Group C cars grip under acceleration

I dont want to take anything away from your excelent work doing this testing so far, more maybe point out something, you are comparing the race pace with the fastest laptimes, but i know that the pace somtimes varied massively,and if they sometimes did a quick lap, they did many slow ones to compensate.

In the Suzuka race for example, it was ran in the dry, and without safety cars. So my way of doing it would be, grab the total time that the winner sauber took to do the race, subtract 4 minutes to acount for both pitstops, and then divide by the total number of laps to give the average race lap. Doing this gives me a race lap around the 2 minute mark. Yes you are reading this right. Now, i believe they slowed down the pace in the end because they were well clear, but still, they were clear because they outran everybody before, and we never know if they were running over fuel or not, but when they post their fast laps, they most certainly were.

 

First of all, thanks for chiming in.

An average of 2 minutes sounds about right to be honest, considering that fastest laps were in tje 58s. It's quite tough to compare due to the layout changes the track has gone through, which are not big but definitely impact on laptimes by a lot given the margins we are looking into.

Yes, I also believe their fastest laps were definitely over the quota number. I took the 20% margin mentioned, which sounds quite arbitrary at first, from Heinz Harald Frentzen's Twitter posts, where he explained that engineers would ask drivers to save a given number of fuel, which could be up to 20%, given how the numbers were going depending on the dynamics of the race.

I have not tried to actually hit the quota numbers on track to gauge how much pace is lost in the process, but it would not surprise me if more than a second is lost in the process (trying to lift and coast that is, reducing fuel mapping would come at a much bigger loss!)

Still, while numbers were a surprise finding (close enough for my expectations), the main takeaway for me is how the car drove with the setup limitations proposed

ATM I'm starting test runs with the 962C. I may have some info to share later or tomorrow.

 

I'd not thought of that - not sure how to sync the mech damage to the race length as I run a dedi server through Emperor. But it seems about right as it has happened before - Brabham BT52, 2nd lap blow up after driving like a granny and then downshifting in the redline and BOOM! I tried to induce engine / gearbox blowups on the practice server, and you can abuse the ever-living #### out of the gearbox on downshifts without too much damage. I managed to blow the Sauber engine only by intentionally going from 5th to 1st when at speed. It does feel as if mechanical damage only really operates in the race session.

 

Did some laps on the Porsche 962C. Some interesting findings for sure so far.

Same test scenario as usual: Spa 1993, 17/9/89 in game, real weather, default progressing track, 2 PM time in game. Changing at first only steering lock and brake pressure, and loading 51 liters of fuel. I'm considering that the car is a 1987 spec, which was the last one that was run by the works team in Sprint/HDF form.

Right out of the box, the car feels quite driveable and somewhat easy. There is some turn-in understeer and some power out oversteer, which is to be expected, but perhaps at lower levels than what they should be. Engine is outputting 727 HP, which feels correct: the real engine outputted 740 HP max at 8400 RPM with 1.2 bar of boost pressure. While the red light turns on in the game HUD at that RPM, the power curves show that the max power is achieved actually between 8100 and 8200 RPM. Same happens when you up boost to maximum value, where the engine in game does deliver the 740 HP stated. The boost gauge goes up to 1.15 bar, as stated by other posters before.

Best laptime I achieved on this run was a 05.8, with most laptimes in the 06s range, specially after tyre wear was being felt. It was causing it to oversteer more on power out while keeping the understeer on entry, and it was hard to keep bettering laptimes past halfway, not being able to come close to my PB. As a reference, I experienced similar troubles with the Sauber, with the best lap on this very scenario a 04.3. A predictable 1.5s gap, which seems fine on principle.

So decided to try some changes, but not in the search of maxing performance, but of realism, like with the C9. Front Ride Height comes at 55mm, which is stated in the book I have as the ground clearance of all 956/962 models, so that is a good minimum value (it can be lowered to 50 in game), and rear ride height comes at 66, and can be lowered to 61, which is good. Left it at 65. On the diff side of things, I turned Spool ON. It did not give me the option to deactivate LSD, I don't know if the Spool overrides it? Maybe devs can answer this one? Also, saw that the 962C used fixed spring rates at Sprint setup, coming at 220 N/m at front and 170 N/m at rear. The former is achievable, the rear is not, as the minimum is 190 N/m. So I set the front at 240 N/m (default value), to keep a similar bias.

With these changes, the understeer on entry and oversteer on exit were way more pronounced. And at the beginning of the run, it was slower by some tenths. But as the run went on it kept going faster, as the balance was less affected by tyre wear, and ended up beating my PB by 2 tenths, setting a 05.6. Data showed that for this car, cambers were fine, it just needed a slight reduction of tyre pressures at the front and some damper tweaking.

Maxed boost, took fuel out, reduced rad to 60% and went for a Q run with two hotlaps. Best lap was a 04.8. Probably more pace to be found, but not that much actually.

So now I topped the tank, fitted Hard Tyres, set radiator at 70%, boost at 80%, engine braking at 1, tweaked dampers, and started a run. I couldn't complete it as I had to stop driving, but I did about 1/3 of the tank, and there are some things to ponder already looking at the data.

First, the pace. The run started in 08s, but not long after I was doing 06s, best lap coming at 06.6. That is one second faster than the scenario I tested on the C9!

Power output at 80% boost was roughly 715 HP. But the most telling thing is that the fuel consumption was on average 4.25 Liters. That is right on the Quota + 20% limit stated before, and 0.3 Liters per lap down on the current C9 at 66% boost and 770 HP. Now, I know that we are also talking of a 55 HP gap, but bear in mind a couple historical details: the Porsche 962C in 1987 was still running with the Motronic MPI 1.2 system, while the Sauber C9 in 1989 was already running the Motronic MPI 1.8, a newer version and even superior to the 1.7 that the 962Cs used from 1988 and onwards. One would be led to believe that is normal that a V8 5 liter engine gulps more fuel than a Flat 6 3 liter. But it's just not what was happening in this era of motorsports, where the former was king come race time. I think that a good way to compare them is to set them at similar power outputs (I should try lowering the Sauber to around 50% and should be quite close), and on that scenario, the C9 should take less fuel.

Top Speed Recorded: 297 km/h
Downforce Generated at That Point: -12686 N. Front: -8825; Rear: -3861
Max Downforce Generated: -2805. Front: -1981; Rear: -824. Spot: Eau Rouge/Raidillon Transition. Speed: 277 km/h.

If I'm interpreting these numbers correctly (they are always negative and go up as speed rises), the 962C generates more downforce than the C9. I need to run the C9 again at similar ride heights and recheck values, maybe I'm just neutering it too much by forcing front ride height to be higher.


So far, what I can see is that the Porsche, set with the intentional front ride height restraint and the spool diff, is a blast to drive, extremely fun and seems quite true to real life in my books, I just love it. But it seems to be too effective on the fuel consumption and outright downforce levels, leading it to be a faster car than the Sauber on a realistic endurance scenario. As my best dataset on the C9 comes at Suzuka, I need to move the Porsche there and do a full run right as I have it setup at the moment. Will try to do that between tonight and tomorrow.

 

Just finished a full tank run with the 962C at Suzuka, so here are the numbers.

Track: Kansai Classic
Date in game: 9/4/89
Time in game: 2 PM
Weather: Realistic
Track: Medium rubber
Best Laptime: 1:51.8
Top Speed: 299 km/h
Downforce Registered at Top Speed: -16888 N. Front: -11524; Rear: -5364.
Ride Heights at Top Speed: FL 27.54; FR 27.38; RL 23.79; RR 24.38
Max Downforce Registered: -14694 N. Front -10055; Rear -4640.
Speed at Max DF: 297 km/h
Ride Heights at Max DF: FL 34.58; FR 13.12; RL: 31.79; RR 11.89.
Spot: Right at turn in at 130R, no lifting involved.
Tyre Left % at end of the run: FL 75; FR 62; RL 62; RR 53
Engine Damage % at end of the run: 4%
Avg Fuel per Lap: 3.45L

In terms of describing the car: it understeers a lot, specially on higher fuel loads. But while the tyres do wear down significantly, the car remains quite consistent on its handling, while benefitting from fuel burn and getting lighter. It starts the run slower than the C9, but after a few laps it matches its laptimes, and by the time that you used a third of the tank, it becomes faster.

At its best, the car is faster than the C9 by a huge margin of 2 seconds. They are almost matched in top speed, just a 3 km/h gap, not bad for being 55 HP down. But where I see a massive difference is on ride heights at top speed: the 962C rides 20mm lower than the Sauber, while the rear end squats down, which probably helps the 962 on shedding off drag on a straightline, which is also helped by the rear springs being run at a softer value. This triggers the need of some back and forth testing between the cars, like trialling the 962 again on 10mm higher ride heights, trialling the C9 with softer rear springs and measure its effects, etc. What remains true, to me, are two things:

1. The car produces more downforce than expected while being setup with true to life suspension settings, leading to too fast laptimes, aided by...
2. The car burns too little fuel, here at 0.2 Liters per lap less than the Sauber, and staying inside the fuel saving window of 20% while using 80% of the boost available. +700 HP on races was a luxury that Porsche did not enjoy AFAIK.

These things aside, the car's driving is extremely believable on its current state. I would like for Spool diff and spring rates at 220 front and 170 rear to become fixed on the setup

 

When you refer the spring rates, you are quoting the Haynes book, right? Bear in mind that altough the springs and dampers are there, they dont give you the information of motion ratios (if you know these, disregard what i am saying). So those numbers,and the relationship between front and rear might be (and probably was) different.

 

I don‘t know if a spool should be the only available option for the Porsche. You can probably tune the LSD to get a 100 % locking effect but it had no spool like the 917.

Daily Sportscar published a fantastic series of articles about Porsche and their journey through the development of the 956 and the 962, and apparently Stuck was winning races with the PDK in 1986 already, so I‘m not sure we have the 1987 model.

It‘d probably be OP with a PDK in AMS2 ^^

 

Yes, Haynes book! TBH, I don't understand the damper numbers shown there. And no, I don't have motion ratios. That's an excellent point you make. I know there are better 956/962 books out there that probably have better and more complete information, but unfortunately they are out of my budget.

 

Basing myself on the book I have, spool was a de facto decision by Porsche designers. Their reasoning was that in case of a driveshaft failure, a spool diff car can make it back to the pits for a repair, while a LSD one would be stranded on track and out of the race. Even with the PDK, diffs remained spool on works cars.

 

Looks like it really had a locked diff. Did Singer write a book? He really should consider it if he hasn‘t already, I think people would buy it ^^

 

I think he has book called "How to read a rulebook", and is focused on the times Porsche exploited loopholes to get advantages, the most (in)famous being Le Mans 1994.

 

Quick test to check aero when raising ride heights on the 962C. The maximum AMS2 allows atm is 60 front 81 rear. So I raised front to 60, and set rear at 70, to keep the previous rake. Took fuel out to make outlap + 3 laps simulating end of run (but with new tyres of course) and this is what I've got.

Best Laptime: 1:53.4
Top Speed: 298 km/h
DF at Top Speed: -24040. Front -15805; Rear -8235
Ride Heights at Top Speed: FL: 35,57; FR 35,23; RL 31,96; RR 33,84
Max DF Registered: -19508. Front: -12252; Rear -7256. Spot: Turn-in 130R.
Speed at Max DF: 288 km/h
Ride Heights at Max DF: FL: 42.88; FR 10.25; RL 41.15; RR 7.07

Downforce loss with 5mm higher ride height is massive, and laptime gap is probably on the 2s region, considering I had brand new tyres.

I guess a quick conclusion I reach, there is no easy way out for balacing the cars with just setup range limitation. You can take front aero adjustments of the Sauber out and fixate it at 0.00, and force Spool on the 962C with LSD off, which are both things that would enhance realism, and would be a step forward with a hotfix. But the aero balance of the C9 needs further revisions to have it pushed back more, and I'm pretty certain that the 962C is requiring an overall decrease of downforce while retaining current levels of drag. But being close to end of the line and leaderboard reset, that seems like maybe a work for 1.6 version, IMO.

Next up, Nissan R89C

 

That must certainly be super interesting to read. Looks like he wrote several books, I think I'm going to get one or two. ^^
I'm planning to go to the Porsche museum, hopefully they have some of his books in the shop.

 

Did my first laps with the R89C at Spa, usual test scenario.

Out of the box, the car reminds me more of the 962C than the C9 when driving, and that is a car with a big tendence to understeer a lot, then try to swap ends when you accelerate it more. Not what I expected, considering that the Nissan is of a much newer construction with better materiales and better aero design. The car did not lack downforce, but here it seems lacking on that regard. In laptimes, it's also slower, struggling to match the Porsche's laptimes, but with a much higher tyre wear rate, that makes it tough to keep the pace as the run goes on.

The car is gulping 4.55 Liters per lap at 80% boost, producing around 790 HP at around 7700 RPM, 200 RPM after the rev light goes on in the cockpit. So it seems like the engine was knocked back from the power levels of previous versions, when was producing a power output of a 1990 engine instead of a 1989 one. That, plus ride heights being more believable at the moment, make you feel you are driving a racecar instead of a powerboat as it was up to 1.4.8, where clipping a kerb a bit too violently would send the car flying as the aero effects were temporary reversed.

Something funny I notice: oil temp seems to top at 110ºC instead of keeping on raising as the car accelerates on long straights, like what you see on other cars.

Car comes with a front ride height of 65 and a rear one of 70, which means on default, it's the Group C running the highest ride height with the lowest rake. Front can go down to 60, so I'm setting it like that for the next run, because it really feels like it could use some more rake. Specially given that the springs on this car tend to follow the weight distribution, and are set stiffer at the rear, something that was not the case in the other two. -3.5 front camber seems like too much, and tyre temp spread expresses it, going past 13 degrees at some points, so I'm dialling it back to -3. Rear wheel speeds showed lots of inside wheel spin, so the power angle was dialled back to 30, while the coast was raised to 60, in search of improving turn-in.

With these changes, the car transformed and became A LOT more fun to drive, the best of the 3 in my opinion, kind of a middle ground between the resistance to turn from the Porsche, and the "pointiness" (new and unexisting term? lol) of the Sauber. This car has different tyres in game compared to those two, who share them (and in reality, it was both Nissan and Porsche works teams, together with Jaguar, that run Dunlop Denlocs), and the Nissan's slide more. That was noticeable with the fronts climbing up Raidillon on default, but now it's felt everywhere, specially in and out of the Bus Stop, and it's a pleasure to drive and push. The tyre wear was less noticeable, and laptimes kept on improving, managing a 04.1, which was 1.8 seconds faster than the best achieved before, a mammoth improvement! The car changed from struggling to follow the Porsche, to be able to fight the Sauber. Downforce numbers improved a lot from what I can see on the telemetry too, being now quite close to what the C9 produces.

I will take this car to Suzuka for the customary full tank run to have more data points, but as it stands now, the car is quite good and close to where it should be, main problem being that it's fuel consumption is too similar towards the Sauber. If the latter's is not going to be reduced, then the Nissan's should be increased. Otherwise, it can stay right where it is and I think it's on the right ballpark.

Sharing my adjusted setup via screenshots. I think it's a MUCH better starting point than what it's coming with the game.

 

Instead of taking the R89C for a full run at Suzuka, decided to check where the intruder of the class, the Corvette GTP, is standing against the other 3 at Spa.

The car seems to have lost the magical (and IMO artificial) performance edge it once had, and now has to deal with the actual limitations: high weight and low downforce. It does have tyres that seem to grip more than its rivals, and the balance right out of the box is less understeering than Porsche and Nissan, and has good traction out of corners, but it cannot help itself with longer brake distances, slow reactions when changing directions, and struggling to absorb suspension impacts when hitting kerbs, the car wanting to pogo and lose control. It also wears tyres at a different (and worse) rate than its rivals, because not only eats up the rears faster than the front, but it's also lateral biased at Spa, the right side tyres dying faster than the left ones. All of this makes for a machine that is able to hit decent laptimes, but struggles to keep them on a longer run.

It consumes 4.38 Liters per lap. Below Sauber and Nissan, but higher than the Porsche. This car has no boost adjustment, and default fuel map at 1.00, so it was producing max power, which now is 770 HP at 7500 RPM (Suzuka will give a higher number probably, given it's at almost sea level, compared to the 400 meters above at Spa).

Ride heights come at 65/76, with minimums at 60/71, so I set it to those values. Changed cambers, increased caster, some damper and diff tweaking, and went out again.

Car was much more responsive, and after a few laps I started to improve my PB, but the tyre wear, while less noticeable, was still there and I could see on the HUD that right tyres wore more. Got a 04.8, and I reckon I had a couple tenths more to better it before I got rubber-limited, but kids were not interested in me continuing my run, and instead wanted to test the damage model against the walls

I will have to do full tank runs with each car (repeating runs with the 962C and C9, not limiting ride height, but still respecting spool diff with the former, and 0 front downforce with the latter) to come up with more precise conclusions regarding how the cars stand against each other, and propose something that can be done on a hotfix.

I will say this: there is no question in my mind that Group C class on Automobilista 2, no matter all small historical inaccuracies I may have mentioned, drives better than ever compared to previous versions on this game

 

Progress has been slow for me on this. I managed to get a setup for the Sauber, used the pointer given by Wilks on another thread that the C9 did not use rear anti-roll bar, to get rid of the oversteer, but I'm not exactly happy with the result, even if it's driveable and very fast with ride heights at minimum. I'm still some time off final comparison data. Still trying though.

 

Excellent!

 

How did you get those ride height and DF numbers?

 

Second Monitor's Telemetry Viewer.

 

Yep, after I asked in this thread I did a quick search and it looks like he wrote a few books. Definitely getting one or two.