Cooling Upgrades - which radiator? DD with future turbo
 

So having finished my first major upgrade on the stock 2002 car, a FM happy meal clutch kit, I am looking at the next step - cooling. I will likely be doing engine and maybe diff mounts this weekend (depending on how much of a pain the engine mounts are), but since my stock radiator is starting to turn brown, and since I manage to get the car nice and hot on my 60+ mile commute, and since I may go turbo at some point, I'm looking for major cooling system improvements in the near future.

I've tried to do a fair amount of reading to catch up on things, and at this point it seems like I should most certainly do:

• reroute
• airflow improvements such as sealing off openings around front of radiator (no hood vents for now, it's a DD with maybe some autocross in its future)
• new/better radiator
• maybe new shroud/fan(s)?

Ideally, I'd like things to be as plug-and-play as possible in terms of radiator, fan(s), and shroud. I've got AC and I intend to keep it. No sustained track use in this car's future, 90%+ of its time will be spent cruising at 80mph on highways.
36mm Koyo hypercore? Supermiata? FM crossflow?
From what I've read so far, it seems that the lowest hanging fruit lies in reroute and airflow improvements, so it seems I should be willing to compromise a bit on radiator performance/efficiency for my mild intended usage if that makes things a bit more easy to install, tidy, and intercooler-piping friendly.
Thanks!

Stock radiator until the budget allows for a Supermiata crossflow or the Trackspeed radiator.

He said the stock rad is turning brown. That's when they start to crack. He needs something now.

The rads you're suggesting are total overkill for a street car which is what OP has.

I'm with schroedinger on the rad choice. I did a crap ton of research before deciding not to do a reroute on my car which is 100% street only.

Supermiata crossflow. Easily one of the best radiators available today, both in build quality and cooling capacity, and not profanely expensive.

Hell, I put one on my street car after I got fed up with shitty aftermarket radiators cracking.

Budget for a radiator like Supermiata is not an issue in this case. Would rather buy one now that will be good for a turbo setup with occasional autocross later on down the road, and long highway commutes with AC blasting in 100F summer weather with strong winds. I don't need an all out track setup, but as this is my daily commuter car, I would rather be a little overly cautious and ready for longer-term power-adding upgrades. Based on the throttle usage on the driveto work currently, if I go with a turbo and taller diff ratio later on, I imagine I'll be close to low- or zero-vaccuum highway cruising with the turbo doing a little bit of work.

I'm not terribly worried about my current radiator blowing up in the next couple days/weeks, it's much closer to black than light brown/yellow, and it has been holding up for 65 mile drives home from work in 80+ degree weather (at ~6000' altitude for what that is worth). I will be doing reroute and radiator sometime in the next 2-3 weeks I imagine.

For low power, KoyoRad 36mm Hypercore, or the lower end unit that Supermiata has. Emilio sells it, so that means he endorses it.

If budget is not an issue Supermiata Crossflow is the ticket especially because you want to boost in the future. That's what I run and it works great at our elevation using AC and boosting through the mountains.

Emilio has specifically stated, the Supermiata Crossflow IS NOT for a turbo street car. It was designed for a shrouded, ducted, racecar without air conditioning. Street car with a turbo, air conditioning, and little/no/oem ducting he recmmended the Trackspeed radiator.

Well this is what it says on the website: "Suitable for any power output from stock to the most heavily taxed 300whp turbo track Miatas running factory A/C."

Was just about to say: TSE and never have to touch the radiator ever again. Ever

Im still rocking the csf radiator! Turbo street ducting and no a/c

How timely of me to start this thread today. Noticed a strong gasoline smell in the garage a couple hours after I got home, while looking for a leak I found a nice puddle of coolant and an empty overflow tank... Seems like it may have just boiled over. I had been keeping an eye on the temp gauge all the way home (it was pretty warm out). Gauge stayed slightly below the halfway mark, but last week my OBD scan tool told me the coolant temp was 108C when the gauge was in the same spot, so I suppose it's not out of the question that it got a little too hot. Still gotta figure out the gas smell though...

Is the bottom of the fuel tank shielded from your exhaust? It is on the factory exhaust but often forgotten by first timers adding aftermarket exhausts. I actually heard the fuel boiling without my shielding and the tank was pushing fumes out.

Uh, no. Never made that explicit statement. If quoting me, I prefer cut/paste of actual stuff I wrote.

The Supermiata Crossflow (32mm) does just fine on street turbo cars with A/C. Where larger capacity cores help is in cars with inadequate ducting and airflow for stop and go driving. A larger core is bit of a band-aid in this scenario. Even with a huge 50-70mm crossflow, if you have bad ducting and venting, your car will overheat once moving. We have seen this and that's the reason we built our Crossflow to 32mm. When everything is right, you don't need band-aids. Bullet, running 30 minute w2w races at 400whp in 90° at 85% humidity never strayed above 210° clt with our Crossflow.

I'll have to double check - I'm not sure if anything is missing in my car, this being my first Miata - I'll get some pictures tomorrow. That said, it's a stock exhaust setup. I think the only aftermarket stuff on the car is the clutch I just installed.
The gas smell dissipated and did not return this morning, so I'm guessing it was heat-related. I'll start with simple/obvious/easy stuff like gas cap (it takes a bunch of torque and two hands to get it to click), and I'll inspect the fill neck and see if I can get the canister out.

This is the type of approach I like! Your stickied threads are great reads.
Does your crossflow radiator require much trimming to fit a stock car? I'm guessing I'll need to move the condenser forward. If you've got a link to install instructions that would be super helpful!

The link is on the product page.

http://949racing.com/supermiata/CF_R...uctions_G2.pdf

Excellent, thank you.

I installed this radiator on my stock '04 last year, with aircon and stock fan shrouds.

Both fan shrouds had to be trimmed a bit to fit (it's obvious once they're laying in place,), and the condenser mount has to be moved slightly forward, for which hardware is included. The instructions cover this, and it's important to read them carefully, as I missed a step the first time around and was wondering why I had some rubbing. After I re-read the instructions, it worked perfectly.

Here's a photo of mine from just before I installed it back into the car. I've highlighted in red the cuts I made to the OEM fan shrouds.
https://cimg7.ibsrv.net/gimg/www.mia...a09d7f5328.png

Don't forget the most important part:

I got your back, fam.

FWIW, I ran a $90 ebay radiator for a whole year at 350 whp with AC and no ducting with ZERO overheating. I never saw more than 210* even at the drag strip running mid 11's at 121 mph.

Hell my race car runs the same rad with only the OEM bumper duct and no undertray and no reroute and I have never seen it hot either on track.

No reason to overthink this...

now did you cut/past or copy/paste?

Maybe I used the wrong term, but point is to focus on how much and how fast coolant is coming into contact with rad core and in turn, how much air mass is coming into contact with that core.

conduction

[kuh n-duhk-shuh n]ExamplesWord Origin
See more synonyms for conduction on Thesaurus.comnoun

1. the act of conducting, as of water through a pipe.
2. Physics.
   1. the transfer of heat between two parts of a stationary system, caused by a temperature differencebetween the parts.
   2. transmission through a conductor.
   3. conductivity.
3. Physiology. the carrying of sound waves, electrons, heat, or nerve impulses by a nerve or other tissue.

Thanks for the input, everyone. Ordered the reroute kit and the Supermiata crossflow earlier today.

Good choice. You will be very happy with that purchase. Everything Emilio sells is top notch stuff and he stands by his product.

OK fine. Let's get technical then. Forced convection is what we are talking about since the fluids (air and water) are being forced through the radiator (water pump for the water and a fan or vehicle movement for the air).

That said, at the point where the heat is being transferred from the water to the aluminum to the air, it's conduction. Surface area is what counts.

It's after the fact and I agree with OPs choice. I've got a Koyo Hyper on the turbo daily, and Emilio's crossflow on the track car. The big benefit of the Koyo is that it is completely plug-and-play and cools well. I got it before Emilio's crossflow was available.

But Emilio's crossflow is fantastic. I saw a 15-20 degree drop in on track temperatures vs. a Ron Davis (itself considered a decent radiator). It may even be better than that because I always end up at the thermostat temperature, so there is still excess cooling even when air temperatures exceed 100F. It's also nicely configured for a reroute hose. Only downside is minor fab required as Joe pointed out. Very minor.

Also, as OP discovered, the OEM temperature gauge lies on purpose. You can get up to 215 to 220 before it starts budging from the middle. This prevents warranty returns. If you ever see that gauge indicate hot, it means it!

​​​​​​Also, since it seems like we're being "that guy", that first definition is both poorly worded and a rather confusing choice as, by definition, water flowing through a pipe would be considered forced convection, not conduction, in the world of heat transfer.

That definition was just an online dictionary/thesaurus. It wasn't a technical definition. I think we all get the gist.

You said it not me! (*inset We Didn't Start the Fire soundtrack*)

To cover my ass I'll start by saying that the, by far, most useful data in regards to application specific effectiveness is the empirical data you and Emilio so kindly provided. That being said, the ocd-elf inside me can't not mention the incorrect and/or ill-explained physics in your last post.

The only significant conductive heat transfer (HT) occurring in a radiator is in the path heat must take to go from the inner wall of the (most likely) aluminum tubes to the outer wall/fins (of the same tube). In terms of circuit analogs, the resistance to the transfer of heat in the conduction through the the relatively thin aluminum is many orders of magnitude less than the resistance to HT which is inherent in both the coolant-tube interaction and the tube-air interaction (although in most cases it is the tube-air HT which plays the limiting role in the cooling power of a radiator) for any given temperature delta. This is why everyone makes such a big deal out of construction type and presence of turbulators (or any feature designed to trip free-stream air flow into turbulent reynolds number regimes for enhanced convective HT).

All heat transfer scales with area (conduction, convection , or radiation), this is just how the physical phenomena are described from first principles (e.g. you need to solve surface integrals to find heat transfer of a 2D or 3D object). The important consideration is really, which HT surface area do I maximize? The answer to this question, give the relative heat capacity and density of water vs. air and thin film effects, is almost always the fins/tubes which allow for convective HT between radiator and air (and, if not, the answer is usually the other convective process between coolant and radiator)

Source:
this paper from 1930 about radiators Sci-Hub | Heat transfer in automobile radiators of the tubular type | 10.1016/0735-1933(85)90003-X
this paper on copper nanoparticle doped coolant in radiators Sci-Hub | Experimental Study of Heat Transfer of a Car Radiator with CuO/Ethylene Glycol-Water as a Coolant | 10.1080/01932691.2013.805301
2.5 degrees in mechanical engineering focused on a completely unrelated area.

Where do I buy the nanoparticle coolant?!?!?!
Also, nice explanation :) as I read it I was thinking "huh this person definitely has at least 1 more mechanical engineering degree than me!"

FTFY.

It's probably only good for internet arguments but gotta justify those loans somehow.

Also be warned it's a little pricey: https://www.sigmaaldrich.com/catalog...g=en&region=US

all this talk of cooling and no one even mentioned the hyper re-route

Especially since Spiff mentioned an elf.

Why is this conducting, convecting thing called a Radiator?

There really only are two radiator options from what I see and have experienced. Supermiata CSF Crossflow for light and oem fit or TSE for biggest you can fit. I am really fond of the Supermeowta Crossflow. It is fucking light, looks high quality, fits like oem and kept my msm with undertray and stock fans at normal temps. But now I am adding power I switched to the TSE. Well the price is fairly spread apart between the two $619 and $402, with the TSE needing a aftermarket fan in some cases like mine.. Both are equally baller. But I think they both have their place. The Supermiata Radiator is nearly as small as OEM as far as space it takes up and stock fans bolt up, while the TSE takes up alot more room and will only allow one single oem fan(not w/ RB Sway brace) or slim spal. The supermiata has two drain plugs, which at first seemed handy, but in reality it is two more leak points, where as the TSE has no drain plug(kinda a pain) But no worries to be had. My Supermiata Rad. leaked from both drain pliugs just after a couple drains due to the oring failing, the plugs got to the point no o-ring would seal and I opted for a crush washer, which didn't work. I ended up using a high tack sealant to permanently seal the plugs as I could not get them to stop a slight constant drip. When I opened the box for my TSE radiator I was immediately in shock as to how fucking big it is...lol It is massive, and weighs a bit more than the super. But there is no doubt in my mind I upgraded further from the super. I now have one single 14" spal, no drain plugs to leak, larger cap and throat with a sacrificial anode.If you need room for stock or shrouded fans, IC tubing or want the lightest weight baller radiator, go Super. If you want the extra margin of cooling efficiency and don't worry about space, oem fans(ac) or weight. Go TSE

https://cimg8.ibsrv.net/gimg/www.mia...803c295fc6.png

Reroute kit was out of stock on the Supermiata website when I first placed my order, so I went through FM, not realizing they were backordered too. By the time I realized it, Supermiata had them back in stock, so I ordered through there and got it in two days (thanks Emilio!).
I was getting some boiling a couple weeks ago, so I topped off the coolant and was sure to burp the system really well. I also did a simple length of aluminum tape on the top front edge of the radiator. On the hottest days I still had a little boiling happening, but just a few drops. I have a 1.1bar cap of unknown age, but figured I would just be careful until I got the new radiator and reroute installed.
I hooked up a scan tool this morning to keep an eye on temps. Ambient was ~59F, and doing about 83mph for a 20-minute stretch resulted in ~214F according to the scan tool.
The plan is to get the new radiator and reroute installed tomorrow, and do some ducting at that point to seal up the front of the radiator all the way around. My condenser is not in great shape, and I hit a lot of bugs on my drive to work each day, so it might be easier to just replace it rather than clean it, and try to add some sort of open mesh to mouth the to catch the bigger bugs - but I'll hold off on that for now until I get data for the new radiator/reroute/shrouding. At this point I'm certain the new setup will make a world of difference (especially the ducting - coolant temp seems to rise fairly linearly with speed from what I saw today).

Please remove mystery meat rad cap. Rad caps are a wear item. More than once, I have seen expensive engines done in by failed caps. Cap has two seals. The secondary seal (outer) is the important one. It allows excess pressure to push coolant in to the reservoir. The key is it will allow a vacuum to build in the system which draws coolant in from the overflow tank. If this seal is bad, no vacuum will build when the engine cools down. So on every heat cycle, some coolant is pushed into the overflow res where it evaporates instead of being sucked back into the system. Your coolant level gradually drops until
you have a massive air bubble, which inevitably reaches the head. When it does, it'll pop a head gasket in a few seconds. You'll see the temp gauge spike and 3 sec later a plume of coolant vapor in rear view mirror.

Inspect the seals on the cap regularly. Also make sure your overflow tank is seeing two way flow when hot and cold. Another reason to install a coolant pressure gauge on your track car BTW.

OEM is 1.1 bar. Koyo cap is 1.3 bar.
Our Crossflow ships with a 1.4 bar cap and is also available separately.
https://supermiata.com/CSF-radiator-cap-4540.aspx

I just have to make it home today - next time I drive the car will be with your crossflow radiator and included cap :)

And what is the price for a SuperMiata radiator cap from 949 Racing?
https://cimg0.ibsrv.net/gimg/www.mia...72fdcc5493.png

Okay, new radiator is installed, as well as the reroute kit. It went fairly smoothly, unsurprisingly the trickiest part was working between the firewall and back of the head, but not too terrible. I ended up moving the condenser to the front of its mounting brackets on both sides.

The bleed has been a bit tricky... I cold bled it through the bleed fitting as the manual suggested, until I had a steady stream of water with no bubbles. I burped the upper radiator hose when I did that. I closed it up and then hot bled it using the Lisle funnel suggested in the instructions, with the car level (not nose up). It seemed to keep getting an air pocket in the new thermostat housing, because the gauge on the dash was reading the temp accurately as it warmed up, then it started reading full cold. Radiator got nice and hot, but fan didn't kick on due to the air around the sensor. Even after opening the bleed plug while hot (until dash gauge started showing warm again), it seemed that the air pocket would return to the thermostat housing.

I'm planning to drain the coolant once it cools off (this was just a water fill to flush most of the old coolant out). Should a nose-up fill/cold bleed/hot bleed do the trick this time? Any other suggestions?

Before going crazy over bleeding properly, check that the gauge sensor is grounding properly, as it grounds through its' body to the engine.

I will check that out, but there was definitely air in the housing when I opened the bleed screw hot, and once I got the air out, the gauge came back to life (tried this a couple times).

Okay, good call. I ran a jumper wire from the sensor body to the intake manifold while a helper watched the gauge, and it jumped up. Now to figure out how to ground it permanently...

Not super happy with my grounding strategy, but I think it will hold up okay. I wove some stripped wire through the slots in a hose clamp, then clamped it around the hex fitting of the sensor. Put a spade connector on the other end and routed it to the ground bolt on the throttle body. Did a super thorough bleed with the nose in the air (couple of mild heat cycles).

I did some temporary ducting of the radiator, just the top and bottom sides, using aluminum tape. Started making templates for a permanent solution with cardboard, but I need to go buy some aluminum or plastic or something for the permanent pieces.

I live on top of a ~500ft tall hill, as I left the driveway, temp read ~165. By the time I coasted to the bottom of the hill, temp dropped to 150 lol. Drove about 3 miles at 70mph, temp got up to ~190 and seemed to stabilize. Going back up the hill at 25mph (5-12% gradient for 1.5 miles), got to 195, and then idling in the driveway it gradually climbes to 205. Then the fan turned on, and dropped it back down to 197 before it turned off. No signs of leakage so far, and the system built up a little pressure. We'll see how it does on the commute tomorrow.

NB1&2 gauge sensor is a 3 wire. Power, signal ground. Grounded through the OEM harness. Consult FSM wiring diagram or poke it with a VOM if you have doubts.

If you have anomalous gauge readings, it has nothing to do with the Qmax.

Yeah I just checked the factory service manual, I see that it grounds to a signal ground pin on the ECU ("4O"). I'll do some investigating tomorrow and try to see what's going on, but I think my bandaid fix will get me to work tomorrow.
The reroute and radiator seem to be of fantastic quality, and fitment is pretty darn good. If you want any feedback, I noticed a couple of things that might be helpful to add to the instructions (a few steps I took weren't metioned in the instructions, but it would have been next to impossible to do the install without it - e.g. removing the EGR tube was required to get the old housing clear of the stud).

The gauge half of the NB CLT sensor is not grounded through the harness. It's grounded through the body of the sensor, as HarryB said.

Ah, that makes sense now. I'm guessing that the sealant and/or anodizing is responsible for the insulation of the sensor body. The easy solution probably would have been to sand down the anodized coating from the sensor area of the housing and under a bolt head, but I don't feel like draining and pulling it out...

Anyway, car is running cooler for sure. Last drive to work (Friday), I was at 212-215F with 59F ambient temp. Today with the radiator and reroute, and only top surface sealed off (bottom blew away), 190-192F with 62F ambient. It held dead steady at those temps for 50+ miles at 80-83mph. It should be 92F ambient and sunny this afternoon when I head home, so that will be the true test. I'm sure with halfway decent ducting, I'll never have an issue. Very happy.

EGR tube is too cozy with heater hose, popped a small hole in the heater hose this morning. The straight barb fitting for the heater hose points it towards EGR tube, and there's not much room to route the heater hose differently. Perhaps a 45deg hose barb would do the trick.

https://cimg2.ibsrv.net/gimg/www.mia...3d5e1bb18b.jpg

EGR delete anyone?

Or bend the tube. Aren't they fairly ductile?

I vote mallet.

Rubber doesn't like that kind of heat, also.