The Heatsink with NO FAN...The Sandia Cooler

[pabloottawa]

Either way. It's clear there are some challenges thay have to overcome in order to make it marketable to the PC enthusiast...... For example, as it stands, that damn noise would never make it into my PC case lol.

[pantherx12]

Originally Posted by pabloottawa

Either way. It's clear there are some challenges thay have to overcome in order to make it marketable to the PC enthusiast...... For example, as it stands, that damn noise would never make it into my PC case lol.

The sound is the motor, it's very easy to upgrade the motor to something more silent

Having said that inside a pc case 3-4 foot away from your ear it's likely even this motor would be in audible, they're filming from within a foot the device as far as I can tell.

I never have my ear that close to my pc when it's running

[Steevo]

You know, if we had used this for years they would give a Nobel prize to the man who invented a heatsink with a fan on it.

[rectifryer]

Originally Posted by lilhasselhoffer

~Sigh~

LOL I suppose you have access to all the data supporting this. I'm not even going to go any further with this.

1) Dust collects because there is a fouling lay associated with fluids. The slower a fluid travels the larger this fouling layer is. Functionally, you can prevent dust from building up if the fouling layer depth is less than that of a dust particle. Extremely fast moving air can have this small of a fouling layer rather easily.

Ofcourse it can. Its moving extremely fast. Having seen the effort it takes for a turbine to move air, I doubt a 3k rpm device could create this effect in such a small radius impeller yet large intake. Do you not agree mechanical engineer?

2) See 1. No dust buildup means it cannot be blocked.

I don't agree with your premise.

3) No. They are comparing a larger heatsink, like the heatsinks you see on higher end cooling (think 212, and similar coolers). As explained way earlier, the increase in conductive and convective cooling is huge when the "bearing" material is a flowing fluid. Not hard to see the volumetric efficiencies being 30 times greater than conventional solutions.

When is the "bearing" material not a flowing fluid? Even a conventional setup has flowing air. When comparing efficiency, if the variables aren't directly stated, then I am not going to assume for them. That is a textbook method of selling something

4) STATIONARY. You make two assumptions when you say this. There is no convective cooling, and there are no fans. Air is a fluid, that moves based upon a large number of factors. Coolers generally have a fan to move air, which brings cooler air into the mix so that the difference in air temperatures (heatsink versus ambient) is higher. Higher temperature differences provoke more heat transfer.

This is just irrelevant.

Practicality means understanding the basics. Needless to say, the substantial change in the "basics" of how this cooler works are why it might be difficult to warm up to. Without citing any specifics, I would recommend that you take another look at the assumptions you are working under. A flawed set of assumptions can be interpreted correctly into an incorrect answer. I know I've been guilty of that in the past...

I am well aware that the velocity of fluid impacts the collection of particulates. I am calling bullshit on claims that this device can pull air in at that speed. You are making assumptions about what I said, perhaps I wasn't clear myself. To be specific, I don't believe you can force air through the impeller of this device fast enough at atmospheric pressure to reduce the fouling layer to the point where cleaning won't be an issue. Can other systems do this effectively? Yes yet they have such different specs.

Could I be wrong? Sure, SNL thinks so. Thats reason enough to second guess. Their video still sucks at explaining the device.

Of course, none of us have seen any real data so its all conjecture.

[de.das.dude]

Originally Posted by Random Murderer

When acting as a gas, sure. In this case, the air is acting as a fluid.

do you even know the definition of a fluid?

also the impeller design itself seems like BS on this heatsink. and i call BS on the 150W thing.

[rectifryer]

http://prod.sandia.gov/techlib/acces...010/100258.pdf

That is their abstract and tech doc. This should clear things up for everyone.

[fullinfusion]

Ya wouldn't want to get your finger to close to this thing.

[de.das.dude]

went through the doc. i think it is possible. but we shouldnt expect high performance. just similar/slighty better to stock HSFs.

[Completely Bonkers]

1./ It wont chop your fingers off or snag wires because it will have a "lid" on it. Just as the fins have a bottom plate, there should be a top plate too. (albeit with a hole in the middle). The demo with the blades showing is suboptimal, because it allows air to spill out the top. Design should include top. PATENT ME

2./ Mounting this will be tricky. Look at the video. The motor is underneath competing with the space exactly where you would expect the "hot" CPU to be. You can't mount the motor on top or you will stop the airflow into the vortex. This means it will be a large cumbersome device with heatpipes, or we will need mainboards with holes and mounting positions for motors underneath.

3./ You will get just as much dust, perhaps more even, just not deposited in the fan... but it will still be in the PC case. Just like a twister (tornado)... it is nice an "clean" at the vortex, but rather dirty everywhere else.

[pantherx12]

Originally Posted by Completely Bonkers

2./ Mounting this will be tricky. Look at the video. The motor is underneath competing with the space exactly where you would expect the "hot" CPU to be. You can't mount the motor on top or you will stop the airflow into the vortex. This means it will be a large cumbersome device with heatpipes, or we will need mainboards with holes and mounting positions for motors underneath.

.

Mounting it will be fairly simple, it would just have to be bolted from the opposite side of the motherboard ( screwing into the baseplate/heatspreader the motor sits ontop of)

[lilhasselhoffer]

Time for one last go at this, because its getting tiring to make sense of your opposition.

1) The amount of heat a motor generates is minimal, when compared to what the components it is cooling generate. It's best described simply through construction.

A fan can be constructed out of plastic, with minimal contacts built from metals. Obviously, the fans don't reach a high enough temperature to melt the plastics.

A CPU can be used to ignite thermite. If you want a demonstration, check out the hackaday archives. It burns through metal, and ignites the phosphorous which starts the thermolytic reaction.


2) This is not a tornado. It will never be a tornado. A more apt comparison would be a mixer.

If you set the mixer to high, and blend a beverage (lets say it's a rum and coke, because after arguing this I feel like I need one) you can remove the blender from the beverage without anything on it. If you were to set the blender much lower then you could remove it with parts of the beverage still on it.

In this same way, a fast spinning blade can remain cleaned of dust.

Additionally, set a blender into a beverage, extract it, then turn it on high. Bits of that beverage will be flung throughout the kitchen, which illustrates what any settled particulates will do on this cooler.


3) Bearings get hot. Very, very, hot. I dare you to drive a car for an hour, then touch the grease from the pumpkin. It's scalding hot, and it is what was flowing between the bearings.

Extrapolating, imagine a bearing that could have grease infinitely pumped into the bearing. The grease would remove heat, as it was pulled away from the bearing.

It takes very little to then see this happening in an air bearing. A physical contact (ball bearings, needles/pins, etc...) are replaced by fast moving air. The air bearing is separating the metal bits, and providing an infinite flow of "lubrication" all at once. You get both a long lasting component, and insane cooling through conduction.


4) Most of the heat from the processor will be channeled to the air bearing and sink via heatpipes. Heatpipes use phase change cooling to transfer energy. It's like the bad ass cousin of water coolers.

Take five minutes, and find the phase change energy of water (liquid to gas), versus the heat capacity of water (in a liquid). I'll let those that remember high school physics move on, and give those remaining a quick run down.

The energy required to change a given amount of atoms average temperature one degree is its heat capacity. Lets take a look at water specifically: http://en.wikipedia.org/wiki/Properties_of_water

We can see that the heat capacity of water is about 4.2 J/(g*K) (joules per gram per degree Kelvin). We can see that the vaporization energy for one mol of water is 40.6 kJ.mol at 100 C. What does this mean? (assuming all results are basically standard atmospheric pressure)

You can get water to change from a liquid at one temperature to a liquid at another temperature rather easily. Conversely, changing that water from a liquid to a gas takes a huge amount of energy, and you don't even raise the temperature.

Phase change cooling takes this principal, and changes the pressure around the water. A partial vacuum is created, so that water will boil at a lower temperature. It could boil at 60C, rather than 100C.

So now you've got boiling water, taking a huge amount of energy, and traveling away from the components. You can pass air across the heat pipe, and cool the water back into a liquid. Through both gravity and capillary pressure the liquid is forced back down, and the process is repeated.

This heatpipe can take heat from a component, and transfer it better than conduction ever could. Assuming the other end of the heat pipe is in a high fluid flow situation the process could take almost no time at all. Pulling heat away from the components is then very fast, and basically just an extremely effective evolution of the heatpipe techniques that are already in use today. If it works right now, why is there doubt that it will work in the future?


5) My final question to anyone out there is why is this difficult to understand?

Water cooling uses a fluid, which absorbs and expels heat via conduction. The thermal capacity of water (how much energy it takes to change one degree) is higher than air, so you can get either cooler running systems (same rate of fluid flow as equivalent air cooler), or significantly less fluid flow.

Air coolers pump a fluid across fins, so that the conduction of heat into the air cools the components. Same idea as the water coolers, just a different fluid at play.

Extrapolate that fast moving air acts more like water as a heat transfer medium. This is, greatly oversimplified of course, what is going on. A known fluid, air, is being forced to act more like another fluid, water, currently does by manipulating its other traits (namely velocity). What am I missing, that people seem to be so bent on pointing out?

[fullinfusion]

Way to much to comment on lol.

Thanks for taking the time to post though. Umm ur comment on the wheel bearings! Not true lol.. the bearing must not be set right in order for you not to touch it

Hell my big rig loaded at 100000Lbs the bearing oil is warm not hot and that's going a thousand miles without stopping!

And the rest I didnt read much into it, just to much for the moment lol

Besides I doubt much will come of this technology... That's just old methods (different design thats all) and a new cooling system will evolve in time.

[lilhasselhoffer]

Originally Posted by fullinfusion

Way to much to comment on lol.

Thanks for taking the time to post though. Umm ur comment on the wheel bearings! Not true lol.. the bearing must not be set right in order for you not to touch it

Hell my big rig loaded at 100000Lbs the bearing oil is warm not hot and that's going a thousand miles without stopping!

And the rest I didnt read much into it, just to much for the moment lol

Besides I doubt much will come of this technology... That's just old methods (different design thats all) and a new cooling system will evolve in time.

The economics of size.

I'm talking about a pumpkin (not the wheel bearing, but the large transmission device that the drive shaft attaches to) for a car, rather than that of a big rig.

In a big rig it is easier to have a higher efficiency bearing. This is a simple game of the mechanics of size. To get high efficiency bearings, or any mechanical component really, there are two options:

1) Allow for infinite size. An error of 1/4" on a 300" part is a small percentage. An infinitely large bearing would allow for large amounts of error, without creating a problem. In this example the friction could be nearly zero, allowing for a perfect transmission of energy.

2) Allow for infinite funding. With limitless funding newer technologies, and more time can be spent producing a more efficient device.

Extending the car metaphor:
The transmission system (engine to tires) on a 68 barracuda was a mess. Less than 20% of the power generated by the engine could actually be applied at the tires. This kind of inefficiency was tolerable, because it was the standard for the time.

Conversely, diesel trucks from today share much in common with their ancestors from the 70's and 80's. The same large bearings and transmission systems allow a huge truck to cost efficiently transport goods across the continental masses.

As technology has developed (i.e. capital was invested in research), the barracuda and its brethren have died out. The common vehicles are now efficient, capable of getting twice or more the mileage from the same amount of fuel. This is a direct result of better manufacturing techniques, which cost money to develop.


Back to the original discussion, as I feel I have wandered off.


This is not a new technology. It it the technology we already have, pushed to that next logical step.

We spin hard drives at thousands of RPMs. We have heatpipes, to effectively transfer heat from point to point. We use shearing flows to separate components during manufacturing. All of these technologies are already in place, and Sandia is just now putting everything together.

If Sandia has been working on this for several years, then there are problems somewhere that haven't yet been brought to the outside world. They don't produce things, they develop the technology and license it to other manufacturers. A year between revealing the technology, and not seeing any continued development, means that there's something very wrong.

The problem is that people see that there's no development, and assume it is because of some flawed mechanic. This is, presumably, due how difficult it is to understand how this new mechanism works.

Edit:
Edited for poor english.

[de.das.dude]

Read the section on the hydrodynamic bearing. its the same as how the read/write heads of a HDD "float" on the platter.

one thing CB pointed out is mounting the rotor. i dont think mounting the rotor will be difficult. However, since the rotor is mounted at the centre of the baseplate and the centre is the region of the most heating, there might be some issues with efficiency.

Like i said before, they seem to have tackled all the problems theoretically. The question is how well it will work altogether.

[micropage7]

Originally Posted by de.das.dude

Read the section on the hydrodynamic bearing. its the same as how the read/write heads of a HDD "float" on the platter

with such a small gap you can add a little dust to make it locked
and i questioning about how good heat could get transfered with no direct contact between them

[de.das.dude]

Originally Posted by micropage7

with such a small gap you can add a little dust to make it locked
and i questioning about how good heat could get transfered with no direct contact between them

i dont think that area will be accessible to dust. that will probably be sealed like a ballbearing.

and 0.03mm is really small. its almost the same as the gap between fins and heatpipes that pass through them
also if you look at their experimental darta, air is kinda conductive at that thickness.

[pantherx12]

Originally Posted by de.das.dude

Read the section on the hydrodynamic bearing. its the same as how the read/write heads of a HDD "float" on the platter.

one thing CB pointed out is mounting the rotor. i dont think mounting the rotor will be difficult. However, since the rotor is mounted at the centre of the baseplate and the centre is the region of the most heating, there might be some issues with efficiency.

Like i said before, they seem to have tackled all the problems theoretically. The question is how well it will work altogether.

Uses heatpipes man, so the energy is spread to the entire base plate, you can see from their thermal imaging that it works fairly well as the base plate is all the same temperature : ]

[de.das.dude]

true the thermal image showed that the top of the fins were the hottest. i guess they use a flat heatpipe

So if they HAVE TO use heatpipes. why not go traditional

[Aquinus]

I wouldn't call this fan-less, it's just making the cooler itself act as the fan. I would be really curious to see how well this works and what happens after long periods of use because any device that moves air is susceptible to dust buildup and I'm sure that after a while that air cushion that the cooler rides on might not be as big if dust does build up. I would also imagine that this thing would have to be balanced perfectly to work properly. I don't think I would be an early adopter, I'll let others figure it out.

[pantherx12]

Originally Posted by de.das.dude

So if they HAVE TO use heatpipes. why not go traditional

What do mean, like tower heatsink?

because this new design is more efficient.

It's cooling performance far exceeds that of it's size and weight.


If this thing wasn't better than a conventional heat-sink then they would of stopped production after the first prototype last year


That they are thinking of using this as a condenser inside ACs goes to show how efficient this thing is the radiators in ACs are HUGE. Even small ones are around 20 by 20 cms and about 3-4 cm thick. And two of these things can supposedly replace all that!

Originally Posted by de.das.dude

i dont think that area will be accessible to dust. that will probably be sealed like a ballbearing.

and 0.03mm is really small. its almost the same as the gap between fins and heatpipes that pass through them
also if you look at their experimental darta, air is kinda conductive at that thickness.

The area won't be sealed, it's open so air can spin and rush through it. The speed of the air will force dust out of the air bearing ( if the air doesn't the plate spinning at 2000 rpm occasionally hitting particles will force them out)

And the main reason it works well as a conductor is not the thickness it's the speed of air/constant fresh flow. ( like water flowing through a block, in fact if you imagine this device submerged in water instead you'll probably be able to visualise why this thing is effective. )

[de.das.dude]

"speed of air will force dust ou"

LOLOLOLOLOLOL

so thats what happens when the high speed air goes through heatsink fins.

theoretically its ok. but practically we have to deal with not only dust but lint too. and animal fur etc.

[theeldest]

Regarding the "dustless" claim:

The explanation for this is in the video but seems lots of peeps here missed it. It's due to the centripetal force. <insert whole spinning bucket on a string analogy>

What this means is that any dust particle experiences a force about 100,000x to 150,000x the force of gravity outward from the center. This is why most fans don't collect dust but the heat sinks do.

I think it's pretty realistic for this to be considered dustless, especially when compared to conventional non-passive coolers.

[phanbuey]

Originally Posted by theeldest

Regarding the "dustless" claim:

The explanation for this is in the video but seems lots of peeps here missed it. It's due to the centripetal force. <insert whole spinning bucket on a string analogy>

What this means is that any dust particle experiences a force about 100,000x to 150,000x the force of gravity outward from the center. This is why most fans don't collect dust but the heat sinks do.

I think it's pretty realistic for this to be considered dustless, especially when compared to conventional non-passive coolers.

The explanation is fine, but most fans DO collect dust and MUCH more so than stationary components, in reality. i.e. the back of your videocard is always going to be much cleaner than the fans in your system (unless you have a fan blowing air on to the back of your videocard).

The same principle will apply here. Dust will get in/on this cooler, especially if they put a shroud on it to keep from lopping off people's digits, as that will just act as another surface for dust to latch on to.

[cadaveca]

Originally Posted by phanbuey

The explanation is fine, but most fans DO collect dust and MUCH more so than stationary components, in reality. i.e. the back of your videocard is always going to be much cleaner than the fans in your system (unless you have a fan blowing air on to the back of your videocard).

The same principle will apply here. Dust will get in/on this cooler, especially if they put a shroud on it to keep from lopping off people's digits, as that will just act as another surface for dust to latch on to.

The part that is missing here is that most fans do not rotate at a high enough speed for them to not have dust collect on the surface.

That is part of the deal, they tested to find the optimal speed to prevent dust collection.

[phanbuey]

I don't know... I'm skeptical, i gotta see it to believe it. But having seen 2K rpm fans collect plenty of dust on the blades, my initial gut reaction is to consider this BS.

Also, the fact that they will have a shroud or a grate of any kind to safeguard the user almost guarantees dust.

If they said it was "dust-resistant" or "optimized to collect much less dust" then great! But "dust-less"? No way. Someone in marketing got a bit too excited.

Also not that I care - if this thing cools well, i would buy even if it collected MORE dust .