Nb miata turbo toy.
Have a new mishimoto drop in radiator for it, and also have a 3 core crossbow afco circle track radiator here that's about the perfect size and layout.
Is it worth the effort to make the crossflow work? On my duster it was, but that was compared to a 90s parts store 2 core radiator. Not apples to apples.
Car will be used for the challenge and Street driving and a couple Auto crosses and basic buffoonery.
Admittedly, I'm no expert, but my understanding is Crossflow > Non-Crossflow. I'd do it.
A decade or so ago, I was running my NB turbo track car at 24 psi of boost, 340 at the wheels, using an all-aluminum, but conventional-style brand name radiator. It was fine on the street and autox, but would routinely overheat at the track. I replaced the radiator with a cross-flow option from Trackspeed (no longer available) and it completely solved the track overheating problems.
So the crossflow probably isn't necessary for the use cases you outline, but OTOH Mishimoto has a pretty bad reputation in the Miata turbo community. Whatever route you go, there are a series of other things you'll want to do as well:
- delete AC if possible (having that extra heat exchanger in front really hurts cooling performance)
- Make sure factory-style undertray is installed to avoid high pressure behind the radiator
- Install (fabricate if necessary) the ducting required to ensure that every molecule of air that comes in through front bumper goes through the radiator rather than around it.
- Hood vents are another option to reduce pressure behind the radiator, but I did not find them necessary on my car.
- Be careful with fan shrouds. They work best for improving low speed cooling, and may actually hurt at high speeds.
In reply to codrus (Forum Supporter) :
One of the main driving factors od un-swapping the v6 car to a turbo bp was being able to get air conditioning and cruise back.
Ac is non negotiable.
Your other factors are taken, amd will be implemented! Hood vents dead last though...
Seems like I'm a prime candidate for needing cross flow on a street car so I can keep ac, correct?
I'm shocked any modern car isnt running a cross flow from the factory.
What is the heat transfer science explanation? At first glance, one would think a certain number of square inches of cooling area, with a similar amount of air flow, in a pressurized system, there should be no difference.
A good crossflow is better than an upright radiator. A crap crossflow is not. The fin density is a big factor. We did a whole bunch of testing a while back on this.
https://help.flyinmiata.com/en_us/whats-cooler-than-cool-B1vyGy3QK
And Miatas after 2005 do run factory crossflows IIRC. Is the NA/NB Miata, introduced in 1989, a "modern" car?
All it really boils down to is how efficiently you can distribute the flow through the radiator. Fundamentally you are correct: a certain square inch-age of radiator cools the same all else being the same.
But it can be difficult to get the flow to evenly distribute through the radiator. Certain configurations are better at it than others.
In reply to BA5 :
The amount of time the fluid spends in the heat exchanger is a factor as well.
Keith Tanner said:
A good crossflow is better than an upright radiator. A crap crossflow is not. The fin density is a big factor. We did a whole bunch of testing a while back on this.
https://help.flyinmiata.com/en_us/whats-cooler-than-cool-B1vyGy3QK
And Miatas after 2005 do run factory crossflows IIRC. Is the NA/NB Miata, introduced in 1989, a "modern" car?
I would say anything designed in the 80s isn't modern.
Keith Tanner said:
In reply to BA5 :
The amount of time the fluid spends in the heat exchanger is a factor as well.
People like to say that more rows/cores makes for less efficiency, because the air heats up as it moves through.
But this is also a huge factor: with more volume in the cooler, you get more cooling for the same coolant flow through the engine because it stays in there longer.
This is beyond the fact that the air doesn't heat up to 210 or whatever as soon as it goes past the first row. That second-third-fourth row is still rejecting heat. And also beyond that more rows/cores means less coolant pressure drop through the radiator, so water pump cavitation or even sucking the lower hose shut is less likely.
More area is always good but depth is also good once you max out area.
Another crossflow vs downflow radiator difference is in cooling system pressure. Downflow setups have the pressure cap on the input tank, crossflow has it on the output tank. So crossflow systems will tend to maintain a more steady system pressure, rather than blowing out of the cap a bit at high revs due to water pump pressure hitting the cap. That venting means that when you bring the revs back down, you now have less pressure in the cylinder head passages to prevent boiling which can contribute to cooling problems as the system will be below cap pressure.
Being able to maintain good cooling at higher coolant temps is why some engines use pretty high pressure caps. 13 - 14 psi is common on a lot of older cars, but some setups have gone a good bit higher. My Jeep runs an 18 psi cap from the factory, and the cap on the BMW system doesn't vent until 30 psi (but that engine is intended to run pretty warm).
Pete. (l33t FS) said:
Keith Tanner said:
In reply to BA5 :
The amount of time the fluid spends in the heat exchanger is a factor as well.
People like to say that more rows/cores makes for less efficiency, because the air heats up as it moves through.
But this is also a huge factor: with more volume in the cooler, you get more cooling for the same coolant flow through the engine because it stays in there longer.
This is beyond the fact that the air doesn't heat up to 210 or whatever as soon as it goes past the first row. That second-third-fourth row is still rejecting heat. And also beyond that more rows/cores means less coolant pressure drop through the radiator, so water pump cavitation or even sucking the lower hose shut is less likely.
More area is always good but depth is also good once you max out area.
Our testing also included different thicknesses of radiators. Core design was more important than core thickness for actual heat rejection. The larger capacity of the bigger radiator did slow the temperature rise. Those fat cores are also draggier so it's harder to get air to move through them.
Keith Tanner said:
A good crossflow is better than an upright radiator. A crap crossflow is not. The fin density is a big factor. We did a whole bunch of testing a while back on this.
https://help.flyinmiata.com/en_us/whats-cooler-than-cool-B1vyGy3QK
And Miatas after 2005 do run factory crossflows IIRC. Is the NA/NB Miata, introduced in 1989, a "modern" car?
These things apparently use the same style radiator.


You can see why I confused the NA for a modern vehicle. My mistake.
Most modern cars use vertical flow radiators. You only see cross flow radiators on pickups.
If I had to guess, it's so they can have more flexibility in mounting and hose placement. With a cross flow radiator you are pretty much stuck with four mounting points, and two hose locations. With a vertical flow you can put the mounts and hoses anywhere on the top and bottom.
In reply to Pete. (l33t FS) :
The move to pressurized expansion tanks is likely part of the reason why manufacturers have become more okay with vertical flow radiators again. Getting the pressure cap off the radiator eliminiates one of their disadvantages.
I guess I never looked that closely at Miata radiators. I think I've only ever removed/installed cross-flow radiators. (edit: I've definitely done both, but the majority were cross flow)
Fin size/density, tube diameter/density, #of rows, etc. all see like would be an interesting optimization problem for a fluids/heat transfer class. I would imagine since manufacturers are making these by the millions they are pretty efficient... from a performance per dollar standpoint, but not necessarily optimized otherwise. From an enthusiast standpoint I think we are more interested in performance per volume or frontal area.