The general idea of testing.

[marcush.]

so your point is:better having constant and known errors but repeatability as you may (or may not )have the testing environment under control?

To me real world conditions include cross winds and inconsistencies.and eliminating these in your evaluations will yield unrealistic results only repeatable in a wind tunnel or in your cfd world....

True it´s ALOT of additional work to plan your car carrying all the mounting points for measuring equipment but in the end it is the only way to gain true real world data and open the door for deeper insight what those variables as cross winds changing track temps etc are doing to your aero...It´s a lot of data to sample and analyse but sure the team that does it first will carry a huge advantage very soon given they have the capacity to make use of it all..
I rather think all that 50/60%tunnel work is creating a lot of extra work you could more efficiently use creating a 100%windtunnel model -think about it -you´d have real aero data without having to adjust scaling factors reynolds and whatever and even moreso-you can test the real mckoy -halving your output in terms of molds -as every development step has to be realised in scale + in full scale ....yes you need even bigger working areas in the tunnel with all its implications but the reward is quite high methinks.Of course it´s not allowed these days...So it´s not feasible for F1..

[Lycoming]

I never said which is preferable; merely that both have their place.

Obviously, with crosswinds/turbulence is useful because this is the condition that your car will actually be running in, and it's important to understand how the car behaves in those circumstances. More importantly, real world testing is the only way you know your models (wind tunnel, CFD) are "correct" in the sense that they are accurate enough to be usable. With CFD, error of as much as 30% is not unheard of in something as complex as a racecar.

But, a "clean" environment, like a wind tunnel, produces accurate and repeatable values for things like Cl and Cd. These are values you can then feed into your simulator for a variety of purposes.

The real world is not a controlled environment. Its very chaotic. Gathering lots of data on the chaos does not make it any less chaotic. It's important to understand how your car behaves in the real world, for sure... but I don't think there is much to gain from such a detailed analysis.

And of course, a full scale wind tunnel is better... If you can afford a good one. Although you could argue that RRA is what really prevents it. Sauber's CEO raised a point about the budget cap, specifically that it enables teams to be more creative in their use of resources. For example, Sauber could do full scale wind tunnel tests.

[myurr]

To me real world conditions include cross winds and inconsistencies.and eliminating these in your evaluations will yield unrealistic results only repeatable in a wind tunnel or in your cfd world....

Surely that would depend on whether you could capture cross winds and inconsistencies accurately and incorporate them into your results, otherwise they'll just end up being noise that reduces the quality of all your measurements.

The best, scientific, approach if you want that data is to be able to add in 1 mph cross winds, then 2 mph, then 3, etc. or to hit the car with a gust of wind of certain intensity and direction and see how it affects the stable state of the car. You cannot do that out on the track.

[raymondu999]

Why live? What difference would it make? If they can model, say, the conditions of FP1, they can still model it a few hours later.

[Owen.C93]

Why live? What difference would it make? If they can model, say, the conditions of FP1, they can still model it a few hours later.

Why wait?

[Crucial_Xtreme]

Why wait until fp1? I didn't convey that well, but that was my point. What is so golden about the timing of fp1?

It's not just FP1, it's all live sessions from FP to Quali. There's nothing better than to have another car to check a certain setup and see if it works well. If it does, they can change one or both of the other cars. It's a pretty smart idea.

[raymondu999]

I reiterate my point though. Why wait? Why not do it before fp1?

[Owen.C93]

I reiterate my point though. Why wait? Why not do it before fp1?

Because you don't know the track temp and wind conditions and tyre condition. Once you've run you can make some more accurate assumptions. I'm sure they work on setup in the sim prior to FP1, but they're working on imperfect information.

[marcush.]

you can wade through all the variables beforehand -not knowing what will happen in the session still allows you to simulate every possibility and you just feed in what really happens and get the correct simulation setup to start into the session...,right?

Doing it live during a session hints at their predictions being rather err...crude and they first need real world input and base their simulation on it?

[tok-tokkie]

If they ran the simulator with the same settings as on the car in real time it would be a great verification of the simulator - or expose its failings.
I wonder if that is not where the idea originated.

[marcush.]

If they ran the simulator with the same settings as on the car in real time it would be a great verification of the simulator - or expose its failings.
I wonder if that is not where the idea originated.

sure you start with validation of the simulator ,no doubt about it.

[hardingfv32]

If they ran the simulator with the same settings as on the car in real time it would be a great verification of the simulator - or expose its failings.
I wonder if that is not where the idea originated.

But that could be done at any time. No need to be done immediately.

This has something to do with making better setup decisions/adjustments the next time the car comes to rest and is available for the mechanics to work on.

Brian

[n smikle]

A little nugget I picked up on an AutoSport article this morning: Red Bull runs its simulator live during sessions as a third car to test different configurations Seb & Mark aren't using to see if they're helpful. Pretty interesting.

I think Garry Paffet did this last year - Silverstone.. IIRC.

[Crucial_Xtreme]

A little nugget I picked up on an AutoSport article this morning: Red Bull runs its simulator live during sessions as a third car to test different configurations Seb & Mark aren't using to see if they're helpful. Pretty interesting.

I think Garry Paffet did this last year - Silverstone.. IIRC.

Indeed he did. In the AutoSport article, they said he pulled an all-nighter.

[skgoa]

you can wade through all the variables beforehand -not knowing what will happen in the session still allows you to simulate every possibility and you just feed in what really happens and get the correct simulation setup to start into the session...,right?

No. Simulating every possibility would take longer than the age of the universe and consume all energy in it.

Doing it live during a session hints at their predictions being rather err...crude and they first need real world input and base their simulation on it?

Well, yes. It's still just a simulation, it will never be perfect. And more importantly they simply need to know the conditions during the session to make a sufficiently accurate simulation.