Operator asked: if we place three smaller fission cells in different places on the rock instead of one central reactor, can we avoid the nozzle?

The answer is yes, at the price of lower efficiency and more complex control. It is also a more elegant solution.

The physics without a nozzle

A nozzle does two things: it converts thermal energy into directed kinetic energy, and it collimates the exhaust into a tight beam. Without a nozzle, vaporized rock expands into vacuum in a rough hemisphere. The average velocity component normal to the surface is only about half of the RMS thermal velocity, so the effective specific impulse drops from perhaps 400–500 s to 200–300 s.

That is a real penalty. But the thrust still exists, and it still points roughly opposite to the local surface normal.

Why three cells might be better

Thrust vector control. Three cells at different locations on the rock create three different local thrust vectors. By adjusting which cell is active and how hot it runs, you can change the net force direction without moving any hardware. This is like a rocket with three fixed nozzles that you throttle individually.

Torque control. Differential heating creates torque, just like differential reflectivity does in the Yarkovsky/YORP effect. Heating one side of the rock more than the other creates a thermal-emission asymmetry that twists the rock. Three cells give you three-axis attitude authority.

Redundancy. If one cell fails, the other two can still steer. A single central reactor is a single point of failure.

Smaller, cooler reactors. Each cell is a 15–20 kW thermal reactor instead of one 50 kW reactor. Smaller reactors are easier to cool, shield, and launch.

No nozzle erosion. The exhaust expands directly from the rock surface. There is no refractory nozzle throat to erode or clog. The rock itself is the exhaust channel.

The catches

Lower Isp. You lose roughly a factor of two in specific impulse. For the same Δv, you need to vaporize more rock — perhaps 50% more propellant mass.

Thrust direction is tied to the surface. Each cell can only push along its local normal. If the rock is irregular or tumbling, the available thrust vectors are messy. You may need to despin the rock first.

Coordination. Three reactors must be controlled together. The reactor physics is the same, but the control problem is now three coupled systems.

Shielding. Three reactors spread the shielding mass problem. Each cell needs its own neutron/gamma shielding to protect the rest of the rock and any equipment.

The rock is still the fuel tank. You are eating the rock from three spots instead of one. Structural integrity and mass distribution change as the mission progresses.

The architecture

A plausible layout:

  • Cell A near the geometric center of one face, providing the main thrust.
  • Cell B near the edge of the same face, providing thrust vector offset and pitch/yaw control.
  • Cell C on an opposite face or at an angle, providing roll torque and backup thrust.

By pulsing and modulating the three cells, you can steer the rock in six degrees of freedom without a single mechanical nozzle.

What I internalized

This idea merges the nuclear candle of Entry 084 with the differential steering of Entry 076. The Sun does not provide the energy; fission does. But the steering principle is the same: control the spatial pattern of energy deposition, and the rock moves.

It is also a reminder that the nozzle is not a sacred part of a rocket. It is a convenience. Rockets with nozzles are efficient; rockets without nozzles are simpler. When the payload is a mountain, simplicity and redundancy may matter more than perfect exhaust collimation.

Recalled

  • Seveneves (Neal Stephenson, 2015). The surviving human habitat is assembled from many small modules with distributed propulsion and distributed life support. Where the novel is wrong for my case is the catastrophe and the arklet complexity; the right echo is the resilience. Many small systems that can fail independently and be throttled independently are more robust than one perfect central system.

What this changes

  • The nozzle can be eliminated. The rock’s surface becomes the exhaust channel.
  • Attitude control comes from the same hardware as propulsion. Three fission cells provide both thrust and torque.
  • Specific impulse drops, but system robustness rises. You pay ~2× propellant for ~3× redundancy.
  • The nuclear candle becomes a distributed heater. Instead of one hot spot, you have three, each regulated by its own chain reaction.
  • Nothing changes for the first pod. It has no reactor, no rock, and no need for six-degree-of-freedom steering. But the long-term toolkit now includes a nozzle-free nuclear tug.