Sign in with
Sign up | Sign in
Your question

nanofluids

Last response: in Overclocking
Share
August 5, 2008 6:18:01 PM

I am currently doing a lot of research dealing with the effects of nanofluids on heat transfer and was wondering if anyone else has used nanofluids as the cooling liquid in their liquid cooling systems? I currently have it in mine.

thanks

More about : nanofluids

a c 324 K Overclocking
August 5, 2008 7:04:49 PM

Sounds very interesting...where can you find such a product and its information pages?

Edit: I realize there is a google...I just meant any site in particular...like the stuff you are using, etc?
August 6, 2008 12:25:21 AM

We are looking into starting a company that will make this available. We contacted Toms hardware trying to get a review but haven't heard back. We are looking at some other places for reviews too.. keep an eye out for it. I was just curious if anyone was using it yet.

thanks
August 6, 2008 1:12:06 AM

Well i did some reading it does look promising depending on whats in there,cost!, ambient differential wish reltem had more to and or toms:) 
August 6, 2008 1:31:59 AM

OP, you got a spec sheet or anything?

I am on water, and would be interested in such a thing, price, lifecycle, etc prove ok.
Anonymous
a b K Overclocking
August 6, 2008 1:43:22 AM

what is nanofluid? does it offer any benefit from water? can it be used in standard water cooling loops?

come on please fill us in before you start asking about stuff most people don't know about
August 6, 2008 12:24:44 PM

I will work on a spec sheet. We have been doing research based on different nanofluids flowing over a heated flat plate at local Reynold's numbers from 12,000 - 400,000. We have shown a 16% increase in the amount of heat transfer provided by the fluid. A nanofluid is a fluid that contains a small amount of nano-sized particles - for those of you that don't know. You can find a lot of info on the web and at MIT's engineering dept. My computer at the house has a water cooling unit on it, and the chip temp dropped almost exactly 16% when I put the nanofluids in. Still, we are going to wait to have some sites review it.
a c 324 K Overclocking
August 7, 2008 1:32:29 PM

I would be interested in testing in my watercooling loop, if it is as effective as you say it is. Any way to get my hands on this, or do I have to wait in line like everyone else...? :) 

What information can you give, as in, what are the nanoparticulates used, does it accumulate in pumps, blocks, etc? I didn't know if there was a standard product used...
August 7, 2008 2:36:51 PM

there is no standard used. We did a project for the air force looking into the benefits of adding nanoparticles to the heat transfer fluid. For those of you not aware, a nanometer is 10^-9. It takes 100 nanoparticles stacked to be as thick as a dollar bill - to give you an idea. Our density is too low to have any detrimental effects on the pump or pumping power.
But, as I mentioned, we are going to wait until some reviews come out. Soon.
August 7, 2008 4:09:22 PM

Aprox how long do you think for the info?
August 7, 2008 6:01:55 PM

I'm guessing the heat dissipation properties of nano fluids is similar to basic heatsinks ie. the more surface area (by the nano particles being so small) the more efficiently you can transfer heat?
August 7, 2008 6:19:31 PM

I can't give a time-frame yet.

we found that it improves the convective coefficient. There is a heat balance -> Q(power in) = h(conv coeff)*A(area)*[Tout-Tin]..therefore, with h increasing you can increase the amount of Q and not have a change in the Tout.
August 7, 2008 6:29:13 PM

I've just twigged after a third read through (it's late ;)  ), why use nano fluid as an additive when you could use as a replacement? Is it because it would transfer the heat quickly, but not be effective as water at dissipating it so end up just getting very hot?

*EDIT* Also as the fluid is smaller would it increase the chance of your water loop to leak?
August 7, 2008 7:59:24 PM

the liquid has nanoparticles in it..it wouldn't be an additive. You would use it in replacement of the liquid in the system now. Plus, you have to be careful of how much of the nanoparticles you mix into the liquid. You can't just dump some in and hope for the best..too much and you'll run into pumping issues..too little and nothing will happen. We would supply a total replacement for the liquid in the system now.
Your conv-coeff is increasing..that works on both sides of the system; the heater block and the radiator.
The fluid isn't smaller..if you hook your system up correctly it won't leak..if it leaks with nanofluids it will/would like with water.
a c 324 K Overclocking
August 7, 2008 8:31:11 PM

Send me a sample. :) 
August 7, 2008 9:48:26 PM

Would the 16% decrease in temperature over standard water be represented in the price per litre or will we be paying for the 'tech' as well? I for example wouldn't mind paying maximum £15 per bottle taking into account my system uses almost 2 litre's in a 2 meter loop, if it needed a 6 monthly refresh.

How often would it need to be replaced? 6 monthly would be minimum, yearly would be ideal?

Is the combined fluid non conductive/ anti stain for that rare leak so we can recover/clean our systems to the original state?

Is it anti corrosive ie mixed metal water blocks etc in the loop?

Would it go stagnant if say the computer wasn't switched on for over two weeks? Again don't know the specifics, but over time would the fluids separate if left stagnant?

Can't really think of anything else to ask that others haven't... Tiz asking quite alot, but I'd personally like to know and as my Great Grandad told me as a child looking aimlessly at sweet's in the window, 'if you don't ask you don't get' :D .
Anonymous
a b K Overclocking
August 7, 2008 10:13:56 PM

sounds very neat now that I know a little bit more about it...

I want some!!!
August 8, 2008 11:22:57 AM

Curious if you have done any reliability testing with your fluids and pumps. One of the big issues with water is calcium & copper scavaging on the impeller blades eventually causing failure. Having zillions of nano particles floating around I'm thinking that they may start collecting, over time, on the impellers or on the sides of the vessel. Also, is this stuff RoHas compliant? How do you dispose of it, is it toxic? What happens in a leak event, is the MOBO a complete write off? If the stuff evaporates after a spill, what happens to the nano particles? Do they enter the atmosphere? Is that toxic to breath? You can pump mercury through a loop and get great thermal transfer but you'd probably not want to do that.
August 8, 2008 12:22:07 PM

We have MSDS sheets for the nanofluids. It is not toxic, but I wouldn't drink it. As for pump degradation etc, we have done pump power studies which is related to the density of the fluid, which would indicate how coarse it is. There has been research done on the scouring effects of the nanofluids, which have shown none. Our density does change, but not to the extent to which any pump damage impeller damage occurs, nor to the amount of power required to pump it. If you add a lot of the particles to the liquid and your density is high, then you will have problems. That was part of the research; to find the optimum concentration. When it evaporates the particles do not go into the atmosphere, you would simply have to clean the MoBo off with a rag - like cleaning up any liquid spill. Of course, there is a lot of common sense involved; you don't want to inject it into your veins, or boil it in a pan and breathe in the vapor, drink it, snort it, feed it to your pets, pour it in your fish tank etc etc..just like with antifreeze.
a b K Overclocking
August 8, 2008 2:58:58 PM

Cool work. I wish I did this for my science fair project.
August 8, 2008 4:39:40 PM

closed_deal said:
uses almost 2 litre's in a 2 meter loop, if it needed a 6 monthly refresh.


2 litres?..
August 8, 2008 10:32:25 PM

yeah my loop and res holds just under 2 litre's... isn't that normal??
August 8, 2008 11:23:35 PM

Mine holds about 20oz..not even half a liter..
August 8, 2008 11:48:36 PM

hmmm... I have 2 meter's of 3/8 tubing connected to one PA 120.3 Rad, one PA 120.1 Rad, two 3870 full cover water blocks a 5 1/4 drive bay res and a EK Supreme cpu waterblock. The res alone hold 500ml... so i don't think i'm to far off average :??:  .
August 11, 2008 2:47:45 PM

This looks like some pretty interesting stuff. Are the particles metalic in nature or what? I'm curious here and would love to get my hands on some of this when i put water on my rig this fall
a c 324 K Overclocking
August 12, 2008 2:23:24 PM

I think I am getting close to a gallon...LOL. Actually, not sure how much coolant I am running...prob a lot with those big side tanks.
August 12, 2008 3:38:22 PM

Lol a gallon, mind those side tanks do look sweet, btw why have 2?
a c 324 K Overclocking
August 12, 2008 5:52:08 PM

I originally wanted 1 big one, but then I started to think about the diameter of the tubing...3-4" would stick out quite a bit and get pretty top-heavy. I just went with 2" tubing and made 2.

I originally wanted to mount it/them vertically, but it would have made the flow into the side rad and back difficult and asthetically messy.
August 12, 2008 6:06:47 PM

I agree, seems you made the best out of an awkward situation :D 
August 21, 2008 5:49:02 PM

have you ever crunched thru the numbers to see how much water you actually need? You can get a rough idea with some simple thermo Q=mdot(h2-h1). You just need to know how many watts you want to lift and your temp change. I would just be curious to your mdot.
What is your flow rate now?
August 21, 2008 7:27:00 PM

The way I worked out how much water i needed was to connect everything, fill the res, jump start the pump (MCP355) via a PSU jump start cable i bodged, dry run it until it sucks the water through then kept on filling the res till she were almost full to allow air bubbles to escape for the first 24 hours.

I didn't use math as such, 214W @ load per Gfx card so 428W then going by: http://www.thermochill.com/PATesting/#PA1203 I worked i need just a PA120.3 to dissipate that much heat with 74cfm fan's (although mine are much quieter) and have room to spare to turn PC into silent mode overnight. 65Watts max TDP for the E8400 plus the most important overclocks probs goes to around 110-140W so required a 120.1, these all together suck up quite a large amount of water!

The tubing length is needed to the way i modded my case: http://s232.photobucket.com/albums/ee276/closed_deal/

The pumps rated at a 4.7m del head at 600LPH. I wanted a pump capable to handle slightly restricting 3/8" ID hose with i guessed a 2.5 meter head minimum just to get it moving slightly. My ambient temps range from 14-23C (typical UK weather) Tj at 4.0ghz at 1.33vcore (1.4Bios) is 39C core 1 & 34C core 2 at idle, load 56C Max core 1 & 50C Max via Real Temp 2.70. Gfx Load 43C Max & 35C idle at 850Mhz core 2402 Mem times two for two gfx cards.

Haven't been trained in thermodynamics, I'm a avionics tech (wiggly amps are my thing ;)  ) so please explain to the un-educated :) .
August 21, 2008 8:16:56 PM

Q is your heat load..total in watts. h is the enthalpy of water measured at specific temps in KJ/Kg. You can probably find it online. And, mdot is your flow rate - Kg/s. So, if you know your temps and Q you can solve for mdot..etc etc. This thermo approach is telling you how much heat you need to heat water (or any liquid with know values of h) at a specific flow rate. It doesn't take into consideration a lot of heat transfer stuff, but it will give you a good ballpark number. You pretty much have all numbers. You would have to make some assumptions/guesses at your temps before and after the cooler blocks, but you can assume the water temp coming out of the radiator is the same temp just before the cooling block - which can never be cooler than ambient (ambient would be a BEST case scenario). And, since you know Q, mdot, and T1(which can be used to find h1) you can solve for h2 and get a temperature for your water coming out of the last cooler block. Then, you can slow your Q down and see how that effects your T2. If any of that makes sense...:) 
August 22, 2008 3:56:15 AM

Physics major has become interested. (haven't read the last 16 posts or so, will do that right away) So has the experiment been done yet, if so what were the results?
August 22, 2008 2:24:26 PM

the nanofluids experiment? Our lab experiment has been done, but we are still waiting for the reviewers to post their results. We didn't do our experiment in a computer, but the physics/heat transfer is the same.
August 22, 2008 11:56:15 PM

I see. What was the average temperature decrease when compared to a liquid coolant? Also do you have any sort of data or sheet you could post?
August 23, 2008 4:06:05 AM

we got a 16% increase in heat transfer in the lab, and I got about a 16% temperature drop in my system compared to just water. If you send me your email I can send you the paper that was presented at the ASME Heat Transfer Conference last week. It will be stripped down' meaning that you won't know what the nanofluids/particles are nor how much was used, but the data is there and how we tested it. It will be a PDF file.

thanks for the interest.
August 23, 2008 5:14:47 AM

Well if you sent it to THG it would be nice to hear from someone since they watch all the threads to at least hear them say we are working on it , lol ,we don't care or some thing! sorry if i am to straight to the point i am not trying to be a a$$ or sarcastic, i am old and straight to the point :)  or :( 
August 23, 2008 8:56:29 PM

I filled out all the forms necessary for TomsHardware to review it, I just haven't heard back from them yet. Some others are working on it now tho.
August 24, 2008 9:31:13 AM

TOM???????????????????????????????
August 26, 2008 3:49:02 PM

well?
a b K Overclocking
September 7, 2008 9:08:38 PM

@OP: PM Crashman or Cleeve
September 15, 2008 5:10:43 PM

I got a response, and we hope to get some nanofluid in the mail to him this week for testing.

Stay tuned
a b K Overclocking
September 15, 2008 6:29:11 PM

Good luck.
September 15, 2008 9:09:21 PM

Shadow703793 said:
Good luck.
+1
September 22, 2008 9:40:08 PM

here is a link to LittleOwl's testing. He has ONLY gotten it set up so far and hasn't really done any benchmarking.

http://www.xtremesystems.org/forums/showthread.php?s=af...

he has a few links to some other threads in that one pertaining to the nanofluids.

If you have any questions let me know. We also mailed some to one of the testers here, and, as I have said anandtech got some, as well as another guy at Xtreme. We are currently trying to track down a specific MaximumPC magazine author to do some testing.

thanks
September 22, 2008 10:04:08 PM

Looks good so far, awaiting the non-diluted heat intensive tests :bounce: .
September 22, 2008 10:14:02 PM

the diluting will deduct from the maximum amount of enhancement that he could have achieved. I am going to try to back out the enhancement he should get with his diluted tests to see where it falls in line with our lab tests. I will put a graph up on his thread to show it. Regardless, it should still be interesting to see what everyone gets.
September 25, 2008 6:34:12 PM

the nanofluid looks yucky c)=
make it in uv red ASAP! lol, can't wait for results!
September 25, 2008 6:55:16 PM

We are going to make in UV Blue...sorry..:) ..that is the plan anyway
!