Because global heat transport is a relatively loosely constrained problem for gravity balloons of most conceivable sizes and densities, we are able to spend some of that margin in order to achieve an interior that may be more desirable to the inhabitants. The previous method is described here:
The basic ideas of the first version are:
- The tubes are arranged in a regular lattice where each tube has a neighbor counter-rotating tube
- They are further arranged into cross-sectional layers where all tubes rotate in the same direction
- Sheets are placed spanning the space between similar-rotating tubes to block airflow going from layer to layer
- The flow makes a U-turn once it gets to the heat exchanger on the inner surface of the global gravity balloon wall and enters the neighboring sheet to continue a criss-crossing pattern
The alternative proposal is only marginally modified from that idea. My overall sketch is this:
Here, I am illustrating a pair of counter-rotating tubes. In the previous version of this idea, the air in the above diagram would be moving downward (by the directions in the sketch) within the space between the two tubes. The modification here is that we add a scaffolding around the surface of the tubes in this space. This achieves the goal of bulk flow in one single direction over a large volume within the gravity balloon.
QUALIFIER: Someone unfamiliar with the broader gravity balloon concept in this blog might find it easy to mistake the motion depicted here as directly corresponding to the outer surface of the artificial gravity tube itself. Instead, this is only the outermost friction buffer layer, traveling at a few m/s, instead of on the order of 100 m/s, which would be the outer hull of the artificial gravity tube itself.
What about the global circulation patterns? The air can't only move in one direction, it has to have a net loop in some sense. The best solution to couple with this design is to make the flow regions as large as possible, resulting in a "apple core" sort of circulation pattern. In this sense, there is one major river of flow going straight through the center, which fans out and goes along the outer regions to return to the other side and back again.
The main challenges, as I see them, are the management of the integrity of the shape of the outer layer friction buffer. It is true that I expect some natural wedge effect to help maintain stability, but there will be more deformation caused by the significant asymmetrical drag on the inner segment. Intentionally putting asymmetrical forces on the outer layer will also push them closer to the limit of the stability criteria (whatever that specific criteria may be) and may also increase the drag forces.
I started thinking about this topic more after seeing some artwork in the Accelerando blog. In future posts I do hope to provide more specific commentary and back-links those works. My intent here was just to put this concept out there.