... Same old conversation... Al blocks again.

maxSaleen · 05-25-2005, 08:41 PM

Guys: just give this thread a chance and let me know what you think.

Every element is capable of absorbing/dissipating a certain amount of heat. Copper is clearly better at absorbing and dissipating heat. If all else were to be equal, Copper would clearly be a superior choice for heatsinks.

Aluminum has several advantages over copper, however. Aluminum is much stronger meaning that it can be machined at higher speeds, have tighter tolerances, and have thinner walls. Not to mention it is cheap. If aluminum is used throughout the system, lower costs can be achieved.

This is what got me thinking....

There was recently a thread on the forums about someone who had purchased a heater core that they thought was made of copper. It turns out it was made of brass. The individual was concerned about the effect such an inferior conducter would have on performance. Several members (most of which were "pros") commented that the walls were so thin that the effect would be minimal.

Interesting....

Everybody has sworn high and low that copper is the best material for radiators (for our uses) yet most automotive rads are made of aluminum for cost and durability reasons. I think Marci could probably give us an idea of how much easier it would be to work with Al instead of Cu when such thin walls are used. Think of Koolance systems. Does anyone seriously believe that their radiators are the limiting factor? Heck no. In fact they are fairly nice single pass rads on their own (they used to sell one with 3/8"id compression fittings for about $32, I really wish I had bought one).

Getting to my point....

Maybe we are looking at Al blocks the wrong way. I, personally, don't feel that anyone has ever properly capitalized on the strenghts of Al. Think about it this way: Al has a specific heat that is roughly 2.5 times that of Cu. If one were to make a block out of Cu, and then, with a similar design, create one out of Al with 2.5 times the surface area. Which do you think would perform better? I think the two would be very close assuming flow rate remains constant.

An interesting experiment...

I don't have the time to machine my own blocks. I wish I did but I don't. Hell I hardly have time for the article I'm writing for VH (month overdue). All I have time for is spitballing ideas at you guys. For those who can make blocks I suggest this: Make three jet impignations blocks. The first two being of identical design (very thin base for our puproses ~1mm) one being Al the other being Cu. Then make a third block with the base being as thin as possible (less than 1mm if possible, good luck).Test the three blocks. I would bet that the first and the third would be within 1C of each other.

And what would this prove?

That we should all move to Al jet impignation blocks because some wacko on the proforums with a strange name derived from a car said so? ..... Inhale .... No. Of course not. I'm just suggesting that it might be worth while to investigate the strengths of aluminum and how they can be used in water block design such as thinner walls, more numerous and thinner pins, and possibly different flow patterns (I'm still pushing for 45 degree inlets and outlets). Let's hear what the rest of you have to say.

bigben2k · 05-25-2005, 08:51 PM

...and in that same thread, we also pointed out that if it was possible to use glass to make a heatercore, it probably wouldn't be far off either.

But the processor does something that's unique: it pumps a lot of heat within a very, very small area. So... the different properties of the materials used to make a water block are essentially magnified.

Quote:

Aluminum has several advantages over copper, however. Aluminum is much stronger meaning that it can be machined at higher speeds, have tighter tolerances, and have thinner walls. Not to mention it is cheap. If aluminum is used throughout the system, lower costs can be achieved.

Al is more practical to use to make anything, over copper. I would not call this property "strength". I do believe that a brass core would be more puncture resistant than an Alu core.

maxSaleen · 05-25-2005, 08:56 PM

The fact that something is easier to manufacture is always a strength. The fact that thinner walls and pins can be used with Al is derived from this. Ease of production translates into lower costs. Lower costs translates into a larger market. That would be a win in my book.

bigben2k · 05-25-2005, 09:23 PM

It would indeed, however...

Al is not cutting it anymore. Here's what's being done today, as "best alternative":
(From: http://www.tomshardware.com/cpu/2005...tium_d-06.html )

I recently acquired a couple of these coolers, and the copper base actually extends upwards into a cup shape.

maxSaleen · 05-25-2005, 09:36 PM

Strait Cu isn't cutting it anymore in the enthusiast market. Heatpipes are "must have" on high end air. Evolution is everywhere. I feel that I high surface area Al block with a good flow pattern could easily trounce even the best air cooled heatsinks. If designed properly it could do so at a lower price point.

maxSaleen · 05-26-2005, 06:08 AM

BigBen: what do you think of my jet impignation experiment?

Cathar · 05-26-2005, 08:09 AM

Quote:

Originally Posted by maxSaleen

Strait Cu isn't cutting it anymore in the enthusiast market. Heatpipes are "must have" on high end air. Evolution is everywhere.

There do appear to be a number of heatsinks, all Cu, which give most heat-pipe based heatsinks an extremely good run for their money. eg. Thermalright SLK-900, which happily outperforms the large majority of heat-pipe based heatsinks.

Heatpipes do seem to be primarily useful as a method of weight reduction, rather than superior conductive performance to copper.

As in, copper would most likely still be superior, but the issue is the amount of copper needed to adequately conduct the heat to the large finned surface areas, for which heatpipes can do an almost as effective job, but for MUCH less weight than an all-copper cooler could do.

I'm still of the firm belief that it would be very possible to construct an all-copper non-heatpipe cooler than will outperform the best heatpipe cooler on the market, and within a similar size, but the drawback would be the weight of it.

Marci · 05-26-2005, 09:07 AM

Not forgetting heatpipes have that added "bling" factor to the particular portion of the market that requires bling. I've seen ppl buy sinks purely cos they have a heatpipe, irrelevant of reviews or performance. The presence of a shiny pipe-like affair sticking out of a block and going back in at the top appears to extend willies... anyways...

Quote:

Everybody has sworn high and low that copper is the best material for radiators (for our uses) yet most automotive rads are made of aluminum for cost and durability reasons. I think Marci could probably give us an idea of how much easier it would be to work with Al instead of Cu when such thin walls are used.

More or less all heatercores etc shipped from the ThermoChill factory, whether for automotive or PC, are based on brass cores with copper fins afaik these days. As to the machining benefits of aluminium - not my field... that's the metalwork R&D guys who handle the machinery's department. I'll have words with Andy and see what he says on the issue.

Risky · 05-26-2005, 09:16 AM

Of course if you had a 'desktop' style case with the mobo flat, the weight consideration more or less irrelevant.

bigben2k · 05-26-2005, 11:02 AM

Quote:

Originally Posted by maxSaleen

BigBen: what do you think of my jet impignation experiment?

Where? :shrug:

Ls7corvete · 05-26-2005, 05:15 PM

HHmmm brings up a good point for water blocks anyways. Didnt a few of the storm blocks have bases under 1mm thickness? The differences between copper and alum should be shrinking as the base thickness is shrinking as well.

Combine that with the ability to machine alum faster and smaller channels I think it could be an option.

Everything is a trade off in the end though and with limited production you gotta choose one metal so I think that what was choosen is obvious...

Cathar · 05-26-2005, 06:10 PM

It was the Cascade blocks that had the <1mm base thicknesses.

Storm block's base thickness is quite a deal thicker.

maxSaleen · 05-26-2005, 06:51 PM

BigBen: I meant the proposed experiment.

Everyone: thanks for your input so far.

I'm not suggesting that Al could ever outperform Cu in a block. My train of thought is that when properly designed, we might see a Al block performing along the lines of a maze4. I'd think that such a block would have a substantial advantage over any air cooled setup in terms of noise (depending on rad/fans) and performance. The goal being to have a system made entirely out of Al in order to lower costs a broaden the community.

bigben2k · 05-26-2005, 08:16 PM

Quote:

Originally Posted by maxSaleen

...My train of thought is that when properly designed, we might see a Al block performing along the lines of a maze4. ...

Good luck with that.

maxSaleen · 05-26-2005, 08:38 PM

I detect sarcasm/pesimissim. Yes?

FL3JM · 05-26-2005, 09:20 PM

Quote:

Originally Posted by maxSaleen

I detect sarcasm/pesimissim. Yes?

Well, making an Alu block that will outperform a maze4 will require a substantial amount of luck mate.

jaydee · 05-26-2005, 10:03 PM

Quote:

Originally Posted by FL3JM

Well, making an Alu block that will outperform a maze4 will require a substantial amount of luck mate.

Not really luck. Just some research. I got a couple designs that beat my Maze 4 at all flow rates. The Maze 4 is really not that great of a block.

maxSaleen · 05-26-2005, 10:34 PM

Quote:

I got a couple designs that beat my Maze 4 at all flow rates. The Maze 4 is really not that great of a block.

My point exactly. The Maze 4 would still beat out any air cooled heatsink provided a decent pump and rad are used. I'm thinking that one could throw together a decent system using an Al block, a BIP Al rad, and perhaps a CSP Mag pump for $90-100. If done in mass production, the prices could be driven lower. How much... I don't know. Any ideas on exploiting the strenghts of Al JD? I was thinking about a design with very thin pins and a very thin base. I also would push for 45 degree inlet and outlet.

I remember an article over on Frostytech where they used an Alpha Pal heatsink with a TEC to cool a video card. By mounting the fan at a 45 degree angle, they achieved something like 6-8C lower temps. They attributed this to the way the air exhausted from the heatsink. I'll look for it and post a link tomorrow.

maxSaleen · 05-27-2005, 08:15 PM

Found the link:

http://www.frostytech.com/articlevie...id=1106&page=4

I was wrong. They don't actually say how much the 45degree fan orientation helps. I still think its a good idea though.

JFettig · 05-27-2005, 08:18 PM

Quote:

Originally Posted by FL3JM

Well, making an Alu block that will outperform a maze4 will require a substantial amount of luck mate.

This I have done. its not that difficult. a storm or white water in aluminum would likely do it. I made one a while back that has not evolved into a better design made in CU now.

maxSaleen · 05-29-2005, 06:23 AM

JFettig: so you are confident that with a very thin base and a center inlet would perform within a few degrees C of an identical Cu counterpart?

05-25-2005, 08:41 PM	#1
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	... Same old conversation... Al blocks again. Guys: just give this thread a chance and let me know what you think. Every element is capable of absorbing/dissipating a certain amount of heat. Copper is clearly better at absorbing and dissipating heat. If all else were to be equal, Copper would clearly be a superior choice for heatsinks. Aluminum has several advantages over copper, however. Aluminum is much stronger meaning that it can be machined at higher speeds, have tighter tolerances, and have thinner walls. Not to mention it is cheap. If aluminum is used throughout the system, lower costs can be achieved. This is what got me thinking.... There was recently a thread on the forums about someone who had purchased a heater core that they thought was made of copper. It turns out it was made of brass. The individual was concerned about the effect such an inferior conducter would have on performance. Several members (most of which were "pros") commented that the walls were so thin that the effect would be minimal. Interesting.... Everybody has sworn high and low that copper is the best material for radiators (for our uses) yet most automotive rads are made of aluminum for cost and durability reasons. I think Marci could probably give us an idea of how much easier it would be to work with Al instead of Cu when such thin walls are used. Think of Koolance systems. Does anyone seriously believe that their radiators are the limiting factor? Heck no. In fact they are fairly nice single pass rads on their own (they used to sell one with 3/8"id compression fittings for about $32, I really wish I had bought one). Getting to my point.... Maybe we are looking at Al blocks the wrong way. I, personally, don't feel that anyone has ever properly capitalized on the strenghts of Al. Think about it this way: Al has a specific heat that is roughly 2.5 times that of Cu. If one were to make a block out of Cu, and then, with a similar design, create one out of Al with 2.5 times the surface area. Which do you think would perform better? I think the two would be very close assuming flow rate remains constant. An interesting experiment... I don't have the time to machine my own blocks. I wish I did but I don't. Hell I hardly have time for the article I'm writing for VH (month overdue). All I have time for is spitballing ideas at you guys. For those who can make blocks I suggest this: Make three jet impignations blocks. The first two being of identical design (very thin base for our puproses ~1mm) one being Al the other being Cu. Then make a third block with the base being as thin as possible (less than 1mm if possible, good luck).Test the three blocks. I would bet that the first and the third would be within 1C of each other. And what would this prove? That we should all move to Al jet impignation blocks because some wacko on the proforums with a strange name derived from a car said so? ..... Inhale .... No. Of course not. I'm just suggesting that it might be worth while to investigate the strengths of aluminum and how they can be used in water block design such as thinner walls, more numerous and thinner pins, and possibly different flow patterns (I'm still pushing for 45 degree inlets and outlets). Let's hear what the rest of you have to say. Last edited by maxSaleen; 05-25-2005 at 08:48 PM.

05-25-2005, 08:56 PM	#3
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	The fact that something is easier to manufacture is always a strength. The fact that thinner walls and pins can be used with Al is derived from this. Ease of production translates into lower costs. Lower costs translates into a larger market. That would be a win in my book.

05-25-2005, 09:36 PM	#5
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	Strait Cu isn't cutting it anymore in the enthusiast market. Heatpipes are "must have" on high end air. Evolution is everywhere. I feel that I high surface area Al block with a good flow pattern could easily trounce even the best air cooled heatsinks. If designed properly it could do so at a lower price point.

05-26-2005, 06:08 AM	#6
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	BigBen: what do you think of my jet impignation experiment?

05-26-2005, 09:16 AM	#9
Risky Cooling Neophyte Join Date: Aug 2002 Location: London Posts: 96	Of course if you had a 'desktop' style case with the mobo flat, the weight consideration more or less irrelevant.

05-26-2005, 05:15 PM	#11
Ls7corvete Cooling Savant Join Date: Oct 2004 Location: FL Posts: 164	HHmmm brings up a good point for water blocks anyways. Didnt a few of the storm blocks have bases under 1mm thickness? The differences between copper and alum should be shrinking as the base thickness is shrinking as well. Combine that with the ability to machine alum faster and smaller channels I think it could be an option. Everything is a trade off in the end though and with limited production you gotta choose one metal so I think that what was choosen is obvious...

05-26-2005, 06:10 PM	#12
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	It was the Cascade blocks that had the <1mm base thicknesses. Storm block's base thickness is quite a deal thicker.

05-26-2005, 06:51 PM	#13
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	BigBen: I meant the proposed experiment. Everyone: thanks for your input so far. I'm not suggesting that Al could ever outperform Cu in a block. My train of thought is that when properly designed, we might see a Al block performing along the lines of a maze4. I'd think that such a block would have a substantial advantage over any air cooled setup in terms of noise (depending on rad/fans) and performance. The goal being to have a system made entirely out of Al in order to lower costs a broaden the community.

05-26-2005, 08:38 PM	#15
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	I detect sarcasm/pesimissim. Yes?

05-27-2005, 08:15 PM	#19
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	Found the link: http://www.frostytech.com/articlevie...id=1106&page=4 I was wrong. They don't actually say how much the 45degree fan orientation helps. I still think its a good idea though.

05-29-2005, 06:23 AM	#21
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	JFettig: so you are confident that with a very thin base and a center inlet would perform within a few degrees C of an identical Cu counterpart?

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)