- Email: [email protected]
Innovative Parallel Airflow System for forced-air cooling of strawberries
Accepted Manuscript Title: Innovative parallel airflow system for forced-air cooling of strawberries Author: Habibeh Nalbandi Sadegh Seiiedlou Hamid R. Ghasemzadeh Faramarz Rangbar PII: DOI: Reference:
S0960-3085(16)30102-X http://dx.doi.org/doi:10.1016/j.fbp.2016.09.002 FBP 772
To appear in:
Food and Bioproducts Processing
Received date: Revised date: Accepted date:
11-10-2015 29-8-2016 16-9-2016
Please cite this article as: Nalbandi, Habibeh, Seiiedlou, Sadegh, Ghasemzadeh, Hamid R., Rangbar, Faramarz, Innovative parallel airflow system for forced-air cooling of strawberries.Food and Bioproducts Processing http://dx.doi.org/10.1016/j.fbp.2016.09.002 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Innovative parallel airflow system for forced-air cooling of strawberries
Habibeh Nalbandi1*, Sadegh Seiiedlou1, Hamid R. Ghasemzadeh1, Faramarz Rangbar2 1
Department of Biosystem Engineering, Faculty of Agriculture, University of Tabriz, Tabriz, Iran
2
Department of Mechanical Engineering, Faculty of Mechanical Engineering, University of Tabriz, Tabriz, Iran *
Corresponding Author Email: [email protected]; [email protected] Tell: +9804133392780
1
Highlights
An innovative airflow patterns was introduced for precooling of strawberries.
Cooling process of fruit was simulated inside the new designed package and tray.
The new system was able to distribute the cold air uniformly throughout the packages.
Some improvements were made in homogeneous of strawberry precooling process.
2
Abstract Strawberry is a high value crop and mostly is consumed as a fresh fruit over the world. Precooling of strawberry is one of the postharvest processes that reduces the fruit decay. However, in the traditional cooling system of strawberry considerable differences are observed between the fruit temperature located in the individual packages as well as in the various packages inside the tray. As a result, the self-life of fruit decreases. The aim of this research was introducing the new package and cooling system for precooling of strawberry in which some improvements were made in the uniformity of cooling of strawberry and reduced the fruit decay. The cooling process of strawberry was simulated based on simultaneous airflow and heat transfer process inside the new system and the developed model was validated experimentally, and a good agreement observed between the simulated and measured temperature of fruit. The results of simulator and experimental data showed that the new system had the higher uniformity in the cooling of fruit as compare with the traditional system so that the packages located along the cold air direction received the equal airflow rate. As a result, a difference of 0.84 ºC was measured in the average fruit temperature of the packages after 3 hours of cooling meaning that the differences in the 7/8
th
cooling time of fruit between the packages is negligible. Therefore,
with the same airflow rate of traditional system, the cooling time of strawberries could be reducing by the proposed system significantly. Keywords: Strawberry, Simulation, Innovative parallel airflow system, Precooling, Package design.
3
Nomenclature CFD Cp k MPAS PAS P t T u µ ρ Subscripts p a
Computational Fluid Dynamics Specific heat capacity (J/(kg.ºC)) Thermal conductivity (W/(m.ºC)) Modified Parallel Airflow System Outward vector normal to the surface Parallel Airflow System Pressure (Pa) Time (s) Temperature (ºC) Velocity (m/s) Dynamic viscosity (Pa.s) Density (kg/m3) Product Air
1. Introduction Strawberry is a high value crop and mostly is consumed as fresh fruit, i.e. over 75% of the product is delivered by fresh produce market. However, it is one of the most perishable fruit and is susceptible to mechanical damage, microbial decay and water loss. These prevent the product, in some cases, reaching the consumer at its optimal quality after transport and distribution (Anderson, et al., 2004; Manganaris, et al., 2007). Good temperature management is a key factor for delaying product deterioration, maintaining product quality and extending its shelf life. The cooling of fruit and vegetables from ambient temperature down to a proper temperature controls the microbial activity, retards the ripening process, and reduces respiration, wilting and shriveling due to moisture loss. Two hours delay in cooling of strawberry is long enough to reduce the proportion of marketable fruit dramatically (Kader, 2002). Among various precooling techniques, forced-air cooling is widely accepted as an appropriate method for cooling of most fruit such as strawberry (Becker, et al., 1996; Brosnan and Sun, 4
2001; Chakraverty and Paul, 2001; Anderson, et al., 2004; Castro, et al., 2005; Tutara, et al., 2009). Poor airflow distribution at different locations in the package leads to considerable heterogeneities in the final temperature of product (Alvarez and Flick, 1999a; Amara, et al., 2004; Dehghannya, et al., 2011). Many researchers have used some experimental methods to study the factors that affect the airflow pattern and cooling efficiency (Alvarez and Flick, 1999a, and 1999b; Amara, et al., 2004; Anderson, et al., 2004; Castro, et al., 2004; Vigneault, et al., 2006; Kumar, et al., 2008). However, there are some problems associated with experimental studies such as being expensive, time consuming and situation specific (Zou, et al., 2006a). To overcome these problems, mathematical approaches have been developed to model the transport phenomena inside the packages, for predicting the temperature and airflow patterns and cooling homogeneity. The medium inside the packages were considered as a porous medium in most of these models (Zou, et al., 2006a, 2006b; Van der Sman, 2002). This assumption was made due to limitations in computational resources. Even though the computational method is cost-effective, inevitable simplifications made in porous medium approach, lead to considerable errors in predictions (Dehghannya, et al., 2008; Ferrua and Singh, 2009b). Recent advances in the computational resources and decreasing cost of modern computers have made the application of CFD modeling more efficient and popular and provided more powerful tool for solving the fundamental continuity, momentum and heat transfer equations and predicting the airflow patterns and temperature variation in the packages. The mathematical modeling capability of predicting the cooling process of fruit within the packages has been used as an ideal approach by many researchers, especially when the container-to-product equivalent
5
diameter ratio was under 10, e.g. strawberry packages. In such situation, the porous medium approach could not be used (Dehghannya, et al., 2010; Ferrua and Singh, 2011). Some researchers studied the cooling process of strawberry using various package and tray designs as well as pallet arrangements (Anderson, et al., 2004; Ferrua and Singh, 2007, 2009a, 2009b, 2009c, 2009d, 2011). In the forced-air cooling system that is shown in the Fig. 1-a, the cold air is forced across each palletized row (Fig. 1-b and 1-c), through the vents in the trays and packages (I in Fig 1-d) due to the negative relative pressure created by the system suction fan. Therefore, the cold air enters from the right side of each row and the warmer air exits from the left side (Fig. 1-d). There are some problems associated with this method. Firstly, the air temperature entering to the various packages is not the same as it travels along the packages. The warmer exiting air from package 1 (P1) enters the next one (Fig. 1-d), as a result the air temperature increases gradually as an in-series method. As a result, there was about 6ºC difference in average fruit temperature between P1 and P8 after 1h of cooling process (airflow rate: 1L/(s.kgp)). In the in-series method, modifying the package and/or tray would not improve the rate and uniformity of the cooling process; because the air temperature still rises as it travels along the systems (Ferrua and Singh, 2009d). Therefore, the last package (P8) always receives higher temperature air than the first one (P1). The higher air temperature leads to a lower heat removal rate. Secondly, the most of total cold airflow bypasses the system without removal of heat from the fruit. This significant bypass could be explained by the differences in the shape of packages and trays. Ferrua and Singh (2011) developed a novel forced-air cooling system that split the airflow across each layer of the pallet into two streams of parallel airflow (Fig. 2-a, path 1 and 2). They increased the vent area of tray 2 (left side in Fig. 2-a) 60% larger than the vent of tray 1(right
6
side of Fig. 2-a). In addition, the vent area of the packages that were located inside tray 2 was increased up to 40%. This new system improved relatively the uniformity of cooling of products between two trays and reduced the total cooling time (by 6%), pressure drop and energy consumption of process. Apart from what Ferrua and Singh (2011) achieved, there is still a considerable heterogeneity in the cooling process of strawberries located in various trays as well as packages in each tray. Because, the in-series method of cooling still occurs in each tray, there is heterogeneity in temperature of fruit in the package 1 and 4 and also package 5 and 8 (P1 and P4; P5 and P8 in Fig 2-b). As a result, about 60 minutes difference was measured in 7/8th cooling time of packages 1 and 4 (Ferrua and Singh, 2011). The aims of this work were: (a) Introducing an innovative packaging system for the forced-air cooling of strawberries to increase the effectiveness of process as well as uniformity of cooling process; (b) Numerical modeling and simulating of the precooling process of strawberries inside the new packages and proposed system; (c) Experimental study and validating the developed model and simulator; (d) Improving the various parameter of cooling process such as the uniformity of fruit temperature inside the individual packages, uniformity of the cooling process between the various packages in the system, reducing the cooling time and percentage of bypassed air using the simulator.
2. Materials and methods Any design for package and tray of fresh fruit should improve the uniformity of product cooling. In this work, an innovative Parallel Airflow System (PAS) was introduced to improve the homogeneity of strawberries pre-cooling process. In the proposed system, two separate ducts were designed for introducing the cold air into the individual packages in a way that the exiting
7
heated air can return to air cooling unit without re-entering other packages. One of these ducts (duct 1) was built on the top of the packages (Fig. 3-a, II) and the other one (duct 2) was located between the packages (Fig. 3-b, I). The cold air enters to the duct 1 before entering the packages where it removes heat from the fruit and exits through duct 2 (a detailed description will be presented in the following sections). Package design Due to the strong effect of the package design and airflow pattern on the performance of the Parallel Airflow System, different package designs were considered and their effects on the performance as well as cooling homogeneity have already been investigated by the present authors (Nalbandi, 2015). The best one was used in this work Fig.4. In order to create a duct between the packages (duct 2; Fig. 3-b, I), a trapezoidal cross section was considered for the strawberry package (Fig. 3-b and 3-e). The location and shape of the vents on package walls play an important role on the Parallel Airflow System because of relating the inlet and outlet ducts. The vents can be distributed on the top (Fig. 3-f, I) and lateral walls (Fig. 3-f, II) as inlet and outlet vents. In this design, the opening area was 28% of the total surface of the top and lateral walls of the package. 2.1. Tray design A new design of tray was needed providing a duct on the top of the packages (duct 1; Fig. 3-a). Therefore, the tray walls were assumed taller than the package. Fig. 3-d shows the tray built to carry the packages. Each tray contains eight packages with 2 by 4 arrangements (Fig. 3-c). There was no gap between the packages. The Parallel Airflow System requirements make it necessary to have a different vent design in the front and back walls of tray associated with ducts 1 and 2; side A and B in Fig. 3-d. A
8
detailed description will be presented in the next section. The vent in the side A is related to the duct 1; top of the packages (Fig. 3-a), and the trapezoidal vents in the side B is referred to the duct 2; between the packages (Fig. 3-b). The vent area of the tray was 28% and 13% at side A and B, respectively. 2.2. The airflow paths As previously mentioned, ducts 1 and 2 are used as inlet and outlet of airstreams, respectively. One end of the ducts is closed alternately; the side near the fan in duct 1 (Figs. 3-a, II) and the other side in duct 2 (Fig. 3-b, I). When the circulating suction fan is turned on, the cold airstream enters duct 1, then passes through top openings of packages (I in Fig. 3-f), and exits finally through openings on the walls of the packages (II in Fig. 3-f). The heated airstream returns to cold room for recirculation through duct 2 (Fig. 4). It is obvious that any individual package always receives cooled air directly from cold room and the air exiting the package never enters the next one. This means that the products in the all packages always receive the air with the same temperature and the cooling of the fruit is performed in a parallel manner. 2.3. Simulation of the cooling process of strawberries To evaluate the performance of the Parallel Airflow System, the cooling process of strawberries was simulated based on mathematical modeling of airflow and heat transfer between airstream and the product. To evaluate the efficiency of the Parallel Airflow System through 3D simulation, the tray was partitioned using three middle walls (Fig. 5-a) (the friction on these walls was not considered) and only quarter of tray (Fig. 5-b) containing two packages was considered as the computational domain. 9
2.4.1. Computational domain Strawberry model: A 3D strawberry model equivalent to medium fruit size was created. The area, volume and equivalent diameter of the model are 0.002504 m2, 1.12× 10-5 m3 and 0.028 m, respectively. Package model: The package model with a proper vent distribution was created based on details discussed by Nalbandi (2015). Each package was filled with 39 strawberries. The total weight of fruit was 0.46 kg. In order to avoid the need for finer mesh and undesirable skewness of the mesh elements, individual strawberries were located far enough from each other as well as from the wall so that small gaps were created between them. One-half of the packages was used in the simulation process. Tray model: A quarter of a tray was modeled geometrically. Final computational domain: There is a plane of symmetry in the computational domain. Therefore, a half of this domain was used for simulating process as a final computational domain (Fig. 6). 2.4.2. Mesh generation To establish an optimal mesh, a mesh independence study was performed by changing the mesh density to achieve a mesh that gives a solution with acceptable accuracy (Ho et al., 2010). The average fruit temperature was selected for this purpose and its variation studied at various mesh densities. Four meshes, each containing 96470, 106449, 150890 and 240532 elements were used. The results indicated that mesh independence of numerical solution was obtained when the number of elements was above 150000. Therefore, the mesh with 150890 elements was adequate for accurate numerical prediction. However, to prevent large temperature and velocity gradients, a finer mesh used in the critical areas such as inlet and outlet vents, between the fruit as well as
10
the fruit-air interfaces. The computational domain mesh containing 241706 tetrahedral elements (Lagrange Quadratic) resulted in 1265024 degree of freedom. 2.4.3. Mathematical models For precooling of strawberries, and majority of other horticultural crops, Biot number lies between 0.1 and 10. Within this range, both the internal and external resistances are on the same order and the temperature gradients will appear within the products (Anderson et al., 2004). In simulation of the system, airflow and heat transfer were considered functions of time. 2.4.3.1. Airflow model In the precooling of strawberry the average airflow rate is 1L/(s.kgp) that creates the laminar flow in the packages (Ferrua and Singh, 2009a). In the present work, the airflow rate was considered to be 0.4 L/(s.kgp) and the Reynolds number was calculated based on the hydraulic diameter of the inlet vent. It was about 710. In addition, the maximum Reynolds number was 54 inside the packages. It was calculated based on the hydraulic diameter of free space in the crosssection of the air domain at the various planes of package. Therefore, incompressible fluid and laminar flow were assumed inside and between packages. In addition, the airflow was constant and steady state condition governs during process. However, in order to use the developed model for other conditions in future studies, the mathematical model of airflow was analyzed as a time dependent process. Eqs. 1 and 2 give the mass and momentum conservations for air in the system, respectively. The buoyancy force term is not included in the airflow equation and the air properties were assumed constant. The density and the viscosity were assumed to equal to those of dry air at 0°C. . u 0
ρa
(1)
u T ρ a u .u P . μ a u u t
11
(2)
The airflow model was completed by considering the following boundary conditions. Inlet boundary: laminar inflow was applied to define the inlet boundary condition for airflow model. It was assumed that the air pressure is equal to the atmospheric pressure at distance of 0.5 m from the inlet. Outlet boundary: the constant velocity at the outlet vents of tray was defined. Its direction was perpendicular to the vent. Wall boundaries: a no slip boundary condition was applied to all solid surfaces including inner surfaces of trays, all surfaces of package and fruit surfaces. Symmetry boundary: symmetry boundary condition was used on the plane of symmetry. The boundary conditions on the velocity can be written as: Inlet: p p0
(3)
Outlet: u u 0
(4)
Wall: u 0
(5)
Interface: u 0
(6)
2.4.3.2. Heat transfer model in the product domain Eq. 7, the transient equation that describes the heat transfer process in the strawberries domain. The respiration process was assumed negligible because of low transferred heat. Other assumptions were also made in the model: the thermo-physical properties of strawberry were assumed to remain constant and there would be no moisture loss nor shrinkage during the cooling process (kp=0.57 W/(m.ºC); Cp,p= 3.95 kJ/(kg.ºC); ρp=800 kg/m3). In addition, the initial temperature of strawberry was set to16 ºC. ρ p c p,p
Tp t
. k pTp
(7)
12
For heat flux at the products surface, the following boundary condition was adopted:
k PTP k a Ta . n 0
(8)
2.4.3.3. Heat transfer model in the fluid domain The heat transfer equation within the fluid domain was written as Eq. 9. This equation describes a time dependent process that includes the conduction and convection terms. The air properties were assumed those of dry air at 0°C. In addition, the air temperature was assumed to be 0 ºC. ρ a c p,a
Ta ρ a c p,a u .Ta . k a Ta t
(9)
The appropriate boundary conditions for Eq. 9 are as follows: Inlet: temperature, Ta Ta0
(10)
Outlet: convective flux, k a Ta . n 0
(11)
Wall: thermal insulation, k a Ta . n 0
(12)
Interface: temperature, Ta TP
(13)
Symmetry plane: thermal insulation, k a Ta . n 0
(14)
2.4.4. Numerical solution COMSOL MULTIPHYSICS software, version 3.5, was used to simulate the heat transfer process and airflow using the finite element method. There is a potential numerical instability in the finite element computation of incompressible flows. It refers to the Galerkin formulation. To prevent this numerical instability, isotropic diffusion with a tuning parameter of 0.5 was selected among the various stabilized finite element formulations. The simulation was performed using a personal computer with 32 GB RAM (Processor: Intel (R) Core (TM) i7-2700K CPU@ 3.50
13
GHz 3.50 GHz). A variable time step procedure was used through the simulation. The solution time was about 552 hr. 2.4. Experimental validation of the model To validate the numerical model a forced-air cooling system was used in a cold room where the temperature and relative humidity were 0.1 ºC and 80%, respectively. This system consists of a suction fan to generate a desirable airflow rate and an open tunnel with 50 cm2 cross section and 120 cm length. For each test, the fresh fruit were purchased from the local greenhouse and stored in the refrigeration room at 1 ºC. Fruit were placed at a room with temperature of 16 ºC to equilibrate their temperature before experiments. Strawberries with a uniform shape were packed in the packages and placed inside the tray (Fig. 7). The tray in the Parallel Airflow System configuration or in the Modified Parallel Airflow System configuration (see section 2.6) was located in the center of tunnel and insulated with foam. Thermocouples (K type) were inserted into the center of 8 strawberries and temperature of these points were used for model validation. The temperature was recorded every 1 minute. The exact temperature of airflow entering the system was recorded by a thermocouple located in the front of tunnel. The average temperature of air was 0.3 ºC. The experiments were conducted three times. The moisture loss of fruit was determined by the difference in the package weight before and after cooling. 2.5. Improving the first design of Parallel Airflow System using air restriction plate At some points, it was realized that restricting the duct 1 in the midpoint might improve uniformity of cooling in particular along the package 1. Therefore, the original design Parallel Airflow System, was modified by inserting an Airflow Restriction Plate (ARP) i.e., a plate with hole (Fig. 8). For this Modified Parallel Airflow System (MPAS), a revised computational
14
domain and mesh parameters were defined and final domain containing 242869 mesh elements; Lagrange Quadratic- tetrahedral. 3. Results and discussion 3.1. Original system (Parallel Airflow System) 3.1.1. Simulated Airflow pattern and temperature Air velocity filed is presented in Fig. 9. It is obvious that the air velocity decreased from the right to left of tray. However, the airflow rate entering through the inlet vents increased along the duct 1. Therefore, the vents that were located at downstream (left side) received the largest fraction of total airflow. The airflow rate entering through each inlet vents (I in Fig. 3-f) was calculated using the software and the inlet airflow percentage was calculated for the both packages. As much as 80% of total airflow introduced into the system entered the package 2 through the end vents. Fig. 10 shows the distribution of airflow rates leaving the packages through the outlet vents. It can be seen that a larger proportion of total airflow exited through the vents close to the end side of the packages (the number of vents were shown in Fig. 6). Fig. 11 presents the slice plane plot of air temperature distribution in vertical planes. The vertical plane shows the air temperature distribution in the inlet vents of packages. As expected, the temperature of the air entering into packages is the same at all opening points. 3.1.2. Simulated average fruit temperature in the packages The average fruit temperature in the two packages versus cooling time is shown in Fig. 12, where a difference of 2.8 ºC was seen between the average fruit temperature in the packages after 3 h cooling. In addition, the half-cooling time of fruit in the package 1 and 2 was about 107 and 45 min, respectively. It is clear that the package 2 was cooled faster. This result can be justified by the airflow rate entering the packages. Although both packages received the air with the same
15
temperature, there was a significant difference between the entering airflow rate (package 1 and 2 receive 20% and 80% of total airflow rate, respectively). A lower airflow rate passing through the package leads to larger convective heat transfer resistance within the package. The standard deviation (SD) of the half-cooling time of all fruit as a criterion for uniformity of cooling process are presented in Table 1. There is a considerable difference between the halfcooling time and SD of both packages showing the high heterogeneity of cooling. The effect of airflow distribution on the uniformity of fruit surface temperature is illustrated in Fig. 13. Unexpectedly, the fruit located at the top row was not necessarily cooled faster. The fruit located closer at the rear and bottom of the package was cooled faster. Generally, the fruit located at the second half of the package was cooled faster. One possible justification is poor airflow distribution within the package. About 86% of total air entering the package exit through the outlet openings located at the second half of package. Therefore, there was a large amount of airflow in the second half of the package that led to faster cooling of fruit located in this zone. 3.2. Experimental results of validation Fig. 14 demonstrates the simulated and experimental average temperature of instrumented fruit in the Parallel Airflow System. As this figure indicates, there is good agreement between the simulated and experimental data. The observed temperature differences could be due to the errors caused by inaccurate positioning thermocouples and fruit properties as well as modeling assumptions particularly no moisture loss. The measured moisture loss of fruit was not negligible and it is the major reason of the error, because of the latent heat of vaporization that was taken from the fruit. However, the developed model could be used in future studies of package designing. 3.3. Modified Parallel Airflow System
16
To improve cooling efficiency, the Parallel Airflow System configuration was modified based on the experimental and simulation results. According to fluid mechanics, reducing the amount of air entering the package 2, increases the amount of air entering the package1. The results of simulation are presented in the following paragraphs. 3.3.1. Simulated airflow pattern The results showed that using restriction plate improved air distribution quality between two packages so that the packages 1 and 2 received almost 46 and 54 percent of total airflow, respectively (Fig. 15). Within individual packages, air distribution quality between openings was also improved. This was more remarkable in package 1. The inlet opening located behind the ARP received a lower airflow rate. About 25 % of the total airflow rate entered the package 2 through the inlet vents of its first half. This was almost 43% in the case of the package 1. As observed in the Parallel Airflow System, a larger proportion of the total airflow entering each package exited through outlet vents of the package, which were located in the second half of package. As expected from the airflow pattern, a lower amount of air entering the first half of the packages led to higher convective heat transfer resistance in these zones. Therefore, the air temperature was higher than in other spaces, especially in the case of package 2. 3.3.2. Average fruit temperature in the packages As previously stated, Fig. 12 shows the profile of average fruit temperature in the packages versus cooling time. As shown, the cooling process of fruit inside the both packages was performed more uniformly so that 0.84 ºC difference was observed in the average fruit temperature of the packages 1 and 2 after 3h cooling. This can be attributed to uniformity of the amount and temperature of the air entering the packages.
17
As shown in Table 1, the half-cooling differences between the packages 1 and 2 are lower than for Parallel Airflow System. However, the uniformity of cooling process within the package 2 was not improved in Modified Parallel Airflow System (Fig. 16). Such non-uniformity can be attributed to the poor airflow distribution through the inlet openings. As discussed earlier, the first half of the both packages receives low airflow rate, especially in the package 2. Therefore, more design improvements must be considered.
4. Conclusions In the present work, an innovative parallel airflow system for forced-air cooling of strawberries (Modified Parallel Airflow System) was proposed. The numerical analysis of airflow and heat transfer inside the packages and tray showed that the Modified Parallel Airflow System could improve the uniformity of the cooling process in strawberries. The Modified Parallel Airflow System was able to deliver the equal amount of cold air to the packages at the same temperature, so that the packages 1 and 2 received almost 46 and 54 percent of total air. The cooling process of fruit inside the both packages was performed more uniformly and 0.84 ºC difference was observed in the average fruit temperature of the packages 1 and 2 after 3h of cooling process and a good fit was observed between the simulated and experimental data. Based on the results, this system could be used in industrial forced air-cooling of strawberries. However, the observed heterogeneity inside the individual packages indicated that more design improvements are still needed for package and tray.
18
References Alvarez, G., Flick, D., 1999a. Analysis of heterogeneous cooling of agricultural products inside bins. Part I: aerodynamic study. J. Food Eng. 39, 227–237. Alvarez, G., Flick, D., 1999b. Analysis of heterogeneous cooling of agricultural products inside bins. Part II: thermal study. J. Food Eng. 39, 239–245. Amara, S.B., Laguerre, O., Flick, D., 2004. Experimental study of convective heat transfer during cooling with low air velocity in a stack of objects. Int. J. Thermal Sci. 43, 1213–1221. Anderson, B.A., Sarkar, A., Thompson, J.F., Singh, R.P., 2004. Commercial scale forced air cooling of strawberries. Trans. ASAE, 47 (1), 183-190. Brosnan, T., Sun, D., 2001. Precooling techniques and applications for horticultural products: a review. Int. J. Refrig. 24, 154-170. Becker, B.R., Misra, A., Fricke, B.A., 1996. Bulk refregeration of fruits and vegetables part I: theoretical consierations of heat and mass transfer. HVAC&R Research, 2, 122–134 Chakraverty, A., Paul, S.R., 2001. Postharvest Technology: Cereals, Pulses and Vegetables. Sci. Publ., India. Castro, L.R., Vigneault, C., Cortez, L.A.B., 2005. Cooling performance of horticultural produce in containers with peripheral openings. Postharvest Biol. Technol. 38, 254–261. Castro, L.R., Vigneault, C., Cortez, L.A.B., 2004. Container opening design for horticultural produce cooling efficiency. J. Food, Agri. Envir. 2, 135–140 Dehghannya, J., Ngadi, M., Vigneault, C., 2008. Simultaneous aerodynamic and thermal analysis during cooling of stacked spheres inside ventilated packages. Chemical Eng. and Technology. 31(11), 1651-1659. Dehghannya, J., Ngadi, M., Vigneault, C., 2010. Mathematical Modeling Procedures for Airflow, Heat and Mass Transfer During Forced Convection Cooling of Produce: A Review. Food Eng. Reviews. 2, 227-243. Dehghannya, J., Ngadi, M., Vigneault, C., 2011. Mathematical modeling of airflow and heat transfer during forced convection cooling of produce considering various package vent areas. Food Control. 22, 1393-1399.
Ferrua, M.J., 2007. Modeling the forced-air cooling process of fresh horticultural products. Ph.D. Thesis, University of California, Davis, USA. Ferrua, M.J., Singh, R.P., 2009a. Modeling the forced-air cooling process of fresh strawberry packages. Part I: numerical model. Int. J. Refrig. 32 (2), 335- 348. 19
Ferrua, M.J., Singh, R.P., 2009b. Modeling the forced-air cooling process of fresh strawberry packages. Part II: experimental validation of the flow model. Int. J. Refrig. 32 (2), 349- 358. Ferrua, M.J., Singh, R.P., 2009c. Modeling the forced-air cooling process of fresh strawberry packages. Part III: experimental validation of the energy model. Int. J. Refrig. 32 (2), 359- 368. Ferrua, M.J., Singh, R.P., 2009d. Guidelines for the forced-air cooling process of strawberries. Int. J. Refrig. 32 (8), 1932- 1943. Ferrua, M.J., Singh, R.P., 2011. Improved airflow method and packaging system for forced-air cooling of strawberries. Int. J. Refrig. 34, 1162- 1058. Ho, S.H., Rosario, L., M.Rahman, M., 2010. Numerical simulationoftemperatureandvelocityin a refrigeratedwarehouse. Int. J. Refrig. 33, 1015-1025. Kader, A.A., 2002. Postharvest Technology of Horticultural Crops, thirded. University of California, Division of Agriculture and Natural Resources Publication, Oakland, CA. 3311pp. Kumar, R., Kumar, A., Murthy, U.N., 2008. Heat transfer during forced air precooling of perishable food products. Biosystems Eng. 99, 228 – 233. Manganaris, G.A., Iliasb, I.F., Vasilakakisa, M., Mignanic, I., 2007. The effect of hydrocooling on ripening related quality attributes and cell wall physicochemical properties of sweet cherry fruit (Prunus avium L.). Int. J. Refrig. 30, 1386-1392. Nalbandi, H., 2015. Numerical and experimental study of strawberry pre-cooling by forced-air system. Ph.D. Thesis, University of Tabriz, Tabriz, Iran. Tutar, M., Erdogdu, F., Toka, B., 2009. Computational modeling of airflow patterns and heat transfer prediction through stacked layers’ products in a vented box during cooling. Int. J. Refrig. 32, 295 – 306. Van der Sman, R.G.M., 2002. Prediction of airflow through a vented box by the Darcy Forchheimer equation. J. Food Eng. 55, 49-57. Vigneault, C., Goyette, B., Castro, L.R., 2006. Maximum slat width for cooling efficiency of horticultural produce in wooden crates. Postharvest Biology and Technology. 40, 308–313. Zou, Q., Linus, U.O., McKibbin, R.A., 2006a. CFD modeling system for airflow and heat transfer in ventilated packaging for fresh foods: I. Initial analysis and development of mathematical models. J. Food Eng. 77 (4), 1037- 1047. Zou, Q., Linus, U.O., McKibbin, R.A., 2006b. CFD modeling system for airflow and heat transfer in ventilated packaging for fresh foods: II. Computational solution, software development, and model testing. J. Food Eng. 77 (4), 1048- 1058.
20
Table 1 Uniformity of cooling of strawberries in PAS and MPAS. System PAS
MPAS
Criterion
Package 1
Package 2
Total
Half-cooling time (min)
107
45
76
Standard deviation
40
22
45
Half-cooling time (min)
59
73
66
Standard deviation
26
38
33
21
Fig. 1. The industrial forced-air cooling system of strawberries (Ferrua and Singh 2009d).
22
Fig. 2. The airflow path in the novel forced-air cooling system of strawberries introduced by Ferrua and Singh (2011).
24
Fig. 3. The innovative Parallel Airflow System (PAS)
25
Fig. 4. The air path in the PAS.
27
Fig. 5. Three middle walls and computational domain.
28
Fig. 6. The final computational domain.
29
Fig. 7. The experimental packages and tray.
30
Fig. 8. Air Restriction Plate (ARP) in the Modified Parallel Airflow System.
31
Fig. 9. Air velocity filed in PAS (The figure was rotated as original computational domain to make the results to be more understandable).
32
45
Air flow direction
Airflow rate (×10-6 m3/s)
40
PAS-Package 1
35
PAS- Package 2
30
MPAS- Package 1
25
MPAS- Package 2
20 15 10 5 0 1
2
3
4
5
6
7
8
9
10
11
12
13
Outlet vents Fig. 10. Distribution of airflow leaving the packages through the outlet vents in PAS and MPAS.
33
Fig. 11. The slice plane plot of air temperature distribution after 3 hours in PAS.
34
Average Temperature of package (°C)
18 16
PAS- Package 1
14
PAS- Package 2
12
MPAS-Package 1
10
MPAS- package 2
8 6 4 2 0 0
50
100
150
200
Cooling time (min)
Fig. 12. Average fruit temperature in the packages versus cooling time in PAS and MPAS.
35
Fig. 13. The surface temperature of strawberries in PAS after 60 minutes.
36
Average Temperature of package (°C)
18 16
Simul-Package 1
14
Simul-Package2
12
Exper-Package1
10
Exper-Package2
8 6 4 2 0 0
50
100
150
200
Cooling time (min)
Fig. 14. Simulated (Simul) and experimental (Exper) average temperature of instrumented fruits versus cooling time in MPAS.
37
Fig. 15. The slice plane plot of air velocity in the vertical plane of MPAS.
38
Fig. 16. The surface temperature of strawberries MPAS.
39
S0960-3085(16)30102-X http://dx.doi.org/doi:10.1016/j.fbp.2016.09.002 FBP 772
To appear in:
Food and Bioproducts Processing
Received date: Revised date: Accepted date:
11-10-2015 29-8-2016 16-9-2016
Please cite this article as: Nalbandi, Habibeh, Seiiedlou, Sadegh, Ghasemzadeh, Hamid R., Rangbar, Faramarz, Innovative parallel airflow system for forced-air cooling of strawberries.Food and Bioproducts Processing http://dx.doi.org/10.1016/j.fbp.2016.09.002 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Innovative parallel airflow system for forced-air cooling of strawberries
Habibeh Nalbandi1*, Sadegh Seiiedlou1, Hamid R. Ghasemzadeh1, Faramarz Rangbar2 1
Department of Biosystem Engineering, Faculty of Agriculture, University of Tabriz, Tabriz, Iran
2
Department of Mechanical Engineering, Faculty of Mechanical Engineering, University of Tabriz, Tabriz, Iran *
Corresponding Author Email: [email protected]; [email protected] Tell: +9804133392780
1
Highlights
An innovative airflow patterns was introduced for precooling of strawberries.
Cooling process of fruit was simulated inside the new designed package and tray.
The new system was able to distribute the cold air uniformly throughout the packages.
Some improvements were made in homogeneous of strawberry precooling process.
2
Abstract Strawberry is a high value crop and mostly is consumed as a fresh fruit over the world. Precooling of strawberry is one of the postharvest processes that reduces the fruit decay. However, in the traditional cooling system of strawberry considerable differences are observed between the fruit temperature located in the individual packages as well as in the various packages inside the tray. As a result, the self-life of fruit decreases. The aim of this research was introducing the new package and cooling system for precooling of strawberry in which some improvements were made in the uniformity of cooling of strawberry and reduced the fruit decay. The cooling process of strawberry was simulated based on simultaneous airflow and heat transfer process inside the new system and the developed model was validated experimentally, and a good agreement observed between the simulated and measured temperature of fruit. The results of simulator and experimental data showed that the new system had the higher uniformity in the cooling of fruit as compare with the traditional system so that the packages located along the cold air direction received the equal airflow rate. As a result, a difference of 0.84 ºC was measured in the average fruit temperature of the packages after 3 hours of cooling meaning that the differences in the 7/8
th
cooling time of fruit between the packages is negligible. Therefore,
with the same airflow rate of traditional system, the cooling time of strawberries could be reducing by the proposed system significantly. Keywords: Strawberry, Simulation, Innovative parallel airflow system, Precooling, Package design.
3
Nomenclature CFD Cp k MPAS PAS P t T u µ ρ Subscripts p a
Computational Fluid Dynamics Specific heat capacity (J/(kg.ºC)) Thermal conductivity (W/(m.ºC)) Modified Parallel Airflow System Outward vector normal to the surface Parallel Airflow System Pressure (Pa) Time (s) Temperature (ºC) Velocity (m/s) Dynamic viscosity (Pa.s) Density (kg/m3) Product Air
1. Introduction Strawberry is a high value crop and mostly is consumed as fresh fruit, i.e. over 75% of the product is delivered by fresh produce market. However, it is one of the most perishable fruit and is susceptible to mechanical damage, microbial decay and water loss. These prevent the product, in some cases, reaching the consumer at its optimal quality after transport and distribution (Anderson, et al., 2004; Manganaris, et al., 2007). Good temperature management is a key factor for delaying product deterioration, maintaining product quality and extending its shelf life. The cooling of fruit and vegetables from ambient temperature down to a proper temperature controls the microbial activity, retards the ripening process, and reduces respiration, wilting and shriveling due to moisture loss. Two hours delay in cooling of strawberry is long enough to reduce the proportion of marketable fruit dramatically (Kader, 2002). Among various precooling techniques, forced-air cooling is widely accepted as an appropriate method for cooling of most fruit such as strawberry (Becker, et al., 1996; Brosnan and Sun, 4
2001; Chakraverty and Paul, 2001; Anderson, et al., 2004; Castro, et al., 2005; Tutara, et al., 2009). Poor airflow distribution at different locations in the package leads to considerable heterogeneities in the final temperature of product (Alvarez and Flick, 1999a; Amara, et al., 2004; Dehghannya, et al., 2011). Many researchers have used some experimental methods to study the factors that affect the airflow pattern and cooling efficiency (Alvarez and Flick, 1999a, and 1999b; Amara, et al., 2004; Anderson, et al., 2004; Castro, et al., 2004; Vigneault, et al., 2006; Kumar, et al., 2008). However, there are some problems associated with experimental studies such as being expensive, time consuming and situation specific (Zou, et al., 2006a). To overcome these problems, mathematical approaches have been developed to model the transport phenomena inside the packages, for predicting the temperature and airflow patterns and cooling homogeneity. The medium inside the packages were considered as a porous medium in most of these models (Zou, et al., 2006a, 2006b; Van der Sman, 2002). This assumption was made due to limitations in computational resources. Even though the computational method is cost-effective, inevitable simplifications made in porous medium approach, lead to considerable errors in predictions (Dehghannya, et al., 2008; Ferrua and Singh, 2009b). Recent advances in the computational resources and decreasing cost of modern computers have made the application of CFD modeling more efficient and popular and provided more powerful tool for solving the fundamental continuity, momentum and heat transfer equations and predicting the airflow patterns and temperature variation in the packages. The mathematical modeling capability of predicting the cooling process of fruit within the packages has been used as an ideal approach by many researchers, especially when the container-to-product equivalent
5
diameter ratio was under 10, e.g. strawberry packages. In such situation, the porous medium approach could not be used (Dehghannya, et al., 2010; Ferrua and Singh, 2011). Some researchers studied the cooling process of strawberry using various package and tray designs as well as pallet arrangements (Anderson, et al., 2004; Ferrua and Singh, 2007, 2009a, 2009b, 2009c, 2009d, 2011). In the forced-air cooling system that is shown in the Fig. 1-a, the cold air is forced across each palletized row (Fig. 1-b and 1-c), through the vents in the trays and packages (I in Fig 1-d) due to the negative relative pressure created by the system suction fan. Therefore, the cold air enters from the right side of each row and the warmer air exits from the left side (Fig. 1-d). There are some problems associated with this method. Firstly, the air temperature entering to the various packages is not the same as it travels along the packages. The warmer exiting air from package 1 (P1) enters the next one (Fig. 1-d), as a result the air temperature increases gradually as an in-series method. As a result, there was about 6ºC difference in average fruit temperature between P1 and P8 after 1h of cooling process (airflow rate: 1L/(s.kgp)). In the in-series method, modifying the package and/or tray would not improve the rate and uniformity of the cooling process; because the air temperature still rises as it travels along the systems (Ferrua and Singh, 2009d). Therefore, the last package (P8) always receives higher temperature air than the first one (P1). The higher air temperature leads to a lower heat removal rate. Secondly, the most of total cold airflow bypasses the system without removal of heat from the fruit. This significant bypass could be explained by the differences in the shape of packages and trays. Ferrua and Singh (2011) developed a novel forced-air cooling system that split the airflow across each layer of the pallet into two streams of parallel airflow (Fig. 2-a, path 1 and 2). They increased the vent area of tray 2 (left side in Fig. 2-a) 60% larger than the vent of tray 1(right
6
side of Fig. 2-a). In addition, the vent area of the packages that were located inside tray 2 was increased up to 40%. This new system improved relatively the uniformity of cooling of products between two trays and reduced the total cooling time (by 6%), pressure drop and energy consumption of process. Apart from what Ferrua and Singh (2011) achieved, there is still a considerable heterogeneity in the cooling process of strawberries located in various trays as well as packages in each tray. Because, the in-series method of cooling still occurs in each tray, there is heterogeneity in temperature of fruit in the package 1 and 4 and also package 5 and 8 (P1 and P4; P5 and P8 in Fig 2-b). As a result, about 60 minutes difference was measured in 7/8th cooling time of packages 1 and 4 (Ferrua and Singh, 2011). The aims of this work were: (a) Introducing an innovative packaging system for the forced-air cooling of strawberries to increase the effectiveness of process as well as uniformity of cooling process; (b) Numerical modeling and simulating of the precooling process of strawberries inside the new packages and proposed system; (c) Experimental study and validating the developed model and simulator; (d) Improving the various parameter of cooling process such as the uniformity of fruit temperature inside the individual packages, uniformity of the cooling process between the various packages in the system, reducing the cooling time and percentage of bypassed air using the simulator.
2. Materials and methods Any design for package and tray of fresh fruit should improve the uniformity of product cooling. In this work, an innovative Parallel Airflow System (PAS) was introduced to improve the homogeneity of strawberries pre-cooling process. In the proposed system, two separate ducts were designed for introducing the cold air into the individual packages in a way that the exiting
7
heated air can return to air cooling unit without re-entering other packages. One of these ducts (duct 1) was built on the top of the packages (Fig. 3-a, II) and the other one (duct 2) was located between the packages (Fig. 3-b, I). The cold air enters to the duct 1 before entering the packages where it removes heat from the fruit and exits through duct 2 (a detailed description will be presented in the following sections). Package design Due to the strong effect of the package design and airflow pattern on the performance of the Parallel Airflow System, different package designs were considered and their effects on the performance as well as cooling homogeneity have already been investigated by the present authors (Nalbandi, 2015). The best one was used in this work Fig.4. In order to create a duct between the packages (duct 2; Fig. 3-b, I), a trapezoidal cross section was considered for the strawberry package (Fig. 3-b and 3-e). The location and shape of the vents on package walls play an important role on the Parallel Airflow System because of relating the inlet and outlet ducts. The vents can be distributed on the top (Fig. 3-f, I) and lateral walls (Fig. 3-f, II) as inlet and outlet vents. In this design, the opening area was 28% of the total surface of the top and lateral walls of the package. 2.1. Tray design A new design of tray was needed providing a duct on the top of the packages (duct 1; Fig. 3-a). Therefore, the tray walls were assumed taller than the package. Fig. 3-d shows the tray built to carry the packages. Each tray contains eight packages with 2 by 4 arrangements (Fig. 3-c). There was no gap between the packages. The Parallel Airflow System requirements make it necessary to have a different vent design in the front and back walls of tray associated with ducts 1 and 2; side A and B in Fig. 3-d. A
8
detailed description will be presented in the next section. The vent in the side A is related to the duct 1; top of the packages (Fig. 3-a), and the trapezoidal vents in the side B is referred to the duct 2; between the packages (Fig. 3-b). The vent area of the tray was 28% and 13% at side A and B, respectively. 2.2. The airflow paths As previously mentioned, ducts 1 and 2 are used as inlet and outlet of airstreams, respectively. One end of the ducts is closed alternately; the side near the fan in duct 1 (Figs. 3-a, II) and the other side in duct 2 (Fig. 3-b, I). When the circulating suction fan is turned on, the cold airstream enters duct 1, then passes through top openings of packages (I in Fig. 3-f), and exits finally through openings on the walls of the packages (II in Fig. 3-f). The heated airstream returns to cold room for recirculation through duct 2 (Fig. 4). It is obvious that any individual package always receives cooled air directly from cold room and the air exiting the package never enters the next one. This means that the products in the all packages always receive the air with the same temperature and the cooling of the fruit is performed in a parallel manner. 2.3. Simulation of the cooling process of strawberries To evaluate the performance of the Parallel Airflow System, the cooling process of strawberries was simulated based on mathematical modeling of airflow and heat transfer between airstream and the product. To evaluate the efficiency of the Parallel Airflow System through 3D simulation, the tray was partitioned using three middle walls (Fig. 5-a) (the friction on these walls was not considered) and only quarter of tray (Fig. 5-b) containing two packages was considered as the computational domain. 9
2.4.1. Computational domain Strawberry model: A 3D strawberry model equivalent to medium fruit size was created. The area, volume and equivalent diameter of the model are 0.002504 m2, 1.12× 10-5 m3 and 0.028 m, respectively. Package model: The package model with a proper vent distribution was created based on details discussed by Nalbandi (2015). Each package was filled with 39 strawberries. The total weight of fruit was 0.46 kg. In order to avoid the need for finer mesh and undesirable skewness of the mesh elements, individual strawberries were located far enough from each other as well as from the wall so that small gaps were created between them. One-half of the packages was used in the simulation process. Tray model: A quarter of a tray was modeled geometrically. Final computational domain: There is a plane of symmetry in the computational domain. Therefore, a half of this domain was used for simulating process as a final computational domain (Fig. 6). 2.4.2. Mesh generation To establish an optimal mesh, a mesh independence study was performed by changing the mesh density to achieve a mesh that gives a solution with acceptable accuracy (Ho et al., 2010). The average fruit temperature was selected for this purpose and its variation studied at various mesh densities. Four meshes, each containing 96470, 106449, 150890 and 240532 elements were used. The results indicated that mesh independence of numerical solution was obtained when the number of elements was above 150000. Therefore, the mesh with 150890 elements was adequate for accurate numerical prediction. However, to prevent large temperature and velocity gradients, a finer mesh used in the critical areas such as inlet and outlet vents, between the fruit as well as
10
the fruit-air interfaces. The computational domain mesh containing 241706 tetrahedral elements (Lagrange Quadratic) resulted in 1265024 degree of freedom. 2.4.3. Mathematical models For precooling of strawberries, and majority of other horticultural crops, Biot number lies between 0.1 and 10. Within this range, both the internal and external resistances are on the same order and the temperature gradients will appear within the products (Anderson et al., 2004). In simulation of the system, airflow and heat transfer were considered functions of time. 2.4.3.1. Airflow model In the precooling of strawberry the average airflow rate is 1L/(s.kgp) that creates the laminar flow in the packages (Ferrua and Singh, 2009a). In the present work, the airflow rate was considered to be 0.4 L/(s.kgp) and the Reynolds number was calculated based on the hydraulic diameter of the inlet vent. It was about 710. In addition, the maximum Reynolds number was 54 inside the packages. It was calculated based on the hydraulic diameter of free space in the crosssection of the air domain at the various planes of package. Therefore, incompressible fluid and laminar flow were assumed inside and between packages. In addition, the airflow was constant and steady state condition governs during process. However, in order to use the developed model for other conditions in future studies, the mathematical model of airflow was analyzed as a time dependent process. Eqs. 1 and 2 give the mass and momentum conservations for air in the system, respectively. The buoyancy force term is not included in the airflow equation and the air properties were assumed constant. The density and the viscosity were assumed to equal to those of dry air at 0°C. . u 0
ρa
(1)
u T ρ a u .u P . μ a u u t
11
(2)
The airflow model was completed by considering the following boundary conditions. Inlet boundary: laminar inflow was applied to define the inlet boundary condition for airflow model. It was assumed that the air pressure is equal to the atmospheric pressure at distance of 0.5 m from the inlet. Outlet boundary: the constant velocity at the outlet vents of tray was defined. Its direction was perpendicular to the vent. Wall boundaries: a no slip boundary condition was applied to all solid surfaces including inner surfaces of trays, all surfaces of package and fruit surfaces. Symmetry boundary: symmetry boundary condition was used on the plane of symmetry. The boundary conditions on the velocity can be written as: Inlet: p p0
(3)
Outlet: u u 0
(4)
Wall: u 0
(5)
Interface: u 0
(6)
2.4.3.2. Heat transfer model in the product domain Eq. 7, the transient equation that describes the heat transfer process in the strawberries domain. The respiration process was assumed negligible because of low transferred heat. Other assumptions were also made in the model: the thermo-physical properties of strawberry were assumed to remain constant and there would be no moisture loss nor shrinkage during the cooling process (kp=0.57 W/(m.ºC); Cp,p= 3.95 kJ/(kg.ºC); ρp=800 kg/m3). In addition, the initial temperature of strawberry was set to16 ºC. ρ p c p,p
Tp t
. k pTp
(7)
12
For heat flux at the products surface, the following boundary condition was adopted:
k PTP k a Ta . n 0
(8)
2.4.3.3. Heat transfer model in the fluid domain The heat transfer equation within the fluid domain was written as Eq. 9. This equation describes a time dependent process that includes the conduction and convection terms. The air properties were assumed those of dry air at 0°C. In addition, the air temperature was assumed to be 0 ºC. ρ a c p,a
Ta ρ a c p,a u .Ta . k a Ta t
(9)
The appropriate boundary conditions for Eq. 9 are as follows: Inlet: temperature, Ta Ta0
(10)
Outlet: convective flux, k a Ta . n 0
(11)
Wall: thermal insulation, k a Ta . n 0
(12)
Interface: temperature, Ta TP
(13)
Symmetry plane: thermal insulation, k a Ta . n 0
(14)
2.4.4. Numerical solution COMSOL MULTIPHYSICS software, version 3.5, was used to simulate the heat transfer process and airflow using the finite element method. There is a potential numerical instability in the finite element computation of incompressible flows. It refers to the Galerkin formulation. To prevent this numerical instability, isotropic diffusion with a tuning parameter of 0.5 was selected among the various stabilized finite element formulations. The simulation was performed using a personal computer with 32 GB RAM (Processor: Intel (R) Core (TM) i7-2700K CPU@ 3.50
13
GHz 3.50 GHz). A variable time step procedure was used through the simulation. The solution time was about 552 hr. 2.4. Experimental validation of the model To validate the numerical model a forced-air cooling system was used in a cold room where the temperature and relative humidity were 0.1 ºC and 80%, respectively. This system consists of a suction fan to generate a desirable airflow rate and an open tunnel with 50 cm2 cross section and 120 cm length. For each test, the fresh fruit were purchased from the local greenhouse and stored in the refrigeration room at 1 ºC. Fruit were placed at a room with temperature of 16 ºC to equilibrate their temperature before experiments. Strawberries with a uniform shape were packed in the packages and placed inside the tray (Fig. 7). The tray in the Parallel Airflow System configuration or in the Modified Parallel Airflow System configuration (see section 2.6) was located in the center of tunnel and insulated with foam. Thermocouples (K type) were inserted into the center of 8 strawberries and temperature of these points were used for model validation. The temperature was recorded every 1 minute. The exact temperature of airflow entering the system was recorded by a thermocouple located in the front of tunnel. The average temperature of air was 0.3 ºC. The experiments were conducted three times. The moisture loss of fruit was determined by the difference in the package weight before and after cooling. 2.5. Improving the first design of Parallel Airflow System using air restriction plate At some points, it was realized that restricting the duct 1 in the midpoint might improve uniformity of cooling in particular along the package 1. Therefore, the original design Parallel Airflow System, was modified by inserting an Airflow Restriction Plate (ARP) i.e., a plate with hole (Fig. 8). For this Modified Parallel Airflow System (MPAS), a revised computational
14
domain and mesh parameters were defined and final domain containing 242869 mesh elements; Lagrange Quadratic- tetrahedral. 3. Results and discussion 3.1. Original system (Parallel Airflow System) 3.1.1. Simulated Airflow pattern and temperature Air velocity filed is presented in Fig. 9. It is obvious that the air velocity decreased from the right to left of tray. However, the airflow rate entering through the inlet vents increased along the duct 1. Therefore, the vents that were located at downstream (left side) received the largest fraction of total airflow. The airflow rate entering through each inlet vents (I in Fig. 3-f) was calculated using the software and the inlet airflow percentage was calculated for the both packages. As much as 80% of total airflow introduced into the system entered the package 2 through the end vents. Fig. 10 shows the distribution of airflow rates leaving the packages through the outlet vents. It can be seen that a larger proportion of total airflow exited through the vents close to the end side of the packages (the number of vents were shown in Fig. 6). Fig. 11 presents the slice plane plot of air temperature distribution in vertical planes. The vertical plane shows the air temperature distribution in the inlet vents of packages. As expected, the temperature of the air entering into packages is the same at all opening points. 3.1.2. Simulated average fruit temperature in the packages The average fruit temperature in the two packages versus cooling time is shown in Fig. 12, where a difference of 2.8 ºC was seen between the average fruit temperature in the packages after 3 h cooling. In addition, the half-cooling time of fruit in the package 1 and 2 was about 107 and 45 min, respectively. It is clear that the package 2 was cooled faster. This result can be justified by the airflow rate entering the packages. Although both packages received the air with the same
15
temperature, there was a significant difference between the entering airflow rate (package 1 and 2 receive 20% and 80% of total airflow rate, respectively). A lower airflow rate passing through the package leads to larger convective heat transfer resistance within the package. The standard deviation (SD) of the half-cooling time of all fruit as a criterion for uniformity of cooling process are presented in Table 1. There is a considerable difference between the halfcooling time and SD of both packages showing the high heterogeneity of cooling. The effect of airflow distribution on the uniformity of fruit surface temperature is illustrated in Fig. 13. Unexpectedly, the fruit located at the top row was not necessarily cooled faster. The fruit located closer at the rear and bottom of the package was cooled faster. Generally, the fruit located at the second half of the package was cooled faster. One possible justification is poor airflow distribution within the package. About 86% of total air entering the package exit through the outlet openings located at the second half of package. Therefore, there was a large amount of airflow in the second half of the package that led to faster cooling of fruit located in this zone. 3.2. Experimental results of validation Fig. 14 demonstrates the simulated and experimental average temperature of instrumented fruit in the Parallel Airflow System. As this figure indicates, there is good agreement between the simulated and experimental data. The observed temperature differences could be due to the errors caused by inaccurate positioning thermocouples and fruit properties as well as modeling assumptions particularly no moisture loss. The measured moisture loss of fruit was not negligible and it is the major reason of the error, because of the latent heat of vaporization that was taken from the fruit. However, the developed model could be used in future studies of package designing. 3.3. Modified Parallel Airflow System
16
To improve cooling efficiency, the Parallel Airflow System configuration was modified based on the experimental and simulation results. According to fluid mechanics, reducing the amount of air entering the package 2, increases the amount of air entering the package1. The results of simulation are presented in the following paragraphs. 3.3.1. Simulated airflow pattern The results showed that using restriction plate improved air distribution quality between two packages so that the packages 1 and 2 received almost 46 and 54 percent of total airflow, respectively (Fig. 15). Within individual packages, air distribution quality between openings was also improved. This was more remarkable in package 1. The inlet opening located behind the ARP received a lower airflow rate. About 25 % of the total airflow rate entered the package 2 through the inlet vents of its first half. This was almost 43% in the case of the package 1. As observed in the Parallel Airflow System, a larger proportion of the total airflow entering each package exited through outlet vents of the package, which were located in the second half of package. As expected from the airflow pattern, a lower amount of air entering the first half of the packages led to higher convective heat transfer resistance in these zones. Therefore, the air temperature was higher than in other spaces, especially in the case of package 2. 3.3.2. Average fruit temperature in the packages As previously stated, Fig. 12 shows the profile of average fruit temperature in the packages versus cooling time. As shown, the cooling process of fruit inside the both packages was performed more uniformly so that 0.84 ºC difference was observed in the average fruit temperature of the packages 1 and 2 after 3h cooling. This can be attributed to uniformity of the amount and temperature of the air entering the packages.
17
As shown in Table 1, the half-cooling differences between the packages 1 and 2 are lower than for Parallel Airflow System. However, the uniformity of cooling process within the package 2 was not improved in Modified Parallel Airflow System (Fig. 16). Such non-uniformity can be attributed to the poor airflow distribution through the inlet openings. As discussed earlier, the first half of the both packages receives low airflow rate, especially in the package 2. Therefore, more design improvements must be considered.
4. Conclusions In the present work, an innovative parallel airflow system for forced-air cooling of strawberries (Modified Parallel Airflow System) was proposed. The numerical analysis of airflow and heat transfer inside the packages and tray showed that the Modified Parallel Airflow System could improve the uniformity of the cooling process in strawberries. The Modified Parallel Airflow System was able to deliver the equal amount of cold air to the packages at the same temperature, so that the packages 1 and 2 received almost 46 and 54 percent of total air. The cooling process of fruit inside the both packages was performed more uniformly and 0.84 ºC difference was observed in the average fruit temperature of the packages 1 and 2 after 3h of cooling process and a good fit was observed between the simulated and experimental data. Based on the results, this system could be used in industrial forced air-cooling of strawberries. However, the observed heterogeneity inside the individual packages indicated that more design improvements are still needed for package and tray.
18
References Alvarez, G., Flick, D., 1999a. Analysis of heterogeneous cooling of agricultural products inside bins. Part I: aerodynamic study. J. Food Eng. 39, 227–237. Alvarez, G., Flick, D., 1999b. Analysis of heterogeneous cooling of agricultural products inside bins. Part II: thermal study. J. Food Eng. 39, 239–245. Amara, S.B., Laguerre, O., Flick, D., 2004. Experimental study of convective heat transfer during cooling with low air velocity in a stack of objects. Int. J. Thermal Sci. 43, 1213–1221. Anderson, B.A., Sarkar, A., Thompson, J.F., Singh, R.P., 2004. Commercial scale forced air cooling of strawberries. Trans. ASAE, 47 (1), 183-190. Brosnan, T., Sun, D., 2001. Precooling techniques and applications for horticultural products: a review. Int. J. Refrig. 24, 154-170. Becker, B.R., Misra, A., Fricke, B.A., 1996. Bulk refregeration of fruits and vegetables part I: theoretical consierations of heat and mass transfer. HVAC&R Research, 2, 122–134 Chakraverty, A., Paul, S.R., 2001. Postharvest Technology: Cereals, Pulses and Vegetables. Sci. Publ., India. Castro, L.R., Vigneault, C., Cortez, L.A.B., 2005. Cooling performance of horticultural produce in containers with peripheral openings. Postharvest Biol. Technol. 38, 254–261. Castro, L.R., Vigneault, C., Cortez, L.A.B., 2004. Container opening design for horticultural produce cooling efficiency. J. Food, Agri. Envir. 2, 135–140 Dehghannya, J., Ngadi, M., Vigneault, C., 2008. Simultaneous aerodynamic and thermal analysis during cooling of stacked spheres inside ventilated packages. Chemical Eng. and Technology. 31(11), 1651-1659. Dehghannya, J., Ngadi, M., Vigneault, C., 2010. Mathematical Modeling Procedures for Airflow, Heat and Mass Transfer During Forced Convection Cooling of Produce: A Review. Food Eng. Reviews. 2, 227-243. Dehghannya, J., Ngadi, M., Vigneault, C., 2011. Mathematical modeling of airflow and heat transfer during forced convection cooling of produce considering various package vent areas. Food Control. 22, 1393-1399.
Ferrua, M.J., 2007. Modeling the forced-air cooling process of fresh horticultural products. Ph.D. Thesis, University of California, Davis, USA. Ferrua, M.J., Singh, R.P., 2009a. Modeling the forced-air cooling process of fresh strawberry packages. Part I: numerical model. Int. J. Refrig. 32 (2), 335- 348. 19
Ferrua, M.J., Singh, R.P., 2009b. Modeling the forced-air cooling process of fresh strawberry packages. Part II: experimental validation of the flow model. Int. J. Refrig. 32 (2), 349- 358. Ferrua, M.J., Singh, R.P., 2009c. Modeling the forced-air cooling process of fresh strawberry packages. Part III: experimental validation of the energy model. Int. J. Refrig. 32 (2), 359- 368. Ferrua, M.J., Singh, R.P., 2009d. Guidelines for the forced-air cooling process of strawberries. Int. J. Refrig. 32 (8), 1932- 1943. Ferrua, M.J., Singh, R.P., 2011. Improved airflow method and packaging system for forced-air cooling of strawberries. Int. J. Refrig. 34, 1162- 1058. Ho, S.H., Rosario, L., M.Rahman, M., 2010. Numerical simulationoftemperatureandvelocityin a refrigeratedwarehouse. Int. J. Refrig. 33, 1015-1025. Kader, A.A., 2002. Postharvest Technology of Horticultural Crops, thirded. University of California, Division of Agriculture and Natural Resources Publication, Oakland, CA. 3311pp. Kumar, R., Kumar, A., Murthy, U.N., 2008. Heat transfer during forced air precooling of perishable food products. Biosystems Eng. 99, 228 – 233. Manganaris, G.A., Iliasb, I.F., Vasilakakisa, M., Mignanic, I., 2007. The effect of hydrocooling on ripening related quality attributes and cell wall physicochemical properties of sweet cherry fruit (Prunus avium L.). Int. J. Refrig. 30, 1386-1392. Nalbandi, H., 2015. Numerical and experimental study of strawberry pre-cooling by forced-air system. Ph.D. Thesis, University of Tabriz, Tabriz, Iran. Tutar, M., Erdogdu, F., Toka, B., 2009. Computational modeling of airflow patterns and heat transfer prediction through stacked layers’ products in a vented box during cooling. Int. J. Refrig. 32, 295 – 306. Van der Sman, R.G.M., 2002. Prediction of airflow through a vented box by the Darcy Forchheimer equation. J. Food Eng. 55, 49-57. Vigneault, C., Goyette, B., Castro, L.R., 2006. Maximum slat width for cooling efficiency of horticultural produce in wooden crates. Postharvest Biology and Technology. 40, 308–313. Zou, Q., Linus, U.O., McKibbin, R.A., 2006a. CFD modeling system for airflow and heat transfer in ventilated packaging for fresh foods: I. Initial analysis and development of mathematical models. J. Food Eng. 77 (4), 1037- 1047. Zou, Q., Linus, U.O., McKibbin, R.A., 2006b. CFD modeling system for airflow and heat transfer in ventilated packaging for fresh foods: II. Computational solution, software development, and model testing. J. Food Eng. 77 (4), 1048- 1058.
20
Table 1 Uniformity of cooling of strawberries in PAS and MPAS. System PAS
MPAS
Criterion
Package 1
Package 2
Total
Half-cooling time (min)
107
45
76
Standard deviation
40
22
45
Half-cooling time (min)
59
73
66
Standard deviation
26
38
33
21
Fig. 1. The industrial forced-air cooling system of strawberries (Ferrua and Singh 2009d).
22
Fig. 2. The airflow path in the novel forced-air cooling system of strawberries introduced by Ferrua and Singh (2011).
24
Fig. 3. The innovative Parallel Airflow System (PAS)
25
Fig. 4. The air path in the PAS.
27
Fig. 5. Three middle walls and computational domain.
28
Fig. 6. The final computational domain.
29
Fig. 7. The experimental packages and tray.
30
Fig. 8. Air Restriction Plate (ARP) in the Modified Parallel Airflow System.
31
Fig. 9. Air velocity filed in PAS (The figure was rotated as original computational domain to make the results to be more understandable).
32
45
Air flow direction
Airflow rate (×10-6 m3/s)
40
PAS-Package 1
35
PAS- Package 2
30
MPAS- Package 1
25
MPAS- Package 2
20 15 10 5 0 1
2
3
4
5
6
7
8
9
10
11
12
13
Outlet vents Fig. 10. Distribution of airflow leaving the packages through the outlet vents in PAS and MPAS.
33
Fig. 11. The slice plane plot of air temperature distribution after 3 hours in PAS.
34
Average Temperature of package (°C)
18 16
PAS- Package 1
14
PAS- Package 2
12
MPAS-Package 1
10
MPAS- package 2
8 6 4 2 0 0
50
100
150
200
Cooling time (min)
Fig. 12. Average fruit temperature in the packages versus cooling time in PAS and MPAS.
35
Fig. 13. The surface temperature of strawberries in PAS after 60 minutes.
36
Average Temperature of package (°C)
18 16
Simul-Package 1
14
Simul-Package2
12
Exper-Package1
10
Exper-Package2
8 6 4 2 0 0
50
100
150
200
Cooling time (min)
Fig. 14. Simulated (Simul) and experimental (Exper) average temperature of instrumented fruits versus cooling time in MPAS.
37
Fig. 15. The slice plane plot of air velocity in the vertical plane of MPAS.
38
Fig. 16. The surface temperature of strawberries MPAS.
39