The Physics and Chemistry of Matter

[drunkn.bear]

Very scientific. this explains how each material and method affects the cooling. helpful when looking at specs

 

[dream caster]

Properties of Common Materials


Heat Capacities

Common Metals
Silver 419 W/m-k
Copper 385 W/m-k
Aluminum 210 W/m-k

Metals used for plating
Nickel 60.7 W/m-k
Chromium 69.1 W/m-k

. . .

This is wrong, heat capacity is the quantity of heat energy a material has to absorb to rise its temp one degree, it is one of the properties that make up a good liquid coolant (water has a big heat capacity); the property that you look for in metals is heat conductivity and that is the property you have the values listed up there.

 

[David]

This is wrong, heat capacity is the quantity of heat energy a material has to absorb to rise its temp one degree, it is one of the properties that make up a good liquid coolant (water has a big heat capacity); the property that you look for in metals is heat conductivity and that is the property you have the values listed up there.

Oops.... I should have noticed the units as well *slaps self*.

Edited, thanks.

 

[dream caster]

I think this may be useful:

Energy Conservation, Entropy and active cooling methods:

An active cooling method means you are spending some energy moving heat, after you put this energy into your system it does not vanish, it sticks there losing its order and it appears as heat.
There is some heat that corresponds to inefficiency of your cooling device that is generated within itself; this may go out to the general ambient or go into your system and there is the useful part of its work listed as its power in its specs. This useful work goes directly into the cooling media; in a closed system, like a phase or water cooling system, it is completely used overcoming frictions and finally appears as heat inside your system, upping the needed cooling work; in an open system like an air cooling system, a small part of it goes into your system as friction produced heat and the rest goes out as free energy in cooling media (air).

This applies very much to pumps in water cooling; all the work the pump does moving water finally goes to heat it; so if you use an oversized pump you are only adding heat to system and rising its temp. If you use a small pump say 5W you add that heat production to your system, if you use a 50W pump you add 50W heat (just about what a really hot CPU produces).

 

[ChemJohn]

This applies very much to pumps in water cooling; all the work the pump does moving water finally goes to heat it; so if you use an oversized pump you are only adding heat to system and rising its temp. If you use a small pump say 5W you add that heat production to your system, if you use a 50W pump you add 50W heat (just about what a really hot CPU produces).

This is total rubbish.

Use your brain and think through your statements before spouting such nonsense.

 

[SuperFarStucker]

Hey, its possible that all your pump does is produce heat... we'd call those heaters where I'm from though!

 

[RotKT]

thanks

 

[Captain Helghas]

I'm with ChemJohn on this one. People mistake that all energy must eventually be released as heat, which is not always true. With a water pump the friction of the pump is the only energy transfered as heat. The rest of the power draw of the pump is transfered as kinetic energy to the system.
Brad
Analytical Chem Grad student @ FSU

 

[Liquid3D]

This post is one of the most useful I've ever read on any forums. this is why I've been spending more and more time here. With all due respect it seemed if you wanted to hang with the "Cool Kats" you had to run with the XtremeSystems.org crowd. That site has changed and it seems what used to be a friendly site where information was exchanged, now revolves around the myopic pursuit of being "the fastest."

Understanding the binning process and CPU Fab trends is the basis for the purest overclocking school of thought, unfortunately it's been cheapened by the "wanna be" hordes flocking to identified stores carrying specific batches. One person with too much money and too much ego comes across a great stepping, do they discuss how and why binning led them to this "gift"? No, they throw it in their rig and push it until hero worship is achieved, when all they did was basically get lucky. Almost anyone can claim a speed-title with enough funds and a simple search on Google such as "best stepping for such and such." But very few ever bother to understand the science behind everything we do.

IMHO OCForums is getting back to the roots of overclocking, topics like this are invaluable resources. This knowledge (or at least the desire to learn) seperates the "overclocking poser" from the true lover of science.

Every article I write involves more time researching ancillary topics so I have a better grasp on the subject matter and can communicate that knowledge to my Readers. One of the earliest inspirations for my recent fpray into PC-technical writing was an article I found at Overclocker's which I found myself bopokmarking. i found another at Lost Circuits and eventually read an author I believ to be one of (if not the) very best in this feild; Van Smith of http://www.vanshardware.com/index.htm

Never reading VHJ would be analogus to being a Modern European Literature Major and never reading Jean-Paul Sartre.

ANyway thank you for the topic it's been bookmarked!

 

[BobcatDan]

The nucleus is made up of two types of particle. These are the proton and the neutron. The proton is postively charged. In an atom there are always equal numbers of protons and electrons. Protons and Electrons have equal charge. Neutrons have almost the same mass as Protons but are not charged. These lead to various isotopes of elements.

The mass of an atom is expressed in terms of Atomic Mass Units (AMU) where 1 AMU ~= 1.66 x 10^(-27) kilograms. The mass of an element can also be expressed as grams per mole where one mole is equal to 6.02 x 10^(23) atoms. Thus 1g/mole implies that 1 atom has a mass of 1 AMU, 18g/mole implies 1 atom has a mass of 18 AMU etc.

The nucleus does not always have an equal number protons and neutrons in it (although you do talk about isotopes later). The main example is hydrogen, only has a proton in the nucleus. The isotopes used in fusion reactions, Deuterium and Tritium, have one and two neutrons respectively. Typically, however, for an element to be stable it will have an equal number of protons and neutrons. It is interesting to note that the mass of a neutron is greater than the mass of a proton by the mass of an electron, and that a neutron is in fact simply a proton and electron “stuck” together (this is all an oversimplification).

The explanation on TEC’s is very good, but I thought I might add a point (actually, more like ask a question). As stated before, the electrons in a TEC absorb thermal energy during the “climb” into the higher energy band in the semiconductor, and that they are motivated to this task by the potential applied. My question is how is the heat transfer related to the power applied to the system. Meaning, if the device was 100% efficient (minus the impedance of the circuit) it seems to me that electric power (applied voltage times impedance of the circuit) would be equal to the thermal difference between the two sides. I.E. if you had a TEC attached to your typical 12 lead off of your PS and the impedance was one ohm (nice round number, generated entirely in the device and not any connections to the PS) this would give you 12 watts, would this lead to a 12 watt heat transfer between the two sides of the TEC (I chose to stay in watts here, as it is convenient when relating cooling capacity to heat produced by the CPU)?
Dan

 

[David]

The nucleus does not always have an equal number protons and neutrons in it (although you do talk about isotopes later). The main example is hydrogen, only has a proton in the nucleus. The isotopes used in fusion reactions, Deuterium and Tritium, have one and two neutrons respectively. Typically, however, for an element to be stable it will have an equal number of protons and neutrons. It is interesting to note that the mass of a neutron is greater than the mass of a proton by the mass of an electron, and that a neutron is in fact simply a proton and electron “stuck” together (this is all an oversimplification).

In an atom there are always equal numbers of protons and electrons.

Also: oversimplified yes. However if someone had told me when we first started on atomic structure at high school exactly how electrons were arranged, hunds rules, the bohr model, crystal field splitting, etc etc all at once I probably would have bricked it and dropped Chemistry

 

[David]

Added some stuff about electron arrangement.

 

[Super Nade]

Now we need a Quantum Mechanics sticky from supernade.

I would be most willing to oblige. Maybe an article on quantum computing?

Props to the seldom seen David

 

[David]

I would be most willing to oblige. Maybe an article on quantum computing?

Props to the seldom seen David

Expect to see more of me after May

 

[Bunny771]

Thanks alot David, that was very informative. I remember quite a bit of that from my chem and physics stuff, but about 1/3 of it i had never heard of before. Deffinatley worth while.

 

[Audioaficionado]

The isotopes used in fusion reactions, Deuterium and Tritium, have one and two neutrons respectively.

They actually use 6Li D (lithium-6 deuteride) to make H-bombs. Tritium will decay and have to be replenished plus it is harder to mass the two gasses together to be able to thermofuse them. the ⁶Li D (lithium-6 deuteride) is a stable solid cake and can be stored indefinitely.

http://www.fas.org/nuke/intro/nuke/design.htm
http://en.wikipedia.org/wiki/Nuclear_weapon_design

 

[Fridge]

Nice thread... but what relevance does orbital arrangement have to PC cooling?

 

[David]

Nice thread... but what relevance does orbital arrangement have to PC cooling?

Orbital arrangement governs a lot of bonding theory and also excited molecules. In an excited molecule (eg due to heat for example) the electrons will jump from the HOMO to the LUMO (Highest occupied to lowest unoccupied molecular orbitals, which granted are different from atomic orbitals) but its not really of relevance to computer cooling, it just makes it more complete I guess.

I could write up a bit on thermodynamics I guess...

 

[BobcatDan]

They actually use 6Li D (lithium-6 deuteride) to make H-bombs. Tritium will decay and have to be replenished plus it is harder to mass the two gasses together to be able to thermofuse them. the ⁶Li D (lithium-6 deuteride) is a stable solid cake and can be stored indefinitely.

http://www.fas.org/nuke/intro/nuke/design.htm
http://en.wikipedia.org/wiki/Nuclear_weapon_design

Sorry, I haven’t checked the site in ages, so I didn’t see your response. While your right, lithium 6 deuteride is the material they put in the bomb, it is deuterium and tritium that undergo fusion, per your web link;
“Neutrons from a fission “primary” device bombard the 6 Li in the compound, liberating tritium, which quickly fuses with the nearby deuterium.”
http://www.fas.org/nuke/intro/nuke/design.htm
Dan

 

[batrick]

Water has a heat capacity of ~4200 J/K/kg meaning that 4,200 Joules (4.2kiloJoules) of energy is required to increase the temperature of 1kilogram (which is 1 litre) of water by 1 degree Kelvin.

Smalls things I noticed over a cursory examination: the units for heat capacity is (J/K*kg), which can be read as Joules per Kelvin Kilogram, and 1 kilogram of water is 1 litre at sea level.

Overall pretty good guide