Turbulance equals heat transfer? Radiator design discussion.
 

Ever notice the huge difference between a BIX rad and a chevette heater core? Besides the depth and the fin spacing, a huge difference is in the tubes. BIX cores consist of strait tubes whereas heater core's tubes "zig zag" changing from wide to narrow in a symetrical pattern through the length of the tube. Seeing that turbulance aides in heat transfer in water blocks, I was wondering if the same wasn't true for radiators.

I'll post pics highlighting what I mean when I get a chance.

Sure. (BIP, too).
I'd always thought it had to do with them being designed for different air velocities and temp differentials.
I'd assumed they were trying to introduce turbulence to ensure all air molecules had a chance to have some heat transfered to them (same reason they tend to be 2" or 2.5" thick) and that the issue was making sure that the heat transferred to the air did not drop as air velocity increased.
No science behind this - just assumptions from what little I know about airflow.

Might be worth adding passive cooling while we're at it.

it's true for any turbulent flow. there will be increased mass/heat transfer and mixing over a laminar flow. i too assume they were planning for turbulence although i think the zig zags may have been there to increase the surface area. then again, i have not looked at a heater core lately.

What I would like to see is a round radiator with a fan in the middle. Cap one side of the rad and the fan will pull air through the rad through the sides. Being it is round the water will have more force on the outside of the tubes and theoretcally should help heat transfer to the fins.

Anyone curious enough to re-purpose one of the new Zalman heatsinks reviewed by Joe? http://www.overclockers.com/articles1237/


.

That first one is exactly the concept I was thinking of but larger with water running through it instead of the phase change. I think you would have to cap one end so the air is forced through the fins.

interesting idea. although i wonder about the air flow rates along the length. certainly some areas would experience higher velocity air. maybe one could change the fin lengths to some sort of conical shape?

I was thinking a squirl cage type fan used in blowers maybe... Not sure...

IMHO!

The atribute of greater surface area greatly exceeds the attributes of higher turbulance for our applications.

Adding turbulence can only add restriction to the loop, which hurts the flow rate. We all know that performance in our systems is directly related to the flow rate...

Do a search for Turbulators to see how turbulence is added in most heat exchangers.

hey Jaydee, this has actually been made. Have you ever seen a LandCruiser heatercore? i think that's the one i saw. it's was round just like in the other picture of that heatsink but HUGE. i thought about buying it, but i didn't know how i would rig up the fan's to cool it. this heatcore has to be over 7" across. but it just looked really cool. guess i wanted it for the "cool" factor when it comes to liquid cooling. but didn't know what kind of fan i would put inside.

It would be interesting to see a all copper version. We got a buggy at work that used to pull luggage carts at the airport that also has a round one in it. It is 8-9" though. It is about 4" wide. I don't think it would be to difficult to make one out of copper pipe.

I think my question was misunderstood. I wasn't talking about the heat transfer from the radiator to the air. I was talking about the heat transfer from the water to the surfaces of the radiator.

JD: sweet idea for a rad. You could have a blower fan mounted on the outside of the case and the radiator mounted inside the case. The air would flow from the inside of the case, through the rad, and then be blown out by a blower fan.

That is a crude design, I'm sketching something out right now that includes cooling for the hard drives. I'll share it when I'm done :)