Skip to main content

Complex Geometry Reverse Engineering. Impeller with Hybrid Modeling

Not all industrial geometries are created to be modeled through parametric features. Some — like an impeller, are the direct result of physical phenomena: flow, pressure, energy transformation. In these cases, reverse engineering can also be approached differently.

Instead of starting from the blade as a primary feature, one possible method is to begin from the flow itself. The internal geometry is interpreted as a continuous fluid path: a spiral volume that grows, rotates, and transitions from axial to radial. Within this framework, the blade is not treated as an isolated element, but emerges naturally from the subtraction of that flow. External surfaces can still be defined with clean, controlled parametric CAD. Internal geometry, however, may require a different strategy: SubD reconstruction, continuity interpretation, and final boolean integration. This is hybrid reverse engineering, where the method adapts to the nature of the geometry, not the other way around.

When this approach becomes relevant

This methodology proves particularly useful in scenarios such as:

3D scanning of an impeller
Accurate acquisition of complex geometries, including internal channels and transitional surfaces, without loss of critical detail.

Reverse engineering for engineering use
Transformation of scan data into a stable, editable model suitable for design iteration, simulation, or manufacturing.

Section extraction and technical evaluation
Reliable cross-sections for analysis, comparison, inspection, or performance assessment — based on coherent geometry, not approximations.

Preparation for 3D printing
Clean, watertight geometry ready for production, ensuring the component is not only visually correct, but manufacturable.

Beyond feature-based modeling

The objective is not simply to replicate a shape, but to understand how that shape is generated. When geometry is strongly influenced by physical behaviour, forcing it into standard parametric logic can lead to fragile models, difficult to control and unreliable in downstream processes. A more robust outcome often comes from combining:

  • technical interpretation
  • advanced surface management
  • multiple modeling paradigms used where they make the most sense

If this aligns with your case

Whether starting from:

  • a 3D scan
  • a mesh file
  • or a physical component

the goal is to obtain a model that is technically meaningful, stable, and ready for real-world applications For similar requirements or to discuss a specific case, further information is available on request.

Popular Articles