Let's all cool down with a quick question about radiators

[Theviper1987]

Hi I'm looking at getting an aluminium radiator.

First question

1 row, 2 row, 3 row rads. Would a 3 row radiator cool better than a 1 or 2 row?

Second question

When people go on about a 2 core radiator is that the same as a 2 row radiator or does that mean it flows though the radiator twice?

Thanks

[FridgeFreezer]

I'd imagine surface area would matter most, not number of cores/rows. I'll bet someone's written a book about it before now. In an ideal world the manufacturer might even give numbers for the performance of the radiator in BTU/hr or something like that, allowing you to work out what you need.

Is now a good time to mention aluminium is a worse conductor of heat than copper?

[Boydie]

Providing that the rows (matrix) are staggered then yes, a 3 row is far better and more efficient than a single row.

The trick is

1.) The Material, copper is by far the best matrix material and for the fins. Alloy is the next best but corrodes faster and is not as heat transfer efficient. Yes, I know the surface of the alloy is treated the the cut edges, facing the incoming airflow are not ! If you have an alloy finned radiator go to your local air conditioning specialist and buy a couple of cans of aerosol condenser cleaner, it cleans crud off the fins as well as covering them with a protective film to allay the eventual corrosion.

2.) Airflow, The fins should be spaced apart sufficiently to allow adequate airflow through the radiator given that it will have a determined designed airflow resistance.

3.) Headers. Ideally these should be copper but these days plastic is becoming more common. They are still called "headers" even on cross flow radiators when they are at the sides instead of top and bottom.

If you are going to "play radiators" first acquire, beg, borrow or steal a Vane Anemometer and when the engine is running at idle at NORMAL temperature so that the centrifugal fan is operating, and with the radiator grille off take readings of what air flow there is through the radiator.

To take a good average air flow move the anemometer over the face of the radiator so that you get a good average reading given that it will be higher in areas corresponding to the fan blades, lower in the centre of the fan and at the outer edges of the radiator.

Next speed the engine up to normal running speed of say 2500 and take a second set of readings at just under 4000 RPM (300 Tdi) or 5000 rpm V8) the centrifugal fan will begin to cavitate, its fan blade design follows those fan curves.

Fans. On this subject, when the radiator cooling capacities were designed for the particular engine given the different heat generations and coolant flows the fan manufacturer designed the fan blade curves to cause a given amount of air to flow through a set resistance (the radiator and possible A/C condenser) at a given nominal and maximum engine revolution, around 2500 - 4000 rpm. This fan speed calculation is further complicated as you need to take into consideration the ratio of the crank pulley x the fan pulley diameter.

On my series 1 Disco 300Tdi they are almost the same, even so the fan pulley is slightly smaller than the crank pulley so at 2500 rpm the fan was actually spinning closer to 2800 rpm.

My point is this, to replace the centrifugal fan with a electric one is, in my opinion, crazy as most times these factors wont have been taken into consideration and with an electric fan unless to first test it you dont know how much air it's moving and if it can in fact overcome the resistance of your radiator, normally they are over engineered to give far too much airflow so they need a thermostatic switch to control them and the fan is constantly being switching on and off. A silicon filled centrifugal fan will by its very design "feather" to give the required airflow at the required temperature at any given road speed, it's a well thought out and proven design. Trust me, if electric fans were better and more efficient, Land Rover would use them, currently they are only used on East-West engines as there isn't a mechanical alternative. The other thing for consideration is that 90% of aftermarket electric fans dont like to be submerged in water when you're crossing a deep creek or river ford.

Okay, you have taken your air flow readings on your standard radiator, you can now go to a specialist and with those figures he can select a larger capacity core. I went from the standard 300Tdi cross flow 1/4" 3 row matrix to a 1/2" four row matrix core with larger copper headers,- these had to be hand made due the the additional width of the finished core. The net result is that in 45-50*C ambient temperatures in the Australian desert with the standard centrifugal fan my engine has never run over 90*C.

Finally as my system has no leaks I've switched to "Evans waterless coolant" it has a boiling point of over 165*C and at my engine temperature of 90*C the coolant has not expanded so my header tank, radiator and hoses are under no pressure, meaning of course that there is even less possibility of coolant leakage.

[honitonhobbit]

What he said...

[Theviper1987]

Think I got some night time reading

[lo-fi]

Get a nice copper one from European Radiator Services in Sheffield. They made me a lovely 4 core copper rad with brass tanks to my exact dimensions, all for £300. Comes with a nice "made in Sheffield" sticker too. It's made to fit above the front crossmember on my 109, so smaller than the original series rad, but still copes fine with my V8 with a kenlowe. There simply isn't room for the viscous, or I'd have used it.

https://www.google.co.uk/search?q=european+radiators+sheffield&oq=European+rad&aqs=chrome.3.69i57j0l3.8731j0j4&client=ms-android-oneplus&sourceid=chrome-mobile&ie=UTF-8#fpstate=lie

[Snagger]

Sounds like a brilliant upgrade, Boydie, but by "centrifugal fan", do you mean viscous fan? Centrifugal fans are like those used in SIII, Defender and RRC/Discover heater blowers, where as most radiator fans are axial flow. Did you replace the fan with a large centrifugal one? I haven't seen one, but if they exist, I'd be curious to know how well they work compared to conventional axial fans.

[honitonhobbit]

I read it as a viscous fan.. Iain dates back many years to the time of the centrifugal fan though...

[Bowie69]

Viscous fans actually work because of the centrifugal principle, at least that is my understanding of then, as I split one and serviced it with my dad for his camper van, as new Toyota ones were stupid money.

[Snagger]

Ah, so the viscous hub works with centrifugal bits inside?

[Bowie69]

The bi-metal coil in the front of it opens up a valve when hot, and sends fluid to other parts of the hub, which is forced to the outer edge of the hub and there provides drive.

[Theviper1987]

So is core and row different words for the same thing then.

[ToyRoverlander]

Here's some more info on cooling systems, very informative!

http://www.pirate4x4.com/tech/billavista/Cooling/

[Team Idris]

Ta da ! An expert has arrived

Yep, it's a mine field

My take on it as an off-roader and as Senior Design Engineer in a heat exchanger factory that makes everything from race car rads to big 2m square industrial rads 138mm thick is;

I like copper, especially in off-roaders. It appears to be stronger at the header plate to tube joint. (Sorry work). But ally is fine for everything in truth. And freely available.

Metal conductivity between copper and ally is so small by comparison to the hideous ability of air to shift heat it really isn't worth fussing over.

The toughest car rad you can get is Plate-and-bar. Hardly anyone but Armco smacking drifters need it in cars though. It is literally 'throw it down the stairs" tough.

The thicker you go the less air goes through a rad. Let's say you see 200 Pa on a 30mm deep rad. Double the depth and see 400 Pa. The fan can't do 400Pa so air flow drops to a point where the fan can push air. If you have the option you go bigger.

A layer of stationary water or oil or air is on the inside face of the tubes. The flow is highest in the middle. If you go "double pass" the water speed doubles and that layer reduces. But that only works if your pump can shove the flow in the first place.

A good fan cowl is better than anything ! Make that fan work and every little fin in your rad !

My 1972 transit van has no fan cowl and that is the worst case. Best case is a close fan ring and a sealed cowl with a rubber joiner.

So to sum up: Most things you can do to improve performance takes two steps forward and one back.

For example, if you have a rad with open fin you may reduce rad cooling, but you will reduce under bonnet temperatures, which also cools the engine block better.

[Boydie]

In reply,

yes by centrifugal I was referring to the centrifugal action of the silicon inside the AXIAL fan hub that causes it to spin with a variable degree of drive, the greater the temperature playing on the bimetallic spring, the greater the degree of drive. The only downfall of the Land Rover centrifugal hub is that the bimetallic spring is controlled by the temperature of the air flowing over it. On Kenworth and Mack trucks the bimetallic spring is in the water flow a far better and more temperature accurate arrangement. LR simply didn't have the space to do this.

Next:

Core = Is the assembly of the matrix (these are the flattened tubes the coolant flows through) and the fins joined by the top and bottom headers (or sides in a cross flow radiator).

Team Idris; Spot on, although I'd argue about the heat transfer characteristics of copper and aluminum, the former is better, the later is a lot cheaper.

Snagger and Bowie, Yes the hub is operated by silicon as a fluid much the same as transmission fluid in a torque converter in an automatic gearbox.

Hobbit: I'm way, way older than that, I date back to the days of having young African boys from our farm earn an extra shilling in the evenings when my parents were entertaining by operating sheets of flaxen cloth to swing backwards and forwards to move air to cool the guests in the dining room in my parents farm house in Kenya - and that predates any centrifugal or axial fans

[honitonhobbit]

Was a bit worried at the start of that last paragraph...

[Team Idris]

Boydie

If it's a cheap copper unit then ally will outperform it. Ally rads have tubes covered in a thin layer of high silicon content aluminium. The pure aluminium of the tube melts at 600 degrees centigrade thereabouts, but the rolled-in coating melts at 575-580. It is like the whole surface is covered in aluminium solder. Which is exactly how some copper rads were made. You get an aluminium core that becomes one piece of metal.

However, many copper rad cores were dipped horizontally into a solder bath so that only the first 5mm was mechanically joined. The bit in the middle only works by contact and performance drops off with corrosion. Many classic mini rads are like that. I don't know how many copper rads are like that now, but it's an easy way to save on lead solder, so I bet it still happens.

Fin machines are identical.

One thing worth looking out for on both products is a hemmed edge. The 1mm edge is rolled back on itself and that makes it easier to pressure wash without bending fins. Most rads have a hemmed edge.

https://m.youtube.com/watch?v=Pnc9H-lvldk

As an aside, water rads don't have internal tube fins. The water passes heat readily, so no need. Oil coolers and intercoolers do. Those internal fins really boost performance. They easily put 10% on an oil cooler.

[Bowie69]

My parents old Rover Metro had spirally bits inside the rad tubes... is that what you mean, Idris?

[Boydie]

Bowie, the springs or coils are there to cause the water to spin and so all of the flow comes in contact with the surface of the tube inner surface.

Idris, I dont know about the standard radiator, other than taking pressure drop readings through it and air flows as I described above I never bothered to take much notice of its design, all I knew from bitter experience was that it was totally inadequate for Australian weather conditions.

What I can say of the current 4 row radiator which is now some 6 years old is that the fins go from front to back, they are in a vertical plane as you would expect for a two-pass cross-flow radiator. The core, fins and matrix was plain untinned bright copper. It was painted semi matt black once the headers were made and fitted, these consisted of brass plates* at both ends of the matrix core and copper header "cans" soldered onto them. The original oil cooler was utilised. The brass separation plate, brazed to the brass end plate was drilled with a 3mm by-pass/air vent hole.

* I think these end plates were in the vicinity of about 2.0mm. The header "cans" were handmade out of a flat sheet of 18# copper. The finished radiator was around 13mm wider than the original meaning that the original fan could be retained, this was an essential requirement however the air conditioning condensor had to be relocated to the underside of the roof rack. The original socket for the radiator vent plug was reused.

As I recall the total cost of the radiator was about AUD$650.

[Team Idris]

Bowie69- yes the spirally bits inside the tubes is something you only see on big tubes now. (Inch bore) The best dodge is to dimple the tube which is a lot easier and mixes the water just as well. Speed it up every few inches and it turns it.

Boydie- that could be the cuprobraze process then. The copper-brass equivalent of the brazed aluminium process. That will be a very good radiator and well worth the money.

https://en.m.wikipedia.org/wiki/CuproBraze

[hangover]

I nearly bought a cheap arse alloy radiator off ebay but decided to get my rad rodded/ cleaned instead, was approx. 20% blocked so should be flowing well now.

[Anderzander]

Can I ask for a comment on something ?

For my 200 Tdi Series 1 I was wondering if I could get a rad and intercooler built that fitted under the original header tank to the original sizes.

The thinking being that the more efficient modern cores for both could fit in the large depth / space of the original 50's radiator.

Does that sound feasible ?

[Lewis]

I had my rad purpose built by the company that team idris works for, I don't see why they couldn't do what you're asking