ARCH 655 600 Parametric Design

2025 Fall ARCH 655 600 Parametric Design

Project 2 Evolving with Machines

——Simulation and Optimization of Char House's Daylighting Performance

My Project 2 examined the daylighting performance simulation and optimization of Char House, an industrial heritage site located in Sugar Land. Char House is an iconic structure within the former Imperial Sugar Company complex in Sugar Land, Texas, often described as “the last classic char house in the United States.” Although the company ceased production in 2002 and surrounding buildings were demolished in 2010, the Char House remains standing today as a symbol of precious and rare industrial heritage. Adaptive reuse of Char House would not only preserve and extend this significant industrial legacy but also safeguard its associated historical and community memories.

Char House in 1925

Char House now

This project focuses on the daylighting performance of Char House and employs genetic algorithm tools to explore optimization possibilities for its interior lighting. Currently unused, the quality of Char House's interior daylighting remains unclear. Therefore, this research aims to guide adaptive reuse decisions through parametric optimization while respecting the constraints of industrial heritage.

First, modeling was generated in Rhino, with the fifth and sixth floors selected as test subjects to explore the feasibility of the research process. Subsequently, the ClimateStudio plugin in Grasshopper was used to simulate daylighting conditions for the existing space. ClimateStudio implements the IES LM-83 metrics of sDA (spatial daylight autonomy) and ASE (annual sunlight exposure). The specific metrics are sDA300,50% and ASE1000,250h.

Rhino models of the fifth and sixth floors (east elevation and north elevation)

The fifth and sixth floors were selected as the research targets primarily because they feature a unique overhanging space connecting these two levels. Secondly, the floor between them incorporates numerous circular openings reserved for connecting production equipment pipelines. Thirdly, this interconnected space is one of Char House's defining characteristics. During subsequent adaptive reuse, it can be converted into a public space, providing visitors with a more representative spatial environment for understanding this industrial heritage. The image above shows the actual site scan.（Source: https://projects.realityimt.com/imperialsugarbuilding/index.html）

Daylight Analysis Templates Provided by the Climate Studio Plugin in Grasshopper

The daylighting simulation was performed on the original model. As the result shown in the figure below, under the LM-83 standard, its sDA value reaches 92.23%, indicating excellent natural daylighting performance. However, the ASE value reaches 58%, far exceeding the acceptable threshold of 10%.

To address this outcome, the project introduced external aluminum sunshades/louvers to optimize daylighting for Char House. These aluminum sunshades enhance the building's daylighting performance without compromising the facade, effectively preserving the authenticity of the industrial heritage.

Nodes for generating the Louvers in Grasshopper

The image above shows the Grasshopper cell used to create sunshades/louvers. By applying the logic of dividing contour lines, each window was segmented. Extruding these contour lines yields the aluminum louvers used for simulation. The logic employs only two sliders to set parameters: the Number slider controls the depth of the blade, which equals the division distance of the contour lines. This logic ensures consistent blade dimensions across windows of varying sizes, maintaining visual coherence. The Angle slider controls the rotation angle of the louvers. Additionally, the logic extracts the highest point of each window as the starting point for blades arrangement, proceeding downward. This avoids using the scene's original height as the starting point, which could cause uneven distribution.

To achieve a solution closer to the optimum, the project incorporates a genetic algorithm. Galapagos within Grasshopper serves as the computational engine: the Number slider and Angle slider function as the Genome, while Python code defines the fitness function.

This code calls the sDA and ASE numerical outputs from ClimateStudio simulations. Under the LM-83 standard, sDA is treated as the primary optimization objective, while ASE is subject to an upper bound constraint. If ASE exceeds the maximum acceptable range by 10%, a penalty would be applied: a deduction calculated as the percentage overage multiplied by 5.0. This forces Galapagos to prioritize solutions that both enhance sDA and keep ASE within reasonable limits. A higher final fitness value indicates superior daylight performance. The penalty weight of 5.0 references common penalty function settings in multi-objective optimization, ensuring the magnitude of the penalty matches the primary objective's scoring range.

ClimateStudio + Galapagos cells

This project underwent two rounds of computation. The first round set Max.Stagnant and Population to 20 in Galapagos, with all other parameters at default values. This rapid validation of the workflow yielded the optimal result from the first round. This round ran for 30 generations over 4.5 hours of computation, producing an optimal F value of 86.29. This corresponds to sDA of 86.29%, ASE of 8.57%, a Number slider value of 0.5, and an Angle slider value of 2.

In the second round, the Max. Stagnant and Population were set to 50 and 40, with all other parameters remaining at default values. Compared to the first round, both population size and iteration count were increased to obtain more comprehensive results. This round ran for 38 generations and 12 hours of computation, yielding two optimal F values, both at 88.12%. These correspond to: sDA at 88.12%, ASE at 7.1%, with Number slider at 0.5 and Angle slider at 2; and sDA at 88.12%, ASE at 7.1%, with Number slider at 0.6 and Angle slider at 2.

Results indicate that under the original unshaded conditions, although sDA reached 92.23%, the ASE value hit a staggering 58.07%. This signifies extremely severe thermal radiation and glare risks indoors, failing to meet LEED v4 compliance requirements. After optimizing the louver parameters via genetic algorithms, we achieved a favorable trade-off between performance metrics. The optimized solution successfully controlled ASE within acceptable limits, eliminating the vast majority of direct glare risks. Meanwhile, sDA experienced only a slight reduction, maintaining a high level of 88.17%. This demonstrates that the optimized louver system effectively enhances Char House’s daylighting performance by filtering glare-inducing direct sunlight while maximizing the retention of beneficial diffuse light for interior illumination.

This project provides a reference for the adaptive reuse of Char House's industrial heritage. Based on the spatial characteristics of Char House within the context of industrial heritage reuse, the study combines parametric optimization with climate-based daylight simulation to propose an external shading strategy that enhances indoor lighting quality while respecting the building's historical form. Through a two-stage optimization process, the external louver system effectively balances natural daylighting and glare control without altering the original facade composition or industrial texture. This transforms the previously severely overexposed industrial space into a more suitable and comfortable indoor environment.

YouTube Link: