Because I have to say it as a cooling system design guy.... My vote is air-to-oil

given your constraints, I'd pick B. Having the least number of failure points is what I'd focus on for track duties. It does suck that you're further taxing the cooling system by adding oil heat into it, which is why I still would push for a2o, but I won't fight you on it lol.. thermostats are generally quite robust devices, so I think you are overthinking serviceability needs of them in your A layout, further reinforcing why I'm picking B.

 

The more I modify my car, the more I respect the well balanced compromise the car is as delivered.

I know you said this was mostly a track focused car that happened to be registered, but my vote is for no external oil lines. I have an oil temp gauge and I've never seen it get above 260° measured at the sandwich plate, or in the sump, sven after extended track time at a power track like VIR. That should be well within the safe range for a good synthetic oil. I don't see the need for an oil cooler unless forced induction is in the future of this engine build. It seems like a lot of cost, complication, leak points, weight, and all sorts of other downsides for minimal benefit.

If you can find a way to maximize the effectiveness of a stock style donut cooler, do that. The Mazdaspeed unit sounds great, or maybe even a heavy duty truck that uses larger coolant feed hoses.

For a complete tear up, I'm a fan of Option A. It seems people write off remote mount thermostats because they only think of it in the context of bleed hole style housings and ignore that you can do the heater core bypass off of it.

 

Ahh well ... that car is no more, parted out c2018.

It was placed there so as to get the coolest CLT, also the most stable. My target was to get OT to 80c as quickly as possible on a cold soaked car yo ambient temps of c0-5c. I was not running a reroute, so there did not seem to be any other option. I considered bypass flows to be little better than the flows to the factory toy heat exchanger, that really only left the rad outlet. While it is coolest CLT, while the car is warming up in the pits, the outlet temp is climbing as soon as the thermostat opens, heating the oil - no thermostat in the oil system. I could sit in the car and watch OT climb to 80c (minimum for hard use) not long after CLT reached a stable temp, probably around 90c idling. Job done.

It seemed to work as a cooler when on track, I have no records, but I know I never had my high OT alarm activate - highest ambients would have been high 30s-low 40s. It was the mid size single core, I used AN 12 for the 180* to minimise flow drop, again I had no indication OP was lower than expected, otherwise all AN10.

Likewise, no overheats, running an ebay alloy rad.

Only other photo. Car had no AC, but has PS.

 

This is the feedback I needed as a reminder, thank you! Did a quick mental check... 3 vehicles in the fleet with combined 300,000+ miles on them. I've replaced a GM thermostat on my Colorado (but that's GM cheapness, I'm not counting that) and I've replaced a non-dead Miata thermostat once to go to a reroute. Considering the added 100,000's of miles of my previous vehicles with non-dead thermostats and you're absolutely right.

I am very interested in your tilted radiator setup you previously had on the K24 car (which is now LS?). I will be ducting whatever I end up with, but tilting the radiator forwards is another one of "those things" that are interesting to me. Previously I've found some interference between big sway bars and fans when a thick radiator is used in stock location. Lowering/tilting it seems to be a good way to both improve the function when ducted and make use of free space up there. I still have the crash structure, despite replacing the plastic bumper beam with a Doriyoku "crash bar", so it may get interesting if I try to do too much up there.

Of note: I'm being picky on wording here because SCCA TTB put out a clarification on Max class last year that we aren't supposed to remove safety crash structure like many Miata's do with cutting out the big beam in front of the middle of the radiator to make space for ducting. TTB comments are mostly a "spirit of the rules" thing and that wording is not yet explicitly in the rules, so I'm more or less ignoring it by calling the chonky plastic thing a bumper beam. Also I already stitch welded some of the parts which are definitely part of the crumple zones so... SCCA is just going to have to deal with this one for now. That said, cutting out all of that is probably too much to talk my way around.

• Cost; unfortunately yes. But my other hobbies involves putting a lot of expensive little pieces of lead into the dirt, or endlessly riding around (sometimes crashing) in circles on silly plastic pedal bikes, so... yolo to a point?
• leak points; manageable but mindful of them.
• weight; should be something like +2 lbs all in for plan B considering I removed the factory donut and factory small coolant lines already
• Doing it for the love of the game; priceless. Joking aside, this one is back to point number 3 in my first post. Fluid temps are one of those things I find interesting. That's a big part of why Option A exists at all - because I found it interesting.

Boost isn't in the plan for this car unless something changes I get silly ideas in the future.




Anyway, the more I think it through, the more Option B (which ironically was my original plan. Not sure why I didn't label that as "A", but whatever) is sounding like "good enough overkill" without crossing into "silly overkill". I won't lie, the engine/subframe/accessories is slowly coming together on my workbench and I want it in the damn car so I can work on other things - that may be influencing my opinion here a bit.

 

I totally get overthinking the improvements you make to the car and doing it your own way. I just keep getting burned lately, so my mindset is going even more to the side of keep it factory unless there's a known issue you're fixing, or the upgrade has zero downsides. I was doing a lot of reading on cooling systems over the summer waiting on the machinist to finish my engine. I got burned there too! Anyways, I have my own ideas if you want to optimize and overcomplicate things...

Look into a dual acting, inlet side thermostat. The Honda J series runs one, and I believe the 90's LT1s do as well. It regulates the inlet temperature to the engine, obviously, which is why they're rated to open at around 170 instead of 195. The dual acting element is also a great feature, balancing total flow between the radiator and the bypass. As it reaches wide open, it starts to close off the bypass/heater core flow, forcing all coolant to go through the main radiator where you get vastly better heat transfer. Then you could run a Laminova unit either at the main radiator outlet for 190-200* coolant with the downside of only getting significant coolant flow after the engine warms up, or between the ithermostat and the block for a constant 170*, with the downside of longer lines.

I tried to find a late model car with a dual acting outlet side thermostat with a 195* rating to make my own remote mount style thermostat as you were thinking of for Option A. In the end I went back to the KISS approach with a QMax.

The diagram below doesn't show the regular way a dual acting outlet thermostat would be acting with the engine up to temp in reasonable ambient conditions - open to both radiator and bypass, and balancing flow between them as the coolant temps and engine load dictate. Honda would have a lot more complaints about not having heat in their passenger cars otherwise. But it does a good job of showing the warmup and overheat flow paths. Just pretend "Open (engine at temperature)" actually says "Fully open (cooling system at max, overheating begins)"



Anyways, enough threadjacking, I only enjoy learning about this stuff, Padlock gets paid to do it and he should be your first consultant!

 

I didn't want to recommend the same, but seeing as you did, I'll support you here... Modern synthetic oils are incredibly robust to thermal breakdown. I have routinely tested oils to well beyond numbers that I even want to mention here, but let's just say that if your temp is starting with a "2" then your oil isn't even breaking a sweat. Where you run into problems is viscosity loss as you move up in temp and oil getting thin to the point where journal bearings aren't meeting their minimum film thickness (MFT) requirements. It's a technical detail, but oil temp is more of a bearing specific issue than it is an oil specific issue.

I routinely saw mid to high 200's on my Kswap. I just changed the oil after a couple events as a precaution to small amounts of thermal breakdown and moved on with life. Never felt the need to add a cooler and the new owner of that engine setup is doing the same successfully the last I heard. People will argue the topic, but I see it way too frequently that people are scared of a mid-200 temp. Under worst case conditions like a track, being into the 240F range is completely fine (in my opinion) to ensure moisture (water) boils out of the system (even in the cool spots of the circuit). 280+F is about when many journal bearings start encountering long term durability issues, but I just don't think all that many people are documented to be in that situation. There's a lot of "monkey see, monkey do" in this space. Just because a purpose built track car has an oil cooler, doesn't mean our dual duty (or lower powered track) cars need one. I always like remembering that these systems add weight too. One of the awesome things about miatas is how light they are and keeping them that way by only adding what you need is always in the back of my mind.

Going off of memory here from when I did this a couple years ago, tilting a radiator seems possible without cutting away the crash structure, BUT leaving the crash structure is going to block a lot of the airflow that you'd want the radiator to see... I had a cut crash structure to maximize airflow on my tilted dual pass radiator setup and the car cooled well for me on track.. so.. could your selected radiator mount in place? Yeah, probably. Will it work well if airflow is impacted? You'll likely have to be a guinea pig and test.

 

One last thought before I shut up.

I had an oil cooler for a while on my N/A dual duty car. I had intended to supercharge it, but that never happened, and I eventually installed the cooler anyways because I followed the monkey see, monkey do thought process so common to the car community. The line layout was poorly chosen, with the lines and cooler mounted too high. I believe there was drainback as oil pressure took longer to build after engine start than it did before install and after removal. I am sure that I did more harm than good by having that oil cooler on there, with every start now running the bearings longer with only residual oil, not a pressurized film.

For a low power track car with minimal oil temperature concerns and a focus on increasing bearing life, I'd rather complicate my oiling system with an Accusump to reduce dry starts and g-related pressure dips than add an external cooler to perfectly optimize oil temperature.

 

This is an awesome discussion I didn't really plan to launch into! Yall have given me a good bit to think about. Thanks for that.

This mirrors my suspicions and its why I'm not chomping at the bit to do it right away. In a year or two when I'm itching to work on something over winter? Sure, I'll put it back on the drawing board. First pass will just have a regular old supermiata radiator in the standard orientation.

I am concerned about this with my intended mounting location and this is the single biggest thing which gives me pause. I was thinking of a check valve in the AN oil line, but that couldn't possibly go wrong, right?

I'll give this topic some more time to marinate and think on it. If I can find a giant truck-based donut cooler with 5/8" coolant hookups for my heater core loop, I'll strongly consider that. Y'all are welcome to put your 'tism to use and help me search if you get bored.