Re-Design of the Jet Block.

[jaydee]

First attemps here: http://forums.procooling.com/vbb/sho...&threadid=7220

What I "want" to do is go with the 19 jets in this array that I already tried:


My first attemps have failed pretty miserably. Best I have got was 1C better than the Maze4.

Here is what I am considering for this revision:

Overall base thickness: 3/16" (4.762mm) Original was 1/8"
Cup depth: 1/8" (3.175mm) Original was 1/16"
Cup width: 1/8" (3.175mm) original was .011"ish

Jet ID: 1/16" (1.587mm) Original was .55"
Jet OD: 3/32" (2.381mm) Original was .090"

This time the cups in the base will have vertical walls as I will use an endmill to make the holes. This whole revision will be made on my CNC Mill this time around. So I will have to order tooling and would like to have a general idea of what to order before I order as anything worth buying under 1/8" starts to get expensive.

Any suggestions on changes to what I have above? Remember I have to use standard and not metric for tooling. Once we come up with a set plan I will order the tooling and make it...

[RabidMoose]

A naive question:

How much is the outflow from the center 7 jets constricted in comparison to the outer ring of jets? It seems that this array would result in uneven backpressure and reduced flow to the center of the core.. Would combining the input jets and some channels for outflow possibly be more efficient?

[jaydee]

Flow through all the jets is pretty even. There is plenty of room around the pipes for the water to flow into after exiting the cups in the base.

here is my first version for example:

There is alot of room around the pipes.

I think thats what you were asking?

[RabidMoose]

That's it. Thanks for the explaination.

[hara]

What were you using to make the pipes?

[MMZ_TimeLord]

Jaydee... just a thought... instead of the jets dumping directly into the cups center... how about offsetting them just enough to get the flow to circulate back up and out.

Maybe slamming the water into the cups that small is creating a stagnent area by not being able to get the tubes down in the cups enough.

Just an idea... figgered I'd voice it and be shot down rather than not contribute.

[jaydee]

Originally posted by hara
What were you using to make the pipes?

Laser.

Originally posted by MMZ_TimeLord
Jaydee... just a thought... instead of the jets dumping directly into the cups center... how about offsetting them just enough to get the flow to circulate back up and out.

Maybe slamming the water into the cups that small is creating a stagnent area by not being able to get the tubes down in the cups enough.

I forgot to mention the pipes will go into the cups about 1/64" (.396mm) to start with and they need to be as close to middle of the hole as possible to get the maximum impingment effect and to let the water evenly flow out of the hole.

[jaydee]

Finally found the post I been looking for. Couldn't remember which thread it was in. This explains the impingment effect some.

Originally posted by leejsmith
i made these simulations using real flow and it shows the advantage of the cup base over a drill shape and flat.

first flat base dimple



next drill bit shape base



and last cup base



you can see the cup shape helps the water flow out of the dimple quicker over the same period of time in this case 11 frames with the same water jet in each simulation.

Hope this helps

Lee

From this thread: http://forums.procooling.com/vbb/sho...5&pagenumber=7

[bigben2k]

Originally posted by jaydee116
Here is what I am considering for this revision:

Overall base thickness: 3/16" (4.762mm) Original was 1/8"
Cup depth: 1/8" (3.175mm) Original was 1/16"
Cup width: 1/8" (3.175mm) original was .011"ish

Jet ID: 1/16" (1.587mm) Original was .55"
Jet OD: 3/32" (2.381mm) Original was .090"

That should work.

It would be better if you could make the gap bigger, between the tube OD, and the cup wall, but it should be fine.

See, this is why I keep pushing the brass/copper tubes: you can get tubes with very thin walls, and that's really the key. If you don't get the performance you're after, I think you really ought to give it a shot.

My $0.02

[jaydee]

Originally posted by bigben2k

See, this is why I keep pushing the brass/copper tubes: you can get tubes with very thin walls, and that's really the key. If you don't get the performance you're after, I think you really ought to give it a shot.

My $0.02

I am just trying to beat the freeking Maze 4, which at this point has not happened. My 9 jet block is the best version I made so far which is wildly dissapointing. I would have thought 19 jets would be better than 9 but they proved to not be....

Did we ever find a place for small copper/brass tubes?

[hydrogen18]

try refrigeration places, capillary tube should do

[UNDERBYTE]

As I am an airflow guy my respnse may be off the wall.

I think impact cooling is the way to go. I do think the cup Idea is good although if it was airflow you would lose energy in that configuration.

If I was to design a block I would use an ejector like my pool skimmer . I would make my nozzles part of the heat transfer loop to take advantage of the fluid velocity.

I would impact the block at an angle induce a secondary flow and increase the overall velocity/decrease resistance through the block.

My pool skimmer operates with an ordinary garden hose and has terrific suction at relatively low velocities.

maybe incorporate 10-15 of them in a block.

Just a thought

Another thought- Air cooling has pretty well maxed out in the last couple of years. Water cooling has come a long way at the same time, is there really that much to be gained? what do you calculate as the max limit for your current flow? Like the larger and faster fan game for performance increases on heat sinks maybe the copper is at it limit without more flow (bigger pump)

[sevisehda]

I agree that using small diamter tubing would be an easy approach to make one of these jet blocks. In fact with a little care you could make one of these jet-blocks with a drill press and something to cut the tubing. My local Sears carries small diameter tubing, the only problem I think I'd run into making one of these would be crushes the tube when I tried to cut it.

Underbyte your absolutely right when it comes to loosing alot of energy with a jetblock but it is the most effective method.

Why don't I see more use of Silver in waterblocks? Since almost everyone seams to be using Cu for the base nowadays why not switch that over to silver. The increased conductivity would easily drop temps further.

[bigben2k]

Originally posted by jaydee116
Did we ever find a place for small copper/brass tubes?

Yeah, McMasters.

[jaydee]

Originally posted by sevisehda
I agree that using small diamter tubing would be an easy approach to make one of these jet blocks. In fact with a little care you could make one of these jet-blocks with a drill press and something to cut the tubing. My local Sears carries small diameter tubing, the only problem I think I'd run into making one of these would be crushes the tube when I tried to cut it.

Underbyte your absolutely right when it comes to loosing alot of energy with a jetblock but it is the most effective method.

Why don't I see more use of Silver in waterblocks? Since almost everyone seams to be using Cu for the base nowadays why not switch that over to silver. The increased conductivity would easily drop temps further.

Wouldn't be easy. You have to precisly drill the holes and insert the tubes and find someway to keep the tubes from shooting into the cups in the base under pressure. You still need to mill out an area for water to flow or make a spacer plate an d the holoes in the base need to match up.

Silver? Not much is gained by Silver especially for the cost and availability of it. Can't just order a 2" wide 3" long 3/16" thick peice of silver anywhere. Cathar said he would gain .5C with the White Water in Silver. Not worth it to me. it cost me about $1.50 for a 3/16" peice of copper. Would be around $20+ in Silver for maybe a .5C gain.....

[hydrogen18]

if concerned about crushing small tube when cutting it, use a copper etchant to cut it.

[leejsmith]

the smallest tube i could find is 1/32 OD with 0.004" walls.

http://www.specialshapes.com/brasstubing.asp

MMZtimelord. if the tubes are not central then you wont get the 2nd impingment effect on the cup walls.

JD. when i have time this would be my next attempt.

sorry for the mm.

base 10mm thich
tube ID 1.4mm OD 2mm
19 tubes / cups
tube height above the cup base 7mm
cups 3.5mm 9.5mm deep
spaced 4mm apart leaving 0.5mm between each cup.

[leejsmith]

Originally posted by sevisehda
I agree that using small diamter tubing would be an easy approach to make one of these jet blocks. In fact with a little care you could make one of these jet-blocks with a drill press and something to cut the tubing. My local Sears carries small diameter tubing, the only problem I think I'd run into making one of these would be crushes the tube when I tried to cut it.

Underbyte your absolutely right when it comes to loosing alot of energy with a jetblock but it is the most effective method.

Why don't I see more use of Silver in waterblocks? Since almost everyone seams to be using Cu for the base nowadays why not switch that over to silver. The increased conductivity would easily drop temps further.

do you mean like this




your going to need a cross vice or x-y table too.

this is 2C better than a maze 3 and is my best attemp to date but it needed 4 more tubes for the best jet orifice area.

[hydrogen18]

wow, that looks good...i mean it looks DAMN good....

[t00lb0x]

Originally posted by leejsmith
do you mean like this




your going to need a cross vice or x-y table too.

this is 2C better than a maze 3 and is my best attemp to date but it needed 4 more tubes for the best jet orifice area.

DAMN . What did you ues to make that?

[jaydee]

Originally posted by leejsmith

JD. when i have time this would be my next attempt.

sorry for the mm.

base 10mm thich
tube ID 1.4mm OD 2mm
19 tubes / cups
tube height above the cup base 7mm
cups 3.5mm 9.5mm deep
spaced 4mm apart leaving 0.5mm between each cup.

Damn, thats a pretty thick chunk of copper. I will have to ponder this a little more...

[leejsmith]

Originally posted by jaydee116
Damn, thats a pretty thick chunk of copper. I will have to ponder this a little more...

i only have 10mm and 3mm copper bar and cathar said the height of the tubes should be at least 5* the tube ID. Thats not possible using the 3mm with the 1.4mm ID tubes so i have to use the 10mm.

as long as the tubes sit in the cups you could have 6-8mm. The smaller the tube ID the thinner the base can be. This is anohter reason why smaller cups / tubes work along with larger surface area the base can be thinner.

Hope you can try it i wont have time for a couple of weeks.

[deepblue]

@jd

first up.. i don't get along with inch and that stuff quite so well, so i'll use metric values instead

hm roundbottom cups will surely be much more efficient than any flatbottom cups. you've shown a pretty good example with those little 3ds anims. the hemisphere will guide the water easily along the walls out of the cup whereas the flat bottom will cause massive turbulences that will hinder the water from doing much of anything (good). all in all this might slow down the exiting water quite some...

another point about the pipes. get them to dip as much into the cup as you dare. the ideal distance between cup-button and pipe-bottom should be somewhere around 1mm - so if you're cup is 4mm deep, make sure the pipe reaches about ~3mm into the cup.

the next thing is the strength (width?) of the remaining baseplate below a cup. the thinner, the better - but that might make room for mess-up since if to thin, the baseplate might be deformed when trying to mount it. so i'd advise you to stick with 1mm.
- what i'm working on so far is - if that can be machined by friends of mine - about 0.6-0.8 mm below the centermost cups and up to 1mm of those around that (with pipes that follow suit).
- this of course can only be done with a cnc, since you can't drill that exact by hand...

overall baseplate strength is ~5mm

and for pipes.. well try brass pipes.. there's some that have an ID of 0.6 mm and OD of 1.2mm - others are 0.8/1.5mm (ID/OD). those should be perfect. and to fix them you could either try making the plate for the pipes from copper aswell and drill tight holes into it, trying to pull the shrink/fit thing, or you can use your polycarb and superglue the pipes into.. should work aswell

that should do it..
oh.. cup diameter.. hm let's see... assuming you use the 0.8/1.5 brass pipes, you should make the cups ~2.5mm wide..

[leejsmith]

Originally posted by deepblue
@jd

first up.. i don't get along with inch and that stuff quite so well, so i'll use metric values instead

hm roundbottom cups will surely be much more efficient than any flatbottom cups. you've shown a pretty good example with those little 3ds anims. the hemisphere will guide the water easily along the walls out of the cup whereas the flat bottom will cause massive turbulences that will hinder the water from doing much of anything (good). all in all this might slow down the exiting water quite some...

if the cup base is round the water will just flow out with no impingement happening. This is good for flow rate but not for heat transfer. The cups need to be flat.

another point about the pipes. get them to dip as much into the cup as you dare. the ideal distance between cup-button and pipe-bottom should be somewhere around 1mm - so if you're cup is 4mm deep, make sure the pipe reaches about ~3mm into the cup.

i have 3mm copper and have tried tubes 0.8mmID/1.5mmOD. first i start with them 2.0mm inside the cups and then 1.5 and 0.5. the best results was at 0.5mm. If the tubes are too close to the cup base the jet will not form correctly. For 0.8mm ID the tubes should be at least 4mm above the cup base.

the next thing is the strength (width?) of the remaining baseplate below a cup. the thinner, the better - but that might make room for mess-up since if to thin, the baseplate might be deformed when trying to mount it. so i'd advise you to stick with 1mm.
- what i'm working on so far is - if that can be machined by friends of mine - about 0.6-0.8 mm below the centermost cups and up to 1mm of those around that (with pipes that follow suit).
- this of course can only be done with a cnc, since you can't drill that exact by hand...

I agree the amount of copper above the cpu core should be as thin as possible. Also the amount of copper between each cup should be the same.

overall baseplate strength is ~5mm

and for pipes.. well try brass pipes.. there's some that have an ID of 0.6 mm and OD of 1.2mm - others are 0.8/1.5mm (ID/OD). those should be perfect. and to fix them you could either try making the plate for the pipes from copper aswell and drill tight holes into it, trying to pull the shrink/fit thing, or you can use your polycarb and superglue the pipes into.. should work aswell

using superglue with the brass tubes in polycarb is ok as long as the superglue is ok for use in water Some superglue is not.

that should do it..
oh.. cup diameter.. hm let's see... assuming you use the 0.8/1.5 brass pipes, you should make the cups ~2.5mm wide..

the cups should be 3*the tube ID too little and the water will be sucked out. Too much and the jets wont be able to cool anough of the cup area.

[jaydee]

Originally posted by leejsmith
i only have 10mm and 3mm copper bar and cathar said the height of the tubes should be at least 5* the tube ID. Thats not possible using the 3mm with the 1.4mm ID tubes so i have to use the 10mm.

as long as the tubes sit in the cups you could have 6-8mm. The smaller the tube ID the thinner the base can be. This is anohter reason why smaller cups / tubes work along with larger surface area the base can be thinner.

Hope you can try it i wont have time for a couple of weeks.

There isn't much chance of me getting anything done on this anytime soon either unfortunatly. Still need to finish my Copper Lemon Block to...