The problem
Sail aerodynamics models need checking against physical tests. Rigid models give the most accurate wind tunnel results and are slow to make: hand laid fibreglass with PVC tubing embedded, every tube connected to a transducer by hand. Each is a week of skilled labour, and profile accuracy depends on who laid it up.
What I did
I judged printing methods on fixed criteria: surface tolerance, correct form, freedom from defects, cost and whether the unit could run the equipment. More than 45 CAD designs and ten printed models later, multijet printing was the viable method. A connection base I designed mates the printed pressure channels directly to the transducers, so nobody threads and glues tubes.
The last question was whether printed channels behave like PVC tubes. We tested both in the wind tunnel and again with a loudspeaker, which gives a far higher signal to noise ratio and direct control of amplification. The response matched. The tubes are interchangeable.
What it did not do
No full scale model was produced, because of cost. The method is proven at component scale.
Still to add: photos of the printed models or the test rig, and whether the unit went on to use the method.