When designing a material system within the architectural field, the main purpose and the most basic analysis are about how the structure is able to transfer to the ground loads and forces and how efficient it can be in providing safe utilization by the users and achieving specific architectural solutions (they can be both structural and about aesthetic). Representing the first challenge in construction, structural behavior has always been needing experience, training and scientific studies.

According to these premises, it seems clear why during the last 500 years architectural practice moved from an holistic view that makes no distinction between structure, form and decoration, to a scenario in which the structure configuration comes mainly from an external professional that try to reach that specific architectural will. Nowadays we can detect currents that lead to a renewed vision of the design process, that takes advantage of all the disciplines available.

Computational techniques permit to change considerably the way in which the final shape and its performances will be influenced by modifying parameters as material, dimensions, supports, adding the possibility to create multiple structural relations within the same built environment. Needless to say, processing and scripting are only software and do not have to be considered as intelligent systems, but technical and artificial tools to reach a more productive and flexible working flow. Thanks to these it is possible to have a real time response of structural behaviors according to input parameters that can be modified in different configurations showing how the structure will work. What makes them really useful is the capacity of informing you in terms of data (stiffness, bending stresses, compression, tension, forces flows) that can be used to inform a new different analysis by the previous considerations. Simulating processes permits to save money, time and resources, while giving the freedom to focus on more complex relations.

In addition to the improvement of productivity and resilience of the process, computational techniques are used to create more and more complex relations between the parts of the system while making more tight and evident the connection between design and structure. In this situation, the two terms do not have different meaning or definition, but just the same motivation. That is why during the last years many studies have been developing regards the possibility to modify some material parameters in order to achieve better behaviors that are strictly related to the morphology and functionality of the design. One of the most competitive challenge is to decrease stress on the structural element thanks to its shape, position within the whole, material and self-weight, all of this features changing the approach to the construction.

For instance, the comparison between two structural approach as involving beams or bars in the project showed that, even if in both cases it is possible to conceive an intelligent material system, in absolute terms bars are more efficient. While beams host tension and compression, that merge in bending moments and shear actions, bars structure are conceived in order to present only axial stresses. Since these only moves in one direction we are able to lead the forces flow only in the pathway we want, toward to the ground, just connecting them through joints. From lightweight structures generated by this logic, it is also possible to reinterpret the classical beams through the utilization of bars that translate bending forces in tension and compression, reducing the overall behavior to a more efficient performance. Another strategy often used is exploiting the capacity of materials to be prestressed or preformed in particular shape that can be manipulate accordingly to the aim of the structure. In this sense, it is possible to play with the geometric pattern applied to the surface and directing forces and stresses along determined axis or curves that better fit the design goal.

Conventionally a truss consists of top and bottom members spanned between supports with diagonal members working in compression and tension. The aim of this assignment was to take up different phases of tessellations and finding, with the presence of only the diagonal members, can the geometry structurally function as a 3D truss and have stability.

Tessellations with triangulation were observed from relaxed flat 2D geometry to its peak 3D constellation. Model at 60%, 45% and 30% were analysed with karamba where 60% failed exceptionally to support the span of 20m.

The geometries with 45% opening were expected to perform best given the angle of members. However it has a large displacement and bending moment values as observed in the karamba results.

The tessellation with 30% open performed better than 45% open with a lesser displacement. This model was further taken to analyse with reduction of structural members-A and structural members-B. The analysis showed the structure displaced even less with far less mass with the absence of structural members-B. The cross section of this model was increased to match weight as in the 30% pure tessellation study.

Conclusion, the displacement was reduced by 300% with lesser members spanning the 20m space without member-B.

Analysis

The building is already extremely efficient. It is made by traditional construction bricks which are generated from two lines and three cosine curves, two of them in the XY plane and the top one in the XZ plane.
The utilization graphs show how smart the form is, since the roof, which is the weakest part of the structure (clear at the displacement graph), was folded to distribute the stress among all the surface. At this graph is showed that the folding edges of the roof are the more stressed ones, due to their horizontal. The Forceflow graph show how direct the building conduct the forces to the ground.
The aim of the first modification was to decrease the utilization of the edges of the folded roof, by increasing the amplitude of the lateral curves that generate the building. This was achieved and also was the almost total elimination of the displacement along the roof. In other hand, the forcelines became more curvy, due to the increase of the surface area at the folding walls, which caused the forces to travel in a bigger path.
The objective of the second modification was the same as the first one. This was made increasing the amplitude of the cosine wave of the roof. This modification also decreased the utilization of the roof, but concentrated it near the edges. Also, couldn’t be as efficient of the first modification in the decrease of the displacement, and the middle of the roof continued to be the region with the biggest movement. The only advantage from the first modification is that the shell force flow has a more fluid path.
At the third modification, it was inserted a series of supports as points at the bottom of the cosine curve of the roof, maintaining the original dimensions. This was to simulate a straight beam under the roof. With this support, the main displacement moved from the middle of the roof to its edges, and decreased the utilization of the structure, but maintained at the points in which the surface would touch the beam. Also, this additional beam created two new force flow paths at the roof.
The final modification is a variation of the third one, but instead of a straight beam, it was created supports simulating a curved beam located at the cosine curve at the top. Those supports also helped to divide the most fragile region of the structure, dividing it at two lines, along the roof. This is clear at the displacement, isolines of displacement, and shell force flow graphs. One interesting factor is that since the utilization of the roof was balanced, the inner edges of the folded wall became the most utilized areas of the structure.