Swapping a V8 for a Toyota MR2: Build Your Own Supercar
Please note: This is the first in a series of articles on this topic.
For many years, I have dreamed of owning a mid-engined exotic supercar. Unfortunately, they remained out of my reach financially. I decided that the only way to get one was to build it myself. I’ve always been fascinated by mid-engine sports cars because of their superior handling, braking, and traction over a front-engine sports car, even front-engine cars with a rear transaxle, which have a supposedly optimal weight distribution of 50/ fifty. I like to call them the “dumbbell cars” because their weight distribution is like a dumbbell: heavy at the ends and light in the middle. This is decidedly suboptimal for acceleration, handling, and braking.
Compare this to a mid-engined car, where if you were to make a dumbbell look like the mid-engined car, the weights would slide toward the center. Now, if you want to make the dumbbell or mid-engined car rotate about its vertical axis (called “yaw” in aeronautical terminology), you’ll do it much easier and faster. This is because the traction from the tires does not have to overcome the inertia that a front engine/rear transaxle car would have on each end of the car. The result is that the car will be able to change direction faster and with less tire wear. Peak G-forces will be much higher on a mid-motor because as well, meaning it’s quicker around corners. Rear wheel traction during acceleration is superior as there is more weight on the rear wheels. Strangely, there are plenty of “side effect” advantages to mid-engined cars that the auto press doesn’t mention.
Some examples:
1) The exhaust pipe is usually very short on a mid-engined car (compared to a front-engined car), so the engine has less “pumping losses” to overcome, or resistance to the exhaust going out the exhaust pipe. This means more power. The exhaust system will also be lighter as there is less of it. Dumbbell cars have no advantage here.
2) The rear brakes do a LOT more of stopping vs. a front engine car. When you hit the brakes, the weight is transferred to the front wheels. This means that the rear wheels are unloaded. On front-engine cars, the front brakes do about 80% of the stopping. This is why the disc brakes in the rear took a long time to catch up. They just aren’t needed at the rear. A mid-engined car has a LOT more weight (usually around 55-60%) on the rear wheels. When you hit the brakes, the weight is transferred to the front, so when braking, you can get 50%-60% in front. The dumbbell cars get some of the help the mid-engine car gets, but not as much, because the engine is still in the front and it’s still much heavier than the trans in the rear.
3) the mid-engined car has no driveshaft (unless it’s an AWD car, like R8 or Veyron), so weight is saved here.
Unfortunately, most mid-engine cars are very expensive. Ferrari, Lamborghini, McLaren, Zonda, Koenigsegg, Bugatti, etc. Some of these cars are over a million dollars! Mid-engined cars tend to be harder to work on as well. Changing spark plugs on exotics is a major operation. The McLaren F1 requires dismantling the engine to change the spark plugs!
In the realm of affordable mid-engined sports cars, there are the Pontiac Fieros and the Toyota MR2. In each case, the cars came with 4-cylinder engines. The Fieros also had V6s, but those V6s were seriously underpowered, making a whopping 140 hp. In 1990, Toyota redesigned the MR2 and also improved the power. Base models had 130 hp and the top-end Turbo had 200 hp, which at the time was quite a lot for a car weighing 2700 lbs.
The new bodywork looked very good, very similar to the Ferrari 348 of the time. The build quality was also superior as it was a Toyota after all. I decided to buy a 1993 Toyota MR2 turbo in 2005 with the intention of making a switch from Toyota V6, which had been done by many people up to that point. Around the same time, I discovered that there were some attempts to fit a V8 engine to the earlier MR2 (Generation 1, 1984-1989 body style, or make 1). There were also attempts to install a Toyota/Lexus V8 engine in a MR2 Mark 2. Attempts to convert the V8 to the MR2 Mark 2 were not completed and the owners of the project gave up. The reasons were unclear, but it seemed to stem from the fact that the Toyota V8 was simply too long to fit crosswise in the car, even after severely cutting the car in an attempt to make it fit.
As a mechanical engineer who happens to be a mid-engine sports car nut, I was intrigued by the possibility of fitting a V8 to my MR2 Mark 2. With a powerful V8 engine, the MR2 would transform into a supercar, with supercar performance. . The guys at Fiero have enjoyed swapping out V8 engines on their cars for many years. Fieros have an advantage over MR2s in that the engine bay is wider, allowing for a larger, longer engine, like a V8. Fieros and MR2s have transversely mounted engines. Another advantage the guys at Fiero have had is that the original Getrag transaxle mounts to a Cadillac 4.9L OHV V8 from the late 80’s and early 90’s. The late Cadillac Northstar also mounts without an expensive machined adapter plate to measure.
In late 2007 another V8 was completed by a guy in Europe in a MR2 mark 1 (first generation). The car was incredibly fast and would make cookies in no time. What fun! So I took another hard look at previous attempts to fit a V8 in the MR2 mark 2. What I realized was that they were trying to “keep it in the family” and use a Toyota or Lexus V8. There was really no valid engineering reason to use this power plant. It did not bolt onto any of the MR2 transaxles and was too long. The used Toyota V8 (engine code 1UZ-FE) is approximately 26 inches long from the crankshaft pulley to the rear of the engine or bell interface. This is the critical dimension. Compare this to stock MR2 engines like the 2.0L 3S-GTE turbo engine, which has a critical dimension of 20 inches. This dimension is critical because it fits between the one-piece pseudo-frame rails of the MR2 chassis.
I decided to take a different approach. I started searching the internet for a V8 engine that would fit into the MR2 chassis, preferably uncut, or possibly with only a small amount cut from the MR2 monocoque. My requirements were that it be a V8 with at least 300 horsepower, be available, cost less than $5,000, and be short and narrow enough to fit the MR2. I managed to find one. Audi has an interesting habit of making very short V8 engines. They do this because they want to use their Quattro drivetrain, but at the same time, not compromise on handling too much. Audi seems to prefer longitudinal engine and transmission layouts to transverse ones. The Quattro drivetrain involves a driven front axle, which they had to locate behind the engine. If the motor is too long, it puts too much weight in front of that axle, so they compensate by making a shorter motor. This has the added benefit of allowing Audi to fit this engine into smaller cars that were originally intended to have a 4-cylinder power plant. For my purposes, I found that 1991 through early 2000 Audi V8 engines are about 20.6 inches long in the critical dimension and about 29 inches wide, not including heads or other easily removable items.
I bought a 1997 Audi 4.2L V8 (engine code ABZ) and transaxle and started working on my project. Unfortunately, after much trial and error, I finally decided that the Audi V8 was not suitable for this engine swap. The problem lay in the fact that the engine was always designed to be longitudinal. In my case, with a transverse layout, the correct size shaft had to go along the side of the engine, and Audi didn’t design the engine with that in mind, so there are large portions of the block in the way of that shaft. . The starter, oil cooler/filter, and engine mount are also on the road on that side, however, I did resolve those issues. The nail in the coffin for the Audi was the adapter plate. I determined that the adapter plate required some of the mounting bolts to be located inside the bell housing of the 6-speed transaxle I was using, making it impossible to tighten them. At that point, I decided to change my approach and use a different engine.
Stay tuned for the next articles in this series.