planetf1.com

With limited pre and in season testing, the Simulators are playing a major role in the development of the car. Ferrari say their simulator is outdated so they hired PDR who worked on the McLaren Simulator which is considered state of the art and updated. My question is how does one know its outdated. I mean, its obvious that they would have made it possible to upload new software and stuff when they built it. But what goes into upgrading?

_________________
[b LOADING KILL LIST.......[/b]
--------------------

Just Ferrari trying to find excuses in case they don't have a fast car.
They're already downplaying themselves to make Alonso look better.
It's getting to be disgusting

Wish I was as all knowing as you.....

_________________
Power tends to corrupt and absolute power corrupts absolutely. [Lord Acton]
My own Google Earth Motor Sport file. http://www.mediafire.com/?jzm1ieatytv
Follow me @asphalt_world

Possibly the hardware is out of date, like how P.C.s and Macs from 5 years ago are outdated compared to more recent models. Also, there is only so far you can update a piece of software, eventually it needs to be replaced.

_________________
IV

I wonder if they have cleared out their cookies and defraggled.

And failing that, turn it off and back on again. Works every time.

To be honest, if its out of date (and I presume they mean the hardware, not he software as I would have thought they would have the best software available for their current hardware) why not just replace it? It's not like Ferrari would go bust over buying new simulators.

_________________
Team Vettel, Ricciardo, Bottas and Button.

As Angus said, MOOOOOOOOAR POWER. Probably updates to what moves it and how.

Gee Thank you for your valuable insight.

_________________
[b LOADING KILL LIST.......[/b]
--------------------

LOL

_________________
[b LOADING KILL LIST.......[/b]
--------------------

I'd imagine the teams have to construct the whole simulator themselves - the software, the hardware; all of it. They need a chassis that responds perfectly to what the computer is stating and inputs to setup change that very accurately represent what happens on their race car. Ferrari getting a 'new' simulator isn't just about them going down to the local computer store and getting a new processor, they actually have to develop what they have and that takes time.

Admitting you don't have the best simulator in the world isn't making excuses for failing on the racetrack, it's identifying and acknowledging that you have a weakness in that area and voicing your intent to improve it. That's, apparently, why they hired PDLR. They didn't give him a contract to verify their technology as being out of date.

_________________

Yeah I would assume they mean the hardware is not as like to like with the real thing, hence giving PDR the drive last friday so he can build on this. It might also possibly be the way their software is configured as to how the simulator works.

Its basically GT5 on a giant screen.

The OP's statement is true, in such times using Simulators is critical as it is the source for most base setups as well as development route selection. Basically everything (including drivers) are tested in the simulator, so with the limited testing available simulators are the where to location and to be honest even with unlimited testing Simulators would play a key role instead of teams investing so much in testing only to get few results so it would be worthwhile testing parts in the simulator before even bothering with the expenses, effort and time of making such a part or testing a new setup.

As i have gathered the Redbull and Mclaren Sims are the best in the business. Redbull use a base software of rfactor pro (same company who makes rfactor "game" makes a pro version highly customizable for teams) and then custom make and program every single detail of their car into it using a variety of options compatible with rfactor pro, i don't know much about the hardware but i would guess they buy a multi DOF unit and build from there. Mclaren however custom make the whole thing (not entirely sure about the hardware as well) and i couldn't get what base program they use or did they make the base program but most imply they made it.

As for knowing if you are outdated or not it is quite simple:
1. If your correlation with the real car is about 5% and another team achieves 0.5% correlation it means you are outdated.
2. If your tracks are 98% accurate and other teams are 99.5% accurate then you are outdated
3. If things tested in the sim react marginally different in reality than the sim then you are outdated.

Outdated is not necessarily hardware, but could also mean bad software or bad programming all De laRosa said is that the Mclaren simulator is better, what causes that only the Ferrari engineers know after getting a more detailed feedback from Pedro.


I really do hope i try out the Mclaren sim one day, and if i am lucky enough hopefully i will be working on developing it.

The simulators are quite advanced....video game on steroids. Please look at these links below for more info. So very easy to be outdated, yes by hardware technologies, all to acquire as much real-world data hence lack of true testing.

http://www.f1fanatic.co.uk/2010/01/17/driving-the-f1-simulator-which-gives-teams-the-best-way-to-cheat/

http://www.jamesallenonf1.com/2010/10/video-exclusive-inside-an-f1-teams-driving-simulator/

Redbull Sim with DC:
http://www.youtube.com/watch?v=oWVogW9cnDc&feature=related

Although, I do believe a couple teams have a custom system, designed and upgraded in-house. I thought Ferrari were using the Toyota sim in Cologne as it is difficult, understandably, to upgrade a sim during a season and pre-season testing? I recall Ferrari having problems with their wind tunnel as well.

I have been hearing nice things coming from the Curden simulators

Cruden are a top flight outfit, and their rigs are awesome.

http://youtu.be/DrdrymIwscw
http://youtu.be/EsAzk1R9Fzo

But as you mentioned, correlation is critical. The teams are not there to play games, but to use them as development tools. Such companies as Cruden could be contracted to contribute their expertise, build some components, and integrate software.

Engineers and developers study concepts on workstation computers, then if they are deemed to have potential, the concept is refined and then tested on the simulator. If it is still considered valid, then parts may then be constructed, to be tested (wind tunnel) and then integrated in the car.

And then that's where correlation becomes important, the simulators have to match up with the real world results. If, for example a front wing component offers a 5% increase in downforce and a 2% reduction in drag, it would turn out to be a huge waste of time and resources if by the time it was on the car it did not perform as expected.

With the dire shortage in testing, simulation becomes more and more important.

http://www.autosport.com/news/report.php/id/111960

Thoughts?

Although it is my line of work, i do agree with a lot that he says

Ferrari are only saying this because their simulator isn't as good as McLaren's. Also, they own their own race track next to their factory, and build their own engines, so their testing costs are considerably less expensive than other teams. Not to mention, they employ three ex-F1 drivers (Fisichella, Kobayashi, and de la Rosa) that could serve as test drivers. Take what you read with a grain of salt. He (rightfully) doesn't care about what's best for F1 - he cares about what's best for Ferrari.

so a man that has been in F1 longer than anyone less Bernie only cares about Ferrari? Really?
Ferrari is saying this because the fans want a better show. he has made terrific points.

So because Mclaren has a better simulator that its why he is saying these things?? Ferrari is at the top if not the highest budget in F1. if it was just a matter of a simulator don't you think they can buy the best? Of course they can. If Mclaren has the best simulator they will fight against testing based upon your logic, which would make Mclaren the same position you are saying Ferrari is trying to do.

Simulators are way more expensive in the long run than testing. testing you have weather conditions, you have daylight considerations. When unlimited testing was around Ferrari didn't test 24/7. they tested as much as other teams if I remember correctly it was usually 2-3 days per week in between GP's as much as other top teams.

_________________
One.
The best song ever written....thanks Bono
I am the Number 1Tifosi

Inappropriate post removed.

Inappropriate post removed.

_________________
One.
The best song ever written....thanks Bono
I am the Number 1Tifosi

Inappropriate post removed.

Inappropriate post removed.

_________________
One.
The best song ever written....thanks Bono
I am the Number 1Tifosi

Inappropriate post removed.

viewtopic.php?f=6&t=3225&p=101108#p101108

As for the attacks, stop now. Sanctions will follow shortly: this is not the general chat section, we will not tolerate petty arguments in here.

_________________
Forum rules: viewtopic.php?f=3&t=14979

Please report forum problems to us, via PM/Feedback Thread. Screenshots will also help.

I think it is fair enough that Ferrari have that view, their testing facilities are very close and thus cheaper than what most other teams pay, their Simulator is not top notch and have some calibration issues.

The point i do agree with however is the rising cost of using Simulators, Teams seem to be investing heavily in them but still i don't think the costs are anywhere near that of travelling and testing (even inside the country moving things around and booking a track is VERY expensive).

But still on track testing is the real deal and will always show what's what as Simulators are made to model that and not to design it, and there always will be deficits. I remember reading somewhere from a team engineer that before they laser scanned tracks they would notice a loss of grip between Simulators and real data and they would just model it as that area of tarmac has less grip than other places, After laser scanning the track it was found that there was a minor bump their that upset the suspension and thus lowered grip! Simulators will always be lacking one place or the other compared to on track testing.

Thing is though, Simulators are highly efficient, teams design a part and test it before attempting to make it and thus save plenty of money and time which are very critical. Although worth noting that Simulators are for the long run while testing is only for a short period of time.

I am all for Limited testing so that the cars can be made better, but i would like F1 teams to work on Simulators as well. They are used in many engineering areas and pioneered in Aerospace usage for safety, money, and time constraints and i can say the same should be made for F1.

In Summary: Testing YES, Unlimited NO, Simulators YES YES

i am in the software business and we sold a huge system to a major airline manufacture. This company used our software in their simulators. These simulators are good enough for commercial airline pilots to get certified on. When I was talking to one of the engineers who helped develop the simulator we discussed F1 as both of us are serious F1 fans. his response was in the air it is easy because there is no chemical composition to touch the wind to the plane meaning the track. each track is different with different compounds and different grip levels this is almost impossible to factor in hence the reason why every race teams have a difference in testing live rather than simulations.

I firmly think testing is cheaper in the long run than simulators, its my opinion based upon knowledge of technology and costs over time. Testing again is easier as most teams are based near a F1 circuit less maybe 2. costs of an engineer to develop software and maintain these complex systems are more expensive than a race engineer. you need more software developers than race engineers. again this is my opinion based upon my knowledge.

technology today in F 1 is huge I don't see simulators going away however I would regulate it this way. you have 24 hours in a day, I don't think simulations are limited in todays agreements so perhaps a way to do it is that you can run 24 hrs in either simulator or track or a combination. this way investments that are made can continue and a team has an option to use what they think is best. This way you can use best of both worlds.

_________________
One.
The best song ever written....thanks Bono
I am the Number 1Tifosi

I really, really don't get into the "it is easier on planes" bit. It is very very very difficult to model aircrafts in a simulators, wind changes, areas of low pressure, there is a whole science in aviation which is just one part when modelling an aircraft. Modelling vehicles with tyre data is definitely (IMO) a lot easier. and when the pilot loses control it is nearly unpredictable what will happen, much more than Automotives.

As for your point on track changes, it is the same for simulators as in testing. Temp. drops, Wind changes, tyre wear,,,etc. all change from circuit to circuit and even if teams test 5 days prior to an event on a single qualifying session the wind may change direction and they will be clueless (as has happened plenty of times). and you can't have the teams do tests on EVERY track on the calendar anyway as this would be plenty of times more expensive that work in the simulator.

Also Aircraft sim is not just for commercial planes, any decent military will have a sim to train its fighter pilots on flying their jets.

I Firmly believe that Simulators are the way to go, with limited testing to its aid.

I always wondered why more teams didn't attempt to build their own test track or build their premises at or very near to an existing circuit.

_________________
Power tends to corrupt and absolute power corrupts absolutely. [Lord Acton]
My own Google Earth Motor Sport file. http://www.mediafire.com/?jzm1ieatytv
Follow me @asphalt_world

I agree building a test track is a strong way to go!
M.Nader- I can see your point but respectfully disagree with sims for F1 taking control. It really is about costs. I can tell you that the aero company I am working with spends about 80-90 million per year keeping just 1 sim running. that its an all in cost of the facility the software the mainetence etc. now they offset that cost by charging the airlines time in the sim which of course in F1 that isn't the case it is a full cost basis not a revenue generator which testing can be at least partially income generating with fees to public viewing however limited that money may be.

your point on wind changes for F1 sims is another nail inn the coffin for costs---what do you think the cost is to have these calibrations done daily during in season or off season testing?

_________________
One.
The best song ever written....thanks Bono
I am the Number 1Tifosi

90 million a year is peanuts compared to what it replaces. Many, many years ago I recall watching a DC-8 do touch-and-go's all day at a military airport. When you consider the fuel, insurance, and aircraft costs, it is always cheaper to use a sim than a real aircraft for training. And that's the point, that for specific applications simulation is cheaper than doing it in the real world.

But for other applications such as complete system checks and parts reliability, it has to be tested in the real world.

Testing and simulation should compliment real world testing. Unfortunately, since testing is so restricted, simulation is being used far too much. A balance needs to be struck.

Airplanes are easier to model because you don't have to deal with magic tyres and suspension. Less variables = less complex. Simulators do not model in real time - it is impossible to do CFD with current technology in real time. On a very basic level, for airplanes, lift values are pre-calculated and then accessed via a lookup table, using air speed and angle of attack as inputs. With cars, not only do you need everything that an airplane requires (replacing lift with downforce), but you also need to lookup tyre grip, as a function of slip angle, camber, and tyre loading. But tyre loading isn't just an input - it's an output as a function of dynamic weight changes on the suspension. Just keep in mind that these are the very basic requirements: a more realistic simulator would necessitate many more variables like temp, humidity, (air density), wind, etc. (F1 simulators still use lookup tables. The complexity of a simulator is in finer grain lookups and many more inputs to outputs. F1 teams will feed CFD or wind tunnel lift/drag values into their sims).

Oh yeah - the reason why tyres are called "magic" tyres is because nobody really understands grip and the models for tyre grip are contrived, not derived. See: http://en.wikipedia.org/wiki/Hans_B._Pa ... ire_models