Reverse Engineering Automotive Parts

Overview

Reverse engineering and analysing industrial automotive parts can be a challenge, given the complexity of materials used and the need for precision when processing image data. Researchers have used Simpleware ScanIP to overcome these difficulties in a project involving the conversion of CT data of a cylinder head into a high-quality 3D model suitable for inspecting defects and exporting simulation-ready meshes. A particular focus was placed on segmenting valve seats in order to carry out thermal simulation analysis and consider the influence of voids included when casting the aluminium part.

Characteristics:

  • 3D image data obtained from CT scan
  • Visualization and initial segmentation in Simpleware ScanIP
  • Morphological open filter and 3D editing tools used to improve valve segmentation
  • Processed image data meshed in Simpleware +FE Module
  • Simulation-ready meshes exported to LS-DYNA for thermal structure analysis

Thanks to

JSOL Corporation: M. Miyazaki

Visualization and Segmentation

Imaging data of the cylinder head was acquired using an industrial CT scanner at 0.4 mm x 0.4 mm x 0.5 mm resolution, and loaded in ScanIP. Volume rendering, opacity and colour mapping tools were used to visualize and reconstruct the data based on its underlying greyscale values. This led to the detection of small defects/voids in the cast. Further segmentation was carried out to separate valve seats from the cylinder.

3D Editing of Valve Seats

To segment the valves from the cylinder head in ScanIP, a combination of different image processing tools was used. Initial thresholding quickly created separate masks of the individual components. Morphological filters were then applied on selected regions of interest (ROI) to clean up the model. Using ScanIP 3D editing tools, a 3D ROI was created around the valve seats, with the morphological open filter used to remove small areas of noise. These tools enabled a more accurate and clearer multi-part segmentation of the valves from the cylinder head.

Results

The segmented image data was meshed using the Simpleware +FE Module and imported to LS-DYNA for a preliminary analysis of the impact caused by the presence or absence of voids in the cylinder head. This involved comparing the principal stress around the void for cases where a void was present or absent, with consideration made of intake and exhaust temperature. The approach enabled insights into the potential performance of the part under different manufacturing conditions, setting up future research into defect analysis.