Mercedes GP W02

[beelsebob]

simulator technology is far away from predicting reality.The Mercedes simulator is outdated as well and new tools will come into service only later this year..
You cannot simulate sustained high g-forces in the simulator -how would you do a seven second duration sweeper with susttained 3+gs in a simulator ? the sled would have to accelerate at 3 gs for 7seconds to do that ..what a challenge and what a building to house that sled in....

I'm quite sure that ferrari's robot arm simulator could do exactly this. As an aside though, whether the sim can produce 3G turns for 7 seconds for the driver to feel what it's like is quite apart from wether the simulator can figure out how the car will behave. That's what matters here.

[Hangaku]

Sustained GForce simulation can be achieved already, relatively cheaply. How it could be adapted to an F1 simulator is a different matter. Sega made similar hardware to do this many years ago:



http://mysite.verizon.net/res8aiig/R360 ... r_sale.htm

[marcush.]

simulator technology is far away from predicting reality.The Mercedes simulator is outdated as well and new tools will come into service only later this year..
You cannot simulate sustained high g-forces in the simulator -how would you do a seven second duration sweeper with susttained 3+gs in a simulator ? the sled would have to accelerate at 3 gs for 7seconds to do that ..what a challenge and what a building to house that sled in....

I'm quite sure that ferrari's robot arm simulator could do exactly this. As an aside though, whether the sim can produce 3G turns for 7 seconds for the driver to feel what it's like is quite apart from wether the simulator can figure out how the car will behave. That's what matters here.

please let me know how you would produce a sustained 3g acceleration for 7s with an arm which does not rotate more than 360°....
Mercedes new simulator has the robot arm installed on a sled to simulate sustained cornering force .But still it is unable to produce more than 1.something of gs.
The driver feels the gforce .and he surely needs the force feedback to feel the car at the limit.if you tilt the thing to 90^he will feel one g .full stop...any decent road car can go there.It´s just not the same. reed what the guys say ..Glock for example did so much simulator work for Toyota and says it is a nice toy...It works to an extend but it will not give you the feeling at speed so why have the driver in the loop if the feedback of the car is not the real thing.

I think a driver in the loop simulator is just too elaborate and expensive for what you might gain from it.Drivers CAN drive and the best ones don´t need that much time to be able to extract what is up for grabs.

[beelsebob]

simulator technology is far away from predicting reality.The Mercedes simulator is outdated as well and new tools will come into service only later this year..
You cannot simulate sustained high g-forces in the simulator -how would you do a seven second duration sweeper with susttained 3+gs in a simulator ? the sled would have to accelerate at 3 gs for 7seconds to do that ..what a challenge and what a building to house that sled in....

I'm quite sure that ferrari's robot arm simulator could do exactly this. As an aside though, whether the sim can produce 3G turns for 7 seconds for the driver to feel what it's like is quite apart from wether the simulator can figure out how the car will behave. That's what matters here.

please let me know how you would produce a sustained 3g acceleration for 7s with an arm which does not rotate more than 360°....

Who said anything about it not rotating more than 360°? To sustain high G, all you need do is whirl round and round lots. Yes, keeping hydraulic lines untangled is tricky, but it is workable over more than a 360° turn

The driver feels the gforce .and he surely needs the force feedback to feel the car at the limit.

Right, but that doesn't mean that the racing simulation can't go "oh, the tyres have lost adhesion, and fling you off the road... The driver can know he's capable of doing the quick times because the car isn't in a virtual gravel trap. The question is then, not "can the simulator produce enough G force for the driver to feel like he's quick", but "can the simulator simulate the car accurately enough that he will end up in a virtual gravel trap if and only if he would end up in a real gravel trap driving the real thing. This question is entirely unrelated to whether the sim can produce lots of G force.

[jav]

Got to side with Marcush here.. spinning can produce high G forces for sustained time but that alone doesn't address realistic simulation. Accelerating into the spin to generate the side loads required would be hard to correlate with the rapid dynamic transitions from braking to acceleration to direction changes. These types of centrifugal G force generators may be fine for NASA to simulate constant high G lift off forces but i don't believe even Mercs new simulator (or anyone elses) can replicate and syncronize the rapid transitions in F1.

As such- I don't believe Ferrari's simulator would fool any experienced pilot.

[bill shoe]

I'm sure some F1T people understand simulators better than I do, but until they post I can give some general info.

For low frequency motion (~1-6 Hz), the translational motion that people subjectively feel most clearly is jerk (rate of change of accel). Yes, there is no way a relatively static simulator can create the sustained lateral accel of a long 4G turn. However, if the simulator can produce some cleverly timed jerk motion in conjunction with good graphics then the important parts of the motion will come through in a realistic way.

A purely centrifugal device might be useful for driver fitness and endurance. I'm skeptical it would be useful for driving simulation.

[marcush.]

i have checked this a few months ago ..there is a thread somewhere her on f1t about simulator tech .
the robot arm is a kukko ,if I remember correctly..

the question is why would you want to fool the driver into feeling what is not happening?
I fully understand why pilots get nausea in simultors.

Again what´s the benefit of having the driver in the loop when the feedback of the simulator is not realistic?

You may be able to teach a new driver some of what he might experience..but a well seasoned pro will not learn anything from it...
And if this approach is going to help you with setup..maybe..you could run your dynamic sims first and fudge them into the driving simulator and look if the driver is comfortable with the rate of change in carnering etc..but as the feel is never going to be the same the question still is :how does the driver feel the limit in high g cornering or high g braking.Is the rise or change of gforce a factor in his approach or does he rely on slip,force(steering feedback) yaw and pitch and heave movements to feel the limit.Maybe drivers are different in that respect as well.Some need a lot of time to get up to speed others can bang out the ultimate lap without even driving the thing at the limit ever before..Some drivers excelled in the high qualy boost times and could somehow live with the surprises offered to them on these laps others just could not cope.

[Byronrhys]

[Byronrhys]

Our TD in Merc Garms

[gridwalker]

Sustained GForce simulation can be achieved already, relatively cheaply. How it could be adapted to an F1 simulator is a different matter. Sega made similar hardware to do this many years ago

I remember those ... they wouldn't simulate multiple G though. There is no way they could maintain an arc with the constant acceleration necessary to replicate (say) Turkey Turn 8.

[HampusA]

I think it will be possible to create machines that can do more then 360 degrees and produce more then 5 g´s.

I think the real problem occurs is when they want to simulate Maggotts with 5g´s then in a split second 4-5 G´s in the other direction.
That would take some serious power and would probably just snap the arm off in half.

[marcush.]

I think it will be possible to create machines that can do more then 360 degrees and produce more then 5 g´s.

I think the real problem occurs is when they want to simulate Maggotts with 5g´s then in a split second 4-5 G´s in the other direction.
That would take some serious power and would probably just snap the arm off in half.

you can do almost anything .but the pricetag you have to stick on it will bring you quickly back to earth...You need very good reasons for doing this before someone will fork out that kind of money.
If you are moving the arm on a sled you can surely turn the driver 180° to simulate the chnaging side force direction ....The hell will be to build a sled which can accelerate a t 5 gs peak and sustained 3gs for seconds...8seconds at 30m/s² just think about the speed the damn thing has after 8 seconds..I´d say anything else than a centrifuge will just kick you in outer space for the gigantic dimensions you have to allow for.

[ForMuLaOne]

You simply CANNOT build a machine that simulates G forces like they occur in Formula one cars. You cant change from 5Gs to the left to 5Gs to the right within 0.5 seconds. There is neither a technology nor a material that would last this kind of stress.

[beelsebob]

You simply CANNOT build a machine that simulates G forces like they occur in Formula one cars. You cant change from 5Gs to the left to 5Gs to the right within 0.5 seconds. There is neither a technology nor a material that would last this kind of stress.

You're right, no one could ever make a machine that produces the same forces an F1 car produces, after all, it's not like an F1 car is a machine.

[Tim.Wright]

The simulator is always going to be compromised in terms of the accuracy of the forces felt by the driver. Remember, the ONLY way to reproduce the same accelerations (and just as important yaw rate) is for the simulator to follow the same trajectory and speed profile as the actual car.
Tim