The Plan
FA20 meets FA24
The goals for the BRZ are simple; take what worked on my fallen FRS, and address every compromise made. Because as sorted as the FRS was, anybody with a working understanding of how to build a track car could see that it was deeply compromised.
That’s largely because the car was built to break the Ontario Time Attack rulebook - which rewards stock-ish builds. That led to decisions like running no-power mods. Stock header, stock tune, and a measly 197hp. Because even running E85 ultimately proved to be a disadvantage when points were totaled.
Then came the car’s compromised aero setup - Nine Lives Racing rear wing, and no front aero. Sometimes it worked. Other times it just created a soul-crushing amount of understeer. And at 2800lbs, it was surprisingly heavy for a car with so few creature comforts: the result of a steel rollbar and a general lack of focus on weight reduction.
As much as it pains me to admit this, the lap records I set had less to do with sheer pace, and more to do with fresh, super sticky tires - A052 Yokohamas, a total disregard for mechanical sympathy, and my own mortality.
But that’s not to say there won’t be constraints - and the first is big. The BRZ is still going to be a true street car. A truck and a trailer aren’t in the budget, nor is that my style. There’s a beautiful simplicity in tossing a bag of tools in your trunk and heading to the track after work for an evening of lapping. Better yet, there’s no greater feeling than beating up on trailered race cars in your mildly prepped street car.
All of that is to say, it can’t be deafeningly loud, and it needs airbags - which is more than I can say about the BRZ in its current state.
Power
I’ve never been able to justify big horsepower. That’s not to say I can’t appreciate it, every high-horsepower car I’ve driven has left me giggling like a schoolgirl. It’s because there’s always a sacrifice; usually reliability, and I put a lot of track miles on my cars. Adding forced induction to a naturally aspirated car adds heat, complexity, and stress to engine internals that were never designed for it.
Of course, there are aftermarket solutions to almost any problem; intricate radiator and intercooler setups, oil coolers, forged internals, and closed-deck engine blocks. You can pay a company like iAG to build you an FA20 that can support 600whp, and pair it with a Quaife sequential gearbox and 1000hp-rated axles. It’ll be cool as hell, but it will never be as reliable as an equivalent lightly tuned car, or a car that makes as much horsepower from the factory (Corvette anybody?)
For those who undergo that journey regardless, I applaud you. Those folks don’t care. They’re motivated by the challenge that running a car like that on track introduces, and they have sponsors and deep pockets. Built not bought, so the saying goes. I live for a challenge, but I also need a car capable of surviving a road trip across the continental US for a Gridlife event, and the drive home, without breaking a sweat.
Which brings me to engine swaps. You can easily talk yourself into thinking that they are the Goldilocks solution to horsepower. After all - if there’s an argument to be made for using OE parts, why not just drop in a different OE’s engine? That’s a trap. The problem lies with everything needed to get said OE engine running. Aftermarket wiring harnesses, engine mounts, oil and coolant lines, bell housing, and flywheel adapters, just to name a few. All things that are critical to keeping said OE engine alive. These are niche, low-volume parts that seldom see the level of testing that an OE would undergo.
This is where I ran into problems with my 2.4L Ecotec-swapped Miata. My exhaust fell apart, the wiring failed, the aftermarket oil pan leaked, the tune had issues; a cataclysm of failures around the engine.
The only real way to do an “OE” engine swap is to use an engine from the same manufacturer, that comes in the same (or similar) chassis. Think K24 swap in an 8th Gen Civic Si or a 20-valve 4AGE swap in a base model AE86; cases where at most, mild tweaks are needed. That’s where the FA24 comes in.
FA20 meets FA24
The second-generation BRZ somehow managed to be more polarizing than the first. Journalists who lambasted the first-generation car for being gutless seemed to agree that the second-generation car finally had a powertrain worthy of its chassis. That had less to do with the number in the brochure, and more to do with the fact that the FA24 seemed clinically underrated. Dyno testing showed factory cars making 210-215hp to the rear wheels, suggesting the actual crank horsepower rating is closer to 240hp - significantly more than the 228hp promised. More importantly, it made peak torque at 3,700 rpm, 3,100 rpm earlier than the first-generation car. Subaru did their homework this time around.
So it’s no surprise that first-generation owners have been posting about swapping FA24s into their cars as early as 2021, but very few have actually accomplished it.
As far as I can tell, there are currently only two ways to get an FA24 running in a Gen 1 car. You can drop in an FA24 and get it running via a custom engine harness and standalone ECU, but the cost of doing so - especially given that the cost of a running FA24 is well above $7000 as of this writing, would be enormous.
The budget option is to combine an FA24 short block with everything else from an FA20 (oil pan, heads, front cover - everything), and run it using a factory wiring harness and ECU. That comes with other benefits; custom wiring harnesses introduce a lot of uncertainty, and many places require OE ECUs to pass emissions testing. There’s also a reliability argument. Instrumented testing has demonstrated that the FA20 is significantly better at managing oil pressure under lateral load than the FA24. Using an FA20 oil pan and front cover means using an FA20 oil pump and oiling system.
Everybody immediately assumed that an FA20/24 hybrid build was the new Honda K20/K24: a build combining the best of two engines to produce something that is truly motorsport grade. That’s not the case here. There’s no evidence that the FA20 heads flow better than the FA24. Subaru did significant work revising these heads for improved flow. And unlike the K24, the FA24 requires - according to the little information out there about this swap - significant machining to make this work.
Precisely what needs to be done was still a mystery to me - as it is to almost everybody. Only three people have done it - and only one in North America; ASMotorsports - the legends behind many of the S2000s and FRSs that dominate Gridlife’s GTLC series. I studied the few posts that ASM made about their swap religiously, and reached out to ask some clarifying questions while being mindful that I’m arguably asking for trade secrets.
I learned the following: the reasons the FA20 heads need to be machined isn’t to keep compression in check, or even to optimize flow - it’s for piston clearance. Thanks to their significant differences in bore and high compression ratio, the pistons of the FA24 contact the FA20 head.
This begs the question; why not use FA24 heads? I couldn’t confirm this myself, but according to ASM, using FA24 heads only creates other compatibility issues. I can only assume that the FA20 front cover isn’t compatible with FA24 heads, nor are the FA20 camshaft position sensors.
Given the complexity already involved, I want to keep this as OE possible - meaning keeping the bottom-end factory. That said, since the heads will require machining anyway, they’ll be paired with larger intake valves and upgraded valve springs to allow for more RPM.
The goal: 230whp, and a slightly higher redline of 8000rpm. Because who needs a k-swap?