Hey guys. The Audi A1 is available with patch 2.0 now and next set of DLC cars will be available soon, so thought you’d like to see some notes from the physics development of each one. Four great cars from the four rings this time. Very excited for you all to have a go in them.
Audi A1 quattro: Always tough working on these ultra-limited-edition cars as very often data on what was done to build them gets lost to time, but I think we’ve got a winner here. It’s another one of those weird Audi gearboxes with two final drive ratios in a slick 6-speed manual! Suspension for this car is mostly a parts-bin special from other hot Audi (TTS, S3...) cars, so I found good data on springs and some damper plots from a MkV GTI as a baseline and it works great. Engine matches as much detail as is available, weight and balance available from one of those nice Sport Auto Super Tests, and various other details plucked from various magazine reviews. Tuning ranges are close to typical aftermarket parts available for the S3, GTI and others from the same family. Super fun little car. Prefer it to the other small hatches we have by a long shot.
Audi 90 IMSA GTO: Some late-'80s madness here. Suspension geometry sketched up from our 3D model and photo reference; nothing too fancy going on. Gearbox is a mix of parts from the Group B WRC cars that came before it. Lovely little engine (coolest sounds ever) with little turbo lag and good for 720Nm / 720hp. Real car could be set up anywhere from 50:50 to 30:70 on the AWD torque split and I've used our hybrid system to land it somewhere in the middle. I know some people feel icky about doing it this way, but it really is a nice, effective simulation of AWD torque biasing that we can do now before spending the time on code for a fully adjustable, three-differential physics model. Aero is not lovely; typical mid/late-80s stuff here. While Group C cars were getting into underbody tunnels and crazy downforce numbers, the airflow management in GT was still quite crude and, at best, gave very modest downforce numbers. Means the car is very quick, but can feel a little unbalanced in performance. Not great in cornering, but it sure grips up and rockets out of the exits as few others can. Matches up quite well with the Trans-Am Mustang, or new GT3 stuff even. Huge power, light weight and super wide tires can do a lot for you.
Audi R8 LMP900: Nothing too fancy about this one, and maybe that’s why it was so dominant over its lifetime. Simple, effective suspension; aero package with strong downforce at expense of slightly higher drag; powerful engine with little to no turbo-lag. Can run fast laps all day long in this without feeling stressed. No traction control allowed when this car ran back in 2000-2006, so it’s all up to you to manage that 625hp. Great competitor for the Bentley Speed 8. Will have the measure of the BMW V12 LMR at most tracks - there was big progress in LMP900 performance from ‘99 to 2002 - but it could be a fun matchup with the right drivers.
Audi R18 e-tron quattro: This was a particularly exciting car to work on after it won Le Mans last year. The new hybrid regulations provide some very cool engineering challenges and the teams are, unsurprisingly, reluctant to share much data which could be used to work out where advantages may exist. That meant lots of work on our side to examine the data that was available and do some reverse engineering to figure out the details. Here’s an overview of all the essential areas and what we found.
Suspension: We got enough points from the CAD to work out steering geometry, anti-dive and such. General steering feel is a step up from the R18 TDI, IMO, and it handles the varying loads from downforce much better. LMP1 cars are lighter than before now by 30kg (870kg minimum), but the same basic suspension setup works fine from the TDI to e-tron. Not a big enough difference to throw out the basics.
Gearbox: 7-speed units now, and on-board telemetry that Audi broadcasted during Le Mans and other races last year was good enough to do basic speed vs. engine rpm analysis of the ratios. It's a pretty weird looking setup on paper, but drives really well since the diesel V6 has stupid levels of torque. First four gears are very very close together; I think as a plan to work in tandem with the flywheel hybrid system. Shifts were done at 4500rpm with very few exceptions.
Engine: 2014 regs meant no more air restrictor and a new boost limit of 4.0bar. I don't think they're going quite that high on this car it would go beyond the fuel flow limit with a 4.0L engine. Best guess is that it's running closer to 3.0bar like they used previously in the 3.7L version of the engine, and that's probably good for upward of 1100Nm torque. Slightly more than their 'over 800Nm' claim, but then that was always a 'No shit, Sherlock' kind of claim as the also-claimed 540hp means at least 850Nm@4500rpm. My best guess, using what we know from the previous engine, is that they are hitting that 540hp power peak early - around 3500rpm - and then using ECU boost/fuel control to hold it near constant power up to the shift point. Saves fuel (540hp is enough, really) and means that the operating rev range pretty much always has max power available.
Hybrid: This is where it gets interesting. The streaming video/telemetry was good enough to include a meter for hybrid energy storage. Without any other hard info, we have to rely on this to get an idea of how the system is working, and that ain't all bad. We know it's a 2MJ/lap system; watching closely and taking notes, it looks like they run through about 6 full charges of the system per lap of Le Mans. Easy math, then, to figure that in LM settings it is maxing out charge at ~330kJ. Official claims for system storage are over 600kJ and a few even stating up to 1.2MJ, but it doesn't make sense at Le Mans to carry that much at once if you have at least 6 good charging points around the lap and can only use 2MJ. Perhaps they use a higher level of storage at shorter tracks for the shorter races, but we don't have tons of info on that and LM is the main target anyway. Keeping the flywheel at lower speeds would be good for reliability unless you are absolutely certain that harvesting 4MJ per lap is reliably attainable. It seems they may be thinking just that for 2015 and I would agree that the 2MJ class leaves a lot of potential on the table. This year’s race also gave us some confirmation on the 330kJ number as Audi have claimed that they doubled their flywheel storage capacity to around 700kJ in their move to the 4MJ class.
Going back to the telemetry, out of chicanes on the Mulsanne, it is burning through that 330kJ in 5-6s and is emptied by the time the car is at 230kmh for an average system output of around 60kw (80hp). Maybe the system can do a full 170kW output as claimed, but perhaps that's only if used at maximum speed (340(ish)kph) and below that it is torque-limited so not to burn through the 2MJ too fast. We've got it modeled this way, so 170kW is possible but not ever really used as it isn't practical for performance over a whole lap. What you end up with is about the equivalent of an extra 200Nm from the engine accelerating through 5th gear, and that's not too shabby.
Aero: Very little 'official' data here, but a lot can be inferred. Audi have made public claims of lift:drag efficiency of 5:1, and we can be sure they are understating that. At top speed when ride height reduces under downforce load, it is probably closer to 6:1 (for those who aren't aero nerds, this is very very very good). What we do know is that top speeds without a draft at Le Mans were typically recorded in the 315-319kmh range. With a draft, closer to 340. Not very high in the grand scheme of things, and works out to something around 420lbf of drag @ 150mph in LM aero spec. Prototypes can drop 5% or more of their total drag compared to static ride height when squatting down under load at top speed. So what we'll do is take that 420lbf drag figure, add about 5% when at static ride height and figure the 5:1 downforce from there. If overall lap time performance is anything to go by, this has worked out quite well.
Audi were also kind enough to run both aero packages at Spa last year. #3 used the LM package while #1 and #2 used the sprint package with high drag/downforce. #3 was good for nearly 300kmh at the end of Kemmel straight, and our LM setup ends up matching that. Nice. Move to high downforce and top speed on Kemmel reduces to 275-279kmh while keeping similar efficiency. That's downright slow for an LMP1! These new generation cars are getting all of their speed in acceleration and hitting top speeds early on the straights. This is why the higher energy hybrid classes are the way to go and, IMO, were the key to Porsche’s victory this year.
What's it like to drive? Quite a bit different from the 2011-2013-era LMPs we've modeled so far. The narrower tires mean there is a lot less mechanical grip available and you end up hustling the car more at low speed. At the same time, the new aero/chassis rules mean more downforce on less mass, so high speed corners are an even more intense experience. Cruising to hit a good fuel number and maximize hybrid use is a very different experience and makes for a very interesting drive. Best laps here so far are a 3:26.3 at Le Mans, a 2:01.8 at Spa with high downforce and a 2:02.7 with low downforce. Can't wait to hear what you guys think of it.