My earlier post is entirely dependent on a critical assumption, and after giving it some though, I’m not sure the assumption is valid.
Are the bases of the abutments fixed with respect to each other?
If they are, then I still like the opposing arches.
If not, if weighting the center of the span can cause the abutment bases to spread, then I’d abandon hope for efficiency in material use and proceed differently.
In that case, I’d look at an elevation view resembling an inverted “V”, each member being a fully integrated lam beam of sticks on edge.
Start with a pier, a stick laid flat, three short “blocks” laid flat, another full size flat stick, three more blocks, top middle and bottom, another flat stick, and so on until you had a built up near solid strut, one stick long, one stick width wide, and however many stick widths thick as you could afford, all glued and clamped. Same deal for the other side, and for the road deck.
Then from the base of each abutment, a diagonal extending up to slight past the center of the span, again a built up laminated beam. From the top of each pier, another diagonal of similar dimension, and almost the same angle.
Almost the same because of how the forces are going to act upon two different joints. The road deck will want to deflect under load, while the abutments want to spread apart, but the lower diagonals will try to oppose that. If you stopped there, testing to failure would eventually snap the road deck at the weakest point, in the center.
By reproducing the diagonals above the road deck, a load now induces both halves of the structure to rotate about the meeting point of the two upper diagonals. At the center of the road deck, the force changes from a bending moment which tries to break the deck in half, to a near horizontal tension, tangent to the point of rotation, the radius being the distance from the top of the upper diagonals to the road deck.
This will p
