I'd say it's not what you believe, but what you measure, as Ogami Mushashi points out.
I have in front of my eyes the figures from a 1948 study by Shelburne for locked brakes (forward skidding, so to speak), which represents a minimum for friction coefficient, but a realistic proposition for someone who slams on the brakes before ABS. Data was collected in 1947, so it's not pre-war, unless you count Korea.
It has 11 curves or so (the graph is a mess, I swear, made long before Excel graphs), for combinations of worn-good tread, wet-dry surface, asphalt-concrete and three kinds of "regular" tires.
The curves are parabolic against speed, following the empirical relationship (more or less) of:
Stopping distance = Speed squared/K*Friction Coefficient,
where, roughly speaking, K = 30 for distances in feet and speed in mph.
First curious thing (for some, I imagine): the skidding coefficient varies inversely with the square of the speed! It's not constant, as you learn in Road Design 101. I won't delve into it, or this post will become unreadable even for myself...
Second, to answer the question (in two answers, I confess):
a. The maximum measured value, at 20 mph, good tread, dry surface, asphalt is
0.8. Conversely, the minimum measured value, at 50 mph, asphalt, worn tread, wet is
0.3
b. The "legal" maximum values you can reasonably use in design are
0.4 at 20 mph and
0.28 at 70 mph.
Schulze and Beckman arrived basically to the same conclusions in 1962 ("Friction Properties of Pavements at Different Speeds"), but on German pavements. Their maximum-minimum values, measured with a locked wheel trailer, go between
0.9 at 20 mph, dry pavement and
0.17 (or so, it's a graph) for 50 mph, wet.
I emphasize again that this are skidding figures, but it's a start...
Finally, (really!
) for modern cars the figures vary a lot, when you measure them.
Here you have an Excel file I already posted three or four times:
http://ciropabon.googlepages.com/Catalu ... tution.zip
I made it very quickly, using simple equations. The data is for Catalunya, back in 2005, before the last modifications to the track. Reca provided me with the link to a Brembo page for the stopping distances and I calculated the friction coefficients.
I deduced the lateral coefficients from the speeds posted by FIA at their map of the track, using the average geometrical radius, instead of the true car trajectory radius. I measured it in AutoCad and Eagle Point
(sorry, no Catia!). A lot of assumptions, but hey, it's engineering (or something similar).
To sum up, you get a friction coefficient from
4 at the entrances to (old) Elf, Repsol, Seat and La Caixa to
2.5 at Banc Sabadell, with 3.5 at Wurth curve.
For lateral friction coefficients the variation I find is larger: it goes from
1.6 at La Caixa to
3.5 maximum at Wurth. This are maximum values, as the cars use a larger radius trajectory than the actual radius of the curve, and this difference between real and measured radius is larger for tighter curves.
As this year have appeared the trajectories at FIA site, then you could try to do the same exercise measuring the true trajectory radius, which means you'll get lower figures.
Anyway, I don't get anything close to 5 Gs. On the other hand, I'm sure that this number is a peak value, not the effective friction coefficient for the whole curve: I guess this is what the pilot feels, not the effective lateral force that the wheels are able to develop through the curve.
Just in case, here you have the reported entrance speed in Km/h vs radius in m, as I calculated:
