Aluminum 3 Row Radiator

[mikep]

Yes.
And more fin contact per cross sectional (internal) area, etc.

What is the cheapest proportion house to build? Square. More floor area per wall length.

 

[CaliMeatWagon]

I can't wrap my head around this just yet. If you have 3x 3/8" round tubes, squishing them down to oval shouldn't change the surface area in contact with coolant that much. If you have 2x 3/4" tubes, wouldn't that be more surface area in contact with coolant, allowing for more heat transfer? This of course assumes the material, fin density and tube spacing is comparable.

I could be wrong but the way I'm reading it is it's the amount of coolant in the middle of the tube that's not contacting the walls.

I'm thinking of it like cooking a wide. thin steak versus a narrow thick one.

An oval tube would have less distance to the center than a round one

 

[CaliMeatWagon]

Another way to think about it is if you had a 1-inch tube and let's say heat transfer was the most efficient 0.2 in from the wall. You would have a 0.6 in pocket in the middle
But if you were to take that same 1 in tube and flatten it down to where the tallest section was 0.4 in tall then you wouldn't have a pocket in the middle

 

[gsellstr]

But with a larger tube (assuming same shape), not only would you have more flow, but more surface area as well, so more area to absorb heat to transfer to the fins. I agree, there would be a larger center section depending on height of tubes, shape, etc. but wouldn't the notably larger surface area inside the tube counter the slightly larger tube cross-section?

 

[dbarton]

As the tube size increases the cross sectional area increases quadratically (assuming a circle for simplicity). A lower proportion of the coolant is actually in contact with the surface of the tube transferring energy to the air and fins. The volume of coolant flowing through increases, but the actual efficiency of the heat transfer doesn't get better.

I'm just guessing here.

Well I'm not that well versed in thermodynamics and I'm not going to offer guesses. I have some faith in the recommendations that Griffin Radiator gave me when i wanted an overkill radiator. I've purchased two different oversized radiators in the last 14 years from them, both with two rows of 1.25 inch tubes. They offer 1 inch, 1.25 inch and 1.5 inch tubes and they say the cooling capacity increases with the larger tubes.

Ever since I went with Griffin all of the cooling issues I battled over the past 30 years vanished. Just saying. My personal belief is that a huge ONE ROW radiator with a lot of surface area would be the best you can get. But since all radiators must have a compromise, you should choose a reasonably larger surface area and if one row doesn't allow enough coolant volume or the rows can't be 2 or 3 inches wide structurally (because they can't), then you compromise by going up to two rows with at least as much volume.

I paid a lot more for my last Griffin than anyone here is talking about, almost $1000. Overall size 26.5 x 17 x 3 inches with a core size 22 x 17 x 2.68 inches.

If I was shopping today I'd be looking into this from STS for half the price, but since STS doesn't list the tube size, I would still want to know that first.

 

[dbarton]

And for those thinking I can get something in a stock 240 radiator size so much cheaper, 3 rows, all aluminum, etc, so why can't I make that work so I can keep all my money . . .
In 1992 Volvo changed their 2.3 4 cyl non turbo to radiators with a 32% larger core area.
The Turbo got a 56% larger core area.
Did thermodynamics change or did Volvo Corp finally accept that the 240 engineers were right all along when they asked for much larger radiators? Or do we accept that Volvo 240 engineers were not as well educated as 940 engineers?

 

[mikep]

Core area increase is good. More air to interact with fins, more tube and fin in a given core type. But that is different from the tube shape/size discussion.

What we were discussing:
Given a single tube thickness (height of the flattened shape), widening that single tube increases the fluid to tube contact, and volume of fluid in that single tube.

Looking at single, double, triple layer:
that is doubling or tripling the flow of coolant past heat-conductive material for a given area (charts above this post) but with increased air temp (less cooling) for each layer as it goes through.

 

[mikep]

both with two rows of 1.25 inch tubes. They offer 1 inch, 1.25 inch and 1.5 inch tubes and they say the cooling capacity increases with the larger tubes.

Measure the thickness of each tube, and the ID will be smaller. It will be pretty small. Flat, wide tubes, which is good. Lots of contact area.

Regarding total coolant flow, sure. Sometimes you need a lot of coolant flow. It depends on heat output and airflow.
But if your cooling system was a single metal pipe in air from thermostat to water pump inlet, it would have to be super long, and hold s huge amount of coolant to work.
Radiator designs reduce all that to a manageable package.

Compare a single coolant tube (one of many tubes in the radiator) to a single cooling fin on an air-cooled engine.
How do you make it cool better?
More fin area.
How do you cram more fin area in a space? Taller, thinner fins up until they break too easily or until they can’t get air flow or get too heavy or…
Everything in engineering is a compromise.

 

[mikep]

It isn’t any one number. It is the overall package.

Compare radiators to intercoolers.
Uh oh, now we have two problems.
1. Remove heat from the air (as much as the supercharger/turbo jamming air in tightly creates).
2. Move as much air as possible (up to the max needs of the engine).

Radiators have it easy. One mission. Higher relative heat difference from internal to external
Plus, intercoolers are generally air to air, working with a relatively low temp difference (heat transfer increases with an increase in the difference in temp).

 

[dbarton]

Core area increase is good. More air to interact with fins, more tube and fin in a given core type. But that is different from the tube shape/size discussion.

What we were discussing:
Given a single tube thickness (height of the flattened shape), widening that single tube increases the fluid to tube contact, and volume of fluid in that single tube.

Looking at single, double, triple layer:
that is doubling or tripling the flow of coolant past heat-conductive material for a given area (charts above this post) but with increased air temp (less cooling) for each layer as it goes through.

No argument that more rows reduces thermal cooling. That's why I prefer 2 larger rows rather that 3 or 4 small rows. Radiators only come with compromises. A single row with large enough or wide enough tubes is not structurally possible for a radiator with traditional rolled tube and fin construction. It is possible with bar and plate or with heavier extruded tubes, but the problem with those is the material thickness may be double the thickness compared to traditional tubes and that extra thickness always reduces efficiency. So it's back to more compromises.

If we're not going to discuss an increase in core area, then we're not giving the OP the best info. We all know going with a wider radiator is a big step and adds potential cost for a fan and shroud and changes what intercoolers can be used. So then we get stuck again with a standard 240 radiator size and we think making it fatter will work. If all we're going to do is recommend standard 240 size radiators, then where's the improvement going to come from? Are we back to adding more rows?

That Chinese 3-row the OP was asking about is in my opinion a huge mistake. Three rows within only 1 5/8 thickness? Is that three 1/2 inch rows? Good luck with that.

 

[dbarton]

Plus, intercoolers are generally air to air, working with a relatively low temp difference (heat transfer increases with an increase in the difference in temp).

A low temp difference is probably true of intercoolers during low boost driving, but not when higher boost is used.
In my tests years ago with a custom big Spearco IC, I logged 340 degrees F at 15 PSI before the IC and 77 degrees F after the IC at 90 mph while going up a 3 mile grade (70 deg ambient temp). So that's a 263 degree F temp difference. And that's a lot of heat to pass on to the radiator. And I also have an AC condenser. So I'm biased . . . my idea of a good radiator is one that's a lot beefier in size than most people and a stock 240 size is not that..

 

[mikep]

Good point on intercoolers.

And i agree on beefy being better than fragile.
I know a guy who goes through “leaky” mishimotos. I get his old ones, but they never leak for me.

 

[dbarton]

I grabbed this from the Mishimoto site. That's an extruded aluminum tube in the core. It's a nice looking design. It should be really strong because it's a heavier material that rolled tubes. I would think that would not be subject to leaks unless there's a very heavy impact. That's probably the way to go if you want a one tube core. And I think this probably costs a lot more. If this is what he gave you, it's probably pretty good if it's large enough and if it's large enough, that'll probably offset the lesser heat transfer of the thicker aluminum.

I spent some time searching through their site a while back for any existing sizes that'd be good for a 240, but it seems nearly every radiator they offer is 19 inch tall or taller, except a few with wrong inlet/outlet positions. If you have one in mind they offer, I'd like to know about it.

 

[90volvo]

I think this will be a good place to ask the question. What about dual pass and triple pass radiators. I have a dual pass in my 240 now, it seems to cool just fine. I run 180 thermostat and its been hot here in SW MO. In traffic I hit 190 once.

 

[CaliMeatWagon]

Im glad you guys are debating this as it's a lot more insightful than "here's the radiator you want".


And for clarification, this is for 745.

 

[dbarton]

Im glad you guys are debating this as it's a lot more insightful than "here's the radiator you want".


And for clarification, this is for 745.

I had zeroed in on the 17 3/4"×16" dimension you posted for that Chinese radiator, which is about the same size for 240 or 740 up through 1991. And then in 1992 they began with bigger radiators for the 740-940. I'm curious if you've compared the room in your '91 to a '92 in the radiator area and how much they changed it at Volvo.

I've read more than a couple recommendations that say dual pass or triple pass usually requires a higher volume water pump or flow can be too slow. That may be just an opinion. But in my mind, the cooling shouldn't be any different or any better than single pass. It seems to me that a dual pass would be equal to two half-size single pass cores.

 

[CaliMeatWagon]

I had zeroed in on the 17 3/4"×16" dimension you posted for that Chinese radiator, which is about the same size for 240 or 740 up through 1991. And then in 1992 they began with bigger radiators for the 740-940. I'm curious if you've compared the room in your '91 to a '92 in the radiator area and how much they changed it at Volvo.

I've read more that a couple recommendations that say dual pass or triple pass usually requires a higher volume water pump or flow can be too slow. That may be just an opinion. But in my mind, the cooling shouldn't be any different or any better than single pass. It seems to me that a dual pass would be equal to two half-size single pass cores.

I haven't had a chance to compare between the years in person, but I did look on Rock Auto just now and the '92 has a Nissens radiator not listed for a '91.
For '91 you have 65544A and 65520A with a core height of 16.46' and width of 17.72'
For '92 you have 65542A with a core height of 19.61' and width of 23.23'

These are both with the oil cooler.

I wonder if the call for a higher volume pump is to create more flow pressure. Maybe the smaller ones act as a restriction.

 

[gsellstr]

Thankfully Dave, you're in the same boat as me, so you should be able to swap to the 940 style, however that means new IC, condenser, and running an e-fan as the shroud is different. They have those plastic bolt-on blockoff plates in the core support to shrink it down for the smaller radiator, but the support is drilled for the later style usually. If not, a 940 support shouldn't be that hard to find.

 

[2manyturbos]

And for those thinking I can get something in a stock 240 radiator size so much cheaper, 3 rows, all aluminum, etc, so why can't I make that work so I can keep all my money . . .
In 1992 Volvo changed their 2.3 4 cyl non turbo to radiators with a 32% larger core area.
The Turbo got a 56% larger core area.
Did thermodynamics change or did Volvo Corp finally accept that the 240 engineers were right all along when they asked for much larger radiators? Or do we accept that Volvo 240 engineers were not as well educated as 940 engineers?

What I would say happened is the engineers faced the reality that putting a mechanical fan on an engine to cool it is wasteful in many ways. It takes more energy to accelerate it. It takes more energy to constantly drive it. Those larger radiators keep the engine cool by convection alone on cool days. It wasn't that the cars needed more cooling capacity. Volvo went a completely different path in 1992.

 

[Cwazywazy]

What I would say happened is the engineers faced the reality that putting a mechanical fan on an engine to cool it is wasteful in many ways. It takes more energy to accelerate it. It takes more energy to constantly drive it. Those larger radiators keep the engine cool by convection alone on cool days. It wasn't that the cars needed more cooling capacity. Volvo went a completely different path in 1992.

Might as well run a smaller rad if you're gonna have a mechanical fan spinning away all the time I guess? Although those same engineers also gave the D24 cars a wide radiator and clutchless fan. Gotta keep those 80 furious horsepowers ice cold.