Sub-layer

As a first result ...

• natural scaling of frequencies (at constant speeds) and
• adhering to continuity of throughput
  

just gives one single sub-layer per assembly level
See: Math of convergent assembly

But one sub-layer per assembly level strongly deviates form proposed designs because
there is motivation to deviate from that for several reasons.
See: Deliberate slowdown at the lowest assembly level

One main reason is optimization for minimal friction.
As it turns out the global minimum for friction is somewhere around n~B³ sub-layers
A bit more accurately With factors more or less near one:
<math> n_{opt} = \sqrt{A_A/A_T} B^3 C </math>
For details see: Limits to lower friction despite higher bearing area

(TODO: Can this be generalized to sub-level, that is abstracted away from layer geometry?)

Related

• Optimal sublayernumber for minimal friction

---

• Deliberate slowdown at the lowest assembly level
• Limits to lower friction despite higher bearing area
• Compenslow

---

• Higher throughput of smaller machinery