By definition, a cellular solid is an interconnected network of struts or plates that form the faces of cells. Low-density cellular solids appear widely in nature. One example is in the porous areas of human bone, called cancellous bone. Bone organised in this configuration minimises overall weight while providing structural capacity to bear high loads. Based on the many biological models of cellular solids, synthetic materials have been created, such as honeycomb material used in the aerospace industry as lightweight structural components.
While cellular solids are abundant in other fields, there are few examples of architecture that incorporate their use. One prominent example however, is the Water Cube in Beijing, designed for the 2008 Summer Olympic Games. The Water Cube implements the Weaire-Phelan structure, a complex three-dimensional polyhedron structure, which is said to be a packing logic with the least amount of surface area possible. Even though the structure appears to be a random configuration, it actually uses just two different cells of equal volumes packed together, meaning that the structure does not take into account specific external forces from the surrounding environment. Thus the system is unable to represent the complexity and the adaptability of natural cellular solids.
The design started with the creation of a linked network of functions. These functions were then divided into different types of zones, and connected with specific circulation links. A branching network algorithm was then used in order to generate an adaptive network. This same logic was used and developed further to create what can be called a Hierarchical Branching Network algorithm. This algorithm allows the network system to memorize different variables of rules at each branching node and store them in different hierarchies. The first network function is then stored into the desired hierarchies and combined with the previous algorithm. This resulted in an initial set of points consisting of different functions and hierarchies with the ability to maintain connectivity and adapt to the insertion of new functions.
Each point in the hierarchical branching network represents a function; a Voronoi algorithm is used to generate compartmental spatial arrangements from these initial sets of points. However, the limitation of this system is its inability to divide each space into the required area necessary for each programmatic function. Therefore, a Cell Ratio algorithm was developed in order to readjust cell size and meet the specific spatial requirements for each cell.
Natural cellular solids have the ability to adapt to specific external forces within their surrounding environment. In this design, therefore, the sun is simulated in computational experiments as an external environmental force, generating site-specific geometry. Beginning with two-dimensional geometries, an algorithm was developed, called a Global Sun Exposure algorithm, which takes the data of the sun’s path on the site as a parameter in the generation of cell formation, minimizing sun exposure throughout the entire body of the object. The second algorithm, a Local Sun Exposure algorithm, calculates and compares the area of the cell bodies that have a high ratio of sun exposure with those with low sun exposure. The cells then readjust their positions toward the direction of the sun in order to minimize exposure.
Following the generation of two-dimensional patterns, and experiment was conducted by overlaying a rectangular grid as partitioned space in order to understand the principles of transforming this data into three dimensions. Each grid cell was then transformed into a singular unit with a continuous surface, connected with each neighbouring cell’s wall. Two spatial zones were then generated by manipulating this continuous surface in both the horizontal and vertical planes. Based on specific programmatic requirements, cell depth and volume were then specified, creating a differentiated cellular solid structure. The resulting structure was generated with two phases of development: the first phase concerned with cell formation and it adaptability toward specific environment data, while the second phase generated a refined geometry established by specific building needs. This multiple-hierarchical system exploits natural cellular solid structures’ adaptivity and responsiveness while adhering to material and site-specific constraints.