[home] [Personal Program] [Help]
tag
09:50
20 mins
Photogrammetry in Large Format Additive Manufacturing: an alternative tool for machining setup and inspection
Pedro Andre Garcia Camino, Lucas Abia Hof, Simon Joncas
Session: Session 1: Technologies for Additive Manufacturing
Session starts: Tuesday 14 April, 09:50
Presentation starts: 09:50
Room: Main


Pedro Andre Garcia Camino (École de technologie Supérieure)
Lucas Abia Hof (École de technologie Supérieure)
Simon Joncas (École de technologie Supérieure)


Abstract:
Fused Granular Fabrication (FGF) based Large Format Additive Manufacturing (LFAM) combined with a subsequent machining step has recently been proven efficient in producing small to medium sized composite tooling such as moulds and trim jigs. After printing, the part’s working coordinate system (WCS) is generally precisely repositioned in the machining environment with respect to the true printed geometry. This step is currently performed using laser scanners or probes. Before machining, the printed part is also usually laser scanned for quality control purposes (in and after process) to ensure dimensional accuracy. In this paper, photogrammetry is investigated as an alternative solution to laser scanning to determine the printed part WCS and dimensions. This method could lead to a drastic reduction in investment costs and processing time required for both steps. A series of experiments were conducted to qualify photogrammetry as a metrological tool. First, a test part of 300 x 300 x 150 mm was scanned with a Hexagon Absolute ARM 7-AXIS 8525-7 to create a benchmark digital part. Then, the same part was submitted to photogrammetry using a high resolution C1 camera and a custom scaling system. Images were then processed using MetaShape software. Both methods produced STL files that were overlayed using Polywork’s Metrology Suite to compare accuracy. Preliminary results show that an accuracy of ± 1.5 mm can be easily achieved when compared to laser scanning. The primary source of error lies within the scaling artefacts used in the photogrammetry process, and these will be refined. Once this is addressed, an accuracy of ± 0.5 mm is expected. The final paper will present the integration of the photogrammetry system on a CEAD Flexbot large scale 3D printer on which WCS location and printed part measurement will be validated.