tire temperature

[MazdaMatt]

I'm doing a project at school regarding tire temperature. I'll be building a system that uses infrared sensors to read the distribution of heat across the tires of the school's Forumla SAE car. Do any of you have reference to any previous projects or studies regarding tire temperature distribution?

For those interested:
The system will have 3 infrared sensors mounted above the tire reading temperature. There will be an in-car display which will show the temp of the inside edge, center line and outside endge of the tire. There will also be a colored (red-green-yellow) thermometer-like display that will indicate when your tires are in the 'ideal operating range'. And of course, this info will be downloadable so it can be graphed in excel and compared with past information. The purpose is to tune one's suspension/alignment so that tires wear more evenly. even wear means more rubber is being used at any given time. Longer tire life and better grip at the same time!

For the purposes of the project I would love to hear everyone's ideas, opinions and especially references to preexisting information on the topic.

[Steven Scala]

I was crewing for Gianni Biral during his last full Ontario F2000 season. Gianni had rigged an aluminum bracket onto the left rear upright of his Formula 2000. The bracket held three infrared sensors feeding his 30-channel CDS onboard datalogging system. The peak temperatures get so momentarily high at peak .G's that it may be worth considering to let the driver do his thing, and record the data for later interpretation.

A used six-channel logger might be available for reasonable price. Should be more than ample for this exercise.

[Pete@Marcor]

I have done a fair amount of testing with tires and setup. My personal preference is to use a probe type pyrometer, but that is kind of impossible in this situation. The infrared pyrometer usually only reads the outside tread surface temperature, which is not ideal, as air will actually cool the tread surface down fairly quickly.

It is better to get closer to the core of the tread cap.

Regardless, this sounds pretty interesting, and has potential applications to lapping and Solo guys.

[bumblebee]

Complex problem but not unresolvable, previous points are right on, need to measure impact of air cooling as the tire rotates and the impact of vehicle velocity, simplest way without real messy math is to conduct a series of experiments comparing non-moving temperature to moving temp at various speeds to arrive at suitable constants, there have been a number of SAE papers published throught the years on this issue.

We use a physical probe that touches the tire rather than get fancy, we save the data acquisition memory for other issues that we cannot measure in the pit lane.

Steve

[MazdaMatt]

The air cooling brings up an interesting point...

but to be clear, this will be a constant log of temperatures as the driver drives. This is FAR superior to running 10 laps then coming in to check and see how your temperatures are by the time you finish a cooldown lap, get out of your car and walk around with a probe. This info can then be fed into Excel and you'll be able to watch how your tires change temperature distribution through each and every turn and straight away.

Also, we'll be purchasing no pre-built components. This is an engineering design project for fourth year and we're doing the actual design and building of all components.

If any of you have any reference to previous papers regarding this, please link me. I'll search as well.

[W Tarleton]

When I worked for a Formula Mazda dealer we had data from the Factory that had used a setup like that. In car will not be of any use, and the download data will need to have G-load, track position and gas/braking plotted along with the tire temps. Use that and change the toe and camber for several laps to come up with your setup. (wing if you use them) The tepms will suprise you instead of the 250-280 temps you may see 320 F instead.

[bumblebee]

Not really sure how useful the real time data is. Additionally, the sensors become inaccurate in a non-linear way the further they are from the tire surface, ideally you want them real close, then you risk a debris strike. I might have missed some points but a few thoughts are captured below;

I think that the main questions that you want answered in a road course scenario, where there is great variance in track condition/traffic, are:
- is my driver getting the tires hot enough? (biggest problem with amateur drivers)
- is my driver getting the tires too hot? (biggest problem when amatuer drivers get fast)
- are my inflation pressures correct?
- are my initial camber and suspension camber curves adequate in providing a reasonable temperature profile?
All of those questions can be answered in 30-60 seconds after one of your crew takes 12 tire temperatures in pit lane for less than $100.

Real time data perhaps more useful as it relates to tires are:
- did I lock up a wheel under braking? (outcome is flatspotting, difficult to pick up with a tire temp reading, on the track or in the pits but it kills the tire, knowing this you can correct brake balance or coach the driver)
- how much do I spin the rear tires? (real problem in high power GT cars, if you monitor throttle position and wheelspin you get something that the driver can use or re-gear the car)
- how much are the front tires turned versus the direction the car is headed? (great for evaluating over/understeer and a driver that over-drives)

I am not a great driver but approaching optimization of tires from this perspective got me a whole lot better pretty quickly.

Steve

[MazdaMatt]

Woah now... i'm not trying to make an F1 telemetry system. I'm trying to get realtime tire temp measurement. Scope is important, I only have about 5 months and I don't have a million dollar budget or a team of experienced Ph.D vehicle dynamics engineers.

Points regarding tire temperature, very useful to me, thanks. The distance from the tire is definately an important factor. Since this will be attached to a Formula SAE car, I am actually having the design team give me mounting points inside the tire and we will later design a bracket that will come up around the tire and over the surface. I expect to be within 2-4 inches (note: we will be trying to get a sensor with a 1:8 spot:distance ratio, so that will lead to a 1/2-1/4 inch strip of tire that wea re measuring).

The prime goal of this system is to evaluate the distribution of heat from inside edge->center->outside edge of the tire. G meters will not be needed for this comparison because a simple qualitative approach can be taken. "I see that there is a HUGE spike in inside edge temperatures when I was on the back straight, but in turn 5 we don't even get event temps on the loaded wheels, we need to cut back some camber". Since this realtime system will be datalogged, we will have a time vs temperature graph for each strip of tire. Its easy enough to know by your times where you are on the track, or even without times, you could evaluate the changes in temperature and notice when you are turning right, left or going straight.

For those who like to extrapolate future additions to this system, we could even add a G meter, pitch/yaw meter AND GPS. This would allow us to map the path of the car in all 3 axes (including elevation) and also the distribution of weight due to hills, turns, brakes, gas. Wouldn't you love to see a 3d image of a ribon indicating the exact path and angle that your car took through mosport? This ribon would also have 3 colored strips ranging from cool blue to hot red that vary as you take each corner and etc. This would also allow us to compare G's to temperatures, G's to body roll, avg temps to lap times, etc etc etc etc etc etc (etc).

but i degress... i'm making a tire temp system. that's it. Extra discussion is fun and you can sure bet we've discussed ideas far beyond the scope of our project, but its just two students making a circuit.

As for the in-car display, yeah, you'd think its useless to a driver, but we're going to have a thermometer-like display showing average. Some of you may recall when (on a warmup lap) i launched off the DDT and took a scenic drive up a pole... I don't think that would have happened if i had looked at a display and it showed that my tires were still cold. Confidence makes drivers faster, too.

[MazdaMatt]

When I worked for a Formula Mazda dealer we had data from the Factory that had used a setup like that. In car will not be of any use, and the download data will need to have G-load, track position and gas/braking plotted along with the tire temps. Use that and change the toe and camber for several laps to come up with your setup. (wing if you use them) The tepms will suprise you instead of the 250-280 temps you may see 320 F instead.

Any reason that the tires would be 50 degrees above expected? Toyo says that their RA1's get very greasy after 300F. Could it have been a calibration issue or were the tires actually getting to dangerous temperatures?

[Steven Scala]

I was speaking this week with Robb Smith and Peter Carpenko about achieving cornering balance in front-drive race cars. Both Smith and Carpenko prefer to overinflate the rears to achieve balance (with yaw), admitting less rear tire compliance with the road surface. I prefer to flex the rear sidewall and adjust the car's attitude within a wider rear slip angle, whereas Smith and Carpenko both noted the difficulty of balancing on the tire's self aligning torque.

(Smith has credibly been among the fastest Sentra racers, and Carpenko currently pilots the fastest front-drive GT car in the country...)

Anyway, we all would be interested to know the differences in surface versus carcass temperature in overinflated versus underinflated tires, particularly in the case of the often-underloaded rear tires of front-drive racecars.

My HADA peeps would also be interested, I'm sure. :D

[bumblebee]

Matt, good response and it is a good project for a student (I am a prof so I have some perspective). A simple go/no go light (for driver temp) rather than a guage may be best in the cockpit, it is a pretty busy place. You could even do cold/good/too hot light system.

Camber curves are going to have a large effect on where you place your sensors, don't settle on final location until you have moved the sensors around a little to validate the best position. If you have time for a little math you can capture wheel lockup by looking for a heat spike, you can also capture wheelspin the same way.

Make sure that your sensors are robust, vibration kills most of them. Once you get it working, take a look at how much heat an unbalanced tire/wheel at speed generates....

Good luck!

Steve

[MazdaMatt]

Matt, good response and it is a good project for a student (I am a prof so I have some perspective). A simple go/no go light (for driver temp) rather than a guage may be best in the cockpit, it is a pretty busy place. You could even do cold/good/too hot light system.

Camber curves are going to have a large effect on where you place your sensors, don't settle on final location until you have moved the sensors around a little to validate the best position. If you have time for a little math you can capture wheel lockup by looking for a heat spike, you can also capture wheelspin the same way.

Make sure that your sensors are robust, vibration kills most of them. Once you get it working, take a look at how much heat an unbalanced tire/wheel at speed generates....

Good luck!

Steve

1) We will be using about 8 LED's, 3 yellow for too cold, 3 green for the mfg spec 'ideal' temp range and 2 red for too hot. Should make it a nice simple clear indicator for the driver.

2) At least on the SAE car we are going to mounting them directly to the upright, so it will go up and down and turn side to side with wheel. Clever, huh... we were first thiking in front of the tire, but as the wheel turns you point to the wrong place, then we thought above the tire, but as the suspension travels it hits the sensors and you see a few hundred bucks and a few hundred hours of development skitter across the ground in the rearview... so where it is, its basically locked in with the wheel position.

3) Robust... yeah, we are going to spend 100usd on each of the three sensors. These babies are used for industrial process control. If it breaks i'll be paying an unpleasant visit to the vendor. Also, since stuff always flies off of a tire tangent to its surface (simple physics), we will have a shield protecting the body of the sensor, and the lens will be a little bit inset in a tube, so it will be safe.

[miataboi]

other than logging it for looking at temps offline to correct suspension geometry and alignment... (after it's downloaded and interpreted) I see little / no use for the driver....

And in an SAE car... things happen so fast there I doubt he'd even bother looking... then if it's too cold... or too hot... then what?

[MazdaMatt]

SAE driver will not have time to look at tire temps. Temps are always too cold at the start and i'm sure they will be too hot by the end. That is more for the guys back in the shop to decide on alignment and even reconsider tire choices.

After SAE is done with this, i will be ripping it off the car and mounting it to my own where i WILL care about the avg temp meter. As for the LCD display showing the numbers, that's more for the passenger or for test purposes to make sure that things are actually going on.

[malcolm]

regarding the talk of "300+ degrees being too hot" for some tires, remember you are referring to a temperature taken when you are in the pits. 300 deg in the pits likely means you are having peaks of 350~380 in the corners... So when Toyo says 300 is too much, that means that 300 in the pits is too much since the spikes would be much greater.

I like the idea though. Is there any way you can have the sensors closer to the contact patch with a robust sheild to protect from debris? might be more accurate that way. Might also be worth looking into making the arm rotate around the wheel to see if the location makes much of a difference in measured temps... (if there isn't, then you could obviously move it into a higher and safer location).

I prefer my fourth-year project: designing a high-downforce wing for a formula car. :D (I love aero)

[MazdaMatt]

regarding the talk of "300+ degrees being too hot" for some tires, remember you are referring to a temperature taken when you are in the pits. 300 deg in the pits likely means you are having peaks of 350~380 in the corners... So when Toyo says 300 is too much, that means that 300 in the pits is too much since the spikes would be much greater.

I like the idea though. Is there any way you can have the sensors closer to the contact patch with a robust sheild to protect from debris? might be more accurate that way. Might also be worth looking into making the arm rotate around the wheel to see if the location makes much of a difference in measured temps... (if there isn't, then you could obviously move it into a higher and safer location).

I prefer my fourth-year project: designing a high-downforce wing for a formula car. :D (I love aero)

Good point on the tire readings... i guess toyo doesn't expect that they are marketting to auto-techno-geeks like myself. I like your project idea... obviously your a mech eng, yes? I ahve been speaking with the SAE guys and i have told them that we need to be 'Under 8 inches' from the tire. We will design it as closely as we see safe. We wouldn't want to hit a bump, have the bracket wobble and touch a tire that is rotating at 100km/h. Distance from the tire is actually directly related to risk of debris. The further from the tire, the higher the likelyhood of a stone (at 100km/h) entering the siting tube. Closer = safer except when we hit the rubber.

I will keep this thread updated as the project goes on... and for all you greedy chumps trying to make a buck, consider this a copywrite notice. *shakes fist*

[malcolm]

haha, yeah, fourth-year mech-eng at queens. my project is also part of an international collaboration project run by PACE. Schools from all over the world are building different parts of a formula car and then we're all going to utah to assemble it in the summer (it's not an SAE car... it's just a F-Libre car, essentially... hmm, maybe I should convince them to bring it to race at Mosport??).

Yeah, tomorrow I am going to try to get Fluent to work to give me some numbers and pretty CFD pictures. :)

[W Tarleton]

The temps are much higher using a infared tester because it reads surface temp.I spoke with the Goodyear tire engineer John several times about this subject. When they first used an infared tester they could not beleve the temps they saw. John says that a probe is better to find the temps. He even told me to get a close to the cords as you can with the probe. I still think you should spend your time on recording other things rather than building an in-car readout. You could record steering wheel angle and RPM easily. And it would be of more use.

[MazdaMatt]

The project is half done, i'm not going completely change it now.

Thanks for the info from the goodyear eng. The problem with probes is that it is simply impossible to have probes on your tire while you are driving around the course. Sure you can pull into the pits and check your tires, but that won't give you the temperature gradient than you experience at turn 4.

The key is to focus on the goal. This project is going to be used for alignment and suspention tuning. To do this, the focus is on *gradient* of heat across the tire. RPM has nothing to do with that, and to a point, the core temp has nothing to do with that as it is much more consistent. Steering angle may be useful, but my level of analytical skills just don't agree.

Innovation means doing something that isn't already common practice.

[Mike Basden]

...This project is going to be used for alignment and suspention tuning...Steering angle may be useful, but my level of analytical skills just don't agree.

I knew of a CART engineer that did some testing with this some time ago (he's since past on so I don't think he'll be of much help to you) but I did have some notes from my experiences with him. In the end, I think he had 7 sensors across the tread and sampled at a rate that was equivalent to 1/8 rev of the tire @ ~150mph (you can do the math. :)) A lot of these points are rather obvious now and simple for an engineer to figure out, but as a kid, these were great notes! :)

Many points mentioned here are accurate. The tire temps at speed on the carcass will reach into the 300's (degrees F of course) or more depending on the G loading. They fall off extremely quickly once the speed and G's are removed. The rate of this change is non-linear and dependant on the tire composition, soak time at temperature, and ambient cooling characteristics (ie, nature of the car - open wheel vs closed). These temperatures will be very indicative of the tread temperature at depth but the core temperature of the tire has an inherent latency and is less succeptable to transients (makes sense right?). Any data you aquire will likely have to be filtered for high frequency transients caused by the pick up of debris but this depends on your sensor pick-up size and it's response time to resolve the variations of temperature.

Positioning the sensors at the leading edge of the tire (as it comes off the race surface) will provide the most accurate results and help shield the sensors from turbulance. Ensure the mount is very rigid so that the outer sensor maintains an equal distance from the tire or your data will be skewed. Sensors should be perpendicular to the tread surface for accurate readings. He used ~ 1/2" distance from the surface initially but moved them in a crowned alignment to compensate for cetrifugal movement of the carcass at speed (center having more distance than the shoulders).

The last note that I have, which I think is very important to remember for your project is 'Tire temps are just a tool to help you find the best compromise in the set-up. Don't chase your tail trying to make it perfect everywhere, just get it the best overall.'


From my experience I would say that steering is an underestimated piece of suspension tuning. You have to remember that the entire chassis is a dynamic and tunable system; 'system' meaning there are direct relationships between all of the elements comprising the system including the 'rigid' element of the chassis itself. Tuning doesn't stop at castor, camber, toe, springs and tire pressures.

I realize that it may be too much to look at given the confines of a 4th year project, but you should also note in your paper how you came to narrowing your variables and what other areas may have an impact on the design but were not investigated and why.

While most sedan race cars don't have the ability to adjust much as far as steering goes, (usually just rack location fore and aft, and toe) stock cars and formula cars look at and adjust for bump steer. Bump steer impacts the dynamic slip angle of the wheels under compression/extension and introduces dynamic scrub. In more sever cases, it can even contribute to a tire rolling over on itself. If you're only introducing more scrub from the bump, you may see this in the tire temps as exaggerated camber issues. Trying to hit the right camber set-up with bump steer problems will be difficult to home in on since you have another problem that is enhancing the appearant camber issue. If you look at more than OMI temps, you'll see that rolling over heats up the outer/inner edge of the tire very fast, and introduces an area of cooler transients from the reduced contact patch next to the 'roll' resulting in a steep temperature gradient near the shoulder of the tire.

Hope this helps.

[MazdaMatt]

thanks a lot for all that info, and for the tips for my report. i will definately bare those in mind as the project develops.

Also, since this is a public forum and you guys are all considered to be 'market research' and 'experienced racers' I will likely be dropping names in my report. If you don't want to be mentioned for whatever reason, please say so. Your name being in the report does not require you to back up your cedibility nor will you be contacted or likely ever hear about it.

If this report is worth reading in the end, i will probably post a link to it in its entirety when i'm done.

[nscirocco]

check out the IR sensors from :
http://www.solidstateracing.com

i attended a seminar by Claude Rouelle last week .. if you ever have the chance, you have to go!
he showed us acquisition logs of on track IR tire temps and then showed the temps in the pits .. sometimes they were such that the readings in the pits would have been totally backwards to what was happening on the track.

use IR data not to get the actual degrees, but use the data to compare with other runs when changes are made and to study trends.

-nigel.

[Andy]

Hi MazdaMatt,

As per Toyo RA1 spec. the ideal tire temp. is 200 F. As Malcom points out if you are seeing 300 F in the pits add about another X degerees for on track and you are way too hot.
The purpose of realtime tire temps is to see the tire temps in the corners. By the time the car is in the pits the temps are not relevant to on track temp. We are currently working with a small infrared device with on board Asic that is the same component as some of the F1 teams are using. Its size is about 10 mm by 15 mm. It is a stand aloan component and its output can easily be fed to some sort of data recorder. They are very accurate and use a narrow 20 degree beam. It would not be difficult to incorporate some protection in the mount that you fabricate. Our sensor will directly interface with our Motec ADL. Our plan is to use 3 of these sensors per wheel.
Do a little more research re tire temps. Max grip in the corners does not mean even tire temps, quite the opposite. For max grip in the corner you will allways see the inside of the tire approx. 20 F. higher than the outside. Given that you are using enough camber. Camber is your friend in the corners.
The component we are using is less than your $100 US. If you need info PM me directly. Which University are you attending and what year are you in?

Andy

[SE-R Racer]

Hi MazdaMatt,


Do a little more research re tire temps. Max grip in the corners does not mean even tire temps, quite the opposite. For max grip in the corner you will allways see the inside of the tire approx. 20 F. higher than the outside. Given that you are using enough camber. Camber is your friend in the corners.

Andy
Wow 20 degrees? I knew it should be higher on the inside but I thought 5-10 degrees was enough....back to the drawing boards...and some more camber needed :D

[MazdaMatt]

Again, thanks to everyone for the added information.

I am in fourth year (five if you include my internship) at the University of Western Ontario. I'm taking Computer Engineering.

I find it interesting that the greatest degree of grip is found with hotter inside temps at highest G's. I would have expected it to be even, but I assure you that once this system is in place I will not blindly adjust my alignment settings according to an assumption. All assumptions will be tested. The Formula SAE car does have recordable G's, TPS, and even suspension compression. These will all be compared to the tire temperatures to forumlate the final opinions that will be presented with the product. Your tip will be very helpful and I will make a point of testing maximum G's through a full scale of temperature gradients to make sure that we come up with a good conclusion.

It has also become apparent that the sampling rate will need to be higher than we initially thought. Originally we figured that an updated temp on screen and on LED bar each second would be adequate. Clearly a technical analisys of the temps will require a much higher sampling rate, especially considering it is going on a car that will be driven on an autox course.

Solidstateracing has some very interesting products, I will look further into those. Thank you.

Cheaper than 100usd!? PM 4 u!

-M

[Mike Basden]

I suspect that the 'ideal' temp differential is because of the accumulated heat induced by the camber while running in a straight line(?). If you have an even tire temp in the corner, that would suggest that you are actually over working the outside to make up the temperature difference found compared to the inner.

[MazdaMatt]

Clever one... very clever. I'm interested to see how turning after a straight compares to running the car on a skid pad. That'll clear that up :)