While many might consider that you need to use an insertion probe type thermometer to measure tire temps, F1 and NASCAR teams use real-time data acquisition with 3 IR temperature sensors per tire to get a live profile, and they learn several orders of magnitude more than what you might get running around the car in pit lane. I found an interesting application note that shows how to get the internal temperature of the tire by measuring its surface temperature using an infra-red sensor by solving a thermal balance equation since the thermal conductivity of the tire from the inside to the outside is a constant:

http://www.exergen.com/industr...8.htm

Also see:

http://www.exergen.com/industr...3.htm

and here are the Exergen sensors:

http://www.exergen.com/industr...x.htm




or this:



from here:

http://www.omega.com/pptst/OS136.html

Something for a winter project, and you feel you really want to go pro. You should see what a lap of a 3 tire temp profile looks like. Amazing changes go on on those contact patches, and they get way hotter than you would ever measure in the pits with a pin type probe.

 

Until I am in a position to need, own and use real time data collection systems for tire temps then I'll stick with my probe pyrometer and be happy with the data I get and tuning that it helps me with. As budgets and needs change then often do the tools.

 

Interesting information. And here we were arguing not so long ago that you shouldn't use an IR thermometer to accurately measure tyre temp.
I guess I'll keep using mine with confidence this weekend and so for.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by BETO736 &raquo;</TD></TR><TR><TD CLASS="quote">Interesting information. And here we were arguing not so long ago that you shouldn't use an IR thermometer to accurately measure tyre temp.
I guess I'll keep using mine with confidence this weekend and so for. </TD></TR></TABLE>

Your IR thermometer will only be acurate if you're taking the temps while on the track. If you pull into the pits after a cooldown lap it's not going to give you true temps because the outer surface of the tire will cool much more rapidly than the inner portion that the probe tests.

That being said, lets get someone with both to take temps with both at the same time and let us know the results.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by BETO736 &raquo;</TD></TR><TR><TD CLASS="quote">And here we were arguing not so long ago that you shouldn't use an IR thermometer to accurately measure tyre temp.
</TD></TR></TABLE>

People who argue this have generally not actually tried both an IR an a probe type at the same time on the same set of tires. I have both since I have both and it is a fallacy that you don't get good info from an IR type thermometer. First thing is that the tire's internal and surface temps have cooled significantly by he time you come into the pits, and whichever type you use you are measuring the tire temp after it has cooled. Newton's law of cooling says that the rate of heat loss is proportional to the temperature difference (tire to track/ambient), so the tire initially cools very quickly from its peak on track temperature by the time you get into the pits. It is as if the pit temps are a shadow of the on track temps. I suggest everyone try both types and write them down and figure out the thermal balance equation from that article above:

Tcore = K(Tsurface-Tambient) + Tsurface

Where Tcore is what you measure with the insertion type and Tsurface is from the IR type. For example, if

Tc = 200
Ts = 180
Ta = 80

find K = (Tc-Ts)/(Ts-Ta) = (200-180)/(180-80) = 20/100 = 0.20

where K is the slope of the line of a Tc vs Ts graph, Ts being the intercept. Tc measured with the insertion type is always higher than Ts measured with the IR type, but if you know Tambient and K, you know Tc is you know Ts. The difference between the two is always increasing, but within the fairly narrow range or a hot track tire, the change is not that much, and if you enter the data into a spreadsheet, you can get one from the other.

For our example above, here is what you get.

Ts Tc diff
80 80.00 0.00
90 92.00 -2.00
100 104.00 -4.00
110 116.00 -6.00
120 128.00 -8.00
130 140.00 -10.00
140 152.00 -12.00
150 164.00 -14.00
160 176.00 -16.00
170 188.00 -18.00
180 200.00 -20.00
190 212.00 -22.00
200 224.00 -24.00
210 236.00 -26.00
220 248.00 -28.00

Newton's law of cooling at work. Just tell yourself the NASCAR and F1 boys figured this out a long time ago,and they aren't listening to old wife's tales about pins. You will of course have to use both types on a semi-regular basis as the constant K will change as the tire wears since it is getting thinner, and K will be different for different tires. There are a few other factors involved, but try it and see. Equations rule!

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by descartesfool &raquo;</TD></TR><TR><TD CLASS="quote"> You will of course have to use both types on a semi-regular basis as the constant K will change as the tire wears since it is getting thinner, and K will be different for different tires. There are a few other factors involved, but try it and see. Equations rule!</TD></TR></TABLE>

I love a good equasion too, but you're saying that the only way to get a good reading with IR is to continually test with a probe AND do mind numbing calculations? I'm not getting it, if you still need a probe, why even use a IR?

 

I got readings from both at a local test-n-tune a couple of years ago. Guess what? After 5 laps around the skidpad, I measured the same three spots on the tire within 2 degrees of the probe that someone else took a reading with (at the same exact time, too).

Why are we taking temperatures? To see what the difference is between outside, center, and inside edges so we can figure out what our suspension setup is doing. For that, do we need absolutely accurate measurements anyway? No...we need consistency in measurements. MHO, anyway.

 

I just thought I would point out the cost,
about $150 each sensor x12 (3 per tire) = $1800 + data acquisition hardware and software.

I think I will stick with my $90 hand held probe.
However, it would be cool to see this data.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by L8ApexH22 &raquo;</TD></TR><TR><TD CLASS="quote">Why are we taking temperatures? To see what the difference is between outside, center, and inside edges so we can figure out what our suspension setup is doing. For that, do we need absolutely accurate measurements anyway? No...we need consistency in measurements. MHO, anyway.</TD></TR></TABLE>

But won't the hottest part of the tire cool at a more rapid rate than the cooler part of the tire? So, unless you are getting real-time data you can never be certain of the delta between them.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Flux &raquo;</TD></TR><TR><TD CLASS="quote">

But won't the hottest part of the tire cool at a more rapid rate than the cooler part of the tire? So, unless you are getting real-time data you can never be certain of the delta between them.</TD></TR></TABLE>

Yes, the hottest part will cool faster due to convective cooling (air flow) and conductive (tarmac). This is why real time measurement is so valuable. As usual, Descartesfool has discussed an excellent topic to consider.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by L8ApexH22 &raquo;</TD></TR><TR><TD CLASS="quote">I got readings from both at a local test-n-tune a couple of years ago. Guess what? After 5 laps around the skidpad, I measured the same three spots on the tire within 2 degrees of the probe that someone else took a reading with (at the same exact time, too).

Why are we taking temperatures? To see what the difference is between outside, center, and inside edges so we can figure out what our suspension setup is doing. For that, do we need absolutely accurate measurements anyway? No...we need consistency in measurements. MHO, anyway.</TD></TR></TABLE>

In addition, the actual temperature itself is so important. Some tires have a traction sweet spot temperature of around 220 to 260 degrees. Above or below this temperature, the available traction circle is reduced. So knowing the actual temperatures can help us decide, for example, whether to use a larger (or possible smaller) tire to keep the tire temps in the sweet spot.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by 577HondaPrelude &raquo;</TD></TR><TR><TD CLASS="quote">I just thought I would point out the cost,
about $150 each sensor x12 (3 per tire) = $1800 + data acquisition hardware and software.

I think I will stick with my $90 hand held probe.
However, it would be cool to see this data.</TD></TR></TABLE>
Ditto.
I am not at that level, and I don't think I'll ever be considering my age.
I'll stick to my innaccurate readings from my IR thermometer for now.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Johnny Mac &raquo;</TD></TR><TR><TD CLASS="quote">
As usual, Descartesfool has discussed an excellent topic to consider.</TD></TR></TABLE>

You are a gentleman and a scholar!

As stated in Newton's law of cooling, the hottest parts always cool first. That is true for both the surface and the core temperature. The next time you take tire temps, try an experiment. Time yourself to see how long it takes you to insert the probe thermometer into the tire, let is stabilize to a steady value and write it down. Then do this for the 12 temps required for the 4 tires. The go back and do it all again and compare the temps you get on the second go around. You will see that the high temps dropped more than the low temps, and this should prove to you that if they dropped in the time it took to go around the car, they likely dropped from the last hard left corner on the track. There is no kind of miracle with tire temps taken in the pits with either type of thermometer, probe or IR. They do not represent the on track temperature that is the only one important to the tire while it is cornering hard at the limit.

The surface and core temperatures are linearly related to each other (a straight line on a chart of Ts vs Tc), so once you have calibrated K using a probe type, you can just use the IR type to get from a table the core temperature. If K is very small, then the two temperatures will be very similar. Otherwise use a table to get core temps from surface temps. It is not like you are going to always whip out a camber gage while in the pits, jack the car up and change the camber right after you measure tire temps, so you do have some time in most cases to just re-enter the temps on another sheet. All you have to do is add the Diff values given above for a given Tambient and K. IR temps are much faster to measure than probe temps, and a whole lot easier. Many people I know do not take tire temps regularly because it is a pain even if they have a probe type, and they have all been told that you should never use an IR type thermometer. If the Ferrari Formula 1 team uses IR sensors for tire temps, and NASCAR and other very high end pro teams, why can not amateurs do it as well. Live tire temps on track are where it's at, and you could always do one tire at a time with just 3 sensors and a simple data acquisition system. We do what we can. I can understand that there would be many people who do not believe an IR measurement is useful because so many sources have said you must use a pin type. I say try both and see for yourself as you have to try new things sometimes. Feel the speed of the IR way,and just remember to add a correction based on K.

In the example I presented, you could just add 20 degrees to your IR readings of Ts for a Tcore of 176 to 224 and you would be off by a max of 4 degrees if you don't want to add the exact correction. If your K is less, then the error would be less. i.e. IR says it's 180, you add 20 and write down 200. Easy once you calibrate your setup. Plus the IR method is going to be very fast to show you if the temperature profile is incorrect, and if you have to add or remove air from the tire. Usually you want Tc to be slighlty higher on the inside (due to straight line driving on cambered tires) than the center and outside temps, say by 10-20 degrees. So target inside might be 210, middle 205 and outside 200 as an example. Your IR reading will very quickly show you if you are in the right ballpark, even if it reads 20 degrees less due to K, at 190, 185 and 180. If it said 190 200 180, you might want to think about lowering the pressure.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Johnny Mac &raquo;</TD></TR><TR><TD CLASS="quote">Yes, the hottest part will cool faster due to convective cooling (air flow) and conductive (tarmac). This is why real time measurement is so valuable. As usual, Descartesfool has discussed an excellent topic to consider.</TD></TR></TABLE>

A reasonably short period between driving hard and taking the temps should eliminate much of the error with a surface measurement cooling too much. Rubber isn't all that hot of a conductor of heat, and tends to have a fairly high thermal inertia. So if you sit there for a few mins after a run, yes the data is meaningless. If you get pretty quick readings I'd be shocked if they were off by more than about 5 degrees(but should still maintain about the same distribution across the face of the tire given the small delta T between the portions).

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Johnny Mac &raquo;</TD></TR><TR><TD CLASS="quote">In addition, the actual temperature itself is so important. Some tires have a traction sweet spot temperature of around 220 to 260 degrees. Above or below this temperature, the available traction circle is reduced. So knowing the actual temperatures can help us decide, for example, whether to use a larger (or possible smaller) tire to keep the tire temps in the sweet spot.</TD></TR></TABLE>

Yep, and you can't get anything remotely close to the actual temp without real-time IR sensors. It was described to me (by the Data Aquisition engineer for the World Challenge CTS) was that you will see huge temp spikes mid-corner compared to corner exit. During the straight, the temp is significantly lower than during cornering. You are seriously mis-leading yourself if you're trying to correlate the temp you record in the pit with data from a manufacturer (ie. "Hoosier's see max grip at 300F" etc). Now, I think most people use pyros for tuning purposes, where you can change alignment based on relative tire temps you record. That's still a fine way to tune, but you can't back out the max temps that the tire sees without data aquisition. IR tire temps sensors are very difficult to deal with too.

I don't see the reason to try to get core temperatures over surface temps. The maximum temperature is probably at the surface since that's where the friction interface is. The surface does cool faster so eventually the core gets relatively hotter, but the max temp must be at the surface. I'm not going to do the heat transfer calculations though.

I've been trying to get these on a FSAE car with less than great success.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by GSpeedR &raquo;</TD></TR><TR><TD CLASS="quote">
I've been trying to get these on a FSAE car with less than great success.</TD></TR></TABLE>

What specific sensors are you using, with what kind of output? Themocouple, 0-5V, etc? What are the problems you are encountering, mechanical, electrical or both?

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Def &raquo;</TD></TR><TR><TD CLASS="quote">

A reasonably short period between driving hard and taking the temps should eliminate much of the error with a surface measurement cooling too much. Rubber isn't all that hot of a conductor of heat, and tends to have a fairly high thermal inertia. So if you sit there for a few mins after a run, yes the data is meaningless. If you get pretty quick readings I'd be shocked if they were off by more than about 5 degrees(but should still maintain about the same distribution across the face of the tire given the small delta T between the portions).</TD></TR></TABLE>

Yeah, rubber is a pretty good insulator and usually possesses a relatively high coefficient of heat capacity (Cp). The heat capacity is analogous to inertia (usually referring to mass) such that a material with a high Cp changes temperature slower for a given cooling rate. Sometimes, this high Cp causes problems on courses with a disproportionate amount of time spent in cornering. The tire becomes excessively hot throughout the corner - with a resultant loss of grip. And, with not enough time to cool before entering the next corner, the tire stays too hot and will not be able to generate the ideal amount of grip. As a result, the car is slower - usually much slower - than a car with tires operating in the ideal temperature range.

The BTCC EP3 Honda's from a couple of years ago had this problem and the fix was to expose the forward tread surface of the front tires to the airflow. The tires cooled faster and by adjusting the front aerodynamics of the car, the engineers could control, to some degree, the tire temperatures throughout the race. Sure, the drag of the car was negatively affected, but the improvement in corner exit speeds more than made up for the small reduction in acceleration due to this drag.

For the club racer, it is usually too expensive to justify the cost of all this analysis. On the other hand, the professional teams can sometimes budget the cost of this important data gathering.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by descartesfool &raquo;</TD></TR><TR><TD CLASS="quote">
What specific sensors are you using, with what kind of output? Themocouple, 0-5V, etc? What are the problems you are encountering, mechanical, electrical or both?</TD></TR></TABLE>

0-5V Motec IR tire sensors, which are quite $$$.

The problem was more due to mounting. The closer the sensor to the tire, the less noise and better resolution. However, on an open-wheel FSAE with no wings or bodywork near the tires, it is very difficult to get a mount across the tire. It had to be mounted off the upright (come out from the inside the tire and bend around it), and it also had to be on top of the tire or you will ruin the sensors as soon as you turn the wheel. So we had to have a humungous piece of steel to be rigid enough (still a cantilevered beam) and it was exposed to any cone that wished to ruin $600 worth of sensors. We decided it wasn't worth the hit the budget would take and sent them back unused. We lost 3 engines last year so that ended up being a good choice.

From what I've seen the open wheel cars mount off the front wings and rear sidepods while sports cars use the wheel fenders. I'm sure that if I did get them working, I would've run into both electrical and mechanical issues since that seems to be what sensros are all about.

 

http://www.smartire.com/company/index.html

I know it not some 20k f1 tool but it can be useful.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by MidShipCivic &raquo;</TD></TR><TR><TD CLASS="quote">http://www.smartire.com/company/index.html

I know it not some 20k f1 tool but it can be useful.</TD></TR></TABLE>

I think they actually ping the sensor every 1-5 min or something. It's not live, real-time data.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by nonsense &raquo;</TD></TR><TR><TD CLASS="quote">

I think they actually ping the sensor every 1-5 min or something. It's not live, real-time data. </TD></TR></TABLE>

Never said it was.

It's better than nothing, and its much more afforable.

When the vehicle is in motion, data is transmitted every 4-6 minutes unless pressure changes, then the data is transmitted within 5-7 seconds.

Thats good enough.

I'd somehow attach the transmitters to the tire instead of the rim.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by MidShipCivic &raquo;</TD></TR><TR><TD CLASS="quote">I'd somehow attach the transmitters to the tire instead of the rim.</TD></TR></TABLE>

JB Weld?

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by GSpeedR &raquo;</TD></TR><TR><TD CLASS="quote">

JB Weld?</TD></TR></TABLE>

Krazy glue, if thats not enough I'd use an aerospace glue.