A Parametrically Printed Park Pavilion set in San Diego, designed for mixed-use; eating, sitting, hanging out. Inspired by the landscape of Sunset Cliffs.
Meta Caves is a temporary pavilion project developed by Lakzhmy Zaro and Hala Lahlou as part of the Structural Optimization Seminar during the first trimester of the MACAD Master’s program, completed in December 2025. The project explores the design of a temporary pavilion inspired by the sunset cliff’s natural park in San Diego.

The key inspiration for the Meta Caves project, developed using metaballs, was Sunset Cliffs Natural Park in San Diego, particularly its naturally wind-formed cavities and erosion patterns.
Intent + Inspiration

Computational Design System

Form Finding Process
Form finding starts with populating points which becomes the center of our spheres. The spheres are joined together with dendro from whish we get our mesh that is then refitted in the 8x8x6 box. Then the entry and exit are defined. The mesh is quad remeshed with a guide set at 27 inches that is also used to select the bottom faces of the mesh which is offsetted for the seats.

Our main form finding is then controlled via box populate and dendro metaballs. We did this to introduce randomness that mimic the organic geometries of the caves of Sunset Cliffs, where point populate is varied by a random seed and the spheres are given different radii

With our form, we then define our circulation by having points on either sides of our bounding box connected by a line that goes through our form. This allows us to find the closest point on the mesh from which splitting spheres are made from the 2 points.

Structural Analysis: Alpaca Forces & Material
Using Alpaca4D to analyze our structure and calculate mass and deflection to make a safe pavilion.
To analyze the structure using Alpaca, the lattice nodes in contact with the ground are defined as fixed supports, and gravity loads are applied. Wind loads are also introduced, considering that the dominant winds in San Diego come from the west–northwest (WNW) from the Pacific Ocean, with typical velocities ranging between 3 and 7 m/s. For the analysis, a wind speed of 11 m/s is applied to provide an appropriate safety margin. Additionally, seating loads are included to account for user occupancy.
Glass Reinforced Recycled PET was chosen because it delivers superior strength-to-weight ratio compared to steel while reducing transport costs, offering better durability and flexibility than carbon fiber at a more accessible price point.


Galapagos Optimization
The first pass of optimization looks into the overall silhouette of the form, the metaball. It was done using Galapagos, with genomes connected to the point populate seed and the circulation placement.

The second pass of optimization looks into the lattice structure. It was done using Galapagos, with genomes connected to the Crystallon lattice type and the number of mesh layers.


Form Iterations
We explored various form iterations that aligned with the aesthetic intent of the design, while ensuring that each iteration integrated seating within the pavilion.

Most Structurally Optimized Outputs





Conclusion
This project presents a parametrically designed 3D-printed pavilion for San Diego, drawing formal inspiration from the erosion patterns of Sunset Cliffs. The pavilion serves as a multifunctional public space that accommodates dining, seating, and social interaction. Our computational methodology integrated form-finding processes with voxel crystallization through Crystallon, structural analysis via Alpaca4D including wind load considerations, and system optimization using Galapagos. The resulting design demonstrates how advanced parametric tools and additive manufacturing can translate natural landscape phenomena into functional public architecture that responds to both environmental forces and community needs.