Computational Fluid Dynamics

2. Literature Review

Computational Fluid Dynamics is the art and science of analysing and simulating systems in which a fluid flow is of vital interest and in which heat and mass transfer and chemical reaction may take place. Its advantages over conventional experimental studies are substantial reductions in lead times and development cost, availability to study systems where experiments are not possible and ease of performing a large range of parametric studies for optimisation.

2.1 Ventilation in Greenhouse

Mistriotis et al. (1997a) performed systematic analysis of natural ventilation process in greenhouses for no-wind and low-wind-speed conditions. Natural ventilation is governed by pressure differences created at ventilating openings mostly either by the combination of wind, temperature differences or both. They found experimentally that winds stronger than 2 m s-1 was dominating factor for ventilation process, most of author highlighted this phenomenon (Bot, 1983; Papadakis et al., 1996).

Munoz et al. (1999) cited the reduction of temperature in the greenhouses located in warm regions is basis to obtain good control of climate inside. Natural ventilation is the most commonly used as cooling method, because it permits reduction of temperature for cultivation and the internal CO2 deficit, and it provides better environmental working condition at minimal costs.

Oliveira and Younis (2000) concluded that broadly used theory of two-dimensional flow field in greenhouse at mid-span give rise to overestimation of suction pressures on roof of greenhouse as well and on leeward wall (Fig 2.1).

Figure 2.1: Streamlines around the building to showing local separation step: case (a) 2-D predictions done by model and case (b) 2-D predictions done by RSM-SSG model

Baptista et al. (1999) concluded that in natural ventilation, location and type of the greenhouse, location of ventilator openings, and climatic characteristics (wind velocity, wind direction and temperature) influence ventilation rates. The main advantage of natural ventilation is being its low cost of operation. Screening materials are usually very thin and porous that allows solar radiation and outside winds to enter, but this causes alteration in microclimate in response to unexpected changes of ambient conditions.

Campen and Bot (2002), studied the ventilation of a Spanish ‘parral’ greenhouse using three-dimensional computational fluid dynamics (CFD). Investigation of two types of roof openings has been made, the rollup window configuration and another was flap window configuration at ventilation opening. They found that wind speed linked linearly with ventilation rate for both configurations under study without the buoyancy effect. Fig. 2.2, representing numerical ventilation rate was correlated linearly to the given wind speed for both configurations of ventilation openings.

Mistriotis and Briassoulis (2002) numerically predicated external aerodynamic coefficients are only weakly influenced by the position or the size of the openings. The singular behaviour observed at the opening vent is a result of the high pressure gradient along the openings as shown in Fig. 2.3a, which forces the airflow through the narrow openings at high speed (Fig. 2.3b). This type of flow gives rise to high turbulence along the opening vent, which induces strong negative pressure values.

Figure 2.3: Contour presentation of the air flow characteristics for the case A-2 (symmetrical openings with opening ratio 6.7%): (a) pressure distribution (pressure is normalised by the dynamic wind pressure at 4 m height); (b) air velocity component, vx (m s−1), parallel to the wind direction. Negative values of vx indicate circulating air flow

Effect of different ventilation configuration was studied by Bartzanas et al. (2004). They found ventilation rate in limits of 10 to 58 air exchanges per hour for different opening configuration under study. Fig. 2.4a pictures the temperature profile for side opening; due to low air velocity near floor of greenhouse there get high air temperatures in this region which is 4 oC higher than outside air temperature. Air temperature profile for a tunnel greenhouse with ventilation of roof opening (Fig. 2.4b) reached high values, at leeward wall temperature was 6 oC higher than outside air temperature.

Dayan et al. (2004)

Dayan et al. (2004) presented a simple model of ventilation in a commercial greenhouse with rose-growing in it (Fig. 2.5). The three-segment model which contains 10 equations of which three for dry air mass balance, and next three for water vapor (humidity) mass balance moreover, other three consisting of energy balances (one type for each segment), and an last energy balance equation for the plants. The developed model was flexible enough to be updated and calibrated for various conditions in greenhouse including structures of greenhouse, it take online measurements of microclimate parameters such transpiration, air temperatures, leaf temperature, and humidity at many locations.

Scheme and nomenclature of reduced model of a rose-growing greenhouse

Molina-Aiz et al. (2004), studied the numerically effect of wind speed on performance of ventilation for Almer´ıa-type greenhouse. Results show that cold outside air entered the through the two side vent openings of greenhouse, both at windward and leeward (Fig. 2.6). They studied in-depth air velocity parameter values inside the greenhouse, found highest values at the ventilation opening and lowest values in the middle of greenhouse.

Figure 2.6: Vertical section of air velocity vector in the middle of the empty greenhouse when a southwest flowing wind of 5.0ms−1 speed: (a) experimentally measured values by using hot bulb anemometer and (b) simulated obtained airflow vector field (measurements done on 13 July 2003 at 20:15 h)

Jime´nez-Hornero et al. (2005) studied the natural ventilation in greenhouses using a two-dimensional lattice Bathnagar, Groos and Krook (BGK) model. Fig. 2.7 shows the airflows obtained with the lattice model for different rolling ventilator configurations in the cross-section of a multispan greenhouse with tunnel structure. The results from these models are similar to those simulated with the computational fluid dynamics (CFD) technique but with important drawback of this model is the empirically setting of the Smagorinsky constant that requires an additional study about the characteristics of the flow and greenhouse shape.

Figure 2.7: Natural airflows obtained with the lattice model for different rolling ventilator configurations in the cross-section of a multispan greenhouse with tunnel structure

Katsoula et al. (2006) experimentally studies the effect of vent type namely at side, roof or at both and anti-aphid insect screen on the ventilation rate of polyethylene covered greenhouse. Results showed roof and side vent combination was most effective compared to side vent only (reduction of 46 % ventilation) and roof vent only (reduction of 71 % ventilation).

Roy and Boulard (2005) investigated of wind-induced ventilation; establish that the underestimated values of air velocity was given by the standard model, also studied RNG model, which showed improvement in the homogeneity of the greenhouse climatic parameters. Brugger et al. (2005) invested parral-type greenhouse found deviation in different ventilation rates predicated by the standard model and two-scale model.

Ould Khaoua et al. (2006) studied airflow profile and temperature patterns in a compartmentalised glasshouse by using variables as wind speed and roof vent opening (Fig. 2.8).

Figure 2.8: Simulated contours of case (a) the air velocities normalised in reference to wind speed (u=Uh) and case (b) the respective temperature difference (Ti-To; K) across multi-span greenhouse modified ventilation at roof (two roof ventilation) (Wind speed for a 1 ms-1)

Teitel et al. (2008) experimented to naturally ventilated mono-span greenhouse and installed continuous screens on side vents to determine the factors such as ventilation rate, temperature, air velocity distributions, humidity and also studied the energy partitioning of the incoming radiation to greenhouse. Concluded, ventilation rate increases linearly with wind speed. Ratio of temperature and humidity were greater near the roof while compared to near the crop region, but the air velocity found higher values near the crop region when compared to near the roof.

2.2 Microclimate in Greenhouses

Though measuring the gradients of environmental or climatic properties has been of interest since a long ago, yet a few research literatures are available regarding it. Sase et al. (1984) studied the airflow and temperature pattern in a naturally ventilated single-span greenhouse model installed in a wind tunnel and concluded that the temperature distribution in a greenhouse was largely determined by the airflow pattern. Wang and Deltour (1996) evaluated the full-scale distribution of natural ventilation induced airflow in venlo-type greenhouse openings with ultrasonic anemometer. Boulard et al. (1999) carried out experiments in a small-scale model greenhouse without plants and found that most of the temperatures drop between the soil and the roof occurred within small distances above the floor and below the roof. He observed strong temperature gradients in those thin layers, while the temperature inside the whole cavity was constant, with a uniform and steady temperature distribution. Van et al. (1998) studied the influence of temperature gradients on leaf-temperature and plant processes. They confirmed that to attain an improved bio-control and to have more insight into response of the plants to microenvironment must be well aware of temperature distribution surrounding it.

Several studies have been performed related to the calculation of greenhouse microclimate prediction and evaluation. A numerous mathematical as well as empirical models have been developed using various climatic parameters including temperature, humidity, pressure differences, wind velocity and solar radiation. Boulard et al. (1997) estimated pressure field and airflow patterns at the ventilation openings of greenhouse. Wang and Boulard (2000) worked in Mediterranean climate for predicting the microclimate in a naturally ventilated structure of plastic house. Demrati et al. (2001) studied natural ventilation process and microclimatic parameters in large-scale banana greenhouse.

Bartzanas et al. (2002) investigated the effect of an insect proof screen applied at ventilation opening on microclimatic parameter of greenhouse including airflow and temperature profiles by using CFD programs. Numerical simulation suggested screen incorporation resulted in less air velocity in the greenhouse most in crop area as well seen airflow rate reduction to 50%, which resulting in a significant temperature gain. Fig. 2.9a shows air velocity vector inside the greenhouse, a strong flow near floor of the greenhouse was seen and a slow moving circulation seen at roof zone as well observed a counter flowing current corresponding to outside wind. Greenhouse with screen has a similar airflow pattern but with lower values for velocity while improved air mixing in the internal space (Fig. 2.9b).

Figure 2.9: Velocity vector in vertical plan in middle of greenhouse with a) natural opening b) cladding material

Fatnassi et al. (2003) simulated the climatic condition in full-scale greenhouse. Air flow in crop cover as well in insect-proof nets were calculated by porous medium approach. They established air flows from windward end, taking spiral trajectory through crop, rising over crop cover and escaping though leeward part of roof vent (Fig. 2.10). Significant increase in temperature and humidity was fount by introduction of anti-Thrips and anti-Aphids.

Profile representing air flow pattern in the middle of the simulated greenhouse below a roof opening.

Soni et al. (2003) investigated greenhouses with different insect screens and natural ventilation arrangement in tropics. They observed a heat envelope of length 15 cm from outside screen at ventilation opening, tossed a new concept of heat envelope in terms of air temperature gradient, it shows how wide and strong be the temperature gradient field of influence near and around the insect screen, which offers resistance to mass, energy and momentum transportation in greenhouse. Study found less porous screen (-0.88 oC and +5.7 oC) has more effect on inward and outward temperature gradients were then more porous screens (-0.44 oC and +3 oC).

Fatnassi et al. (2006) studied the effect on wind and climatic condition of greenhouse fitted with insect screen. A trend of temperature (Fig. 2.11) and humidity (Fig. 2.12) rise from windward to leeward side of greenhouse has been noted. When compared the unscreened vents and anti-Bemisia net, air humidity and temperature rises twice and three times for anti-Thrips nets.

Figure 2.11: Air temperature inside the simulated greenhouse with configuration of open windward and side openings as well anti-Bemisia nets on the vents. Temperature in K. (inside the greenhouse horizontal line represents crop height)

Figure 2.12: Air humidity field in g kg−1 inside the simulated greenhouse with configuration of open windward and side openings as well with anti-Bemisia insect screen on the vents (inside the greenhouse horizontal line represents crop height)

Kittas and Bartzanas (2007), analysed ventilation roll-up type configuration, found highest air velocity inside the greenhouse area was reached in near the ground surface while observed the lowest values near the roof of greenhouse. Pivoting door type configuration results are exactly opposite pattern, with maximum air velocities near to the greenhouse roof, as air temperature and humidity were reduced first during the process of dehumidification (Fig. 2.13a-b).

Figure 2.13: Velocity vectors in a tunnel greenhouse with side opening with case (a) roll-up type openings and case (b) pivoting door type openings

Bournet et al., (2007), stated that climate distribution and variation in greenhouse was occurred due to mass, momentum and energy movement across greenhouse through ventilation openings. Observation for configurations C3, revels the highest normalised velocity reaches in the section of windward span and then reduces in the rest of three spans, clearly resulting in improper air mixing in this area (Fig. 2.14). The corresponding temperature profile clearly illustrates that configuration A3 and B3 are more efficient compared to configuration C3.

Figure 2.14: Temperature distribution T-Ta in a simulated glasshouse with a vertical side vent at top location. Case A3: symmetric roof vents, case B3: windward roof vents and case C3: leeward roof vents