1. A film about imbalance

The Day the Earth Caught Fire is a 1961 film about what happens when the planet’s thermal balance is thrown off. It is not great science, but it is a useful reminder: systems fail when heat cannot get where it needs to go. The desktop is much smaller than the Earth, but the same rule applies.

This entry is about the thermal path from a cell to the rack across the attachment interface.

2. The heat source and the sink

A compute cell in the desktop may dissipate one or two hundred watts. That heat has to go somewhere. The options are:

  • Conduction to the rack: through the attachment interface and into the rack structure.
  • Radiation from the cell surface: directly to space, if the cell has its own radiator.
  • Fluid loop: if there is a pumped loop carrying heat away.

The simplest first-generation assumption is conduction to the rack, with the rack carrying heat to a shared radiator. That makes the attachment interface part of the thermal system.

3. What the interface must do thermally

The attachment interface must provide a low-resistance thermal path from the cell to the rack. This means:

  • Flat, clean mating surfaces: gaps and oxides increase resistance.
  • Thermal interface material: a pad, grease, or phase-change material that fills microscopic voids.
  • Controlled pressure: the clamping force affects contact resistance.
  • Repeatability: the thermal resistance must remain acceptable after many mate-demate cycles.

These requirements are in tension with the mechanical requirements. A loose interface is easier to insert but has higher thermal resistance. A tight interface conducts better but may wear or jam.

4. What the first test should measure

The ground experiment from entry 326 should include a thermal measurement. A representative heat load on the cell, a temperature sensor on each side of the interface, and a known rack temperature would give the interface thermal resistance. The test should repeat the measurement after many cycles to see if resistance grows.

5. What could go wrong

  • Galling or wear: repeated insertion could damage the mating surfaces, increasing resistance.
  • TIM degradation: the thermal interface material could pump out or degrade.
  • Pressure loss: the latch force could relax over time.
  • Misalignment: a tilted cell could have high local resistance even if it seats.

These are all discoverable in the first test.

6. What this changes

  • The attachment interface is a thermal interface, not just a mechanical one.
  • The first ground experiment must measure thermal resistance and its repeatability.
  • Thermal requirements add constraints to the mechanical design.