Optimized Simulations Using Component Scans

Digital Twins of Lightweight Components Created Through 3D Scans of the Fiber Structure

Whether in automotive and mechanical engineering or in aerospace, lightweight components made of fiber-reinforced plastics must withstand high loads while keeping weight as low as possible. The internal structure of the material – particularly the orientation of the fibers – is crucial to their stability.

More precise simulations and more accurate predictions of the properties of lightweight components are the result of our researchers’ collaboration with the Danish company Xnovo Technology. Together, a project team is investigating how the real, manufacturing-related microstructure of an injection-molded component can be converted into numerical models using scans. The goal is to predict the actual load-bearing capacity of components with significantly greater precision – before they fail in service.

For the project, injection-molded components were digitally scanned using a scanner from Xnovo Technology. The scan thus serves as the basis for digital twins that are as realistic as possible, providing insights into aspects such as the load-bearing capacity of components. While injection molding simulations have primarily served as the basis up to now, the new approach allows for the direct use of real image data. This makes it possible to significantly refine conclusions regarding the fiber orientation within the component.

Background: Limited Options for Characterizing Fiber-Reinforced Plastics

Until now, characterizing fiber-reinforced plastics using imaging techniques has been limited. Typical challenges in visualization include:

  • Conflict between component size and resolution: Larger components can often only be imaged at reduced resolution.
  • Destruction of samples: To make individual fibers visible, components often have to be cut into small samples. In the process, the actual fiber architecture of the entire component is lost.
  • Complex image analysis: The segmentation of individual fibers or fiber bundles from pure absorption data is computationally intensive and associated with uncertainties. This directly affects the quality of material models and simulation results.
Absorption and Fiber Orientation
© Fraunhofer ITWM
Absorption and Fiber Orientation
Full-component 3D Mapping with FiberScanner3D
© Fraunhofer ITWM
Full-component 3D Mapping with FiberScanner3D

FiberScanner 3D: 3D Mapping of Fiber Structure

For reliable digital twins, the three-dimensional distribution of fiber orientation throughout the entire component is crucial. That is why the project partners decided to use a scanner that enables complete 3D mapping and visualization of the fiber structure. This measurement data forms the basis for simulation-ready material and structural models here at Fraunhofer ITWM.

The FiberScanner3D method from Xnovo Technology combines absorption and scattering data from X-ray measurement. This multimodal analysis allows the local fiber orientation distribution to be determined for complete components up to a size of 150 × 100 × 80 mm³.

This opens up the following possibilities:

  • 3D fiber orientation in every voxel
  • Local fiber volume fraction
  • Fiber anisotropy
  • Pore and inclusion analysis for voids and defects
  • Nominal–actual comparison between CAD geometry and actual geometry
  • Mesh generation for high-resolution FE models

Case Study: Digital Twin of an Injection-Molded PPA/CF Component

A case study illustrates the creation of a high-precision digital twin using a unique workflow for 3D fiber mapping of the entire component. The resulting model non-destructively captures the actual manufacturing state of the component, taking into account manufacturing-related deviations in shape and fiber orientation. Both the real component geometry and the three-dimensional distribution of fiber orientations throughout the entire volume are reconstructed from the FiberScanner3D scan.

 

Simulation Results: Influence of Fiber Orientation on Stresses and Forces

The results are clear: Incorporating fiber orientation data has a significant influence on the predicted stresses and reaction forces. The simulation without fiber orientation information significantly underestimates both the maximum stress and the reaction force. Modeling based on FiberScanner3D data predicts a significantly higher stress level.

This demonstrates that the fiber-reinforced composite structure is a key determinant of a component’s torsional strength.

Analysis of the fiber structure of components
© Fraunhofer ITWM
Analysis of the fiber structure of components

Convincing Results

3D scans of the fiber structure provide a highly accurate alternative to injection molding simulations. This is particularly valuable when precise predictions about component performance are required despite unknown manufacturing process parameters.

Finite Element Analysis
© Fraunhofer ITWM
Finite Element Analysis