If you allow the pressure to change then density is a non-issue. If you
want to compare a change in density regarding thermal transfer, you
can't be changing the pressure (and thusly the volume, too). You could
up the pressure to provide whatever density you wanted.
...
No one said they disappear. They move somewhere else. That's why the
atmosphere doesn't have a restricted volume; i.e., we don't have a rigid
dome covering out atmosphere to prevent expansion (and contraction). A
huge meteor hitting the Earth can blow the atmosphere far enough away
that it won't return (too far for gravity to pull it back) due to its
explosion of solids turned into gases that *displaces* the atmosphere.
The atmosphere didn't disappear. It got pushed out to a larger volume.
Yeah, there would be compression (pressure increase) during the
explosion but afterward the impetus velocity imparted to the atmosphere
along with it being too far from Earth would have it drift away.
Pressure is determined by gravity (by the weight of the column of gases
over a geographical point). If it were a column of humid air, the
density is less at any altitude along the column compared to column of
non-humid air. A column of hydrogen or helium would exert even less
pressure from its column of molecules. Humid air contains hydrogen
whereas dry air did not and instead retains its heavier molecules.
By the way, barometric pressure DECREASES with an increase in humidity.
See
http://www.coscosci.com/barometric/barometer.htm. Note as the
pressure decreases (lower height in the mercury pushed up in the other
vertical tube) that there is more chance of precipitation (because there
is more humidity). Since the weight of a volume of air (due to gravity)
depends on its density, lower density means less weight, so less
pressure on the open column of mercury to push it down and then up into
the next sealed tube to compress that fixed quantity of gas.
So pressure is actually decreasing due to the reduce density when
hydrogen replaces nitrogen and/or oxygen. You get both decreased
density and decreased pressure with humid air. If the volume were to
remain the same and you removed molecules then you would realize that
you have reduced density and hence the weight of that volume of air. So
instead of removing molecules, you swap them with lighter molecules.
...
The atmosphere is not a fixed volume.
...
Again coming back to the misconception that the atmosphere is a fixed
volume. You are not adding water (gaseous or liquid) into a box that
has a fixed volume. Adding water vapor (i.e., in its gaseous form)
means having to displace molecules already there. Unlike an experiment
where you might pump the water vapor into a fixed sized box which would
result in upping the pressure inside, you are adding water vapor to the
*atmosphere* which can expand, so the pressure is not increased by
adding more molecules to the same volume but by pushing out the heavier
atoms with lighter ones with a consequential weight change due to the
reduced density.
Turbulence on the surface of a heatsink fin is a positive
thing. It breaks though surface boundary layers to remove
heated air faster, which had slowed upon contact with the
fins' surfaces.
Which goes against what HVAC engineers will claim, that turbulence (to
maintain the same flow rate) is resistance which means you have to add
more energy to the airflow which means its kinetic energy is higher
which means its temperature is higher which means there is less thermal
differential. Also, please explain what you mean by "surface boundary
layers". A molecule of air against a higher temperature surface will
become more kinetic. However, you still want to move that molecule of
air *away* from the surface from which you are extracting heat.
Obviously laminar airflow that is perfectly parallel to the other
surface may never impact that surface to receive kinetic energy from the
collision. But a laminar flow of air against a surface (by angling it
against the surface) doesn't need to itself generate [much] turbulence
so that molecule's average velocity is away from the surface rather than
at it or against the direction of airflow; see
http://www.electronics-cooling.com/Resources/EC_Articles/MAY96/may96_01.htm
where the angle of attack is mentioned. Also, even if the flow were
laminar (or maximal) for the CPU heatsink, you still have to get that
heated air away from the heatsink, and turbulent airflow inside the case
can have that warmed air returned and result in a slow heat exchange.
Even if the airflow were designed to go straight across the box from
front to back with an equal volume delivered across the entire span of
the front of the case, there is no way that the laminar airflow would be
perfectly parallel to every component on the mothboard, expansion card,
chips, CPU, etc.
Another advantage of laminar airflow is less accumulation of dust. The
average velocity of the dust particle is along the path of the intended
airflow instead of impacting against the surfaces (or against the
airflow to let the dust settle). If laminar airflow wasn't important in
increasing thermal transfer, why are tangential fans manufactured (and
which cost a lot more)? Because you cannot afford as much turbulence
within a confined rack-mounted device as you can with the sloppy design
of a desktop box. With turbulent air, you have to impart more energy to
it to get the same flow rate (due to the resistance of the turbulence)
which means it already has more kinetic energy when it reaches whatever
you are trying to cool. When Googling on "+laminar +airflow +cooling",
I often see statements like, "This turbulence creates noise, slows the
airflow before exiting, and reduces effective cooling.
(
http://www.equipmentprotectionmagazine.com/eprints/UAF.htm)". Also,
laminar airflow reduces contamination (dust) because the particulate is
less likely to settle on the surface whereas "turbulent airflow
ventilated system relies on mixing and dilution to remove
contamination". The accumulation of dust (a thermal insulator) easily
outstrips any arguments, er, discussions we've had regarding humidity,
density, and pressure changes.
Even if you think turbulence was necessary for thermal transfer, the
axial fans used on heatsinks will certainly guarantee a turbulent
airflow. However, putting them on top where they have to push their
exhaust against the solid heatsink at the base side of the fins means
you generate backpressure which reduces airflow speed (I believe this
also reduces fan RPM). Whether turbulent or laminar, I would think
having the fan blow across the fins with the air exit leaving the fins
without impacting a 90-degree surface (i.e., no backpressure) would work
much better, but apparently there are design contraints for that method.
Maybe the heat wouldn't transfer fast enough across longer fins, so you
sacrifice efficiency for effected cooling rate and damn the increased
noise from the turbulence.
Yeah, I'm pretty sure we're by ourselves in this subthread. All this
miniscule stuff probably is insignificant to effective cooling although
turbulence inside the case can be and is often pretty bad unless you are
very neat in routing or orienting your cables and other components
inside (within the restrictions of the hardware provided (I've seen a
case get 7 C cooler by moving cables out of the way or making sure the
ribbon cables are inline with the airflow instead of block it). I doubt
slight changes in *absolute* humidity and barometric pressure much
affect computer cooling. Originally I was discussing the *operating*
temperature range for the Athlon. Mentioning that humidity could affect
cooling got us way off track (and I bet another subthread about the
stored energy in the excitation levels of hydrogen's 1 electron in 1
shell with 1 energy level versus the 7 or 8 electrons in multiple shells
with multiple energy levels in each shell for nitrogen and oxygen would
be even more esoteric). The amount of turbulence and its incumbent
problems probably overwhelms any effect due to humidity, small pressure
change, and molecular energy levels. It's been fun, though.
