project: Fukushima Reactor Rebuild

[AmEv]

^Ah.

[licenseless]

Originally Posted by mDust

I don't think that is a property of the oil itself, but of an uncirculated fluid in a tank. Pugets original build hit around 80C due to not being actively cooled because that's where the temp differentials leveled out. Heat transfer is governed by a temp difference and surface area: a 50C processor will dump heat into 40C oil faster than 49C oil given the same surface area. Likewise with the oil to the air...this is why cold, winter air over a watercooling rad can make a huge difference. Greater surface area, such as a radiator, will also increase dissipation rate. I would pump the oil through a rad just like water cooling. Otherwise, the oil can only passively dissipate heat via the air surrounding the sides of the tank. Just be aware that the head and flow rate will be severely decreased due to the extra viscosity. Viscosity decreases as temp rises though too. There's likely some magical, optimal point but that's far beyond the scope of this discussion and my mathematical abilities.

yes, yes, and yes hahahahahahahahah unfortunetly 15 dollars a gallon is what i saw too... so i was expecting to spend about 165... because i want to buy an extra gallon just in case... the computer should disapate about as much as the copper needs to fill up... so i should nee 10 gallons... but we will see.

these are all reasons why i am not sure what pump i am going to use.

the pump i have now is good... BUT... i am not sure its going to be just as good once things get moving... i dont have enough of an understanding of fluid dynamics to calculate at what rate the oil will disipate and at how fast or slow it will need to flow.

i was told though that any bit of cooling will be plenty... and that the setup i am using is good

i do however still have a small oil cooler from a car (hahahaha) that i can use if need be ... i would just put it behind the tank and run 4 80 mm fans on it... and boom cooling accomplished. But i am hoping the copper is over kill and there is no need for the radiator.

the pump is a trial and error issue... i got the pump for next to nothing a year ago... so i am using it... if its not enough i will upgrade and cut new hoses to fit it.

my only real problem here is that it will be ugly... hahahahahaha but it will be fun...

i do have the computer up and running right now out side of the tank and just in open air it is sitting steady at 32C

so i am sure everything will work fine

thanks guys.

[mDust]

Originally Posted by licenseless

the pump i have now is good... BUT... i am not sure its going to be just as good once things get moving... i dont have enough of an understanding of fluid dynamics to calculate at what rate the oil will disipate and at how fast or slow it will need to flow.

I couldn't whip out equations on this topic either so no worries there. All you need to be sure of is that the oil will never approach the CPUs tj maximum. Which is usually stupidly high (80-90C) for Intel chips and around 60-62C for AMD chips. I would try to keep the oil below 50C no matter the ambient air temp or how long the system has been running.

Originally Posted by licenseless

i was told though that any bit of cooling will be plenty... and that the setup i am using is good

For the most part, this is true. It depends on the tj max though. If you put an AMD chip in the original Puget system, it would have died pretty fast.

Originally Posted by licenseless

my only real problem here is that it will be ugly... hahahahahaha but it will be fun...

Once you get it running, hit it with a hammer and sprinkle iodine-131 and caesium-134 on it. It's post-disaster Fukushima! You will win awards... (Too soon?)
As a modder, fun is number one, and making useful or pretty things are 2 and 3...their order depends on whether you are the 'form follows function' type or not.

Originally Posted by licenseless

i do have the computer up and running right now out side of the tank and just in open air it is sitting steady at 32C

so i am sure everything will work fine

It's so cool because heat is transferred from heatsinks to the air, which rises away to the ceiling and draws cooler air towards the computer. If you let the computer run long enough, the room will heat up and the cooling effect will be slightly diminished. You can think of the oil bath as a really tiny room. Cool oil will carry heat away from the heatsinks, but will only rise a few inches to the top of the tank. Oil conducts heat much faster than air just due to the difference in density, but will heat up faster due to the smaller volume. Once submerged, you can expect temps fairly close to ambient for a bit, but it will slowly rise to equilibrium. The amount of active airflow, surface area, materials (copper vs aluminum vs glass vs acrylic) that make up that surface area and ambient air temp determine how low that equilibrium point is.

[licenseless]

Originally Posted by mDust

It's so cool because heat is transferred from heatsinks to the air, which rises away to the ceiling and draws cooler air towards the computer. If you let the computer run long enough, the room will heat up and the cooling effect will be slightly diminished. You can think of the oil bath as a really tiny room. Cool oil will carry heat away from the heatsinks, but will only rise a few inches to the top of the tank. Oil conducts heat much faster than air just due to the difference in density, but will heat up faster due to the smaller volume. Once submerged, you can expect temps fairly close to ambient for a bit, but it will slowly rise to equilibrium. The amount of active airflow, surface area, materials (copper vs aluminum vs glass vs acrylic) that make up that surface area and ambient air temp determine how low that equilibrium point is.

well that was the point of doing so much copper in the cooling tower. There is a a lot of copper... no acrylic ... all glass tank and no aluminum in the cooling line. The copper should do its job... and to make sure it does like i said before i still have an oil cooler from a car i could stash away behind the tank with fans on it to help the cooling process work like it should.

only issue now is i have all these stupid doctors bills from this lyme disease sitting at my desk right now... and priority is set at paying them off before i go to new jersey on the 22nd and 23rd... (and now maybe the 24th... my buddy just won free tickets to the Metallica concert that same weekend i am to be shooting video in new jersey so he wants to know if i can join him the second day) so i might have to wait until july 1st to buy this 160 dollars worth of oil... wow... 160 dollars worth of oil

thats a lot of oil... yet a spinal tap only costs me (after the insurance takes care of pretty much everything) 60 dollars

stupid speculators making oil cost so much hahahahah

anyways thanks for the info there Dust... its reassurance more than anything that i am going to fine hahahahaha

thanks for viewing... more updates coming soon... i will have the lights and stuff installed this weekend

[mDust]

I don't mean to imply doubt. What you have will work as is, but that equilibrium point may be a bit higher than you might want or expect. If that is the case, then you will need to install that intercooler.

The fluids (liquid metals and specialized, extremely low-viscosity oils designed for low temp baths) I was looking into for my sig project would have cost 10s of thousands for a 10gal volume...which ended that line of inquisition pretty damn quick! A couple hundred for mineral oil didn't sound bad at all in retrospect...

Getting sick is unproductive and prevents you from partaking in fun activities if you didn't know...knock it off!

[licenseless]

Originally Posted by mDust

I don't mean to imply doubt. What you have will work as is, but that equilibrium point may be a bit higher than you might want or expect. If that is the case, then you will need to install that intercooler.

The fluids (liquid metals and specialized, extremely low-viscosity oils designed for low temp baths) I was looking into for my sig project would have cost 10s of thousands for a 10gal volume...which ended that line of inquisition pretty damn quick! A couple hundred for mineral oil didn't sound bad at all in retrospect...

Getting sick is unproductive and prevents you from partaking in fun activities if you didn't know...knock it off!

hahahahahahahahahahaha

i will try and stop getting sick... i am normally a healthy dude... but this **** has been fun

i am thinking it will be about 10 degrees higher... but not completely sure where it will sit... im just hoping it doesnt get up to 60c to 70c... but again... i have the radiator for that...

[licenseless]

so like i said... i wanted this built so i could edit video and photography from this weekends event with my monster... well i couldnt get it done in time... had another doctors visit... and i wound up getting free tickets to the metallica concert this past weekend as well... so i needed more money for hotels and what not...

SO... i did some photography from the event... and here it is for your viewing plessure

http://www.everythingdrift.com/blog/?p=5594

i will probably start doing some more work on this comp... after i finish the video from the event...

here is a quick look at what kind of footage i got

http://www.youtube.com/watch?v=j_TPVf88xWA

enjoy guys

[xr4man]

i think he was doing it wrong. you're not supposed to hit the camera right?

[x88x]

Originally Posted by xr4man

i think he was doing it wrong. you're not supposed to hit the camera right?

or the wall...or the other car....

[licenseless]

oh... vid is done



now i can get back to work on some picture project for a guy... and my comp...

stuck 3 fans in front of my computer case while it was rendering... only got up to 45C or so...

but it took an hour to render

everythingdift.com is hosting it too... but they tweaked it for some reason... i dont like it but it still up so what ever.

http://www.everythingdrift.com/blog/?p=5690