CFD simulations for examining natural ventilation in the learning spaces of an educational building with courtyards in Madurai

Abstract

This paper attempts to investigate the potential of courtyards in optimizing natural ventilation and improving comfort levels in the learning spaces of a naturally ventilated educational institution with courtyards in the warm-humid climatic region of Madurai. Field measurements and experimental studies were carried out to predict the indoor and outdoor environmental conditions. The numerical study was carried out using computational fluid dynamics-based simulations using Ansys Fluent as the solver. The main aim of the simulation is to understand the airflow pattern and air velocity fields inside the classrooms surrounding the courtyards for different wind directions. The computational fluid dynamics results were validated by comparing it with the experimental results obtained in the current study and numerical results from other studies. The major findings of the current study suggest that courtyards with an aspect ratio of 1:2, orientations of openings at an angle of 0–20° to the predominant wind directions and the overall percentage of openings between 15 and 30% in buildings in Madurai region can enhance natural ventilation and thus improve thermal comfort of the occupants.

Practical application : Naturally ventilated buildings in warm-humid climates have difficulty in providing thermal comfort to the occupants. CFD tools have been used to predict the ventilation performance of a naturally ventilated educational building with courtyards. The CFD results were helpful in identifying the implication of building design on the indoor air flow pattern. The recommendations given in this paper are applicable to any building type which relies on natural ventilation for thermal comfort provided they have similar building configurations, boundary conditions and weather conditions. The study is intended to help architects and building designers in the effective design of naturally ventilated buildings with respect to its climatic conditions.

Keywords

Natural ventilation learning spaces warm-humid climate courtyards thermal comfort CFD

Introduction

India is a country with diverse geographic conditions and distinct climate zones. Cities have been tremendously increasing in size due to urbanisation. The growing needs for thermal comfort has led to the increased use of electricity generated by fossil fuels for the operation of heating, cooling and ventilation systems in buildings, which makes it responsible for 30%–40% of the global energy demand and 40%–50% of the world carbon emissions.¹

In India, the majority of schools and universities are mechanically ventilated through the usage of fans alone. Unlike natural ventilation, fans only create internal air movement, but they do not bring in fresh air.² Before the invention of mechanical ventilation, buildings were designed to utilise the maximum benefits of fresh air and maintain comfort. Natural ventilation is important because it can enhance fresh air movement without fans. Natural cooling strategies can greatly minimize energy costs and negative environmental impacts with improved thermal satisfaction among the occupants.^3,4 Apart from cooling of indoor spaces, natural ventilation also plays a vital role in maintaining the indoor air quality. It has been observed that the high-temperature levels and low air supply rates are the main reasons for the increase in carbon dioxide levels in classrooms.⁵ High levels of CO₂ can have adverse effects on health and learning.⁶ Thus, the learning environment requires appropriate planning strategies to provide better indoor air quality through proper ventilation.

For the effective design of natural ventilation in buildings, it is imperative to understand the air flow and heat transfer characteristics in and around buildings. The air flow pattern relies on fluctuations in external and internal climatic conditions, the building geometry and existing site conditions.⁷ Many researchers across the world have investigated the air flow pattern, using computational tools and models and proved that ventilation flow rates are influential factors to ensure fresh air supply.^1,8–11 Thus, computational fluid dynamics (CFD) is considered as an effective tool that provides a computer-based modelling to predict the performance of natural ventilation and microclimatic characteristics of the wind. This study has therefore used CFD to analyse the behaviour of wind in relation to the building design. The numerical study was carried out to evaluate the airing performance of a naturally ventilated educational building with courtyards by assuming the wind movement at a very low speed of 1 km/h from two predominant wind directions on a hot day in the overheated period to analyse the implication of courtyards and window orientations pertaining to the local microclimatic conditions. The CFD results were validated by comparing it with the experimental results obtained in the current study and numerical results from other studies. The results were helpful in identifying the implication of building design on natural ventilation. The study is very important because finally recommendations on appropriate proportion of courtyards, orientation and percentage of openings to enhance natural ventilation have been suggested to improve the thermal comfort of occupants in buildings located in the warm-humid climatic region of Madurai, thereby reducing the dependency on mechanical ventilation systems for cooling. The innovative aspect of this study is that both the exterior and interior wind environment has been thoroughly investigated to understand the implication of building design and orientation of openings on the airing performance.

Literature review

An extensive literature study shows that courtyards have attracted the present time designers as an effective passive design solution to improve the thermal and ventilation performance of various types of naturally ventilated buildings in hot climatic regions, but its optimum proportions are still under investigation to reach the desired ratios, in terms of energy efficiency and thermal comfort. Rajapaksha et al. conducted an experimental and numerical investigation on courtyards for passive cooling in warm-humid climate and showed that there was a notable reduction in the indoor air temperature during the daytime of the overheating period.¹¹ The study proved that courtyards can act as efficient modifiers of the indoor climate in warm-humid climatic regions.

Ernest and Ford confirmed that multiple courtyards promoted convective cooling, as well as reduced the cooling energy required.¹² Optimum courtyard proportions can notably reduce the solar heat gain and cooling loads in buildings in hot areas. Courtyards initiate proper air circulation and protect the inhabitants from the unfavourable weather conditions like stormy winds, hot and dusty winds in hot areas.^13–17 Nielsen discovered that the courtyards perform thermally better when the width of the courtyards does not exceed three times the height of the building.¹⁸ They also added that the courtyard proportions can have a remarkable effect on reducing the building solar gain and cooling loads. Salameh and Taleb conducted a study to investigate the potentials of different proportions of courtyards as a passive design solution in school buildings in UAE.¹⁹ In their study, it was proved that courtyards with width 2 × performed better than spaces with no courtyards and width 3×. They also concluded that the air change rates were faster, and CO₂ concentration was lesser in the case of the 2 × width courtyard school.

Research methodology

The research carries an experimental and numerical study to investigate the influence of courtyards and window orientations on natural wind movement and thermal comfort inside the learning spaces of an isolated educational building in Madurai. The numerical investigation has been carried out for two different wind directions observed on 25 March 2016 using CFD as the simulation tool. The numerical results are validated with the experimental results of air flow rates and indoor air quality inside the learning spaces of the base case building. The data required for the numerical simulation were obtained from the field measurements of the outdoor and indoor weather conditions recorded during the overheated period (March–September) 2016, using WATCHDOG 2700 series outdoor weather station, HTC easy-log Temp/RH data loggers. Figure 1 shows the temperature and RH graph generated by the outdoor weather station. The indoor air velocities were derived Lutron vane-type anemometer model no. AM-4201.The indoor air quality (CO₂ levels) was measured using IAQ-Calc hand device. The specifications and accuracy of the instruments used for the experimental study are given in Table 1.

Figure 1.

Recorded data of outdoor temperature and relative humidity ranges in March–April 2016.

Table 1.

Detailed specifications of the experimental study instruments.

Name of instrument and model number	Sensors	Specifications	Accuracy
WATCHDOG 2700 series outdoor weather station	Temp	−32° to 100℃	±0.6℃
	RH	10% to 100%	±3% at 20 and 25℃
	Wind direction	1° increments	±4°
	Wind speed	0.1 to 200 mph	±2 mph
	Dew point	−73° to 60℃	±2℃
	Solar radiation	0 to 1500 W/m²	±5%
	Rain fall	0.01″ (0.25 mm) resolution	±2% at < 2 in (5 cm)
HTC easy-log Temp/ RH data loggers	Temperature	−40° to 70℃	±1℃
HTC easy-log Temp/ RH data loggers	Relative humidity	0 to 100%RH	±3–5%
Anemometer model no. AM-4201	Air velocity	0.4 to 30.0 m/s, 1.4 to 108.0 km/h, 80 to 5910 ft/min., 0.8 to 58.3 knots	±2%
TSI 7525 IAQ-Calc CO₂, temp, Relative Humidity Meter/Datalogger	CO₂	0 to 5000 ppm	1 ± 3.0% of reading or ± 50 ppm
	Temperature	0 to 60℃	±0.6℃
	Relative humidity	5% to 95%	0.1%

Case study building description

The case study building is an isolated building built in 1994 exclusively for architectural education, within the densely vegetated campus of “Thiagarajar College of Engineering”. The building is rectangular in shape with flat roof and oriented towards the east–west axis. The whole building relies on natural ventilation and fans for mechanical ventilation. Figure 2(a) and (b) represents the ground and first floor plans, respectively. The overall dimensions (Length – L, Breadth – B and Height – H) of the building are 72 m, 44 m and 10 m, respectively. There are two courtyards in the building (Figure 2). Courtyard 1 is smaller in size of 15.5 m × 12 m × 12 m (L × B × H) and located on the north-eastern side. Courtyard 2 is larger in size with dimensions 24 m × 14.9 m × 12 m (L × B × H). The aspect ratio of Courtyard 1 is 1:1 and Courtyard 2 is 1:2. The building is occupied from Monday to Friday and regular working hours are from 9:00 to 17:00.

Figure 2.

Floor plans of the institutional building.

Figure 3.

Pressure variation over the institutional building due to airflow from WNW direction.

Figure 4.

Predicted velocity magnitude contours for wind flow from WNW direction.

Experimental study methods

The learning spaces in the case study building were monitored for five consecutive days (Monday to Friday) from 25 March 2016 to 29March 2016. The overview of the learning spaces considered for the study is given in Table 2. The temperature, relative humidity data loggers were fixed inside the classrooms and recorded for five days continuously, whereas the air velocity and indoor air quality measurements in the respective spaces were taken by the hand tools specified in the ‘Research methodology’ section during the morning and afternoon hours between 9.00 am–1.00 pm and 2.30 pm–4.30 pm, respectively, at a periodic interval of 30 min. The anemometer was positioned at a height of 1 m from the base of the room and 1 m away from window openings in the learning spaces as shown in Figure 7, and the air velocity measurements were noted down. The fans were switched off and windows were left open while measuring the indoor air velocity. The air change rates per hour (ACH) and ventilation rates (V_R) of the case study building with NV classrooms were also calculated using the equations given below.²⁰

ACH = 3600 Q / V

(1)

where Q is the volumetric flow rate in m³/s and V is the volume of the room in m³

V_{R =} (ACH \times OD \times H) / 60

(2)

where OD is the occupant density and H is the height of room in feet or metres.

Table 2.

Overview of classroom details in the case study building.

Name of classroom	Occupant Capacity	Ventilation strategy	Room size L × B × H (m)	Window size Width × Height (m)	Area of openings (m²)
Studio type I	40	NV	18 m × 11 × 4	1.3 × 1.95	20
Studio type II	40	NV	12 × 8 × 4	1.3 × 1.95	13
Studio type III	40	NV	15.8 × 5.9 × 4	1.3 × 1.95	21.9
Studio type IV	40	NV	15.8 × 9.6 × 4	1.3 × 1.95	26.5
Lecture hall I	40	NV	8.9 × 5.9 × 4	1.3 × 1.95	17.4
Lecture hall II	40	NV	8.9 × 5.9 × 4	1.3 × 1.95	14.5
Lecture hall III	80	NV	12.8 × 6.6 × 4	1.3 × 1.95	18.6

NV: naturally ventilated.

Experimental study results and discussions

The air velocity measurements derived from the experimental study ranged from an average of 0–0.6 m/s. The air speed at points 2, 4, 6 and 8 were higher ranging from 0.4 to 0.7 m/s than points 1, 3, 5 and 7 of the room with air speeds below 0.3 m/s for both experimental and CFD results. An increase in air movement from 0.0 m/s to 0.7 m/s in warm-humid climates can substantially increase the Predicted Percentage of Satisfaction from 44% to 100% at temperatures around 28.69℃.²¹ The ACH was calculated using equation (1) and ventilation rates inside the classrooms were calculated using equation (1) assuming the air velocity to be 0.1 m/s. The ACH ranges from 2 to 5 and centilation rates range from 4 to 12 l/s for the NV classrooms as shown in Table 3. The ASHRAE standard 62.1-2007 recommends ACH within the ranges 3–5 and ventilation rates ranging from 8 to 10 l/s. From the current study, it shows that the spaces with lower ACH have higher ventilation rates. This is because the volume of space is larger when compared to the occupant density. The results of the study show that the percentage of openings should be between 15 and 30% to obtain the optimum ranges of ACH (shown in Table 3).

Table 3.

Comparative analysis of indoor variables, ACPH and ventilation rates (V_R) in the base case building.

Name of classroom	Level	Occupant capacity	Exposed wall area in (m²)	Area of windows exposed outside. (m²)	WWR %	ACH at 0.1 m/s (Fans OFF)	Ventilation rate V_R at 0.1 m/s Fans OFF (l/s)
Studio type I	GF	40	64.8	10.8	16.7	2.4	12.4
Studio type II	FF	40	113.8	18.0	15.8	4.6	10.9
Studio type III	FF	40	42.5	7.2	16.9	3.7	9.0
Studio type IV	FF	40	164.8	33.8	20.5	3.5	13.7
Lecture hall I	FF	40	32.0	3.6	11.2	4.7	6.2
Lecture hall II	FF	40	21.2	5.4	25	4	5.3
Lecture hall III	GF	80	46.0	9.0	19.5	3.8	4.0

Indoor air quality measurement results and discussions

The CO₂ levels were measured with fans ‘ON’ and fans ‘OFF’ in the learning spaces where the occupancy levels ranged from 30 to 40 using IAQ-Calc hand device. The average carbon dioxide (CO₂) concentrations in the studios and lecture halls varied between 600 and 1500 ppm. Observations showed that in the morning hours at 9.00–9.30 am when the CO₂ level in the outdoor air was 430 ppm, the indoor CO₂ levels were higher by 100–150 ppm in all the learning spaces both in GF and FF. But within 15 min, there was an increase up to 1000 ppm inside the learning spaces. In the afternoon hours at 2.30–4.30 pm, the difference in CO₂ levels between outdoor and indoor air was more than 1000 ppm. It was also found that when the fans were ‘ON’, the CO₂ levels decreased by 100–150 ppm due to movement of air. This shows that usage of fans becomes mandatory in these spaces due to insufficient air movement. Fans offer an important adaptive opportunity for users at high indoor temperatures usually experienced in India.²² At places with weaker circulation of air, it was noticed that CO₂ levels were higher ranging from 1200 ppm to 1500 ppm in the absence of mechanical ventilation through fans. The CO₂ levels were greater falling within the ranges of 1000–1500 ppm during teaching hours with full occupancy than the non-teaching hours with ranges between 500 ppm to 800 ppm. During non-teaching hours, the occupant density was lesser, and thus the results indicate that the indoor environment was being mostly affected by indoor sources than the outdoor air. ISHRAE recommends 500–1200 ppm of CO₂ in indoors.²³ Higher levels of CO₂ may be associated with health problems like dizziness, head ache or tiredness among occupants, and thus should be controlled.

CFD simulation methodology

The numerical simulation on the discretized domain is carried out using ANSYS Fluent. Standard k-Epsilon turbulence model is used to predict the turbulence and viscous effects.²⁴ A velocity-specified domain inlet is used to indicate the air flow direction and speed. The domain outlet is set with ‘pressure outlet’ boundary condition. Based on the observations from the statistical data of wind for Madurai region,²⁵ it was found that the predominant directions of wind during the overheated period were from WNW and NE directions. Thus, from the recorded outdoor weather data at the site level, two different wind directions (WNW and NE) on 25 March 2016 flowing at a very low speed of 1 km/h with reference temperatures of 34.8℃ at 12.30 pm and 37.1℃ at 3.00 pm, respectively (represented as bold numerals in Table 4), have been taken to investigate the architectural design intervention on natural ventilation in the base case building. Hence, flow inlet is specified with same air speed of 1 km/h for both the cases. The predicted variables were used to calculate the turbulence dissipation rate profiles at the inlet of the simulation domain. The reference pressure is taken as 1.01325 bar. Since the working hours of the institution are from 9.00 am to 5.00 pm, the variables considered for the study were taken within the working hours. The solver solves Pressure Based Navier Stokes Equations (PBNS) of continuity and momentum along with turbulence.⁹ The numerical simulation was conducted using a workstation with xenon processor and 32 GB RAM. Each simulation took approximately 72 h to complete the process.

Table 4.

Recorded outdoor environmental data for 25 March 2016.

Date	Solar rad (SRD) (W/m²)	Relative humidity (RH) (%)	Temp. (TMP) (℃)	Wind direction (WND) (deg)	Wind gust (WNG) (km/h)	Wind speed (WNS) (km/h)
25 March 2016, 8:30	387	63.6	29.4	282	6	1
25 March 2016, 9:00	481	60.6	30.1	216	6	0
25 March 2016, 9:30	571	57.5	30.4	217	6	1
25 March 2016, 10:00	728	53.2	31.5	225	6	1
25 March 2016, 10:30	832	47.6	32.3	204	6	1
25 March 2016, 11:00	901	43.2	33.2	207	8	3
25 March 2016, 11:30	968	36.6	33.7	360	9	3
25 March 2016, 12:00	1003	35	34.4	186	14	4
25 March 2016, 12:30	1010	33.9	34.8	280	14	1
25 March 2016, 13:00	999	28.7	35.8	259	11	3
25 March 2016, 13:30	960	27.6	36.1	327	11	3
25 March 2016, 14:00	908	27.7	36.6	288	11	3
25 March 2016, 14:30	850	28.3	36.6	326	17	3
25 March 2016, 15:00	743	27.5	37.1	69	11	1
25 March 2016, 15:30	535	26	37.5	307	12	3
25 March 2016, 16:00	291	26	37.4	312	11	3
25 March 2016, 16:30	157	26	37.3	259	9	1
25 March 2016, 17:00	97	26.2	36.8	334	12	3
25 March 2016, 17:30	64	26.8	35.3	279	19	8
25 March 2016, 18:00	39	28.1	33.9	343	16	6
25 March 2016, 18:30	10	32.9	32.5	252	17	6
25 March 2016, 19:00	0	35	31.2	309	11	1

Source: WATCHDOG 2700 series outdoor weather station.

Validation of the CFD model

The numerical study results were validated with the measurements taken from the field investigation in the current study. Values of' ‘Q' (volumetric air flow rate) derived from the experimental method are decomposed into respective inlet velocities, and they are given as velocity inlet in the CFD domain that matches the domain size. Values of ‘ACH’ are also available from the experiments of Subashini and Thirumaran in Table 3. For the indoor heat source model, an average human body cross-section is given between 0.12 and 0.4 m².²⁶ Mannequins of such cross-section are modelled inside the domain. The average temperature of a human body is 37℃.²⁷ In the solver (Ansys-Fluent), energy equations are invoked, and the surface temperatures of the mannequins are modelled with a temperature of 37℃. The comparison of air velocities predicted by numerical simulation and experimental study showed an error of less than 10% (Figure 8). Further a regression analysis between the predicted and measured air velocity levels also showed a stronger correlation of R²= 0.967 (Figure 9). Thus, the CFD results were found to be in good agreement with the experimental study results.

Figure 5.

(a) Static pressure distributions over the building due to airflow from NE direction; (b) Wake formation behind the building.

Figure 6.

Predicted velocity magnitude contours in GF and FF level for wind flow from NE direction.

Figure 7.

Position of air flow measurement points in Studio type II.

Figure 8.

Comparison between the experimental and the measured air speed at different points.

Figure 9.

Regression between the CFD and experimental study results.

The numerical and experimental study carried by Rajapaksha et al. is used to validate the prediction of wind velocities using CFD.¹¹ The study proved courtyards as an effective passive design element to promote mass heat exchange and reduce of indoor thermal conditions in hot climate zones. The aspect ratios of the courtyards in the present study are 1:1 and 1:2. A numerical study using CFD for similar conditions and aspect ratios, by Almhafdy et al. shows that the courtyard with aspect ratio 1:2 performs thermally better than the one with aspect ratio 1:1.¹⁴ This proves that the results of both the cases are in fairly good agreement. The numerical investigation performed by Derakhshan and Shaker is used to validate the air flow rates based on wind angles. The air flow rates derived from the experimental study ranged from 0.4 m³/s to 1.3 m³/s for openings with wind angles between 0 and 45°. It was found that maximum volume of air flow rate occurred for wind angles from 0 to 15°, and the volume of air flow rate decreased drastically for wind angles greater than 20°.²⁸ Thus, the experimental results were also found to be in good agreement with the results of Derakhshan and Shaker.

Computational domain and mesh generation

The three-dimensional (3D) model of the building was generated using Autodesk Rivet software and then converted into IGES file to be used for CFD analysis. The building model is meshed with unstructured triangular elements. A growth rate of 1.9 million with skewness less than 60 has been specified for the triangular mesh generation from building wall surfaces to the domain. Mesh refinement is given at appropriate locations of the building especially at flow entry zones like windows, doors, etc. The overall mesh count created for the present analysis is around 8 millions. Unstructured tetrahedral meshes of 7.4 million elements with Tet collapse < 0.1 are created around the building. The total number of nodes taken for the computations is around 1.7 millions. The numerical analysis is carried out in ‘opened window’ condition to simulate the natural ventilation. The convergence of the solutions is set to 1 × e-04.

Boundary conditions

A boundary domain with appropriate size is selected, and the fluid volume is discretized with tetra elements. The sides of the flow domain are mentioned with ‘symmetry’ boundary condition. The building surfaces and ground are given with ‘standard wall’ boundary conditions. The domain outlet is set with ‘pressure outlet’ boundary condition. The constant static pressure boundary condition was used at the outlet of the calculation region. A boundary domain of 560 m length, 244 m breadth and 60 m height has been selected.²⁴

CFD simulation results and discussions

The results of both simulations have been demonstrated in the following sections. The optimum directions for placement of windows in naturally ventilated buildings have been analysed and recommended at the end of this study. The influence of courtyard spaces and window orientations has also been analysed and discussed.

Simulation results for outdoor wind movement from west–north–west

The wind flow at a speed of 1 km/h was observed from west–north–west (WNW) direction at 280° on 25 March 2016 at 12.30 pm with a reference temperature of 34.8℃ (Table 4). The central axis of the building is tilted at an angle of 60° from the north. As the wind flow is from WNW, static pressure rise is found on the rear side of the building as shown in Figure 3. The numerical simulations revealed the wake formation in front of the building towards the north-east direction.

Figure 4(a) and (b) demonstrates the air flow pattern in the ground floor and first floor. The simulation results showed variations in the movement of wind in the ground and first floor. It can be visualized that low wind speed areas (low flow velocities under 0.2 m/s) appeared in learning spaces located on the eastern side of the building when the wind flow was from WNW. The CFD results showed notably strong air movement into the spaces in the north-western part of the building, as an effect of staggered plan and openings facing the west (see Figure 4).

Simulation results for outdoor wind movement from the north–east

Numerical simulation for predicting the flow patterns and ventilation is carried out on the same institutional building assuming the same domain and boundary conditions using ANSYS Fluent Solver. The wind flow direction at the speed of 1 km/h is maintained at an angle of strike (69°N) as observed from the recorded data given in Table 4. The reference temperature taken is 37.1℃. In this case, the pressure rise is noticed over the frontal surfaces (Figure 5(a)) and the wake is formed on the north-western side of the institutional building (Figure 5(b)).

Figure 6(a) represents the predicted velocity magnitude contours in the horizontal plane, and Figure 6(b) demonstrates the velocity vectors in 3D in ground floor when the wind flow is from NE (69°N). Figure 6(c) and (d) represents the velocity magnitude contours in 2D and velocity vectors in 3D at first-floor level. Figure 6(a) and (b) shows the high flow velocity vectors entering the building through the portico and reveals the presence of vortex in Courtyard 1 with an aspect ratio of 1:1. There also exists an increase in entrainment velocity through the corridors as the passage is narrow as seen in Figure 6(c). There are also places where weaker recirculation is formed both in ground floor and first floor.

Discussions of CFD results for both wind directions

From the simulation results for both wind directions, it can be deduced that the air circulation inside the building was mainly through the fenestrations and corridors. The study revealed that when the wind movement was from the west, high velocity wind entered the classrooms through the windows facing the west and north. On the other side, very low velocity contours of 0.1 m/s are noticed inside the studios and lecture halls on the eastern side of the building. When the wind flow is from the east, it is observed that there is proper circulation of air in almost all parts of the building (see Figure 6(a) and (c)). This suggests that openings in the eastern side and the northern side are preferred than the openings on the west for Madurai’s climatic condition. It is also noticed that the corridors acts as tunnels for movement of air and help in channelizing the wind across the building. The influence of courtyard in cooling the air can be well noticed in both the cases. The thermal comfort study conducted by Subhashini and Thirumaran shows that spaces around Courtyard 2 showed better thermal performance than spaces around Courtyard 1.²⁰ The reason for this was investigated with the help of CFD simulation results and perceived that in case of Courtyard 1, the longer side was facing the east and west, so there was direct solar radiation and perpendicular movement of hot air through the window openings, whereas Courtyard 2 was aligned in the east–west axis with the longer side facing the north and south with wind angle less than 45°. Thus, there was a reduction in conductive and convective heat transfer through walls and windows in the spaces surrounding Courtyard 2.

Conclusions

This paper presents the optimisation method of natural ventilation in educational buildings by analysing three design parameters: courtyard proportions, window orientations, percentage of openings. It is inferred from the results of the experimental and numerical study that 15–30% of openings on the exterior walls inclined at an angle of 0–20° to the pertaining wind direction should be given in order to achieve the desirable ACH and ventilation rates in naturally ventilated buildings in Madurai. The CFD simulation results clearly demonstrate the effectiveness of the courtyards in promoting cross-ventilation and circulating cooler air into the surrounding spaces for both wind directions considered in the study. Since the predominant wind directions in Madurai are from the west, WNW and north-eastern sides, it is advisable to give protected openings on the west through large horizontal overhangs, or corridors to prevent hot air movement into the learning spaces. In addition to promoting natural ventilation, courtyards also provide mutual shading, thereby reducing heat gain through walls. The results suggest that the optimum length of the courtyard should be less than 3 × and the longer axis of the courtyard should be oriented in the north–south direction to achieve better thermal comfort. The study proves that courtyards are crucial elements of design for fresh air supply in the naturally ventilated buildings in warm-humid climatic conditions.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Calautit

O’Connor

Sofotasiou

, et al. CFD simulation and optimisation of a low energy ventilation and cooling system. Computation 2015; 3: 128–149.

Kishore KN and Jain R. A bioclimatic analysis tool for investigation of the potential of passive cooling and heating strategies in a composite Indian climate. Build Environ 2017; 123: 469–493. doi: 10.1016/j.buildenv.2017.07.023.

Emmerich

Polidoro

Axley

. Impact of adaptive thermal comfort on climatic suitability of natural ventilation in office buildings. Energy Build 2011; 43: 2101–2107.

Wang

Zhao

Kuckelkorna

, et al. Classroom energy efficiency and air environment with displacement natural ventilation in a passive public school building. Energy Build 2014; 70: 258–270.

Mendell

Heath

. Do indoor pollutants and thermal conditions in schools influence student performance? A critical review of the literature. Indoor Air 2005; 15: 27–52.

Croome DJC, Awbi HB, Bakó-Biro ZS, Kochhar N and Williams M. Ventilation rates in schools. Build environ 2008; 43: 362–367.

Ghiabaklou

. Natural ventilation as a design strategy for energy saving. World Acad Sci Eng Technol 2010; 4: 260–265.

Jones

Kirby

. The performance of natural ventilation windcatchers in schools – a comparison between prediction and measurement. Int J Vent 2010; 9: 273–286.

Yang

Cheng

, et al. CFD simulations to examine natural ventilation of a work area in a public building. Int J Mech Aero Ind Mech Manuf Eng 2014; 8: 1186–1191.

10.

Essah

Yao

Short

. Assessing stack ventilation strategies in the continental climate of Beijing using CFD simulations. Int J Vent 2017; 16: 61–80.

11.

Rajapaksha

Nagai

Okumiya

. A ventilated courtyard as a passive cooling strategy in the warm humid tropics. Renew Energy 2003; 28: 1755–1778.

12.

Ernest

Ford

. The role of multiple-courtyards in the promotion of convective cooling. Arch Sci Rev 2012; 55: 241–249.

13.

Zamani

Taleghani

Hoseini

. Courtyards as solutions in green architecture to reduce environmental pollution. Energy Educ Sci Technol Part A: Energy Sci Res 2012; 30: 385–396.

14.

Almhafdy

Ibrahim

Ahmad

, et al. Thermal performance analysis of courtyards in a hot humid climate using computational fluid dynamics CFD method. Procedia – Soc Behav Sci 2015; 170: 474–483.

15.

Manioğlu

Oral

. Effect of courtyard shape factor on heating and cooling energy loads in hot-dry climatic zone. Energy Proc 2015; 78: 2100–2105.

16.

Soflaei

Shokouhian

Shemirani

. Investigation of Iranian traditional courtyard as passive cooling strategy (a field study on BS climate). Int J Sustain Built Environ 2016; 5: 99–113.

17.

Kolozali

. The role of the courtyard in the establishment of architecture in Cyprus. Architect Res 2016; 6: 123–130.

18.

Nielsen

. Stay cool: a design guide for the built environment in hot climates, London: James & James Ltd., 2002.

19.

Salameh M and Taleb H. Courtyard as passive design solution for school buildings in hot area. In: Proceedings of the 2nd world congress on civil, structural, and environmental engineering (CSEE’17), Barcelona, Spain, 2–4 April 2017.

20.

Subhashini

Thirumaran

. A passive design solution to enhance thermal comfort in an educational building in the warm humid climatic zone of Madurai. J Build Eng 2018; 18: 395–407.

21.

Zain

Taib

Baki

SMS

. Hot and humid climate: prospect for thermal comfort in residential building. Int J Desalination 2007; 209: 261–268.

22.

Indraganti

Ooka

Rijal

, et al. Adaptive model of thermal comfort for offices in hot and humid climates of India. Build Environ 2014; 74: 39–53.

23.

ISHRAE. Position paper on Indoor environmental quality. www.ishrae.in/images/ISHRAE-Position-Paper-Indoor-Environmental Quality.pdf (2015, accessed 14 September 2019).

24.

Subhashini S and Thirumaran K. Computational investigation of natural ventilation in an educational building in Madurai, Tamilnadu. In: Proceedings of PLEA 2017 – design to thrive, Edinburgh, UK. 3–5 July 2017, 3, pp.4973–4980.

25.

www.windfinder.com/windstatistics/madurai_airport (accessed 14 September 2019).

26.

Flintoft

Robinson

Melia

, et al. Average absorption cross-section of the human body measured at 1-12 GHz in a reverberant chamber: results of a human volunteer study. Phys Med Biol 2014; 59: 3297–3317.

27.

Geneva

Cuzzo

Fazili

, et al. Normal body temperature: a systematic review. Open Forum Infect Dis 2019; 6: 1–7.

28.

Derakhshan

Shaker

. Numerical study of the cross-ventilation of an isolated building with different opening aspect ratios and locations for various wind directions. Int J Vent 2016; 16: 42–60.