nanofluid

relttem · Oct 16, 2008

I have been involved with an academic research project for the air force that involved fine tuning nanofluids to cool electronics. I currently have some running in my liquid-cooled computer at the house. I was wondering if anyone on here has heard of anything like this? We are trying to see what kind of interest there is. Nanoparticles increase the heat transfer capabilities of the fluid without any detrimental effects to the pump, tubing, etc.

We (my friend and I) sent samples out to some people (TomsHardware, Anandtech, Xtremesystems) in order for them to review. I was going to put a link up to one of the other forums that has some results, but I wasn't sure how that would go over here.

I am just wondering what everyone's thoughts are.

thanks

hellzarmy · Oct 16, 2008

can you post what is in it ?

other than being fluid in motion in the loop, i cant see anything doing better than just the standard destilled water.

im just curious, thats all. i understand if you want to keep it hush hush too

ShadowPho · Oct 16, 2008

mmm... looks cool
http://en.wikipedia.org/wiki/Nanofluidics

supra · Oct 16, 2008

few threads over at xtremesystems on it, seems interesting, at idle its equal to water, but the few people testing it have reported that it performs 1-2c better with higher heatloads

relttem · Oct 16, 2008

these are the other forums that we have been having discussions on. There are a lot of posts with questions and answers, and some test results (the tests are on-going)..But, if you have any questions I will be happy to answer them. We presented a paper at the ASME Summer Heat Transfer Conference in Jacksonville last summer dealing with the research. I can send you a stripped down version of it..basically it hides any information that would let the fluid be replicated. It goes over our test apparatus, methodology and results. Just PM if you want a copy and I will email it to you.

here are the links to the other forums:

http://forums.extremeoverclocking.com/showthread.php?t=303327

http://www.xtremesystems.org/forums/showthread.php?t=202442

http://www.tomshardware.com/forum/248371-29-nanofluids

thanks

jmh547 · Oct 16, 2008

Hellzarmy, I took a nano fab class my last year in college, i know what you are thinking and i will try to explain how it would benefit. With nanofluids you are double the surface area increasing the efficiency of the energy transfer with out adding volume.

Nano science is all about the scaling. When you are cooling you are talking about the exchange of energy, the faster you can exchange the energy the cooler the processor stays. In the terms of heatsinks: the larger the heatsink the better the cooling. Why? because it has more surface area.

Take that same heatsink and double the number of fins but keep the same volume. Basically what you did was doubled the surface area creating a more efficient heatsink.

relttem please correct me if i am wrong...

relttem · Oct 16, 2008

you are definitely on the right track. The nanofluid can remove more heat than plain water. What we found was that the convective heat transfer coefficient increases with the addition of nanoparticles. But, you have to be careful on the amount that is added. That was one of the points of our research. If you add too much the density of the fluid changes, which will have adverse effects on the pump, and pumping power.

the heat transfer occurring at the chip in your basic water loop is governed by the equation (for the most part)

q = hA(T2-T1)

q - how much power your CPU is puting into the liquid (Watt)
h - convective coefficient
A - surface area of the waterblock that the water actually is in contact with
T - temps

We what are hoping the nanofluid will do is let your CPU put more watts(get hotter) into the water (as in the case of overclocking) and the nanofluid remove it without a significant change in temps. Which means you can overclock more.

hellzarmy · Oct 16, 2008

the only thing i have a har dtime understanding is

how can you double teh surface area of a fluid?

while im not saying this product is bad, i just dont see how it can draw heat so much better than water. now i understand fluids have different thermal characteristics, i just dont see how this can do much better than plain destilled water.

ill look into this more to hopefully understand it a bit better.

relttem · Oct 16, 2008

the Nanofluid is not increasing the area of the water. What it is doing is increasing the convective coefficient. If you look at that equation I posted above you can see that as that rises, and your area and (T2-T1) remain constant, the 'q' has to go up. It wouldn't balance if it didn't. If you add enough nanoparticles to the water you can start to change the thermal characteristics - specific heat, thermal conductivity, density. All those come into play in good and bad ways. We optimized the concentration, which was our goal.

mbigna · Oct 17, 2008

Well, as with all things, there are tradeoffs. You can add particulate matter to liquids, and while you may be able to increase the heat capacity of the fluid, you will be changing its density and viscosity. It's always a challenge to find the right balance in order to maximize how quickly you can move heat away from your source.

You say that your substance has no detrimental effect on pumps and tubing, but even plain old water has a detrimental effect (ask the Grand Canyon what it thinks about the Colorado River). When adding [non-dissolving] particulate matter to a fluid, no matter how small, you will increase friction and add some erosive properties over and above the fluid alone.

The main question I have is--how much will this stuff cost to add to your loop. You can get a gallon of distilled/deionized water pretty cheap.

Conumdrum · Oct 17, 2008

I figured out the formula. It's Unobtanium particles and 50% distilled.

Yea, a joke okay, just had to do it, sorry..........

relttem · Oct 17, 2008

mbigna said:
You say that your substance has no detrimental effect on pumps and tubing, but even plain old water has a detrimental effect (ask the Grand Canyon what it thinks about the Colorado River). When adding [non-dissolving] particulate matter to a fluid, no matter how small, you will increase friction and add some erosive properties over and above the fluid alone.

If your system can run for 100million years without having other problem, then you will see some problems due to the nanoparticles. Our density didn't change. We kept the concentration low enough not to change that, so any problems related to density changes will not happen. We also looked into the current draw of the pump related to the nanofluid concentration, which would reflect any changes that the nanoparticles subjected the pump to and found none.

Friction in fluid flow is a function of Velocity, Pressure, density, and pipe diameter. None of those will be different from water. So, your friction, as compared to water, will be the same. Again, if you add enough particulate matter to the fluid things will change and the effects will be seen. But, adding that much nanoparticles was not necessary.

these are all great questions, and were basically the same questions we had going into this.

Conumdrum · Oct 17, 2008

After a test using standard popular WC parts (no aluminum) for a year 24/7 and before/after pics, temp results day one to day 365, and other data then I think it will interest me. It's gonna be hard to reduce temps more than a few C with any liquid that works pretty good.

More important than a few C (due to overradding mainly) is clogging, tubing clouding, UV and dye long term color viability. Cost too.

I await your results and hope you make the magic stuff we all want. I'll buy it when it's proven long term.

MrLarkins · Oct 17, 2008

2 questions I haven't seen answered (maybe I missed them)

1] whats the cost?
2] what about growth? bacteria and algea?

relttem · Oct 17, 2008

there have been no bacterial growth of our nanofluids that have been on the shelf since last summer. BUT, we are going to get that verified. There is a surfactant that inhibits the nanoparticles from agglomerating (sticking together) that we believe inhibits the bacterial growth.
I can't speculate on the cost yet. First we want to see what kind of reviews we get from all the people to determine if we even want to offer it. We really haven't gotten that far. We would first have to determine what quantities we want to sell (12oz, 24oz, 1 liter). My water cooling system only needed 24oz, while some the systems that are testing for us require a liter. We have patents applied for, and we have to get all the operations down so we can guarantee that everyone is getting the same stuff. We have had requests to make it certain colors, but I think we are going to leave that up to each individual. I know you can buy dye for that. There is still some work to do.

hellzarmy · Oct 17, 2008

is this a pure mix type fluid, or can it be mixed with water?

what about additives like dye's are whatever else?

also out of curiousity, what is its freezing point?

relttem · Oct 17, 2008

it is mostly water, so freezing point would be the same. What we are thinking of sending out is a liter, or a small vile that would be mixed with a liter. The mixture we send would be optimized, but you can add more water. Littleowl on one of the links above diluted it.

gsrcrxsi · Oct 18, 2008

IMO you are going to be ultimately limited by how fast the radiator can remove the heat. you can absorb the heat into the liquid as fast as you want but if the radiator cant match it, it wont matter and the total cooling capacity will be up to the radiator.

supra · Oct 18, 2008

the fact is though that many people have more radiator then they really need

i remember when I first started watercooling, the original BIX single 120 was the best radiator we had, almost noone ran multiple loops, everything was cooled with a single 120 radiator, maybe the addition of a second rad in the loop

and its not like cpu's we're running much cooler back then, remember what the early p4's used to put out heat wise? the lowest number I remember seeing was 89 watts stock

relttem · Oct 18, 2008

that is a great point..that, and having too much fluid or the tubes are too long. It is really not very difficult to size all the stuff. You have a few equations that you have to solve, but most of the equipment gives you all the information. So, you just have to plug some numbers in. But, the radiator will dictate how much heat you can get out. Also, the air temp blowing over the radiator. Most radiators can remove some serious heat. After that, you are governed by the convective coefficient, h. That is where the nanofluid comes in.

nanofluid

Registered

Member

Member

Member

Registered

Member

Registered

Member

Registered

Member

Member

Registered

Member

Member

Registered

Member

Registered

Member

Member

Registered

Similar threads