Innovation takes time. Fields like engineering or medicine have needed centuries to reach their current state. The path is not a straight line but involves leaps and revolutions. These leaps do not necessarily replace but add to and reconfigure earlier knowledge, often sparked by new discoveries or technological advancements.
This article focuses on innovation in engineering design, a field that arguably first emerged in primitive form around 5,000-4,000 BCE and is now entering the era of fourth-generation engineering design.
First-Generation Engineering Design
The first designers used experience and tradition to determine what to build and how to build it. The designer relied on what could be seen, touched, or told to understand their world. For thousands of years this process worked and sometimes produced astonishing results. We see fantastic structures like Theopetra, Stonehenge (5,000-4,000 BCE), the Pyramids (2,500BCE), the Parthenon (500BCE), and, later, the countless Roman viaducts, bridges, domes, and roads that dot our landscape. Driving from London to the channel crossing and on to Paris, you are following a Roman road which, though constructed using techniques vastly more sophisticated than Theopetra, was still a product of the same basic paradigm of experience-led, largely theory-less engineering. In this generation, knowledge is shared locally, and innovation is slow, though occasionally genius acts as a catalyst (e.g. Archimedes in Syracuse).
This long era of intuitive design lasted for several millennia and can be thought of as the First Generation in Engineering Design.
Second-Generation Engineering Design
At some point, around the 15th to 16th century, we entered the second generation of engineering design with a burst of technological creativity. For example, easily transportable and sharable technical drawings created a small information revolution that provided an impetus for innovation. The early works of Leonardo Da Vinci are still perfectly intelligible to an engineer today. Technical progress in engineering and design underwent a sudden acceleration, feeding off and in turn contributing to our expanding theoretical understanding of physical processes which lead to the industrial revolution. The modern idea of the “engineer” as a sort of applied scientist who uses a combination of mathematics and experimentation to construct machines and new structures also emerged in this period. The creations of this new class of engineer fundamentally transformed every aspect of our society.
This era constitutes the Second Generation in Engineering Design.
Third-Generation Engineering Design
The next significant paradigm shift comes in the 20th Century with the creation of the digital computer which would provide the basis for modern Computer-aided engineering (CAE). Electronic engineering drawings could be coupled with analysis tools (Finite Difference or Finite Element solutions) to design precisely calibrated products (cars, airplanes, skyscrapers etc.) in an iterative process. Design flows into analysis, which leads to improved design, in a virtuous cycle that we are now so familiar with. Engineers could test their designs, assess failure modes and costs before manufacturing the product. In this, the Third Generation in Engineering Design, engineers can iterate designs with relative ease, new products appear every day, and progress is continuous. In many ways, this is still the world we live and do our work in today.
However, human ingenuity does not stay still and we are now entering the Fourth Generation in Engineering Design, which will be characterised by a completely new, physics-driven engineering process.
Fourth-Generation Engineering Design
In this next generation the role of the engineer will be fundamentally transformed. If previously a designer started with draft designs which could be analysed for optimality, they will now start with physics and fundamental constraints and then use fourth generations systems to autonomously generate optimised designs. The engineer’s focus shifts from step-wise component design to higher level criteria which describe the system:
- Defining design objectives so that it meets the functional, safety and cost requirements.
- Evaluating and modeling the physical environment in which their design will operate.
- Working around limiting constraints, such as those imposed by manufacturing processes.
Making the leap to fourth-generation engineering is just a question of finding the right systems. And that is where ToffeeX comes in. ToffeeX is revolutionary physics-driven generative design software that empowers engineers to create high-performance designs faster and more efficiently by automating the design process using advanced algorithms and optimization techniques.
Key Features of ToffeeX
- Multi-Objective Optimization: ToffeeX can optimize for multiple objectives simultaneously. With ToffeeX, you can design for different performance parameters simultaneously, dispensing with the iterative process of 3rd-generation Engineering Design.
- Fast Turnaround: Traditional design processes can be time-consuming. ToffeeX delivers optimized designs in hours, significantly accelerating the development cycle.
- Simplified Workflow: ToffeeX requires minimal user input. Users don’t need to be experts in specific tools; they simply need to set their design goals and let the physics design for them.
From the idea to the final product in ToffeeX
Unlike traditional CAD and DfAM software, ToffeeX does not rely on human ingenuity to find the right geometries. Instead, it leverages the laws of physics and advanced optimization algorithms to explore possibilities beyond standard design patterns.
ToffeeX eliminates the need for multiple iterations between design and simulation, thus saving time and reducing the need for human expertise.
ToffeeX only requires a few inputs from the users:
- A Design Domain, i.e., the volume inside which the software can design an optimized component.
- A set of Boundary Conditions, to describe the physical phenomena involved during the optimization.
- One or more Optimization Objective(s), to define the goals of the design process
It’s possible to set up the generative design optimization problem with a few simple steps. For example, one of our customer’s required a new active heat sink design. Using ToffeeX we could set up an Optimization problem with the objectives of minimizing pressure loss and CPU temperature. Users could then easily define all the necessary boundary conditions and parameters to run the optimization exercise using our simple and intuitive interface. There’s no need to be an expert in Computational Fluid Dynamics (CFD), ToffeeX does that for you.
Results
- The air-cooled heat sink generated using ToffeeX proved to be 31% more efficient than the conventional extruded design.
- With its multi-objective optimization capabilities, ToffeeX allows users to create designs that can adapt to any requirements, whether it’s extracting more heat or saving energy by minimizing losses.
- Exploring different design solutions simply involves tuning the relative importance of one optimization target compared to the other, along with any other meaningful design constraints.
- Our customer tested four different parameter combinations for this heat sink before selecting the result that satisfied all their requirements. With each design iteration taking approximately 1 hour and 30 minutes, it took less than 7 hours of computational time and no more than 30 minutes of actual engineering time to go from an empty space to a ready-to-print geometry.
Welcome to the Fourth Generation Engineering Design. Let the Physics do the Work.
