This blog post is the continuation of a series introducing the concept of Computational Engineering. In the first two installments, Lin Kayser, the author, set the stage and provided an overview of the technology. In this post, Kayser goes deeper into the topic, exploring the different tiers of the technology stack for Computational Engineering.
Kayser begins by defining Computational Engineering as an emerging discipline that involves the development and application of computational models for engineering. These models, known as Computational Engineering Models (CEM), are logical structures that encode an engineer’s knowledge. The result is an algorithm that can generate various engineering designs based on different input requirements. Additional mathematical models can then be used to analyze the results and create feedback loops.
The author argues that the output of a CEM, while often a traditional 3D geometry, can also be directly used in manufacturing processes, such as CNC paths or sliced files for 3D printing. This allows for a seamless transition from abstract requirements to the production of physical objects.
Kayser then delves into the different tiers of the technology stack for Computational Engineering. He imagines a fictional top tier, a universal Computational Engineering Model, which would allow high-level specifications of physical objects without the need for detailed engineering knowledge. While this model doesn’t currently exist, the author sees a clear path towards its development.
Moving one level down, Kayser discusses specialized CEMs, such as Leap71’s RP/CEM for rocket propulsion. These models are sophisticated sets of algorithms that take high-level specifications in a specific field and produce manufacturable results with minimal user interaction. They use lower levels of the technology stack to automatically create complex geometries. Through parameter modifications and feedback from simulation or tests, specialized CEMs can generate objects within a wide parameter space, honing in on the desired capabilities.
The author emphasizes the importance of continuously refining these algorithms, with Computational Engineers making direct modifications to the code to improve results. A well-implemented CEM can capture the entire body of knowledge in a field, allowing a wide group of engineers to innovate on top of it.
Kayser’s partner, Josefine Lissner, will discuss how to build and apply a CEM in later parts of the article series.
The author also highlights the importance of a generalized computational model for engineering that can eliminate repetitive basic work and provide a library of code for engineers to build specialized solutions on top of. This model, ideally open-sourced, would enable engineers from all over the world to access and utilize a common set of tools and knowledge.
In conclusion, the blog post explores the different tiers of the technology stack for Computational Engineering and emphasizes the potential benefits of implementing these models in various engineering fields. The power of algorithms and software lies in their ability to capture and utilize knowledge, promoting innovation and efficiency in the field of engineering.
The world of engineering is constantly evolving, and one area that has seen significant development in recent years is computational geometry. In traditional CAD (Computer-Aided Design) software, engineers would manually create geometric shapes. However, with the emergence of computational engineering, engineers now have the ability to work on a higher level of abstraction, using generalized shape generators.
At the foundation of computational engineering is the geometry kernel, which is responsible for generating and manipulating geometries. A geometry kernel needs to possess two critical properties: robustness and speed. Unfortunately, many CAD kernels, which were designed for describing simple objects using complex mathematics, are not suited for the world of algorithm-generated complex geometries.
When my company, Hyperganic, was founded in 2015, we initially considered licensing a traditional CAD kernel and building our technology on top of it. However, we quickly realized that these kernels were not appropriate for our needs. Large software libraries, such as CAD kernels, often suffer from reduced robustness compared to simpler software solutions.
This realization led us to redefine our requirements for a geometry kernel in computational engineering. In addition to robustness and speed, we now placed emphasis on having a small footprint and providing a reduced set of functionalities that consistently produce verifiable results. As a result, a compact, robust, and fast geometry kernel is essential for the practice of computational engineering.
It’s important to note that the role of a computational engineer differs significantly from the traditional engineer. Instead of visually drawing objects on a screen or navigating complex networks of nodes, computational engineers primarily write and modify computer code to produce visual objects. This code creation process is increasingly being augmented by Large Language Models (LLMs), such as ChatGPT, to enhance efficiency and productivity.
For engineers who are not yet familiar with coding, I highly recommend learning this valuable skill. Contrary to popular belief, coding is not inherently difficult. In fact, I started coding at the age of 8, and I was by no means a child prodigy. Once you get started, you’ll find that coding is surprisingly simple, especially for engineers who already possess algorithmic thinking skills. Translating your algorithms into code can unlock a world of possibilities and make your work as an engineer more efficient and effective.
In conclusion, computational engineering is an exciting field that is changing the way engineers work. By employing a compact, robust, and fast geometry kernel, computational engineers can generate and manipulate complex geometries with ease. Additionally, embracing coding as a skill can significantly enhance an engineer’s capabilities and open up new opportunities for innovation and problem-solving.
(Note: This blog post is part of a six-part series by Lin Kayser, the CEO of Leap71, introducing the concept of Computational Engineering. Links to the other parts of the series can be found in the editor’s note at the end of this article.)
“Why did the 3D printer go to therapy? Because it had too many layers of unresolved issues!”
0 Comments