Simulator technology

[Richard]

You'd want intolerability, otherwise you end up with two models with duplicated efforts and their own peculiar gremlins.

I imagine teams would want their main design modelling environment (consisting of several sub models for aero, mech, elec, tyres etc). That would include computational simulation (driver out of loop) to see how the sub models interact to explore optimal configuration (probably using evolutionary methods). My firm does this for tall buildings all the time, so I'd be worried if F1 didn't!

Then when it comes to sticking a diver in a play station to verify the human part, you'd want something that picks up the characteristics straight from your main model, you don't want to crunch the numbers all over again every time the driver turns the steering wheel on the simulator.

[silente]

so your opinion is that this link between adams and a driver in the loop simulation is helpful above all beacuse this allows you to take your model from adams and to put it inside the simulator without too much work to do, correct?

[Richard]

I imagine so - that's the usual reason for promoting software interoperability.

About time someone published FIA-XML ?

[Sonic59]

How long does it take to simulate 1 lap on adams?

[silente]

Adams (and its plugs in) is not a real time software. It takes some minutes on my lap top to simulate a complete lap. Time obviously depends on the deatail with which you model your car, your track etc.

[Tim.Wright]

For a very high end simulator you could have the processing power to run adams in real time, but otherwise there is Adams Real-Time to do that for you as well as a multitude of other simulations which run in real-time.

Tim

[Richard]

I would have thought the simulators would use response curves or performance parameters developed by the full design model. They wouldn't be doing real time CFD or dynamic analysis would they???

[Tim.Wright]

Simulation of multibody dynamics of engine, suspension, gearbox etc is all possible in real-time. Ive even seen it being done on a desktop PC.

Any finite element analysis (structural or CFD) reaquires a massive computing effort so probably would not be possible in real time.

However it is possible to run simplifed structural FE models which only represent the lower few modes of a large FE mesh so you could model the suspension arms like that in real-time

Tim

[raymondu999]

I would have thought the simulators would use response curves or performance parameters developed by the full design model. They wouldn't be doing real time CFD or dynamic analysis would they???

That's what I thought too. Referring to my "feeding in numbers from CFD" comment from above. i.e. kind of "at this speed the car has X downforce at the front, Y at the rear" etc

[mx_tifoso]

Axisofoversteer- Ferrari F1 Simulator



[youtube]http://www.youtube.com/watch?v=5T_tXG-89IU[/youtube]

[samoth]

Says who? Best for what?

Information from rFactor Pro

I knew ISI had a very good physics engine for gaming, but I didn't realise they also had a usable product for this market. Interesting.

Yes, for a few years i sat in this simulator in Kerpen, karttrack from M.S. This was the first rfactor pro simulator. An English man produced this of basis rfactor. He said to me it called rfactor pro. It was so hot in the simulator and i was dizzy after few laps. Maybe the english man gave up the rights

[gold333]

Does anyone have knowledge on the Red Bull simulator?

I know it is just Rfactor driver-in-the-loop version, and I think it runs PTWinSim for import & export and probably something like Xsim to run the hydraulics, but I'd love more info on it.

Here's Webber running the sim:

[youtube]http://www.youtube.com/watch?v=0ChGUYXYz-M[/youtube]

The track is clearly Alexandre monaco 1.1, downloadable for Rfactor here:

http://www.rfactorcentral.com/detail.cf ... w%20Layout

[raymondu999]

Are you implying that it's basically a big video game... or?

[Jersey Tom]

That's one way of describing a DIL simulator in general. At some fundamental level there's no difference:

You have some flavor of core simulation code. Maybe it's MBS, maybe it's lumped parameter. Doesn't really matter for the sake of this discussion. Feeding this are a set of human inputs for throttle, steering, and brake. It then spits out some visual feedback and numeric results.

Within both "professional" vehicle simulation as well as the consumer video game market there's a huge spread in specific solutions for that core simulation code - varying degrees of fidelity, complication, and required computing power - depending on what's most important to the particular end user.

[raymondu999]

JT; I get what you're saying and I completely agree. However my point was asking gold333 if he was implying that the simulator in the video was basically "rFactor ++" with a few new snazzes rather than at the fundamental; conceptual level that you're talking about.