I think reducing the high temperatures of that rendering PC is a secondary function of the setup. When they introduced the concept of a whole room watercooling setup (I believe it was on a WAN show a few months ago), they said the principle purpose of this was to reduce room temps. Otherwise they could simply do a normal watercooling setup for each computer.
I'll admit that I don't know what the final setup will look like (i.e. they could still wrap the copper tubing in insulation), but if their setup stays the way it is they will have some real problems down the road.
I do hope I'm wrong, maybe they do plant to insulate the copper tubing and keep the rads in an environment without subfreezing temperatures. I like their content and don't want to see them fail. But so far they have made no mention of any such plans and it is ingrained in me as an engineer to identify shortcomings and potential failure points in a design.
Incorrect, Q (heat transfer) = h (heat transfer coefficient) * A (surface area)* [Ts (surface temperature) - Tinfinity (ambient temperature)].
http://www.engineeringtoolbox.com/overall-heat-transfer-coefficients-d_284.html
Copper (and materials in general) transfer significantly more energy to water than to air.
Edit: I suppose you do get a much larger delta T with air coolers than water coolers, so the heat dissipated would be similar. I guess you are correct because the heat in a cpu/gpu ultimately comes from the internal resistance of the component. I was considering Q to be the variable, though Q is fixed based on the components energy consumption making the only variable Ts.
I was wrong...
What I want to know is what posessed them to put a 6 core i7 with crappy referecne cooler, GTX770 and Quadro 4000 is that wholy inappropraite Silverstone case??
There was a build about it in a video about a year ago, or so. They wanted to make it as small as possible, I think. And also replaced the Heatsink to prevent throttling (ironic, huh) If that's the same case/video I remember.
Might already have been said , but :
He said those where if the PC's needed to be independent , they will not be used .
The thermal loss from copper tubing would be minimal .
The bare copper tubing would dump the heat from the PCs back into the room. Copper is a great heat conductor. They want it to all go outside, so the output pipes should be insulated.
I still think they should leave the intake pipes bare, though. it would cool down the room significantly, bringing the cold water through pipes in the room.
I don't have a problem admitting when I am wrong, though in this case it is particularly embarrassing. A reactor as just a fancy hot water boiler so I took quite a bit of thermo/fluid dynamics and heat transfer when I was a student. In my defense however, reactors are load following meaning the more power you draw from them, the more power they produce.
We all get it wrong sometimes. Just glad some people can accept they are wrong and move on.
the pump will be loud due to the fact that it will be a sump pump or something to run that much fluid at those heights and distances.
But it should be outside, so that shouldn't matter.
I think they should use a positive displacement pump with no relief valve. #sarcasm
All good suggestions, and I agree that the intent may be to add insulation in a later video, though the radiator sitting right outside those cases baffle me.
Again, the issue I have with a heatexchanger on the exterior of a building in Canada is the potential for the coolant to freeze and/or lead to condensation buildup inside the case. Placing the heatexchanger in a climate controlled environment would be ideal, but the reason Linus is building this setup is to avoid having to pay for climate control.
Most PC watercooling setups utilize a reservoir that resides above the pump (also to provide minimum suction head for the pump), and I imagine that filling this loop without a large reservoir would be a pain. As long as they don't fill the reservoir completely full (or keep it too low), they should be fine on the thermal expansion front.
It would only cool down the room in the winter time (or when the outside temperature is less than inside temperature), it would actually heat up the room in the summer time (though perhaps not as much as the computers would). That may be less than desirable on very hot/cold days. T
now my spelling error is immortalized
I think everyone is overestimating how much heat will be transferred to the air from the bare copper pipe. Sure copper is a great conductor but there isn't enough surface area or airflow on the pipe to radiate say an uncomfortable amount of heat.
That last bit is important, part of this is about comfort, which is mostly subjective. For all we know Linus has a completely unobtainable goal for the reduction in room temperature or he just wants the place to be less like a sauna.
Here is something I don't think they thought about. Centrifugal pumps in parallel don't add to the pumpimgability they thing it does.
They could have two pumps in series, one on each side of the heat exchanger (I would assume that is where the largest dp in the loop occurs).
Or they could use a positive displacement pump lol.
Edit: Did they ever disclose their pump configuration?
It sounds to me like a bunch of us just really want to do this. They will tell us what they did whenever they get around to releasing the videos (no idea why they are waiting so fucking long).
"It sounds to me like a bunch of us just really want to do this." LOL +1
I wouldn't be too concerned with that, it's quite a easy fix, first off the coolent is moving around all the time, and there's a lot of it, so that will make it much more difficult to freeze, 2nd, they could just add some alchohol of some sort, whitch would lower the freezing temperature of the coolent dramaticly. So yeah, there's a few different solutions for the problem, and there's more than likely some other things you could do as well, I wouldn't be too concerned about. We've only seen a small part of the project, and I'm fairly sure that before they would start such a large project, they would plan it all out very carefully, and think about most scenarios and problems that they may run into.