Cars [WIP] Caterham

[Seven Smiles]

Basically i got in touch with a contender of the 2017 Academy series, who is getting us references for the model in exchange of his skin ingame
He has a website here: http://www.caterhamr500.co.uk/
He used to own an R500, just sold, and bought some time ago an Academy that he built. (lot of useful pictures from this website already)

Aha! By the way, he's not just a contender, he's leading the Green group, 20 points ahead of Caterham boss Graham Macdonald with 25 points going to the winner of the final round at Silverstone. You picked well!
http://uk.caterhamcars.com/motorsport/academy/standings

 

[JackCY]

And then there's my car http://www.evo.co.uk/reviews/6284/caterham with its 3rd party suspension which handles completely differently to any other Caterham I've driven (though I haven't tried a CSR).

I'm coming late to this, I've only just bought AC, but if you want technical input I can help. I also know a few Caterham racers (including Daniel French whose car you've captured superbly).

Yeah a completely new suspension, wide track with narrow body.

And where a regular car will be bouncing over a series of sharp bumps, this car is unfazed.

Don't even mention that Front is fine on the 1700 mod but rear will definitely bounce especially at around 2Hz when it gets hit at the biggest transmissibility. Driving tracks no problem but test tracks with equally spaced bumps in a series etc. the rear gets upset but that is quite normal for many cars. I could tune it out but then the car will end up as a race car not road car

Aha! By the way, he's not just a contender, he's leading the Green group, 20 points ahead of Caterham boss Graham Macdonald with 25 points going to the winner of the final round at Silverstone. You picked well!
http://uk.caterhamcars.com/motorsport/academy/standings

Yes he does well. I've browsed through many build blogs and scouted for info around, gave Ben links and images for references for his model building and to ask when we need some info. He only had the R500 which was quite impressive and switched to Academy, more fun to drive. Past R400 it gets quite into murder machine territory IMHO.
Any slight increase in power or changes to springs, dampers are easily noticeable. It is such a light car.

Ben has the final version now for all 3 1700ss options: RHD, LHD, Clamshell. When he gets back he will update it and release it.

Any more info on combined_factor for tyres? Do you like and notice it's changes? I have not added it yet, only some other v1.5 params. After working in the progressive springs, handling balance changed a little, it gets stiffer at front with more bump. Handles as expected and similar to using a regular stiffer front spring. It is so tossable with the open diff and live axle at rear, can slide it but it is no step on it and it will drift, almost no car does with open diff.

Default setup is for road use, there is some adjustment left but no dampers on this one. If you want to race it I do recommend doing some tweaks but that goes for all AC cars. The usual tweaks from a real Caterham should work. Testing testing Donington Park, yep seems to work, easy tweak in the setup menu down 1s on a lap. The draft advantage in race should be large 1-2s per lap on Caterham with front windshield, no way to test for me though and nothing I can do about, all about how AC does slipstream and how it's set on server.

 

[JackCY]

Ran a few tracks, this one might be the most challenging and interesting. Hard though since I don't remember it anymore so I can't floor it, been a long time from ST185 launch.
85% dusty surface, no blankets. Pressure, camber, brakes adjusted in setup menu.
Still even at slow slow speed going a touch too fast means going out of track. I had to slow down from my initial runs to even make it crash free, well beside turning into roads that aren't the track

Oh man YT really butchers the quality of all FPS view footage from anything. It can't deal with the foliage at all.

 

[JackCY]

Cruising with traffic AI = murder.

Tricking AI into overtake on inside line:

Collisions are all defined everywhere still AC manages warp the cars inside each other, how, I don't know, is the collision updating in physics or replay that slow?

 

[JackCY]

Can't say I like the COMBINED_FACTOR=2.3 for tyres, seems less defined and muddy feel than the supposedly default 2.0. Other than that I don't notice any less drifting or less inner spinning tyre as was suggested this setting could be for. I'm also not having an issue with GT86 lighting up tyres in 3rd gear with ease, I can shift the weight just like with Caterham and with throttle make the rear tyres slip, nothing wrong with that and both cars have so little power the engine doesn't explode into redline while tyres turn into smoke within split of a second. Will see months later if Kunos starts using this parameter and changes once again all their tyres to a different version.

 

[Leonardo Ratafiá]

oh, I just remember a great jump in a track if you wanna have some fun, try http://www.racedepartment.com/threads/ebisu-minami-drift-stadium.102520/

 

[mclarenf1papa]

Can't say I like the COMBINED_FACTOR=2.3 for tyres, seems less defined and muddy feel than the supposedly default 2.0. Other than that I don't notice any less drifting or less inner spinning tyre as was suggested this setting could be for. I'm also not having an issue with GT86 lighting up tyres in 3rd gear with ease, I can shift the weight just like with Caterham and with throttle make the rear tyres slip, nothing wrong with that and both cars have so little power the engine doesn't explode into redline while tyres turn into smoke within split of a second. Will see months later if Kunos starts using this parameter and changes once again all their tyres to a different version.

Even though you may not notice it, combined grip increases pretty notably with the combined factor parameter (which changes nothing other than that). Stereo and I wrote the code so I can tell you that definitively. I can also definitively tell you that using 2.0 is the exact same as not using the parameter at all (and is the same as previous versions of AC). The 2.3 number comes from matching the data of a road car tire.

 

[JackCY]

Which road car tyre? There are many variations. Some suitable for some cars some for other, different construction and all, not just "rubber" being different. Well I could achieve similar effect on handling by increasing the FRICTION_LIMIT_ANGLE. Skidpad wise same exact speed achieved. I thought the grip is a little better at first but that illusion went away and handling was worse overall, similar time, less predictable and less readable tyre. More grip, nope. Maybe it pushes some limits higher but practically driving, feels worse to drive and available grip isn't better. Might help those monsters like 812 whatever but for a light car with plenty grip, none in my experience. And 812 is hardly road tyres that are found on most road cars.

On Minami drifting... well sliding when possible to turn even there the 2.3 is kind of annoying tyre. Not faster just less fun to drive. Might work for overpowered cars, or less than 2.3. But then why not just tweak other tyre values.

 

[garyjpaterson]

Even though you may not notice it, combined grip increases pretty notably with the combined factor parameter (which changes nothing other than that). Stereo and I wrote the code so I can tell you that definitively. I can also definitively tell you that using 2.0 is the exact same as not using the parameter at all (and is the same as previous versions of AC). The 2.3 number comes from matching the data of a road car tire.

OT so sorry about that, but since its cropped up - That 2.3 value you recommend for street tyres, do you have any info on the specific type of tyre (ie age, chinese ditchfinder, premium brand summer etc), and would that value suffice for all types? Just wondering if you had any insight for old bias ply tyres, standard street tyres etc?
If not no worries

Had some fun testing a while back, going to values lower than 2 really emphasised exactly what it does.

 

[mclarenf1papa]

OT so sorry about that, but since its cropped up - That 2.3 value you recommend for street tyres, do you have any info on the specific type of tyre (ie age, chinese ditchfinder, premium brand summer etc), and would that value suffice for all types? Just wondering if you had any insight for old bias ply tyres, standard street tyres etc?
If not no worries

Had some fun testing a while back, going to values lower than 2 really emphasised exactly what it does.

It was either a semislick or high performance summer tire; good data for this is super rare so I don't have anything besides the one source. On my cars, 2.2-2.4 matches car behavior better than 2.0.

Which road car tyre? There are many variations. Some suitable for some cars some for other, different construction and all, not just "rubber" being different. Well I could achieve similar effect on handling by increasing the FRICTION_LIMIT_ANGLE. Skidpad wise same exact speed achieved. I thought the grip is a little better at first but that illusion went away and handling was worse overall, similar time, less predictable and less readable tyre. More grip, nope. Maybe it pushes some limits higher but practically driving, feels worse to drive and available grip isn't better. Might help those monsters like 812 whatever but for a light car with plenty grip, none in my experience. And 812 is hardly road tyres that are found on most road cars.

On Minami drifting... well sliding when possible to turn even there the 2.3 is kind of annoying tyre. Not faster just less fun to drive. Might work for overpowered cars, or less than 2.3. But then why not just tweak other tyre values.

Let me reiterate. I have the code in front of me. It objectively raises combined grip. Pure lateral and longitudinal remain the same, that's why the skidpad speeds don't change...and no, changing the optimum slip angle does not have the same effect; no other tire parameters perform the same function. For the tires I have data for (racing spec), 2.3 performs much more realistically than 2.0.

 

[JackCY]

Semislick and high performance = very new tyre such as the 812 I don't mind using it, but on old tyres maybe 2.1-2.2, 2.3 is too high. The combined is hard to describe, driving wise it's similar to raising the FRICTION_LIMIT_ANGLE, there is "more grip" in transition but then it doesn't translate into better tyre performance time wise at least not for me, rather the opposite the tyre feels more muddy, baloony, unreadable, similar to what too high FRICTION_LIMIT_ANGLE causes. And in the case of Caterham with open diff it seems to increase understeer from inner rear wheel while the fronts don't get extra grip to counter it. I don't have any fancy external telemetry running for precise numbers, it's how it drives to me just switching tyres.

I'm definitely looking forward to updated physics and tyres, anything that isn't content and graphics, there is a long way to go. Will see what Kunos does with tyres, will there be a mass patch or just some new cars will start using the "experimental" right now parameters.

All I can say is that the tyres in AC are very slidey and unreadable already, realistic? Can't say.

 

[JackCY]

Just how "green" is 2.3? And where is 2.0?

It may have more grip technically, no doubt, I believe you, but in a situation where precision and predictability is better to have it doesn't seem to translate very well into better handling. Without reworking the whole tyre it may not be worth adding so high combined factor for street tyres. I'm simply saying how it feels to drive it, not in absolute numbers how much grip there is. I don't have any real tyre data to compare with grip and behavior wise beside driver experiences.

Where would 2.3 and 2.0 be on this imaginary tyre? The more from center the more grip. As I understand it and how it feels in AC the combined factor 2.3 is making the tyre more green compared to 2.0. Is 2.0 the neutral yellow?

I do not know what code you have in front of you or what is in it precisely, how accurately it has been made to represent something realistic. Most people do not have access to or reverse engineer AC code to know what Kunos has decided to create. At least to me AC is a black box. I would happily browse through the code but like most other sims such codes are proprietary and inaccessible without RE.

 

[Leonardo Ratafiá]

Can't you just use one of the Kunos tires?

 

[garyjpaterson]

Just how "green" is 2.3? And where is 2.0?

It may have more grip technically, no doubt, I believe you, but in a situation where precision and predictability is better to have it doesn't seem to translate very well into better handling. Without reworking the whole tyre it may not be worth adding so high combined factor for street tyres. I'm simply saying how it feels to drive it, not in absolute numbers how much grip there is. I don't have any real tyre data to compare with grip and behavior wise beside driver experiences.

Where would 2.3 and 2.0 be on this imaginary tyre? The more from center the more grip. As I understand it and how it feels in AC the combined factor 2.3 is making the tyre more green compared to 2.0. Is 2.0 the neutral yellow?

I do not know what code you have in front of you or what is in it precisely, how accurately it has been made to represent something realistic. Most people do not have access to or reverse engineer AC code to know what Kunos has decided to create. At least to me AC is a black box. I would happily browse through the code but like most other sims such codes are proprietary and inaccessible without RE.

Use the tyre tester app, it will show you half of the traction circle (its pretty obvious when you open it which line it is).

 

[JackCY]

Can't you just use one of the Kunos tires?

There are no SDK V10 tyres per se. Otherwise what I'm using now is close to what Kunos has made with street tyres of similar type and size and all, they don't use the combined factor yet either.

Plus what are the absolute values of the axis etc. These graphs are quite useless honestly. Completely unlabeled and with some odd stair stepping in them too.

Sure I can see the effect of CF in the graph but that doesn't provide me with any information at all.

 

[mclarenf1papa]

These graphs are quite useless honestly. [...] Sure I can see the effect of CF in the graph but that doesn't provide me with any information at all.

Answers:

Just how "green" is 2.3? And where is 2.0?
[...]
Where would 2.3 and 2.0 be on this imaginary tyre? The more from center the more grip. As I understand it and how it feels in AC the combined factor 2.3 is making the tyre more green compared to 2.0. Is 2.0 the neutral yellow?

The tire app should be pretty understandable if you're familiar with the typically used/referred to curves for tire dynamics (slip angle, slip ratio, camber, etc). The scales are not the same for each curve, but they don't really need to be to understand their effects.

 

[David Dominguez]

It would be better to have an external software to analyze the tyres better, but that app is better than nothing. At least you can see the dynamic slip angle/ratio peaks, the dropoff of the lateral/longitudinal forces and the combined curve.

 

[JackCY]

Answers:

The tire app should be pretty understandable if you're familiar with the typically used/referred to curves for tire dynamics (slip angle, slip ratio, camber, etc). The scales are not the same for each curve, but they don't really need to be to understand their effects.

In other words it's a total mess of a graph that only the author will understand and can read since they know what they assigned to what at what scales. Anyone else is left in the dark. Even a simple legend would go a long way or for the author to learn how to do basic understandable graphs. Yes one can see it bouncing around dynamically, cool effect, useless without scales and knowing what value is how big.

External app for monitoring tyres would be better indeed than this oversimplified unlabeled wanna be graph.

So hard to do labels for Kunos?

This is readable and understandable, but what AC shows is definitely not.

 

[David Dominguez]

You have the peak values of the curves under the graph, you don't need to be a genius to figure out more or less the scale...

 

[mclarenf1papa]

In other words it's a total mess of a graph that only the author will understand and can read since they know what they assigned to what at what scales. Anyone else is left in the dark. Even a simple legend would go a long way or for the author to learn how to do basic understandable graphs. Yes one can see it bouncing around dynamically, cool effect, useless without scales and knowing what value is how big.

External app for monitoring tyres would be better indeed than this oversimplified unlabeled wanna be graph.

So hard to do labels for Kunos?

This is readable and understandable, but what AC shows is definitely not.

As I said in my post, the graphs are all common; you don't need to be a developer to be able to tell what they are.

Two of the seven things are using known formulas or lookup tables so you should know their scale anyway...as David said, the rest are fairly straightforward to figure out.

It's obviously not ideal, but being so excessively negative and condescending about it (and AC's TM in general) isn't going to help anyone.