Simulator technology

[Tim.Wright]

I must admit I can't see the benefit of adding the inconsistancy of a human test driver to a lap simulation...? Any ideas what the person adds? The team's obviously feel there is an advantage...

A good human driver is still one of the best and most efficient post-processors you can use.

Tim

[Jersey Tom]

That's certainly true. For me it comes down to practicality.

Front tire inflation
Rear tire inflation
Front toe
Rear toe
Front camber
Rear camber
Front ride height
Rear ride height
Overall ride rate
Overall roll rate
Roll stiffness split
Front wing setting
Rear wing setting
Differential preload
Diff on-throttle locking rate
Diff off-throttle locking rate

Just a few for starters. Even at an absurdly coarse range of 3 settings per parameter, you have 3^16 = 43 million combinations. Find me a test driver that's going to put down 43 million consistent laps over the course of a week or two. Especially given that on a 1'30" lap, 43 million goes around would take 122 years of non-stop driving. To make it practical and chug through in 3 days, you'd have to be doing 150-200 "laps" a second.

Clearly you're not going to be sweeping a full factorial DOE all the time... but it does illustrate the point. There are severe limitations in running simulations in real time and/or with a real driver, at least for setup work. For driver training there can certainly be a benefit... but even then you better have a damn good virtual representation of the track, chassis, and tires. Even then you're never going to be able to totally predict the track grip as it evolves, the ambient conditions and how they'll effect engine power and aero...

[747heavy]

sure JT,
on the other hand, the one eyed is still king in the land of the blind.
As with every other tool, software or hardware, it has it´s limitations,
but there are also advantages to be had.

So, I guess it comes down, to using it in a sensible way.
A flight simulator will never substitude real flight training/testing,
still it is widely used, and has it´s justification.

[autogyro]

[...] F1 drivers are essential to the sport and earn every penny they make.

[Jersey Tom]

Certainly makes a complete joke out of Americas autonomous drones.

How the hell do you figure? What does this have to do with anything?

[Giblet]

I really don't think the two are related, but whatever.

If a driver out of the loop sim can make a car quick, doesn't mean your drivers can make it go quick in that way.

You can only test so many iterations with no driver until he needs to be in the sim loop.

[Jersey Tom]

In any event, by the time it gets to the point you need the driver to do final setup checks, I'd rather just have him or her in the real car. When you need that level of final precision, I'd be hard pressed to believe any sim, even driver-in-the-loop, is really going to nail it.

[PNSD]

Although rF pro is shown and said to be used, I believe at least for RBR its just for show, and they have their own simulator, and I imagine much more.... simulating?

[Richard]

In any event, by the time it gets to the point you need the driver to do final setup checks, I'd rather just have him or her in the real car. When you need that level of final precision, I'd be hard pressed to believe any sim, even driver-in-the-loop, is really going to nail it.

Waiting to put the driver in the real car would mean you do nothing until Friday practice. Whereas a simulator helps them prepare for the circuit, get a feel for bumps and corners, see how differing set up will affect the handling.

I imagine they try out many different scenarios in the factory which helps them identify the few that will be tried out during the very limited Friday testing.

[Jersey Tom]

Waiting to put the driver in the real car would mean you do nothing until Friday practice. Whereas a simulator helps them prepare for the circuit, get a feel for bumps and corners, see how differing set up will affect the handling.

Assuming the simulator has a very accurate representation of the track geometry on a macro and micro level... dynamic tire properties... track grip level, even appropriate weather conditions. Maybe they have all those things. Maybe not. If the latter.. how much do you want to be practicing and drilling yourself in something that's not accurate?

I'm sure they have some use. It's just that my admittedly limited experience with these things is they're often more for show than anything, even if they're hyped up to be some godly thing.

[Richard]

Yes, I usually attribute the mystical powers claimed for simulators to PR hype.

However, I can see their use for familiarisation of the track. I imagine they also use it to replicate the performance from the previous year. The have enough telemetry data to coach the drivers on where they braked the previous year, the lines they took on the corners etc. You can imagine Massa trying to mimic Alonso's laps to see where differences are. Why look atlines on a graph when you can shadow the chap in virtual reality?

I am also open to believing that it is worthwhile to play with different settings such as turn in and steering rates.

[Steven]

The only thing driving simulators really lack is G-force in corners. Basically anything else can be simulated.

I know for instance that Lewis Hamilton is testing this week in McLaren's simulator (one which is said to have some level of 3D vision) to validate some of the components the team have been developing for their 2011 contender.

[747heavy]

some informations about simulators in F1 and Rfactor pro can be found here,here, and here

........
Opinion remains divided on the true value of simulators.
To get the best from them the driver must feel totally connected with the car, not just in the virtual world where there is no real penalty for errors, but also in real life, where he can relate what happens in the simulator to what happens on the track.
It’s a tool that is only as good as the user’s understanding of it.
“They will never replace actual track testing,” Damerum acknowledges, “but it’s our job to try and make the system as realistic as possible to help develop the driver and car package.

[raymondu999]

I wonder if they can correctly simulate things such as tyre wear and graining though... Those are, after all, quite crucial parts of the race weekend, if not in terms of outright pace, in terms of the pitstop strategy etc

[marcush.]

as i understand dufournier has a tyre wear model in his portfolio....
http://www.dufournier-technologies.com/

http://dufournier.technologies.perso.ne ... 010_GB.pdf