2019 Mclaren F1 Team - Renault

[Emag]

The thing is, you never get 100% accuracy on testing and simulation models. Wind tunnels are good to calculate how much downforce the car can produce, but in reality that number is usually not very accurate on real situations.

That is because the air/wind is not always constant in its parameters and that changes a lot of variables.

Also there is another big thing to take into consideration. F1 cars need their downforce on curves. And when the car is turning, the front tires are angled and the weight of the car shifts to one side. This totally changes the aero balance of the car and wind tunnels and CFD simulations cannot account for it completely.

Aerodynamicists have to do a lot of calculations to make the bodywork generate the same, or close to the same amount of downforce that it is shown by the wind tunnel even when the aero balance changes during corners.

That's where bad correlation comes from. If people get those calculations wrong, then you have big problems because the car is not generating the downforce the wind tunnel says it should and it is not always easy to pinpoint the reason why.

[SmallSoldier]

When the car met expectations and testing showed that correlation was good

Correlation is really unsung hero of the development. Teams that have good correlation can develop their car without testing and be able to improve their car. If your correlation is bad, you are effectively developing blind and with so little testing time in the season, your lead times become huge.

I'd say first priority of any team is solving any kind of correlation issue. I feel that McLaren made the decision to build a new wind tunnel for exactly this reason. By having a new, state of the art wind tunnel, they can get to the bottom of any kind of correlation issue and improve development speed for the car.

You are right... The big advantage of having positive correlation is that you can keep developing... When the correlation isn’t there, you have to spend time and resources to understand what the problem is and where, this slows down your developments (which you don’t want to keep throwing at the car without knowing if they will actually work).

Mclaren’s improvement this year is thanks to it and the fact that they are bringing small upgrades to each track, allowing them to test them properly and avoiding witch hunts if the car doesn’t behave as expected (for example, what happened to Renault and their France update).


Sent from my iPad using Tapatalk

[SmallSoldier]

Genuine question: How could a team get bad correlation? If McLaren have been using the same wind tunnel for years how does the correlation go from good to bad?

Even the best aero tunnel/CFD package cannot simulate the full transients effects happening on the track.

It was not my understanding that their issue last year was a correlation issue.

They just didn't pickup in CDF and the wind tunnel the side pods were too close to the front tires.

But that’s exactly a correlation issue... The expected downforce from their aero package wasn’t providing the modeled results and therefore every part they were putting in the car wasn’t performing as expected... Last year’s car was a great example of correlation issues driving by problems on the actual car versus problems in the model.


Sent from my iPad using Tapatalk

[diffuser]

Even the best aero tunnel/CFD package cannot simulate the full transients effects happening on the track.

It was not my understanding that their issue last year was a correlation issue.

They just didn't pickup in CDF and the wind tunnel the side pods were too close to the front tires.

But that’s exactly a correlation issue... The expected downforce from their aero package wasn’t providing the modeled results and therefore every part they were putting in the car wasn’t performing as expected... Last year’s car was a great example of correlation issues driving by problems on the actual car versus problems in the model.


Sent from my iPad using Tapatalk

Well they said the problem was not on straights but in curves.

[M840TR]

Genuine question: How could a team get bad correlation? If McLaren have been using the same wind tunnel for years how does the correlation go from good to bad?

So if you Develop an update for your car you are going to design it, test it in a simulation. if these numbers are good you are going to build a model a test it in a windtunnel. Again you look at the predicted performance and if the expected gain from simulation and wind tunnel match then you produce the part. If now these expectations are not met on track, you have a correlation problem. so apparently something in your model does not map real life behaviour really well. The correlation between your simulation / model and real life is off, so you have to improve your simulation / model

maybe someone else can give a better explanation then me.... but this is how i only can explain it in normal language

Also I think there can be problem with scale and how it reflects the model being tested. I've heard that Mclaren moved away from their own WT to Toyota's at the moment when F1 cars got 1/3rd larger/longer due to the new regulations and then their wind tunnel was not the best place to test a scale mode in 65% of the real size car.

Yup. And it mainly manifests as the outwash front-wing demanding greater room on the sides.

[SmallSoldier]

It was not my understanding that their issue last year was a correlation issue.

They just didn't pickup in CDF and the wind tunnel the side pods were too close to the front tires.

But that’s exactly a correlation issue... The expected downforce from their aero package wasn’t providing the modeled results and therefore every part they were putting in the car wasn’t performing as expected... Last year’s car was a great example of correlation issues driving by problems on the actual car versus problems in the model.


Sent from my iPad using Tapatalk

Well they said the problem was not on straights but in curves.

Yes, the problem was in regards to how the wake of the front tires was managed by the bargeboards... This made them lose the expected downforce on the turns, in order to solve it, they were forced to run additional downforce on the car with the correspondent drag penalty... But it was still a correlation issue, one that they took too long to find and by the time it was figured out, it didn’t made sense to try and get a B-Spec chassis at that point and instead focus on the MCL34.


Sent from my iPad using Tapatalk

[mwillems]

But that’s exactly a correlation issue... The expected downforce from their aero package wasn’t providing the modeled results and therefore every part they were putting in the car wasn’t performing as expected... Last year’s car was a great example of correlation issues driving by problems on the actual car versus problems in the model.


Sent from my iPad using Tapatalk

Well they said the problem was not on straights but in curves.

Yes, the problem was in regards to how the wake of the front tires was managed by the bargeboards... This made them lose the expected downforce on the turns, in order to solve it, they were forced to run additional downforce on the car with the correspondent drag penalty... But it was still a correlation issue, one that they took too long to find and by the time it was figured out, it didn’t made sense to try and get a B-Spec chassis at that point and instead focus on the MCL34.


Sent from my iPad using Tapatalk

If I recall, it was specifically that the front wheels were too close to the bargeboards and needed to be moved forward, hence the need for a chassis fix. When turning in corners, the bargeboards were not able to manage the turbulent air, and would not have been able to manage the air until the wheels have moved a little further forward, I think.

Ivan Roldan stated it was "not quite a correlation issue" as it was more to do with a behaviour they could not replicate in the tunnel than the tunnel not giving good correlation. It might have been better to describe that issue as sitting somewhere between a chair and a keyboard than the wind tunnel.

[godlameroso]

Genuine question: How could a team get bad correlation? If McLaren have been using the same wind tunnel for years how does the correlation go from good to bad?

Even the best aero tunnel/CFD package cannot simulate the full transients effects happening on the track.

It was not my understanding that their issue last year was a correlation issue.

They just didn't pickup in CDF and the wind tunnel the side pods were too close to the front tires.

I often wonder if when teams say we have "correlation issues" if it's really a euphemism, soft language used to hide the ugly truth. I always keep things like this in the back of my mind when I hear teams talk. There really is only one way to develop the car, the scientific method is best. Use trial and error to find what works what doesn't, use that trial and error to understand the variables at play. This process depends on good quality data, the more the better, data is evolutionary voltage. To get good data you need good validation engineers to test your models(theories), to see if the experimental data matches the model, and if it doesn't to understand why.

The reason why something does or doesn't work is often obvious and takes honest introspection to see clearly. Often times ego gets in the way of seeing this reason, the inconvenience it causes blinds us from the obvious. McLaren thought they could, with a little planning make a good car, and so far it has been very decent. McLaren is now honest with itself, there are still egos there, but these egos are not clouding the truth from being obvious to the detriment of the team or its development.

McLaren has all the talent and capacity to make a front running F1 car, and they're slowly chipping away at the gap to the top as a customer team. This shows they can return to their ways if they keep course, that they didn't suddenly lose the capacity to build a top class F1 car.

[SmallSoldier]

Found on Reddit... Mclaren the most improved team versus previous year on average (the stat should look even better at the second half of the season since that’s when the team struggle the most in 2018).




Sent from my iPad using Tapatalk

[trinidefender]

Found on Reddit... Mclaren the most improved team versus previous year on average (the stat should look even better at the second half of the season since that’s when the team struggle the most in 2018).

https://uploads.tapatalk-cdn.com/201908 ... f0abee.jpg


Sent from my iPad using Tapatalk

Is this percentage or time improvements? Link to the original?

[SmallSoldier]

It’s average improvement time for the best positioned car of each for the season.


Sent from my iPad using Tapatalk

[PhillipM]

It's gonna get even better in the rest of the season given the dev stopped on the last car by now.

[Emag]

https://www.mclaren.com/racing/team/and ... view-2019/

A Q&A with Andreas where he talks about his experience at McLaren so far and the prospects for the future

[FittingMechanics]

https://www.mclaren.com/racing/team/and ... view-2019/

A Q&A with Andreas where he talks about his experience at McLaren so far and the prospects for the future

Interesting things about digital transformation. I'm guessing Seidl sees that McLaren can improve their processes by utilizing modern technologies. This is good, it would be bad if everything was great but the car slow. Hopefully he can improve the team from the inside.

[_cerber1]

Different tire choices for Belgium.

https://twitter.com/pirellisport/status ... 7032931328