Waterblock Design, good?

theetruscan · 03-17-2003, 03:55 AM

I'm posting this on behalf of an acquaintance, any suggestions very welcome. Thanks.

Well Ive been thinking about a WB idea for a while, and I finaly got board and designed one quick and dirty.
http://www.modders.net/uploads/Rastlin/xstorm2.bmp

Thats a quick drawing of what I think would work, the concept is simple, 2 Angled inlets enter, one at the end of each bar with densly packed fins. The incoming water from each channel meets head on in the center of the block causing massively turbulent flow directly over the core of the CPU, and removing heat from the tall pins in the center of the block. Then the turbulent water flow follows the area of lowest pressure, and exits over the channels with the less densly packed fins, leaving the block though dual outlets.

Concepts are.

Direcly opposed streams to induce turbulance in the center of the block.

Angled inlets to improve turbulent flow over inlet fins.

Dual 3/8" inlets and outlets, that are Y connected to 1/2" hose, theory is that this will increase water velocity inside the block, without restricting overall flow severly.

Medium thickness base, A thin base should transfer heat over a smaller spot, causing a higher temp in a smaller area, a thicker base should spread the heat over a larger area, but this leads to a lower temp differential between the cooling surface and the liquid flowing though the block. A medium base thickness should spread the heat well enough to make active cooling happen both in the center of the block, and in the microchannel fins.

Tall Pinfins cut from the base via a cross sawing technique, Widely spaced pin's allow for less restrictive water flow. sharp edges induce turbulance. Tall height increases surface area and thus heat transfer.

So what do you all think of my quick dirty, havnt designed anything in a VERY long time, proof of concept design?

Anyone have the tooling set up to build a prototype block?

Also it was my intention to have a 5mm thick lexan top, and use direct bolting to Socket A or P4 boards (different top for each board)

winewood · 03-17-2003, 08:13 AM

If i may ask:
The X arms... for lack of a better term. If they aren't over anything thats hot, and they are used for directing the water: do they have to be so long?
It is only opinion here, but couldn't you cut those X arms in half or more and bring the fins right up to the middle pin things and accomplish the same thing? They dont seem to be used to take heat out or in, so why the distance?
Another thing: If the middle holes are used for mounting that places the orientation on the board as a +. This makes the arms extending onto the mobo. Will they be hitting anything? I couldnt read the numbers well, maybe im off base here.
Hope this helps you.

Pritorian · 03-17-2003, 08:19 AM

It´s almost the same as Bladrunners Block.
Check it out.

http://forums.procooling.com/vbb/sho...t=made+a+block

-P-

Rastlin · 03-17-2003, 02:37 PM

This is my design.

It's just for theory at the moment.

The (Arms) are the size of a normal socket A cooler footprint is 60X60mm's so the block is actualy somewhat svelt compared to say Swiftech's blocks.

I had intended the block to have a 5mm thick lexan Top with dropdown bolts for mounting, Top A would be Socket A, and Top B would be P4.

Others have also suggested that the arms be shorter, right now this is just a conceptual WB, I dont have the ability to make prototypes for testing or anything.

I went with the longer arms and a 4mm Thick base, on the concept that a thicker base would spread heat more evenly, and thus the fins would be dissipating extra heat, at both the entrance and the exit.

I came to the conclusion that some WB's work well with thick bases, and some with thin bases, because each produce a different effect. a Thick base will spread heat more evenly, so that there is a much larger area of active themral transfer. Where as a very thin base like the one Cathar used in the WWR blocks will not spread heat very far, this leads to one "hotspot" with an intense tempature differential, If this hotspot is lined up with an efficent area of heat transfer like Cathar's inpingment over microchannel fins. then very rapid efficent thermal transfer will happen there. (the larger the tempature differential is the, faster thermal energy is transfered) How ever maze style blocks would by my theory work best with a thicker base plate, because the concept for a maze is basicly to keep the longest path of thermal transfer possible.

Anyways If I had the ability to make prototype blocks I would certianly be modifying the block with the Idea's I have seen, to see how far the design can be tweaked.

The important concept of my block is that the 2 water streams colide in the center thus creating an area of VERY high turbulance, that should enhance thermal transfer at the spot of higest thermal differential.

Look forward to seeing what else could be said about the block.

Particuarly I would love to see someone make a Prototype and see how it performs.

bigben2k · 03-17-2003, 02:51 PM

Quote:

Originally posted by Rastlin
This is my design.

It's just for theory at the moment.
The important concept of my block is that the 2 water streams colide in the center thus creating an area of VERY high turbulance, that should enhance thermal transfer at the spot of higest thermal differential.

...but if the flows collide in the center, isn't the turbulent area on a perpendicular plane from the baseplate? Is this where the pins come in?

I've struggled with this issue with my block, Radius, but in my design, I opted to put a priority to have the water strike the baseplate, instead of the fins, because the thermal simulation indicated that the baseplate was hotter. Someone pointed out that it would perform worse if the flow was reversed, similarly to what you propose.:shrug:

Rastlin · 03-17-2003, 03:24 PM

I had actualy orginaly thought up something similar to your radius block, but the problem that I couldnt seem to find a satisfactory awnser to was that the water wouldnt flow to the outlet evenly, and in the area outside the fins Iexpect you will basicly going to get stangnet pools of warm water, and poor water flow.

Also I dont expect the turbulance to form a plane at all, where the 2 streams collide I expect the water to swirl around rapidly, striking the pins and the baseplate, If you have ever seen a rapid in a river you get some idea how I think the water will "tumble around it's self" in the center of the block. The pins are there to add to the turbulance, and to increase the area of active thermal transfer, You will notice that I went with pins, because they are easy to make by cross cuting.

As far as your Radius design goes, I would look at using a cross cuting style, so that you end up with a hollow where you have your baseplate strike, and then mabey 20~30 wedge shaped fins, that fit together like the pieces of a pie.

Rastlin · 03-17-2003, 06:03 PM

Actualy I see you have made a lot more progress with the Radius block than I had thought, including making it more pratical to construct, and eliminating the deadspots in the flow. Impressive, good luck with it.

bigben2k · 03-17-2003, 06:51 PM

Thanks.

I like the pie idea, but that's been done by Gone_fishin (or was it Morphling1?). In any case, I must have the fins that cross over the core: it's my only "edge".

I do however have another idea...later.

What pump did you have in mind?

Rastlin · 03-17-2003, 07:01 PM

http://pub6.picturehost.co.uk/Rastli...er%20Block.JPG

Someone was nice enough to do a 3D render of the block. Helps everyone see what it's supposta look like, I was thinking that the inlets should be angled at 45 degree's directing the coolant to flow towards the center. they would obviously need to be machined off so that they were flat with the Lexan top though.

Rastlin · 03-17-2003, 08:50 PM

http://www.modders.net/uploads/Rastlin/Xstorm3.bmp

sorry to post so much, but I tweaked it a little more, it's now on revision 3, Anyone think this one looks better?

hydrogen18 · 03-17-2003, 08:57 PM

all i can say is...try it, otherwise, you will never know whether it is feasible and functional..

LiquidRulez · 03-17-2003, 09:33 PM

Quote:

Originally posted by hydrogen18
all i can say is...try it, otherwise, you will never know whether it is feasible and functional..

I second that.
Looks good, but will be a bitch to mount.

Rastlin · 03-17-2003, 09:40 PM

I think it should mount easily if I had the tools to make it, Figured on a Lexan top, that would be shaped to allow for drop holes, and direct mounting to the motherboard holes. with 8 mounting screws and an O ring for a gasket, I dont think that leaking should be a problem, and with the screw holes outside the gasket, I dont think that Corrison would be a problem either.

LiquidRulez · 03-17-2003, 10:02 PM

Well what are the dimensions of the channels?
Or did I miss something??

I would step down the area above the core and leave the X channels at your desired thickness....thinner base for this type design would be better IMO

theetruscan · 03-17-2003, 10:11 PM

Quote:

Originally posted by LiquidRulez
I would step down the area above the core and leave the X channels at your desired thickness....thinner base for this type design would be better IMO

It seems to me that if you were to do that you would need to ramp the X channels to avoid stagnant pockets of water around the drop-off areas. Is that right?

LiquidRulez · 03-18-2003, 02:42 AM

Quote:

Originally posted by theetruscan
It seems to me that if you were to do that you would need to ramp the X channels to avoid stagnant pockets of water around the drop-off areas. Is that right?

My bad..I though that was clear.

Rastlin · 03-18-2003, 09:44 AM

Could I get a few drawings here about where you think water flow will stop? and how you think I should avoid it?

Rastlin · 03-19-2003, 06:06 PM

Anyone out there that could machine a version of this block? I would realy like to have a prototype in existance just to see if the concept works well.

jaydee · 03-19-2003, 08:05 PM

All those micro channles in the X arms will be a waste of time. If your center inlet is made right with as optimal base thickness as possible all the heat should be disipated long before it reaches those micro channels. And what little heat that may make it there would be insignificant.

Looks like a good start though.

Rastlin · 03-20-2003, 06:30 AM

There is no central inlet, instead there are inlets at opposing ends. of the arms, the 2 with the higher fin count have the inlets, and the 2 with the lower fincounts have the outlets, and I have a few more revisions to make, Ive come up with some better ideas. Going to put a pyramid centered under the inlets, instead of angling them.

winewood · 03-20-2003, 07:31 AM

ahhhhhh I failed to see that as well. By running the water down the right X arms to the left X arms, the only thing I could think of for improving performance would be to narrow the middle path so that the water runs over the die area as much as possible without being able to take a path of less resistance around it.

In English: Take the middle dividers close in to guide flow to center.

Rastlin · 03-20-2003, 07:43 AM

I had designed the block for cross cutting, and that was actualy one of my planed revisions winewood, though I think i will keep the gap between the center and the outlets, I was going to run the inlets right into the center forest of pins, in my next drawing.

03-17-2003, 03:55 AM	#1
theetruscan Cooling Neophyte Join Date: Jan 2003 Location: OR/CA/NY Posts: 81	Waterblock Design, good? I'm posting this on behalf of an acquaintance, any suggestions very welcome. Thanks. Well Ive been thinking about a WB idea for a while, and I finaly got board and designed one quick and dirty. http://www.modders.net/uploads/Rastlin/xstorm2.bmp Thats a quick drawing of what I think would work, the concept is simple, 2 Angled inlets enter, one at the end of each bar with densly packed fins. The incoming water from each channel meets head on in the center of the block causing massively turbulent flow directly over the core of the CPU, and removing heat from the tall pins in the center of the block. Then the turbulent water flow follows the area of lowest pressure, and exits over the channels with the less densly packed fins, leaving the block though dual outlets. Concepts are. Direcly opposed streams to induce turbulance in the center of the block. Angled inlets to improve turbulent flow over inlet fins. Dual 3/8" inlets and outlets, that are Y connected to 1/2" hose, theory is that this will increase water velocity inside the block, without restricting overall flow severly. Medium thickness base, A thin base should transfer heat over a smaller spot, causing a higher temp in a smaller area, a thicker base should spread the heat over a larger area, but this leads to a lower temp differential between the cooling surface and the liquid flowing though the block. A medium base thickness should spread the heat well enough to make active cooling happen both in the center of the block, and in the microchannel fins. Tall Pinfins cut from the base via a cross sawing technique, Widely spaced pin's allow for less restrictive water flow. sharp edges induce turbulance. Tall height increases surface area and thus heat transfer. So what do you all think of my quick dirty, havnt designed anything in a VERY long time, proof of concept design? Anyone have the tooling set up to build a prototype block? Also it was my intention to have a 5mm thick lexan top, and use direct bolting to Socket A or P4 boards (different top for each board) Last edited by theetruscan; 03-17-2003 at 06:04 AM.

03-17-2003, 08:13 AM	#2
winewood Cooling Savant Join Date: Nov 2002 Location: in my chair Posts: 574	If i may ask: The X arms... for lack of a better term. If they aren't over anything thats hot, and they are used for directing the water: do they have to be so long? It is only opinion here, but couldn't you cut those X arms in half or more and bring the fins right up to the middle pin things and accomplish the same thing? They dont seem to be used to take heat out or in, so why the distance? Another thing: If the middle holes are used for mounting that places the orientation on the board as a +. This makes the arms extending onto the mobo. Will they be hitting anything? I couldnt read the numbers well, maybe im off base here. Hope this helps you. __________________ -winewood-

03-17-2003, 08:19 AM	#3
Pritorian Cooling Savant Join Date: Sep 2002 Location: Norway Posts: 201	It´s almost the same as Bladrunners Block. Check it out. http://forums.procooling.com/vbb/sho...t=made+a+block -P- __________________ Boo!

03-17-2003, 06:51 PM	#8
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Thanks. I like the pie idea, but that's been done by Gone_fishin (or was it Morphling1?). In any case, I must have the fins that cross over the core: it's my only "edge". I do however have another idea...later. What pump did you have in mind? __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

03-20-2003, 07:31 AM	#21
winewood Cooling Savant Join Date: Nov 2002 Location: in my chair Posts: 574	ahhhhhh I failed to see that as well. By running the water down the right X arms to the left X arms, the only thing I could think of for improving performance would be to narrow the middle path so that the water runs over the die area as much as possible without being able to take a path of less resistance around it. In English: Take the middle dividers close in to guide flow to center. __________________ -winewood-

03-17-2003, 02:37 PM	#4
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	This is my design. It's just for theory at the moment. The (Arms) are the size of a normal socket A cooler footprint is 60X60mm's so the block is actualy somewhat svelt compared to say Swiftech's blocks. I had intended the block to have a 5mm thick lexan Top with dropdown bolts for mounting, Top A would be Socket A, and Top B would be P4. Others have also suggested that the arms be shorter, right now this is just a conceptual WB, I dont have the ability to make prototypes for testing or anything. I went with the longer arms and a 4mm Thick base, on the concept that a thicker base would spread heat more evenly, and thus the fins would be dissipating extra heat, at both the entrance and the exit. I came to the conclusion that some WB's work well with thick bases, and some with thin bases, because each produce a different effect. a Thick base will spread heat more evenly, so that there is a much larger area of active themral transfer. Where as a very thin base like the one Cathar used in the WWR blocks will not spread heat very far, this leads to one "hotspot" with an intense tempature differential, If this hotspot is lined up with an efficent area of heat transfer like Cathar's inpingment over microchannel fins. then very rapid efficent thermal transfer will happen there. (the larger the tempature differential is the, faster thermal energy is transfered) How ever maze style blocks would by my theory work best with a thicker base plate, because the concept for a maze is basicly to keep the longest path of thermal transfer possible. Anyways If I had the ability to make prototype blocks I would certianly be modifying the block with the Idea's I have seen, to see how far the design can be tweaked. The important concept of my block is that the 2 water streams colide in the center thus creating an area of VERY high turbulance, that should enhance thermal transfer at the spot of higest thermal differential. Look forward to seeing what else could be said about the block. Particuarly I would love to see someone make a Prototype and see how it performs.

03-17-2003, 03:24 PM	#6
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	I had actualy orginaly thought up something similar to your radius block, but the problem that I couldnt seem to find a satisfactory awnser to was that the water wouldnt flow to the outlet evenly, and in the area outside the fins Iexpect you will basicly going to get stangnet pools of warm water, and poor water flow. Also I dont expect the turbulance to form a plane at all, where the 2 streams collide I expect the water to swirl around rapidly, striking the pins and the baseplate, If you have ever seen a rapid in a river you get some idea how I think the water will "tumble around it's self" in the center of the block. The pins are there to add to the turbulance, and to increase the area of active thermal transfer, You will notice that I went with pins, because they are easy to make by cross cuting. As far as your Radius design goes, I would look at using a cross cuting style, so that you end up with a hollow where you have your baseplate strike, and then mabey 20~30 wedge shaped fins, that fit together like the pieces of a pie.

03-17-2003, 06:03 PM	#7
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	Actualy I see you have made a lot more progress with the Radius block than I had thought, including making it more pratical to construct, and eliminating the deadspots in the flow. Impressive, good luck with it.

03-17-2003, 07:01 PM	#9
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	http://pub6.picturehost.co.uk/Rastli...er%20Block.JPG Someone was nice enough to do a 3D render of the block. Helps everyone see what it's supposta look like, I was thinking that the inlets should be angled at 45 degree's directing the coolant to flow towards the center. they would obviously need to be machined off so that they were flat with the Lexan top though.

03-17-2003, 08:50 PM	#10
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	http://www.modders.net/uploads/Rastlin/Xstorm3.bmp sorry to post so much, but I tweaked it a little more, it's now on revision 3, Anyone think this one looks better?

03-17-2003, 08:57 PM	#11
hydrogen18 Cooling Savant Join Date: Feb 2003 Location: USA Posts: 248	all i can say is...try it, otherwise, you will never know whether it is feasible and functional..

03-17-2003, 09:40 PM	#13
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	I think it should mount easily if I had the tools to make it, Figured on a Lexan top, that would be shaped to allow for drop holes, and direct mounting to the motherboard holes. with 8 mounting screws and an O ring for a gasket, I dont think that leaking should be a problem, and with the screw holes outside the gasket, I dont think that Corrison would be a problem either.

03-17-2003, 10:02 PM	#14
LiquidRulez Cooling Savant Join Date: Sep 2002 Location: In Hell Posts: 322	Well what are the dimensions of the channels? Or did I miss something?? I would step down the area above the core and leave the X channels at your desired thickness....thinner base for this type design would be better IMO

03-18-2003, 09:44 AM	#17
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	Could I get a few drawings here about where you think water flow will stop? and how you think I should avoid it?

03-19-2003, 06:06 PM	#18
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	Anyone out there that could machine a version of this block? I would realy like to have a prototype in existance just to see if the concept works well.

03-19-2003, 08:05 PM	#19
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	All those micro channles in the X arms will be a waste of time. If your center inlet is made right with as optimal base thickness as possible all the heat should be disipated long before it reaches those micro channels. And what little heat that may make it there would be insignificant. Looks like a good start though.

03-20-2003, 06:30 AM	#20
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	There is no central inlet, instead there are inlets at opposing ends. of the arms, the 2 with the higher fin count have the inlets, and the 2 with the lower fincounts have the outlets, and I have a few more revisions to make, Ive come up with some better ideas. Going to put a pyramid centered under the inlets, instead of angling them.

03-20-2003, 07:43 AM	#22
Rastlin Cooling Neophyte Join Date: Mar 2003 Location: Michigan Posts: 22	I had designed the block for cross cutting, and that was actualy one of my planed revisions winewood, though I think i will keep the gap between the center and the outlets, I was going to run the inlets right into the center forest of pins, in my next drawing.

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