Study on Skywell Shape in Huizhou Traditional Architecture Based on Outdoor Wind Environment Simulation

Abstract

This study was conducted in the context of the latest Chinese policy on “double carbon”. First, we obtained building skywell and meteorological data parameters through a site survey and measurements. We applied the PHOENICS software to simulate and analyze the wind environment of a traditional building skywell. Secondly, the outdoor wind environment of typical building skywells could be simulated and evaluated one by one. Finally, using the method of controlling the variables and by combining typical buildings and skywell-scale layouts, the study summarized and compared the wind environment of the skywell under different scale combinations from three aspects: building skywell shape, skywell scale ratio, and skywell door opening. The following conclusions were drawn: (1) Among the four skywell shapes, the wind environment inside of the skywell was best in the HUI shape. (2) The wind environment inside of the skywell was best in the simulated skywell width-to-height ratio D/H values of 0.2–0.6; the AO shape D/H value was equal to 0.3; and the best wind environment in the skywell occurred when the D/H value of the HUI shape was equal to 0.4. (3) The wind environment in the skywell was best in the range of 1–1.5 for the aspect ratio W/L in the HUI-shaped building skywell (when the width-to-height D/H ratio was 0.4). (4) The opening of the door of the residential building had a great impact on the wind environment of the skywell.

1. Introduction

1.1. Huizhou Architecture Skywell

The skywell is a typical feature of Huizhou architecture and an important space; thus, there is the saying that “all halls are wells” [1,2]. In traditional Huizhou architecture, the skywell, as a transition space between the interior and exterior [3], not only has positive significance for the lighting and ventilation of the building interior but also contains a lot of traditional Huizhou culture. Huizhou is famous for Huizhou merchants, who do not lack money and pay more attention to the construction of their own houses and the comfort and beauty of their houses; they set up skywells in their houses to acquire the aura of the “unity of heaven and man” and the pride of “standing on top of the sky” [4]. On sunny days, the sun shines through the skywell to the front of the hall and the rooms, which means “sprinkling gold” [5]; in rainy and snowy weather, rainwater flows down from the eaves around the skywell and sinks into the tank below the skywell, which means “flowing silver” and is a metaphor for wealth, also known as the “four waters to the hall” [6]. In the skywell, combined with the through corridors and open halls, when the outdoor wind speed is high, wind pressure ventilation makes up the majority of the wind entering into the room from the skywell out, reducing the amount of indoor ventilation. When the outdoor wind is still, the skywell-shaped thermal pressure, which promotes indoor and outdoor ventilation, thus forms a complete ventilation system and plays the role of “hiding wind and gathering air”.

1.2. Huizhou Architecture: Skywell Shape

According to the research statistics, the traditional group of village architecture in the Huizhou area is compact in layout and variable in form. The layout of a single house is mostly formed in the shape of a rectangle or, within a nearly rectangular courtyard, is symmetrically arranged around the central axis, and the village buildings generally contain three rooms with wide faces, a hall in the middle, two side rooms, and a skywell connected to the hall and entrances [7]. According to the location and layout of the skywell, the plan types of traditional village architecture in Huizhou can be summarized into the following four forms: AO-shaped, HUI-shaped, H-shaped, and RI-shaped (Table 1).

Table 1. Architectural layout of traditional villages in Huizhou.

Form	Features	Example	Legend
AO-shaped (Three-in-one)	The most basic, plain layout unit is composed of a skywell, a hall, and left- and right-wing rooms. The skywell forms the center. This layout is very simple. It is generally square, and the depth of the bay is roughly the same. This plain shape is also called “One Seal”.	Dunren SuTing LiuJi Dongyuan Youqing
HUI-shaped (Four-in-one)	Additionally called “Courtyard Dwellings”, this kind of skywell shape is configured by rooms and roofs facing in four directions, and is a three-bay and two-deep building; for the upper and lower halls, two groups of three room types are combined in opposite directions; the middle is the Mingtang skywell, and generally, the upper hall ground is higher than the lower hall; and the lower hall depth is shallower than the upper hall, and both sides of the skywell are connected by compartments, also called the “upper and lower opposite hall”.	Yingfu Gaozuo Jishan Shuguangming
H-shaped (Three rooms and two courtyards)	It consists of two “triple courtyard” mirrors, with two three-room, back-to-back combinations in the middle and a skywell at the top and bottom. This layout form can increase space; the middle hall has a double slope ridge, so this form is also known as “a ridge over two halls”.	Panxianxiong Panmaotai Qingminju Shushiming
RI-shaped (Three rooms and three courtyards)	It is a combination of two HUI shapes, which are arranged in a “triple depth” according to the central axis, and the H shape can increase the space.	Shulimin Guanting Wujinhua

1.3. Proportion of Skywell in Huizhou Architecture

The enclosed interface of the skywell in the traditional village architecture of the Huizhou region is: top interface, vertical interface, and bottom interface. The characteristics of these three surfaces are as follows. The first point is the top interface of the skywell space of Huizhou houses: the opening of the skywell will face the sky, and the eaves on the roof of the building will extend into the range where the opening is located, so the opening and the “eaves” form the top interface. The second point is the vertical interface of the Huizhou folk house skywell space: the horizontal interface of a Huizhou folk house skywell is in the space; to define the space from a horizontal perspective, the medium of the enclosed “surface” in the horizontal direction will make people feel it first, the horizontal interface can block the line of sight, and the vertical interface is the outer wall of the room adjacent to the horizontal interface. The vertical interface is the exterior wall of the room adjacent to the horizontal interface. The third point is the bottom interface of the skywell space of Huizhou houses: this interface needs to consider rainwater in the planning and design stage, which enters from the top of the building and needs to be drained at the bottom, so the top interface and the bottom interface are in a “symbiotic relationship”; the bottom interface is also the base of the skywell space, which can accommodate various functions of the internal courtyard [8]. In this paper, we focus on the vertical enclosure scale of the vertical interface of the skywell.

In this paper, the vertical enclosure scale of the vertical interface of the skywell was measured by the ratio of the width of the enclosure interface and the height of the enclosure interface, D/H, which has been described above. The ratio of the width-to-height ratio D/H of Yoshinobu Ashihara was mainly focused on 1. For the architectural skywell, when the D/H value is greater than 1, the architectural skywell space shows openness, and when the D/H value is less than 1, the architectural skywell space shows closedness. In order to investigate the scale ratio of architectural skywells in traditional villages in the Huizhou region, the authors counted the D/H values of the skywells of representative residential buildings in Huizhou houses and Dongyang houses in Zhejiang Province as follows: the skywells of Huizhou houses are mainly narrow and long, while the skywells of Dongyang houses are mainly open (Table 2).

Table 2. Representative D/H value of Huizhou residential houses.

D/H	Example	D/H	Example	D/H	Example
0.35	Cheng Sui Lin Residential	0.63	Ba Weizhu Residential 1	0.26	Ningyuan Hall 1
0.37	Bi Shunsheng Residential	0.28	Ba Weizhu Residential 2	0.40	Ningyuan Hall 2
0.36	Jiang Zhonglai Residential	0.24	Ba Weizhu Residential 3	0.42	Yude Hall 1
0.47	Cunai Hall	0.38	Yizhen Hall, Yilan Pavilion 1	0.43	Yude Hall 2
0.43	Bao Xunzheng Residential	0.52	Yizhen Hall, Yilan Pavilion 2	0.41	Yude Hall 3
0.43	Bao Shumin Residential	0.47	Yizhen Hall, Yilan Pavilion 3	0.54	Fang Jinrong Hall
0.43	Yuan He Hall	0.70	Tianxin Hall	0.22	Jingzhao Di
0.46	Ba Fudao Transportation	0.50	Jiushitong Hall	0.48	Liangwang Hall
0.64	Pan zaojin Residential	0.57	Panxianxiong dwelling	0.50	Pan Runsheng Residence

According to Table 2, the width-to-height ratio D/H of traditional dwellings in Huizhou is generally low, and the maximum D/H value does not exceed 0.7. In the data, the D/H value is mainly 0.3–0.4. Through field investigation, it can be found that the plain of traditional dwellings in Huizhou is a long rectangle, and the space of the skywell is narrow and tall. The residential buildings in Dongyang, Zhejiang Province consist mainly of two floors. Table 3 shows that their D/H values are basically greater than 2. The plain of Dongyang folk houses is approximately square, and their skywell space is open and bright. The courtyard spaces of traditional folk houses in Huizhou and Dongyang folk houses in Zhejiang are the two extremes of the courtyard space enclosure scale of traditional Chinese folk houses, and the D/H value of most other traditional folk houses is between 1 and 2. Through field measurement, the author found that the enclosure scale of the skywell is only one aspect that affects the feelings of people living in it. The interior space of the skywell of Huizhou folk houses is rich and colorful, and there are many influencing factors.

Table 3. Representative D/H value of Zhejiang Dongyang residential houses.

D/H	Example	D/H	Example	D/H	Example
2.35	Yeshi Flower Hall	2.28	Wei Residence	2.25	Chengdong Flower Hall
2.42	Ruizhi Hall 1	2.33	Ruizhi Hall 2	2.4	Yijing Hall
2.11	Wuben Hall 1	2.22	Wuben Hall 2

1.4. Research Basis

Most of the studies on skywells have focused on natural ventilation, the thermal environment, the height, and the opening area [9,10,11,12,13,14,15,16,17,18]. Studying the indoor environment of skywell dwellings, Yang Yang et al. [4] showed that the internal space of traditional skywell dwellings is comfortable and the transition space of skywells has a significant cooling effect through the summer in the thermal environment of skywell dwellings in the Huizhou region. Huang, Zhijia et al. [19] concluded that back-shaped skywells in traditional dwellings have better control over natural ventilation than one-character skywells and can achieve the effect of both increasing natural ventilation in the transitional season and reducing outdoor cold air intrusion in winter. Zeng Zhihui et al. [20] concluded that a high and narrow skywell has a significant effect on the transition space and can effectively reduce and delay the effect of outdoor air temperature on the interior, and the more skywells there are, the better the ventilation effect is; in a study of the effect of different geometric skywell sizes on the effect of indoor natural ventilation, Liu Sheng et al. [21] quantitatively analyzed the relationship between different forms of skywell, whether to provide additional ventilation openings, and the effect of indoor natural ventilation. Lin Borong et al. [22] analyzed the effect of different sizes of courtyard space and the height of the house on the natural ventilation effect of the courtyard. Mousli K et al. [23] showed that skywells had been widely used in the Middle East as a passive ventilation and cooling measure. Al-Hemiddi N A et al. [24] studied the effect of skywells on the indoor cooling in summer by measuring skywell buildings. Kubota T et al. [9], in their study of the indoor thermal environment of traditional skywell buildings in Chinatown in Malaysia’s hot and humid climate zone, concluded that the evaporation of water and transpiration of plants at the bottom of the skywell can take away some of the heat from the room, and the effect of the skywell on the internal wind environment of residential buildings is also influenced by the form of building construction and the temperature of the wall of the envelope. Micallef D et al. [10] conducted a study on the through-air of skywell buildings; the results showed that the ventilation rate increases with the increase in skywell height and the airflow flows from outside to inside the skywell.

In a study of natural ventilation simulation and the optimization of skywell dwellings, Kobayashi et al.’s [25] evaluation of the ventilation performance of monitor roofs in a residential area was based on simplified estimation and CFD analysis. Gou, S et al. [26] studied the climate response strategies of traditional dwellings in ancient villages in hot summer and cold winter regions; it was found that utilizing buffer spaces such as internal balconies and patios can effectively promote natural ventilation in buildings. Zhong et al. [27] conducted a study entitled Numerical Investigations on Natural Ventilation in Atria of China’s Southern Yangtze Vernacular Dwellings; this study presented a numerical investigation of natural ventilation in a skywell of Southern Yangtze dwellings using a validated computational fluid dynamics (CFD) model. The Reynolds-averaged Navier–Stokes (RANS) modeling approach with the RNG k-ε turbulence model was used for the numerical simulation performed in Open FOAM. Kotol et al. [28] carried out a study of current ventilation strategies in Greenlandic dwellings. Cardoso et al. [29] studied a labelling strategy to define the airtightness performance ranges of naturally ventilated skywells: an application used in Southern Europe. Prakash et al. [30] studied the ventilation performance of low-rise courtyard buildings with various courtyard shape factors and roof topologies.

Liu Xiangmei et al. [31] quantitatively studied the effect of skywell scale on airflow velocity in the courtyard space by applying certain transformations to the length, width, and proportion of skywell scale, and arrived at the optimal skywell scale for natural ventilation. Chen Qiuyu et al. [32] simulated the effect of different skywell shapes and changing the location of the ventilation openings on the indoor ventilation effect, and the results showed that a long vertical skywell has the best comprehensive effect on the indoor environment, and additional ventilation openings on the windward side can significantly improve the indoor ventilation effect. Qian Wei et al. [33] conducted a study on the optimal design of a skywell dwelling, taking into account the natural ventilation and lighting in the room. Rajapaksha I et al. [11] analyzed indoor environmental parameters for five different skywell opening area conditions in a temperate and humid region of Sri Lanka and showed that a proper skywell design can achieve the natural ventilation needed for the building, thus improving the thermal comfort of the building.

This study uses simulation experiments through the PHOENICS (2019) software to quantitatively analyze the relationship between the outdoor wind environment and skywell morphology of traditional Huizhou buildings, and proposes optimization strategies for the skywell morphology of traditional Huizhou Village buildings. The optimization strategies proposed in the research results will provide more valuable theoretical and quantitative criteria for the design and renovation of architectural skywells in traditional Huizhou villages. In-depth research on the outdoor wind environment condition of traditional villages in China is conducive to the inheritance of traditional village culture; the condensation, excavation, and inheritance of Huizhou traditional village architectural planning and design ideas and techniques; and the exploration of construction techniques that can both adapt to modern architecture and retain the characteristics of a traditional architectural style.

2. Research Method

2.1. Research Location

Through the analysis of the traditional villages in the Huizhou region and due to the limitation of measurement and the fact that the “RI” plain form was a combination of the “AO” plain form, in this paper, the main measured skywell forms were: the “AO” plain form, the “H” plain form, and the “HUI” plain form. The authors selected the main research objects of this paper as Xidi Village (29°54′ N 118° E), Yuyuan Village (28.77° N 119.66° E), and Zhifeng Village (29.28° N 117.67° E) on the basis of the research summary [34]. The three villages are in Southeastern China with similar latitudes and longitudes, and they all lie in the subtropical monsoon climate zone.

2.2. Case Selection

The authors researched and measured three traditional villages in Huizhou: Xidi Village, Yuyuan Village, and Zhifeng Village, and selected eight groups of buildings with different skywell shapes for a comparative study in Figure 1.

Figure 1. Location of three village.

Of all the skywell houses in Xidi Village, about 80% were AO-shaped skywell houses, 11% were HUI-shaped skywell houses, and 9% were H-shaped skywell houses. One of each type was chosen for further investigation—Yingfu dwelling (HUI-shaped), Lufu dwelling (H-shaped), and Dunren dwelling (AO-shaped). The locations of these dwellings in Xidi Village are shown in Figure 2. The basic information of three monitored dwellings in Xidi Village is summarized in Table 4: Dunren dwelling (floor area 174 m²), Lufu dwelling (floor area 147 m²), and Yingfu dwelling (floor area 185 m²).

Figure 2. Research object orientation map of Xidi Village.

Table 4. Summary on the study of skywell architecture in Xidi Village.

	Dunren Dwelling	Lufu Dwelling	Yingfu Dwelling
building type	AO-shaped	H-shaped	HUI-shaped
year built	1722	1684	1664
floor area (m2)	174	147	185
floors	two-story	three-story	two-story
no. of skywells	three	two	one
size of skywell L × W × H (m)	main skywell: 4.3 × 2.15 × 5.26 two small skywells: 1.45 × 1.2 × 7.6	west side: 2.3 × 1.3 × 7.5 east side: 3.3 × 1.35 × 7.8	4.2 × 1.8 × 6.8
area of skywell (m2)	main skywell: 9.25 two small skywells: 1.74	west side: 2.99 east side: 4.46	7.56

In Zhifeng Village, about 90% of skywell houses were H-shaped (two AO-shaped houses or one AO-shaped and one HUI-shaped house combined together). The locations of these dwellings in Zhifeng Village are shown in Figure 3.

Figure 3. Research object orientation map of Zhifeng Village.

Two typical houses—Panmaotai dwelling (H-shaped) and Panxianxiong dwelling (H-shaped)—were chosen for further investigation. Basic information about these monitored dwellings is summarized in Table 5: Panxianxiong dwelling (floor area 198 m²) and Panmaotai dwelling (floor area 162 m²).

Table 5. Summary of the study of the courtyard buildings in Zhifeng Village.

	Panxianxiong Dwelling	Panmaotai Dwelling
building type	combined AO-shaped and HUI-shaped	H-shaped
year built	1900s	1900s
floor area (m2)	198	162
floors	three-story	three-story
no. of skywells	two	two
size of skywell L × W × H (m)	south side: 1.1 × 0.75 × 6.2 north side: 1.8 × 0.75 × 6.3	east side: 1.1 × 0.6 × 6 west side: 1.1 × 0.6 × 6
area of skywell (m2)	south side: 0.83 north side: 1.35	east side: 0.66 west side: 0.66

Two main types of dwellings can be found: AO-shaped and HUI-shaped. After calculation, there is about 81% AO-shaped and 13% HUI-shaped within all the courtyard houses in Yuyuan Village. The locations of these dwellings in Yuyuan Village are shown in Figure 4.

Figure 4. Research object orientation map of Yuyuan Village.

Three typical Chinese traditional houses—Yufengfa dwelling (AO-shaped), Shuting dwelling (AO-shaped), and Gaozuo dwelling (HUI-shaped)—were chosen for further investigation. Basic information on the three studied dwellings in Yuyuan Village is summarized in Table 6: Shuting dwelling (floor area 261 m²), Yufengfa dwelling (floor area 288 m²), and Gaozuo dwelling (floor area 573 m²).

Table 6. Summary of the study on the skywell buildings in Yuyuan Village.

	Shiting Dwelling	Yufengfa Dwelling	Gaozuo Dwelling
building type	AO-shaped	AO-shaped	HUI-shaped
year built	1796–1820	1800	1796–1820
floor area (m2)	261	288	573
floors	two-story	two-story	two-story
no. of skywells	one	one	one
size of skywell L × W × H (m)	6.1 × 5.2 × 5.3	9.2 × 2.95 × 4.9	9.3 × 5.8 × 5
area of skywell (m2)	31.72	27.14	53.94

2.3. Natural Climate Characteristics of Huizhou Region

The historical meteorological data of Huizhou show that the east wind and northeast wind are the dominant winds in the Huizhou area, with the east wind prevailing in spring and summer and the northeast wind in autumn and winter. Table 7 shows a summary of climate data for the three villages (av.temp. = Average temperature and av. RH = Average relative humidity).

Table 7. Summary of climate data in Xidi Village, Zhifeng Village, and Yuyuan Village.

Season	Climate Data	Xidi Village	Zhifeng Village	Yuyuan Village
summer (Jun. to Aug.)   hot and humid	hottest month (av.temp.)	July (29.6 °C)	July (29.0 °C)	July (28.5 °C)
	max. temp.	35.8 °C	36.6 °C	36.4 °C
	av. temp.	28 °C	27.7 °C	27.2 °C
	prevailing wind	northwest and south	north and northeast	northeast and southwest
	av. RH 9 a.m.	82%	73%	76%
	av. RH 3 p.m.	62%	63%	65%
autumn (Sep. to Nov.)   warm and humid	av. temp.	18.2 °C	18.9 °C	18.7 °C
	max/min. temp.	35.1 °C/6 °C	34.6 °C/5.9 °C	34.3 °C/4.1 °C
	prevailing wind	southeast and northwest	south and east	southeast
	av. RH 9 a.m.	86%	72%	75%
	av. RH 3 p.m.	52%	56%	60%
winter (Dec. to Feb.)   cold and humid	coldest month (av.temp.)	January (3.9 °C)	January (4.9 °C)	January (5.1 °C)
	min. temp.	−0.7 °C	−3 °C	−0.8 °C
	av. temp.	5.1 °C	6.6 °C	6.5 °C
	prevailing wind	northeast and south	northeast	northeast and east
	av. RH 9 a.m.	87%	73%	75%
	av. RH 3 p.m.	58%	55%	66%
spring (Mar. to May)   mild and humid	av. temp.	16 °C	16.5 °C	17.1 °C
	max/min. temp.	30.0 °C/1.0 °C	29.8 °C/0.8 °C	31.5 °C/3.4 °C
	prevailing wind	south	west	east
	av. RH 9 a.m.	82%	79%	76%
	av. RH 3 p.m.	59%	67%	64%

2.4. Wind Environment Evaluation Criteria

This paper takes traditional villages in the Huizhou region as the research object and combines domestic and international evaluation standards of wind environments [35], which are mainly based on wind speed threshold, pollutant deposition, and pedestrian comfort [36], involving wind frequency, wind speed ratio, wind pressure, and wind speed magnitude [37,38]. As shown in Table 8, this paper combines the above evaluation indexes and the characteristics of climate change in the Huizhou region in four seasons, to evaluate the outdoor wind environment of traditional villages in Huizhou in winter and summer, mainly choosing the three indexes of wind speed, wind vortex, and the wind amplification coefficient.

Table 8. The evaluation standard of Huizhou Village wind environment adopted by this paper.

Season	Wind Speed (1.5 H) (1.5 H Stands for 1.5 m Pedestrian Height)	Wind Vortex	Wind Speed Amplification	Comfort and Air Pollutant Concentration
Summer	0.60 m/s ≤ V ≤ 5 m/s	Avoid	<2	Comfortable; the concentration of air pollutants decreases with the increase in wind speed
Winter	V ≤ 5 m/s	Avoid	<2	In cold weather, the concentration of air pollutants decreases with the increase in wind speed

3. Numerical Simulation Methods

3.1. Model Establishment

In the building outdoor wind environment simulation, a three-dimensional model needed to be built according to the actual size of the building, which was required to present the maximum details within the building boundary. The field distribution, however, without affecting the flow around the building, could be reasonably simplified in the actual modeling in order to improve the computational convergence speed by ignoring the small concavities on the building surface, and the entity approximating the cubic building was directly simplified to a regular cube. Take Figure 5 as an example.

Figure 5. Simulation model of Arcadia residential area.

3.2. Calculation Area Setting

In this paper, the size of the calculation area was set based on the Green Building Design Standard. The specific settings are as follows: building coverage area—below 3% of the whole calculation domain area; horizontal calculation area—the area with the target building as the center of the circle and 5 H as the radius; calculation area above the building—greater than 3 H (H is the height of the main body of the building) [39].

3.3. Computational Meshing

In the PHOENICS software, the quality of mesh generation largely affects the accuracy and stability of the simulation results [40]. To ensure the accuracy and computational efficiency of the simulation results, the mesh division must be reasonable [41]. According to the existing Japanese AIJ regulations, the stretching ratio of two adjacent grids must be under 1.3, and the minimum resolution of the grid is 1/10 of the building scale. In Jiangsu Province [42] and the Beijing Green Building Design Standard [43], it is stipulated that: the grid division at a 1.5 m or 2 m height in the pedestrian area next to the building should be 10 grids or more; the key observation area is placed at the third grid or more on the ground. Combining with the above regulations, this paper reasonably divided the calculation grid, taking Xidi Village as an example in Figure 6.

Figure 6. The calculated domain grid of Arcadia residential area.

3.4. Calculation Boundary Condition Setting

The reasonable setting of boundary conditions is an important prerequisite for the accuracy of simulation results. Among a large number of studies on the setting of boundary conditions for wind environments, the AIJ of the Architectural Institute of Japan and the European COST analyzed numerous simulations and summarized the guidelines for setting boundary conditions and simulation settings applicable to outdoor wind environment simulations [44], which are the most representative studies. In this paper, we mainly refer to the standard of AIJ of Japan to set the boundary conditions.

Wind field type selection: the WIND method was used for the adjustment of the calculation area that needed to meet the following requirements: a building coverage area below 3% of the entire calculation domain area; horizontal calculation area: the target building as the center of the circle and 5 H as the radius of the area; and calculation area above the building: greater than 3 H (H is the height of the main body of the building). The calculation area of the WIND method is much smaller than the INLET–OUTLET method, and the calculation volume is greatly reduced [45]. Therefore, this paper took the WIND method to simulate the complex wind direction under various working conditions.

Entrance condition setting: the wind speed decreases when the wind is close to the ground due to ground friction and generates vertical gradient changes, thus forming urban gradient wind, as shown in Figure 7.

Figure 7. Contour profiles of different regions of the city.

The exponential rate was used as the expression for the wind speed profile in the software simulation process:

$$\frac{U_z}{U_s}={\left(\frac{Z}{Z_s}\right)}^{\alpha }$$

where $Z_s$is the reference height (m), usually 10 m, $U_s$is the average wind speed at the reference height (m/s); $U_z$is the average wind speed at height Z (m/s). The index $\mathsf{\alpha }$is the degree of wind speed reduction near the ground surface, and the increase in the index$\alpha$ represents the increase in the degree of ground roughness. According to the classification of ground roughness and the corresponding index $\alpha$value given in the Code for Structural Loading of Buildings (GB50009-2012 [46]), as shown in Table 9, the value of 0.3 was taken in this paper according to the actual situation.

Table 9. Index values of different terrain ground roughness.

Ground Type	Applicable Area	Roughness Index Value	Gradient Wind Height (m)
A	Offshore areas, lakeshore, and desert areas	0.12	300
B	Fields, hills, small and medium-sized cities, and suburbs of large cities	0.15	350
C	Large urban areas with dense buildings	0.22	450
D	Urban areas with dense clusters of buildings and taller houses	0.30	550

The turbulence model and fluid material were selected: Normally, the KEMODE model is selected for outdoor wind simulation. The fluid material is usually selected under “0” gas (gases).

The number of iterations and step length setting: The initial number of iterations of the software is 100 and the step length is 1. The researcher should adjust it according to different research situations. In this paper, the number of iterations was set to 2000, and the step length was 1.

The convergence parameters were set: First, the solution converges sufficiently to stop the computation, and second, it needs to be determined that the value of the specified observation point no longer changes or the root mean square residual is less than 10 × 10⁻⁴. In this paper, the convergence criterion was set to 0.01%. The European COST proposes an iterative convergence determination criterion of 0.001 for industrial fields, which is more demanding. Therefore, the iterative residual can be reduced to 10⁻⁴ as the convergence determination for the simulation of the wind environment around the building. In addition, the target variable needs to be detected, and the solution can be judged to be convergent when the variable value is constant or oscillates in a small range around a certain value.

4. Results and Discussion

4.1. The Influence of Skywell Shape on the Outdoor Wind Environment of Buildings

In this simulation, the eight groups of models were AO-shaped, H-shaped, and HUI-shaped, and the average wind speed of 1.9 m/s in the Huizhou area in summer was the wind parameter used in this simulation. The simulated wind direction was the same as the measured wind direction, which was northeast wind. In this paper, we studied the wind speed cloud and wind speed vector diagram on the Z-axis slice at the pedestrian height (1.5 m) and used the wind speed evaluation index to evaluate the simulation results. The simulation results are as follows in Table 10.

Table 10. Wind speed nephogram of each study object’s skywell.

(a). Cloud Chart of Wind Speed at the Pedestrian Height of Dunren dwelling	(b). Cloud Chart of Wind Speed at Pedestrian Height of the Yufengfa dwelling
(c). Cloud Chart of Wind Speed at Pedestrian Height of Shuting dwelling	(d). Cloud Chart of Wind Speed at Pedestrian Height of Lufu dwelling
(e). Cloud Chart of Wind Speed at Pedestrian Height of Panmaotai dwelling	(f). Cloud Chart of Wind Speed at Pedestrian Height of Panxianxiong dwelling
(g). Cloud Chart of Wind Speed at the Traveling Height of Yingfu dwelling	(h). Cloud Chart of Wind Speed at Pedestrian Height of Gaozuo dwelling

According to the wind speed nephogram of eight groups of different skywell configurations simulated in Table 10, the following can be concluded. AO-shaped architecture: the wind speed of the Dunren skywell is 0.15 m/s–1.37 m/s, and the measured average wind speed is 0.29 m/s; the wind speed of the Yufengfa skywell is 0.18 m/s–1.69 m/s, and the measured average wind speed is 0.50 m/s; and the wind speed of the Shuting skywell is 0.34 m/s–1.69 m/s, and the measured average wind speed is 0.57 m/s. H-shaped buildings: the wind speed of the skywell of the Lufu dwelling is 0 m/s–0.79 m/s, and the measured average wind speed is 0.018 m/s (glass sheltered); the wind speed of the skywell of the Panxianxiong dwelling is 0 m/s–0.34 m/s, and the measured average wind speed is 0.28 m/s; and the wind speed of the skywell of the Panmaotai dwelling is 0 m/s–0.39 m/s, and the measured average wind speed is 0.12 m/s. HUI-shaped: the wind speed of the Yingfu dwelling skywell is 0.79 m/s–1.84 m/s, and the measured average wind speed is 0.22 m/s (covered by rain cloth); the wind speed of the Gaozuo dwelling skywell is 0.16 m/s–0.94 m/s, and the measured average wind speed is 0.21 m/s. The above data are summarized in Table 11.

Table 11. Distribution of wind speed in courtyard.

Building Case	Wind Speed Range	Maximum Wind Speed	Plain Shape of Skywell
Dunren Hall	0.15 m/s < V < 1.37 m/s	V = 1.37 m/s	AO-shaped
Liuji Hall	0.18 m/s < V < 1.69 m/s	V = 1.69 m/s	AO-shaped
Shuting Hall	0.34 m/s < V < 1.69 m/s	V = 1.69 m/s	AO-shaped
Lvfu Hall	0 m/s < V < 0.79 m/s	V = 2.14 m/s	H-shaped
Panmaotai	0 m/s < V < 0.39 m/s	V = 0.39 m/s	H-shaped
Panxianxiong	0 m/s < V < 0.34 m/s	V = 0.34 m/s	H-shaped
Yingfu Dwelling	0.79 m/s < V < 1.84 m/s	V = 1.84 m/s	HUI-shaped
Gaozuo Hall	0.16 m/s < V < 0.94 m/s	V = 0.94 m/s	HUI-shaped

According to the above simulation data, the most ideal wind speed in the skywell of an AO-shaped building is that of the Shuting dwelling. In general, the low pressure in the skywell will cause the wind speed to reverse at the opening of the counter wind surface. Vortex whirlwind will appear vertically in the skywell, resulting in a decrease in the overall wind speed in the skywell. At the same time, the area of wind stagnation in the skywell where the wind speed is less than 0.60 m/s will be expanded. The most ideal wind speed in the skywell of an H-shaped building was in the Lufu dwelling. As the Lufu dwelling was covered with a glass roof in the actual state, the measured average data were obviously smaller than the simulation data. The most ideal wind speed in the skywell of an HUI-shaped building was in the Yingfu dwelling. The area of the skywell of the Gaozuo dwelling was significantly larger than that of the Lufu dwelling. The reason for the poor average wind speed was that there were doors on the north and east sides of the Lufu dwelling, and the prevailing wind direction was convenient for entering the skywell.

4.2. Influence of Skywell Proportion on Outdoor Wind Environment

According to Table 4, Table 5 and Table 6, the courtyard buildings in each village can be summarized as follows. AO-shaped buildings: the width/height ratio D/H value of the courtyard in the Shuting dwelling is 0.98 > the width/height ratio D/H value of the Yufengfa dwelling is 0.59 > the width/height ratio D/H value of the courtyard in Dunren is 0.41 and 0.18; H-shaped building: the width/height ratio D/H value of the skywell in the Lufu dwelling is 0.17 > the width/height ratio D/H value of the skywell in the Panxianxiong dwelling is 0.12 > the width/height ratio D/H value of the skywell in the Panmaotai dwelling is 0.1; and HUI-shaped building: the height/width ratio D/H value of the Yingfu dwelling skywell is 0.26 > the wind speed of the Gaozuo skywell is 1.16. It can be seen from the simulation results that the width-to-height ratio of the three types of planar buildings, AO-shaped buildings, and H-shaped buildings is in direct proportion to the wind speed; that is, the greater the D/H value, the greater the wind speed. It can be seen that the key value of Yoshinobu Ashihara D/H theory is concentrated on 1. When the width-to-height ratio D/H value is greater than 1, the space will be open, and when the width-to-height ratio D/H value is less than 1, the space will be closed. According to the authors’ field survey, the width-to-height ratio D/H of traditional dwellings in Huizhou was generally small, and the maximum value in the measured ratio data was not more than 0.7, at basically between 0.3 and 0.4. Since the four forms of Huizhou courtyard dwellings evolved from AO-shaped and HUI-shaped, this simulation selected AO-shaped architecture and HUI-shaped architecture as the prototype for simulation. In the previous part of this section, the measured residential standards (the same bay, depth, and story height) were used to establish a model as a quantitative measure, so as to simulate the two kinds of skywell shape and try to explore the skywell shapes that are both universal and conform to the comfort wind field, while studying the changes in the skywell wind environment.

(1)

Influence of the ratio of width to height of the skywell on the outdoor wind environment of the building skywell

The control variable method was adopted in the simulation. Under the condition that the height of the bay depth of the residential building remains unchanged, two working conditions were used to simulate the influence of the courtyard building form and the D/H value of the courtyard width/height ratio on the outdoor wind environment. The prototype of the residential building in this simulation has a bay of 13 m, a depth of 10 m, and a height of 9 m. In hot summer and cold winter areas, we can guide and strengthen the dominant wind direction in summer by reasonably planning and arranging different shapes of the building skywell. One of the main reasons for setting skywells in traditional village buildings in Huizhou is the ventilation and heat dissipation in summer. In this paper, the prevailing wind direction in summer is northeast, and the wind speed is 1.9 m/s.

Working condition 1: wind field of AO-shaped and HUI-shaped buildings when D/H = 0.2, D/H = 0.3, D/H = 0.4, D/H = 0.5, and D/H = 0.6 (Table 12).

Table 12. Wind field of each D/H value skywell in AO-shaped.

Skywell Shape D/H Value	Z-Axis Slice Wind Speed Nephogram	Z-Axis Slice Wind Speed Vector Diagram
AO-shaped D/H = 0.2
AO-shaped D/H = 0.3
AO-shaped D/H = 0.4
AO-shaped D/H = 0.5
AO-shaped D/H = 0.6

By analyzing the simulation results, it was found that the wind speed inside the model skywell under different aspect ratios was less than 2 m/s at a 1.5 m pedestrian height, and the wind speed amplification coefficient was less than 1.69, meeting the Green Building Evaluation Standard. In the AO-shaped skywell, when D/H = 0.2, the wind speed in the skywell was 0.20 m/s–1.23 m/s, the wind speed distribution was uneven, and the probability of the wind speed being greater than 0.60 m/s was 37%. When D/H = 0.3, the wind speed in the skywell was 0.48 m/s–1.42 m/s, the wind speed was evenly distributed, and the probability of the wind speed being greater than 0.60 m/s was 79%. When D/H = 0.4, the wind speed in the skywell was 0.33 m/s–1.26 m/s, the wind speed distribution was relatively uniform, and the probability of the wind speed being greater than 0.60 m/s was 63%. When D/H = 0.5, the wind speed in the skywell was 0.24 m/s–1.12 m/s, the wind speed distribution was uneven and there was a wind vortex, and the probability of the wind speed being greater than 0.60 m/s was 47%. When D/H = 0.6, the wind speed in the skywell was 0.22 m/s–1.01 m/s, the wind speed was evenly distributed, two wind whirlpools appeared, the area of the calm wind area was large, and the probability of the wind speed being greater than 0.60 m/s was 23%. From the above simulation data, it can be concluded that when the internal width/height ratio D/H value of the AO-shaped building skywell is equal to 0.3, the wind field in the skywell will be the best. It can also be verified that “the D/H value of the width-to-height ratio of traditional dwellings in Huizhou is generally small, and the maximum value of the measured ratio data does not exceed 0.7, basically between 0.3 and 0.4”. This is in line with the objective law. The above simulation analysis is summarized in Table 13.

Table 13. Range and probability of wind fields at different scales in AO-shaped form skywell.

D/H Value of Skywell	Wind Speed Range of Skywell	Probability of Even Distribution of Air Speed in Skywell
D/H = 0.2	0.20 m/s < V < 1.23 m/s	V > 0.60 m/s = 37%
D/H = 0.3	0.48 m/s < V < 1.42 m/s	V > 0.60 m/s = 79%
D/H = 0.4	0.33 m/s < V < 1.26 m/s	V > 0.60 m/s = 63%
D/H = 0.5	0.24 m/s < V < 1.12 m/s	V > 0.60 m/s = 47%
D/H = 0.6	0.22 m/s < V < 1.01 m/s	V > 0.60 m/s = 23%

By analyzing the simulation results in Table 14, it was found that the wind speed inside the model skywell under different aspect ratios was less than 2 m/s at a 1.5 m pedestrian height, and the wind speed amplification coefficient was less than 1.69, meeting the Green Building Evaluation Standard. When D/H = 0.2, the wind speed in the skywell was 0.32 m/s–1.26 m/s and the wind speed distribution was relatively uniform under the HUI-shaped form. The probability of the wind speed being greater than 0.60 m/s was 68%. When D/H = 0.3, the wind speed in the skywell was 0.19 m/s–1.24 m/s, which is more evenly distributed than the wind speed. The probability of the wind speed being greater than 0.60 m/s was 59%. When D/H = 0.4, the wind speed in the skywell was 0.37 m/s–1.46 m/s, the wind speed distribution was relatively uniform, and the probability of the wind speed being greater than 0.60 m/s was 85%. When D/H = 0.5, the wind speed in the skywell was 0.20 m/s–1.35 m/s, the wind speed distribution was uneven and there was a wind vortex, and the probability of the wind speed being greater than 0.60 m/s was 74%. When D/H = 0.6, the wind speed in the skywell was 0.21 m/s–1.38 m/s, the wind speed was evenly distributed, the area of the calm wind zone was large, and the probability of the wind speed being greater than 0.60 m/s was 65%. From the above simulation data, it can be concluded that: when the internal width/height ratio D/H value of the skywell of the HUI-shaped building is equal to 0.4, the wind field in the skywell is the best. It can also be verified that “the D/H value of the width-to-height ratio of traditional dwellings in Huizhou is generally small, and the maximum value in the measured ratio data does not exceed 0.7, which is basically between 0.3 and 0.4”. This conforms to the objective laws. The above simulation analysis is summarized in Table 15.

Table 14. Wind field of each D/H value skywell in HUI-shaped form.

Skywell Shape D/H Value	Z-Axis Slice Wind Speed Nephogram	Z-Axis Slice Wind Speed Vector Diagram
“HUI” type D/H = 0.2
“HUI” type D/H = 0.3
“HUI” type D/H = 0.4
“HUI” type D/H = 0.5
“HUI” type D/H = 0.6

Table 15. Range and probability of wind fields at different scales in HUI-shaped form skywell.

D/H Value of Skywell	Wind Speed Range of Skywell	Probability of Even Distribution of Air Speed in Skywell
D/H = 0.2	0.32 m/s < V < 1.26 m/s	V > 0.60 m/s = 68%
D/H = 0.3	0.19 m/s < V < 1.24 m/s	V > 0.60 m/s = 59%
D/H = 0.4	0.37 m/s < V < 1.46 m/s	V > 0.60 m/s = 85%
D/H = 0.5	0.20 m/s < V < 1.35 m/s	V > 0.60 m/s = 74%
D/H = 0.6	0.21 m/s < V < 1.38 m/s	V > 0.60 m/s = 65%

Working condition 2: when the skywell plain shape is AO-shaped or HUI-shaped, D/H = 0.2, D/H = 0.3, D/H = 0.4, D/H = 0.5, and D/H = 0.6 (Table 16).

Table 16. Air field conditions of various skywells with the same D/H value.

Skywell Shape D/H Value	Z-Axis Slice Wind Speed Nephogram AO-Shaped	Z-Axis Slice Wind Speed Nephogram HUI-Shaped
D/H = 0.2
D/H = 0.3
D/H = 0.4
D/H = 0.5
D/H = 0.6

By analyzing the simulation results, it was found that the wind speed inside the two model skywells under different aspect ratios was less than 2 m/s at a 1.5 m pedestrian height, and the wind speed amplification coefficient was less than 1.69, meeting the Green Building Evaluation Standard. From the wind speed distribution of each wind speed cloud chart in Table 5, Table 6, Table 7, Table 8, Table 9 and Table 10, it can be seen that although the simulation results meet the evaluation criteria, there are also many calm and stagnant areas. Therefore, this paper evaluates the distribution uniformity of the wind speed cloud chart based on the probability that the wind speed in the skywell is greater than 0.60 m/s. The above simulation analysis is summarized in Table 17.

Table 17. Wind field distribution in different forms of skywell.

Yard Shape, D/H Value	AO-Shaped	HUI-Shaped
D/H = 0.2	V > 0.60 m/s = 38%	V > 0.60 m/s = 68%
D/H = 0.3	V > 0.60 m/s = 79%	V > 0.60 m/s = 59%
D/H = 0.4	V > 0.60 m/s = 63%	V > 0.60 m/s = 85%
D/H = 0.5	V > 0.60 m/s = 47%	V > 0.60 m/s = 74%
D/H = 0.6	V > 0.60 m/s = 23%	V > 0.60 m/s = 65%

It can be seen intuitively from the above table that the wind speed in the HUI-shaped skywell is obviously better than that in the AO-shaped skywell.

(2)

Influence of the length/width ratio of the skywell on the outdoor wind environment of the building skywell

It can be seen from the above simulation results that different building skywell width-to-height ratios can have different impacts on the building wind environment, so we can optimize the wind environment by adjusting the building skywell ratio. From the above simulation comparison and analysis, it can be found that the wind speed in the HUI-shaped skywell was obviously better than that in the AO-shaped skywell, and the wind speed in the skywell was the best when the width-to-height ratio of the HUI-shaped skywell was equal to 0.4. To ensure the preciseness of the research, this paper discusses the impact of the length-to-height ratio on the outdoor wind environment of the skywell building under the building prototype of the HUI-shaped skywell with a width-to-height ratio of 0.4. According to the survey, the range of the long side of the skywell is generally 5–8 m. Therefore, the L/W ratio of the length to width of the skywell was set as 1, 1.5, 2, 2.5, and 3 to study the impact of the length of the skywell on the outdoor air environment of the skywell building.

By analyzing the simulation results in Table 18, it was found that the wind speed inside the two model skywells under different aspect ratios was less than 2 m/s at a 1.5 m pedestrian height, and the wind speed amplification coefficient was less than 1.69, meeting the Green Building Evaluation Standard. When L/W = 1, the wind speed in the skywell was 0.65 m/s–1.74 m/s, and the wind speed distribution was relatively uniform under the HUI-shaped form. The probability of the wind speed being greater than 0.60 m/s was 90%. When L/W = 1.5, the wind speed in the skywell was 0.48 m/s–1.89 m/s, the wind speed was evenly distributed, and the probability of the wind speed being greater than 0.60 m/s was 84%. When L/W = 2, the wind speed in the skywell was 0.34 m/s–1.74 m/s, the wind speed distribution was relatively uniform, there was a very-small-scale wind vortex in the skywell, and the probability that the wind speed is greater than 0.60 m/s was 80%. When L/W = 2.5, the wind speed in the skywell was 0.34 m/s–1.84 m/s, the wind speed distribution was uneven and there was a wind vortex, and the probability of the wind speed being greater than 0.60 m/s was 78%. When L/W = 3, the wind speed in the skywell was 0.20 m/s–1.84 m/s, the wind speed distribution was uneven, two wind whirlpools appeared, the area of the calm wind area was large, and the probability of the wind speed being greater than 0.60 m/s was 68%. To facilitate a comparison, the above simulation results are summarized in Table 19.

Table 18. Wind field with aspect ratio of skywell in HUI-shaped form.

Skywell Shape L/W Value	Z-Axis Slice Wind Speed Nephogram	Z-Axis Slice Wind Speed Vector Diagram
HUI-shaped L/W = 1
HUI-shaped L/W = 1.5
HUI-shaped L/W = 2
HUI-shaped L/W = 2.5
HUI-shaped L/W = 3

Table 19. Range and probability of wind field for skywell aspect ratio of HUI-shaped form.

L/W Value of Skywell	Wind Speed Range of Skywell	Probability of Even Distribution of Air Speed in Skywell
L/W = 1	0.65 m/s < V < 1.74 m/s	V > 0.60 m/s = 90%
L/W = 1.5	0.48 m/s < V < 1.89 m/s	V > 0.60 m/s = 84%
L/W = 2	0.34 m/s < V < 1.74 m/s	V > 0.60 m/s = 80%
L/W = 2.5	0.34 m/s < V < 1.84 m/s	V > 0.60 m/s = 78%
L/W = 3	0.20 m/s < V < 1.84 m/s	V > 0.60 m/s = 68%

From the above simulation data, it can be concluded that at least one wind vortex appeared in the wind field in the skywell when the length/width ratio W/L value was 2–3, and the larger the ratio is, the larger the area of the stagnant wind area and the calm wind area will be, and the wind environment in the skywell will be disordered. When the length/width ratio L/W value was 1–2, the wind speed distribution in the courtyard was relatively uniform, and there were basically no wind whirlpools and quiet wind areas, which meets the outdoor wind environment standards in summer, and the wind environment was good. Therefore, it is recommended that the HUI-shaped building courtyard (width/height D/H value ratio 0.4) is the best in the length/width ratio L/W value range of 1–1.5.

4.3. The Influence of the Building Door Opening on the Outdoor Air Environment of the Skywell

It can be seen from the above simulation results that the wind speed in the skywell is the best when the width-to-height ratio of the HUI-shaped building is 0.4. Because the building door is the key to connect the building roadway and skywell, this paper discusses the impact of the opening of the building door on the outdoor wind environment of the skywell building under the building prototype of the HUI-shaped building with a width-to-height ratio of 0.4.

By analyzing the simulation results in Table 20, it was found that the wind speed inside the model skywell with different doors open was less than 2 m/s at a 1.5 m pedestrian height, and the wind speed amplification coefficient was less than 1.69, meeting the Green Building Evaluation Standard. When all the building doors were closed, the air speed of the skywell was 0.01 m/s–0.17 m/s, the air field of the skywell was basically in a calm state, and the probability that the wind speed was greater than 0.60 m/s was 0%. At this time, the air environment of the skywell was extremely poor. When one door of the building was closed, the wind speed of the skywell was 0.32 m/s–1.11 m/s, the wind speed distribution was very uneven, and the probability of the wind speed being greater than 0.60 m/s was 37%. When all were opened, the wind speed in the skywell was 0.37 m/s–1.46 m/s, and the probability of the wind speed being greater than 0.60 m/s was 85%. The wind speed distribution was relatively uniform. To facilitate a comparison, the above simulation results are summarized in Table 21.

Table 20. The situation of the air field of the skywell in the open state of different doors in HUI-shaped form.

Opening State of Building Door	Z-Axis Slice Wind Speed Nephogram	Z-Axis Slice Wind Speed Vector Diagram
All doors closed
One door closed
All doors open

Table 21. The situation of the air field of the skywell in the open state of different doors in HUI-shaped building.

Opening State of Building Door	Wind Speed Range of Skywell	Probability of Even Distribution of Air Speed in Skywell
All doors closed	0.01 m/s < V < 0.17 m/s	V > 0.60 m/s = 0%
One door closed	0.32 m/s < V < 1.11 m/s	V > 0.60 m/s = 37%
All doors open	0.37 m/s < V < 1.46 m/s	V > 0.60 m/s = 85%

From the above simulation data, it can be concluded that: when the building doors are all closed, a large area of stagnant wind area and quiet wind area appear in the skywell of an HUI-shaped building (the ratio of width to height D/H is 0.4), and the wind environment in the skywell is disordered. When one door of the building is opened, the wind speed distribution in the air field in the courtyard is uneven, and there are some wind whirlpools and quiet wind areas, so the wind environment is poor. When the building door skywell is fully open, the internal wind speed of the skywell is large and evenly distributed, meeting the outdoor wind environment standard in summer, and the wind environment is good. Therefore, this paper suggests that in summer, residents open the front and rear doors of the building to promote the entry of tunnel air into the building and strengthen natural ventilation; in winter, all the front and rear doors of the building can be closed to prevent cold wind from entering the building.

5. Conclusions

Through summarizing the impact of the traditional plain shape of a village building skywell, the skywell scale ratio, and skywell door openings on the outdoor wind environment in Huizhou, we can draw the following conclusions:

(1)

Among the three skywell shapes measured by the author, the width-to-height ratio D/H values of the AO-shaped architecture were in the following order: Shuting dwelling > Yufengfa dwelling > Dunren dwelling; the width/height ratio D/H values of the H-shaped buildings were in the following order: Lufu dwelling > Panxianxiong dwelling > Panmaotai dwelling; and the height/width ratio D/H values of the HUI-shaped buildings were in the following order: Yingfu dwelling > Gaozuo dwelling.

(2)

In the study of the effect of aspect ratio on the wind environment of the skywell, the following conclusions were drawn: the wind field in the skywell of the AO-shaped buildings will be optimal when the aspect ratio D/H is equal to 0.3; the wind field in the skywell of the HUI-shaped buildings will be optimal when the aspect ratio D/H is equal to 0.4; and the wind speed in the skywell of the HUI-shaped buildings is significantly better than that in the skywell of the AO-shaped buildings for the same aspect ratio. The wind field in the skywell will be best when the D/H ratio is equal to 0.4; the wind speed in the skywell of the HUI-shaped buildings is significantly better than that in the skywell of the AO-shaped buildings under the same width-to-height ratio.

(3)

In the study of the influence of the aspect ratio of the skywell on the outdoor wind environment of the building skywell, it was found that the HUI shape of the building skywell (width-to-height D/H value ratio of 0.4) is best within the aspect ratio L/W value range of 1–1.5.

(4)

In the study of the impact of the opening of building doors on the skywell wind environment, it was found that: opening the front and rear doors of the building can promote alley wind to enter the building courtyard. The skywell wind environment is good and can strengthen natural ventilation. When closing one door of the building, part of the alley wind will be blown into the skywell, while the airflow circulation is low because there is a static wind area. When closing all the front and rear doors of the building, the skywell is almost in a static wind state because the skywell wind environment is very poor.