Water Cooling Estimator Spreadsheet

[AndreaBZ]

Thank you.........
Unfortunately I do not know.

[AndreaBZ]

Spreadsheet V.4.00 2009!!
************************************************** *************************

We have included almost all the Pump tested by Skinnee Labs (SKL).
The radiators XSPC RX 120 (SKL)
Pump Iwaki RD30 24V ! tested by Martinn's Liquid Labs (MLL)
WB Enzotech Luna Rev.A tested by Nexthardware (NEX)
......... And more


Push Pull in the Spreadsheet





Download V.4.00 2009 in the first page!


************************************************** *************************

[Hondacity]

andreabz

you're work is great,

I've been testing the core7 particularly the 940...and its hot...been also observing power drawn when i'm running occt or prime95..here's a screenshot..

at idle(connected to my ups the kil-a-watt >psu>mobo cpu ram and raptor no fans or pump connected)

at load


My question regarding the spreadsheet, With everything in its place, if I set my temperatures to say 60C with known components(cpu +cpu waterblock +pump +radiator etc), can we calculate how much total heat in terms of wattage that my cpu is producing?

so far intel's tdp is very good..i've had e8600 a q9550..and both are in the 50-150w range...but this core i7 is a beast...as shown in my pics...but thats power from the wall...computing for power supply load....i'd assume 80%..

but i only have one question with the spreadsheet.. if i change the cpu temperature can we reverse calculate the heat load..in terms of wattage?

[AndreaBZ]

andreabz

you're work is great,

I've been testing the core7 particularly the 940...and its hot...been also observing power drawn when i'm running occt or prime95..here's a screenshot..

at idle(connected to my ups the kil-a-watt >psu>mobo cpu ram and raptor no fans or pump connected)
---
at load
--
My question regarding the spreadsheet, With everything in its place, if I set my temperatures to say 60C with known components(cpu +cpu waterblock +pump +radiator etc), can we calculate how much total heat in terms of wattage that my cpu is producing?

so far intel's tdp is very good..i've had e8600 a q9550..and both are in the 50-150w range...but this core i7 is a beast...as shown in my pics...but thats power from the wall...computing for power supply load....i'd assume 80%..

but i only have one question with the spreadsheet.. if i change the cpu temperature can we reverse calculate the heat load..in terms of wattage?

The "Coeff.* Watt CPU"
Are not the watts products by the CPU But (watt * a coefficient "k")
About this I wrote this article :

http://www.xtremesystems.org/forums/...0&postcount=41

Therefore it is not possible to calculate with the Spreadsheet the watts products by the CPU.
-------------------

[Hondacity]

eurika!!! thanks for the link...i finally have a grasp on the coeficient "k"

[Hondacity]

the total flow is not correct when adding a second radiator...

pls have a look, thanks

[Mech0z]

Year I had an issue when adding another rad, flow would not change at all.

And if possible I would like to see a 3. rad option

[AndreaBZ]

the total flow is not correct when adding a second radiator...

pls have a look, thanks

you are right!
!!!!!!!!!
happens in the last version.
system immediately.
Thank you for an.

[AndreaBZ]

Correct version :

http://ultrashare.net/hosting/fl/c1b...ea_V-4.01-2009

................... is OK ?

Between a little comes the version 5
Equations flow rate vs. pressure drop improved , refined.
Perhaps the radiators XSPC RX series tested by skinne labs.
You have tried to verify if my formula push pull is correct ?

Thanks for your collaboration.

[AndreaBZ]

Spreadsheet V.5.00 2009
coming soon

I have inserted in the spreadsheet the radiators Swiftech MCR320, ThermoChill PA120.3, XSPC RX360

[AndreaBZ]

Spreadsheet V. 5.01 2009
Download available in the first page.

************************************************** ******************************

[Mech0z]

What is it the:
Watt Radiatore 300
Means?

[AndreaBZ]

What is it the:
Watt Radiatore 300
Means?

300 is heat load in watts from radiator
= watt CPU + watt VGA + watt pump (10 - 20 watt)
+ watt chipset mainboard NB SB and mosfet every 30 watt .

[Mech0z]

300 is heat load in watts from radiator
= watt CPU + watt VGA + watt pump (10 - 20 watt)
+ watt chipset mainboard NB SB and mosfet every 30 watt .

So its the total watt load on the radiator? Or is the CPU watt taken from the Coeff.* Watt CPU

[AndreaBZ]

the total watt load on the radiator

Coeff.* Watt CPU is watt CPU * a Coefficient whic may be more or less of 1.
The coefficient is needed to make uniform and comparable data of the various "tester".
Practically...To read in spreadsheet the same temperatures that indicates "core temp".

[bobo5195]

************************************************** *************************
Theory, to me, on mathematical extrapolation of the data on the efficiency of radiators.
Consider the structure of 2 radiators identical.
The surface of the radiators and the number (size) of the fans different .

Mathematical formula for radiators (same structure):
C°/W A = thermal resistance of the radiator with 3 fans ........ height of surface rad.A= hA
C°/W B = thermal resistance of the radiator with 2 fans ........height of surface rad.B= hB
formula:
C°/W B = C°/W A*(((3fans/2fans)+(hA/hB))/2) = C°W A*(((2/3)+(hA/hB))/2)
C°/W A = C°/W B*(((2fans/3fans)+(hB/hB))/2) = C°W A*(((3/2)+(hB/hA))/2)


************************************************** *************************
In spreadsheet the radiators calculated mathematically are listed with 2 asterisks "**"
************************************************** *************************

Some examples:

This the chart of the radiators
Swiftech MCR 320 MLL e Swiftech MCR 220 MLL (Tested by Martin):



This the chart of the radiators
Swiftech MCR 320 MLL and Swiftech MCR 220** extrapolated from the radiator Swiftech MCR 320 MLL



This finally the chart of the radiators
Swiftech MCR 220 MLL and Swiftech MCR 220** extrapolated from the radiator Swiftech MCR 320 MLL


Seem coincidental .
************************************************** *************************
Still. ............
XSPC RS 240 MLL and XSPC RS 240 ** extrapolated from the radiator XSPC RS 360 MLL

Too these almost coinciding.
************************************************** *************************
Another example of extrapolated data........
This the chart of radiators:
Feser X-Changer 480 MLL
Feser X-Changer 480 Next


This the chart of the radiator:
Feser X-Changer 480**extrapolated from radiator:Feser X-Changer 360 Next

This the chart of radiators:
Feser X-Changer 480 Next
Feser X-Changer 480** (extrapolated)

.....very similar.

This the chart of radiators:
Feser X-Changer 480 MLL
Feser X-Changer 480** (extrapolated)

Practically the same!
A further demonstration of validity of "extrapolated data".
************************************************** *************************
In this example instead .........XSPC RS 120 MLL and XSPC RS 120 ** extrapolated from the radiator XSPC RS 240 MLL


Are not coincidental.
In my opinion it is preferable the chart extracted from the data extrapolated mathematically,
preferable to that which is obtained from measurements of Martin Liquid Lab (MLL).
............
This is the chart of three radiators XSPC tested by Martin (MLL)
and of the radiator XSPC 120 ** "extrapolated" from the data of the radiator XSPC 240 MLL.

To it seems me that the chart of the radiator XSPC 120 ** "extrapolated" is more in line with the other 2.
The chart of the radiator XSPC 120 ** "extrapolated" is more realistic.
************************************************** *************************

This is correct as a first order interpolation given what radiators do. The actual formula for radiators is something called Log mean temperature differentail. Which would say that the water entering the last fan of a 3 fan radiator is actually cooler so will have less heat transfered away from it.

Look at the wikipedia entry for crossflow heat exchangers as an example
http://en.wikipedia.org/wiki/Heat_exchanger

Compact heat exchangers by Kays and landon is the holy bible on this subject despite it being from the 50's there are some newer computational methods being developed though
http://www.amazon.com/Compact-Heat-E.../dp/1575240602


Me and bill A and then martin had an argument about this your approximation is good because if you don't do this it is hard to model in excel without a good equation solver.

With available equipment and temperature differentails in the water cooling scene it is impossible to calculate anything accurately as the test data is not 100% right.

[AndreaBZ]

***************


Unfortunately "Babylon" the translation software that i use gave me a text almost incomprehensible.


For the radiators I use a formula for the type .........Y = a*X^n with n < 0 and > -1 .



I think the same used by Martin.

Is a function easily manageable with Excel and, to me, quite correctly describes the phenomenon physical.

According to me , it is very valid if the difference between the water temperature and
the ambient temperature is not too high.

The data extrapolated in this way provide data very similar to the data measured.

I have given above some examples.

The difference is of two tenths of a degree in the cases more unfavourable.

Of course this is a nearest,depends on the accuracy of formula and the accuracy of data from extrapolate.

Another example of extrapolation:
Fans in push pull mode .......

This is a chart published by Martin



I indicated the two temperatures interesting.

This is the formula I have published long ago :

http://www.xtremesystems.org/forums/...19&postcount=6

************************************************** *************************
Radiators with fans in push pull

**************
See :
Mathematical extrapolation of the data on the efficiency of radiators.

http://www.xtremesystems.org/forums/...0&postcount=32



Consider the structure of 2 radiators identical.
The surface of the radiators and the number of the fans different .

Mathematical formula for radiators (same structure):
C°/W A = thermal resistance of the radiator with 3 fans ........ height of surface rad.A= hA
C°/W B = thermal resistance of the radiator with 2 fans .........height of surface rad.B= hB
formula:
C°/W B = C°/W A*(((3fans/2fans)+(hA/hB))/2) = C°W A*(((2/3)+(hA/hB))/2)
C°/W A = C°/W B*(((2fans/3fans)+(hB/hB))/2) = C°W B*(((3/2)+(hB/hA))/2)

Analogously ..............

Push Pull .........
The structure of 2 radiators : identical
The surface of the radiators : identical
The number of the fans = X*2

Mathematical formula for radiators (same structure):
C°/W A = thermal resistance of the radiator with X*2 fans ....... height of surface = h
C°/W B = thermal resistance of the radiator with X fans .........height of surface = h

formula:

C°/W A = C°/W B*(((Xfans/X*2fans) + (h/h))/2)
= C°/W B*((( 1/2) + (1))/2)
= C°/W B*(0.5+1)/2
= C°/W B*0,75

C°/W A = C°/W B*0,75



....... 10,1° * 0,75 = 7,575°........very similar to the data published by Martin.


Is always pleasant and interesting exchange views on this mysterious world.
......

[bobo5195]

You may want to read this. Hopefully the many postings will clarify what is talked about...

http://forums.procooling.com/vbb/showthread.php?t=12306

[Power_VANO]

Great work you are doing here Andrea

[azza21]

Mate i can't seem to get this spreadsheet from the links provided, anychance of a new link? Also hows the Spreadsheet V.5.00 version coming along?

Az

[AndreaBZ]

Mate i can't seem to get this spreadsheet from the links provided, anychance of a new link? Also hows the Spreadsheet V.5.00 version coming along?

Az

New link In the first page

[AndreaBZ]

************************************************** *************************

************************************************** *************************

[Mech0z]

Do you think it would be possible to use the data from martins fantesting to be able to pick which fans your using and then see how much difference there is between each fan. That would be super sweet, and you would proberly only have to add the most popular ones

[AndreaBZ]

Do you think it would be possible to use the data from martins fantesting to be able to pick which fans your using and then see how much difference there is between each fan. That would be super sweet, and you would proberly only have to add the most popular ones

I am working
A preview .......


I wait for the test of Skinnee on the radiator feser X changer

Then the update of the spreadsheet arrives

[AndreaBZ]

*************************************************