Converting the data obtained in the CT scan to .stl files for 3D printing has not been without its challenges. Dr Fiona Brock, Lecturer in Applied Analytical Techniques at Cranfield University, listed them for us:
- the maximum height of an object in a single scan is 12cm, so the mouthpiece and the main body of the flute had to be scanned in 2 and 3 sections, respectively, and the files digitally ‘stitched’ together for printing.
- prior to printing, the data files had to be ‘cleaned up’ to remove noise, which could have resulted in random data points being printed.
- the mouthpiece, with a wooden block inserted within it, created additional challenges to the main body and base of the flute. Although the ivory surface could be determined and digitally cleaned up easily for printing, the wood block itself created several problems. Firstly, the wood was very low-density material relative to the ivory (and even more than occasional specks of high-density lead solder present within the instrument). This made it very difficult to identify and isolate the external surface of the wood, especially as the internal voids (from natural vessels within the wood and beetle damage) could not be isolated by standard thresholding techniques from the surrounding air leaving all the surface detail within the wood block for printing would have generated an extremely large data file which could not be processed by most laptops/PCs, and which would have taken a very long time to print. Furthermore, the standard benchtop 3D printer available at Cranfield has a nozzle diameter of 0.4mm, meaning that any detail smaller than this could not be printed. For these reasons, the wood block would ideally have been digitally isolated from the ivory and printed separately before being inserted back into the printed mouthpiece, but this was
- if similar objects were to be scanned for 3D printing in the future, it might be advisable to scan twice, at different working conditions, one optimised for the wood and one for the ivory (in a similar way to which medical CT scanner use two settings, one for bone and one for soft tissue).
The file for the wooden block is currently with Engineering Science at Oxford University for final tweaking – we are hoping for a playable printout of the original ivory recorder soon!