That is interesting. I never tried to overcome understeer in iR with more lock. I would think that would probably work up to a point in most situations in real life. For one thing, you're taking energy out of the car by doing that, and therefore reducing the tendancy to understeer. I guess it would depend on the load on the tire, the slip angle, and just how much "too fast" you're travelling. Are you talking soley about entry? (no or little throttle, possibly some brakes).
With default setups, I get some understeer in iR as well, but it usually solves itself intuitively. What's odd is that in LFS, an understeer situation is very hard to come out of - but an oversteer situation is not. Once the front goes, you're in trouble. If the back goes, no problem. iR takes much more focus on oversteer, and much less on understeer - sort of backwards in terms of the degree of required input and "how it feels".
But, as you say... different cars, tracks (especially surfaces), setups etc. Fun to play with though!
This is the only thing that inherantly pisses me off about LFS at the moment... Curious to see what's coming down the pipe.
Oh um, yes indeed *cough* of course. I was, uh, thinking that too!
i rather hope hell choose to use some type of physical model (as lfs seems to) and tune its parameters until it fits the measured data well
any regression derived model is essentially the same as table based models
Just a thought, but i think the new physics could make some cars faster, FWD and AWD especially. And FXR could get its chance to be competitive in the GTR class, which, i think, would be appreciated by many. Same for RB4. Just my 5 cents, we'll need to wait ofcourse, what devs will come up with.
Entry is bad, yes. Little to no throttle, possibly brakes, or just coming off brakes. I don't find it completely unrealistic, but it isn't exactly sharp either. When you are on throttle, its worse of course. Throttle steering also seems to be way too pronounced in most of the iRacing cars that I have tried.
Yes, steering more when understeering in real life can bring you out of understeer, but you come out because you are scrubbing the tires, using them as a brake which of course transfers some weight. It will destroy your tires in a short while. With iRacing, it seems to be pretty much required once the weight distribution evens out after braking.
I would agree with that. LFS is very hard to stop understeering once it starts and it does seem to be easier, in general, to catch oversteer. iRacing seems to be similar to other games in that oversteer can be extremely difficult to catch, even in mild situations where understeer is easier to recover from.
New physics could definitely change the balance on the cars but it isn't going to make the FXR more competitive in my opinion. It already can accelerate with full power pretty much no problem. It *may* help with turn-in on FW/AW drive cars but I would suspect that it would be pretty balanced in that way for all cars.
Well, I'm waiting happily and patiently for the next patch, no matter when it gets released.
It was good to read this piece of information though... it gives us some clues on what they are working on. But whatever it is, I'll be happy to get it and play with it on the right time.
Was it just me or has anyone else read "I'm licking this thread" ?
I iR, I generally find myself coming out of corners with a hint of oversteer much like in LFS. That could be just because I try and do things the proper way no matter what though. I've never noticed it being necessary to mess with understeer, and my lap times are respectable... ish.. (hit a 48 in the Solstice at Laguna, that's not bad). Maybe I'm not 100% sure what you mean by "required".
Understeer should be easier to overcome since it doesn't yaw the car more; the whole point is the car yaws less than you're wanting. In LFS with oversteer, I really never feel that anything is out of hand or even dangerous... Kind of like "Yawn, oh, woops look at that... I'm at 45 degrees... bah, no big deal". In iR, if the back steps out, it's perfectly savable as long as you're on the ball and paying attention. It requires driver input and calculation. I would never compare it to poo like ISI put out, where pretty much any slip angle is certain doom. You won't see anyone (successfully) drifting in GTR for example.
As a side note, I'm willing to wager that when Scawen finishes messing with the physics on this iteration, that throttle steer will be much more prevalent in LFS than it is now - mark that down on your calendar.
If that doesn't turn out to be the case, then I reserve the right to remove this post however.
Hehe... and what is a algorithm for that? Doing a regression model with least square method based on a functional relation. This equation can be every physical model. You just define the paramters by a regression analysis .
i mean a physical model suitable for simulation eg brush models and other types of simplified fem not behavioural models like pacejka which is what youre suggesting
Yes but how do you get the paramters of your simplified FEM model? If you want to "re-engineer" given data like the folks of iRacing are doing as someone said.
I oversteer coming out of corners with iRacing too... mostly because of all the lock I have in trying to make the corner.
What i mean by required is that if I am slow enough to make my initial turn in point, mostly off throttle and my arc is good to hit the apex, if I want to stay on that path, I have to increase steering lock substantially with iRacing or my arc will wash out and I won't hit the apex. This is especially bad when I am using a touch of maintenance throttle to keep the back end pinned down. I found that I can't use maintenance throttle in the Solstice. However, with other cars I have to use it because of the throttle steering and then the steering lock goes way up.
But, seriously, the iRacing car will drift out nicely with a tad of throttle induced oversteer coming out of a corner. LFS does this as well.
LFS does throttle steering very well. It is the setups that generally wash all of it out.
Driving the spec racer Ford in iRacing and the Barber car is a bit nuts. The throttle steer in both of those is way happy. I swear I could make it around a lot of tracks without turning the wheel more than a degree. Throttle steer is good and it is realistic, I'm just saying that iRacing goes a bit too far with some of the cars. It's a car model thing though, not a tire physics thing I think.
Anyway, I am not sure if any of that is right or wrong, I am just saying that is how I see it.
you dont need a functional relation to compare your models results with the measurements though
all you need is a mmse approach on the probably smooth model data compared against the measured datasets on a virtual tyre test rig
Yes, but what is your virtual tyre test rig? If its not a couple of equations I dont really get the point, then you still dont have something you can build a simulation up on. At least not one you can run live. I assume that also LFS is build upon equations and i assume also that these are relatively often approximations. I dont really understand (i dont say that its not because of my limited knowledge ). Maybe we should stop bothering the thread and if you are so kind, you can explain me by private message...
As far as I know (and our local tyre guru Todd seems to agree on this IIRC) LFS uses some kind of physical model, e.g. some kind of simplified finite element approach, that doesn't model the tyre as one equation but instead as a number of small elements interacting with each other and the road.
The thing about these approaches is, that they show what is essentially emergent behaviour, which is hard or impossible to predict, particularly in fields like tyre simulation where the models that work well and the data needed to fine tune models isn't easily or publicly available.
So what you need to do is code something that will put your model under exactly the same conditions as real tyre tests and then compare against real world results while tuning your model parameters.
I only remember one, thank Sam for that. (Aero conversation - shakes fist at Sam) What was the other one? Something to do with sound?
By the way; to anyone that might think I'm taking this conversation in an iRacing direction - I enjoy both these sims and simply enjoy observing and discussing their qualities with likeminded folk. Right now, either one of these sims has features (and lack of features) that can ruin the other one for you .
It's going to be an interesting next half a year in simracing what with Scawen working on some things, and iRacing announcing tire model improvments, night racing, some awesome road racing & tin top content, and drive train modelling.
Kunos is dead, SimBin is worthless, but at least now we have two worthwhile projects poised to present some work worth investigating.
Hah,
Fair enough. When you say that, I do remember that I have a much harder time being smooth in iRacing than I used to in LFS when I was driving it a lot. Not in the Solstice, but in the Late Model for sure. I wish they'd make a LM road series, that car is a hoot on road courses.
I'll have to test that... Thinking about it I don't recall needing more lock before I hit the apex, if anything I'm oversteering a tiny bit before I even hit it most of the time. Could be that I'm just being a goon though I guess.
Really? With the solstice I pretty much brake, ease off whilst turning in initially, and then get back on pretty much full throttle all the way around the turn. I don't notice any understeer doing that
Indeed, I agree. Or at least I do when LFS is using a bastardized setup. The car doesn't behave the same with a similar setup as one would use in iR however, that's my beef. I remember many an LFS race coming around with neutral steering on the throttle; good times.
- I'm thinking it's the odd setups that actually dial it in TBH.
Heh, hyperbole aside, I think that the Legends car is the epitome of throttle steer. Get a road setup, take it around Limerock and tell me what you think of them apples!
That being said, you have the SR Ford drivers in the iR forum saying that say that indeed is how the car behaves - so who knows.
Yes i remember, its this 40 (something) parts the tyre is divided in. Right? But since this isnt to much, i guess that there are some equations how these 40 elements are attached to each other. And i think this equations are kind of important also, not just how each one handles. But this is not going into a proper way, in the end... we both think the same. There is developed model, and not a one equation model which includes tables to work on, but something more FEM like. Up to now, iRacing is using the ISI way, a map of data, like an engine map. Right?
I think the understeer, oversteer LFS/iRacing discussion is more interesting.
I was hoping for a tyre physics update. It's better than any content, because all the current content gets new that way.
Hopefully to much grip longitudinally with much wheelspin will be solved as well as the generally a bit icy feeling and low grip problem for a lot of situations.
If both are solved, this will be a great patch which makes LFS all new somehow.
All the cars will behave different then ...
.
.