a result, the products inside the cabinet are stored at temperatures higher that the recommended ones. Moreover, the infiltration of exterior humid air inside the case results in additional ice formation over the evaporator coils, thus reducing the heat transfer and the efficiency of the refrigeration system; therefore, the defrosting cycles become more frequent, further increasing the power consumption.
Some estimations show that 72-75% of the cooling load is used in order to counteract the effect of warm air infiltrations through the air curtain ( Figure 18) and can reach even 90% if the operating conditions are not suitable [11]; 50% of the power consumption of a supermarket is due to the refrigeration and freezing cabinets.
CFD simulation was applied to a vertical display cabinet with four shelves, and in order to evaluate the temperature gradient, the following stages were taken into account: preprocessing -geometry set-up and design of the discretization scheme; processing-introduction of the boundary conditions and calculation; post-processing-visualization of the velocity and temperature fields.

CFD simulation and experimental tests
A refrigeration vertical cabinet with four shelves (Figure 19) was the basis of the simulation; the dimensions of the cabinet (L Â W Â H) are 1900 Â 796 Â 1911 mm.
The axial fans (2) induce the airflow over the evaporator coil (3), placed at the bottom front part of the cabinet. A limited amount of air is fed into the unit, passing through the perforated plate behind the shelves (4), while the most significant amount of air flows through the horizontal     Table 2 summarizes the transducers used in the experiments; the placement of the temperature, humidity, and velocity sensors in front of each shelf is shown in Figure 22a; temperature, humidity, and velocity sensors were also placed at the exit of the horizontal grill (GPA, Figure 22b); all the transducers were connected to a central data acquisition unit (Figure 22c). The sensors were placed in three vertical planes along the cabinet length: left, middle, and right (1, 2, 3, Figure 23).
Figures 24 and 25 present the experimental results referring to the air velocity and temperature at the exit from the horizontal grill (GPA). The experimental data were then filtered using a C+ + program; Figures 26 and 27 show the filtered results.

The numeric model
The general equation for the incompressible flow of a fluid for a dependent variable Φ (Φ =1for the continuity equation; Φ = v for the momentum equation; Φ = t for the energy equation) is [12]: The model is based on the following assumptions: • air is considered a perfect gas; • the latent heat of condensation for the humidity in the air curtain is neglected; • the air curtain is considered as a jet; • air velocity and temperature at the horizontal grill are modeled as polynomial functions; • the effect of air flowing through the holes of the perforated plate behind the shelves is neglected.
The model is based on two equations (one for the turbulent kinetic energy and the other one for the dissipation rate of the turbulent kinetic energy), using the SIMPLE algorithm. The calculation is an iterative process, using the pressure-velocity coupling algorithm, in which the momentum and continuity equations, based on pressure, is solved simultaneously, and the terms referring to the pressure gradient and mass flow rate are discretized implicitly.

Boundary conditions
The boundary conditions were imposed by the pre-processing program Gambit and completed with functions and values in the Fluent CFD simulation program [12]. The boundary conditions also took into account the conditions inside the climatic chamber (t env =25 C, Φ env = 60%, v air = 0.2 m/s, according to EN-ISO 23953); as there are different boundary conditions for the horizontal grill (GPA) and base grill (GRA), these are presented separately.

Boundary conditions for GPA
The velocity and temperature functions, for the three vertical planes considered (left, middle, and right), were the ones presented above. The air turbulence was specified using the turbulence intensity [12]: Refrigeration I t ¼ 0:16 Á Re Dh À0:125 Á 100: (12) Turbulence intensity at GPA exit had the following values: I t = 2.66 (left plane); I t = 1.83 (middle plane); I t = 2.16 (right plane).
A boundary condition was also imposed for the lateral plane (0, Figure 23); here, air velocity was constant over the entire section (0.2 m/s), the flow was laminar, and air temperature was 25 C.

Boundary conditions for GRA
The boundary conditions for the base grill take into account the average experimental values: over the entire length of the grill, air velocity is 1.7 m/s, and air temperature is +9.2 C. The air turbulence parameters were defined as above.

Boundary conditions for heat flow rate
The boundary conditions for the heat flow rate take into account the conductive heat transfer through the cabinet walls and the heat generated by the illumination system of the display case.
The conductive heat transfer was defined using Fourier's law; the overall heat transfer coefficient was calculated based on the thermal conductivity of each individual layer of the respective wall. The conductive heat transfer rates were as follows: 6 W/m 2 for the ceiling of the cabinet; 7 W/m 2 for the bottom of the cabinet; and 7.63 W/m 2 for the side walls.
Fluorescent lamps, type OSRAM L58W20, were used for the illumination of the appliance; the corresponding heat flow rate was 10 W/m 2 .

Boundary conditions for the walls
The cabinet walls were modeled only in the wall-air contact areas. The average temperatures were obtained experimentally, as follows: evaporator fins-0.95 C; interior walls +7 C; shelves +5 C.

Product thermal load
According to the EN-ISO 23953 standard, the load is simulated using a gel-type substance (tylose), with thermal properties similar to the ones of beef meat. However, in the present simulation, there was no product load inside the cabinet.

Processing
In the processing stage, the time step was set at 0.2 s, the number of steps was 1800, and the maximum number of iterations for one step was 10.
The simulation was performed on a Pentium IV, DualCore 6400, 2.4 GHz system with 4 GB RAM; for 6 min of simulation and a total number of 18,000 iterations, the computing time was approximately 12 h.
In order to prevent the equation coefficients from changing too quickly, the change in dependent variables from one iteration to another was slowed by "relaxing" them. Thus, the linear relaxation method was used for the control of the CFD simulation in order to maintain its stability; Table 3 presents the values of the relaxation factors for the different physical parameters.
In this CFD simulation, the convergence criterion was selected so that the residues remained lower than 10 À6 .

Post-processing
Post-processing is the final stage of the CFD simulation, aiming to display the temperature and velocity fields, as well as the streamlines in the simulation domain. This stage is useful in the intermediate phases of the simulation, allowing its calibration based on the experimental data; at the end of the simulation, the final temperature and velocity fields and the streamlines are presented in a graphical form.
Figures 28-30 present the temperature fields in the three vertical planes taken into account (see Figure 23); the results show that, for the lower shelves, the temperature is with at least 4 C higher than the one required (5 C, according to ISO 23954-2:2005), which means that this area is not adequate for the storage of refrigerated goods.   Table 3. Linear relaxation factors.
Refrigeration between 0.4 and 0.6 m/s; these values were significantly lower than the average value of 1.7 m/s, which was considered as a boundary condition for GRA, showing the non-uniformity of the air curtaininthisarea.  The streamlines presented in Figure 33 complete the picture of the air curtain in terms of air velocity, while Figures 34 and 35 display the streamlines in terms of air temperature; the results can be summarized as follows:

Refrigeration 68
• the turbulence and temperature are higher at the back of the cabinet; • the infiltration rate through the air curtain is higher in the lower part of the cabinet, through the left and right planes; • the temperature difference between the upper and lower shelves is 6 C.