Turning Tall Ship 3D Scan Data into a CAD Model with 3D Systems Geomagic
In 2012, engineer Christian Stark took his family on a sailing trip around the San Juan Islands in the Pacific Northwest in Washington State, and fell in love with the 90 year-old 127 foot Tall Ship, the Schooner Zodiac. On a second cruise as volunteer crew, Christian was discussing the ship’s sailing performance and the skipper noted that the ship’s performance on a portside tack to a starboard tack were perceptibly different. In addition, the skipper mentioned to Christian additional port side structural damage that had been repaired in an earlier renovation.
Christian’s engineer brain went into action. He started asking about blueprints, about hull scans, proposing the idea to actually measure the difference in shape between port and starboard. However, the blueprints had been lost in a fire at the shipbuilders many years ago and the only drawings that existed were approximated hull line drawings created for CoastGuard certification. And these drawings assumed that both sides of the hull were exactly the same. How could the hull shapes be easily and quickly determined? In the way that such things go, a passenger on that second cruise turned out to be a 3D scanning expert from Toronto, Canada, and by the time the cruise was over a plan was in place. Ross Nairn from Toronto would come and scan the ship when it was in dry dock in March 2013, and Christian would work the data to create blueprints.
Using a FARO Focus 3D scanner donated by TSC Inc., in Toronto Canada, Ross scanned the boat while in both the water and then in dry dock as well as conducting thorough scanning of the interior of the ship. With more than 2 billion 3D data points on hand, Christian would now have to try and turn this into a CAD model. Geomagic Design X, a comprehensive reverse engineering software product from 3D Systems, combines 3D scan data processing with history-based CAD to create fully-functioning CAD solid models. Starting with this software, Christian successfully cleaned the data and resampled it so that the final point cloud was reduced to about 26 million 3D points.
To see the full story, please select the PDF below.