Morning William - FEA on a part basis will be spot on with stiffness if it is constrained and loaded realistically. In simsolid I can model the friction in connections and the bolt loads so it will model assembly deflections accurately. In your case it will need to be a bonded connection? How does F360 deal with assemblies? If the connection is bonded I assume it to be 50% efficient a a first guess.
Using close pack fibres that are well stacked (hexagonal close pack or square close pack) theoretically you can get 90% plus but in reality 60% Vf is the top figure. Random bits I expect to be 40% Vf. Hard to get to 50%. Using CF will be uneconomical for you unless you can get scrap. I'd use a random scrap stack. Need specialist software to design laminates and FE laminates I use Strand7.
You need to get some aluminium swarf, place in a known volume container on a scale weigh the swarf then fill the swarf volume with water and weigh. You can then convert weight fractions to volume fractions and you will know the Vf of the system. Same with sand if you go the sand route.
With glass and CF laminates under 1atm pressure usually works out Vf=50%. Autoclaves at 6bar can get to 60% Vf in laminates. Internal hydraulic pressure and friction within the laminate prevent squeezing it much further. It is possible to go further, I did read Boeing went a lot further but you do need resin around fibres and the autoclave technology becomes complex. Peter
https://www.compositesworld.com/news...s-to-tennessee
there are recyclers in europe as well