Tire pressures ... very good article by Neal Tovsen

[Reijo]

A very well written article by Neal Tovsen here:

http://www.conecoach.com/2015/03/11/tire-pressures-finding-the-sweet-spot/

Reijo

[E6Cueman]

I think it's helpful to start some thinking... but is there is a problem?

What if my car is stuck with zero camber and a terrible camber curve because it's an old car with a mac strut.  Is 55psi really the answer?  Doesn't that pretty much guarantee I don't ever get to use the full potential of the tire (not in acceleration, not in braking, not in cornering). 

Also, if I happen to be one of those angry drivers who just keeps turning the wheel when the car doesn't do what I want, I might be wearing the names off my sidewall no matter what.

[MurrayPeterson]

What if my car is stuck with zero camber and a terrible camber curve because it's an old car with a mac strut.  Is 55psi really the answer?  Doesn't that pretty much guarantee I don't ever get to use the full potential of the tire (not in acceleration, not in braking, not in cornering). 

I am not sure that I see a better solution.  If my car had huge horsepower, then tip toeing ariund corners to get more power down would be fine.  But on "normal" cars, the best solution appears to be setting your pressures to allow maximum cornering -- if that happens to be 55 psi, it sucks.  But I still think it would be the fastest overall solution. 

[Reijo]

Yes, I remember a 2002 Nissan Sentra SER - Spec-V that I ran in the 2003 Nationals ... 55 psi was too much ... lost too much ... but 53 psi worked .. Mac suspension ... street tires without enough sidewall stiffness ... that is what worked.  not ideal for some scenarios but somehow there was a compromise with roll-over and other consideration factors ... We had maxed' the camber but it was not adjustable ... even loosened all the suspension bolts and "pulled" the struts in ... and actually gained 0.5 deg. of -'ve camber ...   That made a noticeable difference.


R

[Sprockett]

Is he suggesting varying pressures using the triangle on the sidewall as a measure of where the tire should roll over to?

I'm pretty sure the triangle is there to point to where the wear bars are located in the tread? Kinda of irrelevant to rollover?

[MurrayPeterson]

The triangle does point to the location of the wear bars, but it is also placed at the point where the tread meets the sidewall.  Pretty good place to stop rollover.

[Reijo]

I believe he mentioned that as a potential starting point when playing with air pressures.

Personally, I'd start with around 1/8 to 1/4" rollover .... e.g. very little rollover.   I know some like less pressure ... but I like to optimize right off and run as much pressure as possible (sort of knife edge feel) before starting to lose performance.....but also so the pressure does not increase much beyond the optimum during a particular run ....

And, every car/tire is a bit different and also varies with suspension set-up etc.

Biggest point, as I'm sure you know, is to not be riding on the sidewall .... nor too high so the tire is riding on the middle of the thread...

R

[Sprockett]

The triangle does point to the location of the wear bars, but it is also placed at the point where the tread meets the sidewall.  Pretty good place to stop rollover.

I don't think there is any standard on where the triangle is placed on relative to usable tread on a tire so as a universal point of reference across all tires, they seem all but completely useless.

[majormojo]

Viewing tire rollover is a nice place to start and is good to help a beginner get some starting pressures. But to properly understand what the tire is doing and what / how much suspension adjustment is required, you really need to look at tire temp measurements. It's hard with autocross, because we often don't get tires up to optimal temps and virtually nobody invests in real-time data acquisition for tires, so we have approximations done after the runs, but it's still more valuable than just looking at the tire edges and assuming that no rollover = max grip.