copper vs aluminum
 

hopefully a simple question. How much do you really lose by using aluminum for your waterblock construction over copper?

the only thing I can tell you is that the thermal conductivity of copper is close to twice the thermal conductivity of aluminium.

Copper is better. The results will be different depending on the block, but all in all copper is better. I still use AL blocks like the one I made in my avatar and it works great. Running a XP1600+@1740mhz 2.0Vcore just fine with it. Temps are low. Aluminum is definatly usable, but copper will shave a few degrees off most any block in any design. I like aluminum because it is easy to work with and works just fine for what I need it to do.

maybe i should have been a bit more specific in my question. I know that copper does transfer heat better, i was more curious when dealing with these types of water blocks does the cost justify the couple degree difference that it will make in the end? plus as previously stated aluminum is much easier to machine than copper is.

It depends on what your overall goal is. If you want to coolest running block or someting quiter than air and still cools decent. I opt for aluminum as I don't need the extra 2-5C as the overclocks I get I am plenty happy with and do not require the extra 25-50mhz I might get with copper. As that sure the hell isn't going to show me any noticable diference in anything I run. I got 340mhz overclock with an aluminum block and temps under 42C load. I am good with that.

Now if you cannot live without that extra mhz then use copper. The extra cost for Copper isn't all that bad really. Usually about tripple. But still thats less than $10 for enough to do a water block. I would use Copper more if it was easier to mill. As of now I have no justification to do so.

it all depends on what you want to acheive.

well, could you tell us what you plan to use your block for?

Another thing you may want to consider is if you have CU anywhere else in your system, the ALU will corrode.

Nick C, fair enough question. I've just recently started looking into WC as a solution. I have available to me plenty of machining equipment that making it is not a problem. The design part is not either (i will probably ask some stupid questions though in the near future) i will do that in autocad 2k2 then post the gcode out and machine it myself. I guess the main reason for me asking was because i readily have aluminum to do the project at nearly no cost, however copper is a different story. I guess i'm being a bit cheap on the whole thing and was more curious if i am going to toast my cpu or gpu if i design a waterblock for them and run from aluminum rather than copper. if you have any more questions about what i'm planning please ask and i hope this helps clear up my question.

You can use Aluminium or Copper, but we would highly recommend Copper if you're going to have a large heatload (from overclocking/overvolting).

Theoretically, you can use either metal, but the design should be slightly different (i.e. baseplate thickness, fin width,...) in order to compensate for the different heat density/thermal resistance, and if you look at the figures, you'll find that Alu will give you a lower heat dissipation than copper, but you should be able to compensate for that, a bit, with a big pump.

The main concern (for me) is avoiding galvanic corrosion, which comes from using different metals (block and rad). Although I will have some brass parts in my rig (from the heatercore), it is a very small source of corrosion, compared to a copper/brass and aluminium combo. The heatercore will eventually be removed, as I switch to another cooling solution.

bigben, thanks for the info, now a couple other questions. First i'm guessing that the wall thickness of aluminum for the fins should be thinner correct? and secondly what should the base thickness be also. I looked around and all i saw was thicknesses for copper. if there is already a thread covering this for aluminum please point me there because i probably overlooked it by accident. thanks.

This is all going to depend on the rest of the equipment used. Like the pump ect.. I have had good luck with 1/8" and 3/16" for ther base with Aluminum with "MY" designs. Other designs will vary and I put ZERO scientific calculations into my blocks as I have no reason to waste the time to. Like I said I got 340mhz overclock at 42C (note temps are from onboard sensor on the 8K7A+ taken from MBM, but who cares how acurate they are with 340mhz OC!!) load at 2.0Vcore stable as a rock on a AL block in avatar and a 170gph pump. I do not require the wasted time in calculations! :D

But for the more educated like Bigben, Cathar, unregistered, myv65, and a bunch of others around here that are capable of figuring the science out and are willing to then more power to them. I hope they can help you your design. I have always been a trial and error person and it has served me very well so far.

If you have aluminum at your disposal then make a block or 2 with it. The corrosion is a mute point. It takes forever for any damage to occure. The worst part is the slime that builds up. I been using methonal in the form of windsheild washer fluid religously for a while now and it works well to keep it down. I use a Copper heater core on one and a brass heatercore on the other and the AL blocks. Havn't any big issues yet but I do change the water every other week because I am always changing things around.

Good Luck.

yet another question. OK this drawing took a total of like 45 seconds to do so be kind. is this a descent layout for a waterblock or should i look at copying or imitating cathar's design in a waterblock?

thanks

In my limited reasearch big long channels are not the answer for performance cooling.
My Lemon Block has outperformed my E Designs and anything I have made with channels.
http://www.customcooledpc.com/lemonsig.jpg

Cathar's has the same basic idea of getting rid of the channels with a strait through design in an attempt to cool the area where it needs to be cooled which is right over the core of the CPU. Channles just can't get that done effectivly. They take heat from the whole block and remove it as apposed to removing the heat from the source. At least that is how I think he designed it by looking at the block. Adding the Jet concept was and is an exellent idea.

First thanks all for the help so far, you guys here are geat. jaydee116 i'm a bit curious if i could either see a bit closer picture if you dont mind someone making a pseudo copy of that block, or if you wouldn't mind even more if you did a cad drawing of it, if i could get that i would be extremely greatful. anyway all the help i can get would be great, reinventing the wheel is a waste of time imho.

here is the CAD drawing for anyone who is interested. I will see if I can attach it for download.

The yellow is basically what the channles will look like after milling. The white lines are the tool paths. Disregard the green circles as I don't think they are in the right spot for mounting. They might be but I can't remember. :D

Also here are a few close up pics. The outer chanhnel is 3/8" wide and deep. The pins are only 3/16" tall and go all the way to the top of the block and make contact with the top peice and they are sitting on a 3/16" base. The material is 1/2" thick by 2" wide by 3" long.

http://www.customcooledpc.com/aluminumWB5/002.JPG
http://www.customcooledpc.com/aluminumWB5/004.JPG
I been wanting to make one out of Copper but havn't got around to it. I think the pins would be more usefull as copper should be able to transfer the heat more effectivly than AL. but it does work good for mw the way it is.

Forgot the attachment. :D

cool thanks a million for both the pics and the cad drawing. I might modify the design a bit depending on what tooling i have (or dont have for that matter) I might play around with the design a bit too. nothing scientific or anything just goofing around to see what i can come up with. one more question for ya, have you considered making arcing channels in there kinda like cothars design but in a leamon shape? that might give you more surface ares on the cooling fins and it might actually cause some aggitation in the coolent too.

Yeah I have, just havn't had time to move forward with anything new lately. Do as you wish to the design. If you can get anyone to help with actual numbers all the better. I just do not feel like spending countless ours learning the science of flow and fliud dynamics to make a water block a few C cooler than it is now. :D

Here is another drawing I came up with even before cathars block! :eek: Just never had the tooling to do it.

here is what i was contemplating by channels, hacked up your drawing really quick in autocad so the numbers aren't correct, if anyone want's me to redo it correctly let me know :P anyway here is a jpg of what i was thinking. the one main problem i see with doing it this way is that to get flow correctly the openings on the end fins will need to be larger than the inner ones to get flow correct, but i could be way off with even the thinking of this design.

Actually, I would like to point out a small correction to the ALU CU argument, I don't know if it will change much though. Copper absorbs heat better, and Alu will dissipate better. So Copper on the bottom, with Aluminum to draw out the heat should be best. A block of either in whole would agreeably have entirely different concepts of design. If I am wrong, I appologize.

Take in consideration the newer swiftech air block that uses a copper bottom w/alu pins to release the heat.

Some facts;
 

Aluminum Properties

Discovered 1827, Woehler
Atomic Weight 26.981539
Atomic Number 13
Single Isotope, 27
Crystal Structure: face centered cubic
Density 2.702 g/cm**3
Resistivity 2.650e-8 ohm-meter
Thermal conductivity 237 W/m/K
Molar volume 10.061 cm**3/mole
Coefficient of Linear Thermal Expansion 23.2e-6 /K
Molar heat capacity 24.35 J/mole/K
Melting point 933.52K
Boiling point 2740K

Copper Properties

Naturally occurring element
Atomic Weight 63.546
Atomic Number 29
Two isotopes:
63 (69.17%)
65 (30.83%)
Crystal Structure: face centered cubic
Density 8.92 g/cm**3
Resistivity 1.673e-8 ohm-meter
Thermal conductivity 401 W/m/K
Molar volume 6.712 cm**3/mole
Coefficient of Linear Thermal Expansion 16.6e-6 /K
Molar heat capacity 24.435 J/mole/K
Melting point 1336.6 K
Boiling point 2840 K

its a myth, go read billA's new articles at o/c.com

I wish someone would put a bullet in the head of this myth once and for all....

actually Dave Smith's article on amdmb addresses that issue much more directly

yea, find that myth guy and shoot him

And to add to that, if you add to different meterials together that are not molecularly bonded you have another joint. That joint will most likely remove any potential gain by using the 2 materials together as you will loose heat transferability just like the TIM joint on CPU to HSF or CPU to water blocks.

Correct me if I am wrong.

And to for a comment to the redesigned Lemon Block I am not sure how well that would work. I seen a block on the market already very similar to that but in copper and only one inlet and outlet.

for some marketing fluff related to the metal's joint
http://www.theinquirer.net/?article=6907