Thanks, that's a great find. I need to think a bit more about those box keels they have designed and modelled. Each is 60mm high (deep) and 50mm wide. The skins that make up these boxes are 2mm thick aramid reinforced plastic (epoxy, I assume). Inside the box cavity is some supergoo "Theta-6" silicone that is tough and strong (at least for a silicone). As near as I can tell, they are attempting to make the keels "tough" and energy absorbent by including this silicone and using aramid (which certainly does hang together well after it crumples) to make a keel beam that can absorb a lot of energy >when< it crumples. It will be ductile, like a steel tube rather than a stiff. frangible composite component, I guess. And that sounds good. But: Each beam is 197.5cm long, and by my math the aramid epoxy skins weigh 1.2kg (2.6 lbs) each. The silicone density is 1.06g/cm3, and there are 5,088cm in each beam, for a silicone mass of 5.4 kg (11.9 lbs). So, by weight, there's 450% more silicone than composite skin in each keel. Also, aramid is very weak in compression (in some cases, weaker than the epoxy matrix). That could be a problem in this end-loaded column. And, in their modeling these keels do crumple and the occupant safety area is badly compromised (though not as badly as in the "before" case where the glider had no keels/lower longerons at all).
What I'd like to know (and what they didn't model): What if you'd replaced that silicone with light XPS foam or a corrugated CF/epoxy web and used the remaining saved weight to triple the thickness of those walls? And used CF/epoxy to build them? Put a light layer of aramid outside to contain the shards, in case of fracture, but I'd bet there's a good chance that the keels would remain intact and so would the passenger area. Do the energy absorbing with the nose cone, good harnesses, etc. and keep the flail space intact.
Anyway, they have an interesting approach. There are some promising references at the end of the article. I'll try to track them down and see what I probably misinterpreted.
Mark