How topology optimisation works
Where the name comes from
Topology
tópos means place. lógos means study. Together, the study of place or placement study.
Optimisation
The act of making the most effective use of a situation or a resource.
Three ways to optimise a part
Not all optimisation is the same. Topology is the most powerful of the three because it is free to change the entire shape.
Sizing
Changes a dimension, like the thickness of a beam or a cross-section, while the shape stays fixed.
Shape
Moves features around, like the position of a hole, without adding or removing material wholesale.
Topology
Redesigns the whole shape, including where holes and voids appear. The greatest freedom and the biggest weight savings.
The working principle
The design space is divided into a mesh of small elements. Each element is asked a simple question. For the loads on this part, do you carry stress, or are you along for the ride. Elements that carry load stay. Elements that do little are removed. Run that judgement across the whole mesh, again and again, and material collects along the true load paths while the rest disappears. The load can be mechanical or thermal. The logic is the same.
What comes out is not a neat part. It is a raw density field that shows where material wants to be. Turning that into something manufacturable is the engineering, and it is the step software cannot do for you.
Why it speeds up development
The commercial reason to optimise early is fewer expensive loops later. Getting the shape close to right before CAD takes redesign steps out of the process.
The Method Step by Step
This is the real workflow behind an optimised part. The renders below are from the award-winning motorcycle swingarm, taken through each step.
1
Define the design space
First decide what is allowed to change and what is not. Bearing bores, mounting faces, and interfaces to other parts are fixed. Everything in between is space the optimiser is free to fill or empty. Get this wrong and the best possible answer is still out of reach.
2
Establish the load cases
Now describe every way the part is loaded in service. For a swingarm that means braking, accelerating, cornering, and hitting obstacles, each as a separate case. This is the hard part, and it is where most optimisations quietly fail. The maths is exact. The loads are an estimate. If the estimate is wrong, the part is confidently optimised for a life it will never live.
3
Set the optimisation targets
Decide what winning looks like. Minimum weight, a specific stiffness, or a target natural frequency to avoid resonance. The targets shape the answer as much as the loads do, so they are chosen deliberately against what the part has to achieve.
4
Apply the manufacturing constraints
A shape you cannot build is worthless. Casting and forging need draft angles and minimum wall thickness. Additive manufacturing removes most of those limits and opens up organic form. The production method is decided here, not after, because it changes what the optimiser is allowed to produce.
5
Analyse the result
The optimiser returns a raw shape, all load paths and large voids, that looks unmistakably like bone. This is the payoff of the method made visible. It is not yet a part. It is the honest answer to the question you asked, and it has to be read with judgement before it becomes geometry.
6
Detail it into a real part
The final step is engineering the raw result into a clean, buildable component with smooth surfaces and integrated functionality. For the swingarm, sixteen separate parts collapsed into one, with the brake system integrated into the structure. This is where the method turns into hardware.