I bought a whole range of different thin threads from suppliers on eBay, as well as rating our sewing box at home. I then measured the load:extension curves and worked out how many threads per cm width would be required to give the target material properties.
For practical purposes, I aimed for threads that needed a spacings of several mm – they would have to be arranged by hand in the final wing and spacing closer than that was a recipe for some nasty tangles.
Once I had the properties of the individual threads, I laid them up in a latex matrix and then tested the resultant strips of reinforced latex. They ended up looking like this:
Then the challenge was to build a wing with these properties, and with a reinforcing layout that followed the published illustrations. The pictures below show on elf my first attempts and also the wing spar. Note how there is very little anteroposterior (a-p henceforth) curvature in this – it reflects as far as is possible this lengths of straight carbon fibre rod the shape seen in fossilised specimens. However, I think that the combined wing + membrane would have had considerable a-p curvature (see Palmer & Dyke 2011 and Chapter 6 of the PhD for the reasons behind this.). I therefore made the wing membranes with a plausible a-p curvature and when they are stitched to the wing spar, they force it to deflect and this reload puts tension into the membrane.
