Page 147 - JCAU-6-2

Journal of Chinese
Architecture and Urbanism RuiXue Multi-Hall in reciprocal structures

Figure 3. Boundary opening optimization. Source: Drawing by Yingzi Hu

primary entrance (E1) was reduced from 8.2 m to 3 m. The considered four primary parameters involved in the initial
openings E3–E6 are situated in the transitional region of shell formation: ground anchor points, axial anchor points,
the outdoor gray space and are not enclosed by additional elasticity, and load. In addition, the parameter governing the
vertical elements. E3, E5, and E6 serve as secondary boundary pull-back curve was introduced. Subsequently,
exhibition entrances that establish visual connections by numerically adjusting these parameters, as depicted
with the enclosed courtyard and the expansive skyline. To in Figure 5, the internal spatial design of the building was
achieve this, the net height of these openings was reduced refined based on empirical and structural considerations.
from 3.8 m to 2.4 m, a sufficient dimension for secondary It is important to note that ground anchor points, axial
passages and visual communication. anchor points, and the pull-back curve remain constant

E4, on the other hand, is positioned at the steps of and do not undergo further numerical adjustments. The
the outdoor theater grandstand. The overall slope of the remaining parameters of load (W), length deformation
opening mirrors the step-by-step retreat of the grandstand coefficient (L), and elasticity (S) are selected as the three
and is designed to accommodate gatherings of people in adjustable factors for comparative optimization.
the courtyard. Consequently, the height of this opening Before modifying the shell surface parameters, the
was adjusted from the initially calculated value of 5.6 m to height-to-span ratio for the primary exhibition hall was
a net height of approximately 3.9 m, ensuring suitability for established at 1:5 to achieve an optimal spatial effect. With
viewing and performance. a main exhibition hall span of 34 m, the estimated height
After optimizing the height function of the boundary of the north main shell space, excluding the preset 0.6 m
control curves, it is essential to further rationalize the of the structural layer, amounts to approximately 7.4 m.
three-dimensional boundary. This rationalization process Subsequently, for the four smaller subspaces (S3, S4, S5,
ensures that each segment of the curve boundary can exist and S7), the forming height needs to be maintained around
on a tilted or horizontal plane in one dimension, facilitating 3.2 m after accounting for the structural layer (0.6–0.8 m) to
subsequent construction positioning. This process can be meet the desired net space height requirements effectively.
achieved by following the rationalization steps typically The parameter W, which represents roof load,
applied to plane boundaries. It involves smoothing and was initially studied while keeping other parameters
simplifying the three-dimensional curve boundaries constant. Three values were considered: 60, 80, and 100.
through the continuous blending of low-order rational It was observed that the shell space obtained when the W
curves. In the end, a combination of three types of curves parameter was set to 80 was the most reasonable. Under the
is utilized, which includes three-degree plane curves, two- other values, the shell space was either too low or too short,
degree circular arcs, and straight-line segments (Figure 4). making it difficult to accommodate internal functions
An algorithm-parameter analysis was initiated to optimize and leading to spatial redundancy. Once the roof load
the internal space within the shell structure. This analysis parameter stabilized at W = 80, the length deformation

Volume 6 Issue 2 (2024) 5 https://doi.org/10.36922/jcau.1635