Journal title : Investigation into the daylight performance of expanded-metal shading through parametric design and multi-objective optimisation in Japan
Author : Rendy Perdana Khidmat, Hiroatsu Fukuda, Kustiani, Beta Paramita, Ma Qingsong, Agus Hariyadi
Accepted July Received 1 November 2021; Received in revised form 3 February 2022; Accepted 15 February 2022. Journal of Building Engineering 51 (2022)
https://doi.org/10.1016/j.jobe.2022.104241
Abstract :
This paper presents a parametric and Multi-Objective Optimisation approach to investigate daylight performance of the expanded-metal shading depicting the sky conditions in Kitakyushu, Japan. The study is an initial design method utilising a robust computational calculation aiming to verify expanded metal as an environmentally friendly building material concerning its function as a shading device. The proposed platform generatively iterates the logic of expanded metal as dynamic design variables, such as Bond, Strand, Length, Height and Angle, to achieve the optimisation goals of minimising Annual Sunlight Exposure (ASE) while simultaneously maximising Spatial Daylight Autonomy, Useful Daylight Illuminance (UDI) and the openness of the expanded metal profile. Ladybug and Honeybee were utilised in a parametric platform, Grasshopper, to perform daylight simulation, while Octopus was employed to optimise the results. A thorough analysis was conducted to determine the significance of each expanded-metal parameter in achieving the daylight and View objectives. The results showed that adopting the proposed framework successfully fulfilled the LEED v4.1 daylight requirements by reducing ASE by 100% and improving UDI by approximately 50% over the baseline model. Furthermore, the iteration process revealed a wide range of aesthetic patterns, providing another factor for designers to consider during the design decision-making process.
Research Methodologies :
The initial phase of the research started by formulating the research problem and defining the daylight metrics as the objectives. The second phase was establishing the modelling processes, the baseline model and the expanded-metal parametric definition. In this phase, the overall function of the modelling components was linked with one parametric-definition arrangement. The third phase was the making of daylight and the View-simulation system. The established model in the previous phase would be converted into simulated objects and undergo daylight simulation. Up to this point, the physical properties of the model, the climate and the urban context were incorporated. The fourth phase was the optimisation process, in which the system generated all possible design solutions and came up with the presumably preferable solution based on the optimisation setting. The process iterated design possibilities by multiplying the combination of each dynamic parameter movement that drove the expanded-metal shading-pattern configuration. Last, the process concluded by analysing and interpreting the data gained from the optimisation processes, in which close observation was conducted along with sensitivity analysis to investigate the role of each parameter.
Result and Conclusion :
Expanded-metal shading is a reliable material for supporting a passive daylighting design strategy. By iterating the design parameters based on the expanded-metal components defined in the early design stages, the designer can generate design solutions that go beyond what was envisioned before the design processes began. The success of the form-finding process can be observed where the optimisation procedure generated 3176 design solutions by incorporating relatively short simulations. These 10 design ideas demonstrated performance targets that meet the international daylight standard LEED V4.1 in ASE and sDA. The best design solution identified from the sensitivity analysis towards the 155 Pareto-front individuals was model number 1728.
This model was equipped with parameter configurations of 4-cm Bond, 12-cm Length/LW, 12-cm Height/SW, 4-cm Strand/W and 5◦ in Angle. . Compared to the baseline model, the best design showed a reduction in ASE1000,250 by 100%, sDA300/50% by about 4%, View by about 34% and an improvement in UDI100–2000lx by about 50%. The DGP test showed that the best design solution (rank one) performed a 0.081 lower DGP value, improving the category from an intolerable glare to a disturbing glare.