Awatif: A New Data Structure for Structural Engineering
At its core, Awatif is a data structure. It's born from a fundamental question: Why, in an age of advanced computation, do we still lack a simple and effective data format for exchanging structural models?
This isn't just an academic problem. A clear data structure is the essential infrastructure for collaboration, whether with other humans or with AI. In the end, it is all about data: ideas can be represented in texts, visuals in pixels, music as signals. Yet, when it comes to geometry, we haven't agreed on a unified structure. Formats like IFC are often unnecessarily complicated, hindering progress rather than enabling it.
Awatif addresses this fundamental problem by drawing deep intuition from mathematics.
The Awatif Data Model: From Idea to Analysis
In mathematics, geometry has two primary representations: smooth and discrete. Awatif builds its data model on a clear distinction between these levels, creating a logical hierarchy from engineering intent down to the numbers a solver can read.
1. The Semantic (Parametric) Level
This is the "higher object representation"—the pure engineering idea. It's not geometry yet; it's the parametric definition of a component.
- Examples: A
Trussdefined by its span and height, aBuildingassembled from floors, or aRoofdefined by its structural system. - Purpose: This is where you expose parameters for parametric modeling, allowing for rapid design exploration and definition of high-level assemblies.
2. The Ideal (Smooth) Level
This is the "perfect," mathematical geometry generated from the semantic level. It's what designers and engineers think of as the "true" shape.
- Examples: The
primitives—perfectly straight lines, polygons (triangles, rectangles), and the "bent and twisted" forms they can take (curves, splines, NURBS, or subdivision surfaces). - Purpose: This level represents the exact, ideal geometry of the structural parts before any approximation is made for analysis.
3. The Discrete (Meshed) Level
This is the approximation of the ideal geometry, broken down into a "mesh" of simple elements. This is the only representation a computer can use to run physics simulations.
- Examples: A collection of 1D line elements or a 2D mesh of triangles.
- Purpose: This is the level for Finite Element Analysis (FEA). As you noted, we most often discretize (or mesh) the smooth representation to run physics.
You can always convert between these levels: meshing transforms the Ideal level into the Discrete level, and interpolation can reconstruct an Ideal-level surface from Discrete-level data.
Our Vision: Open, Collaborative Engineering
This three-level procedure is common in most FEM applications. However, in Awatif, we standardize it for structural engineering to finally enable transparent and robust collaboration.
This standard is the foundation of an open-source initiative. We are building examples to link the entire process, from analysis and design to the automated generation of reports and drawings. We encourage the community to build their own parametric models, custom reports, and drawing templates, sharing them in an easy, straightforward way.
With this vision, we hope to break the harmful monopoly that many structural engineering software companies maintain. We aim to make advanced FEM solvers and robust design libraries affordable and accessible, all verified and loved by a global community.
🚀 See it in Action
Wanna see all this in action? Check out the Awatif repository: https://github.com/madil4/awatif