Custom Acrylic w/ Novec 7000

[jaov2k]

I've seen the youtubes on the Novec 7000 stuff and found it to be informative. But I was wondering why they didn't just fill the entire acquarium to the top, inserting part of the heat sink fins into the fluid?

Is it necessary to leave room at the top for the expansion of the fluid at boiling or is it ok to completely fill the container?

I'm thinking about building a horizontal acrylic case, filling it with Novec 7000 all the way to the top, ala http://www.iceotope.com, making the top panel of 16oz copper (~0.032" sheet [0.81mm]). Possibly brazzing aluminum fins to the copper, sealing contact surface edges with silicone to avoid electrolisys since the presence of an electrolyte is avoided. Correct me if my understanding of the electrolisys process is wrong here, please.

I'm trying to build a pc with no moving parts, cooling included. So no fans on the fins.

Thanks

 

[inzone]

I have come across different types of immersion computer builds and you can use different fluids , currently Novec 7000 may be an expensive liquid to use and experiment with. Maybe that's why on the youtube video they didn't go all the way to the top with the Novec 7000. An immersed Pc has been proven to work and heat has been controlled. I use water cooling and have been for many years and once while I was in the middle of changing radiators I used a fish tank as a resivoir/radiator with a submersable pump (fisf tank variety) and while it worked there was a lot of condensation from the hot water and the fact that there were no fans to cool the water the liquid reached a temp and stayed there. So I found that by using a large tank of water to cool with there was no way to cool the water and it slowly raised the temp of the water and the temps of the hardware were higher then when using radiators and fans. So in your build you may want to figure a way to get some cooling effect for the liquid. The fins that your suggesting may or may not do it but that will be something you will find out if you end up building it.

 

[jaov2k]

I guess I'm going to have to determine if a copper sheet or even a plate will be required to keep up with the boiling fluid. Such that it condenses as quick as it evaporates upon touching the copper that is at room temp, 22C. I don't anticipate the CPU getting any hotter than 75C. If it shoots higher them it will be temporarily. To render a scene or something.

I think I have to google calculating heat dissipation rate of copper... Wikipedia here I come?

 

[inzone]

I doubt that you will get the liquid to a boiling point and if you do that will be too high for the cpu and it will shut down. 90c is the thermal threshold of a cpu so tha's 194f and just under the boiling point. (212f and 99.98c) Even with a large amount of liquid it will eventually get to a high temp and while it won't get to a boiling point it will still be higher then with a radiator and fans.

 

[jaov2k]

I doubt that you will get the liquid to a boiling point and if you do that will be too high for the cpu and it will shut down. 90c is the thermal threshold of a cpu so tha's 194f and just under the boiling point. (212f and 99.98c) Even with a large amount of liquid it will eventually get to a high temp and while it won't get to a boiling point it will still be higher then with a radiator and fans.

But novec 7000 boils at 34C... maybe I'm not following what you are saying... I understand that no cpu should maintain a 90C temp for any extended amount of time.

The goal is to achieve a passive cooling solution by using the characteristics of the chemical. I know that adding secondary cooling solutions would only help control the temps, but I'm trying to use things which are passive and do not consume additional power or noise.

I'm thinking perhaps introducing heat pipes which would recirculate the vapor to the bottom and sides of the tank where temps are lower, condensing the fluid for application once again... I gotta think about this one though...

The constant is that the room that it will sit in will always be 22C. Therefore the fluid will always condense given the proper conductor and time. Maybe I'm all messed up...

At this point in the project I'm going to have to decide how to adequately incorporate the novec 7000 into the design. It's whats at the heart of this whole exercise. The goal is to utilize this product's low boiling point as the primary temp regulating medium to the whole system, and only introduce passive cooling methods as secondary cooling systems to aid the primary such as heat sinks, heat pipes, water reservoirs, more heat sinks.

I also haven't done the cost forecast of this little idea. I suppose that while the metals, acrylic sheets, and of course, the novec 7000 have a heafty up front cost. They will eventually pay themselves off by not having to power fans, pumps, even a condenser. And of course the noise and install complexity of other liquid cooling methods. Also, it should be taken into account that this is a 100%, ALL components, are immersed and cooled, versus specific ones. So there's lots to consider, but I have hopes.

Plus the "cool" factor is HUGE!!! No pun intended.

Thanks for the feedback thus far guys.

 

[inzone]

Guys ? So far I'm the only one responding. So when you say no moving parts are you also not going to be using any pumps of any kind? I can understand your thought process to not have a conventional cooling design and to do this with liquid and no radiators and fans but to have no movement with the liquid will make it harder to cool.
With a phase change cooling system a gas is used that has a certian temp in which it changes from vapor to liquid and then back again to vapor and this process can be used to cool certian things like and air conditioner cools a room and a closed loop setup can be used to cool a copper block that absobs heat from a cpu. You could think of this as a large condenser and have the novec 7000 change from a liquid to a vapor when it reaches it's boiling point and then back again as it's circulated to the bottom of the tank. However this process would require a pump. I'm just spitballing here since your looking for ideas on how to do this and maybe something like this will be useable or maybe not.

 

[jaov2k]

Oops... Sorry about the "guys" reference. Not sure why I assumed a different responder.

I got some ideas which I can use conventional steam to test out. I want to see if I can spin a small turbine that in turn would drive the pump. Nothing fancy. Just dreaming out loud... I appreciate you letting me bounce things off you.

I think it's doable since there is already a power source. It's just a matter of spinning it to get the most usage out of it.

 

[inzone]

Innovation is always a good thing and you can come up with a new way of doing things by doing things like this. For me I like to modify things and I apply that to my computer case and I try to see what I can do to fit things and maske it better. I have done some heavy modifacations to my current case and it still looks good and you would think it came like that. So that's why I can relate to what your trying to do.

 

[jaov2k]

I've made a bit of progress on this, establishing two options: the 3M and the Dupont. Both have pro's and con's; the 3M offers performance at a premium, not just for the fluid but is also design and manufacturing is more difficult to work with since it doesn't play well with elastomers. Dupont on the other hand is more affordable, does play well with a range of other materials, buuuuuut its performance isn't as good.

At this point I need to confirm that several looped heat pipes within the enclosure will be enough to mitigate the heat across the lower half of fluid. I know that sounds unusual. Once I complete the sketches I'll post it for critique.

 

[fir3murd3r3r]

From this Youtube video:http://www.youtube.com/watch?v=Z_X_hgtlJpA it looks like they use the space to allow the fluid to condense on the radiator to allow the heat to be dissipated to the air. I have not looked in to this fluid enough but they could be doing this because the fluid has a property that allows a large amount of heat to be removed from the system once is condenses.

As to your project, I was thinking of if your ever wanted to overclock and run really cool, try adding a liquid cooling system, such as a water cooler or even if you really want/can try getting a radiator that you can mount outside and pipe the heat outside instead of back into the room (this might even save you more money on that aspect since 700Ws (just to give an idea) can put off about the same amount of btus as a small space heater).

Just trying to help you get ideas if you haven't completed the project already.