Sign in with
Sign up | Sign in
Your question

Radiator - thick or thin

Last response: in Overclocking
Share
December 7, 2011 5:35:32 PM

well.. honestly there are a lot of radiators out there..

can someone explain me the performance diff between -

1. thick rad (~50-60mm) medium fin density (~10-15fpi)
2. thick rad (~50-60mm) low fin density (~5-7 fpi)
3. thin rad (~30mm) high fin density (~20-30 fpi)

all the rads are of same size.. i know fan will matter cos of their rpm /static pressure/flow but still could someone, like tabulate it?

also, on a completely diff note.. has anyone tried using car coolant in their loop? any benefit/drawback over distilled water?

More about : radiator thick thin

a c 324 K Overclocking
December 7, 2011 5:40:38 PM

It all has to do with surface area...more FPI means more surface area for dissipation.

What thick rads are 10-15 FPI? Can you link 1-2?

As for thick rads with lower FPI and thin rads with higher FPI- again, its surface area. I guess it really depends on what you are looking for as a comparison?

You can use automotive coolant- I don't really see the need unless you run mixed metals in your loop or intend on running components outside in the winter (or in a location prone to freezing temps). Otherwise, there are no performance benefits to doing so.
m
0
l
December 8, 2011 6:31:05 AM

strongdc said:
well.. honestly there are a lot of radiators out there..

can someone explain me the performance diff between -

1. thick rad (~50-60mm) medium fin density (~10-15fpi)
2. thick rad (~50-60mm) low fin density (~5-7 fpi)
3. thin rad (~30mm) high fin density (~20-30 fpi)

all the rads are of same size.. i know fan will matter cos of their rpm /static pressure/flow but still could someone, like tabulate it?

also, on a completely diff note.. has anyone tried using car coolant in their loop? any benefit/drawback over distilled water?

A larger thin rad is better than a smaller thick rad, thicker rads will need higher rpm fans.
m
0
l
Related resources
December 8, 2011 7:55:40 AM

@rubix : phobya and ek rads are 11 and 10 fpi resp.

@homeboy2 : the rads i'm considering are same size.

but the ques still remains.. which rad is better??
a thick rad will hv less fpi but larger fins while a thin rad will hv high fpi but smaller fins.. so i guess the surface area will be nearly same..
m
0
l
a c 324 K Overclocking
December 8, 2011 12:31:59 PM

Yeah, I was thinking EK or Magicool might run 10-12 FPI...I was going to check...I guess I didn't recall Phobya being in the same ballpark. Be careful with your generalization- not all thin or thick rads are designed the same. There are thin rads that are both low FPI and others that are higher FPI (Swiftech, low. Koolance and BIX run 20-30 FPI on some rads of the same thickness).
m
0
l
December 8, 2011 12:51:20 PM

all the more reasons to distribute them and classify them into categories. that is something i believe would help anyone going for water cooling to select a radiator.. all the stickies give info on the size of radiators one should go for, but which type of radiator? no idea.. :|
m
0
l
a c 324 K Overclocking
December 8, 2011 1:03:56 PM

To be completely honest, you'll find that most rads that run 7-12 FPI will perform around the same. FPI isn't the only thing that contributes to performance; some manufacturers use different size tubes, more tubes or different metal thicknesses...this all accounts for subtle differences. When you stop to compare radiators of the same thickness with lower FPI (7-12) and those with higher FPI (20-30) the difference maker is fans. However, there typically is only 10-15% difference (at most) in total performance between these differences when using fans designed for each.

When it comes to the biggest differences in radiator performance, simple size and volume of the rad itself is the largest factor over FPI or tubing design. Outside of that, it is a lot of complicated thermodynamics and heat transfer properties that need to be taken into account for the specific performance of one radiator design to another.

At the end of the day, most watercooling radiators from reputable manufacturers will provide moderately similar performance based on size and fin density alone.

You have, however, given me a lot of things to mull over in attempt to try and come up with some intricately detailed ways of determining specific qualities for radiator size and performance characteristics. The only problem is, will there really be a realistic way to do this and not lose my mind in the process. :) 
m
0
l
December 8, 2011 1:22:54 PM

It seems I recall reading on another forum where it all comes down to the type of fan used.
Higher fpi means you will need a higher static pressure (higher cfm?) to push through the area and have the outgoing air cfm as high as you would like it to be.
Whereas a lower fpi means you can run slower, lower static pressure fans and still achieve the outgoing cfm rates you desire.
m
0
l
a c 324 K Overclocking
December 8, 2011 1:34:01 PM

Static pressure and CFM are different, but often confused as being related 1:1. To generalize as you stated, typically higher CFM fans ARE capable of higher static pressure simply because of the power needed to push the rated CFM they are listed at. However, 2 fans that might advertise 90 CFM might have very different static pressure properties.
m
0
l
December 8, 2011 6:14:55 PM

rubix_1011 said:
To be completely honest, you'll find that most rads that run 7-12 FPI will perform around the same. FPI isn't the only thing that contributes to performance; some manufacturers use different size tubes, more tubes or different metal thicknesses...this all accounts for subtle differences. When you stop to compare radiators of the same thickness with lower FPI (7-12) and those with higher FPI (20-30) the difference maker is fans. However, there typically is only 10-15% difference (at most) in total performance between these differences when using fans designed for each.

When it comes to the biggest differences in radiator performance, simple size and volume of the rad itself is the largest factor over FPI or tubing design. Outside of that, it is a lot of complicated thermodynamics and heat transfer properties that need to be taken into account for the specific performance of one radiator design to another.

At the end of the day, most watercooling radiators from reputable manufacturers will provide moderately similar performance based on size and fin density alone.

You have, however, given me a lot of things to mull over in attempt to try and come up with some intricately detailed ways of determining specific qualities for radiator size and performance characteristics. The only problem is, will there really be a realistic way to do this and not lose my mind in the process. :) 



try not to get lost :p 

guess my engineering background has made my decision making to be entirely based on proven facts...
not dat its bad.. but now i find it hard to make choices based just on taste, preference.. :( 
m
0
l
a c 324 K Overclocking
December 8, 2011 6:23:52 PM

No, I completely understand; being an IT systems engineer, you are always driven for answers to be yes or no...not maybe's.

I was trying to imply that the majority of radiators produced are going to perform well with other peer rads relative to the size/dimensions being shared. Even when it comes to fin density, you don't get major data swings either way. With so many choices, it does mean there is a lot to decide on, but the most important concerns should be getting rads that fit your implementation, and if you have enough rad performance to effectively dissipate the heat watts your hardware is dumping into the loop.
m
0
l
December 8, 2011 6:53:27 PM

silverthorn said:
It seems I recall reading on another forum where it all comes down to the type of fan used.
Higher fpi means you will need a higher static pressure (higher cfm?) to push through the area and have the outgoing air cfm as high as you would like it to be.
Whereas a lower fpi means you can run slower, lower static pressure fans and still achieve the outgoing cfm rates you desire.


cfm is a measure of flow, which is different from static pressure. static pressure is the pressure diff at inlet and outlet of fan (or pump) when there is no flow of the fluid (that's why its called static pressure).

the rated cfms are flows (i guess) at no obstruction to the fan. when you install a radiator (= obstruction) flow reduces. More the restriction lower the flow. so you need a high static pressure fan to keep the air moving. its just like when you add a restrictive block to your system, flow rates drop and you need a more powerful (higher static pressure) pump to make the liquid go around.

usually high cfm fans have higher static pressure but not necessarily. similarly, high rpm fans have high cfm but not necessarily
these depend on the actual design and construction of blades/fins of the fan which varies from manufacturer to manufacturer.
m
0
l
December 8, 2011 6:59:04 PM

rubix_1011 said:
No, I completely understand; being an IT systems engineer, you are always driven for answers to be yes or no...not maybe's.

I was trying to imply that the majority of radiators produced are going to perform well with other peer rads relative to the size/dimensions being shared. Even when it comes to fin density, you don't get major data swings either way. With so many choices, it does mean there is a lot to decide on, but the most important concerns should be getting rads that fit your implementation, and if you have enough rad performance to effectively dissipate the heat watts your hardware is dumping into the loop.


so it basically boils down to -
1. thermal requirement - single, double,etc.
2. physical constraints - size that fits

beyond these its just a matter of personal taste, prices and availability of a particular product.


PS - one thing that varies is the fan spacing. this can be a major factor when fighting a radiator inside the cabinet.
m
0
l
a c 324 K Overclocking
December 8, 2011 7:05:54 PM

You are completely correct with that list. This is by far the best way anyone can agree on radiator choice for a given build. I know it is by far less-scientific than a lot of people wish it was (including me) but it really comes down to these.

You make a very good point on the fan spacing issue...this was something I overlooked. It really is one of the last form size variances that isn't standardized in watercooling and can cause some headaches if you aren't careful.
m
0
l
December 8, 2011 7:18:56 PM

rubix_1011 said:
You are completely correct with that list. This is by far the best way anyone can agree on radiator choice for a given build. I know it is by far less-scientific than a lot of people wish it was (including me) but it really comes down to these.


"far less-scientific" :pt1cable: 


rubix_1011 said:
You make a very good point on the fan spacing issue...this was something I overlooked. It really is one of the last form size variances that isn't standardized in watercooling and can cause some headaches if you aren't careful.


atleast u can get the data from the radiator manufacture's website. but on the cabinet side, i haven't seen any company sharing such data. :( 
i have a Lancool K62 and the fan spacing is 20mm between the top two 140mm fans. this i got by measuring. this data is not available anywhere on the lancool site. :(  that limited my choice of rad to black ice. all other brands have 16mm fan spacing.
m
0
l
a c 324 K Overclocking
December 8, 2011 7:26:25 PM

Drilling holes is an option, but you really have to take your time and get your measurements down. Also, depending on how you plan on mounting it, you might be able to use mount holes on the 'corners' of the rad if the fans would be mounted on the opposite side. There are a couple options that would be easy to determine if you wanted to go with other options...but most would require drilling a few well-placed holes it seems.
m
0
l
December 8, 2011 7:36:54 PM

rubix_1011 said:
Drilling holes is an option, but you really have to take your time and get your measurements down. Also, depending on how you plan on mounting it, you might be able to use mount holes on the 'corners' of the rad if the fans would be mounted on the opposite side. There are a couple options that would be easy to determine if you wanted to go with other options...but most would require drilling a few well-placed holes it seems.


once you decide to pick up the drill , there's nothing stopping you from fitting anything anywhere ;)  it opens up a whole new world. :D 
but lets acknowledge d fact that case modding is not for the faint of heart.
m
0
l
December 8, 2011 7:42:10 PM

noob2222 said:
http://skinneelabs.com/2011-2012-radiator-comparison/5/
Thicker rads will normally have less pressure drop/flow restriction than thin rads.



hmmm.. i doubt if that ll affect d pump much..


noob2222 said:
http://skinneelabs.com/2011-2012-radiator-comparison/5/
As far as water vs antifreeze, its about heat transfer.

http://www.koolance.com/technical/cooling101/002.html

Water transfers heat over twice as efficient as glycol. lower transfer rates means you need more radiator to have the same performance.


if i use a mix of distilled water + car coolant. does it have biocidal properties? or i still hv 2 use a kill coil??
m
0
l
a c 324 K Overclocking
December 8, 2011 7:44:40 PM

Quote:
Thicker rads will normally have less pressure drop/flow restriction than thin rads


Rads are the lowest restriction components in a loop, anyway, and restriction is never really an issue with them. Your CPU block is likely the most restrictive component by far, unless you run MOSFET or RAM blocks. I'd also be willing to bet that the differences between flow restrictions in any rad is negligible for any loop- the only reason thicker rads are less restrictive than thin rads is the simple fact that they contain larger tubes due to the increased thickness...this allows more flow. Rad restriction numbers are purely scientific and really doesn't impact flow, overall.

Glycol would only be recommended in a mixed metals environment- something you don't want, anyway. Otherwise, unless you are running your rad outside your window in the winter, you don't need or want it...no performance benefit. Distilled and a killcoil are cheaper than a jug of antifreeze.
m
0
l
December 8, 2011 7:51:27 PM

rubix_1011 said:

Glycol would only be recommended in a mixed metals environment- something you don't want, anyway. Otherwise, unless you are running your rad outside your window in the winter, you don't need or want it...no performance benefit. Distilled and a killcoil are cheaper than a jug of antifreeze.


i'm in Goa, India. and i doubt if the temp goes below 15C, even at the chillest of times :D  mostly its around 35C thought d day and 20C at night.
m
0
l
a c 324 K Overclocking
December 8, 2011 7:52:56 PM

Ah, well then I would say you shouldn't need to worry about freezing temperatures, then. :) 
m
0
l
!