Designing a facade by biomimicry science to effectively control natural light in buildings (Glare analysis)

The increasingly popular design trend of glazed facades using daylight in buildings has made it essential for innovations to eliminate the unnecessary intrusion of sunlight in spaces. This study aims to create a kinetic façade pattern by referring DNA structure and photosynthetic behaviour to mimic biomimicry science characteristics in Wallacei evolutionary software for generating possible patterns. Daylight glare thresholds were determined as an essential factor for user productive work. Comparison of three-building envelope potential was made for preparing spaces (zones A, B) in Bangkok condition; without a façade, with a static facade, and with a kinetic façade. DIVA software was used to analyse glare in terms of daylight glare probability (DGP). First, DGP (without façade) for zones A and B were 100% and 55%, or intolerable glare. Second, DGP (static façade) for zones A and B were 59%, 30%; zone A was intolerable, and B imperceptible. Third, DGP (kinetic façade) for zones A and B were 28% (imperceptible glare). Therefore, a kinetic façade has a high potential for protecting against unsuitable glare. These findings may serve as preliminary evidence for understanding kinetic façade potential for self-adjustment by light intensity to improve quality of life for occupant use of spaces.


Introduction
Buildings are now designed with either exaggerated proportions of transparent materials or large open spaces within a building envelope [1], [2]. For instance, in the images of Suvarnabhumi International Airport ( Figure 1) and a working space (Figure 2), there are various issues that can occur from excessive natural light, one of which is "intolerable glare". Glare directly affects human activities that use spaces such as these as working spaces, creation spaces, and relaxing spaces, yet this is particularly so in working spaces since people must focus their eyesight [3]. Natural light is crucial since it can make work more effective. In 2011, a poll was conducted among 135 architects, lighting designers, and consultants, which found that over 80 percent of participants deemed natural light to be either an important or extremely important design factor ( Figure 3). Glare factor from daylight was simulated for participants in various fields, while the simulations were integrated with luminance contrast measurements and direct sunlight identification in the users' proximity in a workspace [4]. Natural light is related to human health, especially eyestrain which is a common condition that occurs when the eyes get tired from intense use, such as staring at computer screens and other digital devices, IOP Publishing doi: 10.1088/1757-899X/1148/1/012002 2 reading books, or working. Almost all such problems occur as a consequence of unsuitable light, including natural and artificial light that can result in eye discomfort [5]. These issues arise since open areas of wall are unsuitable for a controlled climate. Therefore, the use of spaces are inefficient to ensure that natural light remains suitable for human activities in such spaces. This issue prompted the researcher's interest to suitably control the atmosphere in buildings, specifically the glare factor which affects human comfort, by using the faç ade to control this problem.      Figure 4 separates this into two issues, insufficient light and inappropriate glare. First, an insufficient light issue affects both the use of spaces and human health aspects. The muse of space aspect is then divided into three issues: (1) unusable space; (2) (2) physical health, including eye strain or eye discomfort, and headache. Second, problematic glare issues also affect both aspects [6]. Nevertheless, some minor issues are not included, namely the accidental and mental health aspects.

Study purpose
The research studies and mimics natural form and movement. This study focuses on DNA structure and the phototropism phenomenon to find a distinctive point by interpreting its physical appearance. Kinetic faç ade technology is developed to effectively controlled a suitable interior atmosphere that has natural light in indoor spaces in order to encourage work efficiency and human health. The purposes of this study are as follows:   Figure 5 presents the overall research methodology to determine how a faç ade directly affects the glare factor. The method is divided into two phases, the faç ade design phase and the natural light analysis phase. The faç ade is validated in three types of building envelope, that is, kinetic faç ade, static faç ade, and building without a faç ade (only glazed faç ade) through testing with the DIVA software that analyses in terms of natural light. This research focuses on glare analysis to improve visual quality in the space. Rhino and Grasshopper computational methods were used to design faç ade optimisations for the simulation. Rhino software is used to build 3D forms while Grasshopper software uses two plug-ins -DIVA to analyse natural light and Wallacei to generate faç ade patterns. The kinetic faç ade integrates with biomimicry science to design the physical and behavioural components. The physical component uses characteristics of DNA structure to generate probability patterns and phototropism behaviours of plants, and is applied in the faç ade movement to respond to sunlight. Then, both inspirations are interpreted into faç ade patterns to generate probability patterns in Wallacei software and the potential of each faç ade type is analysed in DIVA software to optimise for the glare factor. Data was collected to compare the effectiveness of building envelopes with a static faç ade, kinetic faç ade, and without a faç ade to simulate the use of space in the same zone by focusing on the glare factor. From Figure 5, the variables are as follows:

Dependent variables
 Natural light, focused on the glare factor.

Analyse potential faç ade patterns (Phase 1)
This section analyses potential faç ade patterns inspired by physical DNA and phototropism behaviour to construct the strip form. Wallacei is used to generate potential faç ade patterns, which is an evolutionary algorithm software to generate algorithms forms. Evolving algorithms are a series of optimisation algorithms that imitate the mechanisms of natural evolution [7].  Photograph: Ann Ronan Pictures/Getty Image Figure 6 shows the processes that mimic strip DNA. This study mimics two parts for a flexible movement purpose; the physical DNA structure and phototropism behaviour. Phototropism is most commonly found in plants and involves the development of an organism in response to a light stimulus (external factor) by photoreceptors related to signalling mechanisms. According to external factors, plants adaptively change their growth characteristics according to the amount of accessible sunlight [8]. This mechanism is crucial for interpreting the faç ade strip mechanism since it must adjust itself to the sunlight, similar to plants.
Phenotypes are measurable traits of an individual that derive from the association of its genotype (total genetic inheritance) with the environment. Features that may be observed include behaviour, biochemical properties, colour, shape, and size, for example the bird species depicted in Figure 7 [9]. Thus, the phenotype's simple definition are external appearances or characteristics that are easy to observe.  Figure 8(a) shows the strip phenotype probability patterns, resulting in 450 solutions from generation 0 to generation 8. Each generation has different and repetitive patterns compared to other generations. For instance, generation 0 and generation 5 have some strips with the same patterns. The strip patterns that occur from the phenotype are different, resulting in various strip phenotypes. The strip phenotype occurs from mimicking DNA strip and phototropism phenomenon to simulate the characteristic of strip patterns for application in the kinetic faç ade model. The generation of new solutions was stopped after 450 solutions had been generated, since when looking at the standard deviation trend graphs ( Figure 8[b]), the final generation had a smaller slope than the first generation. This means that the strip phenotype repeated the same patterns and has no benefit for the purpose of this research. Thus, the final generation at which the generation process was paused was generation 8. A low standard deviation factor indicates that most values are similar to each other (less variation within the population). In contrast, a high standard deviation means that the values are farther from the mean (greater variation within the population) [10]. When looking at the standard deviation graph, early generations had more varied strip patterns than the final generation, as Figure 9 shows the probability output in 50 solutions of strip phenotype in generation 3, indicating greater repetition of strip patterns than the early generation. However, it can be seen that this generation has various strip phenotypes when observing the twist angle characteristics that respond to sunlight.    Figure 10 summarises the strip patterns from generation 0 to generation 8 that were applied in the kinetic faç ade model. These patterns were selected since they respond to sunlight differently, similar to phototropism behaviours. The various angles of the strip patterns have differing potential to block sunlight at different times. Thus, it is beneficial for adapting to user needs in various situations when living in spaces. When there is less light intensity, the strip will be more twisted than the high light intensity, similar to a strip 7 as in Figure 10.  Figure 11 presents the probability of faç ade patterns in generation 0. The solutions that the researcher exported is 50 out of 1150 solutions. Indeed, the phenotype generates into 23 generations, including generation 0 to generation 22. When comparing the kinetic faç ade from generation 0 to generation 22, the results are similar to generating DNA strips when considering standard deviation, fitness values, and standard deviation value trendline which are consistent. Therefore, this output can predict that the kinetic faç ade can twist or adjust to sunlight, similar to the strips in Figure 10. Figure 12 shows the graph trend that can predict the possibility of the kinetic faç ade characteristic. First, standard deviation represents the distribution of a set of values from the mean [11]. The chart's objective is to present and analyse the levels of variation and convergence for each generation in the population. Increased variation is represented through a 'flat' curve, while increased convergence is represented through a 'narrow' curve. Thus, it can predict that each generation's faç ade patterns are independent since each generation's curve is similar. Nevertheless, each generated faç ade has a range of patterns, for instance the early generation's faç ade patterns have various phenotypes when compared to the last generation. Second, the fitness values chart analyses the fitness values for each fitness objective independently across the entire population. The aim is to visualise how the solutions perform in relation to one another, both within each generation and across the population. When analysing the graph, most of the generation is independent since most line graphs are relatively stable. However, the early generation has a fluctuating line which means that the phenotypes or external appearance are more diverse than the stable line. Third, the standard deviation trendline chart presents the standard deviation value for each fitness objective independently, and for each generation across the entire simulation from start to finish. The aim is to highlight specific trends in the variation and convergence of each generation across the population. Thus, after analysis, the graph has a similar meaning to the standard deviation graph. This illustrates that the early generated patterns were more varied than the last generation or the early generation has greater diversity than the last generation.

Analysis of natural light ambient (Phase 2)
This part analyses the building envelope in working spaces with a static faç ade, kinetic faç ade, and only glazed panels (without faç ade). The kinetic faç ade refers to cases generated in phase 1 to simulate effective glare protection. Figure 13(a) illustrates the setting model, which is the working space for simulation by the node. The node or dot is crucial for calculating the glare factor, hence the researcher set the node position close to the human eye at a distance of approximately 1.10 metres when doing work activities as in Figure 13(b). Figure 13. Model setting to simulate the glare factor using the node.   Glare occurs when part of an interior space is significantly brighter than the room's luminosity. Harm can occur when the eyes are exposed to a more intense light than they are normally adapted to. The most prevalent cause of glare arises from the light source and window being in the same visual area, either directly, through reflections, or both simultaneously [12]. Daylight glare probability (DGP) was established in a new glare prediction model since existing glare indices did not accurately estimate the degree of daylight glare irritation in a working atmosphere with regular work practices and diverse non-uniform sources of glare, such as the Venetian blind system [13], [14]. GDP was derived from laboratory results in private office spaces containing human test subjects.  [15]. Glare types that negatively affect humans are intolerable glare and disturbing glare. Thus, designers should avoid and protect against these types of DGP since they will occur if spaces have an exaggerated open space or glazed faç ade. Figure 14 shows the 3D model and zoning for simulating the glare factor for views A and B. These zones were selected since they have different light intensity, with view A receiving greater light intensity than view B which is a narrow space. The simulation must control the input data as follows: location; climate; and period. This study is based in Bangkok, Thailand, with clear skies and during the Winter Solstice period (22 December) between 7.00 AM and 5.00 PM. During the Winter Solstice period in Thailand, the building can receive more direct sun than during other periods, making it suitable for testing. Figure 15 illustrates a glare analysis with a building model without a faç ade. For View A, the result is intolerable glare and the daylight glare index versus the percentage of persons disturbed (DGP) is 100%. View B shows the glare analysis in another area, which resulted in intolerable glare with a DGP of 55%.      Figure 19. Glare analysis with a building model (kinetic faç ade). Figure 18 shows the model and the zone of views A and B to simulate the glare factor. In Figure 19, view A shows the glare analysis with the kinetic faç ade building model, which indicates imperceptible glare and a DGP of 28%. View B illustrates glare analysis with the result of imperceptible glare and a DGP of 28%. Figure 20 illustrates a comparison of the three-building envelopes in terms of glare factor. The result is that a kinetic faç ade can better protect against the glare factor than a static faç ade, which is useful in some areas and inappropriate glare in some periods (not stable glare factorview B).

Figure 20.
Comparing the potential of the three-building envelopes in terms of glare factor.

Conclusion
One crucial factor to encourage users to work effectively in working spaces is the suitable glare factor, since this can negatively affect people and reduce their working effectiveness. Thus, this study focused on simulating the potential of a kinetic faç ade by comparing it with a static faç ade and a building without a faç ade (only glazed panels). The kinetic faç ade uses biomimicry science to attain a faç ade pattern by mimicking nature, both in physical appearance and behaviour. The kinetic faç ade was inspired by nature, including DNA structure and the phototropism behaviour of plants. Thus, these inspirations must interpret a physical appearance for simulation in the evolution stage to discover the potential faç ade patterns by using Wallacei. After simulating the faç ade patterns, the next process simulated each building envelope's effectiveness in the preparing space, from views A and B (  Figure 15, Figure 17, and Figure 19). First, the DGP of the building without a faç ade from views A and B was 100% and 55%, indicating intolerable glare and unsuitability for use as a space. Second, the DGP of a building with a static faç ade from views A and B was 59% and 30%, with view A having intolerable glare and view B subject to imperceptible glare, indicating that the space is suitable for users only from view B (narrow space). Third, the DGP of a building with a kinetic faç ade from views A and B was 28%, with both views subject to imperceptible glare. Hence, both areas are suitable as working spaces. The building with a kinetic faç ade has the potential to protect against glare. Furthermore, if the faç ade receives less natural light, it will twist itself to receive outdoor light following the same pattern as the seventh strip in Figure 10. Thus, the kinetic faç ade is a crucial element that should be installed in buildings with exaggerated proportions of open spaces and glazed panels.  49400  44200  39000  33800  28600  23400  18200  13000  7800  2600   Lux   36100  32300  28500  24700  20900  17100  13300  9500  5700  1900