Custom waterblock

[ying]

Hi! im not new in the watercooling world, so i decided to create my own custom waterblock for complement my mod. Actually i don't know how many money i can spend in, but first i would like to make a good design.

This is my actual design:

This is the waterblock:

http://i220.photobucket.com/albums/dd163/ying_27/blokefoto4-1.png

They are mirochannels of 0.5mm*38mm and 3mm of deep. In total there are 30 microchannels.

The material of the waterblock will be copper and acetal for the top. The top:

http://i220.photobucket.com/albums/dd163/ying_27/blokefoto3-1.png

Its simple but i think it will perform well.

This is how looks the full waterblock:

http://i220.photobucket.com/albums/dd163/ying_27/blokefoto2-1.png

The performance of my design will be as a water block from the market?

Should I need to redesign the waterblock to add more microchannels?

Whats the hole size for the inlet and the outlet for 1/4"g fittings? and another noob question, for m4 screws should i need to make holes of 4mm of diameter? or i should make them smaller for the thread? (3,5mm of diameter should be ok?)

For the design i use autocad and solidworks, but i haven't the solidworks flow simulation so will the solidworks floxpress be useful too?

Thanks for helping me =)!

[Dr_Swizz]

Interesting. I am actally currently making my own waterblock too. :-)

For M4 you should drill holes that are a bit smaller (I don't remember the eaxct hole size, perhaps 3.6mm? I can look it up). I would advice you to either use helicoil instead of threads in directly in the acetal or drill holes through the top and use nuts on top of the block (will look more ugly, but will be much stronger). Alternetively you could use some kind thread that is more coarse than mm thread.

[Dr_Swizz]

Interesting. I am actally currently making my own waterblock too. :-)

For M4 you should drill holes that are a bit smaller (I don't remember the eaxct hole size, perhaps 3.6mm? I can look it up). I would advice you to either use helicoil instead of threads in directly in the acetal or drill holes through the top and use nuts on top of the block (will look more ugly, but will be much stronger). Alternetively you could use some kind thread that is more coarse than mm thread.

[ying]

This is the new version of the waterblock. Im trying to make the waterblock more lighter.

http://i220.photobucket.com/albums/dd163/ying_27/bloke5.png

In a moments i will post the simulation with the floxpress.

The simulation with the floxpress (i know, this simulation isn't really good, but i haven't the flow simulation software and i don't know how to setup the test xd):

http://i220.photobucket.com/albums/dd163/ying_27/bloke5simulacion1.png

http://i220.photobucket.com/albums/dd163/ying_27/bloke5simulacion2.png

[ying]

Interesting. I am actally currently making my own waterblock too. :-)

For M4 you should drill holes that are a bit smaller (I don't remember the eaxct hole size, perhaps 3.6mm? I can look it up). I would advice you to either use helicoil instead of threads in directly in the acetal or drill holes through the top and use nuts on top of the block (will look more ugly, but will be much stronger). Alternetively you could use some kind thread that is more coarse than mm thread.

Hey thanks for commenting in my post ;). Can you post here your block too? it will be interesting to see other custom designs =).

[Dr_Swizz]

I do not have access to any CNC equipment so I have not made any drawing or sketches of how my block will look when it is finnished. The base plate has many tiny fins like most other modern water block. I will try to make the top similar to HK 3.0, EK Supreme HF etc.
I honestly do not know exactly how I will design all the channels in the interior of block. I was thinking of maybe making a "dummy" top of styrofoam and gradually hollow it out to see what how it will look.

[ying]

New update! Optimizing the flow of the waterblock:

http://i220.photobucket.com/albums/dd163/ying_27/bloke6foto1.png

http://i220.photobucket.com/albums/dd163/ying_27/bloke6foto2.png

http://i220.photobucket.com/albums/dd163/ying_27/bloke6foto3.png

[Dr_Swizz]

Nice.
Have experimented moving the inlet & outlet holes horizontally? If so, how what effects does it have on the waterflow? What happens if you move the inner hole towards the middle of the block?
Would it help if you added a third hole and had the inlet in the middle of the block and two outlets on the sides of the block?

[ying]

Nice.
Have experimented moving the inlet & outlet holes horizontally? If so, how what effects does it have on the waterflow? What happens if you move the inner hole towards the middle of the block?
Would it help if you added a third hole and had the inlet in the middle of the block and two outlets on the sides of the block?

Yes, the problem of having the inlet in the center is that the flow of the water isn't as equal as having it like me (in my design).

A third hole with the inlet in the center will be the best option, but actually, a third hole can be annoying when mounting the system.

The cons of having the hole in the center is that you will not able to use compression fittings of 13/19 tubes but you can make as the swiftech design, the block that the center hole can be moved but this will affect to the full flow, then after all my simulations i learned that the best you can do is to add the less restrictions possible.

I'm noob on this themes so, can someone pro check my design?

thnx =)!
PS: Dr_Swizz, come on, make a sketch of your waterblock ;), this will help you to know the measure of all the parts.

[relttem]

what are the velocities going thru the channels, and I'd try to get rid of all that excess copper

[ying]

Wow thank for the response

Actually i don't know how exactly works the floxpress of the solidworks (and the flowsimulation less).

Whats the normal presure in a watercooling system? The software told me to introduce the preassure in the outlet, what value should i put?

I'm making this questions to make the little "test" as the most realistic as possible and then i can told you the velocity.

And about the design, i'm still working in it, but first i prefer to finish the design of the flow.

Thx for the comment :D

[relttem]

you need to define an 'inlet' pressure. You can do that by picking a pump and finding out what the pressure outlet of the pump is. That'll be a start anyway. If you want to define an outlet pressure you would need to know more about the system so you can take into account all the other pressure drops. By defining the outlet pressure you are defining the amount of pressure drop thru the block that is allowable that would let your pump still work.

[ying]

With one d5 (but in my sistem i have 2 d5 in single loop):

http://i220.photobucket.com/albums/dd163/ying_27/Sinttulo12345.png

[Manicdan]

what about one channel along the center instead of 4?

im currently worried that the center of the chip is not going to receive that much flow compared to the outside. i dont expect this to change much in real life, but could be 1-2C difference by removing that hotspot

and when you build it i think you should start small and then increase the opening of the channel a little at a time and track changes. then after youve tested 5-6 sizes you can find out the best size.

(i am not an engineer at all, and i could be completely wrong)

[relttem]

your flow software should give a pressure drop too..how is that looking?

[ying]

your flow software should give a pressure drop too..how is that looking?

I'm using the floxpress and isn't the best tool for that (in my opinion xd).

I need to find some one who have solidworks flow simulation and then test the block for me.

[relttem]

well, your best bet would be FLUENT.. SolidWorks stuff is pretty good, but nothing beats FLUENT. Just based on the velocity profile above, you could change the diameter of some of the microchannels to get a better balance of the velocities and ultimately the pressure drop.

[ying]

well, your best bet would be FLUENT.. SolidWorks stuff is pretty good, but nothing beats FLUENT. Just based on the velocity profile above, you could change the diameter of some of the microchannels to get a better balance of the velocities and ultimately the pressure drop.

wow thanks for the info ;) then the fluent is more easy?

I will try this one, thx for the software name =)

[relttem]

no, FLUENT is not easy at all..it is impressive tho. The best way to teach yourself FLUENT is go thru the tutorials.