Is this even better than regular heater core?

[OTMOPO3OK]

Hi guys i found this heater core and i have a question...
is it better than regular core where water makes a loop?
(goes down and back up--i mean a core with tubes on one side)
but in this one it just runs thru once...meaning it flows twice slower than the other core(above) since all the water devides among all the tubes...in regular core it flows only in half of the core and back in other half....
so which one is better...can anyone advise....just from theoretical point of view....
thank you...
http://www.heatercore4u.com/2-149.htm

[Since87]

I suspect that might be an extremely good performer. (In some systems particularly.)

I'm assuming you'd chop the pipes down past those elbows.

I don't know the details of that heatercore, but if all else is equal to more common heatercores, I would expect it to have about 1/4 the resistance to water flow that a typical heatercore of the same size has.

This might be an extremely good HC to pair with a very low resistance block like an MCW462. In most systems with MCW462's, the low flow resistance (of the block) is of little value because the rad/HC limits the flowrate greatly.

I suspect this might be a good heater core to pair with any block though. It's unfortunate the intake and outlet are not in opposite corners. I suspect the water flow will be disributed somewhat unevenly through the flat tubes. The mismatch may be small enough, that it's not of much significance though.

Edit: BTW, having the water flowing more slowly in the heater core tubes is not desireable in and of itself. The potential value of this heatercore is in the probably reduced pressure drop compared to other cores of similar size.

I wish I'd known that this HC existed earlier.

[Blackeagle]

A core like that could also, if placed right, reduce your total length of lines in a system, and shorter lines would reduce your pressure drop as well.

edit:
An even better cooling one is 2-342, same idea but more total area if you have space for it. And on 2-342 the inlet/outlet are on opposite corners.

Gotta like it.

[OTMOPO3OK]

thank you for the info guys

[Since87]

Originally posted by Blackeagle

An even better cooling one is 2-342, same idea but more total area if you have space for it. And on 2-342 the inlet/outlet are on opposite corners.

Dammit. A friend of mine who builds guitars did some slumming and built me a nice wooden shroud box for my heatercore. Now I'm very interested in this one.

I'd be willing to chip in $10 towards getting one and shipping it to BillA for his testing if he would be willing. Anyone else?

Edit: link

[Spamz0r]

that heator website is the ugliest thing i have ever seen !

[breathe_nitrox]

Another HC with the inlet/outlet on oposite sides (single pass) is the 2-617 (Ford truck). THe nice thing about this one, at least for my set up is the pipes coming striaght off of the hc, no elbows.

[Blackeagle]

SpamzOr,

Polish a brass core with brasso or other metal polish and seal and they'd lood alright.

Nitrox,

How & where in case would you set up 2-617 so as to avoid the use of 90's with it?

I'd mount the 2-342 rad on the case floor of a wide case. With the pump back under the drive bays and the rad extending out in front of the motherboard slightly. This would offer reduced head loss by being single pass and shortening lines.

Leaves the question of just how much differance the above advantages would make in frow rates.

Since87,

I'll help with that testing gladly. But would we not need to also provide a 2-199 for a equal sized core of the "standard type" for a proper comparison? A very good idea if BillA would be interested.

And would this not help even with a high head loss waterblock, (White Water, remember BillA's suggestion to match the White Water with a system otherwise set up for minimal head loss?)? At least then you'd be dropping head where it would do more for temps, although a high flow block would perhaps gain more.

A interesting thread.

BE

[Since87]

Originally posted by Blackeagle

And would this not help even with a high head loss waterblock, (White Water, remember BillA's suggestion to match the White Water with a system otherwise set up for minimal head loss?)? At least then you'd be dropping head where it would do more for temps, although a high flow block would perhaps gain more.

It would depend on how much the speed of water flow through the flat tubes of this heatercore was reduced, compared to a similar core with both barbs on one end, and the effect that the change in flow speed had on cooling.

I would guess that this might be a good heatercore to pair with any block, but without the actual measurement data I couldn't say that with a lot of confidence.

Maybe later this evening I'll do some more detailed performance guesstimations, but I've had a four year old sitting on my shoulders while writing most of this post, and it's time to play Candyland.

[breathe_nitrox]

Originally posted by Blackeagle
Nitrox,

How & where in case would you set up 2-617 so as to avoid the use of 90's with it?

BE [/b]

Sorry I should have been more clear. I'll be setting the HC up externally. If your putting it in the case then the 2-617 would probably be a poor choice. I just jumped on the thread because I also like the single pass solution, and was intrigued that others were thinking along the same lines.

[Since87]

I did some simulations of different hypothetical systems to show the what might be expected using one of these heater cores.

Because the only HC I've seen flow resistance data on is BillA's "Big Momma", the most reasonable thing I can do is propose a hypothetical HC which has barbs and tanks like a 2-342 but is otherwise identical to a "Big Momma". I've dubbed this hypothetical HC "Fast Momma".

I've assumed that "Fast Momma" would have 1/4 the flow resistance of "Big Momma" because the flow restriction of "Fast Momma" can be thought of as twice as many flat tubes of half the length. Big Momma as tested by Bill had 3/8" hose barbs which skews the picture somewhat but I'm ignoring that as well as a couple dozen other things in this analysis.

The following graphs show a hypothetical pump PQ curve, (lets call the pump a Sim 1250) a pressure drop vs flowrate curve for a system composed of waterblock and HC, (with tubing totally neglected) and a pressure drop vs flowrate curve for the waterblock on it's own.





By identifying the flowrate at the point where the 'system' curve intersects the pump curve, you can determine the flowrate through the system.

The flowrate of White Water will be improved somewhat by pairing it up with 'Fast Momma'.

WW w/ BM flowrate ~ 5.9 lpm
WW w/ FM flowrate ~ 6.9 lpm
improvement ~ 17%

Now let's look at a Swiftech MCW462-UH paired with both heater cores.





In this case there is clearly a much bigger difference:

UH w/ BM flowrate ~ 8.6 lpm
UH w/ FM flowrate ~ 13.3 lpm
improvement ~ 55%

[BillA]

liquid side dps are quite simple to run
the air side is much more complex
and the dissipation is a super chore (due to the air side pressure AND temp control)
- I have all the equip but it takes ~4 days to setup and calibrate
and some of it is presently occupied testing wbs for JoeC's Roundup

deeppow has loaned me a Chevette core

what do you want ?

[Blackeagle]

Hi BillA,

We have been talking about the flow advantage to be found in the use of a heater core that is single pass vs two pass in cores of the same size, total numbers of tubes ect.

Cores 2-342 & 2-199 in this case. Both are 9.5" X 6.125" X 2" but the 2-342 has the inlet on one end and the outlet diagonally across at the other end. While 2-199 is the more often seen dual pass core with the inlet & outlet at the same end.

In a cooling loop if you alternated using the above two rads with the rest of the loop remaining the same the single pass 2-342 should give a higher system flow due to the lower head loss with the single pass rad. But at the same time the flow increase gained in the use of the single pass rad would in part be lost to the resulting higher velocity in the system lines. But also the loss to increased velocity could be minimized by care taken in locating the single pass rad in the loop so as to shorten the total length of lines.

So the idea we were disscussing was if we purchased and shipped to you the two rads in question, would you be interested in running them in your testing rig to find the amount of gain?

I guess what it amounts to in a way is a extension of phaestus's work on improving the flow of a system by care in choice in fittings and this would be care in choice of rads. A way of optimizing a higher flowing system without high cost when setting one up. And without a super pump.

Interested?

[BillA]

liquid side only ?
sure
send something for the return postage, eh ?

[Since87]

Originally posted by unregistered

what do you want ?

LOL

Data, data, and more data and preferably for free.

But being a bit more reasonable...

I would be interested in seeing your test results for the 2-342 heatercore, but perhaps it would be more generally valuable to modify a Chevette heatercore such that the flow through all the flat tubes was in the same direction. (Outlet on 'bottom'.) The 2-342 is bigger than most would want to deal with.

If you will be setting up to test rads at some point and would be willing to test such a modded heatercore, perhaps I (and maybe other willing contributors) could pay Dodge Viper to pick up a Chevette heatercore and put an additional barb on the bottom, and ship it to you.

(I haven't been paying close attention, but my impression is that DV has a bit of a business adding barbs to cores. I don't know if doing as I described would be crossing a DIY vs commercial line. If so, maybe someone here with the metal working skills could add a barb.)

Since you would be doing the lion's share of the work though, I think a better question is, "What seems reasonable to you?"

[Since87]

I didn't realize the question was 'What do you want tested?'

I'd say liquid side only would be enough to give very useful data.
Particularly in the case of a modded Chevette core if you were going to do the works on deeppow's core.

I already have a core that is 9.5 X 6.125 X 2. So if Black Eagle wants to buy a 2-342 and send it to you, I'll send mine as well.

What's your preference? Modded Chevette is a single test set rather than two additional test sets.

[BillA]

liquid dps are easy enough
I cannot mod deeppow's
send 'em as you want them tested
if dv has a business he should throw me a $20/'his' rad(s)
std models for the return postage

addy on site in sig

[Blackeagle]

As I said before I'll chip in. Matter of fact I'll pay for the 2-342 myself and it's shipping. Anyone willing to go for the 2-199?????

A second point of possible interest would be the use of a rad with 2.5" core vs a 2" core. In this case the 2.5" thick rad would need to be matched to the 2" core on the basis of water volume as close as possible.

The thicker core due to shorter & 25% larger tubes would gain in flow due to the much lower resistence. But would this offer gains when you consider the increased resistence to air flow of the thicker rad? And would you also be reducing the fin surface area at least by a small amount?

BillA's question about air vs liquid rates got me to thinking about this one. But this one will have to wait as that would be a heck of a lot of testing for Bill. Unless BillA wants to do a "update article" for JoeC on the optimum heater cores for best flow and cooling balance.=) Or perhaps JoeK would be interested as it would make a good companion to phaestus's article.

This has become a real interesting thread, and it keeps getting better.

since87,

Reason I offered to pay for the larger rad entirly if we used the bigger ones is I think I'd use it in a system. It will (I think) fit in floor area of the case I want for my next build. The Chevette core is smaller than I want for my use. But they are pretty populer. The 2-199 is the same one JP is useing in his XLR 120 rads he sells with those "Scoop" shrounds he makes. So perhaps we can find a member interested in buying that one, he could get a pre made shroud that way form JP. Any one interested???

[Blackeagle]

So which one(s) of several sizes should we send?

Which one(s) would you prefer working with BillA?

[Since87]

Originally posted by Blackeagle

since87,

Reason I offered to pay for the larger rad entirly if we used the bigger ones is I think I'd use it in a system. It will (I think) fit in floor area of the case I want for my next build. The Chevette core is smaller than I want for my use. But they are pretty populer. The 2-199 is the same one JP is useing in his XLR 120 rads he sells with those "Scoop" shrounds he makes. So perhaps we can find a member interested in buying that one, he could get a pre made shroud that way form JP. Any one interested???

I mentioned somewhere above that I've got a heatercore the same size as the 2-342 but with two 5/8" barbs on top. I don't know the part # of mine, but I'll try to figure out which one it is. In any case I believe it is close enough to be a worthwhile comparison. Since you want the 2-342 anyway, why don't we each just send our cores to Bill.

[BillA]

all the same to me
inlet and outlet should be the same size for most accuracy
(I 'calibrate' the connections in same size pairs)

[Since87]

I'm pretty sure that mine is a 2-274, so it's actually 9.875 long rather than 9.5 long.

Close enough for useful comparison though IMO. Same number of flat tubes.

Bill,

Is connecting directly to the 5/8" tubing of the core a problem for you? I wasn't planning on putting barbs on, but I can if it makes things easier and/or it makes the comparison of the two cores more relevant. I wouldn't want to go smaller than 5/8" barbs though.

BE,

The 2-342 has a 3/4" and a 5/8" connection.

Were you planning on adding barbs?

[Blackeagle]

Cool, I'll stop in at the local auto parts shop tomorow and see about getting that rad. From the way it looks in the pix at Fedco I think the 3/4" is only at the end of the one line. If I cut both those long extended lines off at the first bend I think it will be a 5/8" tube at both ends. Soon as I get it here and mod the fittings I ship it to BillA. I'll post as soon as I have it.

[Blackeagle]

since87,

No barbs unless BillA has to have them. I'll use it with just the 5/8" tubes on it with 5/8" lines for what I have in mind. But if BillA wants barbs I could add some, then remove later.

[BillA]

Cu pipe (WITH hose clamps) will always yield a better ID than a barb
my std connection/hose sizes (ID) are 5/8, 1/2, and 3/8
but the connection resistance (to be deducted) is measured with pairs of the same size
- so mixed connection sizes are extrapolated
no biggie, for sure