Solo Heat

» Posted by on Aug 6, 2012 in Current projects and Tech Notes, RB6 Formula Ford | 2 comments

Solo Heat

When a tire is slipping on the pavement as a result of hard cornering, braking, or accelerating, that slipping heats the tread rubber. A tire generates the most grip when that heating is uniform across the width of the tread because every point on the tread generates the same force on the ground. That heating is a good thing because racing tread rubber generates the most grip at a temperature that is higher than ambient. Of course there can be too much of a good thing, too. Autocross tires are designed to perform best at a lower temperature than track racing tires because an autocross run is over before a pace lap is halfway around the track.

The common method of measuring tire temperature is a needle probe pyrometer that is stuck into the tread as soon as the car stops. Normally, 3 temperature measurements are taken on each tire: inside, middle, and outside. Comparing the three temperatures across a tire provides guidance for optimizing tire pressure and camber settings, and comparing temperatures from front to rear is an indication of cornering balance. The upsides of a tire pyrometer are that it’s relatively inexpensive and easy to use. The downsides are that tire temperatures change rapidly, the measurements only reflect the time-weighted average conditions over the most recent segment of the track, and the temps can only be measured after the car is stopped in a safe place.

Of course there is a fancier way to measure tire temps: a series of non-contact infrared temperature sensors mounted on the car, with the sensor outputs recorded by a data logger. The upside of this technique is near-continuous, real time measurement of the tread surface temperature. That’s what matters the most, and it’s what the photo above shows. Freddie has one front and one rear tire instrumented with 3 sensors across each tire tread. The sensors are fastened to a square steel tube mount that is bolted to the upright, so steering and suspension travel don’t affect the relative positions of the sensors and wheel. The mounts are symmetric, so they can go on either side of the car. They are on the left front and left rear at the moment. Now we get to find out what’s really going on!

If you plan to use these sensors, here are a couple of notes: 1) The field of view is quite a bit larger than the data sheet shows. My first autocross with the sensors on 2 weeks ago resulted in bad data because the holes in the mount were sized to clear the advertised cone angle, but part of what the sensor saw was the mount. 2) The internal processor in the sensor only updates the output signal at about 3 Hz, so even though the logging rate was 100 Hz, the data has large stair steps. Of course the signal updates among the sensors are not synchronized, so the data for each sensor steps up or down independently of the timing of the other sensors.

This weekend, I ran another autocross at El Toro with much larger holes in the mounts, and got good data. There are a lot of real surprises in the data, and we all get to see them.

But first, here’s a look at the course that was generated by logged GPS data:

Here’s the video. The cross is at the start line and the map is color coded with speed. My maximum speed was 66 mph and minimum was 29 mph. The run time was 57.296 sec. Since there were more long left hand corners, the sensors would have provided more valuable data if they were mounted on the right side, but there was no way to know that before hand. I don’t have all four tires instrumented because each sensor plus wiring harness extension costs $220.

While I’m in the process of re-learning my driving skills, I’m running used Hoosier R35A compound road racing tires, so my maximum cornering acceleration was only 1.4 g. That’s because the tires weren’t up to their operating temperature for road course running. I’ve seen 1.9 g on flat road course pavement on hot R35As. Others have reported similar grip while autocrossing on R25Bs. When it makes sense, I’ll buy a set of Hoosier R25B compound autocross tires.

A really interesting thing I saw in the temp data was on the way to the start line for my first run:

Sunny/shady tire temps

The sunny side of each tire was 30°F warmer than the shady side. Now take a look at the data in the shut down area after my first run:

Sun/shade post run

5° of that solar heating was still there after the run! I predict that I have just started a new fashion trend in the grid, so I’ll name it: the Solar Roll.

Here are the 3 front tire temp sensors and lateral acceleration from my best run:

Front Temps Run 3

Color codes: Yellow = Inside front, Gold = Middle front, Red = Outside front, Gray = Lateral acceleration, right turn is positive.

The middle of the tire was only 70°F hotter at the end of the run that at the start, and it only warmed up that much because of the long braking zone and the long right hander at the end. For this run, the static camber was -0.5° and the pre-run pressure was 16.5 psi. If the pressure were set optimally, the middle temperature would be right between the inside and outside temps while cornering to the right. Since it’s higher than that average, the optimum pressure is lower than 16.5. How much lower? More runs at lower pressures will answer that question. If the camber were set optimally, the inside and outside temps would be the same. Since the outside temp was lower, I have too much negative camber. That’s a surprise since I’m only running -0.5°. So, I’ll run less negative camber next time, and I’ll keep adjusting camber and pressure until the three temps across each tire are the same while cornering with the sensors on the outside tires.

It’s interesting that the inside front tire either does nothing or gets cooler while cornering. In this case, the left front tire is the inside tire while turning left, which is negative on the graph above. That’s an indication of how lightly it’s loaded due to weight transfer. From this, it’s easy to conclude that the primary influence of the inside front tire is to tune the loads on the other tires, and it doesn’t contribute much if anything to cornering grip. I believe that this is why Ackermann steering is such a numb adjustment on cars that don’t produce massive aero downforce. It takes a huge change in Ackermann to produce a noticeable difference on most cars.

Here is the front tire temp data again, with longitudinal acceleration shown in green:

Front temps and Longitudinal Acceleration

The front inside and middle temps responded to braking at 81°F per second! Those rapid spikes were a result of near-locked brakes. The front outside temp didn’t follow suit unless the car was also turning right while braking. That’s another indication that Freddie has too much static negative camber.

I locked the left front a couple of times during earlier runs, and the front temps instantly shot past the 300°F measurement range of the sensor. I have seen data from other cars that showed the same result due to massive wheelspin.

The rear tire temps were very similar to the fronts in most respects:

Rear Tire Temps and Lateral Acceleration

Color codes: Light blue = Inside rear, Medium blue = Middle rear, Dark blue = Outside rear, Gray = Lateral acceleration, right turn is positive

For this run, the rear static camber was -0.5° and the pre-run pressure was 18.5 psi. Again, the data shows that Freddie had a bit too much negative camber and a lot too much tire pressure. More experimentation is in order here!

That’s OK by me, because my policy is to never run exactly the same setup twice: Every time the car comes to a stop, I change one thing, sometimes two or more if they are in unrelated areas. That makes every run or outing a development opportunity. Also, during every road course outing, I will adjust at least one bar and the brake bias while under way.

It’s interesting that the rear inside temp didn’t respond to left turns at all. It would seem that the primary influence of the inside rear tire is very similar to that of the inside front, but with the additional burden of resisting acceleration wheelspin since FF rules require an open differential.

Again, here are the rear temps with longitudinal acceleration shown in green:

Rear Tire Temps and Longitudinal Acceleration

Aside from a very small wheelspin spike at the start, the rear temps didn’t respond strongly to longitudinal acceleration or braking. It makes sense that the wheelspin spike went away very quickly since the surface heat was soaked up radially through the thickness of the tread rubber, which was much cooler.

Because the front temps did respond to braking and there was no evidence of a locked rear tire, the brake bias was front-heavy.

Comparing front to rear temps shows a happy story:

Front to Rear Tire Temp Comparison

Since the temps stayed nearly the same through the full run with the exception of the hard braking zones, the cornering balance of the car looked reasonably close to balanced. It felt loose in steady state cornering, which could be real or it could be a false perception on my part.

Just for your amusement, here is what the tire temps looked like during a spin:

Tire Temps During a Spin

Ummm, yes I meant to do that, just for your education. See how generous I am?

Despite that spin, I’m happy to report that my driving skill has advanced to the level of basic competence, meaning that I’m not horribly embarrassed and completely disgusted with myself every time I get out of the car. I managed to drive two runs that weren’t complete disasters, and I had fun doing it. There is a very long way to go before I’m competitive again, but I finally have a reasonable baseline to build on.

Update: More data! This is a graph of front middle temperature versus lateral acceleration. The outer bounds of lateral acceleration show a clear trend of increasing grip with increasing temperature. That’s not a surprise at all since the R35A compound is intended for road racing. An autocross run doesn’t last long enough to get the tire up into its designed operating temperature range.

Here is longitudinal acceleration vs front middle temp. Because there was only one straight-line braking zone, there isn’t enough data to show a clear trend.

2 Comments

  1. Hi Neil,

    Great article. I tried this a couple years back on my (now sold) Citation. Made a big mistake and had the inside and outside sensors too wide and so ruined my data. Only got to try once because the sensors were rentals. So it was great to see your data. Thanks so much!

    I see you’re not registered for Nationals. Did you decide to put it off for a year? Even though I’m not in the class anymore it is certain that the guys would get a real kick out of being able to talk with you and check out your car. If you plan to go, register soon or the reg fee goes way up.

    I’m in BMod this year, figuring out my Cheetah and trying to improve front downforce (big honking wing on the rear is probably too big). Also trying to get my driving into gear after forgetting how to drive without taking a break!

    Hope to see you in Lincoln but if not, thanks for your website.

    Sincerely,
    Jim

    • Thanks, Jim!

      My overly ambitious early plan was to run the solo nationals this year and the runoffs next year. Since other obligations have limited my track time, my pace of driving improvement has not kept up with The Plan. Also, my employment situation at the moment doesn’t mesh well with the time required to tow more than halfway across the country and back. I’m motivated, but I’m not going to go unless I have a viable shot at the gold. I’ll show up when I’m ready.