Automobilista 2 V1.5 Physics Development Update

about +1s with the "low diff" setup.
1:25.2 with high diff
1:26.1 with low diff

note: one-shot lap, maybe in the end, exploiting the pros of both setups, the difference could come down to some tenths, maybe 0.5s, but not sliding is generally safer but slower.
note2: I did not check tires consumtpion/overheating at all, I am expecting it to be quite higher on setup 1.

 

Cause in my experience dropping a bit the diff is ok but your second test is excessive and it shows in the video. Something intermediate is probably the better option.

 

Yeah, often the default setup feels like the differential is quite locked and when cornering, kinda wants the car to do a bit of drifting. Changing it too open makes the car too unstable though. Adjusting the diff is one of the things I often try first with setup (though usually it's not the number of clutches).

 

Edit: spelling/typing
yes totally agree.
The two tests were just "scientific trials" to prove a point. and I was successful indeed.
I am now thinking about potential differences with low preload / mid-high clutch plates or high preload / few clutch plates.

One thing: differently to what common opinion is, in the second setup (low preload low clutch number) I found also breaking and corner entry to be more stable.
Common agreement is that more diff lock makes braking and corner entry more stable, but I think it is a matter of driving dynamics and driving style.
e.g.: if I feel during braking and corner entry that the car struggles to turn in (case of high diff lock), I automatically apply more trail braking inducing slight oversteer/instability, so I "feel" that too much diff lock for me results in throttle oversteer but also braking/entry oversteer.
Same note for action on throttle, common knowledge says that locked diff results in power understeer (since both tires provide thrust torque in tangential direction to the corner), but this is true only until the tire grip is not (even slightly) exceeded. Afterwards oversteer is induced.

My conclusion: all the "guidelines/wikis" must be taken with a grain of salt and accomodated to driving style, driving dynamics and personal feel.

 

Just yesterday I joined online race late in the qualifying, it was street track but the name now escapes me. I felt the car was diff was too locked, so I opened it a bit (both clutch and preload on the gt4 ginetta) and then opened it a bit more just before the race... went too far and it definitely made the car really unstable during braking and corner entry (and I'm quite smooth on brakes and cautious in general in my driving). Lets just say, it was so unstable that I didn't finish the race...

 

It depends a lot on what BB you are using and what car/track. Had a few laps yesterday to go on the leaderboards for GT1 at Monza and it's pretty sensitive when you go down on locking at turn in given the little aero and probably a bit the nature of the car with its engine braking.
I ended up reducing pre-load (probably I will test even less preload today) and fine tuning with coast ramp. Still the car is pretty sensitive to downshifting and braking while turning in.
Definitely there is very little traction out of slow corners and it feels a tad too much for me.
Other people on the leaderboard went down to 70 pre-load. Car had a 150/160 or so baseline...
With the amount of grip from the tires it is impossible to use such a pre-load in low speed corners.

 

Did you try decreasing clutch number by 1 click? Or increasing the ramps by 10-15 or more?
Because preload (as the name says) is the static preload for the clutches and thsi gives a certain amount of lock when there is no torque coming from the shaft to the differential. This actually acts only when there is no torque (positive or negative) imposed to/from the engine, which is a very very brief moment in driving dynamics (e.g. very low speed maneuvering). Other than this, the ramps + clutches will play the main role in the diff locking.
If you feel uncomfortable in braking/downshifting (negative torque) or in acceleration (positive torque), then my understanding is that preload does not matter much or not at all. I woudl try increase substantially the ramps, or lowering clutch number (more dramatic effect).

 

Correction!
The preload actually acts as a starting point for the locking ramp, actually shifting it left/right.
Somethign like this, in my mind.

What is not clear to me is how to link the preload to the locking percentage.
I don't know how much initial locking percentage I have with 100N preload, or how much more with 200 or 300N.

 

The pre-load is just as you show above--a minimum. With that minimum, the car will never revert to an open diff situation even in neutral state that you describe above. If your problem is in that neutral or transitional state, then adjusting the pre-load could help. If you have normal diff-related balance issues, then just adjust the normal diff settings.

Remember, lots of vehicles have LSDs that are not so conveniently adjustable as what we have in the game. Reiza has "closed" some adjustments here and there, but there is still often more adjustability than in real life, which is a good thing when some of the "connecting tissue" physics are still being cemented into place.

 

Yea I was pleasantly surprised to see a lot of the tuning options in the Formula Inters reduced and greyed out. Would like to see more of this in other cars as some cars don't have that level of adjustments IRL.

 

Agreed, but the basic physics has to be very solid before doing so.

There was a reason you could pick diffs like magic from an open buffet in PC 2. It allowed you to get the car to handle the way you wanted, even if it wasn't realistic or the same as the real car. Over time, cars in AMS 2 are getting closer and closer to behaving and responding exactly like their real counterparts. This is the purpose of simulation, but with things as complicated as race cars and vehicle dynamics, better not to close doors until you are sure you won't need to go back through them

 

To all those who complain about handling, grip, car behavior, I do strongly suggest to play with setups.
I honestly have no memory of a sim so sensitive to 1 click of arb or 5mm of ride height.
I am really really surprised and astonished.
One clear example is the sauber c9. We all know its tendency to slip the rears. Especially out of slow corners.
Well, lower the rear 5mm, 1 click less of rear ARB, increase 15 points the power and coast ramps of lsd. Et voilà. Not a puppy for sure, but a treat to drive.

For the third day in a row... Chapeau Reiza!

 

It is astonishing that we can feel and recognize such small changes, isn't it!!!

I don't think I am telling tales out of school that with every single vehicle in the game getting a major tire update (again!), there has not been time to optimize every default set-up in the game. We should not hesitate to fine tune, even with just the very basics, if a car is not behaving as well as we would like.

 

Agree on the sensitivity to settings changes. That is why I am telling you that Porsche GT1 is far too prone to lose traction that people are forced to reduce a lot the diff locking (we are talking about going from 160 to 110 or even 70, not slight adjustments).
I do my own setups and try my best to be in the top 10 of leaderboards for GT1s at certain tracks, usually with a setup made by myself. Usually I made it there.
Will continue to explore the setup obviously to see what other workarounds are there but so far the two tracks I have tried were very much the same thing. Locking being far far too extreme vs the available grip.
So there are two things: either the grip is insufficient or the locking is excessive for some reason. Possibly a bit of both.
At low speed there is little impact from aero and ride height. So mechanical grip settings (including suspensions) are the only thing that can help

 

Yeah it is.
But often its just because we exactly know we have made the changes.
And exactly what the changes is and what the expected outcome should be.
Thats the reason in scientific tests neither the person who is under test or the person who execute the test must know if the test is real or just executed through some placebo stuff.

ByTheWay: Its called double blind safety.

 

It would be great when the day finally comes and you actually get a genuine "race engineer" in a sim. Something whereby you can indicate your issues from a vast array of structured options.. eg "Oversteer in sweeping bends on exit" or "understeer on entry slow corners". Then the "engineer" (game) adjusts your setup for you accordingly.

These options could be structured in a way to not be overwhelming for selection.. ie Select "Oversteer" and then drill down further into more detailed options related to oversteer.

There could be a page summarising the adjustments made and you just confirm or cancel or maybe even be given the option to tweak these further before committing to them. Perhaps even some adjustment option alternatives , ie "This set of adjustments A" or "This set of Adjustments B" (more than one way to skin a cat). Then, out you go to test the results. (might need an "undo last set of adjustments "option)

This would also be a great teaching aid to general setup and its effects.

 

Sounds like what PC2 had.

 

Hey, you're right. I completely forgot about that. Thought I had a great idea for once.. Seems it was someone else's idea after all

 

I get your point - but...
But my guess is that the same kind of people who disgust to use any effort to tune setups will be the same kind of people who will let down this multiple choice - because they never get the excact setup outcome they are after.
In RL its similiar to many skilled drivers who really have a hard time to get their race engineer to understand what they mean. And want.