1/32 fin space?

CheeseBall · 02-19-2003, 07:55 PM

I have been looking for an endmill to cut the kind of fins I want and the smallest w/ the longest legth of cut was 5/64 w/ 5/16 LOC ($16.00 double end). So I started looking for a better alternatives. And at the shop I found some "metal saw blades". Basically you hook one onto the mill, run it very slowly, and cut away. They are about 1/32 wide. So just curious: how small is too small when it comes to the spacing in between fins for a WB?

Also, w/ 1/32" space how big should I make the fins themselves. I can't make 'em too small because they will start bending away from the saw blade as you cut more fins...

One last thing. W/ these fins should I just make the base 1/8" thick?

Water will be coming in right on top of fins and coming out both sides...

Thanks,
Bryan

bigben2k · 02-19-2003, 08:13 PM

That's a good question.

From Cathar's work, he found that a fin to channel ratio between 0.75 and 1.5 is about ideal (given some specific conditions/parameters, i.e. water as a fluid, I think...).

I considered saw blades for my block. Morphling1 actually uses them.

"Too small" is when the spacing is so tight that the hydraulic resistance is relatively the same, regardless of the fin height, but I don't think that 1/32 is anywhere near that.

It might be a good idea to run some calcs, with different flow rates.

Neomoses · 02-20-2003, 02:33 AM

Bryan, I'll be making a WB or 2 this weekend with 1/32" channels. I'll let you know how it goes.
~the other Bryan (yes, we're the only ones who spell it right.)

8-Ball · 02-20-2003, 05:39 AM

Would 0.35mm be too small a channel width, with fins of the same dimensions.

8-ball

8-Ball · 02-20-2003, 05:42 AM

0.35mm = 0.0138" (1/72")

Channels would be around 3-5mm deep.

8-ball

murray13 · 02-20-2003, 09:03 AM

Just as an example for a block with 20 channels .35mm wide and 5mm tall flowing 2.5gpm you will have 43 feet of head. For the same setup flowing 1gpm you will have 7.9 feet of head.

8-Ball · 02-20-2003, 10:17 AM

So we can safely say that it will be a low flow design then.

8-ball

8-Ball · 02-20-2003, 10:19 AM

How do you calculate this.

I would be looking at 70 3.5mm channels, each 50mm long. Either centre inlet or just flowing from one end to the other.

8-ball

8-Ball · 02-20-2003, 10:21 AM

For those wondering why so many channels, it would produce a cooled section 49mm x 50mm, designed to cool a Hammer with a 40mm x 40mm heatspreader fitted.

8-ball

murray13 · 02-20-2003, 03:04 PM

I assume you ment .35, also the longer you make the channels the worse the backressure. And more channels won't decrease the backpressure it will only allow more gpm.

To do this right you need to choose a pump and know it's specs before you start designing the block.

CheeseBall · 02-20-2003, 09:16 PM

Ok, so if I am going with a 295GPH pump w/ 69" max head. 1/32" channels, 1/16" fins (maybe smaller) and about 7 fins. What am I looking at?

Another thing. I want the fin hight to be 3/8". Maybe w/ this hight I should make 1/16" channels?

Here is the latest drawing of the block!.

Do you think I should do something different? What are you suggestions?

Blackeagle · 02-21-2003, 01:35 AM

8-ball,

If you read the thread over at Extreme systems by Morphling1 he said he tried real narrow channels and got higher temps. But I don't recall just how narrow his were. I know he had 14 channels at one point though compared to the 10 he uses now.

I'd say Cheeseballs 1/32" should work pretty well. And all that extra wall friction if not for better cooling will just increas your head loss. Even reduced to say 30 channels that is a heck of a lot of small cuts.

Will need a heck of a pump.

8-Ball · 02-21-2003, 04:21 AM

No cutting required, and I know it would need a big pump, but then I wouldn't be aiming for high flow rates with this one. I'm trying to design a block that would work efficiently with not to high a flow rate, so other blocks in the loop don't become such a limiting factor.

I haven't finalised the design yet and I'm not exactly sure if it will work. However, due to me having finals this summer, it's gonna have to wait until then.

As a side note, I might be tempted to go for 0.45mm channels and fins, or maybe 0.45mm channels and 0.35mm fins. Will try out different designs.

8-ball

Cathar · 02-21-2003, 04:06 PM

What's basically being attempted is is something like the first stage prototype of the White Water

Yes, it works well. Getting the fin width, fin height, channel width, number of channels, base-thickness, etc, are all critical to success/tuning of the design.

Neomoses · 03-04-2003, 05:40 AM

I'm getting very good results with 1/32" channels and 1/32" fins. I have 14 fins that are 0.15" tall and 2" long on a 2"x3" WB. If you have the time, give it a shot. Here's a pic.

hara · 03-04-2003, 08:12 AM

Is that variable depth i see there?

OTMOPO3OK · 03-04-2003, 05:07 PM

Neomoses!
"I have 14 fins that are 0.15" tall and 2" long"

is that right? 0.15" tall??? isn't that too small if i get it correctly your base is like 1/4" thick at least or more....
am i wrong?

CheeseBall · 03-04-2003, 11:30 PM

Thanks for the post NeoMoses. Perfect timing too. I might start cutting the actual fins tommorrow (wed.) or mon.

But after seeing your block I have a few questions about it:
1) How does the in-flow and out-flow work w/ such a wide finned area?
2) Did you use a slitting saw? If so the "corners" of your channels mush be curved instead of square corners. More info on that please.
3) And like OTMOPO3OK asked: how thick is the base? I am planning on 1/16".

My block will have 9 fins (1/32 channels, 1/32+ fins), about 1" long, but HOPEFULLY .375" tall. Do you think I'll be able to cut that deep w/out the fins bending on me too much?

Nice purple die also.

Neomoses · 03-05-2003, 06:26 AM

Quote:

is that right? 0.15" tall??? isn't that too small if i get it correctly your base is like 1/4" thick at least or more....
am i wrong?

That is correct. I'm running a Nemesis II (still a prototype) right now, and getting temps of 31idle/36 load. The nemesis II base is 2"x3"x 0.25" thick overall, with 15 channels 0.15" deep and 0.031" wide, leaving a 0.10" baseplate thickness. What you see pictured is a prototype block made from 5/8" thick copper barstock. it's still using 1/32" channels, but they are cut to 0.25" deep, and the base will be shaved down incrementally to find the optimal BP thickness.

1) How does the in-flow and out-flow work w/ such a wide finned area?
there's just a channel cut across the top to spread the flow to the outer channels. See Here The evo uses the same base, but has a nozzle to focus the jet of water.

2) Did you use a slitting saw? If so the "corners" of your channels mush be curved instead of square corners. More info on that please.

Yes, these are cut with a slitting saw, and the edges slope upwards. As many have found, the ends of the fins are not doing much, if anything, to cool down the chip. all the interesting stuff is happening in the middle.

HOPEFULLY .375" tall. Do you think I'll be able to cut that deep w/out the fins bending on me too much?

If you want to keep the channels straight, I would recommend no more than 0.125" per pass with a 1/32" slitting saw. If you try cutting much deeper than that, the tool starts to bend. You should take it in 3 cuts, 0.125" each.

Nice purple die also.
Thanks, it's Dykem blue layout fluid. It gives very good contrast with copper and aluminum.

Hope this helps clear some things up.

CheeseBall · 03-05-2003, 11:07 PM

Thanks again for the info.

Thoughs channels cut in the top piece are a good idea. I never noticed 'em on the nemesis. I think I will use them and add a few more fins. More fins, groovey.

If I may ask. Why did you go to shorter fins? Was the performance alot better w/ shorter ones?

One last thing. What is your ambient temp? Thoughs are damn good temps, CONGRATS!

Dykem... I have used the stuff also. But when I heard the stuff called "Dykem" I didn't think that was it's name. hehe I just thought thats what my friend was calling it to be funny like "Dike 'em". Uhh yeah, my bad.

Neomoses · 03-06-2003, 02:53 AM

Why did you go to shorter fins? Was the performance alot better w/ shorter ones?
I originally used the short fins because of limitations in manufacturing and because I wanted very high fluid velocity through the block. I haven't gotten around to testing the block with the 0.25" deep channels yet, but I have noticed that going from 0.050" channels to 0.031" channels noticeably improved performance. I believe the extra surface area is the main reason.

My ambient temps usually hover around 21-23C. Thanks for the congrats. My CPU thanks me every day, too!

Tuff · 03-06-2003, 08:17 AM

I like the idea and all but......what kind of pressure are the fins going to be able to take? Over an extended period of time..will they start to fold over or break away?

Tuff

Neomoses · 03-06-2003, 12:34 PM

No pump that I've seen used for watercooling is going to bend over these fins. Maybe if your pump's pushing a couple hundred psi you should worry.

Even the 0.016" fins that I made were plenty stiff. Remember that they're only 0.150" tall, so they're not flimsy at all. If you have a feeler gage, take a look at 0.015" and 0.030". Even though the materials are different (steel vs. copper) you'll get some idea of how stiff they really are.

02-19-2003, 07:55 PM	#1
CheeseBall Cooling Savant Join Date: Jan 2003 Location: Colorado Posts: 176	1/32 fin space? I have been looking for an endmill to cut the kind of fins I want and the smallest w/ the longest legth of cut was 5/64 w/ 5/16 LOC ($16.00 double end). So I started looking for a better alternatives. And at the shop I found some "metal saw blades". Basically you hook one onto the mill, run it very slowly, and cut away. They are about 1/32 wide. So just curious: how small is too small when it comes to the spacing in between fins for a WB? Also, w/ 1/32" space how big should I make the fins themselves. I can't make 'em too small because they will start bending away from the saw blade as you cut more fins... One last thing. W/ these fins should I just make the base 1/8" thick? Water will be coming in right on top of fins and coming out both sides... Thanks, Bryan Last edited by CheeseBall; 02-19-2003 at 08:20 PM.

02-19-2003, 08:13 PM	#2
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	That's a good question. From Cathar's work, he found that a fin to channel ratio between 0.75 and 1.5 is about ideal (given some specific conditions/parameters, i.e. water as a fluid, I think...). I considered saw blades for my block. Morphling1 actually uses them. "Too small" is when the spacing is so tight that the hydraulic resistance is relatively the same, regardless of the fin height, but I don't think that 1/32 is anywhere near that. It might be a good idea to run some calcs, with different flow rates. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

02-20-2003, 09:16 PM	#11
CheeseBall Cooling Savant Join Date: Jan 2003 Location: Colorado Posts: 176	Ok, so if I am going with a 295GPH pump w/ 69" max head. 1/32" channels, 1/16" fins (maybe smaller) and about 7 fins. What am I looking at? Another thing. I want the fin hight to be 3/8". Maybe w/ this hight I should make 1/16" channels? Here is the latest drawing of the block!. Do you think I should do something different? What are you suggestions? Last edited by CheeseBall; 02-20-2003 at 10:30 PM.

03-04-2003, 08:12 AM	#16
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	Is that variable depth i see there? __________________ - Every great HD crash day is the day before back-up day. - My Past System - "Better to reign in Hell than serve in Heaven." - Milton, Paradise Lost. - FMZ

03-04-2003, 05:07 PM	#17
OTMOPO3OK Cooling Neophyte Join Date: May 2002 Location: Brooklyn, NY Posts: 45	Neomoses! "I have 14 fins that are 0.15" tall and 2" long" is that right? 0.15" tall??? isn't that too small if i get it correctly your base is like 1/4" thick at least or more.... am i wrong? __________________ Keyboard not found. Press F1 to continue...

02-20-2003, 02:33 AM	#3
Neomoses Cooling Neophyte Join Date: Nov 2002 Location: Rolla, MO Posts: 51	Bryan, I'll be making a WB or 2 this weekend with 1/32" channels. I'll let you know how it goes. ~the other Bryan (yes, we're the only ones who spell it right.)

02-20-2003, 05:39 AM	#4
8-Ball Cooling Savant Join Date: Feb 2002 Location: Oxford University, UK Posts: 452	Would 0.35mm be too small a channel width, with fins of the same dimensions. 8-ball

02-20-2003, 05:42 AM	#5
8-Ball Cooling Savant Join Date: Feb 2002 Location: Oxford University, UK Posts: 452	0.35mm = 0.0138" (1/72") Channels would be around 3-5mm deep. 8-ball

02-20-2003, 09:03 AM	#6
murray13 Cooling Savant Join Date: Sep 2002 Location: Omaha, NE USA Posts: 216	Just as an example for a block with 20 channels .35mm wide and 5mm tall flowing 2.5gpm you will have 43 feet of head. For the same setup flowing 1gpm you will have 7.9 feet of head.

02-20-2003, 10:17 AM	#7
8-Ball Cooling Savant Join Date: Feb 2002 Location: Oxford University, UK Posts: 452	So we can safely say that it will be a low flow design then. 8-ball

02-20-2003, 10:19 AM	#8
8-Ball Cooling Savant Join Date: Feb 2002 Location: Oxford University, UK Posts: 452	How do you calculate this. I would be looking at 70 3.5mm channels, each 50mm long. Either centre inlet or just flowing from one end to the other. 8-ball

02-20-2003, 10:21 AM	#9
8-Ball Cooling Savant Join Date: Feb 2002 Location: Oxford University, UK Posts: 452	For those wondering why so many channels, it would produce a cooled section 49mm x 50mm, designed to cool a Hammer with a 40mm x 40mm heatspreader fitted. 8-ball

02-20-2003, 03:04 PM	#10
murray13 Cooling Savant Join Date: Sep 2002 Location: Omaha, NE USA Posts: 216	I assume you ment .35, also the longer you make the channels the worse the backressure. And more channels won't decrease the backpressure it will only allow more gpm. To do this right you need to choose a pump and know it's specs before you start designing the block.

02-21-2003, 01:35 AM	#12
Blackeagle Thermophile Join Date: Oct 2002 Location: U.S.A = Michigan Posts: 1,243	8-ball, If you read the thread over at Extreme systems by Morphling1 he said he tried real narrow channels and got higher temps. But I don't recall just how narrow his were. I know he had 14 channels at one point though compared to the 10 he uses now. I'd say Cheeseballs 1/32" should work pretty well. And all that extra wall friction if not for better cooling will just increas your head loss. Even reduced to say 30 channels that is a heck of a lot of small cuts. Will need a heck of a pump.

02-21-2003, 04:21 AM	#13
8-Ball Cooling Savant Join Date: Feb 2002 Location: Oxford University, UK Posts: 452	No cutting required, and I know it would need a big pump, but then I wouldn't be aiming for high flow rates with this one. I'm trying to design a block that would work efficiently with not to high a flow rate, so other blocks in the loop don't become such a limiting factor. I haven't finalised the design yet and I'm not exactly sure if it will work. However, due to me having finals this summer, it's gonna have to wait until then. As a side note, I might be tempted to go for 0.45mm channels and fins, or maybe 0.45mm channels and 0.35mm fins. Will try out different designs. 8-ball

02-21-2003, 04:06 PM	#14
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	What's basically being attempted is is something like the first stage prototype of the White Water Yes, it works well. Getting the fin width, fin height, channel width, number of channels, base-thickness, etc, are all critical to success/tuning of the design.

03-04-2003, 05:40 AM	#15
Neomoses Cooling Neophyte Join Date: Nov 2002 Location: Rolla, MO Posts: 51	I'm getting very good results with 1/32" channels and 1/32" fins. I have 14 fins that are 0.15" tall and 2" long on a 2"x3" WB. If you have the time, give it a shot. Here's a pic.

03-04-2003, 11:30 PM	#18
CheeseBall Cooling Savant Join Date: Jan 2003 Location: Colorado Posts: 176	Thanks for the post NeoMoses. Perfect timing too. I might start cutting the actual fins tommorrow (wed.) or mon. But after seeing your block I have a few questions about it: 1) How does the in-flow and out-flow work w/ such a wide finned area? 2) Did you use a slitting saw? If so the "corners" of your channels mush be curved instead of square corners. More info on that please. 3) And like OTMOPO3OK asked: how thick is the base? I am planning on 1/16". My block will have 9 fins (1/32 channels, 1/32+ fins), about 1" long, but HOPEFULLY .375" tall. Do you think I'll be able to cut that deep w/out the fins bending on me too much? Nice purple die also.

03-05-2003, 11:07 PM	#20
CheeseBall Cooling Savant Join Date: Jan 2003 Location: Colorado Posts: 176	Thanks again for the info. Thoughs channels cut in the top piece are a good idea. I never noticed 'em on the nemesis. I think I will use them and add a few more fins. More fins, groovey. If I may ask. Why did you go to shorter fins? Was the performance alot better w/ shorter ones? One last thing. What is your ambient temp? Thoughs are damn good temps, CONGRATS! Dykem... I have used the stuff also. But when I heard the stuff called "Dykem" I didn't think that was it's name. hehe I just thought thats what my friend was calling it to be funny like "Dike 'em". Uhh yeah, my bad.

03-06-2003, 02:53 AM	#21
Neomoses Cooling Neophyte Join Date: Nov 2002 Location: Rolla, MO Posts: 51	Why did you go to shorter fins? Was the performance alot better w/ shorter ones? I originally used the short fins because of limitations in manufacturing and because I wanted very high fluid velocity through the block. I haven't gotten around to testing the block with the 0.25" deep channels yet, but I have noticed that going from 0.050" channels to 0.031" channels noticeably improved performance. I believe the extra surface area is the main reason. My ambient temps usually hover around 21-23C. Thanks for the congrats. My CPU thanks me every day, too!

03-06-2003, 08:17 AM	#22
Tuff Cooling Savant Join Date: Feb 2003 Location: Canada Posts: 141	I like the idea and all but......what kind of pressure are the fins going to be able to take? Over an extended period of time..will they start to fold over or break away? Tuff

03-06-2003, 12:34 PM	#23
Neomoses Cooling Neophyte Join Date: Nov 2002 Location: Rolla, MO Posts: 51	No pump that I've seen used for watercooling is going to bend over these fins. Maybe if your pump's pushing a couple hundred psi you should worry. Even the 0.016" fins that I made were plenty stiff. Remember that they're only 0.150" tall, so they're not flimsy at all. If you have a feeler gage, take a look at 0.015" and 0.030". Even though the materials are different (steel vs. copper) you'll get some idea of how stiff they really are.

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