Simulation-Driven Design, Generative Design, and Topology Optimization

Traditional methodologies in engineering and product design have changed drastically over the last generations and are becoming more inclusive of advanced, computer-aided techniques and concepts. Among these techniques and concepts, generative design, simulation-driven design, and topology optimization are common buzzwords we hear often.

Oftentimes these terms are used interchangeably or inconsistently at best. In this blog post, we will explore these terms, their differences, benefits, and applications in greater depth, then end with how Altair solutions can aid in these processes.

 Simulation-Driven Design

Simulation-driven design is an approach that integrates simulation and analysis directly into the design process rather than following a traditional, linear design process. This methodology allows engineers and designers to evaluate and optimize their designs through simulations at every stage of development, rather than relying solely on final analysis at the end of a project or physical prototypes.

A typical linear design cycle might look something like:

Fig 1: Typical Linear Design Cycle

This is a very simplistic idea of what a common design cycle might look like, but it does a good job of illustrating a linear process that relies on one stage to be completed before the next step can take place. The time it takes to test a finished product linearly, then return to the design stage happens many times during the design cycle, with each iteration leading to longer time to market.

With the concept of Simulation-Driven Design, the addition of simulation early in a design cycle can drastically reduce the backend design changes that are sometimes required with a result aiming at getting to market much quicker with a higher-quality, more innovative product.

Fig 2: Simulation and Testing occurring in same stage as design.

This isn’t a new idea and has been around in the industry for quite some time, but we wanted to be sure to define this concept a little more clearly since it has been applied inconsistently across the internet in the past.

Generative Design

Generative design is an approach that leverages artificial intelligence (AI) and machine learning algorithms to explore a multitude of design possibilities. Unlike traditional design, which relies heavily on the designer's intuition and experience, generative design begins with the definition of design parameters, specific constraints, and goals. Generative Design combines simulation and validation results to analyze and generate new design ideas before a final design is chosen and sent to manufacturing.

Fig 3: Generative Design Workflow

This process can result in more “out-of-the-box” design ideas, since it’s not relying on tribal design knowledge based on years of experience. It can also lead to less feasible design ideas if the proper variables aren’t considered up front. Like most simulation tools the term “Garbage in, Garbage out” can apply here. i.e. The more accurately you can set up your input parameters, the better result you will get in the end.

Fig 4: ABI Research Names Altair as Overall Leader in Generative Design Software Suppliers

Topology Optimization

Whereas the first two terms were general concepts and processes that are commonplace, Topology Optimization is an actual mathematical approach to engineering that focuses on optimizing the material layout within a given design space for maximum performance. It is particularly valuable for creating lightweight yet strong structures. Essentially, it helps a designer reduce the mass or volume of their design, while still maintaining the design intent.

How Topology Optimization Works

Define the Design Space: Engineers specify the allowable volume or space where the material can exist.

Set Objectives and Constraints: Objectives such as minimizing weight or maximizing stiffness are defined, along with constraints like load conditions and manufacturing limitations.

Optimization Process: The software iteratively removes unnecessary material using simulation results, while maintaining structural integrity, resulting in an optimized design.

While generative design, simulation-driven design, and topology optimization each have distinct processes and benefits, they are not mutually exclusive. In fact, these methodologies can complement each other to create highly optimized and innovative designs as seen below in Fig 5.

Fig 5: Relationship between the three concepts in a design cycle

Altair Solutions that Aid in these Processes

As mentioned above, these concepts are all things that Altair is actively supporting and looking for ways to build into the future. Let’s look at some of the existing tools in their portfolio and how they are being used in these processes.

 Altair HyperWorks

HyperWorks is the flagship product for Altair because of its strength in preprocessing and solving high-end CAE problems. One of the solvers in the platform, OptiStruct, is an industry leader when it comes to Topology Optimization. It performs Topology, Size, and Shape Optimization; Multi-Material Optimization; Topography Optimization; and many more techniques, all of which help engineers develop innovative and efficient structural designs that meet specific performance criteria.

Fig 6: Multi-Material Optimization in Altair OptiStruct

Also included in HyperWorks is Altair HyperStudy, a design exploration and optimization tool that supports traditional validation processes and Topology Optimization by allowing users to perform design of experiments (DOE) to ensure the optimal design. HyperStudy is multi-disciplinary and solver neutral, so seamlessly fits into most workflows.

Fig 7: Examples of Design Exploration results using Altair HyperStudy

 Altair DesignAI

DesignAI is similar to HyperStudy in that it excels as a design exploration tool. The difference between the two is that DesignAI is a cloud native app that creates optimal datasets, auto-selects the best machine learning model and enables the user to do quick what-if studies for fast, collaborative design improvements.

Fig 8: Altair DesignAI Interface with results

Altair Inspire

Inspire is a powerful, yet simple-to-use tool that includes the ability to create traditional CAD geometry, run simulation studies (including CFD), and now includes the ability to create implicit geometry for lattice creation. One thing Inspire has always been known for is that it includes the OptiStruct topology optimization technology, so not only can users create innovative new designs, but can also use the leading technology for optimizing them.

Fig 9: A complex lattice structure model with simulation results in Altair Inspire

Altair SimSolid

SimSolid was built with the term Simulation-Driven Design in mind. Its solver technology does not require the user to create a mesh or simplify geometry, it runs the simulation on the native or neutral CAD file. This speeds up the design iteration process, making it easier to evaluate multiple design options quickly.

Fig 10: Large assembly Model in Altair SimSolid

Hopefully this has helped clear up these commonly used terms and helped you understand how they relate to the design and testing cycles, but also how they relate to each other. As always, if you have any other questions, don’t hesitate to reach out.