Some in the aviation trade press have speculated that this means Boeing must be looking at an elliptically shaped fuselage cross-section for the 797.
This could allow it to be wider though not as deep as today’s widebodies. Such a cross-section could reduce the overall size of the jet, though at the cost of cutting the available cargo space below the passenger cabin.
That's not correct. An elliptical fuselage (wider than taller) with the same height of a single aisle fuselage like the 757 and the width of say a 767 will have less cargo space than a 767 but more than a 757, and of course the length will also be a factor.
One of the issues with any non-circular fuselage cross-section is that when you pressurize it internally it tries to become circular.
For an elliptical cross-section you would need a tensile element across the short axis of the ellipse to prevent this from happening.
I would guess that double decker planes (where the "ellipse" is oriented vertically) probably solve this problem by using the floor(s) as the tensile member(s). (Verbal?)
A tensile vertical element will help, but would not fully solve the issue. In a simplified model (if we assume that no flexure solicitation exists where the left and right side meet), the left and right side will tend to to become circular segments themselves. So while the tensile vertical element will help reduce the flexure tension caused by the pressurization, it will not eliminate it.
That's why the DC-9/MD-80/90/B717 family (and also the A380), don't have a vertical elliptic shape, but a double-lobe (2 circular segments, one on top of the other) (triple for the A380) and the horizontal tensile element is exactly where both lobes meet, and that's the floor (both floors for the A380). In this case, yes, the tensile element practically eliminates all the flexure caused by these lobes trying to become circular, because they are already circular.
So maybe Boeing is thinking in a double-lobe, side to side? Or a quadruple lobe (2 lobes side to side but connected by two lobes in the top and bottom to eliminate the top and bottom connection, that would be an angle protruding "inside"the fuselage)?
This is a design that I thought myself in my times of an aeronautical engineer dreamer. It came together with a 2 - 3 - 3 - 2 abreast configuration that gives great pax comfort, with no pax having to climb over 2 other folks to get to the aisle, and with only 20% of middle seats. Until one day I realized that the airlines would quickly eliminate the middle aisle and make it 3 - 5 - 3 to gain 10% of pax capacity and converting the best seat config available in the worst one.
In the case of a horizontally oriented ellipse a vertical tensile member would help to handle the pressure-induced loads.
So maybe they're thinking of some kind of laterally split cabin with a "wall" in the middle, and a single aisle each side........
I would hate the window seat with the sight our of the window being another passenger next to me
Now, you don't need a wall. 2 beams at the tom and bottom of the fuselage connected by tensile rods every several feet would be enough.
And, finally, you don't NEED any of that. There are lots of pressurized planes that have non-round cross sections and don't have things crossing the fuselage to help handle the pressurization loads. You can make the fuselage structure to handle those loads by itself. Will it tend to become circular? Yes, in the same way that all the fuselages section ahead and back of the wing tend to flex down on its owns weight. And while that would add weight to the structure, composite materials help a lot with that. One of the great shapes is a rounded square, composed by 4 circular segments of big radius at the top and bottom. This gives a lot of space, and you need to reinforce only the 4 points where the 4 sections meet.