Lift gas sloshing

One of the issues with vehicle design for airships that must traverse the lower 30 km of the atmosphere is that the lift gas needed to remain aloft near peak altitude compresses to a small fraction of the ship's volume near sea level. A roughly cigar-shaped vehicle will be pitch unstable if the lift gas isn't stopped from shifting form one end of the vehicle to the other much like the bubble does in a carpenter's spirit level. 

When the bubble's container is given a larger diameter near the center, the bubble stays mostly in the center if the level is.. well.. roughly level. For an airship, movement of the bubble is equivalent to a shift of the bubble of lift gas. When this happens, the center of buoyancy on the vehicle shifts. Pitch stability requires, therefore, some mechanism to push the bubble back to where it belongs. Typically, this is done with two ballonettes (balloon within a balloon) that act like diaphragms on either side of the cell's center. In a spirit level, we can imagine them at either end acting like pistons.

An airship operating in the lower 30 km of the atmosphere, however, requires the ballonettes to expand to occupy most of the volume the lift gas would occupy at altitude. Failure to do this ensures the outer envelope of the airship will experience a pressure difference crushing it. Some of this pressure can be handled by the airframe, but at quite a cost to the mass of the vehicle.

A serious engineering challenge, though, is that a large ballonette can shift much like the lift gas bubble if the ballonette is not secured against the airframe or load envelope. Even then, the air inside the ballonette is not fixed. Pitch instability is still an issue until the interior of an airship is subdivided into many smaller cells by baffles that limit the run of lift gas and the ballonette's air. The more finely the subdivisions are made, the less attitude instability is built into the design. Obviously, this comes at as a trade-off with the vehicle's mass.

So... what to do?

One possibility is to subdivide the ballonette by having more than one in a lift gas cell. To imagine this, look at the spirit level again. Two piston's at either end push the bubble back to the center. Now add another structure around the inner wall near the middle that would allow for the cell's diameter to be restricted like a clogged artery. This would split the lift bubble at low altitude, but not at high altitude when the bubble expands. Splitting the bubble creates a small potential barrier to overcome during small pitch changes.

This is the configuration I've chosen to model. The airship isn't cigar shaped, thus the load envelope itself contributes to preventing pitch instability. Cells near the front and back have a strongly curved ceiling preventing lift gas bubbles from running. Cells in the middle on the wings have no sloping ceiling on that cross-section, so the pistons at each end are much more necessary.

The load envelope is essentially two toroids joined at a chord with ribbing toroids kept at high pressure to provide some shape stability. 

Two keel trusses on top and bottom are contained within two similar toroids tied along the hull chord by a vertical baffle. At altitude, the baffle is expected to be under tension, but not a lot. At sea level, the baffle must be placed under tension by pumping air into the load envelope and/or ballonettes to retain the outer envelope's shape. Hull ribbing toroids will not be sufficient.

Lift gas cells are subdivisions of the hull toroids around the polar angle with vertex at the hull's center. Baffles separate the cells and support a toroidal 'piston' ballonette that presses against two cells when pressurized. These ballonettes CAN push the lift gas around a bit, most mostly they squeeze it into a smaller volume thus reducing its impact on the center of buoyancy. The extra mass of the gas in a ballonette ALSO shifts the vehicle's center of gravity, so both effects contribute to managing vehicle attitude.

With some of the cells, I've sketched in a larger rib around the center that can be pressurized to split lift gas bubbles. I don't have them all drawn in yet, but they will be soon. They are currently sized to contain a maximum volume equivalent to a small ring truss running around the inside of the hull. This truss would provide a protected path between upper and lower ring trusses, but I'm still sketching in what that might be used for. Most cells will not have this truss, though, as it is probably over kill. Its mass does not contribute enough to shape stability to justify it. [Considerably more truss-work would be required to turn this vehicle into a rigid airship.]