Analysing Surface Deviations in Reverse Engineering Workflows Using 3D Printing and Non-Contact 3D Scanning
Rojas, Joakin Haug (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025052515781
https://urn.fi/URN:NBN:fi:amk-2025052515781
Tiivistelmä
This study investigates the surface accuracy of reverse engineering processes using non-contact three-dimensional (3D) scanning and 3D printing, focusing on the surface deviations introduced throughout the process. It aimed to quantify and compare surface deviations between a Computer-Aided Design model and both coordinate measuring machine (CMM) and non-contact 3D scan data, to analyse whether reverse engineering can reliably reproduce surface-accurate objects. Three test artifacts were 3D printed on a Creality Ender-3 Pro, and three on a Creality K2 Plus and measured using a Mitutoyo CRYSTA-Apex V series CMM and an EinScan HX 3D scanner. GOM Inspect was used to compare measurements and generate surface deviation maps and values.
Surface deviations were found throughout the reverse engineering workflow. CMM data showed that the K2 Plus produced more surface-accurate test artifacts than the Ender-3 Pro, with absolute mean surface deviations (zero-centred) of 152.86 µm and 279.74 µm, respectively. Non-contact 3D scans showed a deviation of 185.58 µm for the K2 Plus and 127.50 µm for the Ender-3 Pro. The K2 Plus deviation increase suggests cumulative error. The Ender-3 Pro results consistently exhibited incomplete scan data and were interpreted with caution throughout the analysis. Average surface deviation values showed a bias towards negative deviations (-75.34 µm, -51.46 µm, -166.21 µm, -109.82 µm), indicating measurements and scans often fell short of the CAD surface. A reduction in average positive surface deviation between CMM and non-contact results for the K2 Plus suggests that meshing algorithms may have smoothed out localised surface spikes. Deviation mapping confirmed that curved features showed the highest deviation and were most prone to missing data.
The reverse engineering workflow produced acceptable accuracy for applications not requiring tight tolerances. The study recognised the limitations of GOM Inspect and made recommendations for experimental adjustments, such as mesh filtering and optimising noncontact 3D scanning parameters.
Surface deviations were found throughout the reverse engineering workflow. CMM data showed that the K2 Plus produced more surface-accurate test artifacts than the Ender-3 Pro, with absolute mean surface deviations (zero-centred) of 152.86 µm and 279.74 µm, respectively. Non-contact 3D scans showed a deviation of 185.58 µm for the K2 Plus and 127.50 µm for the Ender-3 Pro. The K2 Plus deviation increase suggests cumulative error. The Ender-3 Pro results consistently exhibited incomplete scan data and were interpreted with caution throughout the analysis. Average surface deviation values showed a bias towards negative deviations (-75.34 µm, -51.46 µm, -166.21 µm, -109.82 µm), indicating measurements and scans often fell short of the CAD surface. A reduction in average positive surface deviation between CMM and non-contact results for the K2 Plus suggests that meshing algorithms may have smoothed out localised surface spikes. Deviation mapping confirmed that curved features showed the highest deviation and were most prone to missing data.
The reverse engineering workflow produced acceptable accuracy for applications not requiring tight tolerances. The study recognised the limitations of GOM Inspect and made recommendations for experimental adjustments, such as mesh filtering and optimising noncontact 3D scanning parameters.