Mclaren ATLAS Question

[atc4ksaw]

Hello everyone. My first post here.

I race on iRacing and have gotten extremely interested on the technical aspects of chassis tuning. iRacing has an agreement with Mclaren for support with their data anaylizing program, ATLAS. I've fooled around with it; but I would like to learn more about what that program is telling me vs what I want to be looking at that demonstrates a good race machine.

I'd be happy with a book or guide where I can learn for myself or any other help in determining what I'm looking at.

Thank you very much,

Scott B.

[Jersey Tom]

Basically, you want a primer on racecar engineering. That's an awfully big subject to cover. I'm not going to touch it just yet.

Here's something to consider though - as it seems you've kind of alluded to, ATLAS won't tell you what makes a good racecar (or driver) or how to find speed. You / the engineer / whoever has to decide or otherwise know what it is you're trying to achieve with the car. Often it is a choice of what you think is best, and often it can be based off past experience. What's good for an open wheel car may not be relevant for a stock car, and what's good on a real race car may not always be the case in a sim.

Also worth noting is that the driver is as much if not more of the total performance package than the car setup. That may even be the easier starting point - thinking about and coming up with a way of identifying where you (driver) can improve.

[Nando]

Now this is weird...

Earlier today i sat and watched the first page of F1T wondering if i should make a post about ATLAS since i want to start building my own setups in iRacing..
If you don´t mind maybe we can share the thread and step by step understand more about the program?

Right now i´m simply using ATLAS to get proper tire temps and pressures over a Qual and Race stint.

Maybe JT knows this,

How do i know if i run to much camber? should i compare the maximum temps the tire reaches at any point of the circuit or should i look at the average temp over a lap?

I assume it´s the average temp and that it should not be any more then 10 degrees across the surface?

(right click view in another window for full size)


The bar is simply showing the temps at any point of the track, real time.

But the box below shows,

real time temp
Celsius
Maximum temp
Average temp (i think)

[Jersey Tom]

How do i know if i run to much camber? should i compare the maximum temps the tire reaches at any point of the circuit or should i look at the average temp over a lap?

I assume it´s the average temp and that it should not be any more then 10 degrees across the surface?

Well, what's your reasoning to that conclusion? To my point earlier - these are simulated tires, not real tires. Who is to say they want no more than 10 degrees across the surface? Who is to say that that's truly the best for real tires anyway?

Comes down to what you call "too much" camber - and that's a choice you have to make. What will the car do (or not do) if there's too much? How might you measure that quantitatively?

[Nando]

Well, what's your reasoning to that conclusion?

run of the mill talk i guess,

To my point earlier - these are simulated tires, not real tires. Who is to say they want no more than 10 degrees across the surface?

Dave Kaemmer, the guy who created the tire model said you would want no more then 10-15 degrees across the surface,
So my question was more if it was by looking at peak temps on a particular tire of if it´s by looking at the average temps over a full lap.

Who is to say that that's truly the best for real tires anyway?

See first quote,

Comes down to what you call "too much" camber - and that's a choice you have to make. What will the car do (or not do) if there's too much? How might you measure that quantitatively?

Too much camber and you will shred the inside part of the tire i would presume.
And i believe you can have so much camber that it will start to become counter-productive?

Sure they might not be 100% like reality but i´m fairly sure it´s not that far of in terms of temps and pressures?
If you had a telemetry file from real life and overlaid it, it would not look that much different.
that´s what certain V8 supercar drivers have said when they overlaid their telemetry.

[Jersey Tom]

Too much camber and you will shred the inside part of the tire i would presume.
And i believe you can have so much camber that it will start to become counter-productive?

Tire temperature is ultimately not car performance. So tire temperatures aside... how could one measure the point where it's counter-productive? What are basic measures of car performance?

If you identify those, would it not make sense to sweep camber and air pressure and measure the car performance directly rather than try to do it with temperatures? Sim tires are free, so use as many sets as you like!

[Nando]

Tire temperature is ultimately not car performance.

That´s true, but so you are saying you can´t learn anything about your camber settings (or what path to take)by reading into your temps across the surface of the carcass?

So tire temperatures aside... how could one measure the point where it's counter-productive? What are basic measures of car performance?

Good question, i guess lap time is one if you are consistent and have fully learnt the inside and out of the track to the point where there´s no "aha" moments in terms of lines etc.

If you identify those, would it not make sense to sweep camber and air pressure and measure the car performance directly rather than try to do it with temperatures? Sim tires are free, so use as many sets as you like!

So basically go through most if not all settings and basically throw away ATLAS as it´s useless anyways?
There has to be shorter ways to figuring out what works and what doesn´t then going through that cycle, surely ATLAS can help with that?

[Jersey Tom]

That´s true, but so you are saying you can´t learn anything about your camber settings (or what path to take)by reading into your temps across the surface of the carcass?

Double edged sword.

So tire temperatures aside... how could one measure the point where it's counter-productive? What are basic measures of car performance?

Good question, i guess lap time is one if you are consistent and have fully learnt the inside and out of the track to the point where there´s no "aha" moments in terms of lines etc.

How about lateral acceleration? Doesn't require knowing the inside and out of a track, or even following a good line.

So basically go through most if not all settings and basically throw away ATLAS as it´s useless anyways?

Didn't say that. See above. ATLAS is just a tool - a data viewer. Need to know the right questions to ask when interrogating the data... and even if it's the right tool for the job.

Keep in mind also (and to the point of right tool for the job) that some of the better functionality of ATLAS is impractical or impossible for use in something like iRacing or rFactor or whatever. The 'TL' in ATLAS stands for 'Telemetry Linked'. It's pretty nice being able to see data in real time during an event - that's a key feature. Doing visualization of historical sessions can be done in any number of tools - ATLAS, MATLAB, Excel...

[Nando]

Please expand on the double edged sword, i think i know what you mean but a more expanded view on why it´s that is always nice.

So more camber until i start seeing the lateral acceleration decrease?
Do you think running on a centripetal circuit? (basically a round parking lot that is perfectly 100% smooth)
Edit: skid pad is the word i´m looking for.
Is that a good place to tweak setups on or is it better to tweak setups at the actual track?
Is there anything i can learn about the setup or generally make it better by running there?


What type of functionality would that be more specifically?

[Jersey Tom]

Please expand on the double edged sword, i think i know what you mean but a more expanded view on why it´s that is always nice.

So more camber until i start seeing the lateral acceleration decrease?
Do you think running on a centripetal circuit? (basically a round parking lot that is perfectly 100% smooth)
Edit: skid pad is the word i´m looking for.
Is that a good place to tweak setups on or is it better to tweak setups at the actual track?
Is there anything i can learn about the setup or generally make it better by running there?


What type of functionality would that be more specifically?

In no particular order...

ATLAS functionality - like I was getting to, you can use ATLAS to view live streaming telemetry (as opposed to previously acquired data). Go through data straight off the car during a session. Don't quite get that aspect from sim racing. Otherwise, viewing historical data.. you can do it in any number of different programs. Plotting Y vs X is plotting Y vs X is plotting Y vs X.

Double edged sword... to my point earlier, you can learn things in the real world which don't apply in the sim world. You can learn things in the sim world which don't apply in the real world. You can read other opinions of the real world and car performance which may or may not be applicable to what you're doing.

10 degrees across the surface of a tire. Why? Why is 10 the magic number? Is that the same for open wheel cars as opposed to sports cars? Stock cars? If you took the exact same car setup and ran it on a Figure 8 track (lot of cornering, little or no straights) and compared it to a run on a track like Indy (lot of straights with brief bits of cornering)... do you think you'd get the same average temps? Did the iRacing developers dive into the n-th level of physical tire modeling and somehow 10 deg spread is the best for their model? Or did they, knowing that your average racer is going to try and tune pressure and temp to hit some spread bogey intentionally tweak their tire model to have that behavior?

What to look for... is a skidpad better than a full circuit? Maybe. Is a circuit better than a skidpad? Maybe. Is lateral acceleration a good performance indicator? Maybe. I offered it as an idea. The point is that these are all choices that you have to make or things you have to explore. A good engineer keeps a notebook, certainly. A key performance indicator might be temperature spread across a carcass - but you have to figure out what that best number is. 10-15 deg might be the number circulated by amateur and semi-pro guys, but is that good enough? Is that even relevant to your specific use? Same could be said of symmetric damper histograms or any number of things. All items which warrant further exploration - and ATLAS may be a tool to meet that end. Going in with preconceived notions of what's best.. that's where you can get bit in the rear.

[Nando]

I´m getting answers i do not like but obviously understand


if we switch it around then.. assume you are going to a new track, how would you go about building a setup, generally speaking?
Skip gearing, i can´t change the gearing on the Mclaren since it´s locked by FIA 2012 regulations.

Would ride-height be the first thing to look at? Getting as low as possible at the front then figuring out the optimal rake?
Or downforce levels first?

[Jersey Tom]

Depends. Have to know what the car is sensitive to. Some cars are very sensitive to front ride height and not particularly sensitive to rear. Others may be extremely rake sensitive. Some tires are very temperature sensitive, and others it's almost inconsequential by comparison. Some cars may be more aero balance sensitive than mechanical, or more mechanical than aero, or a balance between the two.

So the first step is gathering some knowledge about the car itself - maybe by doing some broad testing up front. Then once you have your check list of big factors to small factors... work in that order any time you go to a new track.

May not even have to change setup much from track to track, really.

[Nando]

Will do, thanks.