camber testing- results attached

[nomis3613]

Edit 11 Sep 2013: Please see this post below for updated testing: https://www.gtplanet.net/forum/showthread.php?p=8811231#post8811231 The results are slightly different

In hindsight, it appears that the original testing and analysis (below the line in this post) had the following problems:
- Acceleration testing affected by AT gearbox hitting the rev limiter for some tests
- Drawing conclusions for brake test results without considering measurement accuracy (ie most of the results are within 1 metre)
- Similarly for the understeer results, focussing on peaks and troughs due to measurement accuracy, rather than looking at the overall trend.

Woops!

--------------------------------

Hi folks,
I've been doing some testing on the effects of camber. Thought I'd share them in case anyone is interested. (sorry if the graphs don't work, I have tried to make them compatible with most programs, but I am using excel 2007 so I hope it worked)

Basically I tested braking distance, cornering grip and acceleration time for camber angles between 0 and 2 degrees to see if I could find any general trends (the first test went up to 4.2 degrees, but nothing interesting happens up there, so after that I only went up to 2 degrees because the tests are very time consuming). Tests were conducted on the Plymouth Superbird (first with SM tyres, then with RH) and Opel Speedster Turbo (RM tyres).

Acceleration grip was tested using 0-400m times.

Braking grip was tested by measuring stopping distance (using the data logger) from 250-100 km/h. The Opel Speedster brakes are not powerful enough to cause the rear tyres to skid (hence the test results would not be accurate) so I was unable to test rear braking on this car. Anyway, the rear braking tests are not very good in general because the results are so close to each other.

Cornering grip for the front tyres was tested by lap times at Motegi Twin Ring. So that I could plot all the results on the same graph (which turned out to be confusing so I removed it anyway), the numbers in the table are lap times minus 30 seconds (eg 13.3 = 43.3s lap time). The car was set up to heavily understeer, so that any increase in front grip would lower the lap time.

Cornering grip for the rear tyres was not possible. I tried to do a similar test with a heavily oversteering car, but the lap times were all over the place (and I kept spinning!). Also, an oversteering car is more agile, so an increase in rear grip would not always mean a faster lap.

Feel free to question my testing methods, no offence will be taken! And I'm interested to see what trends you folks find.

Simon

 

[Parnelli Bone]

As usual, I'm always interested. At work now, though, and we're not supposed to download stuff, or even unzip files.

 

[sukerkin]

An interesting way to try and codify Camber as a tuning variable.

Years of experimenting have shown me that, unless you have the car set-up fairly well to begin and are using low grip tyres, with then it is very hard to detect any effects resulting from camber change.

It will be fascinating to read your conclusions once you finish your research. I tried something similar a few years ago and came to the conclusion that the changes were so small compared to my favourite variables {damper and spring tuning} that anything ball-park was good enough .

 

[nomis3613]

As usual, I'm always interested. At work now, though, and we're not supposed to download stuff, or even unzip files.

Cheers, I'm looking forward to your comments.

...came to the conclusion that the changes were so small compared to my favourite variables {damper and spring tuning}...

Yeah, they don't have much affect on the overall balance. But I reckon one might as well start the tune off with the tyres working as efficiently as possible (before using springs, stabilisers, etc to distribute the available grip between the front and rear ends).

 

[nomis3613]

Edit 13 September 2013: This is all wrong! Updated results below: https://www.gtplanet.net/forum/showthread.php?p=8811231#post8811231

--------------------
Hi,
So, using the results in the first post, here are my theories about GT4 camber:

- there is a camber angle where acceleration grip, braking grip and corner grip are all at their maximum. (sometimes one of the types of grip has more than 1 angle where it's maximised, but there's always a camber angle where all 3 are maximised.) This makes tuning much easier- find the maximum of one type of grip and you have found it for every type!
- the optimum rear angle is about 0.5 deg less than the optimum front angle. So once you've found one, the other end is much quicker to find.
- racing tyres have slightly lower optimum angles than sports tyres
- optimum camber angles are usually* below 2.0 deg.
* if anyone wants to confirm this by testing it for every car in the game, be my guest!

I was also wondering if maybe the cornering peak is always at the 2nd peak for acceleration/braking, but I guess there's not enough data yet to prove this is true.

So the way I'll be tuning for camber now is as follows
1) tune front angle using understeer method (although this test is time consuming, it is very reliable and this way you are directly measuring the most important effect of camber)
2a) for rwd cars, check the rear angle using acceleration (this test is very quick, and reliable if you do it right. The range of angles tested can be reduced by focussing around 0.5 less than the front angle)
2b) for fwd cars, check the rear angle by measuring the braking distance (I'm not a fan of rear braking tests because the camber has minimal effect, but there aren't really other options. Luckily the rear camber angle isn't as critical for fwd, because the cars usually have an understeer problem anyway)

 

[nostuner03]

I was looking at a porsche 911 gt2 from the 90's and noticed that it had huge amounts of camber front and rear. It made me wonder if that was a contributing factor to why my Ruf btr handled like crap. So i added the same of camber(based on what i seen on the video)which was around 4.0+ degrees all around. It improved the way the car cornered, braked, and accelerated.

 

[sukerkin]

Never be afraid to try things observed in the real world or to think outside of the accepted wisdom when all the logical adjustments have failed .

It is by such a route that I found that raising the ride height at the rear for a catastrophically over-steering MR made things better .

 

[dudejo]

is it possible that LESS camber is needed with better tires? that's what i understand from the excel file, anyway...

 

[nostuner03]

Yeah. From my experience though, it depends on the vehicle, drivetrain, and power...

 

[nomis3613]

- optimum camber angles are usually* below 2.0 deg.

So i added the same of camber(based on what i seen on the video)which was around 4.0+ degrees all around. It improved the way the car cornered, braked, and accelerated.

Yeah, I probably shouldn't have said the above, since I've only tested a handful of cars. I deliberately included an old car (the Plymouth) in the testing to see if it would throw up strange results (it didn't), but still the range of cars I've tested is very limited.

In GT4, there are always exceptions to the tuning "rules". And like Sukerkin says, don't be afraid to try a crazy setting, it might just work!

 

[nomis3613]

I was looking at a porsche 911 gt2 from the 90's and noticed that it had huge amounts of camber front and rear. It made me wonder if that was a contributing factor to why my Ruf btr handled like crap. So i added the same of camber(based on what i seen on the video)which was around 4.0+ degrees all around. It improved the way the car cornered, braked, and accelerated.

Just wondering if you are able to dig up which upgrades and settings you were using (especially the type of tyres)? Here are my results, I used most of the engine upgrades, default tyres (sport medium IIRC), FC suspension and left all of the settings as default.

I couldn't find any improvement in grip about 4 deg. Not that I'm saying your settings are "wrong", whatever works for you is best.

 

[nomis3613]

Has anyone noticed that increasing front camber on a rwd car increases turn-in but doesn't reduce (perhaps even increases) power understeer? I'm not really sure, but I have a feeling this was happening as I was testing front camber on a RUF. Just wondering, has anyone noticed anything similar?

Or, theoretically, is there anything in either real life physics, or our understanding of the GT4 model to suggest this might happen?

 

[Leonidae@MFT]

Power understeer in a rear-engined car is just natural since the light front end doesn't have any weight to push the tyres on the road, even less so when higher camber value is applied. You need to remember how acceleration and deceleration effect on thee weight shifting.

 

[nomis3613]

Power understeer in a rear-engined car is just natural since the light front end doesn't have any weight to push the tyres on the road, even less so when higher camber value is applied. You need to remember how acceleration and deceleration effect on thee weight shifting.

Of course!! Thanks heaps. I'll keep a lookout for the same effect on the rear tyres for front-engined cars when braking.

 

[zgoldmen]

These tests are great and all, but there is still a big part missing as i race cars in real life, and in some cases i have seen race cars use up to 9 degrees of camber, really what the camber is adapting to is the type of track but how camber works from my understanding is that the lower the car is and of course the springs would have to be a little stiffer as you go down, the less camber is needed as the tire does not roll over on itself as much, however the softer or stickier the tire is, the more able the tire will be able to roll over as you will go faster through corners, so with that the car may need to be stiffer and have slightly more camber, generally what you are trying to do is balance the front and rear, for a front engine front wheels drive car a general starting point in real life is to have the back completely straight and have the front cambered in 2-3 degrees and 0 toe. What this is essentially doing is taking some of the stress off of the back of the car since it will understeer massively otherwise. However on the porsches this does not work the same way. There is no set way to determine a cars camber settings without testing because each car has a completely different way of handling, and although not in the game tire structure also has a big part of camber, but the track is one of the biggest factors, a track with fast sweeping corners will require more camber than one of slower corners. Think about what you are actually doing to the tires. And basically if you camber the tires in too much, say 12 degrees for a honda civic, what will happen is that you will constantly be on an edge of the tire, rather than through the corner the tire flattening out and having full contact patch.

 

[nomis3613]

Hi zgoldmen, thanks for your post (sorry I didn't reply earlier, I didn't realise anyone had replied).

When you talk about the "tyre rolling over itself", do you mean a) sidewall flex (which I know very little about) or b) camber-gain due to suspension (which I also know very little about!)?

These days I prefer to use braking and acceleration tests to measure the optimum camber angle (because these tests are much more repeatable than trying to measure cornering grip). But this assumes that the three types of grip (cornering, braking, acceleration) are effected by camber the same way in GT4. I've tested this in the past, but still not 100% sure it's correct.

What methods do other people use tune the camber angle?

 

[GTsail]

Greetings nomis3613 -
I appreciate all the work you've done with your tuning guides. I've read most of them and they really add to our knowledge base.👍

For camber tuning, I think you are putting too much emphasis on accelerating and stopping tests, and not enough emphasis on speed thru corners.

I think what you've found is that a car will stop the best with a small amount of camber because this would mean that more of the tire is in contact with the track.

I think what is missing is how much grip is gained or lost because of the camber angle going around a turn.

From all the driving that I have done, I've found that having more camber than stock at certain tracks, really improves my cornering speeds and therefore significantly lowers my lap times. Two tracks that spring to mind are the High Speed Ring and the Motegi Superspeedway. At both of these tracks, having more camber really helps my cornering speeds.

Alternatively, at some tracks I find that excessive camber is detrimental to lap times even if it might increase cornering speeds a little. An example is New York, where I find that I don't want the camber angle to be very high. My theory at NY is that due to the high number of sharp 90-degree turns, having the car tuned for accelerating and stopping, is what's best for fast lap times.


When I race a Championship Series, I always adjust the camber a little between each track to make is suit the track. I have not done any specific testing of this, but its just what I've been finding after all the miles I've driven in various GT4 cars.

Respectfully,
GTsail

 

[nomis3613]

From all the driving that I have done, I've found that having more camber than stock at certain tracks, really improves my cornering speeds and therefore significantly lowers my lap times. Two tracks that spring to mind are the High Speed Ring and the Motegi Superspeedway. At both of these tracks, having more camber really helps my cornering speeds.

Alternatively, at some tracks I find that excessive camber is detrimental to lap times even if it might increase cornering speeds a little. An example is New York, where I find that I don't want the camber angle to be very high. My theory at NY is that due to the high number of sharp 90-degree turns, having the car tuned for accelerating and stopping, is what's best for fast lap times.

Hi GTsail, thanks for sharing your thoughts. Yeah I'm now thinking that my cornering test results were dodgy and a few people are saying "higher camber = better high-speed cornering", so clearly this is something I need to re-test. Thanks for pointing me in the right direction.

 

[nomis3613]

In case anyone still keeps an eye on this dusty GT4 forum, I've repeated the camber testing and got quite different results! This time the test cars were Honda Integra DC5 (FF) and Lotus Elise (MR).

Traction
Measured at the dragstrip with AT gearbox and traction control. With careful tuning of power, gearing and traction control; the results are accurate to the nearest 0.001s. (if the AT gearbox allows the engine to hit the rev limiter, then consistency goes out the window)

0-400m results:

A 0.02 second difference probably means the difference in traction won't be noticeable on the track, but for what it's worth there is a slight decrease in traction as camber is increased.

Braking
Same as my OP (from years ago!), braking is measured at Twin Montegi ring using the data logger. I brake from 210km/h to 80km/s, then measure the distance for 200-100km/h. The data logger updates the speed at 0.5m intervals, therefore the accuracy for these tests is 1 metre (0.5m for both the initial and final positions). To improve the accuracy, the distances shown are average results from a couple of tests, however I would still be reluctant to claim an accuracy of better than 1 metre.

200-100km/h braking results:

All results are within 1 metre, therefore I believe the effect of camber (between 0-4 degrees) on braking is negligable.

Lateral grip
Now for the interesting stuff But also the hardest to measure.

Various methods were tried (see Raw Data section), but in the end I stuck with the same method as last time: set the car up to moderately understeer, then measure lap times at Montegi Speedway. Each setup was run until I felt the best lap reflected the car's potential with me at the wheel, which was usually after 5-10 laps. I wish I was a more consistent driver! Small differences in line could improve the lap time by 0.2 seconds, so unfortunately some setups will have their time set by a "fluke lap" and others won't. Therefore I estimate the accuracy to be 0.2 seconds.

Front lateral grip results:

The results are quite wonky, we'll need to look at the raw data (below) before making too many conclusions. But the graphs indicate:
- zero camber is not optimal for cornering
- sports tyres work best at or below 1.6 degrees
- racing tyres perform well all the way up to 4 degrees

Rear lateral grip results:
Ummm...errrr.. Given how hard it was to get consistent results for dragging an understeering car around by its nose, I suggest it would be impossible to get meaningful measurements of rear grip!

Raw data
The best result for each test is shown in bold. Dubious results are in italics.
Racing Qualify (R5) tyres- FF results: (the Elise didn't have enough horsepower to provide meaningful results on these tyres)

Note 1: I suspect the Integra didn't have enough horsepower to push these grippy tyres to their limit at Montegi Speedway. Therefore the grip gained from camber settings would not show up as lap time improvements.

Racing Medium (R3) tyres- FF on the left, MR on the right:

Note 2: Due to lack of horsepower from the Elise, Racing Super Hard tyres were used for this test instead. I suspect even these tyres were too grippy to start with, for the results to be meaningful.

Sports Soft (S3) tyres- FF results: (I didn't bother testing the MR with these tyres, since other results showed the same trends for Sports Softand Sports Medium)

Sports Medium (S2) tyres- FF on the left, MR on the right:

Just a quick comment on the acceleration testing, for one setup I noticed a strange dip at 3.2 degrees. But re-testing the setup with a small gearing change brought back the linear trend seen for all other setups. Therefore I believe this anomoly was caused by the balance between the car's powerband and the available traction (eg increasing the grip was bogging the engine but 3.2 degrees was a "sweet spot" of traction and slip for this setup).

Also, I should justify why this time I tested at intervals of 0.8 degrees, instead of 0.2 degrees for my earlier test. I repeated an understeer test from 0.4-1.2 degrees in 0.2 degree increments. The results corresponded well to testing conducted in 0.8 degree increments, ie following the same linear trend. So I believe that the peaks troughs seen in the earlier testing were due to measurement accuracy rather than any real phenomenon.

More lateral grip stuff
The failed methods I tried were:
- Lap time as various circuits. Unfortunately the variation between laps was far greater than the variation between settings.
- Maximum speed to hold a given radius at the skidpad. All results were within 2km/h of each other, so this was also not accurate enough. (the minimal effect of camber on skidpad cornering suggests camber has little effects either at low speed (~65km/h) or on unbanked roads.

Another issue with Montegi is that the corners are banked to varying degrees. Subjectively, it felt like racing tyres at high camber gave more grip on banked (cambered) roads but the trade-off was less grip for flat corners. This trade-off may explain the lack of a trend for racing tyres.

So, my (not very well supported) theories for how camber affects lateral grip for racing tyres are:
- The optimum range is 0.8-4.0 degrees.
- Within this range, the effect of camber is minimal.
- Higher camber may be effective at high speed (or perhaps banked) turns, but the trade-off is less grip for low speed (or perhaps unbanked) turns.

 

[shotamagee]

Always good to have additional info added to the forum, esp as I'm attempting the MR Championship in an Elise at the moment.

I normally start with 3.0 front & 2.0 or less rear camber settings for cars with an understeer bias on Normal or Sport Tyres, no hard data to support this, its just what feels like it helps "turn with the rear" a little more.
Though now I may try reducing these values as per your data to see if that is better.

 

[nomis3613]

Wow, someone read this! Thanks shotamagee, all the best for the MR Championship.

 

[BobK]

Not to worry, he's not the only one still reading.

 

[Bopop4]

Lurking as well.

(Damn, I really need to increase the camber on my cars.)

 

[uncle_pollo]

i leave camber and toe to zero, fronts sofyer and higher than rears ... on the loosey goosey side