Design Optimisation

Good enough is rarely good enough. You need a product that is lighter yet just as strong, a process that runs more efficiently, or a component that fits a tighter design space without compromising performance. Trial-and-error prototyping is slow and expensive — simulation-driven optimisation lets you explore hundreds of design variants systematically and find the best solution in a fraction of the time. We combine FEA, CFD and advanced optimisation algorithms to get you there.

Problems we solve

Our clients turn to optimisation when a single analysis is not enough — when the question is not just "does my design pass?" but "what is the best design?". Typical projects include:

  • Weight reduction — remove material where it is not needed while maintaining or improving strength, stiffness or fatigue life targets.
  • Performance maximisation — improve flow efficiency, heat transfer rates or structural stiffness by systematically varying geometry, materials or operating conditions.
  • Design space exploration — understand how your design responds to changes in key parameters and identify which variables have the greatest impact on performance.
  • Fitting a design into a confined space — find the optimal shape or material layout when packaging constraints leave little room for conventional design approaches.
  • Reducing manufacturing cost — demonstrate through simulation that thinner plates, fewer stiffeners or a different material grade still satisfy all requirements.
  • Robust design under uncertainty — ensure that your design performs reliably even when material properties, manufacturing tolerances and service loads vary within their real-world ranges.

Parametric optimisation

Parametric optimisation is the workhorse of simulation-driven design. We define the parameters you want to vary — dimensions from CAD, material properties, loads, wall thicknesses, even discrete choices like commercially available plate sizes — together with the performance targets you want to achieve. The optimiser then automatically runs and evaluates the necessary simulations to find the combination of parameter values that best meets your objectives.

This approach works with any physics we simulate: structural FEA (minimise weight for a given strength target), CFD (minimise pressure drop for a given flow rate), thermal analysis (minimise peak temperature), or a combination of these for multi-physics problems.

Parametric optimisation results showing radial stress contour plots for multiple flywheel design variants
Parametric optimisation of a flywheel. Each variant represents a different combination of geometric parameters, and the radial stress contour plots reveal which design best balances strength and weight.

Topology optimisation

When you do not yet know what shape your component should have — or when you want a fundamentally new concept rather than an incremental improvement — topology optimisation provides the answer. Starting from a maximum design envelope, the algorithm removes material that does not contribute significantly to the structural performance, producing an optimised material distribution that meets your stiffness, strength or frequency targets at minimum weight.

The resulting organic shapes are ideally suited for additive manufacturing (3D printing) and casting processes, and often outperform conventional designs by a significant margin. Topology optimisation is particularly powerful in aerospace, automotive and medical device applications where every gram of saved weight translates directly into performance or cost benefits.

Design of Experiments and sensitivity analysis

When your design has many input parameters, evaluating every possible combination is impractical. Design of Experiments (DOE) techniques allow us to cover the full design space with a scientifically selected, much smaller set of simulations — without sacrificing insight into parameter interactions.

From the DOE results we build response surfaces: mathematical models that predict how your design will respond to any parameter change, instantly and without running additional simulations. These response surfaces drive sensitivity analysis (which parameters matter most?), trade-off studies (how does weight relate to stiffness?) and goal-driven optimisation (find the parameter set that simultaneously satisfies all your targets).

3D response surface showing how an output variable changes as a function of two design parameters
A response surface visualises how a performance output varies with two design parameters. The surface enables instant exploration of the full design space and guides the optimiser to the best solution.

Robust design and Six-Sigma analysis

An optimised design that works perfectly under nominal conditions can still fail when real-world scatter enters the picture: material batches that vary, manufacturing tolerances that stack up, service loads that differ from specification. Six-Sigma and Monte Carlo methods quantify the probability that your design meets its performance targets across the full range of expected variations.

The result is a robust design — one that is not just optimal on paper, but reliable in practice. We identify which sources of variability contribute most to performance scatter and recommend where to tighten tolerances (or where you can afford to relax them) for the best balance between cost and reliability.

What you receive

Every optimisation project results in a clear report documenting the design parameters, the optimisation setup, the explored design space (with sensitivity charts and response surfaces where applicable), the recommended optimal design and how it compares to your baseline. We provide all the information you need to implement the improved design directly into your development process.

Ready to get more out of your design?

Whether you want to reduce weight, improve performance, explore a new concept with topology optimisation or ensure robustness against manufacturing scatter — our optimisation specialists will set up the right approach for your project.

Get in touch for a free initial consultation. We will review your design challenge, identify the optimisation potential and provide you with a clear project proposal.

 Contact us  or call us at +32 478 618 118

Frequently asked questions

Common questions about simulation-driven design optimisation.

We use the full Ansys simulation suite — Ansys Mechanical for structural optimisation, Ansys Fluent and Ansys CFX for flow-based optimisation, and Ansys Aqwa and Autodyn for specialised applications. For fatigue-driven optimisation we couple these with nCode DesignLife. We also use Matlab and Python extensively for custom optimisation workflows, DOE automation and response surface generation. The real value, however, lies in formulating the right optimisation problem — choosing meaningful objectives, constraints and parameters — which is where engineering judgement matters far more than the software.

Parametric optimisation improves a design that already has a defined shape by varying specific parameters — dimensions, wall thicknesses, fillet radii, material choices. You start with a concept and make it better. Topology optimisation finds the optimal material layout from scratch within a given design space. You start with a maximum envelope and let the algorithm determine where material should be placed. Parametric optimisation refines; topology optimisation invents.

That depends on the number of parameters and how they interact. A Design of Experiments (DOE) approach selects a scientifically determined set of simulations that covers the full design space efficiently — often tens to a few hundred runs, even when the theoretical number of combinations is enormous. From these runs we build response surfaces that predict the behaviour for any parameter combination instantly, without additional simulations.

Yes. Multi-objective optimisation is common — for example minimising weight while maximising stiffness, or reducing pressure drop while maintaining mixing quality. When objectives conflict, the result is a set of Pareto-optimal solutions that represent the best possible trade-offs. We present these trade-offs clearly so you can make an informed decision based on your priorities.

A design that is optimal under nominal conditions may fail when real-world variations enter the picture: material batches that differ, manufacturing tolerances that stack up, or loads that deviate from specification. Robust design (Six-Sigma) quantifies this by running the optimisation across a range of expected input variations using Monte Carlo methods. The result is a design that performs reliably in practice, not just on paper. It also identifies which sources of variability matter most, guiding decisions on where to tighten or relax tolerances.

Topology optimisation produces organic, sometimes complex shapes. These are directly suitable for additive manufacturing (3D printing) and investment casting. For conventional manufacturing (milling, sheet metal), the raw topology result serves as a concept that guides the engineer toward a simplified, manufacturable design that retains the performance benefits. We can apply manufacturing constraints during the optimisation (e.g. draw direction for casting, minimum member size) to obtain results that are more practical from the start.

A parametric study with a few design variables and a single physics type can be completed in one to two weeks. Larger DOE studies, multi-objective optimisations or topology optimisations with subsequent design refinement typically take three to six weeks. The main drivers are the number of parameters, the simulation time per design point and whether the underlying FEA or CFD model already exists.