-   Going Faster (
-   -   What the Mazda RX-8 should have been (new autocross car build) (

John V 01-07-2018 12:20 PM

What the Mazda RX-8 should have been (new autocross car build)
I'm not 100% sure where this is going to go yet, but for the two or three of you that are interested I'll take a crack at documenting the work to get this thing together.

First of all, I have an irrational love of the rotary engine. I blame Car and Driver. In 1992, they broke the story of the upcoming FD-chassis 3rd gen RX-7, and by all appearances that was going to be a landmark car both for Mazda and for sports car enthusiasts. Given that I was 14 when that magazine landed in my mailbox, I had nary a hope of getting into a third gen anytime soon. But that article also had a little blurb on the second-gen RX-7 and while I had seen these cars on the road I knew next to nothing about them (okay, I knew next to nothing about cars beyond 0-60 times, skidpad g's and braking distances). When I landed a second gen turbo as my first car two years later, I was immediately a member of the Wankel-engine fanclub. That car taught me more than I ever expected about the love and frustration of working on cars.

Fast forward to 2004 and our very own Clyde offered me a ride in his first-year RX-8 at a local autocross. I had actually intended on buying the RX-8 when I got my first real job out of school in 2002, but delay after delay meant I had to postpone that goal, and I took the opportunity to buy one of my other high school dream cars, an E36 M3. I loved the M3, and to be honest my interest in the RX-8 had long-since waned, after the "scandal" of Mazda over-rating the RX-8's naturally aspirated rotary hit the press.

But hopping into that car was a revelation. It turns! It stops! It'll even do both at the same time! The wonders of a low center of gravity, the stiffest chassis on the market, double wishbones with great geometry. I was pretty torn. Clyde and I drove that car together for a few years and I've never forgotten the fun we had. I also never forgot just how phenomenal the chassis was and just how underwhelming the engine was.

I couldn't quite shake the thought of getting one as a street car and putting a real engine in it. LS V8? Seems like a great idea on paper, but it's been done and I'm not sure the character of the engine fits the car. Maybe something that revs a bit faster and puts a little less weight on the nose (mostly it's the weight of the beefy transmission that needs to go behind an LS, but either way it's weight). I had also thought for years about doing an autocross car using an RX-8 as a starting platform. There is one class that seemed to make perfect sense: Street Modified.

For those not familiar with SCCA Solo classing (basically everyone on the planet?) Street Modified is a class for four seat cars where the drivetrain is essentially free. The class was based around one rule: Make as much power as you want, using an engine made by the manufacturer of the chassis. i.e., Mazda power in a Mazda. The rule set obviously consists of more than that, but that's the basic premise. So my first thought two or three years ago was why not drop in a 3rd-gen RX-7 13B? The formula for making (relatively) reliable power (in the 450hp range is generally understood to be required for competitiveness) is known using the 13B-REW. There are kits to get the engine to bolt up to the RX-8 and it would be relatively straightforward to build.

But I got to reading the rulebook and was reminded that in SM, there is a minimum competition weight that is based on drive configuration, induction type and engine type. Long story short, a 13B turbo rotary has to weigh at least 2,700lbs. The same car with a 2.5L four cylinder turbo piston engine can weigh 120lbs less. On top of that, 13B-REW parts are pretty expensive and reliable tuning on a rotary is challenging. My mind was made up. I went hunting for an RX-8.

John V 01-07-2018 12:26 PM

I found this gem on Craigslist on a trip down to North Carolina back in the early fall. I was heading down for a wedding and decided to browse CL to see if anything interesting popped up. It was listed for $2,100 and looked to be in nice shape. Talking to the owner, he bought it thinking it would be a good project car for him but he got in over his head swapping the (dead) engine out of it.

I got him down to $2k even and towed it home. It's actually a really nice car. It's got 140k on the clock and has absolutely zero rust anywhere. Perfect project car. It came with a bunch of speed parts, which I promptly sold and recouped the purchase price of the car.

John V 01-07-2018 12:39 PM

The car came with the dead motor, but it wasn't installed in the car. I sold that for parts pretty quickly and got started on removing all of the rotary-related stuff. The stock 6-speed transmission is not compatible with the Mazda piston engine so I removed that. The engine bay was impressively clean already. The ABS unit is on the opposite side from an MX-5 (the MX-5's relevance to this project will become clear shortly) and I can tell already it's going to interfere with the intake manifold on the four cylinder.

The front subframe and steering rack won't work for me. The RX-8 uses an electrically-assisted rack and it's enormous. The mounts are also set up for a rotary and would take a lot of work to adapt to a piston engine, so I don't think I'll be using it.

John V 01-07-2018 12:44 PM

The rear subframe I'll use, but it had to come out to get solid bushings (the stock bushings are very soft rubber) and all the rear links will be converted to spherical bearings (again, very soft rubber bushings).

This is a view from the rear looking forward, just showing how clean the underside of the car is. I love working on this car - the engineering is so well done compared to the E46 where everything is kind of a pain.

All of the emissions-related stuff is coming out to save weight and reduce complexity, including this surprisingly heavy charcoal canister / gasoline vapor recovery box.

John V 01-07-2018 12:49 PM

I started on the rear subframe first, because I didn't have anything I needed to start working on the engine swap. A fellow competitor sent me a complete set of solid aluminum rear subframe bushings in exchange for me making him some delrin bushings for his car, so I got those installed. Removing the stock bushings was an absolute bear, but pressing in the aluminum ones just required some Acme-screw all-thread and some wheel bearing anvils.

The rear subframe with all the mounting bushings installed. You'll notice that the diff and axles are gone. The stock diff is certainly strong enough to handle the desired power level, but the available ratios are all tailored to the NA rotary and are far too short for a lower-revving piston engine. It's also a really heavy unit, and the axles aren't the strongest, so I will be going a different direction here.

John V 01-07-2018 12:52 PM

So, the rear diff. The stock diff is something like a 4.44:1. For a rotary with a 9,000 RPM redline or whatever that's fine, but the 2.5L MZR (aka Ford Duratec) that I'll be using won't want to rev past 7,000 regularly, so this is way too tall. I was searching for something with around a 3.73:1 ratio to give me 75MPH at 7,000 RPM in second gear. Also, this thing is a pig.

82 lbs!

John V 01-07-2018 12:56 PM

Someone has already figured this out, because LS V8 swaps into RX-8s and RX-7s are becoming more common. The Ford 8.8 IRS is an aluminum-cased differential that has 10+ ratios available from 3.07 up to 4.11:1, and they're cheap and plentiful as they were installed in thousands of Explorers between 2007 and 2010. They also have a ton of different off the shelf options for LSD units, and if I happen to grenade one somewhere in the middle of nowhere (doubtful, but I've broken a lot of stuff I shouldn't have been able to over the years) it won't be hard to find one. I picked one up for $75 from a salvage yard and they even loaded it into my truck for me :)

68lbs. Not a huge difference, but for a much stronger differential it's a nice side benefit. I fully expected it to be heavier than the RX-8 unit.

John V 01-07-2018 01:07 PM

So... I mentioned an MX-5. Interestingly, the third-generation MX-5, the NC chassis, is almost identical to the RX-8 from the firewall forward and from the rear glass backwards. The NC's front and rear subrames will bolt directly into the RX-8, and vice versa. The inboard suspension mounting locations are in exactly the same location between the two cars - the only differences are the engine mounts and the steering rack mounts. The two cars use different uprights, hubs, brakes, and control arms, as the Miata has a narrower wheelbase and less weight.

I figured I'd use this to my advantage. I know that the RX-8's firewall is a different shape than the MX-5 as it's designed for a rotary and not the taller piston engine. The MX-5's 2.0L engine is very similar to the 2.3L and 2.5L versions that were offered in various other Ford and Mazda products, but with far less robust internals. I don't want to use the 2.0L as it's not remotely strong enough to handle the power I'll be throwing at it, but given that it's the only engine in the MZR family that's installed in a rear-wheel drive chassis (excluding the Ford Ranger which used a 2.5L version of the MZR, but with truck-oriented manifolds and tuning) I thought it'd be very handy to have an NC around for parts scavenging.

So I bought one.

And that's where we are now. Today, the goal is to lower the engine, transmission and subframe out of the NC as one unit and start taking measurements. I can already tell that it won't directly bolt into the RX-8 because the engine setback is so extreme on the MX-5. The top of the 2.0L will hit the top of the RX-8's firewall. The even-taller 2.5L that I'll eventually pick up will be even worse. But as a starting point, it'll be helpful to have everything out of a running NC to get going on converting wiring, which will be far from plug 'n play.

To convert a 2.5L engine (out of one of a million Fusion, Mazda 6, Mazda 3 or similar cars) to a rear wheel drive configuration requires many parts off of an NC Miata, so this 2.0L will donate a lot of good stuff. Looking at how it's installed in the MX-5, I think I can easily lower it 2-3 inches and move it 4-5 inches forward to make it fit in the RX-8. Obviously moving the engine forward isn't ideal for handling, and perhaps a rotary would have upset the balance of the chassis less. But I'm not allowed to hack up the firewall, and I really do think this will be the best long-term solution to making this car fast and reliable. So I'm committed to blasphemy... and ending my long love affair with the rotary.

JST 01-07-2018 01:29 PM

Thanks for the write up! Interesting stuff.

Plus there’s a part of me that really likes the idea of you being forced to use Ford parts. ;)

Josh (PA) 01-07-2018 02:14 PM

very interesting. Maybe I didn't comprehend, but what motor are you using and where is it coming from?

All times are GMT -4. The time now is 04:47 PM.

Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.
Forums © 2003-2008, 'Mudgeon Enterprises - Site hosting by AYN & Associates, LLC