CFD Assisted Study of Multi-Chapels Greenhouse Vents Openings Effect on Inside Airflow Circulation and Microclimate Patterns

The aim of this work is to study and quantify the air mass flow exchanged between inside and outside of the greenhouse, in order to determine the ventilation openings layout and the design effect on greenhouse airflow and microclimate distribution. The study was conducted over a 945 m multi-chapels arched greenhouse with a polyethylene cover and has thirteen crop rows oriented from north to south; the greenhouse was equipped with side wall and roof vents openings. A simulation was performed using different arrangements and configurations of ventilation openings with the same wind direction. Numerical simulation has been adopted in three dimensions (CFD), using the Fluent computer code which relies on the resolution of the Navier-Stokes equations. These equations were solved in the presence of the turbulence model (k ε) and the Boussinesq model equation adopted to incorporate buoyancy forces. The effects of solar and atmospheric radiation were included by solving the radiative transfer equation (RTE), using Discrete Ordinate (DO) model. The effects of the roof openings, the presence of anti-insect screens and crops orientation were investigated and quantified. In a 3-span greenhouse with an anti-aphid insect screen in the vent openings, combining roof and sidewall vents gave a ventilation rate per unit opening area that was 1.4 times more than with only side vents. In the latter case, the difference of temperature between the inside and the outside of the greenhouse was greater than 3 ̊C. Numerical simulations with an anti-insect screen having a porosity of 56% showed that the air exchange rate with combined ventilation was reduced by 48%. Finally, the paper focused on the effect of vent arrangement on the efficiency of the ventilation and the distribution of the microclimate inside the greenhouse. Results showed that computed ventilation rates varied from 53.43 to 70.95 kg/s, whereas temperature differences How to cite this paper: Senhaji, A., Mouqallid, M. and Majdoubi, H. (2019) CFD Assisted Study of Multi-Chapels Greenhouse Vents Openings Effect on Inside Airflow Circulation and Microclimate Patterns. Open Journal of Fluid Dynamics, 9, 119-139. https://doi.org/10.4236/ojfd.2019.92009 Received: May 6, 2019 Accepted: June 9, 2019 Published: June 12, 2019 Copyright © 2019 by author(s) and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access


Introduction
Natural ventilation is considered to be one of the most important factors in the greenhouse environment [1] because it directly affects the greenhouse air temperature, humidity and CO 2 concentration. The Mediterranean region which is characterized by a large radiating flow, such efficient air conditioning is crucial to reduce the temperature of the greenhouse air and eliminate excess humidity.
The ability to maintain the desired climate in greenhouse production system depends on the design and performance of the ventilation system [2]. Natural ventilation is used by the vast majority of growers in the Mediterranean region because it is the best economical way to regulate microclimate of greenhouses. However, the control of airflow with natural ventilation is limited, so it is necessary to analyze the effectiveness of natural ventilation.
The arrangement of the openings may vary from a greenhouse to another. For most greenhouses, the openings are arranged, continuously or alternatively, along the side walls and/or the roof (Venlo-type greenhouses). However, in some cases, the greenhouse may have only holes in the roof (parral type) or discontinuous openings obtained by separating the plastic cover on either side of a tunnel greenhouse.
To prevent crop pests (whitefly and thrips, for example), especially in the Mediterranean regions, the openings must be equipped with insect-proof screens that strongly influence the ventilation process.
Recently Majdoubi et al. [3] have shown that the insect screen reduced the greenhouse ventilation rate by 46%, and the tomato rows that were oriented perpendicular to the prevailing air movement through the greenhouse reduced the ventilation rate by 50%.
Kittas and Bartzanas [4] have studied the efficiency of two configurations concerning ventilation openings on the greenhouse microclimate during the dehumidification process. The results of the simulations performed for an outside wind direction perpendicular to the greenhouse axis, show clearly the influence of ventilation openings configurations on the velocity, temperature and humidity distribution inside the greenhouse.
In recent years, computational fluid dynamics (CFD) has become an important tool that is widely used in many fields. In numerical modelling of microclimate greenhouse, many researchers have used CFD to study the effects of Kacira et al. [7] studied the effect of vent configuration in naturally ventilated greenhouses, they observed that the maximum greenhouse ventilation rates were achieved when rollup type side vents were used in the side walls. The rollup side vents considerably enhanced the ventilation rate in the plant canopy zone. More recently, Bartzanas et al. [8] studied the influence of vent arrangement on windward ventilation of a tunnel greenhouse using commercial fluid dynamics code. They showed that the largest ventilation rate did not necessarily correspond to the best greenhouse air temperature and velocity distribution.
Other studies are interested in the effects of specific elements and of outdoor weather conditions on natural ventilation [9] [10] [11] [12]. Few of these studies have examined the effect of vent used for cooling and dehumidification [9] [13] [14] but their findings provide valuable guidance for the management of the ventilation system; However, they are limited to the specific structures and local climatic conditions.
Results published by Bournet and Boulard [13] shows main factors influencing the movement of air inside the greenhouse -in terms of ventilation efficiency inside it-are examined on the geometry of the greenhouse and the opening arrangement. Other parameters affecting the ventilation, such as wind speed and direction, the addition of anti-insect proofs or shade screens, and interactions with culture, are also discussed.
Another aspect of ventilation investigated concerned the effects of anti-insect screens placed over the vent openings. Insect screening reduces ventilation, which in turn causes air temperature and relative humidity to rise significantly.
Molina-Aiz et al. [15] observed a 50% reduction of the ventilation rate for a screen porosity of 39%. For a porosity of 50%, Bartzanas et al. [16] also simulated an induced 50% loss of ventilation efficiency. Considering screens with a 25% porosity, Baeza et al. [9] simulated a reduction of the ventilation rate comprised between 77% and 87%, depending on the number of spans for a parral greenhouse equipped with both sidewall and roof vents. Fatnassi et al. [17] observed that the ventilation rates with anti-thrip (porosity 19%) or anti-aphid (porosity 56%) screens represented 41% and 53%, respectively, of the flow without a net.
The numerical simulation (CFD) is now more developed, integrating the radiative exchange between the atmosphere and the greenhouse environment [5] [18] as well as the transfer of water vapor and heat between culture and air [19] [20] [21].
However extensive researches in this field, a few studies have included humidity and radiative mechanisms by solving the Radiative Transfer Equation (RTE) for luminance in the greenhouse. Studies using CFD on the effect of the crop rows orientation and the effect of openings arrangement with anti-insect proofs on ventilation and inside climate in a greenhouse are even rarer.
Spurred on by these important issues, the aim of this work is to study the ar- It seeks a better understanding of the behavior of the greenhouse type to be studied and also the management of the microclimate inside the greenhouse. Airflows established inside the greenhouse were analyzed for different types of geometries. The impact of the combination of openings, the presence of insect screens, the width and number of the openings, the height of the greenhouse and the external wind conditions are carefully examined.

Governing Equations
The CFD methods can explicitly calculate the velocity field and the associated temperature field of flow by numerically solving the corresponding transport equations. The three-dimensional conservation equations describing the transport phenomena for steady flows are of the general form [22]: Φ represents the concentration of the transport quantity in three momentum conservation equations and the scalars mass and energy conservation equations.
u, v and w are the components of velocity vector; Γ is the diffusion coefficient and S Φ is the source term [18]. The governing equations are discretized following the procedure described by Patankar [22]. This consists of integrating the governing equations over a control volume. CFD code Fluent 6.3.2 was used to solve Equation (1), using the finite volume numerical scheme to solve the equations of conservation for the different transported quantities in the flow around and in the greenhouse (mass, momentum and energy).
The turbulent stress is modelled using the k-ε model. In Equation (1), Φ also represents the turbulent kinetic energy k (m 2 /s 2 ) and dissipation of the kinetic energy ε (m 2 /s 3 ). The Boussinesq model was also activated to take account the effect of gravity, which means that the buoyancy force due to the differences of the density of air is added as a source term in the momentum equation [2].

Ventilation Model
There are two possibilities to ventilate greenhouses: where V r is the ventilation rate (h −1 ), ϕ m the mass flow rate (kg/s 1 ), V g the greenhouse volume and ρ the density of air.

Turbulence Model
The effect of turbulence on the flow was implemented via the standard k-ε model. According to Nebbali et al. [24], k-ε model gives the lowest error value and it can be chosen as it represents a good compromise between the complexity of calculation and realism in the simulation of turbulence. The standard k-ε model is a semi-empirical model based on two equations, the turbulent kinetic energy (k) and dissipation rate (ε) [25]. The model constants C 1ε , C 2ε , C µ , and

Flow through Insect Screens and Plants
To take account of dynamic effects induced by the insect screens (placed over the vents) and the crop, we can model them by means of the porous medium approach governed by the Darcy-Forchheimer equation.
The flow of air through a screens porous jump, used as boundary conditions, can be expressed as: The values of coefficients, K(m 2 ) the permeability of the porous medium and Y(−) the non-linear momentum loss coefficient, were obtained from the following equations [27]: where φ is the porosity of the porous medium.
The crops were modeled as a rectangular block (with dimensions 22.8 mL × 1 mW × 2 mH) porous media approach by the addition of a momentum source term to the standard fluid flow equations. In this case, the source term was described as [26]: C 1 = 1/K is the viscous resistance (m −2 ), C 2 is the inertial resistance factor (m −1 ).

Radiation Model
Many studies have successfully applied computational fluid dynamics (CFD) in the numerical simulation of a greenhouse in order to get an overview of the resulting climate inside the greenhouse. In the majority of the studies that used CFD, the effect of solar and thermal radiations was taken into account by setting The radiative heat transfer was also calculated by using the non-gray discrete ordinates (DO) radiation model. Considering the ray direction of S , the radiative transfer equation for spectral intensity ( ) , I λ r S , can be written as [26]: where A present the rate of increase in radiation intensity, B is the loss by absorption and out-scattering, C is the gain by emission, and D is the gain by in-scattering. The no-gray model was activated by dividing the radiative spec- In the present study, the power absorption coefficient was chosen in such a way as to get a transmittance of 0.8 in the short wavelength part of the spectrum and to obtain a zero transmittance in the range of long wavelengths. a λ is computed from the absorptivity a using the following relationship in accordance to the media thickness e [28]: The beam direction and irradiation were computed by a solar calculator according to a given position, date and time. The flow iteration was set to 10 per radiation iteration. The discrete ordinate (DO) model solution required material thermal and optical parameters, as summarized in Table 1.

Meshes and Boundary Conditions
The size of the outdoor domain was much larger meshed than the greenhouse ( Figure 1 A substantial increase in the ventilation rate (11%) between the first and second grids was observed. A small increase in the ventilation rate (1%) between the second and third grids indicates that the grid resolution has almost no influence on the solution. Finally, the grid with 1,282,350 cells was chosen.
A mixed heat transfer boundary condition (combination of radiation and convection with convective heat transfer coefficient, h = 8 W/m 2 K [28]) is applied at the ground, with a diffusive radiation-opaque material. Also, the semi-transparent roof cover zone has a fluid region on each side; it is called a "two-sided wall". To couple the two sides of the wall, a coupled thermal condition is selected (This option will appear in the Wall panel only when the wall is a two-sided wall). No additional thermal boundary conditions are required, because the solver will calculate heat transfer directly from the solution in the adjacent cells, whereas the side walls were treated as coupled and opaque material. At the inlet of the computational domain a logarithmic inlet velocity profile was considered. The profile was linked to the CFD main module using the user-defined function (UDF).

Numerical Procedure
This study utilized the CFD model using software package (Fluent 6.3) to predict natural ventilation and turbulent airflow patterns inside and outside the multi-span greenhouse. The computer used was a PC with a 2 Intel processors 2.6 GHz and 12 GRAM, this software is capable of modeled simulations higher than 100,000 calculation cells.
The Green-Gauss cell based on semi-implicit method for pressure-linked equations (SIMPLE) was adopted to solve the coupled pressure-momentum equations.
The standard scheme was used for the pressure discretization and a second-order upwind discretization scheme was selected for the momentum,

Simulation of the Ventilation Rate
The mass flow rate through each vent opining surface is computed by summing the value of density multiplied by the dot product of the facet area vector and the facet velocity vector [26].
The ventilation rate of the greenhouse was deducted from the mass flow rate using the following Relation (2) described above in Section 2.2

Vent Configurations Used for Simulations
Three roof-opening and four side wall opening configurations of vent openings were considered in this study ( Figure 3): Open Journal of Fluid Dynamics

Results
In a first step, we used the developing model to investigate the effects of crop rows orientation, the effects of anti-insect screen (used in the vent openings) and the effects of side vents opening(parallel and perpendicular to the incoming vent) on the flow rate, airflow, and temperature patterns in a three-span greenhouse. The effect of the combination of roof openings with side openings in a screening greenhouse was carefully investigated during this step. In order to examine the influence of these parameters on greenhouse microclimate the configuration (C1) has been used with and without roof openings.
The second step was to represented solar radiation, result temperature, result humidity, the total flux at a different element of the house and to evaluate the climate resulted in the greenhouse (in terms of thermal, hydric and dynamic fields), and to study the performance of the aeration system for all cases studied.

Effect of Insect Screens, Roof Openings and Crops Rows Orientation
Within most crop ecosystems, some pests have no effective biological control agents. However, the anti-insect screen is the one commonly used by growers in Mediterranean countries to limit the entrance of the white fly and aphids species.  [19]. The placement of insect proof on the ventilation openings reduce the main air velocity in the greenhouse between 86% and 56%

A. Senhaji et al. Open Journal of Fluid Dynamics
(with and without roof openings respectively) and increased the main humidity by 11%.   Compared with lateral ventilation only, roof ventilation has improved the climate in terms of air exchange rate and climate uniformity, because the air inside is better mixed than in the case of lateral ventilation only.
The influence of the orientation of the crop rows is also very important; a summary of this influence on the efficiency of ventilation in the greenhouse is presented in Table 3, for the greenhouse studied without roof opening.
The crop rows orientation perpendicular to the prevailing air circulation reduces the greenhouse ventilation rate by 24%, the air velocity by 25% and increase the average temperature by 1˚C. However, the combination of crop rows perpendicular and insect proofs effects reduce the greenhouse ventilation rate by 72%, the velocity by 56% and increase the average temperature by 6˚C.   Table 4 shows the total average heat flux in the roof, sidewall, outside ground, inside ground and soil under crop. It is similar for each case and represents a high value on the roof compared to the other elements, which exhibits the highest radiative flux. A positive value indicates the flux exited roof wall to inside air.

Radiation Output
The other interior surface of the greenhouse structure has low values, explaining the solar radiation flux received by them. Similar results were obtained by an experimental study of Majdoubi et al. [21].

Inside Climate
The convective flux at the soil surface and at the cover is the key parameter to characterize the conditions of the greenhouse microclimate that is influenced by the ventilation rate.

Effect of Vent Arrangements on Airflow and Temperature Patterns
Velocity profile and corresponding temperature for each configuration, at crops level, are shown in Figure 9(a) and Figure 9(b) respectively. The first observation is that the velocity (Figure 9(a)), in upstream greenhouse, is lower for con-

Dynamic, Thermal and Hydric Field at Middle Plan
The dynamic field was represented in Figure 11.

Ventilation Efficiency
The efficiency of the ventilation was considered by reducing the temperature difference T i -T 0 between inside and outside and by reducing normalized velocity inside the greenhouse. The homogeneity of the temperature, humidity and velocity distribution has been evaluated by reducing the standard deviation , also the uniformity is calculated by the formula given below. A summary of the main results for the five configurations is presented in Table 6.
The standard deviation of a specified variable on a surface is computed using the mathematical expression below [26]: ( )  (11) where x is the cell value of the selected variables at each facet and x 0 is the mean value of x.
The uniformity index represents how a specified field variable varies over a surface.
The area-weighted uniformity index ( γ ) of a specified field variable x is calculated using the following equation [26]:  (12) where I is the facet index of a surface with n facet, and a x is the average value of the field variable over the surface.
The best configuration was thus a compromise between a high ventilation rate and good uniformity of the climatic conditions inside greenhouse especially of the crop.

Conclusions
The influence of vent arrangement on windward ventilation of a multi spans greenhouse was numerically investigated using computational fluid dynamics code. The numerical model integrates solar and atmospheric radiation by solving the RTE, that is to say, instead of setting thermal condition (specific wall temperatures or heat fluxes) at the physical boundaries of the greenhouse itself, the effects of solar and thermal radiation are taken into account by setting radiative conditions at the limits of the calculation domain. These results showed that the presence of a crop would act as a barrier to the flow of ventilation air, the crop row orientation relative to the sidewall vents have strong influences on ventilation airflow. The use of an anti-aphid screen reduces the mean air velocity inside the greenhouse and increased the average temperature by 5℃ compared to the values for a greenhouse without screen.
For five configurations of ventilation openings were investigated, the ventilation rates, the airflow, humidity and temperatures patterns are evaluated. These results indicate that the highest ventilation rates are not always the best criterion for evaluating the performance of different ventilation systems in greenhouses. The best criteria are the air velocities in the crop, the temperature difference between inside and outside, the absolute humidity (HA), homogeneity and uniformity of the climate parameters. For the configurations studied in this work, the above criteria show that the best configuration is to place the west side opening at the bottom, the eastern side opening at the top and to close the medium roof opening.