91³Ô¹ÏÍø

Simpleware Case Study: Non-Destructive Inspection of AM Parts Using CT Data

Metal Additive Manufacturing (AM) has many advantages for producing topologically complex designs that cannot be easily achieved with traditional methods. However, non-destructive inspection and testing of the same components is often challenging due to inaccessible features, leading to missed defects and increased performance testing. As a result, parts may be scrapped, wasting time, resources, and costs, and meaning that an AM part¡¯s real-world performance can differ from its original CAD design. In this project, a non-destructive testing process based on computed tomography (CT) data was used to analyze a ¡°hot box¡± heat exchanger - a good example of a complex AM part.

Overview

Metal Additive Manufacturing (AM) has many advantages for producing topologically complex designs that cannot be easily achieved with traditional methods. However, non-destructive inspection and testing of the same components is often challenging due to inaccessible features, leading to missed defects and increased performance testing. As a result, parts may be scrapped, wasting time, resources, and costs, and meaning that an AM part¡¯s real-world performance can differ from its original CAD design. In this project, a non-destructive testing process based on computed tomography (CT) data was used to analyze a ¡°hot box¡± heat exchanger - a good example of a complex AM part.

Highlights

  • Synopsys, Nikon, Uni Huddersfield, MTC, and HiETA worked together on an industrial use study (3-in-1 project)
  • Image-based method allows non-destructive inspection of AM parts
  • Simulation in COMSOL Multiphysics?  enabled insights into ¡°as-designed¡± vs. ¡°as-built¡± parts
  • Workflow promises cost and resource benefits

Reference

  • . UK Research and Innovation
  • Turner, N., Brierley, N. and Townsend, A., 2019. . 9th Conference on Industrial Computed Tomography (iCT) Padova, Italy, 13-15 February 2019.

Thanks to

Nathanael Turner, Advanced Research Engineer, NDT, The Manufacturing Technology Centre (MTC)

¡°Simpleware software was intuitive to use and the in-built tools made it easy to process complex data sets and perform useful analysis, such as wall thickness calculations in the click of a button.¡±

Scanning the AM Part

The ¡°hot box¡± used in this study is a test jig used to determine performance of a particular structure before a customized heat exchanger is created. The part has a lattice structure through which air passes, as well as cross-corrugated channels for liquid coolant, making it a complex design to create using conventional manufacturing. In addition, physical inspection of the interior of the ¡°hot box¡± is impossible without cutting open the part.

Simpleware AS Ortho segmentation of a hip with implant

Figure 1. ¡°Hot box¡± heat exchanger (image courtesy of HiETA Technologies Ltd.)

To solve this problem, the ¡°hot box¡± was first manufactured using AM from AlSi10Mg by HiETA Technologies, before being CT scanned at The Manufacturing Technology Centre (MTC). Image reconstruction made it possible to visualize and quantify defects within the part, and to set up the next stages of an image-based workflow.

3D Image Processing

The 3D image data of the part was imported to Simpleware ScanIP to carry out image processing and meshing. Automated segmentation tools in the software were used to determine surfaces, while a local surface correct filter was applied to counter beam hardening effects. To understand the difference between the ¡°as-built¡± and the ¡°as-designed¡± versions of the parts, the Simpleware Surface Deviation tool was used to compare the CAD surface and AM surface.

Simpleware AS Ortho segmentation of a hip with implant

Figure 2. Volume rendering (grey) used for initial inspection showing powder build-up in the base of the ¡°hot box¡±, and automated segmentation tools used in Simpleware ScanIP to generate the initial image-based model (yellow)

This method identified three areas of deviation, including trapped powder and differences from the designed lattice structure. Having processed the image data, Simpleware software was then used to export a full volumetric mesh for CFD analysis. The generated mesh included three parts: metal, fluid, and air, with assigned boundaries also added to model inlets and outlets for fluid flow regions.

Thermal Simulation

The mesh was imported to COMSOL Multiphysics? for simulation and comparison of thermal behavior, including coupled heat transfer and laminar flow. The temperature distribution the coolant flow through the channels is shown in Figure 3, with deviations in the ¡°as-built¡± geometry most evident towards the base of the heat exchanger. Furthermore, the overall cooling from inlet to outlet regions is greater in the ¡°as-built¡± geometry, and less uniform in the vertical axis. As a result, the study was able to identify that the ¡°as-built¡± part created using AM performs worse than the original, ¡°as-designed¡± part, the result of geometric differences.

Simpleware AS Ortho segmentation of a hip with implant

Figure 3. Thermal simulation showing difference between ¡°as-designed¡± CAD based simulation and ¡°as-built¡± image-based simulation

Conclusion

This industrial use case is valuable for demonstrating the importance of using X-ray CT and image-based modelling to enhance understanding of what occurs between a CAD design and the actual manufactured part. Using this workflow, manufactures can close the design loop and potentially save on costly retesting by understanding how unexpected defects and inconsistencies affect real-world performance from this simulation data.

Simpleware AS Ortho segmentation of a hip with implant

Figure 4. Additive manufactured ¡°hot box¡± heat exchanger

Special thanks to our project partners: Xtek Systems (Nikon Metrology), The Future Metrology Hub at the University of Huddersfield, The Manufacturing Technology Centre (MTC), and HiETA Technologies, and to UK Research and Innovation for funding this project.

Simpleware AS Ortho segmentation of a hip with implant

Any Questions?

Do you have any questions about this case study or how to use Simpleware software for your own workflows?