Painted rad bad for cooling?

davidm71

[H]ard|Gawd
Joined
Feb 11, 2004
Messages
1,568
Guys,

Was wondering I have a xspc rx360 v1 rad that i spray painted ten years ago and was never happy with the performance. Nevermind that had a perforation that I had fixed by an autobody place. So this rad should I retire it? Or would it be safe to use paint thinner to clean up the fins of paint?

Thanks
 
Guys,

Was wondering I have a xspc rx360 v1 rad that i spray painted ten years ago and was never happy with the performance. Nevermind that had a perforation that I had fixed by an autobody place. So this rad should I retire it? Or would it be safe to use paint thinner to clean up the fins of paint?

Thanks


Did you paint just the exterior, or the fins as well?
 
Paint is an insulator. Unless you bought paint specifically for electric or heat transfer... it will massively hurt performance.
 
Depending on how much time you have on your hands, you could always strip it and re-paint it with the proper thermal paint.

It's probably easier to replace it though. If you already have all the other parts, a radiator doesn't make a huge cost difference.
 
Paint is an insulator. Unless you bought paint specifically for electric or heat transfer... it will massively hurt performance.
that's BULLSHIT. yes, paint is an insulator, however a thin coat or 2 of paint will not effect temps not even 1c. Been painting rads for builds 10+ years (never paint core that is common sense) never seen a temp difference and I have proven this time and time again. To say "Massissivly hurt performace" is a fallacy on your part
 
Most rads you buy today are painted or powder coated. I have 3 copper rads from different companies sitting around here and all of them have painted fins (the XSPC and Black Ice ones I have are powder coated). The only way to get around that is to buy an unpainted bare rad. Just be aware you won't get that from anyone that has a known reputation in the cooling business. Of it's sim,ply painted you can of course strip it. Powder coating is a little tougher to get off.
 
Most rads you buy today are painted or powder coated. I have 3 copper rads from different companies sitting around here and all of them have painted fins (the XSPC and Black Ice ones I have are powder coated). The only way to get around that is to buy an unpainted bare rad. Just be aware you won't get that from anyone that has a known reputation in the cooling business. Of it's sim,ply painted you can of course strip it. Powder coating is a little tougher to get off.


Every radiator I've bought had been painted on the outside, but the fins are bare copper.

I mistook it for rust when I first got them:

63464_IMG_20160710_173030.jpg
 
that's BULLSHIT. yes, paint is an insulator, however a thin coat or 2 of paint will not effect temps not even 1c. Been painting rads for builds 10+ years (never paint core that is common sense) never seen a temp difference and I have proven this time and time again. To say "Massissivly hurt performace" is a fallacy on your part
I would think painting the fins would have a greater impact. I’ll stand corrected.
 
Every radiator I've bought had been painted on the outside, but the fins are bare copper.

I mistook it for rust when I first got them:

View attachment 126319
That's the opposite of the rads I have owned. I have 3x 360mm rads now - one an XSPC EX360, The next a Hardware Labs Nemesis 360 and the third a Chinese no-name copper rad. All of them have painted cores and there are literally dozens of reviews out there for the first two that show they have really decent performance. So honestly I am not all that sure if the paint takes performance way. I haven't ever had any problems with painted core fins.
 
That's the opposite of the rads I have owned. I have 3x 360mm rads now - one an XSPC EX360, The next a Hardware Labs Nemesis 360 and the third a Chinese no-name copper rad. All of them have painted cores and there are literally dozens of reviews out there for the first two that show they have really decent performance. So honestly I am not all that sure if the paint takes performance way. I haven't ever had any problems with painted core fins.


I could see if a proper thermally conductive powder coat were used it would only have a minimum impact on heat transference.

Problem is, OP is talking about a radiator he spray painted himself, under - probably - less than ideal conditions. I mean, you might luck out ad have this work without issues, but I'd argue its safer not to.
 
Last edited:
Yeah agreed. I doubt paint from a rattle-can is going to be all that thermally conductive.
 
Yeah agreed. I doubt paint from a rattle-can is going to be all that thermally conductive.
Agreed. There's "painted" and then there's "painted," if ya catch my drift.

Even so, I wonder how much impact it has. I've got a spare XSPC rad laying around that I've considered stripping entirely. I wonder if it's experiment time!
 
Agreed. There's "painted" and then there's "painted," if ya catch my drift.

Even so, I wonder how much impact it has. I've got a spare XSPC rad laying around that I've considered stripping entirely. I wonder if it's experiment time!

If you do it, I'd be curious what you find!
 
Agreed. There's "painted" and then there's "painted," if ya catch my drift.

Even so, I wonder how much impact it has. I've got a spare XSPC rad laying around that I've considered stripping entirely. I wonder if it's experiment time!
I think it's a good idea.

If you're going to strip it anyhow, you're already in it for labor of doing that work. To strip more paint won't really incur that much addition time.

The time will be in the test, breakdown, paint, re-test.

To that end, if I may suggest, aside from the "givens" like using the same hardware, same environment and ambient temp, and system load...
-Testing it how it is, right now, from XSPC. Call this "Baseline A".
-Test it with a couple coats of rattle-can spray paint for "worst-case scenario", don't bother stripping XSPC's IMO, since we have Baseline A to extrapolate the temperature difference of your painting over it. Call this "Ghetto Mod Painted". "Hell, use a bright color to really drive home the label of Ghetto lol)
-Strip it of all its paint and now test it again. From that we'll get an idea of how much of an impact XSPC's paint impacts things (likely negligible). Call this "Baseline B".
-Last, paint it the way you planned to all along, and test it again. Call that "Proper Paint Mod" or something. Hopefully you had planned on using the aforementioned "thermal paint" for this, to really "science the shit out of this", since that tells us how worth-it using that paint is.

Now if you really want to go above and beyond even that amount of work, testing inbetween layers of paint would be interesting, but I'm guessing a waste of time. However, prior to stripping it off, go actually-ghetto and just bomb the hell out of it with a thick-ass "I don't have a clue how to paint" layer of paint. lol But seriously, that'd really give us an idea of how impacting the paint is.


My assumptions:
Painting will have an impact.BUT the impact is going to only be on the temperature prior to heat-soak! It'll cause the initial-laod temp to rise higher by maybe 2.5C (I bet 4C if you do do a super-heavy coat test), but once it hits the heat-soaked limit, it'll probably only be 1C avg, maybe 1.5C.
Stripped totally down to the metal, this will possibly vary. I don't doubt that the temps will be marginally better.
However, I could see it being a fair bit better IF the stripper used is caustic enough to etch and pit the copper or aluminum (whatever your rad's fins are made from), which in turn would both increase surface area and could also result in more turbulent air... Granted, that's definitely an extreme "maybe", as who knows how much that minute increase in surface area or air turbulence will be, and on top of that if it'd be enough to impact temps

"Engage your safety squints! Aaaaand CORNTACT!!" (points for anyone who gets that reference lol)
 
I think it's a good idea.

If you're going to strip it anyhow, you're already in it for labor of doing that work. To strip more paint won't really incur that much addition time.

The time will be in the test, breakdown, paint, re-test.

To that end, if I may suggest, aside from the "givens" like using the same hardware, same environment and ambient temp, and system load...
-Testing it how it is, right now, from XSPC. Call this "Baseline A".
-Test it with a couple coats of rattle-can spray paint for "worst-case scenario", don't bother stripping XSPC's IMO, since we have Baseline A to extrapolate the temperature difference of your painting over it. Call this "Ghetto Mod Painted". "Hell, use a bright color to really drive home the label of Ghetto lol)
-Strip it of all its paint and now test it again. From that we'll get an idea of how much of an impact XSPC's paint impacts things (likely negligible). Call this "Baseline B".
-Last, paint it the way you planned to all along, and test it again. Call that "Proper Paint Mod" or something. Hopefully you had planned on using the aforementioned "thermal paint" for this, to really "science the shit out of this", since that tells us how worth-it using that paint is.

Now if you really want to go above and beyond even that amount of work, testing inbetween layers of paint would be interesting, but I'm guessing a waste of time. However, prior to stripping it off, go actually-ghetto and just bomb the hell out of it with a thick-ass "I don't have a clue how to paint" layer of paint. lol But seriously, that'd really give us an idea of how impacting the paint is.


My assumptions:
Painting will have an impact.BUT the impact is going to only be on the temperature prior to heat-soak! It'll cause the initial-laod temp to rise higher by maybe 2.5C (I bet 4C if you do do a super-heavy coat test), but once it hits the heat-soaked limit, it'll probably only be 1C avg, maybe 1.5C.
Stripped totally down to the metal, this will possibly vary. I don't doubt that the temps will be marginally better.
However, I could see it being a fair bit better IF the stripper used is caustic enough to etch and pit the copper or aluminum (whatever your rad's fins are made from), which in turn would both increase surface area and could also result in more turbulent air... Granted, that's definitely an extreme "maybe", as who knows how much that minute increase in surface area or air turbulence will be, and on top of that if it'd be enough to impact temps

"Engage your safety squints! Aaaaand CORNTACT!!" (points for anyone who gets that reference lol)
A nice thorough approach. I'll dig through my remaining gear and see what I've got to test this with. Unfortunately, I've sold off all my processors and motherboards, so I'd really only be able to test this by plumbing the rad up to my main rig.

That wouldn't be TOO hard to do but it would be a bit of hassle. We'll see if I have time for it hahaha
 
A nice thorough approach. I'll dig through my remaining gear and see what I've got to test this with. Unfortunately, I've sold off all my processors and motherboards, so I'd really only be able to test this by plumbing the rad up to my main rig.

That wouldn't be TOO hard to do but it would be a bit of hassle. We'll see if I have time for it hahaha
Could always take the "FrostyTech" approach of strapping it to a known-wattage hotplate, along with a thermocouple... that's assuming you have digitial multimeter with a thermocouple port. heh

:\ If I had disposable funds, I could ship you my old Phenom II X6 1090T, 790FX board and some RAM. (assuming it all even works still)
 
Could always take the "FrostyTech" approach of strapping it to a known-wattage hotplate, along with a thermocouple... that's assuming you have digitial multimeter with a thermocouple port. heh

:\ If I had disposable funds, I could ship you my old Phenom II X6 1090T, 790FX board and some RAM. (assuming it all even works still)
That's actually not a bad idea, the hotplate. Then you can apply a constant thermal load to the system. Actually, using a setup like that I can use a coolant temp stop plug probe and read it with my Aquaero.

I've got a standalone pump/res setup that I use for cleaning components and a spare CPU block.

Jankiness in the name of science!
 
That's actually not a bad idea, the hotplate. Then you can apply a constant thermal load to the system. Actually, using a setup like that I can use a coolant temp stop plug probe and read it with my Aquaero.

I've got a standalone pump/res setup that I use for cleaning components and a spare CPU block.

Jankiness in the name of science!

Or you could just have an open radiator with fans running at a known (low) speed on it.

monitor the temps, and pour a distilled water of a known (heated) temp into one of the openings until full.

Measure how much time it takes for temperature to drop.

This seems the simplest to me.
 
Or you could just have an open radiator with fans running at a known (low) speed on it.

monitor the temps, and pour a distilled water of a known (heated) temp into one of the openings until full.

Measure how much time it takes for temperature to drop.

This seems the simplest to me.
Wouldn't work, or at least my automotive side says it won't.
While filling it with the hot water you're either going to have to have the bottom open a little, which is going to skew results, or you're going to end up with water pockets, which will REALLY skew the results.

I could very well be wrong though, might not be as big of an issue with a computer watercooling rad, at least provided it has a bleeder valve on the top tank.
 
that's BULLSHIT. yes, paint is an insulator, however a thin coat or 2 of paint will not effect temps not even 1c. Been painting rads for builds 10+ years (never paint core that is common sense) never seen a temp difference and I have proven this time and time again. To say "Massissivly hurt performace" is a fallacy on your part

Every time I've painted the fins on a radiator, even slightly, I've had degraded performance. There were a few builds that were "radiator fins" out, that needed to be white or black. Even the thinnest paint layers caused a few degrees rise in temps. 1-3c maximum. But when using on the same build, without re-seating either blocks (typically when I do case mods, I don't bother re-seating or cleaning my blocks out) it's a farirly good, albeit imprecise, benchmark.

And use a TEC and waterblock to measure radiator capacity. By far the cheapest method. Many 200w TEC's run on 12v. :p
 
Back
Top