Cooling
Cooling is both relevant in both
- mundane and important practical systems and
- especially in systems that aim to go to the absolute limits of what is possible
The latter by nature goes in quite speculative territory as
it is not compatible with conservative estimation and exploratory engineering.
The latter relates to a common misconception about the limits of power density:
Taking stated numbers too literally
See page: Limits of power density imposed by limits of cooling
Contents
Practical cooling in gem-gum factories
Heat that needs to be removed originates from:
- dissipation from (more or less intentional) snapping instabilities
- more subtle inefficiencies in piezomechanosynthesis
- frictive dissipation – Only partly recuperable by heat engine due to conversion of free energy into bound energy.
- squeeze-out of entropy – heat from moving disorder in positional space to disorder in impulse space (aka heat) (fully recuperable)
Temperatures to cool to:
For reliability of piezomechanosynthesis lower is better.
Below some point there are diminishing returns though.
Optimal might likely be somewhere around liquid nitrogen. Maybe a bit lower.
Room-temperature works bit is a bit too high for good reliability.
Liquid helium temperatures are likely overkill.
Anticooling (i.e. energy recuperative warming/heating):
To get the energy back that was invested to cool below room temperature and
to not freeze clean room air in further up assembly levels
the so far assembled parts-fragments need to be warmed up again before proceeding to higher assembly levels that
no longer involve (unguided) aligning of atomic bonds.
Basically one needs to recuperate the cooling energy by heat engine.
Given layer geometry of assembly level (assembly layers) as a possible self suggesting geometry
cooling and anticooling forms a "cooling sandwich".
From an surface to volume ratio this thin layer sandwich geometry is far from optional.
Maybe hinting on that an other more batch processing
and serializing in-between geometry might be worth considering.
Increasing fast track transport distances increases frictive losses though so its a tradeoff optimization.
Also depending on how efficient piezomechanosynthesis can be made as it likely dominates over frictive losses.
(TODO: Investigate possible assembly level geometries taking this into account too.)
Taking cooling to its limit
See main page: The limits of cooling
Warning! you are moving into more speculative areas.
Application cases may include:
- Hyper high throughput microcomponent recomposition
- Carriage particle accelerators (for space rockets)
- Nuclear fusion
- ... military horrors :(
Strategies that could be employed include:
- Choice of best materials as the thermal mass (this is beyond metamaterial emulatability)
- Choice of best materials as the thermal conductors (this is beyond metamaterial emulatability)
- Choice of best materials for applying eventual tricks (this is beyond metamaterial emulatability)
- Thermal mass capsules transported on tracks providing good thermal contact.
- Trick: Active squeeze-out of degrees of freedom from the perspective of the equipartitioning theorem
(wiki-TODO: discuss these in detail)
Choice of materials
- Best for thermal conduction seems diamond
- Best for thermal resilience/refractoryness (and still very good for conduction) seems moissanite (gem quality silicon carbide)
- Best for thermal mass seems water high pressure can be easily safely contained by nanoencapsulations
Cooling by capsule transport & its optimization
Thermal energy transport via heat conduction can be very fast.
Especially across short distances. But longer macroscale distances is slows down. Nonlinearly.
One way to counter that is to physically transport thermal masses.
There are some interesting tradeoffs/optimizations to make.
Thus there should be:
- some optimal cooling transport speed depending on details of design choices.
- some optimal scale of thermal-mass-transport-capsules
Optimizing speed:
- PRO: Higher speed means faster heat removal. Trivial.
- CON: Speeding up to high speeds means higher dissipative losses from (wearless) friction
leading to even more heat needing removal.
Unfortunately very low friction suspension (like means of levitation) are not an option at the hot side
due to them providing insufficient thermal contact.
Optimizing scale:
- PRO: Bigger thermal-mass-capsules means less thermally conductive bearing area per thermal mass
- CON: Bigger thermal-mass-capsules means (nonlinear) increase in charge-up-time of these capsules
Squeezeout of degrees of freedom of thermal energy
See related topic: entropomechanical converter
Note: The following is wasteful as it actively increases the devaluation of energy (irreversibly converting free to bond energy).
This seems only useful for boosting performance (not efficiency) while keeping system volume small.
Basically the idea is to accelerate the dumping of heat on the cold side.
- thermal-mass-capsules contain bundles of chain molecules attached to two end-plates
- hot side: bundle-capsules arrive stretched out thermal – normal suck-up
- hot side: end plates are released (leaving machine phase) further thermal suck-up => loss of potentially recuperable free energy
- transport: bundles-capsules are transported to cold side
- cold side: bundles-capsules are pulled apart by force (injection of free energy) accelerating the release of stored thermal energy due to further elevated temperature
- transport back and repeat
By pulling the chain molecules straight they loose degrees of freedom (equipartitioning theorem).
With conservation and fast adiabatic pulling the same energy concentrate in fewer DOFs. Temperature rises accordingly.
Higher temperature means higher thermal gradient means higher speed of thermal flow.
As desired and as is the point of the whole exercise.
Another alternate way to look at the chain molecule stretching process is to interpret it as a forcing of disorder
from position space into impulse space.
Limits:
The chain molecules need to be both
- thermally resilient and
- highly flexible.
This is only possible to a limited degree.
Silicones might be the best bet allowing temperatures to about ~600K or ~300°C
Unfortunately emulation by mechanical metamaterials is not viable.
While chains made of moissanite (SiC) crystolecules would be superbly resistant against very high temperatures,
they would likely not provide a sufficient volumetric density of mechanically squeezable thermal degrees of freedom
due to the size of the individual chain-links being at least several dozens or rather hundreds atoms each.
Related
- Hyper high throughput microcomponent recomposition
- APM and nuclear technology: Fusion
- Rocket engines and AP technology: Carriage particle accelerators
