GPU/NB Water block (development thread) - Page 3

tex707 · 11-07-2003, 05:34 AM

.....and a cross-section and an assembly as well...don't know about the evenness of the flow with this design, though...

JokerF15 · 11-07-2003, 07:46 AM

With this design, wouldn't 'hot' water get stuck in the corners underneath the angled barb (not underneath but in the corner), it seems if it comes in at an angle it points straight to the other barb.

Wouldn't that cause the rest of the waterflow to be essentially slower through the block since, water will take the least restrictive path to its destination.

Maybe i'm just wrong =).

I would try to get the barbs to come in from the top in the corners, you shouldn't have much problem implementing this on a NB block.

Now for a GPU block, I'd have the barbs come in on the top side, and have a small divider through the middle of the block creating a small channel.

Than again this are just opinions.

Pic for reference of statements i posted before.

tex707 · 11-07-2003, 09:45 AM

Quote:

Originally posted by JokerF15
With this design, wouldn't 'hot' water get stuck in the corners underneath the angled barb (not underneath but in the corner), it seems if it comes in at an angle it points straight to the other barb.

Wouldn't that cause the rest of the waterflow to be essentially slower through the block since, water will take the least restrictive path to its destination.

Maybe i'm just wrong =).

I would try to get the barbs to come in from the top in the corners, you shouldn't have much problem implementing this on a NB block.

Now for a GPU block, I'd have the barbs come in on the top side, and have a small divider through the middle of the block creating a small channel.

Than again this are just opinions.

Pic for reference of statements i posted before.

I really can't tell about the flow...the angled barb is the inlet one, of course....we'll se what other more experienced people than me are going to say...

About the pic...that's the block I've seen!

JokerF15 · 11-07-2003, 12:56 PM

The bad thing about the block that I posted is, it is not done proper, it is 3/8'' first of all (not bad, but the channel area does not calculate to 3/8'' ID hose area =\.). Looks wise it is a good block, performance wise, it kills the flow, and doesn't do the greatest job.

If done proper, I think it can be an excellent design.

Thanks.

-JokerF15

Blackeagle · 11-07-2003, 02:26 PM

Tex 707,

Thank you for your efforts, your CAD work is really good. Most of all the transparant ones.

What I have in mind is differant. I'm sure you've seen blocks that have the inlet & oulet on one side of the block and these don't take up any space to speak of which won't then block use of extra card slots.

What I have in mind is that the side of the block where those barbs are mounted is cut at a sloped angle so the barbs would rise over the RAM higher allowing use of heat sinks. It might not be possible to cut the base at an angle as well as the top or at least not all the blocks side, so as to not affect the cooling pin area of the block & it's O-ring.

Your last two pics/drawings gave me a differant view point regarding this idea. I don't think it will work as well with the thick base involved here. I do think if it were made from a air pin sink that perhaps then it would as the base would be much thinner with the pins rising out of the thinner base. Will take some thick plexi for the top to make it work even there, but it might.

Thanks again for your fine CAD work.

jaydee · 11-07-2003, 03:00 PM

Sorry for my lack of response here guys. I been having to much fun in the New DD block thread. Not to often I find someone that can make an ass of themselfs better than I can! Had to play a little.

Anyway I will be pretty much gone untill Sunday afternoon. I will try and pop in a few replies though. I got to get back to work, then pick the kid up after work, help my brother move stuff tomorrow, work on my freshly rebuilt website (which I hope tex707 will sign up to, need someone with that awsome rendering capablility!), and spend some time with the kid.

Keep up the good work guys, interesting stuff!

tex707 · 11-07-2003, 03:25 PM

Quote:

Originally posted by Blackeagle
Tex 707,

Thank you for your efforts, your CAD work is really good. Most of all the transparant ones.

What I have in mind is differant. I'm sure you've seen blocks that have the inlet & oulet on one side of the block and these don't take up any space to speak of which won't then block use of extra card slots.

What I have in mind is that the side of the block where those barbs are mounted is cut at a sloped angle so the barbs would rise over the RAM higher allowing use of heat sinks. It might not be possible to cut the base at an angle as well as the top or at least not all the blocks side, so as to not affect the cooling pin area of the block & it's O-ring.

Your last two pics/drawings gave me a differant view point regarding this idea. I don't think it will work as well with the thick base involved here. I do think if it were made from a air pin sink that perhaps then it would as the base would be much thinner with the pins rising out of the thinner base. Will take some thick plexi for the top to make it work even there, but it might.

Thanks again for your fine CAD work.

It looks like I'm not following you quite well...are we talking about GPU or NB waterblocks here? NB waterblocks are, AFAIK, perfect with perpendicularly mounted barbs...GPU blocks, however need barbs to be mounted parallel or at an angle.

There is a 3D model of the waterblock I am using on my video card presented on the previous page....the hoses clear the RAM heatsinks due to a pretty complicated barb mounting design. Please take a look and tell me if you are referring to that particular problem.

Thank you for the kind words regarding my CAD work...I think that I have more to offer in that field.

Yo-DUH_87 · 11-07-2003, 05:58 PM

PostNuke, better than e107

Whatever floats your boat :shrug:

Why don't you make the gfx one like the Swiftech MCX(?) 40, with the barbs comming out the side?

jaydee · 11-07-2003, 06:46 PM

Quote:

Originally posted by Yo-DUH_87
PostNuke, better than e107

Whatever floats your boat :shrug:

Why don't you make the gfx one like the Swiftech MCX(?) 40, with the barbs comming out the side?

To many problems with e-107. Half the stuff doesn't work and my requests for help goes unanswered. Not to mention security is not so good, and he has to update it every freeking week, which means I have to. The thing is still beta bigtime. Everything works on PostNuke, and the security of the latest release is supposed to be good. Don't have to update every week and it loads faster. :shrug: Got it on both my sites now, see no reason to change. I like the forums better with postnuke.

I don't need the barbs coming out the side on mine and it would add another 3/4" to the already 1/2" Might get a little heavy being they will be all copper. other than that sure. Danger Den has a nice side mounted one aswell..

Blackeagle · 11-07-2003, 10:25 PM

Tex707,

Yes, I'm talking about a GPU design, I agree with you on the NB.

JD, I know the DD block your refering to and it's 45 degree angled barbs are better than the 90's some designs use. I also relize this isn't needed for your currant project. I posted it as being of possible interest to others and perhaps you as well (for the future).

I think that if the side of the block were cut at a 35-45 degree angle and then the barbs installed there it would be a improvement over many designs now out. Should offer a bit lower resistence while at the same time improveing the cooling by a small amount.

Not try'n to hijack the thread, so I'll not post any more on it, but this has offered me some things to consider. I may not have a mill to make blocks with. But I'm thinking I've enough tools to have a go at something.

Tex707

Thanks again.

jaydee · 11-07-2003, 10:33 PM

Quote:

Originally posted by Blackeagle

Not try'n to hijack the thread, so I'll not post any more on it, but this has offered me some things to consider. I may not have a mill to make blocks with. But I'm thinking I've enough tools to have a go at something.

Carry on! it is good idea's. Mine is already set in stone for this project, but that is just this project. Keep the idea's coming.

Also this was the DD block I was refering to. http://www.dangerden.com/mall/Blocks/maze4gpu.asp

pHaestus · 11-07-2003, 10:34 PM

What about something similar to xjinn's CPU block for a video card instead?

http://www.procooling.com/articles/a...t-team-wc3.jpg

Nothing as elaborate for sure (fewer larger channels would be my suggestion), but it would be fairly low in restriction with all the parallel channels and it would be easy to put 45 degree barbs on the ends to raise the barbs off the ramsinks.

jaydee · 11-07-2003, 10:42 PM

Quote:

Originally posted by Yo-DUH_87

Why don't you make the gfx one like the Swiftech MCX(?) 40, with the barbs comming out the side?

I am going to think about this some more though. I want to use 1/2" connector. I can solder on a 1/2" copper pipe... I will think about it.

pH, everytime that pic is brought up I get the shivers of tryng to drill that bitch out! That was some impressive work. Something else to ponder for sure...

tex707 · 11-08-2003, 01:50 AM

Quote:

Originally posted by Blackeagle

<snip>

I think that if the side of the block were cut at a 35-45 degree angle and then the barbs installed there it would be a improvement over many designs now out. Should offer a bit lower resistence while at the same time improveing the cooling by a small amount.

<snip>

OK, I'll try to model this....just one more question, is your idea to cut the Plexi top at an angle, or both the top and the copper base?

tex707 · 11-08-2003, 06:50 AM

Here is a simple stress analysis of the Plexi top…in case some of you have wondered how much load it can take, i.e. how much are we are allowed to tighten the nuts that press the springs against the lower washer and the protruding mounting parts of the Plexi top itself.

Yield strength for Plexiglass is 5.5e+7 N_m2, while the maximum stress barely exceeds 3e+6 N_m2 with the force of 50N (11.24 lbf) applied with each spring. This force is, of course, far higher in intensity than necessary to make an excellent contact between the NB and the waterblock.

The maximal displacement (at the tip of the protruding Plexi part) is a mere 0.03 mm.

All this data is provided for a 8mm thick Plexi top and is not to be taken as an exact calculation because it depends on a meshing size, which should be smaller than I have used. However, it takes too much time to make an analysis that way.

Attached are screenshots of a deformed Plexi top - extremely exaggerated, of course, for demonstration purposes only.

jaydee · 11-08-2003, 05:54 PM

I don't think it will bend like that. You have to remember the bolt going through the mobo will keep it from bending. If anything happens it will just break.

Also here is a pic of the block modified to fit my ASUS A7V8X-X.

tex707 · 11-08-2003, 06:15 PM

Quote:

Originally posted by jaydee116
I don't think it will bend like that. You have to remember the bolt going through the mobo will keep it from bending. If anything happens it will just break.

Also here is a pic of the block modified to fit my ASUS A7V8X-X.

Oh, but it will...the bolt has nothing to do with this...this is, as I've stated previously: "...extremely exaggerated, of course, for demonstration purposes only."...this analysis guarantees it will not break under 50N load per spring...since the displacement is only 0.03mm, it is obvious that the bolt is not going to come into contact with the Plexi top.

About the modification for the ASUS board...there is really a big difference between "standard" ABIT and that ASUS of yours...

jaydee · 11-08-2003, 06:26 PM

Quote:

Originally posted by tex707
Oh, but it will...the bolt has nothing to do with this...this is, as I've stated previously: "...extremely exaggerated, of course, for demonstration purposes only."...this analysis guarantees it will not break under 50N load per spring...since the displacement is only 0.03mm, it is obvious that the bolt is not going to come into contact with the Plexi top.

About the modification for the ASUS board...there is really a big difference between "standard" ABIT and that ASUS of yours...

I mean the mounting bolt going through that hole where the peice is bending. If it starts to bend it will hit the bolt on the inside of the hole and keep it from bending. You can't bend a peice of plastic with a steel shaft going through it to easy. Unless of course the holes are way wider than they need be. Non the less it doesn't matter. 3/8" plexy is plenty strong.

I am not sure why ASUS strayed from the standard, but all their newer boards seem to be like this. I got a dead ASUS A7V333 and it is the same as the ASUS A7V8X-X.

tex707 · 11-08-2003, 06:34 PM

Quote:

Originally posted by jaydee116
I mean the mounting bolt going through that hole where the peice is bending. If it starts to bend it will hit the bolt on the inside of the hole and keep it from bending. You can't bend a peice of plastic with a steel shaft going through it to easy. Unless of course the holes are way wider than they need be. Non the less it doesn't matter. 3/8" plexy is plenty strong.

I am not sure why ASUS strayed from the standard, but all their newer boards seem to be like this. I got a dead ASUS A7V333 and it is the same as the ASUS A7V8X-X.

I understand what you're saying....and you're right, absolutely....but the screenshot is meant to show the WAY it bends, not the magnitude....since the maximal displacement is 0.03mm from horizontal plane the Plexi is definitely NOT going to come into contact with the bolt.

And...please remind me not to purchase an ASUS motherboard....

tex707 · 11-08-2003, 06:48 PM

Here's what I mean about the displacement:

jaydee · 11-08-2003, 06:56 PM

Quote:

Originally posted by tex707
Here's what I mean about the displacement:

Ahhh, I see now. Nice!

jaydee · 11-08-2003, 06:58 PM

Quote:

Originally posted by tex707

And...please remind me not to purchase an ASUS motherboard....

Yes, don't. I like my ABIT KD7 MUCH MUCH better! It is just a much nicer and easier to work with board. I was very disapointed with the A7V8X-X. it is still a good board, but just not up there with Abit anymore.l

pHaestus · 11-08-2003, 07:11 PM

A7V8X was a crappy board. The Asus boards that are good are best in class (A7V-133 and A7N8X leap to mind) but they also have released some stinkers.

The Asus A7V-133 is the best motherboard I have EVER owned period.

jaydee · 11-08-2003, 07:21 PM

Quote:

Originally posted by pHaestus

The Asus A7V-133 is the best motherboard I have EVER owned period.

My first Asus board was the original A7V with the KT133 chipset and NOT the KT133A chipset, is that the one your refering to? If so I agree. Mine was a damn nice board. Still regret selling it.... I got the Abit KT7A right after I sold it thought and that was a good board to. I dropped a hard drive onthe South bridge and killed the hard drive controller.

I loved overclocking Durons on those boards.

Blackeagle · 11-08-2003, 08:10 PM

What I thinking at this point Tex707 is to cut the Lexan 1/2" top (one side) at an angle. Combined with a thin based block with pins raised up out of it I think this can work.

Pins would be approx. .2" (5mm) and the top would be .5" Lexan,
with a thinner 1/8-3/16" base.

I'll rout a square area for the pins into the Lexan top. Although I may also use a copper 1/2"=>3/8" bell reducer as the inlet to get a bit more water velocity to impact the base with, & a bit of 1/2" copper tube for the open low restriction outlet.

A question I still have.

Is the .2" pin hieght a good length? I'm open to advice on this as I picked this .2 height as it would be handy to work with along with a 1/2" top of Lexan.

11-07-2003, 05:58 PM	#58
Yo-DUH_87 Cooling Savant Join Date: Mar 2002 Location: Oregon, USA Posts: 230	PostNuke, better than e107 Whatever floats your boat :shrug: Why don't you make the gfx one like the Swiftech MCX(?) 40, with the barbs comming out the side? __________________ Signing out... Yo-DUH_87 If it works, fix it until it's broke! Then, after it's broke, add duct tape! Affordable webhosting!

11-08-2003, 08:10 PM	#75
Blackeagle Thermophile Join Date: Oct 2002 Location: U.S.A = Michigan Posts: 1,243	What I thinking at this point Tex707 is to cut the Lexan 1/2" top (one side) at an angle. Combined with a thin based block with pins raised up out of it I think this can work. Pins would be approx. .2" (5mm) and the top would be .5" Lexan, with a thinner 1/8-3/16" base. I'll rout a square area for the pins into the Lexan top. Although I may also use a copper 1/2"=>3/8" bell reducer as the inlet to get a bit more water velocity to impact the base with, & a bit of 1/2" copper tube for the open low restriction outlet. A question I still have. Is the .2" pin hieght a good length? I'm open to advice on this as I picked this .2 height as it would be handy to work with along with a 1/2" top of Lexan. Last edited by Blackeagle; 11-08-2003 at 08:40 PM.

11-07-2003, 12:56 PM	#54
JokerF15 Cooling Neophyte Join Date: Mar 2002 Posts: 9	The bad thing about the block that I posted is, it is not done proper, it is 3/8'' first of all (not bad, but the channel area does not calculate to 3/8'' ID hose area =\.). Looks wise it is a good block, performance wise, it kills the flow, and doesn't do the greatest job. If done proper, I think it can be an excellent design. Thanks. -JokerF15

11-07-2003, 02:26 PM	#55
Blackeagle Thermophile Join Date: Oct 2002 Location: U.S.A = Michigan Posts: 1,243	Tex 707, Thank you for your efforts, your CAD work is really good. Most of all the transparant ones. What I have in mind is differant. I'm sure you've seen blocks that have the inlet & oulet on one side of the block and these don't take up any space to speak of which won't then block use of extra card slots. What I have in mind is that the side of the block where those barbs are mounted is cut at a sloped angle so the barbs would rise over the RAM higher allowing use of heat sinks. It might not be possible to cut the base at an angle as well as the top or at least not all the blocks side, so as to not affect the cooling pin area of the block & it's O-ring. Your last two pics/drawings gave me a differant view point regarding this idea. I don't think it will work as well with the thick base involved here. I do think if it were made from a air pin sink that perhaps then it would as the base would be much thinner with the pins rising out of the thinner base. Will take some thick plexi for the top to make it work even there, but it might. Thanks again for your fine CAD work.

11-07-2003, 03:00 PM	#56
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	Sorry for my lack of response here guys. I been having to much fun in the New DD block thread. Not to often I find someone that can make an ass of themselfs better than I can! Had to play a little. Anyway I will be pretty much gone untill Sunday afternoon. I will try and pop in a few replies though. I got to get back to work, then pick the kid up after work, help my brother move stuff tomorrow, work on my freshly rebuilt website (which I hope tex707 will sign up to, need someone with that awsome rendering capablility!), and spend some time with the kid. Keep up the good work guys, interesting stuff!

11-07-2003, 10:25 PM	#60
Blackeagle Thermophile Join Date: Oct 2002 Location: U.S.A = Michigan Posts: 1,243	Tex707, Yes, I'm talking about a GPU design, I agree with you on the NB. JD, I know the DD block your refering to and it's 45 degree angled barbs are better than the 90's some designs use. I also relize this isn't needed for your currant project. I posted it as being of possible interest to others and perhaps you as well (for the future). I think that if the side of the block were cut at a 35-45 degree angle and then the barbs installed there it would be a improvement over many designs now out. Should offer a bit lower resistence while at the same time improveing the cooling by a small amount. Not try'n to hijack the thread, so I'll not post any more on it, but this has offered me some things to consider. I may not have a mill to make blocks with. But I'm thinking I've enough tools to have a go at something. Tex707 Thanks again.

11-07-2003, 10:34 PM	#62
pHaestus Big Player Making Big Money Join Date: Aug 2001 Location: irc.lostgeek.com #procooling.com Posts: 4,782	What about something similar to xjinn's CPU block for a video card instead? http://www.procooling.com/articles/a...t-team-wc3.jpg Nothing as elaborate for sure (fewer larger channels would be my suggestion), but it would be fairly low in restriction with all the parallel channels and it would be easy to put 45 degree barbs on the ends to raise the barbs off the ramsinks.

11-08-2003, 07:11 PM	#73
pHaestus Big Player Making Big Money Join Date: Aug 2001 Location: irc.lostgeek.com #procooling.com Posts: 4,782	A7V8X was a crappy board. The Asus boards that are good are best in class (A7V-133 and A7N8X leap to mind) but they also have released some stinkers. The Asus A7V-133 is the best motherboard I have EVER owned period.

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