Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Custom 62mm TEC Water block
New Posts  All Forums:Forum Nav:

Custom 62mm TEC Water block - Page 24  

post #231 of 680
Quote:
Originally Posted by Scarlet Infidel View Post
This inspired me to find more pictures of waterblock internals. Ran across this thread which is just amazing.
LINK
We have that here to..Hall of Blocks: CPU *Updated Often* ..The V6 Stinger with the Buck knife under it is one of mine.

This is another good thread to look through if you haven't already seen it...BlockHeads: The Thread
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
E8600 4545Mhz*9 w/1.29V DFI DK P45 XFX 8800 GS D9GKX & D9GMH 
Hard DriveOptical DriveOSPower
2x Seagate 160 Perp. RD-1 2x Asus Vista Premium Corsair HX 620 
Case
Open Bench 
  hide details  
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
E8600 4545Mhz*9 w/1.29V DFI DK P45 XFX 8800 GS D9GKX & D9GMH 
Hard DriveOptical DriveOSPower
2x Seagate 160 Perp. RD-1 2x Asus Vista Premium Corsair HX 620 
Case
Open Bench 
  hide details  
post #232 of 680
Thread Starter 
Quote:
Originally Posted by ChielScape View Post
results are looking very interesting, awesome

EDIT:
i just remembered the Stinger internals... may be of interest to you.
Yeah i'll try it with cut outs on the bottom .. im getting there...

hay thank for the help with the cut outs ' i wouldn't have used them if it wasn't for your help
post #233 of 680
Thread Starter 
Quote:
Originally Posted by alk View Post
Very interesting. You see how the more turbulent areas correspond to a large increase in the heat flux in the design with the rippled channels? Why not make all the channels have these cut outs to increase the heat flux further, afterall, the ultimate purpose of the waterblock is to remove as much heat as possible from whatever you are cooling!

Just to let some readers know the theory behind this, it is all to do with what is known as the heat transfer coefficient.

Now, the heat transfer coefficient basically tells you how much heat is transferred per degree difference in temperature and available surface area, and has the units Watts/m^2 Kelvin.

Turbulence corresponds to an increase in the Reynolds number, and convection data for a forced convection situation can always be expressed as Nu=k*(Re^n)*(Pr^m) where Nu is the Nusselt number, Re the reynolds number, Pr the Prandtl number, and k n and m are constants.

With Laminar flow over the fins, a near static surface boundary layer is formed which results from lower Reynolds numbers. This lowers the Nusselt number, which is directly proportional to the heat transfer coefficient.

So all in all, turbulence, whilst it can reduce flow rate by increasing restrictiveness, if you can find a good balance between the turbulence and flow rate you will be able to get the best optimised block.

These cut outs appear to do 2 useful things for the performance of the block:

1) Increase channel surface area.

2) Increase fluid turbulence, and thus increasing heat flux.

But the compromise here is with reduced fluid flow rate.

Ultra, I suggest you begin experimenting with these cutouts, and systematically recording the number of cutouts, and the heat flux through the block. So you do something like set up an Excel spreadsheet, and enter in the number of cutouts in the design, and the overall heat flux (Or maybe just the max heat flux and you can assign a correction factor to it based on what percentage of the channels the high heat flux covers). Now you should plot the number of cutouts against the heat flux.

Now as the turbulence inscreases in the block and the mass flow begins to fall, you should notice a parabolic relationship which rises as turbulence increases but will fall once the the mass flow becomes too low. This will give you a maximum value which will indicate the ideal number of cutouts!

You dont even have to do that many simulations! Just add in say 6 cut outs every time then records your results, then add in a parabolic trendline in excel which will give you a good idea of how many of these cut outs you need!

I've used a similar technique when using FEA which helped save a huge ammount of time with design. I hope that this technique could help you!

alk

i intend too .. i haven't in the past because it takes long to try new ideas, if i have to apply it to the whole block.
post #234 of 680
Thread Starter 
Quote:
Originally Posted by omaryunus View Post
thanks ultra cant believe i missed that... looking grat though cant wait to see you make one of these hopefully by the time i get back to the states youll have a good setup up and going and ill be able to get a new block from you then i can have a stinger adn an ultra block =D
yeah the block is going really well, the one i made at the beginning worked awesomely. It worked better than the theoretically best case.. And this new block may be 30% better than that one.
Edited by Ultrasonic2 (muffy) - 7/30/09 at 1:24pm
post #235 of 680
Thread Starter 
This is one you can compare with other pics



LL
LL
LL
Edited by Ultrasonic2 (muffy) - 7/30/09 at 11:53pm
post #236 of 680
Thread Starter 


post #237 of 680
not sure what you posted in your last post but it didnt load for me...

everything else so far looks great. 30% sounds nice.
post #238 of 680
yeh those images aren't working
POLARity
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 920 D0 @ 4GHz MSI Big Bang Xpower 2x 5850 @ 765MHz/1120MHz 6GB Corsair Dominator DDR3 CL7 1600MHz 
Hard DriveOSMonitorKeyboard
Corsair 64GB SSD | 4x Samsung F3 1TB RAID10 Win7 Professional 64bit Samsung P2770H 27" Roccat Arvo 
PowerCaseMouseMouse Pad
Enermax Revolution 850W POLARity TJ07 Roccat Kone Roccat Sota 
  hide details  
POLARity
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 920 D0 @ 4GHz MSI Big Bang Xpower 2x 5850 @ 765MHz/1120MHz 6GB Corsair Dominator DDR3 CL7 1600MHz 
Hard DriveOSMonitorKeyboard
Corsair 64GB SSD | 4x Samsung F3 1TB RAID10 Win7 Professional 64bit Samsung P2770H 27" Roccat Arvo 
PowerCaseMouseMouse Pad
Enermax Revolution 850W POLARity TJ07 Roccat Kone Roccat Sota 
  hide details  
post #239 of 680
Thread Starter 
Quote:
Originally Posted by oliverw92 View Post
yeh those images aren't working
yeah Overclock.net crashed last night while posting . so it was time for bed for me.

Here You Go. i dont actually trust the below numbers i sure it as 28.4c last night ?.. im gonna spend a bit of time trying to free it up again without touching the cut outs



LL
LL
LL
Edited by Ultrasonic2 (muffy) - 7/31/09 at 1:30pm
post #240 of 680
Thread Starter 
Quote:
Originally Posted by ChielScape View Post
not sure what you posted in your last post but it didnt load for me...

everything else so far looks great. 30% sounds nice.
Actually doing the numbers so far i dont think it will be that much
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Peltiers / TEC
This thread is locked  
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Custom 62mm TEC Water block