Sorry, I got distracted by OkStupid for a couple days. Let’s get back to our previously scheduled content, which was RWD vs. FWD. Today, I did an experiment on Center of Gravity, which I modified from the original RWD 51.5 to FWD at 65, 60, and 55.
I’m doing all of my game launches from Content Manager, so the testing situation has changed a tiny bit. Here are the details.
- Assists, nannies, whatever: everything off
- Time 09:00
- Weather: Clear
- Track: Optimum
- Temperature: 20.0 C
- Wind: 0 km/h
- Ballast: 0 kg
- Restrictor: 0%
- Setup: base with Street 90s (SV) tires
The vehicles are all created by modifying the NA Miata with a build script, which is located at https://github.com/iankorf/acexperiments. The only changes made from RWD to FWD are the following:
- ARBs are swapped front and rear
- Track width is swapped front and rear
- Center of gravity is changed: 55, 60, 65
- Brake bias is changed with CoG: 70, 74, 78
Nothing else is changed. There are several things I could have changed, and you could argue I should have changed.
- Weight was kept the same despite FWD vehicles being lighter
- Power was kept the same despite FWD vehicles having less driveline loss
- Tire pressures and suspension settings were kept the same despite having more weight on the front tires
- Alignment wasn’t changed because I don’t know how to optimize that for FWD vs RWD
The FWD cars had no advantages, and I think you could argue had some disadvantages. A racer would normally tune their suspension and tire pressures to match their weight distribution. This is not exactly a fair fight.
I conducted the test by driving 20 laps in a row at Brands Hatch Indy and then selecting the 10 best for further analysis.
OK, so this came as a bit of a surprise, but FWD can be faster than RWD. When FWD is at 60/40 weight distribution, it is very similar to RWD at 51.5/48.5. But when the FWD balance is 55/45 it is faster than RWD. Also, at 65/35 it is slower than RWD. You might argue that I’m a better FWD driver than RWD and that’s the reason for this. But I don’t think so. In simulators, I spend much more time driving RWD than FWD. The 55/45 FWD wasn’t just a little faster than RWD. The slowest lap I did in the 55/45 was a 1:02.920. This is faster than every RWD lap.
In the graph below:
- Red is RWD (because all Miatas should be red)
- Blue is FWD 65/35 (because my Yaris is blue and is probably ~65/35)
- Black is FWD 60/40
- Green is FWD 55/45
The difficulty in driving 65/35 is that it understeers and has low mid-corner speed. To compensate for the lack of grip, I took a slightly different line starting at 2500 feet, sacrificing the exit so I could get well set up for the critical downhill corner. This is why the blue line deviates a bit from the others. What was really surprising was how much faster the FWD cars were in Surtees. I usually think of Graham Hill Bend as the most important corner on the track.
Conclusions and Thoughts
Can you imagine a FWD Miata with a 55/45 balance? It would be faster, lighter, and cheaper than its RWD counterpart. On the other hand, it probably wouldn’t do so great in the 0-60 tests, and nobody in their right mind would buy a FWD roadster. Well, I’m not your usual car enthusiast, so please make mine in British Racing Green.
I’ve actually fantasized about making a FWD for Lemons racing. Not because I thought it would be any better, but because it’s such a stupid idea. Nobody swaps a RWD car to FWD, and the last car you would do that with would be a Miata. Now it’s starting to make some actual sense.
Since my Yaris has such a forward weight distribution, I’m wondering if it would be faster if I added ballast to the rear. In a drag race, it would go slower, but on a skid pad it would go faster. Is a race track more like a drag strip or a skid pad? Certainly it depends on the track. This kind of experiment is easy enough to do in Assetto Corsa. Maybe that’s my next experiment? Check back and find out.
4 thoughts on “RWD vs FWD: Part 3 – Center of Gravity”
In AC, Optimum is way to sticky. I’ve found an initial grip of 90 that never improves provided sim data very close to real life in a Porsche GT4.
Track builders get to choose the level of grip of the track and car builders get to choose the grip of their tires. So it’s not surprising that you need to make modifications to make a sim fit reality. It could also be your driving that is different in sim and reality. You might need to look at data traces in Race Studio Analysis rather than lap times to figure that out.
I use TrackAttack since it can load data from any source I’ve tried. There is a TrackAttack plug-in for AC and I currently use Apex Pro for track data. It is also possible to get TrackAttack data from other drivers although matching cars is difficult. That makes comparisons pretty inconclusive.
TrackAttack is fine except that the free version is so limited now. Can’t see anything but the speed trace IIRC. The great thing about sim driving is that you can make sure everything is exactly the same, and therefore truly compare the drivers.