SIZING
Gauge and cross-section
Optimize skin gauges, web thicknesses, and stiffener sections across the whole structure in one pass. Analytic gradients keep full sizing passes cheap enough to run over and over.
Structural optimization, rebuilt around iteration.
Size, shape, and grow aircraft structures in one scriptable workflow. PGD takes you from parametric model to sized design on your own hardware, with every gradient analytic and every input validated before it solves.
What it does
Run three design phases in one pipeline.
SHAPE
Geometry that stays buildable
Move the structural layout itself with shape variables that live at the sketch level. Geometric constraints keep every candidate valid, buildable geometry while the optimizer explores.
TOPOLOGY
Generative internal layout
Grow internal member layouts directly from the load paths with topological derivatives and stress-aligned Voronoi seeding. Sizing then turns the winning layout into a dimensioned design.
Compose the modes into one strategy: interleave sizing, topology, and shape cycles in a single run, and end on a layout that is already sized and buildable.
Why teams switch
What's different.
- Validate before you solve
- Catch a typo’d key or a dangling reference while you’re still authoring, when it costs seconds to fix. A model that passes validation runs in the solver and means what you said.
- Keep the workflow you already have
- Import and export Nastran-format bulk data, and open field output directly in ParaView. Models are plain, schema-validated documents you author from Python, so they belong in version control and code review like the rest of your work.
- Explore more of the design space
- Search the design space instead of polishing a single starting point. Analytic adjoint sensitivities and a parallel in-house finite-element solver keep each optimization cycle short.
- Keep your data on your hardware
- Run everything on your own machines. There is no cloud dependency and no license server to phone home, so your geometry never leaves your network.
Where we're headed
Built first for aircraft structures.
The methods are general.
We started with the hardest case we know: wing boxes and airframe structure, where buckling, frequency, and thermal constraints all bite at once. The same solver, sensitivities, and layout engine extend to other structural domains. If you're optimizing structure outside aerospace, tell us about your problem.
See it on your structure.
Request a demo and bring a model. The software makes a better case running on your problem than we could make in slides.