Thought you guys would like this....

[maxSaleen]

Link:
http://www.swiftnets.com/Technical/A...erformance.pdf

Thoughts....?

[Joe]

I find it quite interesting.

Some of the conclusions that can be drawn are similar to what you get from Bills radiator testing he did on OC'rs a good long time ago.

I found some of the curves of interest.

[Les]

A brave attempt to rationalise performance with radiators physical characteristics.
The conclusions* are supported by the evidence.
Although I have niggling doubts about the superiority of MCR80-QP over BI Micro being entirely explained by one being single row and the other dual ("This superiority of a single row over a dual row is due to the more efficient use of the fan’s output".).


Think radiator testing and wb testing are at similar crisis points.
Relations between measured data sets and performance are not understood.
The radiator situation is illustrated by the can we agree on the basis of "C" in C/W ? and calibrating' the DUT instrument connections discussions.

The WB situation by the often heated arguments over Testing Method, with particular reference to Apogee versus Storm

Suggest the only way forward is analysis. The interaction between DUT and the Test-Bench has to be analyzed. In particular the relation of Sensor Values to Real Values and Real Values to "Figure of Merit(FOM)"
Jonathan(Incoherent) and I are actively exploring WB /Heat-source/Sensor-Reading/FOM . Bill appears to be alone(with the exception of UNDERBYTE's kind offer) in the radiator labyrinth.
Think, at the moment, any testing can only be used for banking. The difficulty for the tester is that everything has to be recorded.


* "Dual row radiators and heater cores have a greater heat rejection capability than single row, but to utilize the higher capability stronger fans must be used to overcome the airflow resistance."


Edit: Added second radiator link

Edit 2: Replaced truncated and misleading quote of conclusion by full sentence

Edited: In the light of Bill's comments

[bobo5195]

Quite liked it as an article, at least its secondary data to back up bill's old work. Well done Swiftech!

Les dual pass rads are inherently more efficient Kays and London say this, and a simple thought experiment can prove it
A dual pass radiator can be thought of as two singe pass radiators back to back hence they are bound to be more efficient.

[Les]

Originally Posted by bobo5195

....Les dual pass rads are inherently more efficient Kays and London say this, ....

Pass ?
Was talking row.

[gmat]

The top performers are so close in terms of performance that another factor becomes more important IMHO: the price. Also practical mounting solutions can be a factor for many people (who are not modders). Excellent document overall, which provides solid data foundations to common watercooling knowledge.

[BillA]

rather surprised to see the article now, have to wonder why ?

Les
others have had difficulty with my 'conclusions' also (suggestive of being wrong ??)
"However although the conclusion* is supported all could equally well be explained by other variations within the classifications. For example: MCR80-QP vs BI Micro(Graphs 15-19) the difference in thickness and Frontal Area are not considered and could equally explain the superiority ("This superiority of a single row over a dual row is due to the more efficient use of the fan’s output".)."

see Table 1, pg 3; both single and dual row can be compared in 80 and 120mm
I did consider the sizes, they are ordered so in the table
a shame I do not have the data to normalize, but remember that finished units are being compared;
-> not an attempt to define relative core efficiency, rather unit performance (extrapolated to . . . , yea)

care to show how the dimensional data in Table 1 would support your comment ?

bobo
K&L have definitions which may differ from those used by WCers
(row/tube/core, side to side/front to back, all confused wrt automotive nomenclature)
you need to follow our descriptions

the data should be considered primary relative to my previous articles content in terms of accuracy
my present rad testing is 'better' only due to more rigorous system calibration with a quartz thermometer (which HAS revealed differences only now appreciated)

[BillA]

Originally Posted by gmat

The top performers are so close in terms of performance that another factor becomes more important IMHO: the price. Also practical mounting solutions can be a factor for many people (who are not modders). Excellent document overall, which provides solid data foundations to common watercooling knowledge.

quite correct, the dissipation difference in the performance of the top units (of the same size) is indistingusable by any user
rads are now all about price/reliability/features

another consideration not addressed in this testing, unit to unit variation
I've tested multiples of 3 mfgrs' 120mm rads and seem to see ~5% variation (too small a sample)
one can appreciate that a different sample selection could see a reordered ranking

[Les]

Originally Posted by BillA

Les
........care to show how the dimensional data in Table 1 would support your comment ?....

Cannot support. Have not done the analysis. Would prefer not to at the moment.
Would an edit and rephrase suffice?:
"The conclusions are supported supported by the evidence.
Although I have niggling doubts about the superiority of MCR80-QP over BI Micro being entirely explained by one being single row and the other dual ("This superiority of a single row over a dual row is due to the more efficient use of the fan’s output".).
Edited: In the light of Bill's comments"

[BillA]

sloooow
why edit ?, when you wish take a look at those #s - think will be found ~ok

[bobo5195]

bill there are reasons why automotive peeps use their wording. Also consider adding noise to price/reliability/ features

Pass = row in WC wording.

In automotive/ engineering notation; An extra pass is when the output from the rad (hot water) is fed back through the radiator again, creating an effective second rad in series with the first. An extra row would refer to thickening the rad by adding another virtual rad behind the first using a second row of fins. This create parallel flow paths and is akin (although not exactly equal performance wise) with doubling the thickness of the rad.

Keeping the two distinct is important because adding an extra row is a cheap and effective means of getting a little extra performance on the water side at the expense of a vast decrease in air side performance as the rad is a lot thicker. Two passes drastically improve overall efficiency as they make much better use of the air coming through the rad. They degrade water side flow rate but give the air a much better temperature gradient to work with.

You can have much more than 2 passes but it is generally not optimal, cost and air side performance wise. If memory serves me right power stations use many (7 or more) in some cases as efficiency is very important and capital costs and space are not important.

[Joe]

I thought pass was = to how many passes the water made going from side to side? And rows were just called rows..

[BillA]

just observing that there are different words describing the same things

no, not at all correct and you do need to ground your comments in reality
Pass does not equal row, pass refers to the # of circuits (passes) the fluid makes across the face of the rad

i.e. the original BI was a 1 row (single core) 4 pass rad (4 in series, unequal air)
the next BI was 1 row 2 pass (2 in series, mirror air w/unequal flow)
the BIX is a 2 row 2 pass (two 2 row in series)
and now they have both 1 and 2 row in both 1 and 2 pass configurations

over here, in the automotive world they use 'core' for row (of tubes) which may be in-line or offset
the rows may be in series or parallel, or divided into segments; all of which present different 'cross-flow' configurations wrt the air side
I have tested tube flow configurations not commercially offered (translate: followed theory out the window) with no discernable effect

feel free to opine, I did suggest to use our terminology - which means to understand it, no ?

[j813]

In PC cooling it's better to use single pass as to have faster flow?

[Joe]

it depends what you want from your cooling, if you want silent then you need a different setup than a noisy. Just read the graphs its all there

[bobo5195]

i think my powers are slipping bill and i should do some more reading grrrrrr, youre right just ive not being doing alot on WCing of late amd my knowledge is slipping.

Im trying to get an artical done for the wiki on jet impingement but never seem to have time.

Pc water cooling is so air starved that im not sure that complicated setups matter all that much and its more how much air you can get to your fins.